Toxicological Housekeeping Genes: Do They Really Keep the House

It is assumed that the expression of housekeeping genes is constant regardless of experimental conditions. In toxicology, this assumption has indeed b...
0 downloads 9 Views 123KB Size
Download PDF
Environ. Sci. Technol. 2006, 40, 7944-7949

Toxicological Housekeeping Genes: Do They Really Keep the House? AUGUSTINE ARUKWE* Department of Biology, Norwegian University of Science and Technology (NTNU), Høgskoleringen 5, N-7491 Trondheim, Norway

It is assumed that the expression of housekeeping genes is constant regardless of experimental conditions. In toxicology, this assumption has indeed become a misconception of reasonable concern, as these so-called housekeeping genes vary considerably across different experimental conditions and thereby lead to an erroneous interpretation of the expression profile of a target gene. Given that the choice of reference gene will ultimately influence statistical interpretation of toxicological data, it is essential to validate potential reference genes prior to their use, to establish their suitability for a specific experimental purpose. Therefore, the aim of this study is to quantitatively evaluate the most commonly used housekeeping genes in toxicology research for their suitability as reference endpoints, and thus provide toxicology researchers who have little experience in molecular biology but find themselves interested or involved with gene expression analysis with a summary of information necessary for re-evaluating their procedures. We show that the expression pattern of β-actin, β-tubulin, 18S ribosomal RNA (18S rRNA), and elongation factor-1R (EF-1R), representing commonly used housekeeping genes in toxicology, was modulated on the basis of random exposure condition and time, in both in vivo and in vitro test systems of Atlantic salmon (Salmo salar). Based on the data presented herein and several other reports by other researchers, there are very few biological justifications to refer to anything as a housekeeping gene in real-time PCR assays for toxicological research. However, given the absolute need for normalization genes to correct for sampleto-sample variations, the choice of internal control gene should be determined empirically on the basis of the individual exposure condition and by the individual researcher.

(becasue it is only in this phase that amplification is extremely reproducible) (4). Housekeeping gene is defined as a gene involved in basic functions needed for the sustenance of the cell. Thus, housekeeping genes are defined by specific gene promoter elements and are expressed constitutively in every cell and used in a normal cell to maintain basic cellular functions. Yet, it does not necessarily mean that the expressions of housekeeping genes are not regulated in cells. Therefore, the use of housekeeping genes in molecular toxicology relies on the assumption that their levels of expression remain the same from cell to cell, sample to sample, and exposure to exposure, a situation that has so far proven difficult to demonstrate in toxicology (2, 3). As a consequence, the use of a housekeeping gene for relative or absolute mRNA quantification should generally fulfill the following criteria: (i) it should exhibit constitutive and nonregulated expression in the sample types investigated, (ii) the detection should be RNA-specific where a pseudogene- and DNA-free amplification should be realized by stringent primer design (5), and (iii) it should be expressed in a range similar to the target gene in the samples to be analyzed (2, 4, 6). Housekeeping genes are often used as reference nucleic acids and as controls for variables that may affect real-time PCR reactions. Because of their assumed constant expression, some researchers use housekeeping genes to normalize gene expression in several array (gene chip) experiments to monitor the relative changes of gene expression between control and test samples. In toxicology, housekeeping genes are ubiquitously used, and only on very few situations have these genes been put into test to ascertain whether they change between exposure conditions. Researchers have ignored the obvious question asked in the present evaluation, whether these so-called housekeeping genes really keep the house of toxicology. Given that the choice of reference gene will ultimately influence statistical interpretation of toxicology, it is essential to validate potential reference genes to establish their suitability for a specific experimental purpose. Therefore, the aim of this study is to quantitatively evaluate the most commonly used housekeeping genes in toxicology research for their suitability as reference endpoints, and thus provide toxicology researchers who have little experience in molecular biology but find themselves interested or involved with gene expression analysis with a summary of information necessary for re-evaluating their procedures. In this regard, we evaluated the expression patterns of the four most commonly used (β-actin, β-tubulin, 18S ribosomal RNA (18S rRNA), and elongation factor-1R (EF-1R)) housekeeping genes in toxicology in both in vivo and in vitro test systems.

