1 23

Tumor BiologyTumor Markers, Tumor Targeting andTranslational Cancer Research ISSN 1010-4283 Tumor Biol.DOI 10.1007/s13277-014-2420-0

COX-2 overexpression and -8473 T/Cpolymorphism in 3′ UTR in non-small celllung cancer

Imtiyaz A. Bhat, Roohi Rasool, IqbalQasim, Khalid Z. Masoodi, ShabeerA. Paul, Bashir A. Bhat, FarooqA. Ganaie, Sheikh A. Aziz, et al.

1 23

Your article is protected by copyright and all

rights are held exclusively by International

Society of Oncology and BioMarkers (ISOBM).

This e-offprint is for personal use only

and shall not be self-archived in electronic

repositories. If you wish to self-archive your

article, please use the accepted manuscript

version for posting on your own website. You

may further deposit the accepted manuscript

version in any repository, provided it is only

made publicly available 12 months after

official publication or later and provided

acknowledgement is given to the original

source of publication and a link is inserted

to the published article on Springer's

website. The link must be accompanied by

the following text: "The final publication is

available at link.springer.com”.

RESEARCH ARTICLE

COX-2 overexpression and -8473 T/C polymorphism in 3′ UTRin non-small cell lung cancer

Imtiyaz A. Bhat & Roohi Rasool & Iqbal Qasim &

Khalid Z. Masoodi & Shabeer A. Paul & Bashir A. Bhat &Farooq A. Ganaie & Sheikh A. Aziz & Zafar A. Shah

Received: 15 April 2014 /Accepted: 29 July 2014# International Society of Oncology and BioMarkers (ISOBM) 2014

Abstract A new class of compounds targeting cyclooxygen-ase 2 (COX-2) together with other different clinically usedtherapeutic strategies has recently shown a promise for thechemoprevention of several solid tumors including lung can-cer. The aim was to study the possible role of COX-2 -8473T/C NP and its expression in the pathogenesis of non-smallcell lung cancer. One hundred ninety non-small cell lungcancer (NSCLC) patients and 200 healthy age-, sex-, andsmoking-matched controls were used for polymorphic analy-sis, and 48 histopathologically confirmed NSCLC patientswere analyzed for COX-2 messenger RNA (mRNA) andprotein expression. Our results showed that the frequencies

of variant genotypes 8473 CT/CC were significantly lesscommon in the cases (30.0 %) than in the controls (36 %),suggesting that the 8473C variant allele is related with lowersusceptibility in NSCLC (OR=0.79, 95 % CI 0.54–1.4).However, the frequency of COX-2 -8473 TC and CC geno-types were significantly associated with age in NSCLC (P=0.02). Quantitative real-time expression analysis showed asignificant increase in the COX-2 mRNA in tumor tissues ascompared to their adjacent normal tissues [delta cycle thresh-old (ΔCT) = 9.25 ± 4.67 vs 5.63 ± 3.85, P= 0.0001].Multivariate logistic regression analyses revealed that theCOX-2 expression was associated significantly with age(P=0.044). Also, an increasing trend was observed in stagesI and II and in female patients compared to stages III and IVand male patients, respectively, but no statistical significancewas observed. However, COX-2mRNA expression shown noassociation with the -8473C variant allele. Our findings indi-cate that the COX-2 T8473C polymorphismmay contribute toNSCLC cancer susceptibility in the Kashmiri population,while our expression analysis revealed a significant increaseof COX-2 in tumor tissues as compared to their adjacentnormal tissues, suggesting that it could become an importanttherapeutic marker in NSCLC in the future.

Keywords Polymorphism . Restriction fragment lengthpolymorphism (RFLP) . Non-small cell lung cancer .

RT-PCR .Western blotting

Introduction

Cyclooxygenase 2 (COX-2) is the key enzyme of prostaglan-din synthesis in inflamed tissues [1], and it also supports themediation of prostaglandin implication in inflammation andangiogenesis and supports the growth of several solid tumors[2–4]. COX-2 expression has been reported in many human

I. A. Bhat : R. Rasool : I. Qasim : Z. A. Shah (*)Department of Immunology and Molecular Medicine,Sher-I-Kashmir Institute of Medical Sciences, Soura, Srinagar,Kashmir 190011, Indiae-mail: zaffaramin@gmail.com

I. A. Bhate-mail: imty82@gmail.com

K. Z. MasoodiDepartment of Urology Research Laboratories, University ofPittsburgh Medical Center, Pittsburgh, PA 15232, USA

S. A. PaulDepartment of General Medicnine, Sheri Kashmir Institute ofMedical Scinces, Medical College, Bemina , Srinagar, India

B. A. BhatDepartment of Plastic Surgery, Sher I Kashmir Institute Of MedicalSciences, Srinagar 190011, India

F. A. GanaieDepartment of Cardiovascular and Thoracic Surgery, Sher-I-KashmirInstitute of Medical Sciences, Srinagar 190011, India

S. A. AzizDepartment of Medical Oncology, Sher-I-Kashmir Institute ofMedical Sciences, Srinagar 190011, India

Tumor Biol.DOI 10.1007/s13277-014-2420-0

Author's personal copy

cancers, including colorectal, breast, ovarian, uterine cervix,lung, head, and neck [3, 5–9] cancers. In most of the cancers,COX-2 actively supports cancer growth [3, 6, 7] and itsoverexpression acts as an independent prognostic factor inlung and oral cancer, osteosarcoma, multiple myeloma, andcolorectal and ovarian cancer [5, 8, 10–15].

