Subscribe Now Subscribe Today
Research Article

Evaluation of Reference Genes for Normalization of Quantitative Real Time PCR in Non-Small Cell Lung Cancer

Chiaki Endo, Zhifu Sun, Julian R. Molina, John R. Attewell and Ping Yang
Facebook Twitter Digg Reddit Linkedin StumbleUpon E-mail

To compare gene expression levels of lung cancer by real time quantitative reverse transcription polymerase chain reaction (qRT-PCR), choosing appropriate control genes for normalization of RNA input is important. We examined whether frequently used control genes are really appropriate. Eighteen lung cancer tissue samples were used in this study. Gene expression levels were measured by qRT-PCR and compared to the data generated by a DNA microarray experiment. The expression levels of four housekeeping genes that are most frequently used as control genes for qRT-PCR published reports in lung cancer were examined. Four test genes (TP53, laminin γ2, TGFα and cathepsin L2) were also examined. These genes had significantly higher expression in the high aggressive disease group. When taken as single genes, the expression levels of these so called control genes (glyceraldehydes-3phosphate dehydrogenase, TATA box binding protein, β actin and β2 microgloblin) were not consistent among the tested samples. However, the geometric mean of the threshold cycle of the TATA box binding protein and β2 microglobulin showed less variation than that of the single control gene. These results indicate that normalization by multiple control genes is more appropriate for comparison of gene expression levels of lung cancer when using qRT-PCR analysis.

Related Articles in ASCI
Search in Google Scholar
View Citation
Report Citation

  How to cite this article:

Chiaki Endo, Zhifu Sun, Julian R. Molina, John R. Attewell and Ping Yang, 2007. Evaluation of Reference Genes for Normalization of Quantitative Real Time PCR in Non-Small Cell Lung Cancer. International Journal of Cancer Research, 3: 1-12.

DOI: 10.3923/ijcr.2007.1.12



Researchers have investigated the differences between gene expression profiles of non-small cell lung cancer by stage, histology, or other factors using quantitative real time polymerase chain reaction (qRT-PCR). When comparing gene expression levels between samples, internal control genes are most frequently used to normalize the mRNA fraction. Ideally, the reference mRNA should be expressed at a constant level among all samples across all groups under comparison. Therefore, the housekeeping gene appears to be a good reference, although no single housekeeping gene has been a perfect candidate in all situations (Lossos et al., 2003; Beillard et al., 2003; Lion, 2001; Lupberger et al., 2002) and none has been validated in control experiments.

In the published literature, some genes such as TATA box binding protein, Abelson (ABL1), and protein kinase cGMP-dependent type 1 have been recommended as control genes for qRT-PCR of lymphoid tissues after proper validation (Lossos et al., 2003; Beillard et al., 2003). To our knowledge, no reports have been found regarding control genes for accurate normalization in lung cancer. In almost all studies of lung cancer, housekeeping genes have been used without proper validation of their presumed stability of expression. Thus, we evaluated four housekeeping genes frequently used as references in gene expression studies of lung cancer to examine whether or not they are really appropriate for normalization of RNA fraction.

Materials and Methods

Sample Selection
In this study, the same samples were examined using both qRT-PCR and DNA microarray analysis. Samples were separated into two-comparison groups, high vs. low aggressive disease cases as defined in a previous study (Yang et al., 2004). Briefly, all analyzed samples came from patients with stage I Squamous Cell Carcinoma (SCC) of the lung that have undergone surgical resection at Mayo Clinic in Rochester, Minnesota. The high aggressive disease group consisted of patients who died within two years; the low aggressive disease group consisted of patients who lived beyond four years. Eligible study subjects were prospectively diagnosed between 1997 and 2000 and followed through 2003. Patients with high and low aggressive disease were one-to-one matched on age, gender, race, tumor histology, size, smoking status and treatment. Eight high aggressive disease samples and ten low aggressive disease samples were included in our analysis. Eight of the ten low aggressive samples were also matched to the samples in the high-aggressive group. Tumors of each patient were snap frozen after resection and stored at -80°C. Demographic data of the 18 patients are shown in Table 1. Matched pairs were from case 1 and 2 to case 15 and 16. Case 17 and 18 were in the low aggressive group and had no pairs, respectively.