Materials and Methods Introduction The analysis of gene expression patterns using quantitative (real-time) reverse-transcriptase polymerase chain reaction (qRT-PCR) has become an essential requirement in determining the amount of gene expression in toxicological and biological samples (1). Real-time PCR assays used for quantitative RT-PCR combine the best attributes of both relative and competitive (end-point) RT-PCR in that they are accurate, precise, capable of high throughput, and relatively easy to perform (2, 3). For the sake of accuracy and precision, it is necessary to collect quantitative data at a point when every sample is in the exponential phase of amplification * Phone: +47 73 596265; fax: +47 73 596100; e-mail: arukwe@ bio.ntnu.no. 7944

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 24, 2006

Chemicals and Reagents. Thyroxine (T4), 1,1-dichloro-2,2bis(4-chlorophnyl)ethane (p,p′-DDE or DDE), DNA polymerase I-Klenow fragment, random hexamer primers, dimethyl sulfoxide (DMSO), 100× penicillin-streptomycinneomycin solution, collagenase, bovine serum albumin (BSA), N-[2-hydroxyethyl]piperazine-N′-[2-ethanesulfonic acid] (HEPES), ethyleneglycol-bis-(β-aminoethylether)-N,N′-tetraacetic acid (EGTA), and 0.4% trypan blue were purchased from Sigma Aldrich (St. Louis, MO). 4-Nonylphenol (NP; 85% of p-isomers) was purchased from Fluka Chemika-Biochemika (Buchs, Switzerland). 3,3′,4,4′-Tetrachlorobiphenyl (TCB or PCB-77; 99.7% pure) was purchased from Dr. Ehrenstorfer GmbH (Augsburg, Germany). iScript cDNA synthesis kit and iTAQSYBR Green supermix with ROX were purchased from Bio-Rad Laboratories (Hercules, CA), GeneR10.1021/es0615223 CCC: $33.50

 2006 American Chemical Society Published on Web 11/15/2006

uler 100-bp DNA ladder, and dNTPs from Fermentas (St. Leon-Rot, Germany). Trizol reagent for RNA purification and TA cloning kit, dubelco minimum essential medium (DMEM) with non-essential amino acid and without phenol red, fetal bovine serum (FBS), and L-glutamine were purchased from Gibco-Invitrogen Life Technologies (Carlsbad, CA). Fish and Treatment. Juvenile immature Atlantic salmon (mean weight and length ( standard deviation: 10 ( 2 g and 9 ( 1.5 cm of n ) 6, respectively) were obtained from Stjørdal hatcheries (Meråker, Norway) and kept 500 L aquaria at 10 ( 0.5 °C and 12:12 h photoperiod at the Department of Biology, Norwegian University of Science and Technology (Trondheim, Norway) animal holding facilities in Trondheim. Fish were starved during the experiment. The fish were divided into four groups in separate 70 L tanks and exposed once to waterborne doses of DDE (10 µg/L) and T4 (50 µg/L) under static condition, singly and in combination using dimethyl sulfoxide (DMSO) as carrier (70 µL or 0.001% of total tank volume). The control group was exposed to the DMSO. The water was not changed during the treatment; after 5 days the fish were anesthetized with tricaine methane sulfonate (MS222; 0.5 mg/mL), and brain, kidney, and liver tissues were removed and homogenized in Trizol (Invitrogen). Collagenase Perfusion and Isolation of Hepatocyte. Juvenile and immature Atlantic salmon (Salmo salar) of approximately 400-500 g were kept at the animal holding facilities at the Biology Department, NTNU. Fish were supplied with continuously running saltwater at a constant temperature of 10 °C. Prior to liver perfusion, all glassware and instruments were autoclaved. Solutions were filtration sterilized by using 0.22 µm Millipore filter (Millipore AS, Oslo, Norway). Hepatocytes were isolated by a two-step perfusion as described by Berry and Friend (7) and modified by Andersson et al. (8, 9). Thereafter, cells were washed three times with serum-free medium and finally resuspended in complete medium. Hepatocyte Culture and Exposure. After collagenase perfusion and isolation of hepatocytes, viability of cells was determined by the trypan blue exclusion method. A cell viability value of >90% was a criterion for further use of the cells. Cells were plated in triplicates on a 35 mm Primaria culture plates (Becton Dickinson Labware, USA) at the recommended density for monolayer cells of 5 × 106 cells in 3 mL of DMEM medium (without phenol red) containing 2.5% (v/v) FBS, 0.3 g/L glutamine, and 1% (v/v) antibiotic/ antimycotic. Hepatocytes were cultured at 10 °C (fish holding temperature) in a sterile incubator without O2/CO2 for 48 h prior to chemical exposure. Cells were exposed in triplicates to the test chemicals, DMSO (control) at 0.1% or 10 µL/L medium, NP (5 µM), TCB (1 µM), or combined NP and TCB at respective concentrations. The test chemicals were dissolved in DMSO and added to the cell culture after 48 h of pre-culture at the respective concentrations. The final concentration of DMSO in the sample never exceeded 0.1% (v/v) because DMSO may exert a cytotoxic effect at higher concentration. Media were collected and replaced after 24 h of exposure. The experiment was performed to test the effect of compounds and time. For the time study, cells were harvested at 12, 24, 48, and 72 h after exposure, while cells for the compound-dependent effect were harvested at 48 h postexposure for total RNA isolation. Blank sample that was incubated without the carrier solvent (DMSO) was harvested at 48 h after exposure. RNA Purification and cDNA Synthesis. Total RNA was isolated using the Trizol reagent according to manufacturer’s protocol, and RNA concentrations were determined using a NanoDrop ND-1000 UV-vis spectrophotometer (NanoDrop Technologies, Wilmington, DE). Total cDNA for the realtime PCR reactions was generated from 1 µg of DNase treated total RNA from all samples using poly-T primers from the

