非小细胞肺癌对吉非替尼治疗敏感的必要因素

摘自：Biomedicine & Pharmacotherapy 65 (2011) 151–156

作者：Yi-Fan Zhao a,1, Chong-Ren Wang a,1, Yan-Ming Wu a, Sheng-Lin Ma b, Yuan Ji c,**, Yan-Jun Lu a,*

肺癌是目前全世界癌症的首要死因。吉非替尼作为一种选择性表皮生长因子受体（EGFR）酪氨酸激酶抑制剂，被广泛用于非小细胞肺癌（NSCLC）的治疗。本研究表明吉非替尼可以降低磷酸化蛋白激酶B（p-Akt）水平，同时升高p21水平、抑制周期蛋白依赖性蛋白激酶2/4（cdk2/4）和细胞周期蛋白E/D1（cyclinE/D1）活性，造成G1期阻滞，从而阻断NSCLC细胞细胞周期进程，抑制细胞生长。吉非替尼通过诱导的p21蛋白稳定性，而不是增加RNA蓄积使p21水平升高。此外，本研究发现从姜黄科温郁金中提取的天然药物β-榄香烯能够通过调节细胞中p21的水平从而恢复耐药细胞对吉非替尼的敏感性。以上结果表明吉非替尼与β-榄香烯联用可能能够更好的治疗对吉非替尼耐药的NSCLC。p21对于NSCLC细胞对吉非替尼治疗敏感的必要性也在p21过度表达和敲除的研究中得到进一步的证实。基于以上结果，我们认为p21是NSCLC细胞对吉非替尼治疗敏感的必要因素。

Gefitinib is a small-molecule quinazoline derivative that wasdeveloped as a tyrosine kinase inhibitor (TKI) of the EGF receptor(EGFR). EGFR is known to promote growth of cells and functions asan oncogene and expressed in up to 80–90% of non-small cell lungcancer (NSCLC) [1]. However, it is now clear that there are limitedsubgroups of (12–18%) NSCLC patients who derive particularbenefit from this treatment [2]. Gefitinib has been recentlyregistered as first-line treatment of NSCLC patients with EGFRactivating mutations. This approval is based on the data of thePhase III IPASS study, which demonstrated superior progression-free survival, greater objective response rate, improved tolerabilityand significant quality of life benefits for gefitinib compared tocarboplatin/paclitaxel doublet chemotherapy in clinically selectedfirst-line patients in Asia [3–5]. 

Mutations and amplification of theEGFR gene, and other molecules such as phosphorylated Akt andErbB-3 expression have been well described as markers of a betteroutcome in patients treated with Gefitinib [6–8]. However,molecular approaches linked to Gefitinib sensitivity in NSCLCare still unknown. Furthermore, correlation between some of thepredictors and clinical benefit is urgently required.

In the other hand, even in cases sensitive to Gefitinib, resistanceis acquired through continuous drug administration. Additionaltreatments for cases of NSCLC relapsing with treatment Gefitinibare urgently needed. Beta-elemene, a natural plant drug extractedfrom Curcuma wenyujin, has been used as an antitumor drug fordifferent tumors, including NSCLC via mechanism that inhibitsRas/Mapk signaling and cell cycle progression [9,10]. Beta-elemeneacts synergistically with cytotoxic drugs against a variety of tumorcells, and the observation that beta-elemene is able to do so indrug-resistant patients and, thereby, overcome drug resistance isespecially provocative.

Alterations in cell cycle control are a universal feature of lungcancers. P21, the product of the WAF1/CIP1/SDI1 gene, is aninhibitor of cyclin-dependent kinases and is activated throughp53-dependent or p53-independent pathways [11]. Studies haveclearly indicated that p21 plays an important role in regulation ofthe cell cycle, especially in G1 arrest [12,13]. Recent studies haveshowed that p21 plays important role in antiproliferative effort ofGefitinib [14–17]. However, the clinical significance of p21expression in NSCLC response to Gefitinib remains unclear.

In this study, the correlations between p21 expression and theincidence of Gefitinib-induced cytostasis, proliferative activity,and p53 status in the cell lines were assessed. Our data implicatedthat p21 plays an important role in NSCLC response to Gefitinib.

Cell cycle analysis by FACS

Cells were washed twice with PBS, trypsinized and resuspendedin PBS containing 0.1% Triton X-100 and RNase (1 mg/ml) (Sigma).The cell suspension was incubated at 37 8C for 30 minutes.Propidium iodide (molecular probes) was added at a finalconcentration of 50 mg/ml and the cell suspension was kept at4 8C for 1 hour. The cells were filtered and the cell cycle was analysedby flow cytometry with the FACScan system (Becton Dickinson).

