Gene expression analysis in superficial bladder cancer: comparison of two suitable endogenous reference genes
Running title: Gene expression normalization in superficial bladder cancer
Authors: Paula A. Videira1, Dário Ligeiro2, Manuela Correia1, Hélder Trindade1,
Affiliation of the authors: 1
Departamento de Imunologia FCM-UNL, Campo Mártires da Pátria 130, 1169-056
Lisboa, Portugal. 2
Centro de Histocompatibilidade do Sul, Lisboa, Portugal.
To whom correspondence should be addressed: Paula A. Videira, Departamento de Imunologia FCM-UNL, Campo Mártires da Pátria 130, 1169-056 Lisboa, Portugal. Telephone: +351 218 803 045; Fax: +351 218853480 Email:
[email protected]
1
Keywords Superficial bladder cancer; normalization; endogenous control gene; real-time reverse transcription PCR; β-actin; GAPDH
Abstract Gene expression studies demand that data is normalized to an endogenous control gene. The selection of this control is not straightforward, in particular for tumour tissue samples. We have measured the expression of genes engaged in the immune response, in matched samples of superficial bladder cancer relative to normal urothelium. The βactin and GAPDH genes were selected from a panel of common reference genes as the most stably expressed, and were, hence, used for normalization. Even so, we found that in some samples, the immunity genes expression, when normalized to either β-actin or to GAPDH was opposing. The combination of both reference genes was clearly more adequate to be considered. Nevertheless, the gene expression studies should be regarded as qualitative rather than quantitative. Our study stresses the required prudence when selecting endogenous control genes. The gold standard normalization is most probably obtained with the mean expression of multiple endogenous controls.
2
Introduction The fluorescence-based real-time reverse transcription PCR (RT-PCR) is a widely used technology for the quantification of mRNA level, and is becoming a critical tool for basic and clinical research [1]. RT-PCR is now applied to identify genes which correlate with disease and the inflammatory response to disease. Its advances in cancer have allowed the identification of tumour markers and have improved the tumour diagnosis and prognosis [2]. For the RT-PCR, it is mandatory to relate, i.e., to normalize each gene expression to a stably expressed reference gene, referred as endogenous control gene. The ideal endogenous control is usually an housekeeping gene, which means its expression should be sufficiently abundant, constant across samples and be unaffected by experimental treatments [1, 3]. Including an invariant endogenous control gene in the assay, corrects for sample to sample variations in RT-PCR efficiency and sample amount differences; which is essential for the reliability of any relative RT-PCR experiment [4]. Nevertheless, the selection of the appropriate endogenous control is not straightforward. The genes described as having an invariant expression by the great majority of the studies, have also been shown to be cell or tissue dependent [4] and to vary in certain circumstances, such as cell proliferation [5], hypoxia or experimental treatment [6] and neoplastic cell growth [7, 8]. Therefore, when it comes to the identification of genes with altered expression in tumours, the use of endogenous control genes for normalization may have increased pitfalls. In the present study, we have analysed the expression of ten genes in samples of superficial bladder cancer relative to normal urothelium. The study was initiated in
3
December 2005 and aimed, originally, to study the expression of genes engaged in the immune response in superficial bladder cancer. The samples were obtained from thirty patients. For normalization, we have analysed a panel of eleven common endogenous control genes, and the GAPDH and β-actin genes were selected for showing the most constant and stable expression in these tissues. Even though, the data here presented shows the unanticipated variations in relative gene expression found in some samples, when using these endogenous control genes alone or combined.
Materials and methods Patient tissue samples Matched pairs of cancerous bladder tissue and normal urothelium samples were obtained from 30 patients with a mean age of 69,3 years (range 45 to 88) with diagnosed superficial bladder cancer (tumour staged at pTa and pT1, according to the International Union Against Cancer - UICC) after transurethral resection of the tumours. Tissue samples were immediately immersed in RNAlater® RNA Stabilization Reagent, stored at 4º C, overnight and then preserved until further processing at -20º C. Informed consent was obtained from all patients and the use of the tissue material was approved by the local ethics committee.