iScript cDNA Synthesis Kit as described by the manufacturer (Bio-Rad). Primer Optimization, Cloning, and Sequencing. PCR primers (given in 5′-3′ direction) for amplification of β-actin; forward-CCCCCTCAACCCCAAAGCCA and reverse-CGAGACATCAAGGAGAAGCT (324 bp), β-tubulin; forwardATGGGGACAGTGACCTTCAG and reverse-GAGCTGGAAACCTTGCAGAC (296 bp), EF-1R; forward-GATCCAGAAGGAGGTCACCA and reverse- TTACGTTCGACCTTCCATCC (250 bp), 18S-rRNA; and forward- CGGACACGGAAAGGATTG and reverse-AGCCCCGGACATCTAAGG (243 bp) PCR-products were designed from conserved regions of the studied genes using Primer3 software http://www.bioinformatics.vg/index.shtml. Prior to all real-time PCR reactions, all primer pairs were used in titration reactions to determine optimal primer pair concentrations and their optimal annealing temperatures, and real-time PCR was run using reverse transcriptase reactions without enzyme. All chosen primer pair concentrations used at the selected annealing temperatures gave a single band pattern for the expected amplicon size in all reactions (9). Quantitative (Real-Time) PCR. Quantitative (real-time) PCR was used for evaluating gene expression profiles. For each treatment, the expression of individual gene targets was analyzed using the Mx3000P REAL-TIME PCR SYSTEM (Stratagene, La Jolla, CA). Each 25 µL DNA amplification reaction contained 12.5 µL of iTAQSYBR Green Supermix with ROX (Bio-Rad), 1 µL of cDNA, and 200 nM of each forward and reverse primers. The three-step real-time PCR program included an enzyme activation step at 95 °C (5 min) and 40 cycles of 95 °C (30 s), 60 °C (30 s), and 72 °C (30 s). After amplification, a melting curve was generated using the following program: 95 °C (30 s), 55 °C (30 s), and 95 °C (30 s) to confirm the specificity of each amplification. Controls lacking cDNA template were included to determine the specificity of target cDNA amplification. Cycle threshold (Ct) values obtained were converted into mRNA copy number using standard plots of Ct versus log copy number. The criterion for using the standard curve is based on equal amplification efficiency with unknown samples, and this is usually checked prior to extrapolating unknown samples to the standard curve. The standard plots were generated for each target sequence using known amounts of plasmid containing the amplicon of interest. Data obtained from triplicate runs for target cDNA amplification were averaged and expressed as ng/µg of initial total RNA used for reverse transcriptase (cDNA) reaction. Statistical Analyses. Statistical analysis was performed with GraphPad Prism, version 2.1 (GraphPad Software). Significant differences between control and exposure groups were performed using one-way analysis of variance (ANOVA) followed by the Tukey Kramer multiple comparison test, after checking for data normality and homogeneity of variances. The level of significance was set at p ) 0.05 unless otherwise stated.

Results Optimization of Real-Time PCR Amplification. Real-time PCR assay was optimized using different concentrations of forward and reverse in generating a titration curve for best primer pair concentration and different annealing temperatures. We used the SYBR Green method for absolute quantification of housekeeping gene expression patterns. A typical SYBR green experiment has two important phases, the amplification phase, which corresponds to the PCR portion that generates double-stranded DNA (dsDNA), and dissociation phase that represents the melting of the dsDNA to single-stranded DNA (ssDNA) through a stepwise increase in temperature. Changes in fluorescence data are recorded at each temperature step. Therefore, the magnitude of the VOL. 40, NO. 24, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7945

FIGURE 1. Expression of β-tubulin (A), elongation factor-1 (EF-1r: B), and β-actin mRNA in juvenile Atlantic salmon exposed to carrier solvent control (DMSO), thyroxin (T4), and DDE, singly and in combination. Messenger RNA (mRNA) expression levels were quantified using real-time PCR with specific primer pairs. Data are given as means ( standard error of mean (SEM of n ) 6). Different letters denote exposure group means that are significantly different, using ANOVA (p < 0.05). reduction in fluorescence intensity of the SYBR green dye as a result of its release from the dsDNA provides an indicator of the amount of dsDNA dissociated at each step of the dissociation curve. Thus, a single fluorescence peak will subsequently represent single product amplification. In Vivo Expression of Housekeeping Genes. The in vivo and in vitro exposure conditions presented in this study represent random experiments using chemicals that are under systematic investigations in our laboratory using fish and frog as model species (9-12). Exposure to T4 and DDE, singly and in combination, caused slight, but nonsignificant elevation of liver β-tubulin mRNA levels (Figure 1A). Exposure of juvenile salmon to T4 did not affect transcriptional change in liver EF-1R mRNA levels, as compared to the control (Figure 1B). DDE exposure singly and in combination with T4 produced significant elevation of liver EF-1R mRNA expression (Figure 1B). For β-actin, T4 induced transcriptional changes of mRNA levels, as compared to the control (Figure 1C). When compared to the control group, DDE exposure alone did not affect β-actin mRNA expression, while combination exposure with T4 resulted in significant liver β-actin mRNA expression, as compared to the control (Figure 1C). The 18S rRNA reached maximum amplification at 5 PCR 7946

