Contributed equally
Quantitative reverse transcription PCR (qRT-PCR) is becoming increasingly important in the effort to gain insight into the molecular mechanisms underlying adipogenesis. However, the expression profile of a target gene may be misinterpreted due to the unstable expression of the reference genes under different experimental conditions. Therefore, in this study, we investigated the expression stability of 10 commonly used reference genes during 3T3-L1 adipocyte differentiation. The mRNA expression levels of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and transferrin receptor (TFRC) significantly increased during the course of 3T3-L1 adipocyte differentiation, which was decreased by berberine, an inhibitor of adipogenesis. Three popular algorithms, GeNorm, NormFinder and BestKeeper, identified 18 ribosomal RNA and hydroxymethylbilane synthase (HMBS) as the most stable reference genes, while GAPDH and TFRC were the least stable ones. Peptidylprolyl isomerase A [PIPA (cyclophilin A)], ribosomal protein, large, P0 (36-B4), beta-2-microglobulin (B2M), α1-tubulin, hypoxanthine-guanine phosphoribosyltransferase (HPRT) and β-actin showed relatively stable expression levels. The choice of reference genes with various expression stabilities exerted a profound influence on the expression profiles of 2 target genes, peroxisome proliferator-activated receptor (PPAR)γ2 and C/EBPα. In addition, western blot analysis revealed that the increased protein expression of GAPDH was markedly inhibited by berberine during adipocyte differentiation. This study highlights the importance of selecting suitable reference genes for qRT-PCR studies of gene expression during the process of adipogenesis.
Obesity has become a widespread issue in modern society. When energy input exceeds energy expenditure, adipose tissue mass increases by adipocyte hyperplasia and hypertrophy (
Quantitative reverse transcription-polymerase chain reaction (qRT-PCR) is a powerful and efficient means of rapidly comparing gene expression patterns between different developmental stages and experimental conditions (
In this study, 3 popular algorithms, GeNorm (
Dulbecco’s modified Eagle’s medium (DMEM) and other culture reagents were obtained from Gibco Life Technologies (Grand Island, NY, USA). The cell culture plates were purchased from Nalge Nunc International (Roskilde, Denmark). Human insulin (HumulinR) was obtained from Eli Lilly S.A.S. (Fegersheim, France). Bovine serum albumin (BSA), IBMX and dexamethasone were purchased from Sigma (St. Louis, MO, USA). Anti-actin, anti-GAPDH, anti-α1-tubulin (tubulin), anti-mouse IgG and anti-rabbit IgG conjugated with horseradish peroxidase were obtained from Cell Signaling Technology (Beverly, MA, USA). Murine-derived 3T3-L1 preadipocytes were purchased from the American Type Culture Collection (ATCC; Rockville, MD, USA). Berberine was obtained from the National Institute for the Control of Pharmaceutical and Biological Products (Beijing, China).
3T3-L1 preadipocytes were grown and passaged in DMEM containing 25 mM glucose plus 10% fetal bovine serum (FBS). For adipocyte differentiation, 2-day post-confluent cells were placed in 10% FBS-DMEM with 250 nM dexamethasone, 0.5 mM IBMX and 1 μg/ml insulin. After 2 days, the medium was changed to 10% FBS-DMEM containing 1 μg/ml insulin alone for 2 additional days and was replaced with 10% FBS-DMEM. Thereafter, the medium was changed every 2 days.
3T3-L1 preadipocytes induced to differentiate for various days were washed with phosphate-buffered saline (PBS), fixed with 4% paraformaldehyde in 0.1 M phosphate buffer, pH 7.4 for 15 min at room temperature, and washed 3 times with deionized water. A mixture of Oil Red O (0.6% Oil Red O dye in isopropanol) and water at a 6:4 ratio was layered on the cells for 10 min, followed by hematoxylin counterstaining.
Total RNA was isolated using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) and reverse transcribed from random primers (Promega, Madison, WI, USA) according to the manufacturer’s instructions. Real-time PCR was performed on a Roche LightCycler 480 system using SYBR Premix Ex Taq™ (Takara, Otsu, Japan) in a final volume of 20 μl. The conditions for real-time PCR were as follows: denaturation at 95°C for 10 sec, 40 cycles at 95°C for 5 sec, and 60°C for 31 sec. A melting curve was built in the temperature range of 60–95°C at the end of amplification. The primer sequences used for real-time PCR are presented in
Cells in 6-well plates were washed twice with ice-cold PBS and placed immediately in lysis buffer containing 1 mM phenylmethylsulfonyl fluoride (PMSF), protease inhibitor cocktail I (Calbiochem/EMD Miliipore, Billerica, MA, USA) and phosphatase inhibitor cocktail V (Merck, Darmstadt, Germany). Lysates were gently mixed for 10 min at 4°C and then centrifuged at 13,000 × g for 15 min at 4°C. The protein concentration of the extracts was determined according to the method of Bradford, using BSA as the standard. Samples were separated by SDS-PAGE on 8% polyacrylamide gels and transferred onto PVDF-Plus membranes (Bio-Rad, Hercules, CA, USA). Primary antibodies were detected with donkey anti-rabbit at 1:2,000 for 1 h at room temperature. The blotted membrane was developed with ECL Advance (Cell Signaling Technology, Boston, MA, USA) and imaged with a LAS-4000 Super CCD Remote Control Science Imaging System (Fuji, Tokyo, Japan).