RNA Isolation and Characterization Preserved tissue samples (20 to 50 mg wet weight) were potter homogenized and the total RNA was isolated using the GenElute Mammalian Total RNA Purification kit and DNAase treatment (Sigma), according to the manufacturer’s instructions. RNA concentrations and A260-to-A280 ratios were measured in a spectrophotometer and only samples with A260/A280 ratios between 1,9 and 2,1 were considered further.
4
Between 250 to 500 ng of total RNA were reverse transcribed with random primers, using the High Capacity cDNA Archive Kit (Applied Biosystems, Foster City, CA).
Expression of endogenous control genes The expression of endogenous control genes was assessed using the TaqMan Human Endogenous Control Plate (Applied Biosystems) according to manufacturer’s protocol. This plate evaluates the expression of eleven common endogenous control genes [18S rRNA (18S), Acidic ribosomal protein (ARP), Beta-actin (β-actin), Cyclophilin (CYC), Glyceraldehyde-3-phosphate dehydrogenase (GAPDH), Phosphoglycerokinase (PGK), β2-Microglobulin (B2m), β-Glucoronidase (GUS), Hypoxanthine ribosyl transferase (HPRT), Transcription factor IID TATA binding protein (TBP) and Transferrin receptor (TfR)]. Matched samples of cancerous and normal urothelium of three randomly chosen patients were used for testing. In these assays, exactly 500 ng of total RNA was reverse transcribed into cDNA as described.
Gene Expression Measurements RT-PCR was performed in a 7500 Fast system (Applied Biosystems) using TaqManFast Universal PCR Master Mix, primers and Taqman probes provided by Applied Biosystems and following the manufacturer’s instructions. The analyzed genes were MIP-1α (Hs00234142_m1), MCP-1 (Hs00234140_m1), MCP-2 (Hs00271615_m1), IP10 (Hs00171042_m1), MIG (Hs00171065_m1), CD1A (Hs00233332_m1), CD1B (Hs00233507_m1), CD1C (Hs00233509_m1), CD1D (Hs00174321_m1) and CD1E (Hs00229421_m1). Each reaction was performed in duplicate. All genes, including the endogenous controls, and the matched samples of cancerous and noncancerous tissues were always analyzed in the same run to exclude between-run variations. The number of
5
PCR cycles needed to reach the fluorescence threshold in each sample was defined as the cycle threshold (Ct). The relative expression for each gene was calculated by using the formula 2-∆∆Ct, according to the method described by Livak and Schmittgen [9]. ∆Ct stands for the difference between the Ct of the amplification curve of the target gene and that of the endogenous control gene. The efficiency of the amplification reaction for each primerprobe was above 95% (as determined by the manufacturer).
Data analysis Data was analysed for statistical significance using the Student’s t-test and ANOVA Oneway (GraphPad Prism Version 4.0). Correlations between variables were tested using the Spearman correlation analysis. P values < 0,05 were considered as significant. To calculate the expression stability of the candidate endogenous control genes, the freely available
software
packages
geNorm™,
http://medgen.ugent.be/~jvdesomp/genorm/],
version NormFinder18
3.5
[[10], [[11],
http://www.mdl.dk/publicationsnormfinder.htm] and BestKeeper1 [[12], http://www.genequantification.de/bestkeeper.html] were used.