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 24, 2006

cycles, and this is also true after 10 000 times dilution of the cDNA; the data were therefore not included here, but will be discussed below. In Vitro Expression of β-Actin. The variations of β-actin were studied as a result of both chemical exposure and time of exposure. Exposure of primary culture of salmon hepatocytes to carrier solvent (DMSO or experimental control), NP and TCB, singly and also in combination produced an exposure-specific pattern of β-actin (Figure 2). Hepatocytes harvested after the corresponding DMSO time-interval in Figure 2A were used as the control in β-actin gene expression for the respective chemical exposures. For DMSO, a 2.8-, 2.5-, and 3-fold increase was observed at 24, 48, and 72 h postexposure, respectively (Figure 2A). Exposure of hepatocytes to NP produced a 1.5-fold decrease at 24 h postexposure, and, thereafter, while a 2.5-fold was observed at 48 h, no change in β-actin expression was observed at 72 h after exposure to NP (Figure 2B). For TCB, a 4.3-fold increase of β-actin at 24 h was observed, then no change at 48 h, and thereafter a 6-fold increase at 72 h after exposure (Figure 2C). When hepatocytes were exposed to NP and TCB in combination for 24 h, a 4.5-fold increase in β-actin mRNA was observed (Figure 2C). While no change was observed at

FIGURE 2. Time and exposure transcriptional changes in β-actin mRNA levels in primary culture of Atlantic salmon hepatocytes exposed to carrier solvent (DMSO: A), nonylphenol (NP: B), 3,3′,4,4′-tetrachlorobiphenyl (TCB: C), and combined NP and TCB (D) sampled at 12, 24, 48, and 72 h after exposure. The abundance levels of β-actin transcripts also changed depending on time and chemical. Messenger RNA (mRNA) expression levels were quantified using real-time PCR with specific primer pairs. Data are given as means ( standard error of mean (SEM of n ) 3). Different letters denote exposure group means that are significantly different, using ANOVA (p < 0.05). Note that the control in (A) is blank sample (minus DMSO), while the control in (B-D) is DMSO exposed at the corresponding time interval. 48 h, combined NP and TCB exposure produced a 2-fold increase of β-actin at 72 h postexposure, as compared to 12 h (Figure 2D).

Discussion The accuracy of real-time PCR experiments is highly dependent on PCR efficiency. A reasonable efficiency should be at least 80%. Poor primer quality is the leading cause for poor PCR efficiency. In this case, the PCR amplification curve usually reaches a plateau early, and the final fluorescence intensity is significantly lower than that of most other PCRs. Our real-time PCR is based on the validated absolute quantification method (13). Absolute quantification uses serial dilution of standards with known concentrations to generate a standard curve. The standard curve produces a linear relationship between Ct values and initial amounts of total RNA or cDNA, allowing the determination of the concentration of unknown samples on the basis of their Ct values (14). We generated a standard curve for absolute quantification using a cloned plasmid containing the amplicon of interest for the respective genes. In calculating the amount of DNA in the standard curve, only the concentration of the insert was used. A critical step in using the standard curve is the equal amplification efficiency between known (plasmid with amplicon or insert) and unknown (experimental) samples, and a single product peak in the standard and unknown samples. In our real-time experiments, an amplification efficiency should be >90% to be classified as acceptable. Modulation of in Vivo and in Vitro Housekeeping Genes. Evaluation of β-tubulin, EF-1R, β-actin, and 18S rRNA showed that these normally used housekeeping genes in toxicological research were affected after exposure of salmon to T4 and DDE, singly and also in combination. Similar effects were observed in vitro, where exposure of primary salmon