For stability comparisons of candidate reference genes, 3 validation software programs, GeNorm, NormFinder and BestKeeper, were used according to their original publications (
For obtaining accurate and reliable results of gene expression, attention should be paid to well-differentiated adipocytes, equal sample size, intact RNA and efficient primers apart from reference genes (
Another essential strategy suggested for normalizing qRT-PCR data is to use a suitable internal reference gene whose expression should not change with treatment or under different experimental conditions (
To identify the most suitable set of genes for normalization during 3T3-L1 adipocyte differentiation, the expression stability of the 10 candidate reference genes was analyzed using the GeNorm, NormFinder and BestKeeper programs. The GeNorm algorithm determines expression stability (M) through a pair-wise comparison of one candidate reference gene and all other candidate genes, independent of the level of gene expression for each sample (
NormFinder was designed to calculate stability by using the combined estimation of intra- and intergroup expression variations of the analyzed genes (
BestKeeper calculates the percentage coefficient of variation (CV) and standard deviation (SD) using the crossing point (CP) value of each candidate gene (
Taken together, the software analysis results indicated that GAPDH and TFRC ranked as the least stable reference genes, while 18S and HMBS ranked as the most stable ones (
To illustrate the impact of reference gene selection on the determination of target gene expression levels, the expression profiles of 2 key transcription factors in adipocyte differentiation, PPARγ2 and C/EBPα (
Berberine, one of the major constituents of the Chinese herb, Rhizoma coptidis, has been reported to improve insulin resistance and reduce hyperglycemia (
In addition, we further detected the protein expression levels of 3 commonly used reference genes. Western blot analysis revealed that GAPDH protein expression markedly increased at 5 and 7 days after induction, while there were no obvious changes in actin and tubulin protein expression throughout adipocyte differentiation. GAPDH protein expression levels were inhibited by berberine, but not those of actin and tubulin (
qRT-PCR has become one of the most popular techniques for quantifying mRNA levels, particularly for low-abundance genes. For the accurate and reliable analysis of target gene expression, the normalization of qRT-PCR data with suitable internal reference genes is required (
Firstly, we evaluated the differentiation status of 3T3-L1 adipocytes using Oil Red O staining. The results revealed that 3T3-L1 adipocytes were filled with lipid 7 days after the induction of differentiation, along with a significant increase in the PPARγ2 mRNA level. Therefore, these cells were considered to be fully differentiated. Secondly, the samples of various groups were simultaneously extracted, reversed and amplified to reduce intersample variability. We directly compared the CT values of all candidate reference genes in 3T3-L1 adipocytes differentiated for different periods of time (different number of days). As ΔCT values are exponents, all validation data were converted to ‘fold changes’ using the 2−ΔCT method for raw data. Accordingly, a ΔCT value of −0.5 and +0.5 is equivalent to 0.7- and 1.4-fold changes in relative gene expression, respectively. Using the criterion of ΔCT ≤±0.5 as a delimiter of reference gene suitability, we identified 8 reference genes (actin, PIPA, 18S, tubulin, 36-b4, HMBS, HPRT and B2M) suitable for use in target gene normalization. However, GAPDH and TFRC mRNA expression levels fluctuated with (ΔCT >1.0) after 3T3-L1 adipocyte differentiation, showing dramatic increases as compared to before induction. Thus, the 2 genes were inappropriate as endogenous reference genes under this experimental condition.
As a simple comparison of CT values revealed an ‘overall expression variation’ for the candidate reference genes, we further evaluated the expression stability of the 10 candidate reference genes using 3 software programs. Despite the differences we found in the reference gene ranking when using different statistical approaches, there was a general agreement among the methods for the determination of the most stable and unstable genes. The GeNorm and NormFinder identified 18S and HMBS as the most stable genes, while BestKeeper recommended HMBS and PIPA as the most stable ones. The 3 software programs all identified GAPDH and TFRC as the least stable genes, which confirmed the simple comparison of the CT value according to the guideline described in the study by Gorzelniak (
GAPDH is a classic glycolytic enzyme and has been traditionally considered as a reference gene, the levels of which are considered remain stable under different manipulations. Two previous studies have demonstrated that GAPDH is a stably expressed gene in human adipose cells across a wide range of experimental settings (
Insulin elicits a redistribution in TFRC expression in 3T3-L1 adipocytes through an increase in the rate constant for receptor externalization (
The expression profiles of target genes can be markedly influenced depending on the choice of normalization genes. If the wrong reference gene is selected, it can result in false findings (
This study clearly demonstrates the critical importance of reference gene validation for adipocyte differentiation studies, and highlights GAPDH and TFRC as unsuitable reference genes under this condition. 18S and HMBS are the most suitable reference genes for normalizing target genes during the course of 3T3-L1 adipocyte differentiation. The other 6 reference genes (actin, PIPA, Tubulin, 36-B4, HPRT and B2M) also showed relatively stable expression levels.