Results Selection of reference genes In order to select two of the most appropriated endogenous control gene to initiate the study of the relative gene expression in superficial bladder cancer, we have first performed a MEDLINE search for the most common genes used by others to normalize the gene expression in bladder cancer or other neoplasms. According to our search, the β-actin or the GAPDH genes were the most frequently used for normalization in the
6
majority of neoplasms, including in bladder cancer measurements [13-15]. As in recent publications, these genes are presented as inappropriate for general bladder cancer (tumours staged between pTa and pT4) [11, 16], we have analysed few patient samples, for the expression of eleven genes, commonly used as reference genes. In order to rank the reference genes, based on their expression stability, determining the most stable genes for normalization, we have used three different methods, the geNorm, NormFinder and BestKeeper software. geNorm calculates the gene expression stability measure, M, for a reference gene as the average pairwise variation for that gene with all other tested reference genes. Among the analysed reference genes, the most stable were ranked as follows, according to the lowest M value (Table 1): GAPDH > β-actin > TfR> HPRT> TBP> PGK> CYC> ARP> GUS> B2m> 18S. All these genes reach a reasonable high expression stability with M values less than 1, below the default limit of M = 1,5 [10]. NormFinder is another tool which estimates the intra- and intergroup variation and combines both results in a stability value for each reference gene [11]. Genes with the lowest stability value are the most stable within the group studied. The results of the NormFinder analysis applied to our data have given a ranking identical to the geNorm (Table 1). According to these results, the 18S is the most unsuited reference gene and GAPDH is the most stable gene, closely followed by the β-actin. BestKeeper uses a pair-wise correlation analysis of all pairs of candidate genes [12]. In a first analysis of the data, based on the inspection of raw Ct values, BestKeeper calculates variations of all reference gene (SD [± Ct]) and a low SD is expected for useful reference genes. In a second analysis, BestKeeper estimates the correlations in the expression levels between all the possible candidates. As this algorithm determines
7
the best suited standards, out of only ten candidates, we have excluded the 18S gene due to its worst ranking with the previous tools. The BestKeeper analysis has also revealed some similarity with the geNorm and NormFinder analysis. In fact, GAPDH has the lowest SD [± Ct] value and a good correlation coefficient with the remaining analysed genes (Table 1). Although B2m and GUS genes present SD [± Ct] values lower than the β-actin gene, those genes have a very low correlation coefficient β-actin is, therefore, the second best suited reference gene, according to the BestKeeper analysis. In accordance with to the obtained results, we have chosen GAPDH and β-actin as the best endogenous control genes to normalize our gene expression study.
Expression levels of target genes normalized with GAPDH and/ or β-actin genes The expression of ten genes related with the immune response in bladder cancer was analysed in 30 patients. The study was encompassed by genes coding for the CD1A, CD1B, CD1C, CD1D and CD1E molecules, which are important molecules for presenting lipid antigens [17]; and also genes coding for chemokines, MCP-1 and MCP2, also known as monocyte chemoattractant proteins [18], MIP-1α, the macrophage inflammatory protein-1-alpha [19], MIG, the monokine induced by gamma-Interferon and IP-10, the Interferon-inducible protein-10 [20]. The gene expression profiles in cancerous bladder tissue relative to normal urothelium were compared using either the β-actin or the GAPDH gene for normalization and the results revealed a significantly different variance (p < 0,001) for all the analysed genes (Fig. 1). When using the geometric mean of the expression of β-actin and GAPDH to normalize the data, the overall profile was also significantly different (p < 0,05) from the expression profiles obtained using β-actin or GAPDH genes individually, in all genes except for the MCP-1 and MCP-2 genes (Fig. 1).
8
Still, it is interesting to note that, whereas the cytokine genes present a divergent profile, all CD1 molecules are downregulated in a great majority of the cancerous bladder tissue relative to normal urothelium (Fig. 1). These results are probably related with fact that tumourigenesis usually induces the downregulation of antigen presenting molecules in the tumour tissue. This downregulation is known to have an important contribution to the tumour escape to the immune surveillance [21].