hepatocytes culture to NP and TCB, singly and in combination, produced exposure- and time-specific expression patterns of β-actin gene. Two housekeeping genes, β-tubulin and EF-1R, showed the least variation in the in vivo exposure condition. Despite the low variations of these genes, their expression patterns were exposure-dependent. In eukaryotes, the elongation factor 1 plays an integral role in translation by catalyzing guanosine triphosphate (GTP)-dependent binding of aminoacyl-transfer RNA (tRNA) to the acceptor site at the ribosome. Generally, the protein is involved in a broad diversity of cellular functions and represents an abundance of about 1-3% of total cytoplasmic protein content of the cell. In different tissues of salmon, EF-1A and EF-1B were ranked as the most stable housekeeping genes (15). Microtubules are involved in several basic cellular processes such as segregation of genetic material, intracellular transport, maintenance of cell shape, positioning of cell organelles, extracellular transport by means of cilia, and movement of cells by means of flagella and cilia (16). Tubulins (R and β subunits) are heterodimeric globular protein monomers that are a structural component of a microtubule, where each heterodimer binds two molecules of GTP nucleotide (17). There are conflicting data on the suitability of elongation factor and β-tubulin as reference genes in realPCR assays. While some studies observed condition-dependent variations of these genes (18-20), others have used and proposed these genes as suitable normalization references in real-time PCR assays (21-23). As a major component of the protein moiety encoding the ubiquitous cytoskeleton protein that provides cellular support and determines shape, β-actin is the most abundant intracellular protein in eukaryotic cells (24). It is the most commonly used and indeed the first gene to be used as an internal standard (5). Despite the fact that several recent VOL. 40, NO. 24, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7947

reports have criticized the use of β-actin as a housekeeping gene, it is still advocated as a quantitative reference for toxicological real-time PCR assays (23, 25-29). In our laboratory, we have consistently observed that the levels of β-actin transcription always vary depending on experimental manipulation. In the data presented in this report, changes in β-actin gene expression were dependent on time and chemical. For example, the abundance of β-actin transcripts is TCB > NP + TCB > DMSO > NP in the time-dependent experiment. This is in accordance with other findings reporting the instability of β-actin in human-based experimental samples (30, 31). Similar effects have been reported in various porcine tissues (32), canine myocardium (33), and multiple tissues of Atlantic salmon, Salmo salar (15). Furthermore, pseudogenes may interfere with the interpretation of results and primers commonly used in the realtime PCR detection (34, 35). In a study using specific oligonucleotide primers previously designed to amplify both genomic DNA and mRNA transcript from paraffin-embedded pathological tissues, Dakhama et al. (35) found that these primers amplified β-actin products both with and without reverse transcriptase reactions. Further investigation of the nature of this product by these authors showed that the PCR product originated from the amplification of DNA, not RNA. The authors therefore conclude that the amplified 154-bp β-actin product generated by these primers cannot be exclusively considered as β-actin RNA product and should not be used to assess successful extraction of RNA, to ascertain its integrity, or to normalize for the total amount of RNA assayed by RT-PCR from paraffin-embedded tissue (35). In another study by Malarstig et al. (36), the effect of lipopolysaccharide (LPS) treatment on the expression of the housekeeping genes, β-actin and β-2-microglobulin (β-2MG), in seven separate leukocyte isolations showed that β-actin expression was modulated by LPS treatment. Elsewhere, the activation of LPS-induced monocytes involves both β- and γ-actin, which implies that β-actin is less suitable as endogenous control in experimental systems involving LPS (37). Several toxicological studies have used 18S rRNA as a normalization gene (15, 28, 38, 39). In other studies, Eleaume et al. (40) evaluated the absolute quantification of three transcripts of interest and showed that they gave results similar to their relative quantification expressed versus 16S rRNA. The authors therefore concluded the suitable use of 16S rRNA as internal standard in real-time PCR quantification of staphylococcal gene expression during the stationary phase of growth (40). In our studies, 18S rRNA usually reaches maximum amplification between 5 and 7 PCR cycles, even after 10.000 dilution of the cDNA sample. Ribosomal RNA represents up to 80-90% of total cellular RNA. Several studies have shown that rRNAs show less variation under experimental conditions that affect mRNA expression (reviewed by ref 2). As an RNA species, 18S rRNA has all necessary biological predispositions to change with stress and particularly chemical stress (20, 21). Furthermore, the mere dilution of the 18S rRNA cDNA prior to amplification undermines the whole concept of normalization. Implications for the Interpretation of Toxicological Data. Stress from contaminant exposure at various levels of biological organization is generally a complex process, which can involve numerous related and unrelated biochemical processes. Regardless of the method used, normalization with an internal control gene, usually a purported housekeeping gene, is desirable to correct the sample-to-sample variations that may arise in the process of mRNA quantification experiments (41). Usually, it is assumed that the expression of housekeeping genes is constant regardless of experimental conditions. In toxicology and as demonstrated in the present study, this assumption has indeed become a misconception 7948