Control Gene Selection
We found 25 articles about lung cancer studies published from April 2003 through March 2004, in which target gene expression levels were quantified relative to a housekeeping gene. β actin (ACTB) was used as a control gene in nine articles (Schneider et al., 2004; Marchetti et al., 2004; Rosell et al., 2004; Marchevsky et al., 2004; Takahashi et al., 2004; Inui et al., 2003; Senchenko et al., 2003; Harden et al., 2003; Inui et al., 2003a, b), glyceraldehyde-3-phosphate dehydrogenase (GAPDH) in seven (Whiteside et al., 2004; Shi et al., 2004; Gu et al., 2004; Cho et al., 2004; Xue Jun et al., 2003; Kawabata et al., 2003; Hu et al., 2003), TATA box binding protein (TBP) in three (Chen et al., 2004; Agathanggelou et al., 2003; Kao et al., 2003) and β2 microglobulin (B2M) (Falleni et al., 2003), β globin (Wu et al., 2003), β glucoronidase (Player et al., 2003), hypoxanthin phosphoriboxyl transferase (Erovic et al., 2003), porphobilinogen deaminase (Steenbergen et al., 2004) and type II collagen (Marchevsky et al., 2004) in one each.

Table 1: Patients’ characteristics
Image for - Evaluation of Reference Genes for Normalization of Quantitative Real Time PCR in Non-Small Cell Lung Cancer
*Stage IA = pT1N0M0 and IB = pT2N0M0. **Survival time is calculated from the day of surgery. ***Low aggressive was defined as surviving more than 4 years and high aggressive was defined as surviving less than 2 years. Case 2 relapsed a year after surgery and received chemoradiotherapy and therefore is considered to be in the high aggressive group despite the survival was longer than 2 years. Case 1 and 2 to Case 15 and 16 were one-to-one matched pairs on age, gender, smoking status, tumor size, tumor histology (all were squamous cell carcinomas), tumor stage and treatment (all underwent complete resections)

However, justification for the choice of a control gene was not provided in any of the 25 articles. Thus, we selected the four most commonly used housekeeping genes (ACTB, GAPDH, TBP, and B2M) to compare the variation of expression levels among samples and to identify the most suitable gene or gene combinations to be used as normalization.

Test Gene Selection
Gene-expression levels of four test genes, TP53, laminin γ2 (LAMC2), TGFα (TGFA), and cathespin L2 (CTSL2) were examined by qRT-PCR with normalization to the control genes. Test genes were randomly selected from a list of genes with significantly higher expression on the high aggressive disease group as determined by data generated from a DNA microarray experiment (Yang et al., 2004).

RNA Extraction
Thirty mm3 of each tissue were sectioned at 20 or 35 μm thickness, collected in buffer RLT (Qiagen, Valencia, CA) supplemented with β-mercaptoethanol and homogenized using PT 1200C rotor/stator (Kinematica AG, Luzern, Switzerland). Total RNA was isolated using the RNeasy kit (Qiagen, Valencia, CA) following the manufacturer’s specifications. Quality and quantity of RNA samples were controlled by spectrophotometry and the Agilent 2100 Bioanalyzer. Hematoxylin-Eosin stained frozen tissue sections were made to confirm that more than 90% of the tested tissue sample consisted of tumor cells (Yang et al., 2004).

Real Time PCR
Total RNA was treated with DNase I (Invitrogen Life Technologies, Inc., Carlsbad, CA). Oligo-dT primers and SuperScript II Reverse Transcriptase (Invitrogen Life Technologies, Inc.) were used in reverse transcription, following the manufacturer’s specifications. The qRT-PCR was performed using a template from the reverse transcription reaction by the ABI Prism 7700 Sequence Detection System (Applied Biosystems, Foster City, CA) with SYBR Green I Mastermix or TaqMan Universal Mastermix (Applied Biosystems). Each PCR reaction mixture contained 1 μL of template (50 ng RNA input) in a final volume of 25 μL with final concentrations of 1xSYBR Green I or 1xTaqMan Mastermix. Thermal cycling conditions comprised an initial incubation at 50°C for two min, DNA polymerase activation at 95°C for 10 min and 40 cycles of denaturation at 95°C for 15 sec and annealing at 60°C for one min. Each measurement was performed in duplicate and the threshold cycle (Ct) was determined.

Sequences of primers and dual labeled probes were as follows:
GAPDH (107 bp), forward: 5’-CATCCATGACAACTTTGGTATCGT-3’, reverse: 5’-CCATCACGCCACAGTTTCC-3’, TaqMan probe: 5’-AAGGACTCATGACCACAGTCCATGCCA-3’; TBP (110 bp), forward: 5’-ACCAGGTGATGCCCTTCTGT-3’, reverse: 5’-GCAGCACGGTATGAGCAACT-3’; ACTB (130 bp), forward: 5’-TCCTCTCCCAAGTCCACACA-3’, reverse: 5’-GGCACGAAGGCTCATCATT-3’; B2M (99 bp), forward: 5’-AGGAGGGCTGGCAACTTAGA-3’, reverse: 5’-GCTTTGAGTGCAAGAGATTGAAGA-3’; TP53 (139 bp), forward: 5’-GCTGGCATTTGCACCTACCT-3’, reverse: 5’- GGGAGAGGGAGAGGGAACA-3’; LAMC2 (98 bp), forward: 5’-TCGGCAGGTTCCCTTACATT-’, reverse: 5’-TTGGCCCACAGACCTTTACC-3’; TGFA (91 bp), forward: 5’- CCTGTGTTTCCTTGCCCTTT-3’, reverse: 5’- CCAGGTGTGTGGCAGCTTAC-3’; CTSL2 (117 bp), forward: 5’- GCCATTCGTCCTTCCAGTTC-3’, reverse: 5’- TCGAATTTGCTCCTTCAAAGC-3’.

