Skip to main content

Evaluation of reference genes for real-time quantitative PCR studies in Candida glabrata following azole treatment



The selection of stable and suitable reference genes for real-time quantitative PCR (RT-qPCR) is a crucial prerequisite for reliable gene expression analysis under different experimental conditions. The present study aimed to identify reference genes as internal controls for gene expression studies by RT-qPCR in azole-stimulated Candida glabrata.


The expression stability of 16 reference genes under fluconazole stress was evaluated using fold change and standard deviation computations with the hkgFinder tool. Our data revealed that the mRNA expression levels of three ribosomal RNAs (RDN5.8, RDN18, and RDN25) remained stable in response to fluconazole, while PGK1, UBC7, and UBC13 mRNAs showed only approximately 2.9-, 3.0-, and 2.5-fold induction by azole, respectively. By contrast, mRNA levels of the other 10 reference genes (ACT1, EF1α, GAPDH, PPIA, RPL2A, RPL10, RPL13A, SDHA, TUB1, and UBC4) were dramatically increased in C. glabrata following antifungal treatment, exhibiting changes ranging from 4.5- to 32.7-fold. We also assessed the expression stability of these reference genes using the 2-ΔΔCT method and three other software packages. The stability rankings of the reference genes by geNorm and the 2-ΔΔCT method were identical to those by hkgFinder, whereas the stability rankings by BestKeeper and NormFinder were notably different. We then validated the suitability of six candidate reference genes (ACT1, PGK1, RDN5.8, RDN18, UBC7, and UBC13) as internal controls for ten target genes in this system using the comparative CT method. Our validation experiments passed for all six reference genes analyzed except RDN18, where the amplification efficiency of RDN18 was different from that of the ten target genes. Finally, we demonstrated that the relative quantification of target gene expression varied according to the endogenous control used, highlighting the importance of the choice of internal controls in such experiments.


We recommend the use of RDN5.8, UBC13, and PGK1 alone or the combination of RDN5.8 plus UBC13 or PGK1 as reference genes for RT-qPCR analysis of gene expression in C. glabrata following azole treatment. In contrast, we show that ACT1 and other commonly used reference genes (GAPDH, PPIA, RPL13A, TUB1, etc.) were not validated as good internal controls in the current model.


The investigation of gene expression has become increasingly prevalent in numerous animal, human, microorganism, and plant studies[15]. The quantitation of gene expression requires sensitive, precise, and reproducible measurements for specific mRNA sequences. Generally, gene expression levels can be determined by a variety of techniques, including Northern blotting, RNase protection assay, semi-quantitative reverse-transcription PCR, and real-time quantitative PCR (RT-qPCR)[4]. RT-qPCR has gained favor as it is a highly sensitive, accurate, and fast technique that offers high-throughput and the ability to detect low-abundance mRNAs[6] and quantify mRNA copy number[7]. Thus, RT-qPCR has been used for countless different applications[15].

One of the main uses of RT-qPCR, when coupled with reverse transcription, is to measure gene expression at the mRNA level in various biological samples. However, there is substantial technical variability associated with RT-qPCR, arising from inherent differences in samples, sample collection, RNA degradation and extraction efficiency, quantity and quality of input RNA, reverse transcription and PCR efficiency, and pipetting accuracy or error. Researchers have employed a number of strategies to normalize their data, including normalization to (i) genomic DNA, (ii) total RNA, (iii) an external standard, and (iv) a reference gene. The most common practice is to normalize to an internal control gene termed a reference gene. A reference gene is subject to the same errors in cDNA preparation as the gene of interest, making it an excellent normalizing control. However, selection of an inappropriate reference gene can add large unpredictable error to the analysis and result in incorrect estimates[8]. The ideal reference gene should have a stable RNA transcription level under different experimental conditions and be sufficiently abundant across different tissues and cell types. However, it has become apparent that such an ideal reference gene has not yet been identified[9]. The most commonly used reference genes, including β-actin, cyclophilin, GAPDH, tubulin, and 18S and 28S ribosomal RNAs, have shown variable expression levels in different cells and tissues under different conditions, and therefore they are unsuitable for normalization purposes owing to large measurement error[6, 1033]. Hence, it is no longer acceptable to arbitrarily select any reference gene for normalization; it must be demonstrated that the reference gene of choice is suitable for the experiment in question.

In recent decades, Candida glabrata has emerged as the second most common cause of invasive fungal infection[1, 34, 35]. Azoles such as fluconazole are the first-line drugs for the treatment of fungal infections caused by C. glabrata. However, resistance to azoles can arise rapidly in C. glabrata during treatment of patients with azoles[36]. An increasing body of evidence has implicated ATP-binding cassette transporters (e.g., Cdr1 and Pdr1) and sterol biosynthetic enzymes (e.g., Erg3 and Erg11) in azole resistance in C. glabrata in both clinical and laboratory settings[1, 3437]. The expression of these genes in C. glabrata in response to azoles is not completely understood. Therefore, we set out to establish an in vitro model for investigating azole-inducible gene expression in C. glabrata, using RT-qPCR. For reliable gene expression analysis, a compulsory step is the selection of good reference genes for normalization; however, no validated reference genes have been reported for the relative quantification of the mRNA expression profile in C. glabrata following exposure to azoles.

We have been using ACT1 as the internal control for gene expression analysis by RT-qPCR in clinical isolates of C. glabrata in the absence of drug challenges[1]. Other researchers also use ACT1 as the reference gene for azole-inducible gene expression studies by slot blotting in Candida species[34, 37, 38]. However, the suitability of ACT1 in studies of azole-inducible gene expression in C. glabrata has not been validated. In this work, we evaluated 16 reference genes to establish their suitability as control genes for normalization and identified a set of genes that are suitable for quantitative gene expression analysis by RT-qPCR in C. glabrata following fluconazole treatment.


Cell culture and drug treatment

All five C. glabrata strains (Table1) used in the present study were grown in YPD broth (Difco Laboratories, Detroit, MI, USA) at 30°C with shaking at 225 rpm. The ura3 mutant Cg84u did not grow in minimal (MIN) medium, unless supplemented with 20 μg/ml of uracil (Sigma-Aldrich, St. Louis, MO, USA).

Table 1 Candida glabrata strains used in this study

Fluconazole (Euroasian Chemicals Private Ltd., Fort, Mumbai, India) was added to cultures of each strain at a final concentration of 200 μg/ml, followed by continued incubation with shaking for 2 h. Cell cultures without fluconazole treatment served as controls.

The susceptibility of each C. glabrata strain to fluconazole was determined on YPD agar medium using an E-test (AB Biodisk, Solna, Sweden) according to the manufacturer’s instructions (Table1).

RNA isolation and reverse transcription

Total RNA was extracted from C. glabrata logarithmic-phase cultures grown in YPD broth, using TRIzol reagent (Invitrogen, Life Technologies, Grand Island, NY, USA) according to the manufacturer’s instructions. The concentration and purity of the RNA was determined using a UV spectrophotometer (NanoDrop 2000C; ThermoFisher Scientific, Waltham, MA, USA) by measuring the absorbance at 230 (OD230), 260 (OD260) and 280 nm (OD280). The OD260nm/OD280nm of the samples, reflecting the average purity, ranged from 1.80 to 2.05, and the OD260nm/OD230nm was in the range of 2.00–2.60. The integrity of the RNA was further checked in a selected subset of samples by electrophoresis through 1% denaturing and non-denaturing agarose gels.