Cells and transfection

Human lung cancer cell lines pc-9 and H1299 cells and theGefitinib-resistant cell line pc-9-ZD cells were cultured in DMEM(Hyclone) supplemented with 10% fetal bovine serum (Hyclone).MYC tagged p21 expression vector was generated as previouslydescribed [18]. ShRNA sequence against p21 (#1, CTT CGA CTT TGTCAC CGA G; #2, GAC CAT GTG GAC CTG TCA C) were clone intopSUPER-EGFP1 constructs (OligoEngine). The pSUPER-Scrambleplasmid (gat ccc cTT CTC CGA ACG TGT CAC GTt tca aga gaA CGTGAC ACG TTC GGA GAA ttt ttg gaa a) was used as the nonsensecontrol [19], which were synthesized by Sangon Ltd. (Shanghai).Cells were transiently transfected with expression constructs usingGeneJammer transfection reagents (Stratagene). For half-lifeexperiments, cells were treated with cycloheximide (CHX, Sigma)at a final concentration of 10 mg/ml.

Measurement of cell death

Cells were seeded into 96-well microtiter plate and treated withGefitinib (10,100,1000 nM) or beta-elemene (40 mg/ml). A 10 ml ofthe CCK-8 solution was added and then incubated 2hoursaccording to the procedure of Cell Counting Kit-8 (DojindoLaboratories). The absorbance was measured at 450 nm using amicroplate reader (BioTeK).

Western Blotting

The cells were prepared in lysis buffer of MC-CelLytics Kit(Shenergy Biocolor). The protein content was determined using theBradford calorimetric assay method (Shenergy Biocolor). Thelysate was resolved by 10% polyacrylamide-sodium lauryl sulfategel electrophoresis and transferred to a Hybond-C Super mem-brane (Amersham). Antibodies used for detection as follows, p21(2946, Cell Signaling), p53 (2524, Cell Signaling), p-Akt (05-669,Upstate), CDK2 (PC44, CalBiochem), CDK4 (sc260, Santa Cruz),cyclinD1 (sc20044, Santa Cruz), cyclinE (sc25303, Santa Cruz) andb-actin (Beyotime). Then, the blot was incubated with a secondaryantibody, IRDye 800 conjugated affinity purified anti-mouse orantirabbit IgG (Rockland Immunochemicals) and detected withOdyssey Infrared Imaging System.

Real-time PCR

Total RNA was extracted by homogenization in 1 mL TRIzolreagent (Invitrogen), followed by chloroform reextraction andisopropanol precipitation. 1 mg of RNA were reverse transcribedusing RevertAidTM M-MuLV Reverse Transcriptase (Fermentas)and random hexamer primer (Fermentas). Real-time PCR was donein a final volume of 20 uL containing 1.6 mL of each cDNA template,1 mL of each primer(10 Mm), and 10 mL of a SYBR Green mastermix (Takara). Primers used were 50-GGCAGACCAGCATGACA GATT-30 (sense) and 50 -GCGGATTAGGGCTTCCTCTT-30 (antisense) forp21(waf/cip1); 50 -CACGATGGAGGGGCCGGACTCATC-30 (sense)and 50 -TAAAG ACCTCTATGCCAACACAGT-30 (antisense) for humanb-actin. The average of p21 gene was normalized to b-actin asendogenous housekeeping gene.