The same samples were used for the microarray analysis as described elsewhere (Yang et al., 2004). Briefly, hybridization, washes, and scanning were performed following the manufacturer’s protocols (Affymetrix, Santa Clara, CA). Two-GeneChip®-array sets from Affymetrix (HG-U133) were used, containing 22,283 (chip A) and 22,645 (chip B) probe sets derived from approximately 33,000 well-substantiated human genes. Reproducibility of the chip-set was examined by repeated tests on the same sample. The correlation coefficient is 0.99 for the two duplicate chip sets (from two different manufacturer lots) tested on the same sample three weeks apart.

The microarray data analysis has been reported previously (Yang et al., 2004). In brief, Affymetrix Microarray Analysis Suite, version 5 (MAS5), was used to process scanned chip images. An expression value (signal) and a detection call for each probe set were calculated according to the One-Step Tukey's Biweight Estimate and detection algorithm (Affymetrix, GeneChip Expression Analysis--Data Analysis Fundamentals, Part No. 701190 Rev. 2, 2002). Each chip was globally scaled with a target value of 1,500.

Data Analysis
For quantitative analysis of qRT-PCR, the amplification process was monitored and each amplification curve consisted of at least three distinct phases (initial lag phase, exponential phase, and plateau phase). The template copy number could be estimated with greater precision from the number of cycles needed for the signal to reach an arbitrary threshold. The threshold must intersect the signal curve in its exponential phase, in which the signal elevation correlates with product accumulation. The threshold cycle (Ct) is the intersection point. The Delta Ct is the difference in threshold cycles between the target and reference. Normalized amount of target (X) is X = K (1+E) –delta Ct, where K is a constant and E is the efficiency of amplification. The value of the delta Ct provides a simple tool to estimate relative changes (Wilhelm et al., 2003).

The coefficient of variation (CV) and the standard deviation (SD) were used as statistical measures of the variation of a set of data from its mean. The CV is the ratio of the SD to the average. When the SD changes with the average, the CV is more useful to summarize the variation as long as the average is greater than zero. Correlation coefficients were determined by rank correlation using a nonparametric Spearman's test since the number of samples was relatively small and the relationship between the Ct values and microarray signals were non-linear. The correlation was considered significant when the p-value was less than 0.05.


Specificity of the PCR amplification for each primer pair was determined using agarose gel electrophoresis and confirmed to give a single amplified band of a corresponding size. The standard curves for GAPDH (y = -3.471x+19.94, r2 = 0.955), TBP (y = -3.636x+25.935, r2 = 0.993), ACTB (y = -3.239x+16.422, r2 = 0.953), and B2M (y = -3.405x+15.772, r2 = 0.989) were almost parallel, which indicates the efficiencies of the amplifications are approximately equal among the four control genes (Hamalainen et al., 2001).

The CV of each control gene was 0.090 for GAPDH, 0.048 for TBP, 0.065 for ACTB, and 0.062 for B2M and the CV values of the four genes were small, with the smallest being for TBP (Fig. 1).

Image for - Evaluation of Reference Genes for Normalization of Quantitative Real Time PCR in Non-Small Cell Lung Cancer
Fig. 1: Ct values of 18 cases with four control genes (TBP, GAPDH, ACTB, and B2M). Each bar represents each sample. Ct: threshold cycle, SD: standard deviation, CV: coefficient of variation

Image for - Evaluation of Reference Genes for Normalization of Quantitative Real Time PCR in Non-Small Cell Lung Cancer
Fig. 2: Ct values of four control genes (TBP, GAPDH, ACTB, and B2M) with 18 cases. Each line represents each control gene. Three lines (TBP, ACTB, and B2M) were nearly parallel and flat and changed within 5 Ct. Ct: threshold cycle

Figure 2 shows variations in control gene expression. Each line represents the data from one control gene. Connecting lines of TBP, ACTB, and B2M are nearly parallel and flat within a 5 Ct difference. This indicates that variations in Ct values could be reflected as inevitable differences like variations in the initial amount of RNA and the efficiency of the cDNA synthesis and the expression levels of these genes (TBP, ACTB and B2M) are relatively constant through all samples. If there are substantial differences of gene expression levels among comparison groups, the connecting lines may not run parallel.