Reverse transcription (RT) was performed on 1 μg of total RNA using a commercially available kit. Prior to RT, the total RNA samples were treated with DNase for 30 min at 37°C (TURBO DNA-free; Ambion, Life Technologies, Grand Island, NY, USA) according to the manufacturer’s instructions. RNA was converted to cDNA using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Life Technologies, Carlsbad, CA, USA). The reaction took place in a thermal cycler (T3 Thermocycler; Biometra, Goettingen, Germany) with a single cycle and incubation periods of 25°C for 10 min, 37°C for 120 min, and 85°C for 5 min. All investigated samples were transcribed with the same reverse transcription reaction conditions. Negative controls, which were run simultaneously, did not contain either RNA (no template control) or no reverse transcriptase (RT negative control), to control for RNA and genomic DNA contamination, respectively.

Primer and probe design

Primers and probes were designed in our laboratory using the primer analysis software Primer Express 3.0 (Applied Biosystems). TaqMan probes were synthesized by Applied Biosystems, and primers were synthesized by Invitrogen/Life Technologies. The primers and TaqMan probes used in the current study were selected to bind specifically to the cDNAs of Cg84u and other C. glabrata strains (Table1). The sequences of TaqMan probes and forward and reverse primers, the gene numbers, and the localization for each PCR assay for the 16 reference genes and 10 target genes assessed in this study are listed in Additional file1, Additional file2, and Additional file3.

RT-qPCR analysis

RT-qPCR for reference gene RNA transcription was performed by SYBR Green chemistry (SYBR Green PCR Master Mix; Applied Biosystems). The increase in fluorescence of the SYBR Green dye was monitored using a 7500 Real-time PCR System (Applied Biosystems). This technique has been successfully used to validate reference gene expression levels in yeast and other cell types[41]. Primers were used at 300 nM each for specific forward and reverse primers and cDNA at 25 ng in 25-μl reactions. Primer sets for the reference genes (Additional file2) were used to amplify the open reading frame (ORF) region of the genes according to the following conditions: one cycle of 50°C × 2 min, 95°C × 10 min; followed by 40 cycles of 95°C × 15 s, 60°C × 1 min; with dissociation (a melting curve) during the last cycle of 95°C × 15 s, 60°C × 1 min, 95°C × 15 s. The dissociation protocol to determine the melting curve from 60°C to 95°C for each PCR product was added after thermocycling to verify that each primer pair produced only a single product. All samples gave only a single peak, indicating a single pure product and no primer/dimer formation. Real-time PCR efficiencies were acquired by amplification of a standardized dilution series of the template cDNA and were determined for each gene as the slope of a linear regression model. PCR efficiency was determined by measuring the CT to a specific threshold for a serial dilution of cDNA. The corresponding real-time PCR efficiencies were then calculated according to the equation: E = 1 0 1 / slope 1 × 100 . All PCRs displayed efficiencies between 94% and 119%.

To study target gene expression, the amplification was detected in real time using TaqMan chemistry (TaqMan Universal PCR Master Mix; Applied Biosystems) according to the manufacturer’s instructions. RT-qPCR was performed in 96-well microtiter plates with a final volume of 25 μl, using a 7500 Real-time PCR System (Applied Biosystems). Primers were used at 300 nM each for specific forward and reverse primers; probes, at 200 nM; and cDNA, at 25 ng in 25-μl reactions. Primer sets and TaqMan probes for the target genes (Additional file3) were used to amplify the ORF region of the genes under the following conditions: one cycle of 50°C × 2 min, 95°C × 10 min; and then 40 cycles of 95°C × 15 s, 60°C × 1 min. The parallel amplification between the reference genes and the target genes was confirmed for each with probe-primer sets. To minimize technical (run-to-run) variation between the samples, all samples were analyzed in the same run for both target genes and reference genes.

Evaluation of reference gene expression stability using four different software packages

Non-normalized gene expression levels from our experimental data were analyzed to evaluate the expression stability of potential reference genes, using four different software programs: hkgFinder[42], geNorm[9, 43], BestKeeper[44], and NormFinder[26]. The hkgFinder software computes the pooled standard deviation (SD) of non-normalized expression data from both phenotypes (i.e., azole-treated and untreated C. glabrata cells), the fold change (FC) values between the two phenotypes, and Student’s t-tests of the log2 fold-change values with Holm-adjusted P-values. The reference genes with the smallest SD and the smallest, non-significant FC are identified as the best potential reference genes (Additional file4). The geNorm software computes a stability value (M) and a pairwise variation (V), which are used to evaluate each individual reference gene candidate or each combination of reference genes, called a normalization factor (NF). The pairwise variability V of two genes j and k is the standard deviation of all log2 ratios of a j /a k , while the stability value M of gene j is the mean of all possible pairwise variations V jk . Graphs of the M values help identify the best individual reference genes, and graphs of the V values identify the optimal number of reference genes for an NF. Note that an earlier version called geNorm Excel was produced as an add-in for MS Excel and it required several hand calculations to convert crossing point (CP) values into relative expression values. That version is now unavailable and the new geNorm PLUS from Biogazelle does not require those hand calculations. The BestKeeper software uses pairwise correlation to determine whether potential reference genes should be included in a BestKeeper Index, which is simply the geometric mean of the CP or cycle threshold (CT) values. The NormFinder software computes a different type of stability value (ρ ig ) based on the intragroup and intergroup variation of the expression data. The software instructions from each package were followed when inputting the RT-qPCR data, fetching the output, and interpreting the analysis results.


Stability of RNA transcription of reference genes in C. glabrata following azole stimulation

In the present study, 16 reference genes were chosen from among commonly used reference genes in published studies with yeast and mammalian cells, paying close attention to selecting genes that belong to different functional classes; their full names, symbols, functions, and gene numbers are listed in Additional file1. Our aim was to identify reference genes with minimal variability under our experimental conditions. To this end, RT-qPCR was used to measure the RNA transcription levels of 16 reference genes in C. glabrata cells following fluconazole treatment. To compare the different RNA transcription levels after azole exposure, the CT values of the reference genes were directly compared between the drug-treated (t) and untreated (ut) samples using the formula: CT Change = CT(ut) − CT(t). The CT is defined as the number of cycles needed for the fluorescence signal to reach a specific threshold level of detection and is inversely correlated with the amount of template cDNA present in the reaction. Thus, a higher value of CT Change indicates lower stability of a reference gene, considering that the expression of a reference gene should not change significantly with azole treatment. As expected, the RNA transcription levels of the reference genes varied (Table2). The three ribosomal RNA subunits RDN5.8, RDN18, and RDN25 were the most stable reference genes, with CT Change values less than 0.5, while UBC13, PGK1, and UBC7 were relatively stable with CT Change values of only around 1.5. By contrast, the other 10 reference genes showed marked variation in response to fluconazole. Among them, the most prominent variation was found in the RNA transcription levels of SDHA, ACT1, and RPL13A; as seen in Table2, the CT Change values of these reference genes were as high as 5.03.