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 24, 2006

of reasonable concern, as these so-called housekeeping genes vary considerably across different experimental conditions and thereby lead to dramatic misinterpretation of the expression profile of a target gene (42, 43). It should be noted that the data presented here are representative of the total observations regarding exposure-dependent variations in the typical toxicological housekeeping genes. To obtain reliable results with biological significance, it is important that realtime PCR data are normalized with a proper internal control. In this regard, prudent reference gene selection is important in evaluating relative gene expression in toxicological samples (44). Based on the data presented in this report and elsewhere, there are very little biological justifications to refer to anything as housekeeping genes in real-time PCR assays for toxicological research. It has been argued that overall study conclusions will remain unchanged despite the variability of reference gene used, because this variability will be similar to controls (45). In a recent study to address the impact of validation with reference genes, including conventional ones, on overall study conclusions, Dheda et al. (1) found that the experimental result outcome was highly dependent on the reference gene chosen. Above all, these authors observed that the presence or absence of statistical differences between exposure groups depended on which reference gene was used in the normalization (1). Therefore, generalization of the housekeeping gene concept in toxicology should be avoided. However, given the absolute need for normalization genes to correct for sample-to-sample variations, the choice of internal control gene should be determined empirically on the basis of the individual exposure condition. This issue of empirical determination of internal control gene could easily result in expensive experimental setup through the evaluation of several genes for their suitability as internal controls. In this regard, absolute quantification using the standard curve generated using cloned plasmid containing the amplicon of interest has proven to be a reliable method in our laboratory. Yet, this method is susceptible to error caused by inadequate quantification of initial RNA samples used for reverse transcriptase reactions and the presence of genomic DNA. These pitfalls could easily be minimized using a very sensitive RNA quantification system and by DNase treatment of RNA samples.

Acknowledgments The Norwegian Research Council (NFR) supported this study financially. The technical assistance of Anne Skjetne Mortensen during exposure and analysis is gratefully appreciated.

Literature Cited (1) Dheda, K.; Huggett, J. F.; Chang, J. S.; Kim, L. U.; Bustin, S. A.; Johnson, M. A.; Rook, G. A.; Zumla, A. The implications of using an inappropriate reference gene for real-time reverse transcription PCR data normalization. Anal. Biochem. 2005, 344, 141143. (2) Bustin, S. A.; Nolan, T. Pitfalls of quantitative real-time reversetranscription polymerase chain reaction. J. Biomol. Technol. 2004, 15, 155-166. (3) Bustin, S. A.; Benes, V.; Nolan, T.; Pfaffl, M. W. Quantitative real-time RT-PCR- a perspective. J. Mol. Endocrinol. 2005, 34, 597-601. (4) Huggett, J.; Dheda, K.; Bustin, S.; Zumla, A. Real-time RT-PCR normalisation; strategies and considerations. Genes Immun. 2005, 6, 279-284. (5) Kreuzer, K. A.; Lass, U.; Landt, O.; Nitsche, A.; Laser, J.; Ellerbrok, H.; Pauli, G.; Huhn, D.; Schmidt, C. A. Highly sensitive and specific fluorescence reverse transcription-PCR assay for the pseudogene-free detection of beta-actin transcripts as quantitative reference. Clin. Chem. 1999, 45, 297-300. (6) Tricarico, C.; Pinzani, P.; Bianchi, S.; Paglierani, M.; Distante, V.; Pazzagli, M.; Bustin, S. A.; Orlando, C. Quantitative real-time reverse transcription polymerase chain reaction: normalization

(7)

(8)

(9)

(10)

(11)

(12)

(13)

(14) (15)

(16)

(17)

(18)

(19)

(20)

(21)

(22)

(23)

(24)

(25)

(26)

(27)