Table 2: Spearman’s correlation coefficient between the microarray data and *delta Ct values of four test genes normalized by four control genes
Image for - Evaluation of Reference Genes for Normalization of Quantitative Real Time PCR in Non-Small Cell Lung Cancer
*The Delta Ct was the Ct value of the test gene minus that of the control gene. **p<0.05 by Spearman’s test. A lower Delta Ct value means a higher gene expression level and a higher value of microarray data means a higher gene expression level. Accordingly, the negative correlation means the same direction of gene expression change. The Delta Ct for TP53 normalized by any control gene showed a statistically significant correlation to the microarray data. The Delta Ct for TGFAα normalized by any control gene but GAPDH and that for CTSL2 normalized by any control gene also showed significant correlation. However, the Delta Ct for LAMC2 normalized by any control gene showed no significant correlation

Image for - Evaluation of Reference Genes for Normalization of Quantitative Real Time PCR in Non-Small Cell Lung Cancer
Fig. 3: Difference of Ct values for four control genes (GAPDH, TBP, ACTB, and B2M) between the matched pairs. Mean difference was less than 2 Ct for all control genes. ACTB showed the smallest mean and SD, both of which were less than 1 Ct. ACTB showed the smallest 95% CI, as well. Ct: threshold cycle, SD: standard deviation, 95% CI: 95% confidential interval

For example, if one gene expressed higher and one gene expressed lower in the high aggressive group and yet another gene expressed constant in both groups, the connected lines of the three genes would not have run parallel or probably have crossed each other.

As our samples were one-to-one matched except for one variable, survival time, control genes are supposedly expressed at the same level between the pairs unless they are associated with survival. Figure 3 shows the difference of gene expression levels (i.e., the difference of the Ct value) between the matched pairs. Since ideally the Ct difference between the pairs is equal or close to zero, the best control gene should have the smaller mean of the Ct difference with a smaller variation. The gene expression differences between pairs had the smallest mean, SD and 95% confidential interval (CI) when normalized by ACTB.

The correlation between qRT-PCR data and microarray data was examined. In general, the microarray data is normalized by multiple control genes and its normalization is thought to be reliable, so the normalized Ct values (i.e., Delta Ct values) should be comparable to the microarray data. Table 2 shows that Delta Ct values of the all test genes except LAMC2 were well correlated with the microarray data.

Image for - Evaluation of Reference Genes for Normalization of Quantitative Real Time PCR in Non-Small Cell Lung Cancer
Fig. 4: Delta Ct values of four test genes in the low and high aggressive groups normalized by four control genes. The error bar represents the mean±standard deviation. All test genes but LAMC2 showed lower Delta Ct values in the high aggressive group. LAMC2 normalized by GAPDH and ACTB showed nearly equal values between the two groups. H: High aggressive group, L: Low aggressive group

Since the four test gene expression levels were significantly higher for the high aggressive group than the low aggressive group in the microarray data, the Delta Ct of the four genes in the high aggressive group is bound to show a smaller value than that of the low aggressive group if the normalization is conducted appropriately. Our results showed this was true regardless of which control gene was used for normalization (Fig. 4). However, LAMC2 normalized by GAPDH or ACTB showed that the values of the Delta Ct between both groups were almost equal, emphasizing the importance of appropriately selecting control genes in order to detect a subtle difference of expression level.


Accurate normalization of gene-expression levels in qRT-PCR analysis is very important for reliable results, especially when the relationship between the clinical findings and the subtle gene-expression level difference is studied. Several systematic studies of normalization procedures have been reported (Tricarico et al., 2002; Vandesompele et al., 2002; Hamalainen et al., 2001; Liu et al., 2002) and extensive analyses in the cancer field have been conducted to evaluate different control genes and to select a suitable control gene for the normalization of expression data (Lossos et al., 2003; Beillard et al., 2003; Lupberger et al., 2002). In contrast, evaluation of control genes has never been investigated for lung cancer. As a standard approach, experiments use only one control gene for sample normalization. This is usually done because of convenience, availability in the lab or because it was done in a similar experiment but, without any validation for selection. No recent published reports were found for control genes in lung cancer, in which the reason for selection of the control gene was stated. In this study, four housekeeping genes were selected as control genes because they have been most frequently used for normalization of qRT-PCR data in lung cancer reports published from April 2003 to March 2004. We compared the Ct values of the control genes with each other and also with their microarray data.