Table 2 Stability of RNA transcription of reference genes in fluconazole-treated C. glabrata as determined by the 2 -ΔΔCT method

To validate the stability of candidate RNA transcription under our experimental conditions, the levels were compared with the RDN5.8 RNA transcription level. We chose to use RDN5.8 as a normalizer because it meets the requirement for both stability and suitability as a reference. First, we calculated the ΔCT between the CT values of reference genes and RDN5.8 from fluconazole-treated (t) and untreated (ut) cells:

ΔCT(t) = CT(t reference) − CT(t RDN5.8) and ΔCT(ut) = CT(ut reference) − CT(ut RDN5.8).

In the second step, we subtracted the change in RNA transcription in untreated samples from the change in treated samples to obtain the ΔΔCT(t):

ΔΔCT(t) = ΔCT(t) − ΔCT(ut).

Thus, ΔΔCT(t) indicates the change in RNA transcription caused by fluconazole treatment after normalization to RNA transcription changes in RDN5.8. A high ΔΔCT(t) value indicates a significant fluconazole-related change in the RNA transcription level of the tested gene. A positive ΔΔCT(t) value indicates down-regulation of transcription, whereas a negative ΔΔCT(t) indicates up-regulation of a gene’s transcription following azole treatment. We then transformed ΔΔCT(t) into a 2-ΔΔCT value, which indicates the fold change in RNA transcription of a reference gene in response to fluconazole as compared with the level in untreated cells. The calculated ΔΔCT(t) and 2-ΔΔCT values of the 16 tested reference genes in drug-treated samples are given in Table2.

Following stimulation with fluconazole, the RNA transcription of SDHA, ACT1, and RPL13A was highly regulated in C. glabrata cells, with changes ranging from 13- to 23-fold compared with transcription in the untreated cells. There was almost no regulation of RDN5.8, RDN18, and RDN25 RNA transcription, while PGK1, UBC7, and UBC13 RNA transcription were only approximately 2-fold induction in response to drug treatment (Table2).

Determination of reference gene expression stability by four different specific software packages

To choose the best reference genes, the reference gene stability was evaluated using four different software packages: hkgFinder, geNorm, BestKeeper, and NormFinder. Each of these software packages uses a slightly different metric to evaluate the candidate reference genes. Our goal was to compare the findings from these four different methods and look for the best-scoring reference genes that might be common to these different methods.

The hkgFinder software identifies the best reference genes by ranking the candidate genes according to their SD and FC values (Table3). Among the 16 potential reference genes, the SDs ranged from 0.19 to 2.76, and the FCs ranged from 1.2 to 32.7. The best three reference gene candidates were RDN18, RDN25, and RDN5.8. The next three best candidate reference genes, which also had reasonable SD and FC values, were UBC13, PGK1 and UBC7.

Table 3 Assessment of reference gene expression stability in fluconazole-treated C. glabrata by using hkgFinder

The geNorm software evaluates reference genes by their M-stability values and V-pairwise variability values. Low M values represent more stable expression and thus the most suitable reference genes (Figure1). The geNorm analysis identified RDN18, RDN25, and RDN5.8 as the three most stable genes; UBC13, PGK1, and UBC7 as relatively stable genes; and RPL13A, ACT1, and SDHA as the three least stable genes under fluconazole treatment in C. glabrata. Interestingly, the ranking of expression stability of the 16 reference genes was identical between the geNorm program and the hkgFinder tool (Figure1; Tables3 and4). The geNorm program also estimates the optimal number of reference genes that could be used in combination as an NF value (Figure2). Each NF was calculated as the geometric mean of the two most stable genes, then the pairwise variability V was computed between NFn and NFn + 1 for n = 2, …, 15. Vandesompele et al.[9] proposed 0.15 as a cutoff value for V below which additional reference genes do not need to be added to the NF. Adding the third gene to the most stable two reference genes, RDN18 and RDN25, produced a V below the cutoff of 0.15, indicating that it would not be necessary to include additional reference genes for normalization; the second best choice of reference gene combination based on geNorm V was the top five most stable genes, i.e., RDN5.8, RDN18, RDN25, UBC13, and PGK1. This reference gene combination showed smaller variation than other gene combinations, with a smaller V of 0.091 (Figure2).

Figure 1
figure 1

Determination of the average expression stability ( M ) of the reference genes using the geNorm program. The reference genes were serially excluded from the analysis, with M representing the mean pairwise variation between an individual reference gene and all other tested reference genes. The reference gene indicated at each point on the x-axis is the one that is to be excluded from the following step. The most stable reference genes are those that are still included, i.e., those that exhibit the lowest M values. Shown is the stability ranking of the reference genes in all samples. Genes are ranked from left to right in order of increasing expression stability (decreasing M value)

Table 4 Comparison of reference gene expression stability* as determined by four different software packages
Figure 2
figure 2

Determination of the optimal number of reference genes as internal references for normalization using geNorm analysis. The geNorm program calculates a normalization factor from at least two reference genes and the mean pairwise variation (V) between every combination of sequential normalization factors in order to determine the minimum number of reference genes required for accurate normalization in the samples. For example, V 5/6 represents the comparison of the normalization factors from five and six reference genes, respectively. On the left-most side is the pairwise variation when the number of reference genes is increased from two to three (V 2/3). Stepwise inclusion of less stable genes generates the subsequent data points. A decrease in the V value indicates a positive effect and means that the added gene should preferably be included for calculation of a reliable normalization factor. The cutoff value for V, below which the inclusion of an additional reference gene does not result in a significant improvement of normalization, was set at 0.15. It was apparent from the analysis of all studied samples that the combination of the two most stable reference genes is the best option and the combination of the five most stable reference genes is the second-best option for accurate normalization

The NormFinder program was also used to rate candidate reference gene stability according to a stability value computed from the intragroup and intergroup expression variability. The least reliable reference genes identified by this program were RPL13A, ACT1, and SDHA, which were identical to the worst reference genes identified by geNorm and hkgFinder analyses (Table4). However, the ranking order of the most stable genes and the relatively stable genes by the NormFinder program was different from that generated by geNorm and hkgFinder (Figure1; Tables3 and4). The geNorm and hkgFinder analyses graded RDN18, RDN25, and RDN5.8 as the most stable reference genes, followed by UBC13, PGK1, and UBC7 based on gene expression stability, whereas NormFinder rated UBC7, PGK1, and UBC13 as the most stable reference genes, followed by RDN5.8, RDN18, and RDN25 (Figure1; Tables3 and4).

Finally, the BestKeeper program was used to grade candidate reference gene stability. This approach permits a comparative analysis across reference genes. Ten reference genes analyzed were correlated and were combined into an index. Subsequently, the correlation between each reference gene and the index was calculated. The best correlations between the reference genes and the BestKeeper index were obtained for UBC7, UBC13, and PGK1 (r = 0.983, 0.978, and 0.974, respectively; Table4). The rankings of the top three and the last three reference genes identified by the BestKeeper program were the same as those generated by the NormFinder analysis, although the order of stability of the other reference genes differed slightly between the two programs (Table4).

Validation of six candidate reference genes using the comparative CT method

Following the identification of the most stable reference genes from the full gene panel of 16 genes, the comparative CT method was used to validate their suitability. The comparative CT method, also referred to as the ΔΔCT method, is a relative quantitation of gene expression between a specific target gene and a reference gene. For the comparative CT method to be valid, the efficiency of the target amplification and the efficiency of the reference (internal control) amplification must be approximately equal, and this must be determined in a validation experiment.