The human COX-2 gene, mapped to chromosome 1q25.2–q25.3, is about 8.3 kb pairs in size and contains 10 exons [5].The COX-2 promoter region consists of several key cis-actingregulatory elements including stimulatory protein 1 (SP1),suggesting involvement of a complex array of factors in itsregulation [6]. COX-2 overexpression promotes and main-tains tumor growth by increasing resistance to the apoptoticsignal [16], inducing angiogenesis [17], and reducing theimmune body response against the tumor [18]. COX-2 is alsoa useful target for chemopreventive and therapeutic interven-tion [19–21]. COX-2 inhibitors have shown their antitumoreffects by increasing apoptosis [16] and antiangiogenic activ-ities [17] and restoring the host immune system [18].

Transcriptional regulation has been shown to be a majormechanism in regulating gene expression. In addition, post-transcriptional mechanisms influencing stability of COX-2messenger RNA (mRNA) are also important. Thus, it ispossible that naturally occurring sequence variations in thepromoter and 3′ untranslated regions of COX-2 gene mightcontribute toward differential COX-2 expression and a sub-stantial degree of interindividual variability in susceptibility tocancer as well as response to the treatment.

Genetic polymorphisms in the COX-2 gene have beendescribed to influence its expression through transcriptionaland posttranscriptional mechanisms. A common SNP(T8473C) in the 3′ untranslated region (UTR) of the COX-2gene has been shown to be associated with the alteration ofmRNA level of the gene, because sequences within the 3′UTR of the COX-2 gene are important for enhancing messen-ger translation as well as for translational silencing [22].

In view of the abovementioned findings, we conceived acase–control study to investigate the polymorphic associationof COX-2 -8473 with non-small cell lung cancer (NSCLC) inour population. Also, mRNA and protein expression of COX-2 was analyzed in lung tumor tissues and their adjacent normaltissues of lung cancer patients by quantitative real-time poly-merase chain reaction (qRT-PCR) and Western blotting,respectively.

Materials and methods

Materials

This study included a cohort of 190 lung cancer patients and200 healthy controls that were age, sex, dwelling, andsmoking matched. All the lung cancer cases selected for the

present investigation were histologically confirmed NSCLCcases. Patients were recruited from Outpatient Department,Department of Medical Oncology, and Department ofCardiovascular and Thoracic Surgery, Sher-I-KashmirInstitute of Medical Sciences (SKIMS), Srinagar betweenApril 2010 and March 2012. The patients having a priorhistory of cancer other than lung cancer and patients whohad received any chemotherapy and/or radiotherapy wereexcluded from the study. All participants of the control groupwere selected from individuals receiving routine medical ex-aminations at SKIMS, with no history of any malignancy orany symptoms of other acute or chronic inflammatory lungdiseases (COPD, asthma), peritonitis, rheumatoid arthritis, orlate stage kidney diseases. Ethical approval was taken fromthe ethical committee of SKIMS, and World MedicalAssociation (Declaration of Helsinki) protocol was followed.All the participants were informed, and written consent wastaken for participation in the present study. Five-milliliterperipheral blood samples were collected in EDTA vials fromlung cancer patients and healthy controls and stored at −80 °Cuntil further use. Tissue samples were stored in RNA later(Sigma-Aldrich) and kept at −80 °C until further use. Adetailed questionnaire was completed for each patient andcontrol. The questionnaire included information on the num-ber of cigarettes smoked daily/quantity of tobacco smokedevery day and the number of years the subject had beensmoking. For smoking status, a person who had smoked

Table 1 Clinicopathologic characteristics of NSCLC patients andhealthy controls used for polymorphic analysis of COX-2

Characteristic Cases (n=190) Controls (n=200) P value

Age (years)

≤50 59 (31.05) 65 (32.5) 0.8>50 131 (68.95) 135 (67.5)

Sex

Male 163 (85.79) 170 (85) 0.8Female 27 (14.21 30 (15)

Smoking status

Nonsmoker 76 (40.0) 84 (42) 0.76Smoke 114 (60.0) 116 (58)

Dweller

Rural 127 (66.80) 129 (64.5) 0.67Urban 63 (33.15) 71 (35.5)

Histology

SCC 140 (73.69) –Othersa 50 (26.32)

Stage

I and II 130 (68.42) –III and IV 60 (31.58)

SCC squamous cell carcinomaaOthers include adenocarcinoma, large cell carcinoma, and broncogeniccarcinoma

Tumor Biol.

Author's personal copy

(cigarette/Hokka) at least once a day for >1 year was regardedas smoker. The cases included 163 (85.79 %) males and 27(14.71 %) females. Fifty-nine (31.05 %) patients were lessthan 50 years of age, and 131 (68.95 %) of the patients wereabove 50 years of age with a mean age of 57.80±10.97. Thecontrol group had 170 (85 %) males and 30 (15 %) females

with 65 (32.5 %) control participants who were above the ageof 50 years and 135 (67.5 %) participants who were less than50 years of age with a mean age of 56.69±12.22. Smokers incases and controls were 114 (60 %) and 116 (58 %), respec-tively, while nonsmokers were 76 (40 %) and 84 (42 %),respectively. NSCLC included 130 (68.42 %) cases who had

Table 2 Distribution of COX-2 -8473 T/C genotypes among non-small cell lung cancer cases and healthy controls

Genotype Cases, N=190 (%) Control, N=200 (%) OR (95 % CI) P value

COX-2 8473 T > C

TT 133 (70) 128 (64) ref.