Although GAPDH varied more than 5 Ct, 32 fold in Fig. 1, similar variation has also been found in other publications evaluating control genes. Ct values of B2M varied more than 5 Ct among non-Hodgkin’s lymphomas (Lossos et al., 2003) and those of TBP varied more than 5 Ct among mononuclear cells (Beillard et al., 2003). In comparison with the variation of the Ct value (Fig. 1-3), TBP, B2M, and ACTB were considered to be more constant than GAPDH. However, there is no single control gene which was constantly expressed in all analyses, although no control gene showed the outlier. GAPDH was not the best choice as a control in any of the analyses reported here.

Moreover, the Delta Cts of the three test genes that were normalized with the four control genes were correlated to the microarray data (Table 2 and Fig. 4). Even when a test gene expression level showed a subtle difference like LAMC2 (the mean Delta Ct difference was less than one Ct), TBP, and B2M showed the same tendency of expression level as in the microarray data (Fig. 4). Its tendency, however, showed no statistical significance, which was due to the small number of samples and the small difference of the mean expression level between the two groups. Only a 2.8 Ct difference can be significantly detected with nine samples in each group if it is assumed that the SD is 2 and the type I error and power are 5 and 80%, respectively. More than 64 samples in each group are needed to detect a 1 Ct difference with 5% type I error and 80% power. Therefore, a set of TBP and B2M should be considered as appropriate control genes for normalization of the qRT-PCR data for lung cancer studies.

Beillard et al. (2003) reported that the selected control genes should fulfill the absence of pseudogenes and high or medium expression. Since the Ct values of three of the control genes reported here (TBP, B2M, and ACTB) were between 15 and 30, these gene-expression levels can be considered not to be very high or low. TBP and B2M have no pseudogenes (Beillard et al., 2003; Vandesompele et al., 2002). Ideally, the control gene expression level would not change over conditions. The gene expression levels of some frequently used control genes have been reported to change under certain conditions. Lossos et al. (2002) reported that alterations in cell morphology associated with tumor progression or transformation were related to changes in ACTB expression. Aviles et al. (1992) reported how B2M expression varied in different lymphomas. GAPDH expression has been reported to exhibit marked variability among normal colon epithelium (Bustin, 2000), prostate cancer (Ripple et al., 1995), during the cell cycle (Mansur et al., 1993), and to correlate to VEGF165/189 expression level (Tricarico et al., 2002). Hypoxia is a common condition in solid tumors and it is known to stimulate GAPDH expression as well as VEGF expression (Graven et al., 1994; Zhong et al., 1999). As smoking and COPD, which are risk factors of lung cancer, are related to hypoxia, the relationship between hypoxia and lung cancer may be more significant than between hypoxia and other solid tumors. Therefore, it is important that normalization using GAPDH for lung cancer studies is carefully considered. Moreover, Vandesompele et al. (2002) reported the possible errors related to the common practice of using only one housekeeping gene for normalization and recommended normalization by the geometric mean of multiple housekeeping genes. In the present study, the CV of the geometric mean of TBP and B2M was 0.043, of TBP and ACTB was 0.058, of B2M and ACTB was 0.058, and of TBP, ACTB and B2M was 0.048, with the CV of the geometric mean of TBP and B2M being the smallest value (Table 3). In Table 2, the geometric mean of TBP and B2M showed good correlation to the microarray data. Moreover, in Fig. 3, the geometric mean of TBP and B2M showed the smallest mean of the Ct difference with the smallest variation. Consequently, normalization of the real time PCR data by the geometric mean of TBP and B2M could be most appropriate.

Table 3: Coefficient of variation (CV) for geometric mean of Ct values Of TBP, ACTB and B2M
Image for - Evaluation of Reference Genes for Normalization of Quantitative Real Time PCR in Non-Small Cell Lung Cancer
Ct: Threshold cycle, CV: Coefficient of variation, The geometric mean of Ct values of TBP and B2M showed the least CV

Since we examined only four housekeeping genes, examination of other housekeeping genes might show more constant expression level. Some investigators have utilized microarray data for choosing control genes. Wikman et al. (2002) chose phopholipase 2A as a reference due to its constant expression at a moderate level in the cDNA array data for lung adenocarcinomas. However, they did not confirm its stability using real time PCR. Hamalainene et al. (2001) selected control genes based on the results of an oligonucleotide microarray for T helper cells and evaluated them using real time PCR.

We examined the validity of the four control genes used frequently for normalization of qRT-PCR data in lung cancer. Our results show that among the four housekeeping genes examined, there is no single control gene that expresses consistently in all analyses. If the microarray data for the same samples can be utilized as in the present study, we may select control genes of qRT-PCR according to the microarray data. But such a situation must be rare in the qRT-PCR study for lung cancer. Based on the data presented in this study, it is suggested that normalization using multiple control genes (geometric mean of TBP and B2M) is more appropriate than a single control gene.