To this end, we first determined the amplification efficiency of 10 target genes (CDR1, PDH1, PDR1, SNQ2, YOR1, ERG2, ERG3, ERG4, ERG10, and ERG11) and six reference genes (ACT1, PGK1, RDN5.8, RDN18, UBC7, and UBC13). Standard curves were generated by plotting the dilutions of the cDNA of each gene against the CT values. The linear correlation coefficient (r2) for all 10 target genes and the six reference genes ranged from 0.98 to 1.0. Based on these slopes of the standard curves, the amplification efficiencies of the cDNA standards, derived from the formula E = (10−1/slope −1) × 100, ranged from 94 to 119%. The CT values of all 16 genes in the samples were within the range of the standard curves. Next, the ΔCT (ΔCT = CTtarget − CTreference) was calculated using the CT values generated from standard curve mass points (target vs. reference gene). These ΔCT values were then plotted versus log10 input amount of cDNA to create a semi-log regression line. The slope of the resulting semi-log regression line was used as a general criterion for passing a validation experiment. In a validation experiment that passes, the absolute value of the slope of ΔCT versus log10 input cDNA would be <;0.1, meaning the two CT versus log10 concentration curves are nearly parallel. As seen in Table5, our validation experiments passed for all reference genes analyzed except RDN18, which had an absolute value >0.1 for the slopes of ΔCT versus log10 input cDNA for all 10 target genes evaluated. Thus, the amplification efficiency of RDN18 was clearly different from that of the ten target genes, whereas the other five reference genes (ACT1, PGK1, RDN5.8, UBC7, and UBC13) had PCR efficiencies that were similar or relatively equivalent to the target amplification efficiencies (Table5).

Table 5 Validation of reference gene suitability as an internal control based on the correlation of amplification efficiency between reference and target genes in fluconazole-treated C. glabrata cells

Comparison of expression levels of inducible target genes using different reference genes for normalization

To test the effect of azole on the expression of pleiotropic drug resistance genes in C. glabrata, we assessed the fluconazole-induced expression of two ABC genes (CDR1 and PDR1) and one ERG gene (ERG4) in five C. glabrata strains, including the PDR1 mutant strain CgB4 (Table1). For comparison, we used both RDN5.8 and ACT1 as references for normalization. As shown in Table6, fluconazole markedly induced increases in ERG4 mRNA levels in all C. glabrata strains examined when normalized to RDN5.8. Fluconazole also significantly increased CDR1 and PDR1 mRNA expression in all of the strains except CgB4, consistent with the critical role of PDR1 in azole-induced transactivation of ABC transporter gene expression, but not ergosterol biosynthesis gene expression, in C. glabrata. In contrast, when using ACT1 as the reference gene for quantification, fluconazole appeared to down-regulate the expression of all three target genes in the five C. glabrata strains (Table6).

Table 6 Comparison of fluconazole-induced target gene expression among five different C. glabrata strains using RDN5.8 or ACT1 as the reference for normalization

Finally, we compared the fluconazole-inducibled mRNA expression levels of four target genes (CDR1, PDR1, ERG4, and ERG10) in C. glabrata (Cg84u strain) after normalizing to different reference genes (ACT1, PGK1, RDN5.8, and UBC13), individually and in pairs. Differences in quantitation were detected according to the reference genes used. As seen in Table7, normalization of the RT-qPCR data against the reference genes suggested as optimal by the four software packages (hkgFinder, geNorm, BestKeeper, and NormFinder) or the 2-ΔΔCT method, gave comparable relative expression levels of the target genes under fluconazole treatment in C. glabrata. However, normalization against ACT1 resulted in relative expression levels of the targets that were substantially different from those normalized using other reference genes, implying that ACT1 is not a suitable reference gene for these studies. Taken together with the data shown above, these results demonstrate that the relative quantification of azole-inducible gene expression varies largely depending on the reference gene and the number of reference genes used for normalization. This highlights the importance of choosing a suitable reference gene or reference gene pair when using RT-qPCR to determine the level of target gene expression in this model system.

Table 7 Comparison of the relative mRNA expression levels* of four target genes in fluconazole-treated C. glabrata when normalized to different reference genes


In any gene expression study, the selection of a valid normalization or internal control gene to correct for differences in RNA sampling is critical in order to avoid misinterpretation of results and to obtain reliable conclusions. When choosing a reference gene as the internal endogenous control for gene expression studies by RT-qPCR, two important criteria must be met. The expression of the reference gene must remain stable throughout the given intervention (i.e., stability), and the amplification efficiency of the reference gene should be similar to that of the genes of interest (i.e., suitability). In the present study, we used five different methods to evaluate 16 reference genes for potential use as internal controls and found that the reference genes performed differently in terms of stability and suitability in C. glabrata cells upon exposure to fluconazole. To our knowledge, this is the first report to validate reference genes as RNA internal references in C. glabrata.

The poor performance of ACT1 in C. glabrata cells was surprising, given that this gene has been used frequently as the reference gene in earlier gene expression studies[1, 34, 37, 38]. Our data clearly demonstrate the unsuitability of ACT1 as an internal control for gene expression studies in C. glabrata following fluconazole treatment. The initial results gained from using ACT1 as the internal control suggested that target gene expression was not up-regulated (Tables6 and7). In fact, the only substantial change caused by azole treatment was a greater increase in ACT1 RNA transcription compared with target gene transcription. While these findings are relevant to our specific study, it appears that numerous other studies have also shown the potential of ACT1 to detrimentally affect the accuracy of results[11, 12, 15, 19, 26, 27, 29, 30, 32, 33, 4547]. We have been using ACT1 as the internal control for quantitation of gene expression by RT-qPCR in clinical isolates of C. glabrata, and we find that this gene works well as the reference in cells without azole or other agent stimulation[1]. However, Edlind and colleagues used ACT1 as the reference gene for azole-inducible gene expression studies by slot blotting in Candida species, and their data clearly show the variation of ACT1 expression in response to azoles in their systems[34, 38]. Thus, our data revealing the instability of ACT1 in C. glabrata following antifungal treatment, combined with evidence from mammalian and other fungus studies, add to the growing body of evidence that ACT1 expression is unstable across various cell types and under different experimental conditions[11, 12, 15, 26, 27, 29, 30, 32, 33, 4547].