Expression variation of the GAPDH and/ or β-actin reference genes Given the differences found for the expression of target genes, when normalized to βactin, GAPDH or both, we have examined the expression variances of these reference genes in the 30 samples. When comparing matched pair samples, we have observed that both genes amplified in a distinct way. In urothelium, both genes have an interval of amplification curve Ct, i.e., the differences between β-actin and GAPDH Ct (Ctβ-actin CtGAPDH) with a median of -0,08, ranging from -3,02 to 1,86 (Fig. 2A). On the other hand, in cancerous bladder tissue, the Ctβ-actin - CtGAPDH have a median of 1,01, ranging from -1,85 to 4,96 (Fig. 2A). Comparing these differences in matched pairs of cancerous and normal patient’s samples, the Ctβ-actin - CtGAPDH in cancerous tissue differed with a median of 1,02 Cts, ranging from -1,98 to 4,72, from the Ctβ-actin CtGAPDH of normal urothelium (Fig. 2B). In fact, 12 out of the 30 samples presented Ctβactin
- CtGAPDH differences between cancerous and normal urothelium higher than one Ct
(Fig. 2B). Among them, the samples 2, 4, 5, 7, 11, 20, 28 and 30, presented Ctβ-actin CtGAPDH differences higher than two Cts (Fig. 2B). Consequently, these samples gave rise to more suspicious results regarding the relative expression of the analyzed genes, obtained after normalizing to β-actin, GAPDH or both (Fig. 1). As an example, in patient 2, the analysis of the CD1A expression, normalized to β-actin gave a two fold
9
increased expression in tumour relative to urothelium. Yet, if the data is normalized with GAPDH, the expression drops to 0,08, which mean 12,5 fold less expression in tumour relative to urothelium (Fig. 1). When considering the geometric mean, the obtained results are 0,38, which mean 2,6 less expression in tumour relative to urothelium. Disregarding the values and in view of the qualitative expression; in other words, considering only, whether the expression is increased or decreased in cancer relative to urothelium, the obtained results are more consensual. As an example, in patient 20, the data of the expression of CD1B, CD1C, CD1D, CD1E, MIP-1α, MCP-1 and MCP-2 genes was always decreased in cancer relative to urothelium, independently of the choice of the endogenous control (Fig. 2).
Discussion The use of suitable endogenous control gene for cell or tissue gene expression measurements is still a matter of debate. The prevalent opinion is to previously test several control genes and select the most constant and the gold standard normalization is now considered to measure the expression of several endogenous control genes and normalize using their mean expression [10]. For the study here presented, we have selected the β-actin and GAPDH genes, based essentially on the results obtained from a test using a panel of endogenous controls. Even so, these genes present significant expression variability in the analysed samples and in, some measurements, the obtained results were contradictory. Our observations have showed two aspects. 1) Caution should always be taken even when using endogenous control genes selected as suitable reference genes from a panel of common housekeeping genes. 2) When using one or two combined controls, the relative
10
expression should be regarded as qualitative rather than quantitative. 3) The most reliable normalization is obtained with the mean expression of more than one endogenous control. Even though, the use of only two control genes may be insufficient and the use of multiple housekeeping genes should lead to more reliable results. The selection of suitable endogenous control genes are costly and time consuming procedures. Therefore, many researchers easily assume, as endogenous control gene, a gene already referred by others. As the β-actin and GAPDH have been the most frequently used genes for normalizing expression measurements, many authors have focused on the study of their reliability as reference genes. As a result, there are nowadays an increasing number of publications reporting the pitfalls of using them as endogenous control [16, 22]. Additionally other genes have been proposed as more accurate housekeeping genes for analysing cancer samples. As an example, in general bladder cancer, two important works report the most suitable reference genes,[10, 11]. Since these works are based on distinct panels of reference genes, each indicate their own most suitable reference gene. So far, the reference genes reported by these authors were not analysed by others in bladder cancer. It is likely that a more profound research reveals that those genes, described as invariable in some studies, are found to be regulated in other studies, even with the same type of analysed tissue. It is reasonable to consider that there is no gold standard for analysing the expression in cancer. Therefore the use of multiple endogenous genes is surely the most accurate approach for gene expression measurements in cancer.