This study has been the first report on the evaluation of control genes for qRT-PCR study in lung cancer, although a relatively small number of samples were used. In leukemia studies, multi-center trials have been conducted in order to evaluate candidate control genes using real time PCR (Beillard et al., 2003). Such trials should be planned for lung cancer studies in the near future.


The authors would like to thank Ms. Susan Ernst for her technical assistance with the manuscript and Drs. Farhad Kosari and Marie Kube for their helpful advice on laboratory experiments. This research was partially supported by: NIH Grant R01 CA 84354 (Yang), Mayo Foundation Funds, and Tohoku University, Department of Thoracic Surgery, Institute of Development, Aging and Cancer, Japan.


1:  Agathanggelou, A., I. Bieche, J. Ahmed-Choudhury, B. Nicke and R. Dammann et al., 2003. Identification of novel gene expression targets for the Ras association domain family 1 (RASSF1A) tumor suppressor gene in non-small cell lung cancer and neuroblastoma. Cancer Res., 63: 5344-5351.
Direct Link  |  

2:  Aviles, A., G. Zepeda, J.C. Diaz-Maqueo, L. Rodriguez, R. Guzman, E.L. Garcia and A. Talavera, 1992. Beta 2 microglobulin level as an indicator of prognosis in diffuse large cell lymphoma. Leuk. Lymphoma, 7: 135-138.
Direct Link  |  

3:  Beillard, E., N. Pallisgaard, V.H. van der Velden, W. Bi, R. Dee and E.J. van der Schoot et al., 2003. Evaluation of candidate control genes for diagnosis and residual disease detection in leukemic patients using real-time quantitative reverse-transcriptase polymerase chain reaction (RQ-PCR): A Europe against cancer program. Leukemia, 17: 2474-2486.
Direct Link  |  

4:  Bustin, S.A., 2000. Absolute quantification of mRNA using real-time reverse transcription polymerase chain reaction assays. J. Mol. Endocrinol., 25: 169-193.
CrossRef  |  PubMed  |  Direct Link  |  

5:  Chen, H.W., S.L. Yu, J.J. Chen, H.N. Li and Y.C. Lin et al., 2004. Anti-invasive gene expression profile of curcumin in lung adenocarcinoma based on a high throughput microarray analysis. Mol. Pharmacol., 65: 99-110.
Direct Link  |  

6:  Cho, N.H., K.P. Hong, S.H. Hong, S. Kang, K.Y. Chung and S.H. Cho, 2004. MMP expression profiling in recurred stage IB lung cancer. Oncogene, 23: 845-851.
CrossRef  |  Direct Link  |  

7:  Erovic, B.M., J. Pammer, D. Hollemann, M. Woegerbauer and S. Geleff et al., 2003. Motility-related protein-1/CD9 expression in head and neck squamous cell carcinoma. Head Neck, 25: 848-857.
Direct Link  |  

8:  Falleni, M., C. Pellegrini, A. Marchetti, B. Oprandi and F. Buttitta, 2003. Survivin gene expression in early-stage non-small cell lung cancer. J. Pathol., 200: 620-626.
Direct Link  |  

9:  Graven, K.K., R.F. Troxler, H. Kornfeld, M.V. Panchenko and H.W. Farber, 1994. Regulation of endothelial cell glyceraldehyde-3-phosphate dehydrogenase expression by hypoxia. J. Biol. Chem., 269: 24446-24453.
Direct Link  |  

10:  Gu, J., L. Zhang, S.G. Swisher, J. Liu, J.A. Roth and B. Fang, 2004. Induction of p53-regulated genes in lung cancer cells: Implications of the mechanism for adenoviral p53-mediated apoptosis. Oncogene, 23: 1300-1307.
Direct Link  |  

11:  Hamalainen, H.K., J.C. Tubman, S. Vikman, T. Kyrola, E. Ylikoski, J.A. Warrington and R. Lahesmaa, 2001. Identification and validation of endogenous reference genes for expression profiling of T helper cell differentiation by quantitative real-time RT-PCR. Anal. Biochem., 299: 63-70.
Direct Link  |  

12:  Harden, S.V., Y. Tokumaru, W.H. Westra, S. Goodman, S.A. Ahrendt, S.C. Yang and D. Sidransky, 2003. Gene promoter hypermethylation in tumors and lymph nodes of stage I lung cancer patients. Clin. Cancer Res., 9: 1370-1375.
Direct Link  |  