Although ACT1 gene expression was variable in response to fluconazole, the three ribosomal RNAs (i.e., RDN5.8, RDN18, and RDN25) remained unaffected and showed stable expression in azole-treated C. glabrata. These results indicate that ribosomal RNA expression offers superior consistency compared with the expression of ACT1 and the other reference genes assessed. The stable expression levels of 18S and 28S rRNAs relative to other reference genes under a variety of experimental conditions has previously been described for numerous systems, including both mammalian and yeast cells[10, 11, 15, 18, 23, 30, 4850]. The levels of ribosomal RNA, which represents 80% of total RNA, are thought to be less likely to vary under conditions that affect the expression of mRNAs because they are transcribed by a distinct RNA polymerase. As an example, Thellin et al. and other groups have recommended the use of 18S or 28S rRNA as an internal control for mRNA quantification studies because mRNA variations are weak and cannot highly modify the total RNA level[10, 11, 15, 18, 30, 39, 4852]. However, our further validation experiments showed that of slopes of ΔCT versus log10 cDNA were sufficiently parallel between RDN5.8 and the target genes, but not between RDN18 or RDN25 and the target genes. With RDN18 and RDN25, the absolute slope values of the ΔCT versus log10 input cDNA lines were >0.1 for all target genes. These data indicate that the amplification efficiency of RDN5.8 was similar to the efficiencies of the target genes, whereas the amplification efficiencies of RDN18 and RDN25 were different from the target gene amplification efficiencies. This may be attributable to the much higher abundance of RDN18 and RDN25 than RDN5.8 compared with target mRNA transcripts, making it difficult to accurately subtract the baseline value in RT-qPCR data analysis. Therefore, although all three ribosomal RNA subunits were stable during fluconazole stimulation, only RDN5.8 may offer a more accurate and suitable alternative to ACT1 as an internal control for gene expression studies in C. glabrata.

GAPDH, a glycolytic enzyme, is encoded by a single gene and has the advantage of being highly conserved across different species[53, 54]. Like 18S rRNA and β-actin, GAPDH has been commonly used as an internal control, often without testing. In the present study, GAPDH showed much higher variability than any of the ribosomal RNAs in fluconazole-treated samples. These data demonstrate that GAPDH is not an appropriate control gene for these studies, as has been pointed out in previous examples, and that it may lead to incorrect results under specific experimental conditions[1215, 19, 2931, 33, 55]. Previous studies have indicated the instability of GAPDH in mammalian systems, and this study broadens the scope of this phenomenon to C. glabrata as well.

PGK1 also plays important roles in the glycolytic pathway, and PGK1 and GAPDH are potentially co-regulated[56]. In our data, however, their potential co-regulation was not significant. PGK1 mRNA levels remained relatively stable, in contrast to the marked variation in GAPDH mRNA levels in C. glabrata cells, following fluconazole challenge. Moreover, our comparative CT calculations showed that the efficiency of PGK1 amplification was approximately equal to the efficiencies of the target gene amplifications. Although PGK1 shows some variation as a reference gene, this may not affect experimental results as long as the intergroup difference being measured is greater than the reference gene variation, that is, a reference gene RNA that has an error of 1 log2 may not be ideal, but it would be sufficient to measure a 2 log2 change in a gene of interest. Thus, it is inferred that PGK1 may be a suitable reference gene for the analysis of expression for genes with higher azole-inducible mRNA levels, such as ERG4 and ERG10.

Ubiquitin is a small regulatory protein that has been found in almost all tissues of eukaryotic organisms. The UBC gene codes for a polyubiquitin precursor protein[57]. Due to its ubiquitous existence in different tissues and cells in eukaryotes, there are an increasing number of studies in the literature using the UBC gene as the internal standard for gene expression analysis in different eukaryotic cell systems[26, 29]. Out of curiosity, we validated three UBC genes (UBC4, UBC7, and UBC13) in this study. Interestingly, we found that UBC7 and UBC13 mRNAs (particularly the latter) were relatively stable in C. glabrata during fluconazole treatment. In addition, our validation experiments demonstrated that the amplification efficiencies of these genes were approximately equal to those of the target genes. These findings indicate that like PGK1, UBC13 and UBC7 may also be suitable internal controls for quantifying the expression of specific genes with higher azole-inducible mRNA levels in C. glabrata, such as some ergosterol biosynthesis genes.

To successfully select reference genes for our studies involving azole treatment, we also investigated seven other reference genes, in addition to the reference genes mentioned above, with a diversity of functions. These reference genes can be generally classified into several groups: transcription-related genes (EF1α), structure/cytoskeleton-related genes (TUB1), protein synthesis-related genes (RPL2A, RPL10, and RPL13A), and finally, genes that cannot be clearly categorized, including PPIA and SDHA. These potential reference genes such as PPIA, RPL13A, and TUB1 are other examples of commonly used internal controls[28, 29, 32]. For example, PPIA has been used as a reference gene because of its remarkable evolutionary conservation and broad cellular and tissue distribution[58]. Although these seven reference genes have been used as internal standards for normalization in countless studies, all of these genes showed an unacceptable variable expression in our model system, with values ranging from a 4.4-fold induction with TUB1 to a 23-fold induction with SDHA after antifungal treatment. Altogether, these results suggest that the choice of internal controls is highly specific to a particular experimental condition, thus highlighting the importance of validating reference genes for each experimental model before commencement of RT-qPCR studies.

Although it is now widely accepted that normalizing to a single reference gene represents a strategy that is simple to use and can control for every stage of the RT-qPCR, some researchers also advocate the use of two or more reference genes, rather than relying on a single RNA transcript[9, 10, 44, 59]. This is a robust method for providing accurate normalization and is consequently preferable when fine measurements are to be made. According to Vandesompele et al.[9], the purpose of normalization is to remove the sampling difference (such as RNA quantity and quality) in order to identify real gene-specific variation. They provided evidence that a conventional normalization strategy based on a single gene can lead to erroneous normalization. However, it is not always possible to measure multiple reference genes because of limited sample availability and cost. Furthermore, even when multiple genes are chosen, the resolution of the particular assay remains dependent on the variability of the chosen reference genes. As to our case, the geNorm analysis using the geometric mean of the expression of the 16 candidate cDNAs suggested the use of RDN5.8, RDN18, and RDN25 in combination or the combination of these three ribosomal RNAs plus UBC13 and PGK1 as the reference control in the current study. However, the geNorm assessment is based solely on the variability of reference genes and does not take other factors into account. For example, we found that although RDN18 and RDN25 were quite stable, their amplification efficiencies were not equal to the amplification efficiencies of all target genes tested; thus, they may not be suitable as internal controls in our system. Therefore, when multiple reference genes are necessary, we believe that the combination of RDN5.8 plus UBC13 and/or PGK1 would be a better choice for quantitation of gene expression by RT-qPCR in C. glabrata following azole stimulation.


In this study, we evaluated 16 reference genes for potential use as internal controls for RT-qPCR analysis of gene expression in C. glabrata (Cg84u strain) following 2 hours of exposure to fluconazole at 200 μg/ml. To our knowledge, this is the first identification and validation of RDN5.8, UBC13, and PGK1 as the most suitable and stably expressed reference genes among the 16 reference genes tested. Therefore, we recommend the use of RDN5.8, UBC13, or PGK1 alone or the geometric mean of these genes as standards for normalization when analyzing differences in gene expression levels in C. glabrata during antifungal treatment. More specifically, RDN5.8 may be a more suitable reference gene for the analysis of expression for genes with lower azole-inducible mRNA levels, while UBC13 and PGK1 may be better internal controls for quantifying the expression of genes with higher azole-inducible mRNA levels in C. glabrata. In contrast, we demonstrated that 10 reference genes commonly used in published reports, including ACT1, GAPDH, PPIA, RPL13A, and TUB1, had significant differences in their expression upon azole challenge, and thus were not validated as good endogenous controls in this model. As a main conclusion, this study emphasizes the importance of evaluation studies for the selection of the most appropriate internal controls for each experimental model used for quantitative expression studies.