CT 53 (27.8) 66 (33) 0.77 (0.5–1.19) 0.02

CC 04 (2.10) 06 (03) 0.64 (0.17–2.32) 0.53

CC + CT 57 72 0.76 (0.49–1.96) 0.23

Allele

T 319 322

C 61 78 0.79 (0.54–1.4) 0.2

TT homozygous wild, TC heterozygous, CC homozygous variant

1 2 3 4 5 M 6 7 8 9 10

50bp

100bp

150bp147 bp124 bp

M 1 2 3 4 5 6 7

147 bp100 bp

500 bp

b M 1 2 3 4 5 6 7 8 9 10 11 12

147 bp124 bp100 bp

300 bp

c

Fig. 1 Representative PCRamplification gel picture of COX-2 -8473 T/C. a Lanes 1–7representative amplified PCRproduct of COX-2 8473 T/C(147 bp). Lane M 100-bp DNAmarker. b RFLP picture of COX-2 -8473 T/C after restrictiondigestion with BclI (3.5 %)agarose gel. Lanes 3 and 11homozygous wild TT (147 bp).Lanes 1, 5, and 10 homozygousvariant CC (124+23 bp). Lanes 2,4, 6–9, and 12 heterozygous TC(147, 124, and 23 bp). LaneM 25-bp DNA marker. c RFLP pictureof COX-2 -8473 T/C after re-striction digestion with BclI(20 %) PAGE gel. Lanes 1, 2, 7,and 9 homozygous wild TT(147 bp). Lane 4 homozygousvariant CC (124+23 bp). Lanes 3,5, 6, and 10 heterozygous TC(147, 124, and 23 bp). LaneM 50-bp DNA marker

Tumor Biol.

Author's personal copy

stages I and II and 60 (3158 %) cases who had stages III andIV. Squamous cell carcinoma (SCC) patients were 140(73.69 %), while others which included large cell carcinoma,bronchogenic carcinoma, and adenocarcinoma were 50(26.32 %). Out of 190 patients, 71 had developed grade 1,76 had developed grade 2, and 43 presented with grade 3 non-small cell lung cancer (Table 1).

Analysis of COX-2 polymorphism

DNA extraction was performed using DNA extraction kit(Qiagen, Hilden, NRW, Germany) according to the manufac-turer’s protocol. DNA content was quantified by spectropho-tometric absorption (NanoDrop Spectrophotometer, BioLab,Scoresby, VIC, Australia). PCR was performed using aniCycler Thermal Cycler (Bio-Rad, Hercules, CA, USA).COX-2 -8473 genotyping was determined using PCR–restric-tion fragment length polymorphism (RFLP). Amplification ofthe target region was carried out by polymerase chain reactionusing the specific forward and reverse primers. Primers weredesigned and selected using Primer3 version 0.4.0 software.The forward and reverse primer sequence used for genotypingwas F-5′-GTTTGAAATTTTAAAGTACTTTTGAT-3 andR-5′ TTTCAAATTATTGTTTCATTGC-3 [22]. The PCR re-action mixture consisted of Taq 1.5 U (Ferments), sense andantisense primers (0.5 μmol/l), MgCl2 (50 mmol/l), dNTP(0.2 mmol/l), and DNA template (1 μg) and was subjectedto an initial denaturing step of 4 min at 95 °C and then35 cycles of denaturing for 30 s at 95 °C, annealing for 30 sat 53 °C, extension for 30 s at 72 °C, and a final extension stepof 10 min at 72 °C. Digestion of the amplified products ofCOX-2 was done by using 10 units of restriction endonucle-ases BclI (New England Biolabs), and they were incubated at50 °C for 16 h. The digested products were checked on 3 %agarose gel/20 % polyacrylamide gel electrophoresis (PAGE),and the RFLP picture forCOX-2 genotype was identified as T/T (124/23 bp), T/C (147/124/23 bp), and C/C (147 bp).

Analysis of COX-2 mRNA expression

Total RNAwas extracted using TRIzol (Sigma-Aldrich, USA)from tumor and adjacent normal tissues of lung cancer pa-tients. Integrity of the mRNAwas checked on 1 % agarose geland quantified using 260/280 ratio. Complementary DNA(cDNA) first strand was synthesized using first-strand cDNAsynthesis kit (Fermentas, USA) according to the manufac-turer’s protocol. Dilution of the cDNAwas performed to geta uniform quantity of cDNA in all samples. Primers weredesigned (Primer Express 3.0 software, Applied Biosystems)for COX-2 and glyceraldehyde 3-phosphate dehydrogenase(GAPDH) (GenBank accession number NM-002046) span-ning exon–exon junctions to avoid the amplification of possi-ble traces of genomic DNA contamination. GAPDHwas used

as an internal gene control. For COX-2, primer sequenceswere as follows: forward F-5 CCGACTACGGCGGACTAATC-3′ and reverse F-5 CGCCTGGTTCTAGGAATAATGG-3′ and for GAPDH, and the primers were as follows: forward5′-GATCCGCATAATCTGCATGGT-3 and reverse 5-GATCCGCATAATCTGCATGGT-3. Quantitative real-time PCR(Agilent Biotechnologies, Germany) was performed for thedetection of COX-2, and GAPDH mRNA RT-PCR was per-formed containingMaxima® SYBRGreen qPCRMaster Mix(two times), and all the samples (unknown and standards)were run in triplicates and accompanied by non-templatecontrol (NTC). After initial denaturation at 95 °C for10 min, the reaction mixtures were subjected to 40 cycles of95 °C for 15 s and 55 °C for 1 min, followed by 1 cycle of95 °C for 15 s, 60 °C for 15 s, and 95 °C for 15 s for meltingcurve for determination of genuine products, contamination ofnonspecific products, and primer dimer. All amplified prod-ucts of real-time polymerase chain reaction were subjected forseparation on 2 % agarose gel electrophoresis for ensuring thecorrect amplification products. Delta cycle threshold (ΔCT)method was used to check the COX-2 relative mRNA expres-sion in tumor tissues and their adjacent normal tissue of lungcancer patients by normalization against GAPDH used as areference gene.