to rRNA or single housekeeping genes is inappropriate for human tissue biopsies. Anal. Biochem. 2002, 309, 293-300. Berry, M. N.; Friend, D. S. High-yield preparations of isolated rat liver parenchymal cells. A biocemical and fine structural study. J. Cell Biol. 1969, 43, 506-520. Andersson, T.; Fo¨rlin, L.; Hansson, T. Biotransformation of 7ethoxycoumarin in isolated perfused rainbow trout liver. Drug Metab. Dispos. 1983, 11, 494-498. Mortensen, A. S.; Tolfsen, C. C.; Arukwe, A. Gene expression patterns in estrogen (nonylphenol) and aryl hydrocarbon receptor agonists (PCB-77) interaction using rainbow Trout (Oncorhynchus mykiss) primary hepatocyte culture. J. Toxicol. Environ. Health A 2006, 69, 1-19. Meucci, V.; Arukwe, A. Transcriptional modulation of brain and hepatic estrogen receptor and P450arom isotypes in juvenile Atlantic salmon (Salmo salar) after waterborne exposure to the xenoestrogen, 4-nonylphenol. Aquat. Toxicol. 2006, 77, 167177. Leiva-Presa, A.; Mortensen, A. S.; Arukwe, A.; Jenssen, B. M. Altered hepatic retinol and CYP26 levels in adult European common frogs (Rana temporaria) exposed to p,p′-DDE. Mar. Environ. Res. 2006, 62 Suppl 1, S10-15. Arukwe, A.; Jenssen, B. M. Differential organ expression patterns of thyroid hormone receptor isoform genes in p,p′-DDE-treated adult male common frog, Rana temporaria. Environ. Toxicol. Pharmacol. 2005, 20, 485-492. Arukwe, A. Modulation of brain steroidogenesis by affecting transcriptional changes of steroidogenic acute regulatory (StAR) protein and cholesterol side chain cleavage (P450scc) in juvenile Atlantic salmon (Salmo salar) is a novel aspect of nonylphenol toxicity. Environ. Sci. Technol. 2005, 39, 9791-9798. Wong, M. L.; Medrano, J. F. Real-time PCR for mRNA quantitation. Biotechniques 2005, 39, 75-85. Olsvik, P. A.; Lie, K. K.; Jordal, A. E.; Nilsen, T. O.; Hordvik, I. Evaluation of potential reference genes in real-time RT-PCR studies of Atlantic salmon. BMC Mol. Biol. 2005, 6, 21. Unger, E.; Bohm, K. J.; Vater, W. Structural diversity and dynamics of microtubules and polymorphic tubulin assemblies. Electron Microsc. Rev. 1990, 3, 355-395. Unger, E.; Bohm, K. J.; Vater, W. Factors regulating microtubule structure-a minireview. Acta Histochem. Suppl. 1986, 33, 8594. Yan, H. Z.; Liou, R. F. Selection of internal control genes for real-time quantitative RT-PCR assays in the oomycete plant pathogen Phytophthora parasitica. Fungal Genet. Biol. 2006, 43, 430-438. von Smolinski, D.; Leverkoehne, I.; von Samson-Himmelstjerna, G.; Gruber, A. D. Impact of formalin-fixation and paraffinembedding on the ratio between mRNA copy numbers of differently expressed genes. Histochem. Cell Biol. 2005, 124, 177-188. Kim, B. R.; Nam, H. Y.; Kim, S. U.; Kim, S. I.; Chang, Y. J. Normalization of reverse transcription quantitative-PCR with housekeeping genes in rice. Biotechnol. Lett. 2003, 25, 18691872. Frost, P.; Nilsen, F. Validation of reference genes for transcription profiling in the salmon louse, Lepeophtheirus salmonis, by quantitative real-time PCR. Vet. Parasitol. 2003, 118, 169-174. Islinger, M.; Yuan, H.; Voelkl, A.; Braunbeck, T. Measurement of vitellogenin gene expression by RT-PCR as a tool to identify endocrine disruption in Japanese medaka (Oryzias latipes). Biomarkers 2002, 7, 80-93. Nicot, N.; Hausman, J. F.; Hoffmann, L.; Evers, D. Housekeeping gene selection for real-time RT-PCR normalization in potato during biotic and abiotic stress. J. Exp. Bot. 2005, 56, 29072914. Bose, A.; Guilherme, A.; Robida, S. I.; Nicoloro, S. M.; Zhou, Q. L.; Jiang, Z. Y.; Pomerleau, D. P.; Czech, M. P. Glucose transporter recycling in response to insulin is facilitated by myosin Myo1c. Nature 2002, 420, 821-824. Mahoney, D. J.; Carey, K.; Fu, M. H.; Snow, R.; Cameron-Smith, D.; Parise, G.; Tarnopolsky, M. A. Real-time RT-PCR analysis of housekeeping genes in human skeletal muscle following acute exercise. Physiol. Genomics 2004, 18, 226-231. Spivack, S. D.; Hurteau, G. J.; Jain, R.; Kumar, S. V.; Aldous, K. M.; Gierthy, J. F.; Kaminsky, L. S. Gene-environment interaction signatures by quantitative mRNA profiling in exfoliated buccal mucosal cells. Cancer Res. 2004, 64, 6805-6813. Celius, T.; Matthews, J. B.; Giesy, J. P.; Zacharewski, T. R. Quantification of rainbow trout (Oncorhynchus mykiss) zona radiata and vitellogenin mRNA levels using real-time PCR after

(28)

(29)

(30)

(31)

(32)

(33)

(34) (35)

(36) (37)

(38)

(39)

(40)

(41) (42)

(43) (44)