Availability and requirements

The hkgFinder software is currently available as an R source script and it can be downloaded for free from the NIAID Exon website [] with a sample data set and complete instructions. It requires installation of R 2.11 or higher on computers using the Microsoft Windows operating system for complete compatibility with its graphic user interface (GUI) elements. Experienced R users should be able to use hkgFinder on an Apple Macintosh or Linux/Unix operating system with some reasonable adjustments. An hkgFinder webtool should be available soon on the NIAID Exon website listed above. The geNorm software is available from Biogazelle [] with a free 15-day trial download as a part of the qBase PLUS software system. The geNorm software has now been integrated into the qBase PLUS software, where calculation of relative quantities and geNorm analysis are combined in a single program to speed up analysis. Presently, many manual pre-calculations are not needed, and cross point (CP) values from RT-qPCR can now be directly used for the gene stability analysis using the qBase PLUS software. The qBase PLUS software currently requires Microsoft Windows XP or above or Apple Mac OS X 10.6 (Snow Leopard) with Java 1.6 or later. Support for Linux is also available, but no requirements are listed on the manufacturer’s website. The BestKeeper software [] is available for free download. Please note that the software requires a password generated by an automatic email response from or It requires the Microsoft Windows operating system and Microsoft Excel, but specific versions are not listed on the manufacturer’s website. The NormFinder software is available for free download from the manufacturer’s website []. It requires Microsoft Windows operating system and Microsoft Excel 2003 or above.



beta-actin (β-actin)


Base pair


Complementary DNA


Candida glabrata


Crossing point


Delta cycle threshold


Coefficient of variation


Elongation factor 1α


fold change


Glyceraldehyde 3-phosphate dehydrogenase


GeNorm stability parameter


Minimum inhibitory concentration


Messenger RNA


Normalization factor


Open reading frame


Polymerase chain reaction


Phosphoglycerate kinase


Peptidylpropyl isomerase A (cyclophilin A)


5.8 S ribosomal RNA


18S ribosomal RNA


25S ribosomal RNA


Ribosomal protein, large, 2A


Ribosomal protein, large, 10


Ribosomal protein, large, 13A


Reverse transcription


Real-time quantitative reverse-transcription PCR


Standard deviation


Succinate dehydrogenase complex, subunit A


Melting temperature


Alpha-tubulin (α-tubulin)


Ubiquitin C


geNorm pairwise variation.


  1. Tsai HF, Sammons LR, Zhang X, Suffis SD, Su Q, Myers TG, Marr KA, Bennett JE: Microarray and molecular analyses of the azole resistance mechanism in Candida glabrata oropharyngeal isolates. Antimicrob Agents Chemother. 2010, 54: 3308-3317. 10.1128/AAC.00535-10

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  2. Bustin SA, Benes V, Nolan T, Pfaffl MW: Quantitative real-time RT-PCR–a perspective. J Mol Endocrinol. 2005, 34: 597-601. 10.1677/jme.1.01755

    Article  CAS  PubMed  Google Scholar 

  3. Kubista M, Andrade JM, Bengtsson M, Forootan A, Jonak J, Lind K, Sindelka R, Sjoback R, Sjogreen B, Strombom L, Stahlberg A, Zoric N: The real-time polymerase chain reaction. Mol Aspects Med. 2006, 27: 95-125. 10.1016/j.mam.2005.12.007

    Article  CAS  PubMed  Google Scholar 

  4. VanGuilder HD, Vrana KE, Freeman WM: Twenty-five years of quantitative PCR for gene expression analysis. Biotechniques. 2008, 44: 619-626. 10.2144/000112776

    Article  CAS  PubMed  Google Scholar 

  5. Wong ML, Medrano JF: Real-time PCR for mRNA quantitation. Biotechniques. 2005, 39: 75-85. 10.2144/05391RV01

    Article  CAS  PubMed  Google Scholar 

  6. Bustin SA: Absolute quantification of mRNA using real-time reverse transcription polymerase chain reaction assays. J Mol Endocrinol. 2000, 25: 169-193. 10.1677/jme.0.0250169

    Article  CAS  PubMed  Google Scholar 

  7. Giulietti A, Overbergh L, Valckx D, Decallonne B, Bouillon R, Mathieu C: An overview of real-time quantitative PCR: applications to quantify cytokine gene expression. Methods. 2001, 25: 386-401. 10.1006/meth.2001.1261

    Article  CAS  PubMed  Google Scholar 

  8. Radonic A, Thulke S, Mackay IM, Landt O, Siegert W, Nitsche A: Guideline to reference gene selection for quantitative real-time PCR. Biochem Biophys Res Commun. 2004, 313: 856-862. 10.1016/j.bbrc.2003.11.177

    Article  CAS  PubMed  Google Scholar 

  9. 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-,

    Article  PubMed  PubMed Central  Google Scholar 

  10. Thellin O, Zorzi W, Lakaye B, De Borman B, Coumans B, Hennen G, Grisar T, Igout A, Heinen E: Housekeeping genes as internal standards: use and limits. J Biotechnol. 1999, 75: 291-295. 10.1016/S0168-1656(99)00163-7

    Article  CAS  PubMed  Google Scholar 

  11. Selvey S, Thompson EW, Matthaei K, Lea RA, Irving MG, Griffiths LR: Beta-actin–an unsuitable internal control for RT-PCR. Mol Cell Probes. 2001, 15: 307-311. 10.1006/mcpr.2001.0376

    Article  CAS  PubMed  Google Scholar 

  12. Glare EM, Divjak M, Bailey MJ, Walters EH: beta-Actin and GAPDH housekeeping gene expression in asthmatic airways is variable and not suitable for normalising mRNA levels. Thorax. 2002, 57: 765-770. 10.1136/thorax.57.9.765

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  13. Barber RD, Harmer DW, Coleman RA, Clark BJ: GAPDH as a housekeeping gene: analysis of GAPDH mRNA expression in a panel of 72 human tissues. Physiol Genomics. 2005, 21: 389-395. 10.1152/physiolgenomics.00025.2005

    Article  CAS  PubMed  Google Scholar 

  14. Suzuki T, Higgins PJ, Crawford DR: Control selection for RNA quantitation. Biotechniques. 2000, 29: 332-337.

    CAS  PubMed  Google Scholar 

  15. 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: 523-526. 10.1006/bbrc.1999.0815

    Article  CAS  PubMed  Google Scholar 

  16. Lee JH, Fitzgerald JB, Dimicco MA, Grodzinsky AJ: Mechanical injury of cartilage explants causes specific time-dependent changes in chondrocyte gene expression. Arthritis Rheum. 2005, 52: 2386-2395. 10.1002/art.21215

    Article  CAS  PubMed  Google Scholar 

  17. Ohl F, Jung M, Xu C, Stephan C, Rabien A, Burkhardt M, Nitsche A, Kristiansen G, Loening SA, Radonic A, Jung K: Gene expression studies in prostate cancer tissue: which reference gene should be selected for normalization?. J Mol Med (Berl). 2005, 83: 1014-1024. 10.1007/s00109-005-0703-z. 10.1007/s00109-005-0703-z

    Article  CAS  Google Scholar 

  18. Schmittgen TD, Zakrajsek BA: Effect of experimental treatment on housekeeping gene expression: validation by real-time, quantitative RT-PCR. J Biochem Biophys Methods. 2000, 46: 69-81. 10.1016/S0165-022X(00)00129-9

    Article  CAS  PubMed  Google Scholar 

  19. Dheda K, Huggett JF, Bustin SA, Johnson MA, Rook G, Zumla A: Validation of housekeeping genes for normalizing RNA expression in real-time PCR. Biotechniques. 2004, 37: 112-114. 116, 118-119

    CAS  PubMed  Google Scholar 

  20. Warrington JA, Nair A, Mahadevappa M, Tsyganskaya M: Comparison of human adult and fetal expression and identification of 535 housekeeping/maintenance genes. Physiol Genomics. 2000, 2: 143-147.