13:  Hu, Y.P., G. Cherton-Horvat, V. Dragowska, S. Baird and R.G. Korneluk et al., 2003. Antisense oligonucleotides targeting XIAP induce apoptosis and enhance chemotherapeutic activity against human lung cancer cells in vitro and in vivo. Clin. Cancer Res., 9: 2826-2836.
Direct Link  |  

14:  Inui, N., K. Kitagawa, S. Miwa, T. Hattori, K. Chida, H. Nakamura and M. Kitagawa, 2003. High expression of Cks1 in human non-small cell lung carcinomas. Biochem. Biophys. Res. Commun., 303: 978-984.
Direct Link  |  

15:  Inui, N., S. Sasaki, T. Suda, K. Chida and H. Nakamura, 2003. The loss of retinoic acid receptor alpha, beta and alcohol dehydrogenase3 expression in non-small cell lung cancer. Respirology, 8: 302-309.
Direct Link  |  

16:  Kao, Y.R., J.Y. Shih, W.C. Wen, Y.P. Ko and B.M. Chen et al., 2003. Tumor-associated antigen L6 and the invasion of human lung cancer cells. Clin. Cancer Res., 9: 2807-2816.
Direct Link  |  

17:  Kawabata, S., M. Oka, H. Soda, K. Shiozawa and K. Nakatomi et al., 2003. Expression and functional analyses of breast cancer resistance protein in lung cancer. Clin. Cancer Res., 9: 3052-3057.
Direct Link  |  

18:  Lion, T., 2001. Current recommendations for positive controls in RT-PCR assays. Leukemia, 15: 1033-1037.
Direct Link  |  

19:  Liu, W. and D.A. Saint, 2002. A new quantitative method of real time reverse transcription polymerase chain reaction assay based on simulation of polymerase chain reaction kinetics. Anal. Biochem., 302: 52-59.
Direct Link  |  

20:  Lossos, I.S., A.A. Alizadeh, M. Diehn, R. Warnke and Y. Thorstenson, 2002. Transformation of follicular lymphoma to diffuse large-cell lymphoma: Alternative patterns with increased or decreased expression of c-myc and its regulated genes. Proc. Natl. Acad. Sci. USA., 99: 8886-8891.
Direct Link  |  

21:  Lossos, I.S., D.K. Czerwinski, M.A. Wechser and R. Levy, 2003. Optimization of quantitative real-time RT-PCR parameters for the study of lymphoid malignancies. Leukemia, 17: 789-795.
Direct Link  |  

22:  Lupberger, J., K.A. Kreuzer, G. Baskaynak, U.R. Peters, P. le Coutre and C.A. Schmidt, 2002. Quantitative analysis of beta-actin, beta-2-microglobulin and porphobilinogen deaminase mRNA and their comparison as control transcripts for RT-PCR. Mol. Cell Probes, 16: 25-30.
Direct Link  |  

23:  Mansur, N.R., K. Meyer-Siegler, J.C. Wurzer and M.A. Sirover, 1993. Cell cycle regulation of the glyceraldehyde-3-phosphate dehydrogenase/uracil DNA glycosylase gene in normal human cells. Nucleic Acids Res., 21: 993-998.
Direct Link  |  

24:  Marchetti, A., F. Barassi, C. Martella, A. Chella, S. Salvatore, A. Castrataro, F. Mucilli, R. Sacco and F. Buttitta, 2004. Down regulation of high in normal-1 (HIN-1) is a frequent event in stage I non-small cell lung cancer and correlates with poor clinical outcome. Clin. Cancer Res., 10: 1338-1343.
Direct Link  |  

25:  Marchevsky, A.M., J.A. Tsou and I.A. Laird-Offringa, 2004. Classification of individual lung cancer cell lines based on DNA methylation markers: Use of linear discriminant analysis and artificial neural networks. J. Mol. Diagn., 6: 28-36.
Direct Link  |  

26:  Player, A., J. Gillespie, T. Fujii, J. Fukuoka, T. Dracheva, D. Meerzaman, K.M. Hong, J. Curran, G. Attoh, W. Travis and J. Jen, 2003. Identification of TDE2 gene and its expression in non-small cell lung cancer. Int. J. Cancer, 107: 238-243.
Direct Link  |  

27:  Ripple, M.O. and G. Wilding, 1995. Alteration of glyceraldehyde-3-phosphate dehydrogenase activity and messenger RNA content by androgen in human prostate carcinoma cells. Cancer Res., 55: 4234-4236.
Direct Link  |  