Table 3 Association of COX-2 -8473 T/C genotypes with clinicopatho-logic characteristics of study subjects (n=190)

Parameters COX-2 8473 T>C Chi-square (χ2) P value

TT CT + CC

Age (years)

≤50 48 11 5.25 0.02>50 85 46

Gender

Male 112 51 0.91 0.37Female 21 06

Smoking status

Nonsmoker 77 37 0.82 0.42Smoker 56 20

Dweller

Rural 88 39 0.09 1.0Urban 45 18

Histology

SCC 97 43 0.13 0.85Othersa 36 14

Pathological grade

I and II 92 38 0.12 0.86III and IV 41 19

SCC squamous cell carcinomaaOthers include adenocarcinoma, large cell carcinoma, and broncogeniccarcinoma

Tumor Biol.

Author's personal copy

Western blotting

Protein extraction was done by NP-40 lysis buffer fortumor tissues and their corresponding normal tissues oflung cancer patients. The proteins were separated usingSDS-PAGE gel electrophoresis and transferred onto thenitrocellulose membranes. Immunoblotting was carriedout by mouse anti-COX-2 monoclonal antibodies at adilution of 1:1,000 (Cat no. 35-8200, Invitrogen) againsta 70 kDa protein followed by incubation with secondaryantibody (anti-mouse IgG; 1:3,000). β-actin was takenas loading control.

Statistical analysis

Statistical analysis was performed using independent ttest and paired t test for continuous variables andPearson’s χ2 test, Fisher’s exact test, or chi-square test.The associations of COX-2 with lung cancer, as a

dependent variable, with different independent variableswere also investigated by logistic regression analysis toaccount for the effect of the different variables. All reportedP values were based on two-sided tests. Significance level wastaken at P<0.05. Statistical tests were performed using thesoftware SPSS 16.0 (SPSS Inc., Chicago, IL, USA) andGraphPad Prism version 5.00 for Windows (GraphPadSoftware, San Diego, CA, USA).

Fig. 2 RT-PCR showing expression of COX-2 using different parameters. The parameters used were as follows: age, sex, cancer type, cancer stage,smoking status, dwelling habitat, and genotype. Bars represent mean±SD (****P<0.0001; **P<0.01)

Table 4 COX-2 mRNA expression in normal lung tissue samples ofnon-small cell lung cancer cases in relation to genotypes

Genotypes COX-2 mRNA expression(mean±SD)

Number P value

TT 9.09±4.816 35 0.62CT + CC 9.84±4.412 13

TT homozygous wild genotype, CT heterozygous, CC homozygousvariant genotype

Tumor Biol.

Author's personal copy

Results

COX-2 -8473 T > C SNP and its correlation with NSCLC

The COX-2 -8473 T/C gene polymorphisms was suc-cessfully conducted in 190 lung cancer cases and 200healthy controls. The genotype distribution COX-2 -8473 T/C polymorphisms in the cases and controls isshown in Table 2 and Fig. 1. The observed genotypefrequencies of controls for the COX-2 -8473 T/C SNPwas in accordance with the Hardy–Weinberg equilibrium(P>0.005). The polymorphism did not achieve signifi-cant difference in the genotype distribution betweencases and controls (Table 2). COX-2 -8473 CC + CT

genotypes showed insignificant association between var-ious clinical pathological variables except age (Table 3).

Overexpression of COX-2 mRNA expression in NSCLC

In the present study, total COX-2 mRNA was detected andquantified in all 48 lung cancer tissues and their adjacentnormal tissues. The COX-2 relative mRNA expression isexpressed as ΔCT (CT GAPDH−CT COX-2). The ΔCTvalues for COX-2 in 48 samples of lung cancer tissue rangedfrom 0.38 to 18.77 with a mean±standard deviation (SD) of9.25±4.67, while the corresponding values in the matchednon-tumorous lung tissue ranged from 0.44 to 12.90, with amean±SD of 5.635±3.85. COX-2 mRNA expression in lungcancer tissue was significantly higher than that in the matchednon-tumorous lung tissue (P<0.0001) (Fig. 2). We observedinsignificant association of COX-2 -8473 (CT+CC) withCOX-2 mRNA expression in our study (P=0.6) (Table 4,Fig. 2). COX-2 mRNA expression was significantly higherin patients with age less than 50 years (P=0.004) in compar-ison to the patients with age greater than 50 years. However,other clinicopathological variables showed insignificant asso-ciation with COX-2 mRNA expression (Table 5, Fig. 2).

The association of COX-2 expression as a dependent var-iable, with tumor histology, sex, smoking status, grade, stage,and age was also evaluated by multivariate logistic regressionanalysis to take into consideration the reciprocal effects of thecovariates investigated. As presented in Table 6, COX-2 ex-pression was found to be independently associated with theage <50 years old, while other clinicopathological character-istics showed no association.

NSCLC shows an increased COX-2 protein expression

COX-2 protein expressionwas checked byWestern blotting toconfirm the results of RT-PCR. COX-2 protein was detectedin 38 (79 %) tumor tissues of non-small cell lung cancerpatients. Ten (21 %) tumor tissues showed negativeexpression, while in normal tissues, only 9 (18.75 %)showed positive expression and 39 (81 %) showednegative expression. The protein expression of COX-2was fourfold higher in tumor tissues than their adjacent

Table 6 Association betweenCOX-2 expression and indepen-dent covariates computed by lo-gistic regression analysis

Variable Coefficient Standard error Odds ratio 95 % CI P value

Sex −2.291 1.489 0.101 0.005–1.872 0.124

Age −2.315 1.151 0.099 010–0.94 0.044

Pathology 0.609 0.950 1.839 0.29–11.84 0.521

Stage −0.923 0.792 0.398 0.08–1.88 0.244

Smoking 0.279 1.025 1.322 0.18–9.86 0.785

Genotype −0.158 0.837 0.854 0.17–4.401 0.850

Table 5 The association of clinicopathological variables and COX-2mRNA expression in lung cancer