(45)

in vivo treatment with estradiol-17β or R-zearalenol. J. Steroid Biochem. Mol. Biol. 2000, 75, 109-119. Hilscherova, K.; Jones, P. D.; Gracia, T.; Newsted, J. L.; Zhang, X.; Sanderson, J. T.; Yu, R. M.; Wu, R. S.; Giesy, J. P. Assessment of the effects of chemicals on the expression of ten steroidogenic genes in the H295R cell line using real-time PCR. Toxicol. Sci. 2004, 81, 78-89. Aluru, N.; Renaud, R.; Leatherland, J. F.; Vijayan, M. M. Ah Receptor-mediated impairment of interrenal steroidogenesis involves StAR protein and P450scc gene attenuation in rainbow trout. Toxicol. Sci. 2005, 84, 260-269. Bas, A.; Forsberg, G.; Hammarstrom, S.; Hammarstrom, M. L. Utility of the housekeeping genes 18S rRNA, beta-actin and glyceraldehyde-3-phosphate-dehydrogenase for normalization in real-time quantitative reverse transcriptase-polymerase chain reaction analysis of gene expression in human T lymphocytes. Scand. J. Immunol. 2004, 59, 566-573. Krussel, J. S.; Huang, H. Y.; Simon, C.; Behr, B.; Pape, A. R.; Wen, Y.; Bielfeld, P.; Polan, M. L. Single blastomeres within human preimplantation embryos express different amounts of messenger ribonucleic acid for β-actin and interleukin-1 receptor type I. J. Clin. Endocrinol. Metab. 1998, 83, 953-959. Foss, D. L.; Baarsch, M. J.; Murtaugh, M. P. Regulation of hypoxanthine phosphoribosyltransferase, glyceraldehyde-3phosphate dehydrogenase and β-actin mRNA expression in porcine immune cells and tissues. Anim. Biotechnol. 1998, 9, 67-78. Carlyle, W. C.; Toher, C. A.; Vandervelde, J. R.; McDonald, K. M.; Homans, D. C.; Cohn, J. N. Changes in β-actin mRNA expression in remodeling canine myocardium. J. Mol. Cell. Cardiol. 1996, 28, 53-63. Mutimer, H.; Deacon, N.; Crowe, S.; Sonza, S. Pitfalls of processed pseudogenes in RT-PCR. BioTechniques 1998, 24, 585-588. Dakhama, A.; Macek, V.; Hogg, J. C.; Hegele, R. G. Amplification of human beta-actin gene by the reverse transcriptase-polymerase chain reaction: implications for assessment of RNA from formalin-fixed, paraffin-embedded material. J. Histochem. Cytochem. 1996, 44, 1205-1207. Malarstig, A.; Tenno, T.; Jossan, S.; Aberg, M.; Siegbahn, A. A quantitative real-time PCR method for tissue factor mRNA. Thromb. Res. 2003, 112, 175-183. Hauschildt, S.; Schwarz, C.; Heine, H.; Ulmer, A. J.; Flad, H. D.; Rietschel, E. T.; Jensen, O. N.; Mann, M. Actin: a target of lipopolysaccharide-induced phosphorylation in human monocytes. Biochem. Biophys. Res. Commun. 1997, 241, 670-674. Sabo-Attwood, T.; Kroll, K. J.; Denslow, N. D. Differential expression of largemouth bass (Micropterus salmoides) estrogen receptor isotypes R, β and γ by estradiol. Mol. Cell. Endocrinol. 2004, 218, 107-118. Lattier, D. L.; Reddy, T. V.; Gordon, D. A.; Lazorchak, J. M.; Smith, M. E.; Williams, D. E.; Wiechman, B.; Flick, R. W.; Miracle, A. L.; Toth, G. P. 17R-ethynylestradiol-induced vitellogenin gene transcription quantified in livers of adult males, larvae, and gills of fathead minnows (Pimephales promelas). Environ. Toxicol. Chem. 2002, 21, 2385-2393. Eleaume, H.; Jabbouri, S. Comparison of two standardisation methods in real-time quantitative RT-PCR to follow Staphylococcus aureus genes expression during in vitro growth. J. Microbiol. Methods 2004, 59, 363-370. Bustin, S. A. Absolute quantification of mRNA using real-time reverse transcription polymerase chain reaction assays. J. Mol. Endocrinol. 2000, 25, 169-193. Lee, P. D.; Sladek, R.; Greenwood, C. M.; Hudson, T. J. Control genes and variability: absence of ubiquitous reference transcripts in diverse mammalian expression studies. Genome Res. 2002, 12, 292-297. Steele, B. K.; Meyers, C.; Ozbun, M. A. Variable expression of some “housekeeping” genes during human keratinocyte differentiation. Anal. Biochem. 2002, 307, 341-347. Zhang, Q. J.; Chadderton, A.; Clark, R. L.; Augustine-Rauch, K. A. Selection of normalizer genes in conducting relative gene expression analysis of embryos. Birth Defects Res. A Clin. Mol. Teratol. 2003, 67, 533-544. Dheda, K.; Huggett, J. F.; Kim, L. U.; Zumla, A. Type 2 cytokines in respiratory syncytial virus bronchiolitis. Am. J. Respir. Crit. Care Med. 2004, 169, 1167-1168; author reply 1168.

Received for review June 27, 2006. Revised manuscript received September 20, 2006. Accepted October 4, 2006. ES0615223 VOL. 40, NO. 24, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7949