28:  Rosell, R., K.D. Danenberg, V. Alberola, G. Bepler, J.J. Sanchez, C. Camps, M. Provencio, D. Isla, M. Taron, P. Diz and A. Artal, 2004. Ribonucleotide reductase messenger RNA expression and survival in gemcitabine/cisplatin-treated advanced non-small cell lung cancer patients. Clin. Cancer Res., 10: 1318-1325.
Direct Link  |  

29:  Schneider, S., J. Yochim, J. Brabender, K. Uchida, K.D. Danenberg, R. Metzger, P.M. Schneider, D. Salonga, A.H. Holscher and P.V. Danenberg, 2004. Osteopontin but not osteonectin messenger RNA expression is a prognostic marker in curatively resected non-small cell lung cancer. Clin. Cancer Res., 10: 1588-1596.
Direct Link  |  

30:  Senchenko, V., J. Liu, E. Braga, N. Mazurenko, W. Loginov, Y. Seryogin, I. Bazov, A. Protopopov, F.L. Kisseljov, V. Kashuba, M.I. Lerman, G. Klein and E.R. Zabarovsky, 2003. Deletion mapping using quantitative real-time PCR identifies two distinct 3p21.3 regions affected in most cervical carcinomas. Oncogene, 22: 2984-2992.
Direct Link  |  

31:  Shi, W., H. Chen, J. Sun, C. Chen, J. Zhao, Y.L. Wang, K.D. Anderson and D. Warburton, 2004. Overexpression of Smurf1 negatively regulates mouse embryonic lung branching morphogenesis by specifically reducing Smad1 and Smad5 proteins. Am. J. Physiol. Lung Cell Mol. Physiol., 286: L293-L300.
Direct Link  |  

32:  Steenbergen, R.D., D. Kramer, B.J. Braakhuis, P.L. Stern, R.H. Verheijen, C.J. Meijer and P.J. Snijders, 2004. TSLC1 gene silencing in cervical cancer cell lines and cervical neoplasia. J. Natl. Cancer Inst., 96: 294-305.
Direct Link  |  

33:  Takahashi, Y., J. Hamada, K. Murakawa, M. Takada and M. Tada et al., 2004. Expression profiles of 39 HOX genes in normal human adult organs and anaplastic thyroid cancer cell lines by quantitative real-time RT-PCR system. Exp. Cell Res., 293: 144-153.
Direct Link  |  

34:  Tricarico, C., P. Pinzani, S. Bianchi, M. Paglierani and V. Distante et al., 2002. Quantitative real-time reverse transcription polymerase chain reaction: Normalization to rRNA or single housekeeping genes is inappropriate for human tissue biopsies. Anal. Biochem., 309: 293-300.
PubMed  |  Direct Link  |  

35:  Vandesompele, J., K. De Preter, F. Pattyn, B. Poppe, N. van Roy, A. De Paepe and F. Speleman, 2002. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol., 3: 1-11.
CrossRef  |  Direct Link  |  

36:  Whiteside, M.A., D.T. Chen, R.A. Desmond, S.A. Abdulkadir and G.L. Johanning, 2004. A novel time-course cDNA microarray analysis method identifies genes associated with the development of cisplatin resistance. Oncogene, 23: 744-752.
Direct Link  |  

37:  Wikman, H., E. Kettunen, J.K. Seppanen, A. Karjalainen, J. Hollmen, S. Anttila and S. Knuutila, 2002. Identification of differentially expressed genes in pulmonary adenocarcinoma by using cDNA array. Oncogene, 21: 5804-5813.
Direct Link  |  

38:  Wilhelm, J. and A. Pingoud, 2003. Real-time polymerase chain reaction. Chembiochem, 4: 1120-1128.

39:  Wu, Y., B. Han, H. Sheng, M. Lin, P.A. Moore, J. Zhang and J. Wu, 2003. Clinical significance of detecting elevated serum DcR3/TR6/M68 in malignant tumor patients. Int. J. Cancer, 105: 724-732.
Direct Link  |  

40:  Xue Jun, H., A. Gemma, Y. Hosoya, K. Matsuda and M. Nara et al., 2003. Reduced transcription of the RB2/p130 gene in human lung cancer. Mol. Carcinog., 38: 124-129.
Direct Link  |  

41:  Yang, P., Z. Sun, M. Aubry, F. Kosari, W. Bamlet, C. Endo, R. Molina and G. Vasmatzis, 2004. Study design considerations in clinical outcome research of lung cancer using microarray analysis. Lung Cancer, 46: 215-226.
Direct Link  |  

42:  Zhong, H. and J.W. Simons, 1999. Direct comparison of GAPDH, beta-actin, cyclophilin and 28S rRNA as internal standards for quantifying RNA levels under hypoxia. Biochem. Biophys. Res. Commun., 259: 523-526.
Direct Link  |  

©  2021 Science Alert. All Rights Reserved