    CAS  PubMed  Google Scholar 

  21. Schmid H, Cohen CD, Henger A, Irrgang S, Schlondorff D, Kretzler M: Validation of endogenous controls for gene expression analysis in microdissected human renal biopsies. Kidney Int. 2003, 64: 356-360. 10.1046/j.1523-1755.2003.00074.x

    Article  CAS  PubMed  Google Scholar 

  22. Wu YY, Rees JL: Variation in epidermal housekeeping gene expression in different pathological states. Acta Derm Venereol. 2000, 80: 2-3. 10.1080/000155500750012397

    Article  CAS  PubMed  Google Scholar 

  23. Sturzenbaum SR, Kille P: Control genes in quantitative molecular biological techniques: the variability of invariance. Comp Biochem Physiol B Biochem Mol Biol. 2001, 130: 281-289. 10.1016/S1096-4959(01)00440-7

    Article  CAS  PubMed  Google Scholar 

  24. Raaijmakers MH, van Emst L, de Witte T, Mensink E, Raymakers RA: Quantitative assessment of gene expression in highly purified hematopoietic cells using real-time reverse transcriptase polymerase chain reaction. Exp Hematol. 2002, 30: 481-487. 10.1016/S0301-472X(02)00787-7

    Article  CAS  PubMed  Google Scholar 

  25. Solanas M, Moral R, Escrich E: Unsuitability of using ribosomal RNA as loading control for Northern blot analyses related to the imbalance between messenger and ribosomal RNA content in rat mammary tumors. Anal Biochem. 2001, 288: 99-102. 10.1006/abio.2000.4889

    Article  CAS  PubMed  Google Scholar 

  26. 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: 5245-5250., 10.1158/0008-5472.CAN-04-0496

    Article  CAS  PubMed  Google Scholar 

  27. Gu C, Chen S, Liu Z, Shan H, Luo H, Guan Z, Chen F: Reference gene selection for quantitative real-time PCR in chrysanthemum subjected to biotic and abiotic stress. Mol Biotechnol. 2011, 49: 192-197. 10.1007/s12033-011-9394-6

    Article  CAS  PubMed  Google Scholar 

  28. Huis R, Hawkins S, Neutelings G: Selection of reference genes for quantitative gene expression normalization in flax (Linum usitatissimum L.). BMC Plant Biol. 2010, 10: 71- 10.1186/1471-2229-10-71

    Article  PubMed  PubMed Central  Google Scholar 

  29. Ren S, Zhang F, Li C, Jia C, Li S, Xi H, Zhang H, Yang L, Wang Y: Selection of housekeeping genes for use in quantitative reverse transcription PCR assays on the murine cornea. Mol Vis. 2010, 16: 1076-1086.

    CAS  PubMed  PubMed Central  Google Scholar 

  30. Goidin D, Mamessier A, Staquet MJ, Schmitt D, Berthier-Vergnes O: Ribosomal 18 S 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. 10.1006/abio.2001.5171

    Article  CAS  PubMed  Google Scholar 

  31. Al-Bader MD, Al-Sarraf HA: Housekeeping gene expression during fetal brain development in the rat-validation by semi-quantitative RT-PCR. Brain Res Dev Brain Res. 2005, 156: 38-45.

    Article  PubMed  Google Scholar 

  32. Radonic A, Thulke S, Bae HG, Muller MA, Siegert W, Nitsche A: Reference gene selection for quantitative real-time PCR analysis in virus infected cells: SARS corona virus, Yellow fever virus, Human Herpesvirus-6. Camelpox virus and Cytomegalovirus infections. Virol J. 2005, 2: 7- 10.1186/1743-422X-2-7

    Article  PubMed  PubMed Central  Google Scholar 

  33. Neuvians TP, Gashaw I, Sauer CG, von Ostau C, Kliesch S, Bergmann M, Hacker A, Grobholz R: Standardization strategy for quantitative PCR in human seminoma and normal testis. J Biotechnol. 2005, 117: 163-171. 10.1016/j.jbiotec.2005.01.011

    Article  CAS  PubMed  Google Scholar 

  34. Vermitsky JP, Edlind TD: Azole resistance in Candida glabrata: coordinate upregulation of multidrug transporters and evidence for a Pdr1-like transcription factor. Antimicrob Agents Chemother. 2004, 48: 3773-3781. 10.1128/AAC.48.10.3773-3781.2004

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  35. Tsai HF, Krol AA, Sarti KE, Bennett JE: Candida glabrata PDR1, a transcriptional regulator of a pleiotropic drug resistance network, mediates azole resistance in clinical isolates and petite mutants. Antimicrob Agents Chemother. 2006, 50: 1384-1392. 10.1128/AAC.50.4.1384-1392.2006

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Bennett JE, Izumikawa K, Marr KA: Mechanism of increased fluconazole resistance in Candida glabrata during prophylaxis. Antimicrob Agents Chemother. 2004, 48: 1773-1777. 10.1128/AAC.48.5.1773-1777.2004

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. Henry KW, Nickels JT, Edlind TD: Upregulation of ERG genes in Candida species by azoles and other sterol biosynthesis inhibitors. Antimicrob Agents Chemother. 2000, 44: 2693-2700. 10.1128/AAC.44.10.2693-2700.2000

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  38. Smith WL, Edlind TD: Histone deacetylase inhibitors enhance Candida albicans sensitivity to azoles and related antifungals: correlation with reduction in CDR and ERG upregulation. Antimicrob Agents Chemother. 2002, 46: 3532-3539. 10.1128/AAC.46.11.3532-3539.2002

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  39. Espinel-Ingroff A, Kish CW, Kerkering TM, Fromtling RA, Bartizal K, Galgiani JN, Villareal K, Pfaller MA, Gerarden T, Rinaldi MG, et al.: Collaborative comparison of broth macrodilution and microdilution antifungal susceptibility tests. J Clin Microbiol. 1992, 30: 3138-3145.