This study was funded by grants from the National Natural Science Foundation of China (30600294, 81070652, 81070617, 81170720, 81261120564 and 81270910).
3-isobutyl-1-methylxanthine
quantitative reverse transcription-polymerase chain reaction
glyceraldehyde-3-phosphate dehydrogenase
β-actin
18 ribosomal RNA
hydroxymethylbilane synthase
transferrin receptor
peroxisome proliferator-activated receptor γ2
CCAAT/enhanced binding protein
bovine serum albumin
fetal bovine serum
phosphate-buffered saline
cyclophilin A
ribosomal protein, large, P0
hypoxanthine-guanine phosphoribosyltransferase
beta-2- microglobulin
3T3-L1 adipocyte differentiation. 3T3-L1 preadipocytes were induced to differentiate into mature adipocytes with MDI (3-isobutyl-1-methylxanthine, dexamethasone and insulin). 3T3-L1 cells on various days of differentiation were stained by Oil Red O.
(A and B) Relative expression levels of candidate reference genes without normalization. Total RNA was isolated from 0 to 9 days after the 3T3-L1 preadipocytes were treated with MDI (3-isobutyl-1-methylxanthine, dexamethasone and insulin) and assessed for reference gene expression by qRT-PCR. Relative quantification analysis of gene expression data was conducted according to the 2−ΔCT method. Data are expressed as the relative fold to day 0. Significance was established at bP<0.01 compared with day 0.
Expression stability analysis of reference genes by GeNorm and NormFinder. Average expression stability values (M) of 10 reference genes were calculated by (A) GeNorm and (B) NormFinder. Reference genes were graphed to illustrate the average gene expression stability on the y-axis and its associated ranking from least to most stable expression on the x-axis. Lower M-values of average expression stability indicate more stable expression.
Relative expression levels of different target genes without and with normalization. Total RNA was isolated from 0 to 9 days after the 3T3-L1 preadipocytes were treated with MDI (3-isobutyl-1-methylxanthine, dexamethasone and insulin) and assessed for reference gene expression by qRT-PCR. Changes in PPARγ2 gene expression (A) without and with (B) normalization to 18S, tubulin, GAPDH, or TFRC are shown. (C) Raw gene expression and (D) normalized expression of C/EBPα. Significance was established at bP<0.01 compared with day 0.
Effects of berberine on mRNA expression levels of candidate reference genes. Total RNA was isolated from the differentiated 3T3-L1 cells (0–9 days) in the presence or absence of 5 μM berberine (BBR). Relative quantification analysis of (A) actin, (B) GAPDH, (C) 18S, (D) TFRC, (E) 36-B4, (F) tubulin, (G) PIPA, (H) HMBS, (I) HPRT, (J) B2M gene expression levels was conducted by the 2−ΔCT method. Data are expressed as the relative fold to day 0. Significance was established at aP<0.05, bP<0.01 compared with day 0.
The protein expression levels of 3 commonly used reference genes. 3T3-L1 preadipocytes were induced to differentiate into mature adipocytes with insulin, 3-isobutyl-1-methylxanthine and dexamethasone. Berberine (BBR) (5 μM) was added from the beginning of induction. The cells at various stages of differentiation were collected for analyzing GAPDH, actin and tubulin protein expression levels by western blot analysis.
Description of reference genes and target genes and their primer sequences used in qRT-PCR.