Clinical parameters COX-2 mRNA expression(mean±SD)

Number P value

Tumor 9.25±4.67 48 0.0001Normal 5.63±3.85 48

Age (years)

≤50 11.92±4.09 16 0.004>50 7.98±4.45 32

Sex

Male 8.90±4.49 38 0.26Female 10.78±5.26 10

Smoking status

Non Smoker 9.78±5.37 15 0.6Smoker 9.07±4.38 33

Histology

SCC 9.23±4.38 35 0.64Othersa 9.94±5.57 13

Dwelling

Rural 9.48±4.08 30 0.72Urban 8.98±5.62 18

Pathological stage

I and II 9.60±4.17 31 0.5III and IV 8.73±5.17 17

SCC Squamous cell carcinomaaOthers include adenocarcinoma, bronchogenic carcinoma, and large cellcarcinoma

Tumor Biol.

Author's personal copy

normal tissues of non-small cell lung cancer patients(P<0.0001) (Table 7, Fig. 3).

Discussion

In this study, we investigated COX-2 polymorphism, COX-2mRNA expression, and COX-2 protein expression of non-small cell lung cancer in North Indian population. The distri-bution analysis of three genotypes in our study revealed theallelic frequency of TT/TT, TT/CT, and CC/CC as 70, 27.8,and 2.10 % in NSCLC patients, respectively. The control

population in our study revealed the allelic distribution ofTT/TT, TT/CT, and CC/CC as 64.0, 33.0, and 3.0 %, respec-tively. Thus, in this study, we observed that the -8473 T ascompared to C allele [0.79 (0.54–1.4)] is related with lowersusceptibility of NSCLC and which is in unison with thealready published studies in Asian population [22–24].However, this is in strong contrast to a significant risk ob-served by Campa et al. [25]. Furthermore, we also found asignificant association between the rare allele (CT + CC) withNSCLC patients older than 50 years of age (χ2=5.52,P<0.05) which has not been reported so far, but no significantassociation was observed with other clinicopathologic vari-ables (Table 3). Given its location in the 3′ UTR of COX-2gene, the 8473 T>C polymorphism is a potential candidatewhich modulates the genetic predisposition for NSCLC. Thethymine (T) to cytosine (C) exchange in an AU-rich elementsregion, which is known to control mRNA stability and deg-radation of several other early intermediate genes encodinginflammatory mediators whose mRNA is very unstable, couldenhance the mRNA transcripts’ stability and ultimately lead toan increased prostaglandin production [26]. However, in our

Normal

Tumor

Normal

Tumor

0

2 0

4 0

6 0

8 0

1 0 0

CO

X2

Pro

tein

Ex

pre

ss

ion

Negative Expression Positive Expression

T1 N1 T2 N2 T3 N3 T4 N4

COX 2

β Actin

Fig. 3 Western blot analysis ofCOX-2 protein in lung tumortissues and their adjacent normaltissues. a Extracts from sampleswere separately run for β-actinprotein expression as loadingcontrols. Lane N protein extractedfrom normal tissue. Lane Tprotein extracted from tumortissue. Membrane was probedwith COX-2 mouse monoclonalantibody specific for COX-2(upper bands) and an anti-β-actinantibody (lower bands) forloading control

Table 7 Western blot analysis of COX-2 protein in lung tumor tissue andadjacent normal tissues

Number Negative expression,N (%)

Positive expression,N (%)

P value

48 (normal) 39 (81.25) 9 (18.75) <0.000148 (tumor) 10 (20.83) 38 (79.16)

Tumor Biol.

Author's personal copy

study group, no significant correlation was observed betweenthe -8473 variant C allele and the mRNA expression with themean delta CT value for wild allele as 9.09±4.81 and forvariant allele as 9.84±4.41 (P>0.05).

The COX-2 is generally expressed at low or undetectablelevels, but inducible in response to environmental stress andinternal stimuli. Increased expression levels of COX-2 mRNAand protein have been detected in different histologic types oflung cancer [27] and in about one third of premalignantlesions, such as atypical adenomatous hyperplasia [27, 28].Moreover, COX-2 has pleiotropic effects on the bronchialepithelium that culminates in malignancy, includingproangiogenesis [29], antiapoptosis [30, 31], and local im-mune suppression [18]. Therefore, alteration in COX-2 ex-pression is believed to be an early event in the development oflung cancer [27, 28]. Recent meta-analysis has shown that theabnormal COX-2 expression has the potential to become theindependent prognostic maker in oral cancer [11], osteosarco-ma [12], multiple myeloma [13], ovarian cancer [15], andcolorectal cancer [14]. Combination of COX-2 inhibitors to-gether with different clinically used therapeutic strategies toimprove the efficiency of future anticancer treatment holds agreat promise particularly in the case of breast cancer COX-2together with prostaglandin E (EP) receptors EP1 and EP3 thatcould be important targets for chemosensitization and inhibi-tion of metastasis in breast cancers that are resistant to che-motherapy [32–35]. Several studies have shown that the em-ployment of selective COX-2 inhibitors can reduce lung tu-mor formation in carcinogen-treated animal models [36, 37]and can block the growth of human lung cancer cell bothin vitro and in vivo [38]. COX-2 expression status could beuseful at early stages to distinguish those with worse progno-sis and has the tendency to become an independent prognosticmarker [10, 39–47]. In our studied group of NSCLC, COX-2mRNAwas expressed at significantly higher levels (P=0.007)in tumor tissues of NSCLC patients as compared to theiradjacent normal tissues (ΔCT for cancerous tissue as 9.25±4.67 vs 5.63±3.85 in adjacent normal lung tissue). Our find-ings are similar to those of previously published observations,which have reported an overexpression of COX-2 protein innon-small cell lung cancer, particularly adenocarcinomas [18,28, 41, 48–50]. In another important case–control study on128 cancer patients and 103 healthy donors, serum free DNAlevels and COX-2mRNA expression in peripheral blood weresignificantly lower in healthy donors than in patients (8.8 vs48.0 ng/ml, z=211.17, P=0.001 for DNA; 1.5 vs 2.0, z=26.02, P=0.001 for COX-2) and were not related to sex,age, or smoking habits in either group [51]. In our study, whenthe histopathology, sex, smoking status, age, pathologicalgrade, and pathological stage were tested in a multivariableanalysis against the COX-2 overexpression, the age was foundto have a significant impact on its expression in NSCLC (P=0.044). The different results may be due to the use of different