    CAS  PubMed  PubMed Central  Google Scholar 

  40. Izumikawa K, Kakeya H, Tsai HF, Grimberg B, Bennett JE: Function of Candida glabrata ABC transporter gene, PDH1. Yeast. 2003, 20: 249-261. 10.1002/yea.962

    Article  CAS  PubMed  Google Scholar 

  41. Liu ZL, Palmquist DE, Ma M, Liu J, Alexander NJ: Application of a master equation for quantitative mRNA analysis using qRT-PCR. J Biotechnol. 2009, 143: 10-16. 10.1016/j.jbiotec.2009.06.006

    Article  CAS  PubMed  Google Scholar 

  42. Skinner J: hkgFinder: Identify housekeeping genes and compute differential expression-User's Manual October 2011. 2011,, Bioinformatics and Computational Biosciences Branch (BCBB), Office of Cyber Infrastructure and Computational Biology (OCICB), National Institute of Allergy and Infectious Disease (NIAID), National Institutes of Health (NIH), Bethesda, Maryland, USA,

    Google Scholar 

  43. Hellemans J, Mortier G, De Paepe A, Speleman F, Vandesompele J: qBase relative quantification framework and software for management and automated analysis of real-time quantitative PCR data. Genome Biol. 2007, 8: R19-, 10.1186/gb-2007-8-2-r19

    Article  PubMed  PubMed Central  Google Scholar 

  44. 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: 509-515.,

    Article  CAS  PubMed  Google Scholar 

  45. Yamada H, Chen D, Monstein HJ, Hakanson R: Effects of fasting on the expression of gastrin, cholecystokinin, and somatostatin genes and of various housekeeping genes in the pancreas and upper digestive tract of rats. Biochem Biophys Res Commun. 1997, 231: 835-838. 10.1006/bbrc.1997.6198

    Article  CAS  PubMed  Google Scholar 

  46. Carlyle WC, Toher CA, Vandervelde JR, McDonald KM, Homans DC, Cohn JN: Changes in beta-actin mRNA expression in remodeling canine myocardium. J Mol Cell Cardiol. 1996, 28: 53-63. 10.1006/jmcc.1996.0006

    Article  CAS  PubMed  Google Scholar 

  47. Foss DL, Baarsch MJ, Murtaugh MP: Regulation of hypoxanthine phosphoribosyltransferase, glyceraldehyde-3-phosphate dehydrogenase and beta-actin mRNA expression in porcine immune cells and tissues. Anim Biotechnol. 1998, 9: 67-78. 10.1080/10495399809525893

    Article  CAS  PubMed  Google Scholar 

  48. Blanquicett C, Johnson MR, Heslin M, Diasio RB: Housekeeping gene variability in normal and carcinomatous colorectal and liver tissues: applications in pharmacogenomic gene expression studies. Anal Biochem. 2002, 303: 209-214. 10.1006/abio.2001.5570

    Article  CAS  PubMed  Google Scholar 

  49. Bond BC, Virley DJ, Cairns NJ, Hunter AJ, Moore GB, Moss SJ, Mudge AW, Walsh FS, Jazin E, Preece P: The quantification of gene expression in an animal model of brain ischaemia using TaqMan real-time RT-PCR. Brain Res Mol Brain Res. 2002, 106: 101-116.

    Article  CAS  PubMed  Google Scholar 

  50. Abruzzo LV, Lee KY, Fuller A, Silverman A, Keating MJ, Medeiros LJ, Coombes KR: Validation of oligonucleotide microarray data using microfluidic low-density arrays: a new statistical method to normalize real-time RT-PCR data. Biotechniques. 2005, 38: 785-792. 10.2144/05385MT01

    Article  CAS  PubMed  Google Scholar 

  51. Burleigh SH: Relative quantitative RT-PCR to study the expression of plant nutrient transporters in arbuscular mycorrhizas. Plant Sci. 2001, 160: 899-904. 10.1016/S0168-9452(00)00460-X

    Article  CAS  PubMed  Google Scholar 

  52. Klok EJ, Wilson IW, Wilson D, Chapman SC, Ewing RM, Somerville SC, Peacock WJ, Dolferus R, Dennis ES: Expression profile analysis of the low-oxygen response in Arabidopsis root cultures. Plant Cell. 2002, 14: 2481-2494. 10.1105/tpc.004747

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  53. Fort P, Marty L, Piechaczyk M, el Sabrouty S, Dani C, Jeanteur P, Blanchard JM: Various rat adult tissues express only one major mRNA species from the glyceraldehyde-3-phosphate-dehydrogenase multigenic family. Nucleic Acids Res. 1985, 13: 1431-1442. 10.1093/nar/13.5.1431

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  54. Bhatia P, Taylor WR, Greenberg AH, Wright JA: Comparison of glyceraldehyde-3-phosphate dehydrogenase and 28 S-ribosomal RNA gene expression as RNA loading controls for northern blot analysis of cell lines of varying malignant potential. Anal Biochem. 1994, 216: 223-226. 10.1006/abio.1994.1028

    Article  CAS  PubMed  Google Scholar 

  55. Brunner AM, Yakovlev IA, Strauss SH: Validating internal controls for quantitative plant gene expression studies. BMC Plant Biol. 2004, 4: 14- 10.1186/1471-2229-4-14

    Article  PubMed  PubMed Central  Google Scholar 

  56. Anderson LE, Gatla N, Carol AA: Enzyme co-localization in pea leaf chloroplasts: glyceraldehyde-3-P dehydrogenase, triose-P isomerase, aldolase and sedoheptulose bisphosphatase. Photosynth Res. 2005, 83: 317-328. 10.1007/s11120-005-0790-2

    Article  CAS  PubMed  Google Scholar 

  57. Kimura Y, Tanaka K: Regulatory mechanisms involved in the control of ubiquitin homeostasis. J Biochem. 2010, 147: 793-798. 10.1093/jb/mvq044

    Article  CAS  PubMed  Google Scholar 

  58. Koletsky AJ, Harding MW, Handschumacher RE: Cyclophilin: distribution and variant properties in normal and neoplastic tissues. J Immunol. 1986, 137: 1054-1059.

    CAS  PubMed  Google Scholar 

  59. de Jonge HJ, Fehrmann RS, de Bont ES, Hofstra RM, Gerbens F, Kamps WA, de Vries EG, van der Zee AG, te Meerman GJ, ter Elst A: Evidence based selection of housekeeping genes. PLoS One. 2007, 2: e898- 10.1371/journal.pone.0000898

    Article  PubMed  PubMed Central  Google Scholar 

Download references


We sincerely thank Huei-Fung Tsai, Jason Noble, and Bryan Walker for useful discussions and gratefully acknowledge Jason Noble for helpful technical assistance. This research was supported by the Intramural Research Program of the National Institute of Allergy and Infectious Diseases, the National Institutes of Health.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Qingdi Quentin Li.

Additional information

Competing interests

The authors declare that they have no competing interests.

Authors’ contributions

QQL conceived of the project, conducted the studies, performed all the experimental procedures, carried out the analysis and interpretation of data, wrote the manuscript, and is the primary author of this paper. JS developed the hkgFinder software and technically helped with the use of other software packages in the present study. JEB participated in the design and coordination of the study and critically reviewed the manuscript. All authors have read and approved the final manuscript.

Electronic supplementary material

Authors’ original submitted files for images

Below are the links to the authors’ original submitted files for images.

Authors’ original file for figure 1

Authors’ original file for figure 2

Rights and permissions

This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Reprints and Permissions

About this article

Cite this article

Li, Q.Q., Skinner, J. & Bennett, J.E. Evaluation of reference genes for real-time quantitative PCR studies in Candida glabrata following azole treatment. BMC Molecular Biol 13, 22 (2012).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI:


  • Candida glabrata
  • Azole resistance gene
  • Fluconazole
  • hkgFinder
  • Housekeeping gene
  • Reference gene
  • RT-qPCR