11
References
1. Bustin SA. Quantification of mRNA using real-time reverse transcription PCR (RTPCR): Trends and problems. J Mol Endocrinol 2002; 29:23-39. 2. Bernard PS, Wittwer CT.Real-time PCR technology for cancer diagnostics. Clin Chem 2002; 48:1178-85. 3. Bustin SA. Absolute quantification of mRNA using real-time reverse transcription polymerase chain reaction assays. J Mol Endocrinol 2000; 25:169-93. 4. Huggett J, Dheda K, Bustin S, Zumla A.Real-time RT-PCR normalisation; strategies and considerations. Genes Immun 2005; 6:279-84. 5. Mansur NR, Meyer-Siegler K, Wurzer JC, Sirover MA.Cell cycle regulation of the glyceraldehyde-3-phosphate dehydrogenase/uracil DNA glycosylase gene in normal human cells. Nucleic Acids Res 1993; 21:993-8. 6. Zhong H, Simons JW.Direct comparison of GAPDH, beta-actin, cyclophilin, and 28S rRNA as internal standards for quantifying RNA levels under hypoxia. Biochem Biophys Res Commun 1999; 259:523-6. 7. Tricarico C, Pinzani P, Bianchi S, Paglierani M, Distante V, Pazzagli M, Bustin SA, Orlando C.Quantitative real-time reverse transcription polymerase chain reaction: Normalization to rRNA or single housekeeping genes is inappropriate for human tissue biopsies. Anal Biochem 2002; 309:293-300. 8. Goidin D, Mamessier A, Staquet MJ, Schmitt D, Berthier-Vergnes O.Ribosomal 18S RNA prevails over glyceraldehyde-3-phosphate dehydrogenase and beta-actin genes as internal standard for quantitative comparison of mRNA levels in invasive and noninvasive human melanoma cell subpopulations. Anal Biochem 2001; 295:17-21.
12
9. Livak KJ, Schmittgen TD.Analysis of relative gene expression data using real-time quantitative PCR and the 2(-delta delta C(T)) method. Methods 2001; 25:402-8. 10. Vandesompele J, De Preter K, Pattyn F, Poppe B, Van Roy N, De Paepe A, Speleman F.Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol 2002; 3:RESEARCH0034. 11. Andersen CL, Jensen JL, Orntoft TF.Normalization of real-time quantitative reverse transcription-PCR data: A model-based variance estimation approach to identify genes suited for normalization, applied to bladder and colon cancer data sets. Cancer Res 2004; 64:5245-50. 12. Pfaffl MW, Tichopad A, Prgomet C, Neuvians TP.Determination of stable housekeeping genes, differentially regulated target genes and sample integrity: BestKeeper--excel-based tool using pair-wise correlations. Biotechnol Lett 2004; 26:509-15. 13. Urist MJ, Di Como CJ, Lu ML, et al.Loss of p63 expression is associated with tumor progression in bladder cancer. Am J Pathol 2002; 161:1199-206. 14. Nomura T, Nakagawa M, Fujita Y, Hanada T, Mimata H, Nomura Y.Clinical significance of thymidylate synthase expression in bladder cancer. Int J Urol 2002; 9:368-76. 15. Gazzaniga P, Gradilone A, Giuliani L, et al.Expression and prognostic significance of LIVIN, SURVIVIN and other apoptosis-related genes in the progression of superficial bladder cancer. Ann Oncol 2003; 14:85-90. 16. Ohl F, Jung M, Radonic A, Sachs M, Loening SA, Jung K.Identification and validation of suitable endogenous reference genes for gene expression studies of human bladder cancer. J Urol 2006; 175:1915-20.
13
17. Roura-Mir C, Moody DB.Sorting out self and microbial lipid antigens for CD1. Microbes Infect 2003; 5:1137-48. 18. Bischoff SC, Krieger M, Brunner T, Dahinden CA.Monocyte chemotactic protein 1 is a potent activator of human basophils. J Exp Med 1992; 175:1271-5. 19. Guan E, Wang J, Norcross MA.Identification of human macrophage inflammatory proteins 1alpha and 1beta as a native secreted heterodimer. J Biol Chem 2001; 276:12404-9. 20. Gasperini S, Marchi M, Calzetti F, et al.Gene expression and production of the monokine induced by IFN-gamma (MIG), IFN-inducible T cell alpha chemoattractant (I-TAC), and IFN-gamma-inducible protein-10 (IP-10) chemokines by human neutrophils. J Immunol 1999; 162:4928-37. 21. Marincola FM, Jaffee EM, Hicklin DJ, Ferrone S.Escape of human solid tumors from T-cell recognition: Molecular mechanisms and functional significance. Adv Immunol 2000; 74:181-273. 22. Bustin SA, Benes V, Nolan T, Pfaffl MW.Quantitative real-time RT-PCR--a perspective. J Mol Endocrinol 2005; 34:597-601.