methods for quantitation of COX-2 mRNA, different primerdesigns resulting in different PCR efficiencies, and relativesmall sample size. COX-2 expression was more in the stages Iand II (ΔCT, 9.60±4.17) as compared to stages III and IV(ΔCT, 8.73±5.17) in our studied group, but no statisticalsignificance was observed. However, a recent meta-analysisshowed a significant increase of COX-2 in stage 1 of NSCLC,suggesting that COX-2 expression could be useful at earlystages to distinguish those with a worse prognosis [47]. Wuet al. have observed co-overexpression of both COX-2 andmPGES-1 to be adversely affecting postoperative overall anddisease-free survivals in NSCLC [52]. Studies have shownthat treatment with NSAIDs decreases cell growth and inducesapoptosis in NSCLC in vitro [28, 38]. Of particular interest isthat the responsiveness to selective COX-2 inhibitors wasinfluenced by the degree of COX-2 expression in advancedNSCLC [28, 53]. Upregulation of COX-2 in the lung cancerpromotes the expression of multidrug resistance-associatedprotein 4 (MRP4) via PGE2-independent pathway and alsoMMP in lung adenocarcinoma cell [54, 55]. The availabilityof selective COX-2 inhibitors like celecoxib gives our resultsadditional importance because increased COX-2 expressionmight represent a direct therapeutic target along with differenttherapeutic strategies like EGFR inhibitors [56] in patientswith NSCLC to further improve the conventional chemother-apeutic treatment modalities. Further studies investigatingselective COX-2 inhibition in patients with NSCLC are war-ranted to gain more insight into the potential clinical useful-ness of this therapeutic approach.

The most interesting finding in our study is the associationbetween high COX-2 mRNA expression levels in more than75 % lung tumor tissues with curatively resected NSCLC.This finding may be clinically important, particularly becauseour study population represents a large number of patientswith different histopathologic subtypes of lung cancer, in all ofwhom curative surgery was achieved. Our observation addsanother step toward the development of an anti-COX-2 drugtherapy for NSCLC.

Acknowledgments The authors gratefully acknowledge the financialsupport provided by the Sher-I-Kashmir Institute of Medical Sciences,Srinagar, for this work. Our thanks are also due to the technical staff,particularly Mr. Fayaz Ahmad and Mr. Rafiq Ahmad of the operationtheater of the Department of Cardiovascular and Thoracic Surgery,SKIMS, Srinagar, who helped us in procuring the tissue samples.

Conflicts of interest None

References

1. Anderson GD, Hauser SD, McGarity KL, Bremer ME, Isakson PC,Gregory SA. Selective inhibition of cyclooxygenase (COX)-2

Tumor Biol.

Author's personal copy

reverses inflammation and expression of COX-2 and interleukin 6 inrat adjuvant arthritis. J Clin Invest. 1996;97(11):2672–9.

2. Williams TJ, Peck MJ. Role of prostaglandin-mediated vasodilata-tion in inflammation. Nature. 1977;270(5637):530–2.

3. Pai R, Soreghan B, Szabo IL, Pavelka M, Baatar D, Tarnawski AS.Prostaglandin E2 transactivates EGF receptor: a novel mechanism forpromoting colon cancer growth and gastrointestinal hypertrophy. NatMed. 2002;8(3):289–93.

4. Salcedo R, Zhang X, Young HA, et al. Angiogenic effects of prosta-glandin E2 are mediated by up-regulation of CXCR4 on humanmicrovascular endothelial cells. Blood. 2003;102(6):1966–77.

5. Kim YB, Kim GE, Cho NH, et al. Overexpression ofcyclooxygenase-2 is associated with a poor prognosis in patientswith squamous cell carcinoma of the uterine cervix treated withradiation and concurrent chemotherapy. Cancer. 2002;95(3):531–9.

6. Zweifel BS, Davis TW, Ornberg RL, Masferrer JL. Direct evidencefor a role of cyclooxygenase 2-derived prostaglandin E2 in humanhead and neck xenograft tumors. Cancer Res. 2002;62(22):6706–11.

7. Dohadwala M, Batra RK, Luo J, et al. Autocrine/paracrine prosta-glandin E2 production by non-small cell lung cancer cells regulatesmatrix metalloproteinase-2 and CD44 in cyclooxygenase-2-dependent invasion. J Biol Chem. 2002;277(52):50828–33.

8. Denkert C, Winzer KJ, Muller BM, et al. Elevated expression ofcyclooxygenase-2 is a negative prognostic factor for disease freesurvival and overall survival in patients with breast carcinoma.Cancer. 2003;97(12):2978–87.

9. Sakamoto A, Yokoyama Y, Umemoto M, et al. Clinical implicationof expression of cyclooxygenase-2 and peroxisome proliferatoractivated-receptor gamma in epithelial ovarian tumours. Br JCancer. 2004;91(4):633–8.