3.1. P21 was elevated by Gefitinib in non-small lung cancer cells thatare sensitive to Gefitinib

Recent studies have led to understand some mechanism ofGefitinib induced cytostasis. Akt activity is commonly reduced byGefitinib in these tumors that are sensitive to Gefitinib. To confirmthis point, we assessed Gefitinib induced cytostasis and Aktactivities in two NSCLC cell lines: pc-9 and H1299 cells. As shownin Fig. 1a, pc-9 cells have an IC50 < 1 mM, indicated as Gefitinib-sensitive cells, whereas H1299 cells have an IC50 > 1 mM,indicated as Gefitinib-resistant cells. Western blotting analysiswas performed using antiphosphor-Akt antibody. As demonstrat-ed in Fig. 1b above, Gefitinib led to a reduction in phosphor-Aktonly in the Gefitinib-sensitive pc-9 cells but not in Gefitinib-resistant H1299 cells. These data suggested that reducing of Aktactivation levels plays an important role in mediating of Gefitinibinduced cytostasis.Studies have shown that antitumor agent suppresses cellgrowth likely related to upregulation of cell cycle inhibitors such asp27 and p21 in NSCLC [20]. Thus, we asked if p21 plays a role inresponse to Gefitinib selectively in NSCLC cell lines whose growthis inhibited by Gefitinib. We examined p21 levels in pc-9 andH1299 cells by Western blotting analysis. As seen again in Fig. 1b,the p21 protein levels were high in pc-9 cells. More, exposure toGefitinib actually increased p21 expression, effects that were mostpronounced in cells which are sensitive to Gefitinib. In contrast, nodetectable p21 could be seen in H1299 cells with or withouttreatment of Gefitinib. We further sought to test some other keycell cycle regulators which showed in Fig. 1b. Thus, we observed asignificant decrease in the levels of cdk2, cdk4, cyclinE andcyclinD1 by Gefitinib in pc-9 cells but these did not occurred toH1299 cells. These data was consistent with results of cell cycleanalysis showed that Gefitinib induced retention of cells in the G1phase, a sharp decrease in the S phase population but no significantchange in the G2/M fraction in pc-9 cells. In contrast, there was noapparent change of cell cycle pattern by treatment of Gefitinib inGefitinib-resistant H1299 cells (Fig. 1c). These data here indicatedthat Gefitinib elevated p21 levels and suppressed cdk2/4 andcyclinE/D1 activities which resulted in impaired cell cycleprogression through G1 arrest. Taken together, Gefitinib treatmentinduced cytostasis through multiple mechanisms such as reducingphosphorylation of Akt activity and suppressing cell cycleprogression by induction of p21 protein in Gefitinib-sensitivepc-9 cells.

We suggested that Gefitinib promoted p21 protein elevationwas, at least partly contribute to its negative regulation of cellprogression by Gefitinib in pc-9 cells. As p21 acts primarily as atranscription target of p53, we next examined the possibility thatthe induction of p21 caused by p53 transcriptional activity in pc-9cells treated with or without Gefitinib, the amount of p53 proteinwas assessed. As shown in Fig. 2a, no apparent correlation betweenp53 levels and p21 levels was seen in, suggesting that the formermolecule is not under the latter’s direct control.

We then evaluated p21 mRNA by real-time RT-PCR analysis todetermine whether mRNA accumulation contributed to elevatedp21 levels by Gefitinib in pc-9 cells. In contrast to the largedifferences in protein levels, p21 mRNA levels were no apparentdifference between with and without treatment of Gefitinib in pc-9cells (Fig. 2b). Therefore, p21 levels were elevated mainly due tothe protein stabilization other than mRNA accumulation.The elevation of p21 by Gefitinib raises a possibility thatGefitinib promotes p21 protein stability in post-transcriptional

pathway in pc-9 cells. Thus, we examined p21 protein in pc-9 cellsby inhibiting protein synthesis with cycloheximide (CHX).Gefitinib treated cells were treated with 10 mg/ml CHX and p21levels assessed by western blotting. Gefitinib treated cellsexpressed elevated p21 levels that did not decline significantlyover the period of CHX treatment; 78% of the p21 remained 2 hoursafter protein synthesis was blocked (Fig. 2c). Taken together, thesedata indicate that Gefitinib caused pronounced elevation of p21possibly through post-transcriptional pathway of inhibition ofprotein degradation in pc-9 cells.

3.2. Requirement of p21 for NSCLC sensitivity to Gefitinib

To determine the restoration of p21 influences Gefitinib-induced cytostasis, we used H1299 cells, which lacks p21expression due to homozygous deletion of the p53 gene. For theexpression of p21, we introduced p21 into the H1299 cells bytransient transfection. As seen in Fig. 3a above, cells transfectedwith vector expression of p21 showed a dramatic induction of p21expression as compared to cells transfected with empty vector.Then, we examined the direct effect of p21 over-expression on thetreatment of H1299 cells with Gefitinib. Upon over-expression ofp21, Gefitinib induced cytostasis was remarkably increased incomparison to that in cells transfected with a control vector(Fig. 3a, bottom).To further examine the role of p21 in the Gefitinib-inducedcytostasis in pc-9 cells, we used a shRNA targeting p21 gene tosilence p21 expression in the cells. As seen in Fig. 3b, cellstransfected with p21 shRNA1 and shRNA2 showed a dramaticreduction of endogenous p21 expression as compared to cellstransfected with nonsense control. Furthermore, silencing ofendogenous p21 expression in the cells increased levels ofCDK2, CDK4, cyclinD1 and CyclinE by Gefitinib in comparison tothe cells transfected with control vector (Fig. 3c). By silencing of