14
Legends to figures
Figure 1. Relative quantification (RQ) of the expression of genes associated with the immune response in 30 patients with superficial bladder cancer. RQ values correspond to the expression of a defined gene in cancerous tissues relative to normal urothelium. Samples were normalized to the geometric mean of β-actin and GAPDH expression (black columns), to the β-actin (grey columns) or to the GAPDH gene (white columns). Values above or below 1 indicate, respectively, overexpression or downregulation of a specific gene in the tumour tissue relative to the healthy tissue.
Figure 2. Differential expression of the β-actin and GAPDH in matched samples of normal urothelium and bladder cancer tissue. A: Differences between β-actin and GAPDH Ct (Ctβ-actin - CtGAPDH) either in urothelium or in tumour tissue. B: Differences between the Ctβ-actin - CtGAPDH of tumour tissue and the Ctβ-actin - CtGAPDH of the matched urothelium of each patient (B).
15
Figure 1 1000,00
100,00
CD1B
CD1A 100,00
10,00
10,00
3
4
5 6
7
8
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
RQ
1 2
RQ
1,00 1,00
1
2
3
4 5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
0,10 0,10
0,01 0,01
0,00
0,00
1000,00 100,00
CD1C CD1D 100,00 10,00
10,00 1,00
4 5
6 7
8
RQ
RQ
1 2 3 1,00
1 2 3
4
5
6
7 8
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
6
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 0,10
0,10
0,01
0,01
0,00
0,00
1000,00
100,00
CD1E
IP-10
100,00
10,00
10,00
1,00
1 2
3 4
5
6 7
8
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
RQ
RQ
1 2 3 1,00
0,10 0,10
0,01 0,01
0,00
0,00
4 5
6 7
8
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Figure 1 (continue)
1000,00
1000,00
MCP-2
MCP-1 100,00
100,00
10,00
RQ
RQ
10,00
1,00
1 2
1,00
3
4
5
6
7
8
5
6
7 8
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
0,10
0,10
0,01
0,01
1000,00
100,00
MIG
MIP1-A
100,00 10,00
RQ
10,00
1,00
1,00
1 2 3 4 5 6 7 8
1 2
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
0,10 0,10
0,01
0,01
0,00
3 4
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Figure 2 A
B
5
6
4
5 Ctβ-actin - CtGAPDH differences
Ct ß-actin - CtGAPDH
3 2 1 0 -1 -2 -3
4 3 2 1 0 1
Urothelium
Tumour
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
-1 -2
-4
2
-3
Matched patient samples
Table 1. Candidate endogenous control genes for normalization and their expression stability values calculated by geNorm, NormFinder and BestKeeper softwares. Gene
geNorm*
NormFinder
18S
0,735
0,488
B2m
0,441
0,252
GUS
0,399
0,211
CYC
0,383
0,176
PGK
0,353
0,150
GAPDH
0,306
0,073
HPRT
0,334
0,131
β-actin
0,307
0,092
ARP
0,390
0,191
TBP
0,337
0,139
TfR
0,328
0,103
BestKeeper**
0,68 0,180 0,72 0,230 1,53 0,975 0,96 0,826 0,54 0,808 0,88 0,896 0,73 0,810 1,10 0,947 0,89 0,751 1,02 0,834
A low stability value corresponds to a high expression stability of the respective gene, therefore to the best endogenous control genes for normalization. * In GeNorm, the stability value is designated as M value. ** Measures of the SD [± Ct], standard deviation of the Ct and correlation coefficients (bellow values) obtained with the Bestkeeper analysis.
1