Gene symbol | Gene name | Accession no. (GenBank) | Gene function | Primer sequences (5′→3′) |
---|---|---|---|---|
Reference genes | ||||
Actin | β-actin | NM_007393.3 | Cytoskeletal structural protein | F: GGCTGTATTCCCCTCCATCG |
GAPDH | Glyceraldehyde-3-phosphate dehydrogenase | NM_008084.2 | Involved in glycolysis and gluconeogenesis | F: AACGACCCCTTCATTGAC |
PIPA | Peptidyl-prolyl isomerase A | NM_008907.1 | Cis-trans isomerization of oligopeptides, accelerate the folding of proteins | F: TGGAGCGTTTTGGGTCCAG3 |
TFRC | Transferrin receptor | NM_011638.4 | Uptake of iron-loaded transferrin into cells | F: GTTTCTGCCAGCCCCTTATTAT |
18S | 18S ribosomal RNA | NR_003278.3 | Eukaryotic small ribosomal subunit | F: ACCGCAGCTAGGAATAATGGA |
Tubulin | α1-tubulin | NM_011653.2 | Microtubules of the eukaryotic cytoskeleton | F: CCAGGGCTTCTTGGTTTTCC |
36-B4 | Ribosomal protein, large, P0 | NM_007475.5 | Protein synthesis | F: TGAGATTCGGGATATGCTGTTGG |
HMBS | Hydroxymethylbilane synthase | NM_001110251.1 | Heme synthesis and porphyrin metabolism | F: ATGAGGGTGATTCGAGTGGG |
HPRT | Hypoxanthine-guanine phosphoribosyltransferase | NM_013556.2 | Purine synthesis | F: TCAGTCAACGGGGGACATAAA |
B2M | beta-2-microglobulin | NM_009735.3 | β chain of MHC class I molecules | F: TTCTGGTGCTTGTCTCACTGA |
Target genes | ||||
PPARγ2 | Peroxisome proliferator-activated receptor | NM_001127330.1 | Induce adipogenesis | F: GCATGGTGCCTTCGCTGA |
C/EBP-α | CCAAT/enhancer binding protein α | NM_007678.3 | Activate the promoters of fat-cell specific genes in adipogenesis | F: CAAGAACAGCAACGAGTACCG |
Detailed expression stability analysis of candidate reference genes by BestKeeper software.
Gene | Actin | GAPDH | PIPA | TFRC | 18S | Tubulin | 36-B4 | HMBS | HPRT | B2M |
---|---|---|---|---|---|---|---|---|---|---|
n | 36 | 36 | 36 | 36 | 36 | 36 | 36 | 36 | 36 | 36 |
geo Mean (CP) | 20.57 | 19.18 | 22.27 | 26.57 | 12.79 | 20.97 | 22.13 | 27.06 | 23.68 | 19.25 |
ar Mean (CP) | 20.59 | 19.23 | 22.27 | 26.60 | 12.80 | 20.97 | 22.14 | 27.06 | 23.70 | 19.25 |
Min (CP) | 19.38 | 17.07 | 21.55 | 24.73 | 12.05 | 20.06 | 20.95 | 26.22 | 22.54 | 18.18 |
Max (CP) | 21.72 | 21.47 | 22.92 | 28.57 | 13.57 | 21.88 | 23.59 | 27.62 | 25.25 | 20.24 |
Std dev. (± CP) | 0.59 | 1.18 | 0.31 | 1.03 | 0.49 | 0.37 | 0.56 | 0.27 | 0.62 | 0.32 |
CV (% CP) | 2.88 | 6.13 | 1.38 | 3.86 | 3.79 | 1.77 | 2.54 | 0.99 | 2.60 | 1.66 |
Min (x-fold) | −2.29 | −4.32 | −1.64 | −3.59 | −1.68 | −1.88 | −2.27 | −1.79 | −2.22 | −2.10 |
Max (x-fold) | 2.21 | 4.87 | 1.57 | 3.99 | 1.71 | 1.87 | 2.76 | 1.47 | 2.96 | 1.98 |
Std dev. (± x-fold) | 1.51 | 2.26 | 1.24 | 2.04 | 1.40 | 1.29 | 1.48 | 1.20 | 1.53 | 1.25 |
Descriptive statistics of 10 reference genes are shown based on their crossing point (CP) values. n, number of samples; geo Mean (CP), geometric mean value of the CP; ar Mean (CP), average mean value of the CP; Min (CP) and Max (CP), extreme values of CP; Std dev. (± CP), standard deviation of the CP; CV (% CP), coefficient of variance expressed as a percentage on the CP level; Min (x-fold) and Max (x-fold), extreme values of expression levels expressed as absolute x-fold over or under regulation coefficient; Std dev. (± x-old), standard deviation of the absolute regulation coefficients.
Ranking of reference gene stability during 3T3-L1 adipocyte differentiation.
Ranking | BestKeeper | GeNorm | NormFinder |
---|---|---|---|
1 | HMBS | 18S | 18S |
2 | PIPA | HMBS | HMBS |
3 | B2M | PIPA | PIPA |
4 | Tubulin | 36-B4 | 36-B4 |
5 | 18S | HPRT | B2M |
6 | 36-B4 | B2M | Tubulin |
7 | Actin | Tubulin | HPRT |
8 | HPRT | Actin | Actin |
9 | TFRC | GAPDH | TFRC |
10 | GAPDH | TFRC | GAPDH |
Reference gene stability is ranked in order of decreasing expression stability.