endogenous p21 expression, Gefitinib induced cytostasis wasremarkably decreased in comparison to that in cells transfectedwith a nonsense control (Fig. 3d). These data are in agreement withthe observation in the cells over-expression of p21 which showedabove implicated that p21 promoted Gefitinib-induced cytostasis.Thus, we propose that p21 is required for Gefitinib-sensitive NSCLCcells.

3.3. P21 reversed sensitivity to Gefitinib in NSCLC with acquiredresistance to Gefitinib

Gefitinib has been tested as monotherapy for patients withrelapsed NSCLC. In cases of NSCLC sensitive to Gefitinib, resistancemight be acquired through continuous drug administration.Additional treatment for cases of NSCLC relapsing duringtreatment with Gefitinib is urgently needed. Beta-elemene, anatural plant drug extracted from Curcuma wenyujin, has beenused as an antitumor drug for different tumors, including NSCLCvia mechanism that inhibits Ras/Mapk signaling and cell cycleprogression [8]. Our clinical trials showed that administration ofGefitinib in combination with beta-elemene for NSCLC which areacquired resistance to Gefitinib, remarkably enhanced the antitu-mor effect as compared to Gefitinib treatment alone (personalcommunication). Thus, we try to determine the molecular basis forthe combination treatment. First, we established acquiredGefitinib-resistant pc-9 cells through continuous exposure ofGefitinib. Resistance against Gefitinib developed about 6 monthsand the relative resistant values reached 4-fold after exposure toGefitinib. We picked the clones of Gefitinib resistant cell linesnamed pc-9-ZD. As it showed in Fig. 4a, pc-9-ZD cells was able tosurvival by > 50% at the concentration of > 1 mM Gefitinibwhereas 84.8% of its parental pc-9 cells was inhibited by that

concentration of Gefitinib. To test the effect of beta-elemene onGefitinib-induced cytostasis in the cells with acquired resistance toGefitinib, we measured the cell death rate in pc-9-ZD cells treatedwith or without beta-elemene in the presence of Gefitinib. As seenin Fig. 4b, beta-elemene reversed sensitivity to Gefitinib treatmentin pc-9-ZD cells. These data suggested that this innovative clinicalapplication of beta-elemene in combination with Gefitinib mayoffer great opportunities for NSCLC patients who are acquiredresistance to Gefitinib.

Studies have reported that antiproliferation induced by beta-elemene was likely dependent on preventing cell cycle progressionby arresting in G0/G1 phase [10]. Therefore, we speculated thatp21 might involve in a mechanism that beta-elemene inhibits cellprogression. To test this hypothesis, we assessed the p21 proteinsexpression in pc-9-ZD cells which incubated with or withoutoptimal concentration of beta-elemene. The results shown inFig. 4c demonstrated that acquired resistance to Gefitinib cells, pc-9-ZD significantly decreased the amount of p21 protein ascompared to its parental pc-9 cells. However, pc-9-ZD cellsexposed to beta-elemene, induced significantly expression of p21protein. These data indicated that beta-elemene induced p21protein in NSCLC cells.Our previously study suggested that pc-9-ZD cells withheightened levels of p-Akt and reduced levels of p21 resistedfurther Gefitinib-induced cytostasis [21]. To assess the directeffects of restoration of p21 influences, we transient transfectedthe pc-9-ZD cells with vector expression of p21. Upon over-expression of p21, Gefitinib-induced cytostasis was remarkablyincreased by Gefitinib in comparison to that in cells transfectedwith a control vector (Fig. 4d). Thus, restoration of p21 seems toreverse sensitivity to Gefitinib treatment in NSCLC cells withacquired resistance to Gefitinib. Taken together, we suggested that p21 promote NSCLC to Gefitinib via a mechanism that involvesp21-dependent inhibition of cdk2 activity and cell cycle arrest inG1 phase.