10. Jiang H, Wang J, Zhao W. Cox-2 in non-small cell lung cancer: ameta-analysis. Clin Chim Acta. 2013;419:26–32.

11. Wang ZM, Liu J, Liu HB, Ye M, Zhang YF, Yang DS. AbnormalCOX2 protein expression may be correlated with poor prognosis inoral cancer: a meta-analysis. BioMed Res Int. 2014;2014:364207.

12. Wang Z, He M, Xiao Z, Wu H, Wu Y. Quantitative assessment of theassociation of COX-2 (Cyclooxygenase-2) immunoexpression withprognosis in human osteosarcoma: a meta-analysis. PLoS One.2013;8(12):e82907.

13. Ladetto M, Vallet S, Trojan A, et al. Cyclooxygenase-2 (COX-2) isfrequently expressed in multiple myeloma and is an independentpredictor of poor outcome. Blood. 2005;105(12):4784–91.

14. Peng L, Zhou Y,Wang Y,Mou H, Zhao Q. Prognostic significance ofCOX-2 immunohistochemical expression in colorectal cancer: ameta-analysis of the literature. PLoS One. 2013;8(3):e58891.

15. Lee JY,Myung SK, SongYS. Prognostic role of cyclooxygenase-2 inepithelial ovarian cancer: a meta-analysis of observational studies.Gynecol Oncol. 2013;129(3):613–9.

16. Liu XH, Yao S, Kirschenbaum A, Levine AC. NS398, a selectivecyclooxygenase-2 inhibitor, induces apoptosis and down-regulatesbcl-2 expression in LNCaP cells. Cancer Res. 1998;58(19):4245–9.

17. Masferrer JL, Koki A, Seibert K. COX-2 inhibitors. A new class ofantiangiogenic agents. Ann N YAcad Sci. 1999;889:84–6.

18. Huang M, Stolina M, Sharma S, et al. Non-small cell lung cancercyclooxygenase-2-dependent regulation of cytokine balance in lym-phocytes and macrophages: up-regulation of interleukin 10 anddown-regulation of interleukin 12 production. Cancer Res.1998;58(6):1208–16.

19. Steinbach G, Lynch PM, Phillips RK, et al. The effect of celecoxib, acyclooxygenase-2 inhibitor, in familial adenomatous polyposis. NEngl J Med. 2000;342(26):1946–52.

20. Crane CH, Mason K, Janjan NA, Milas L. Initial experience com-bining cyclooxygenase-2 inhibition with chemoradiation for locallyadvanced pancreatic cancer. Am J Clin Oncol. 2003;26(4):S81–4.

21. Davis TW, O’Neal JM, Pagel MD, et al. Synergy between celecoxiband radiotherapy results from inhibition of cyclooxygenase-2-derived

prostaglandin E2, a survival factor for tumor and associated vascula-ture. Cancer Res. 2004;64(1):279–85.

22. Hu Z, Miao X, Ma H, et al. A common polymorphism in the3′UTR of cyclooxygenase 2/prostaglandin synthase 2 gene andrisk of lung cancer in a Chinese population. Lung Cancer.2005;48(1):11–7.

23. Zhu W, Wei BB, Shan X, Liu P. 765G>C and 8473 T>C polymor-phisms of COX-2 and cancer risk: a meta-analysis based on 33 case-control studies. Mol Biol Rep. 2010;37(1):277–88.

24. Liu F, He Y, Peng X,WangW, Yang X. Association of the 8473T > Ccyclooxygenase-2 (COX-2) gene polymorphism with lung cancerrisk in Asians. APJCP. 2010;11(5):1257–62.

25. CampaD, Zienolddiny S,Maggini V, SkaugV, HaugenA, Canzian F.Association of a common polymorphism in the cyclooxygenase 2gene with risk of non-small cell lung cancer. Carcinogenesis.2004;25(2):229–35.

26. Chan IH, Tang NL, Leung TF, et al. Association of prostaglandin-endoperoxide synthase 2 gene polymorphisms with asthma and atopyin Chinese children. Allergy. 2007;62(7):802–9.

27. Wolff H, Saukkonen K, Anttila S, Karjalainen A, Vainio H, RistimakiA. Expression of cyclooxygenase-2 in human lung carcinoma.Cancer Res. 1998;58(22):4997–5001.

28. Hida T, Yatabe Y, Achiwa H, et al. Increased expression of cycloox-ygenase 2 occurs frequently in human lung cancers, specifically inadenocarcinomas. Cancer Res. 1998;58(17):3761–4.

29. Tsujii M, Kawano S, Tsuji S, Sawaoka H, Hori M, DuBois RN.Cyclooxygenase regulates angiogenesis induced by colon cancercells. Cell. 1998;93(5):705–16.

30. Tsujii M, DuBois RN. Alterations in cellular adhesion and apoptosisin epithelial cells overexpressing prostaglandin endoperoxide syn-thase 2. Cell. 1995;83(3):493–501.

31. Krysan K, Dalwadi H, Sharma S, Pold M, Dubinett S.Cyclooxygenase 2-dependent expression of survivin is critical forapoptosis resistance in non-small cell lung cancer. Cancer Res.2004;64(18):6359–62.

32. Xiang HG, Xie X, Hu FQ, Xiao HB, Zhang WJ, Chen L.Cyclooxygenase-2 inhibition as a strategy for treating gastric adeno-carcinoma. Oncology Reports, 2014

33. Ramon S, Woeller CF, Phipps RP. The influence of Cox-2 andbioactive lipids on hematological cancers. Curr Angiogenesis.2013;2(2):135–42.