Gefitinib has been approved for the treatment of NSCLCpatients. Recent studies have led to understand some mechanismof Gefitinib induced cytostasis. Current report confirms that Aktpathway is down-regulated in response to Gefitinib only in celllines that are sensitive to Gefitinib (Fig. 1a and b). Thus, reducing ofAkt activation level plays an important role in mediating of cellgrowth-inhibition by Gefitinib.However, the mechanism by which Gefitinib inhibits cellproliferation has not been clearly defined. Cell cycle progression isdriven by CDKs in association with cyclins. P21 has beenconsidered as the most important cell cycle checkpoint proteinby inhibiting CDKs/cyclins activities [18,22]. Here, we showed thatGefitinib elevated p21 levels and suppressed cdk2/4 and cyclinE/D1 activities which resulted in impaired cell cycle progressionthrough G1 arrest in the cells that are sensitive to Gefitinib (Fig. 1band c). These results are consistent with our previous finding thatp21 suppressed cell growth by arresting cell cycle at G1 or G2phase through binding and inhibition of CKD2/4 activity[13,18,21,22], and suggest that p21 plays an important role inmediating of cell growth-inhibition by Gefitinib.P21 transcriptional activities have p53-dependent and -independent pathways. In this case, there was no apparentcorrelation between the level of p21 and that of p53 (Fig. 2a),suggesting that the former molecule is not under the latter’s directcontrol. More, p21 protein stability plays a critical role in its

functions. Studies have shown that p21 is degraded during the cellcycle [23,24]. We find here that the cells that are sensitive toGefitinib enhanced p21 protein stability, which may be theprincipal contributing factor to elevated p21 levels (Fig. 2b and c).The results led us to suggest that Gefitinib may increase p21accumulation in a post-transcriptional manner in pc-9 cells.Effects of CDKs/cyclins on p21 protein stability have beenreported [25]. Complex p21 with CDKs/cyclins is more stable thanthe free form. P21 is a proteasome target that bypasses the need forubiquitylation by binding directly to the C8a subunit of the 20Sproteasome catalytic complex. CDKs/cyclins complex may protectp21 from degradation via association with the C-terminal Cy2 site,which overlaps the C8a-binding site [23]. It is believed that therelative levels of CDKs/cyclins determine the rate of the degrada-tion of p21 protein [26]. Our study suggests that it is likely thatGefitinib decreases CDK2/4 and cyclins activities/levels in the cellsthat are sensitive to Gefitinib, which contributes to increased p21accumulation. Thus, by repressing CDK2/4 and cyclinD1/E activi-ties/levels, it is likely that Gefitinib increases p21 accumulation bymediating p21 binding to CDKs/cyclins, to prevent p21 degrada-tion in a proteasome-dependent manner. Taken together, wesuggest that elevation of p21 plays an important role in mediatingof cell growth-inhibition by Gefitinib.The p21 effect on the cells to response to Gefitinib was furtherconfirmed by p21 over-expression and knockdown studiespointing to a requirement of p21 for the cells sensitive to Gefitinib(Fig. 3a–d). Thus, we propose that p21 is required for Gefitinib-sensitive NSCLC cells.Gefitinib is worldwide used for antitumor in NSCLC patients.However, the efficiency of Gefitinib treatment is limited by thedevelopment of this drug resistance through continuous Gefitinib administration. 

Additional treatments have been needed to improve the acquired drug resistance. Our clinical trials showed thatadministration of Gefitinib in combination with beta-elemene, anatural plant drug extracted from Curcuma wenyujin, remarkablyenhanced the antitumor effect as compared to Gefitinib treatmentalone for NSCLC which are acquired resistance to Gefitinib (personalcommunication). Consistent with our findings in the cell model here,a combined treatment of beta-elemene and Gefitinib reversed thesensitivity to Gefitinib-induced cell growth inhibition in pc-9-ZDcells (Fig. 4a and b), suggesting that it might represent a new strategyto improve the efficacy of therapeutic response to the tumors knownto be resistant against Gefitinib.Moreover, a clear elevation of p21 levels mediating by beta-elemene or by over-expression of p21 reversed the pc-9-ZD cells’sensitivity to Gefitinib treatment (Fig. 4c and d). Together, thesefindings revealed that beta-elemene modulated the elevation ofp21 levels in pc-9-ZD cells, in augmenting an effective therapeuticresponse to Gefitinib treatment.In conclusion, we find that p21 mediate signals that are criticalgrowth inhibition by Gefitinib. Our observations have significantimplications for understanding how loss of p21 contribute to drug-resistance in NSCLC and for developing ways to target aberrantlyactive parts of this growth regulatory pathway. These datahighlight the importance of further studies into the role of p21in NSCLC response to Gefitinib.

The authors declare that they have no conflicts of interestconcerning this article.