34. Kang JH, Song KH, Jeong KC, et al. Involvement of Cox-2 in themetastatic potential of chemotherapy-resistant breast cancer cells.BMC Cancer. 2011;11:334.

35. Lin J, Wu H, Shi H, Pan W, Yu H, Zhu J. Combined inhibition ofepidermal growth factor receptor and cyclooxygenase-2 leads togreater anti-tumor activity of docetaxel in advanced prostate cancer.PLoS One. 2013;8(10):e76169.

36. Rioux N, Castonguay A. Prevention of NNK-induced lung tumori-genesis in A/J mice by acetylsalicylic acid and NS-398. Cancer Res.1998;58(23):5354–60.

37. Masferrer JL, Leahy KM, Koki AT, et al. Antiangiogenic and antitu-mor activities of cyclooxygenase-2 inhibitors. Cancer Res.2000;60(5):1306–11.

38. Hida T, Kozaki K, Muramatsu H, et al. Cyclooxygenase-2 inhibitorinduces apoptosis and enhances cytotoxicity of various anticanceragents in non-small cell lung cancer cell lines. Clin Cancer Res.2000;6(5):2006–11.

39. VanDyke AL, CoteML, Prysak GM, et al. COX-2/EGFR expressionand survival among women with adenocarcinoma of the lung.Carcinogenesis. 2008;29(9):1781–7.

40. Edelman MJ, Watson D, Wang X, et al. Eicosanoid modula-tion in advanced lung cancer: cyclooxygenase-2 expression isa positive predictive factor for celecoxib + chemotherapy—Cancer and Leukemia Group B Trial 30203. J Clin Oncol.2008;26(6):848–55.

Tumor Biol.

Author's personal copy

41. AchiwaH, YatabeY, Hida T, et al. Prognostic significance of elevatedcyclooxygenase 2 expression in primary, resected lung adenocarci-nomas. Clin Cancer Res. 1999;5(5):1001–5.

42. Brabender J, Park J, Metzger R, et al. Prognostic significance ofcyclooxygenase 2 mRNA expression in non-small cell lung cancer.Ann Surg. 2002;235(3):440–3.

43. Kim HS, Youm HR, Lee JS, Min KW, Chung JH, Park CS.Correlation between cyclooxygenase-2 and tumor angiogenesisin non-small cell lung cancer. Lung Cancer. 2003;42(2):163–70.

44. Laga AC, Zander DS, Cagle PT. Prognostic significance of cycloox-ygenase 2 expression in 259 cases of non-small cell lung cancer. ArchPathol Lab Med. 2005;129(9):1113–7.

45. Yuan A, Yu CJ, Shun CT, et al. Total cyclooxygenase-2mRNA levelscorrelate with vascular endothelial growth factor mRNA levels, tu-mor angiogenesis and prognosis in non-small cell lung cancer pa-tients. Int J Cancer. 2005;115(4):545–55.

46. KimGY, Lim SJ, KimYW. Expression of HuR, COX-2, and survivinin lung cancers; cytoplasmic HuR stabilizes cyclooxygenase-2 insquamous cell carcinomas. Mod Pathol. 2011;24(10):1336–47.

47. Zhan P, Qian Q, Yu LK. Prognostic value of COX-2 expression inpatients with non-small cell lung cancer: a systematic review andmeta-analysis. J Thorac Dis. 2013;5(1):40–7.

48. Hosomi Y, Yokose T, Hirose Y, et al. Increased cyclooxygenase 2(COX-2) expression occurs frequently in precursor lesions of humanadenocarcinoma of the lung. Lung Cancer. 2000;30(2):73–81.

49. Ochiai M, Oguri T, Isobe T, Ishioka S, Yamakido M.Cyclooxygenase-2 (COX-2) mRNA expression levels in normal

lung tissues and non-small cell lung cancers. Jpn J Cancer Res.1999;90(12):1338–43.

50. Watkins DN, Lenzo JC, Segal A, Garlepp MJ, Thompson PJ.Expression and localization of cyclo-oxygenase isoforms in non-small cell lung cancer. Eur Respir J. 1999;14(2):412–8.

51. Ulivi P, Mercatali L, Zoli W, et al. Serum free DNA and COX-2mRNA expression in peripheral blood for lung cancer detection.Thorax. 2008;63(9):843–4.

52. Wu YC, Su LJ, Wang HW, et al. Co-overexpression ofcyclooxygenase-2 and microsomal prostaglandin E synthase-1 ad-versely affects the postoperative survival in non-small cell lungcancer. J Thorac Oncol. 2010;5(8):1167–74.

53. Edelman MJ, Hodgson L, Wang X, Kratzke RA, Vokes EE.Cyclooxygenase-2 (COX-2) as a predictive marker for the use ofCOX-2 inhibitors in advanced non-small-cell lung cancer. J ClinOncol. 2012;30(16):2019–20. author reply 2020.

54. Maeng HJ, Lee WJ, Jin QR, Chang JE, Shim WS. Upregulation ofCOX-2 in the lung cancer promotes overexpression of multidrugresistance protein 4 (MRP4) via PGE2-dependent pathway. Eur JPharm Sci. 2014;62:189–96.

55. Zhang J, Luo J, Ni J, et al. MMP-7 is upregulated by COX-2 andpromotes proliferation and invasion of lung adenocarcinoma cells.Eur J Histochem. 2014;58(1):2262.

56. Gadgeel SM, Ali S, Philip PA, Ahmed F, Wozniak A, Sarkar FH.Response to dual blockade of epidermal growth factor receptor(EGFR) and cycloxygenase-2 in nonsmall cell lung cancer may bedependent on the EGFR mutational status of the tumor. Cancer.2007;110(12):2775–84.

Tumor Biol.

Author's personal copy