Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Nov 1.
Published in final edited form as: J Pharmacol Toxicol Methods. 2021 Aug 30;112:107117. doi: 10.1016/j.vascn.2021.107117

Cytochrome P450 3A4 (CYP3A4) protein quantification using capillary western blot technology and total protein normalization

Joseph M Collins a, Danxin Wang a,*
PMCID: PMC8616831  NIHMSID: NIHMS1739658  PMID: 34474151

Abstract

The western blot (WB) is the predominate method for protein quantification, frequently used in pharmacological and toxicological studies. To control for technical variation, WB signals are normalized through immunodetection of an internal standard “house-keeping” gene or total protein quantification via staining of the same blot or a duplicate, sister blot. Increasing evidence suggests that house-keeping genes are subject to change after drug treatment or under disease states, causing protein quantification errors in WB. Recent advances in automated capillary-based WB technologies enable measurement of the protein of interest, internal standards, and total protein in a single capillary. Using this approach, we quantified cytochrome P450 3A4 (CYP3A4) across 179 liver samples and compared normalization by both β-actin and total protein to determine which better functions as an internal standard. CYP3A4 is responsible for metabolism of a wide array of xenobiotics and is known to exhibit large inter-person variation, making it a good candidate to evaluate protein quantification. We observed significant differences in β-actin protein levels between liver samples (~20-fold) and found better correlation between CYP3A4 protein and mRNA using total protein normalization than β-actin, indicating total protein normalization to be less error prone for estimation of CYP3A4. Furthermore, by using total protein normalization, we confirmed significant association between CYP3A4 protein expression and the functional CYP3A4 variant CYP3A4*22, which contains two linked SNPs rs35599367 and rs62471956. Our results indicate that the automatic capillary WB instrument combined with total protein normalization provides a high throughput and robust approach for protein quantification.

Keywords: Cytochrome P450, CYP3A4, western blot, protein normalization, polymorphisms

Introduction:

Quantitative protein expression analysis is one of the critical steps toward understanding molecular events associated with drug therapy, genetic variation, and the development and progression of disease. Although new technology has enabled large-scale proteomic analysis, simultaneously detecting thousands of proteins (Michaels & Wang, 2014; Vogel & Marcotte, 2012), Western Blot (WB) analysis is the predominate methodology for small scale protein detection and quantification and in genetic association studies where a specific protein is quantified in a large number of samples.

Traditional WB involves electrophoretic separation of proteins by their molecular weights, transference of the stratified proteins to a stable membrane, and then the use of antibodies to detect a particular protein of interest. To control for technical variation, such as sample loading differences, transfer efficiency, and detection/visualization, it is common for an internal standard to be measured in parallel. The relative expression of protein(s) of interest is then established through comparison to the internal standard. For this purpose, several “house-keeping” proteins are commonly used, for example: GAPDH, β-actin, or the α-/β-tubulins. These proteins are assumed to be highly and uniformly expressed in the majority of cells, resistant to environmental changes, and to be stable under various disease conditions, thereby making them a suitable internal standard. However, increasing evidence suggests that expression levels of house-keeping genes are not consistent across different tissues, within the same tissue under varying conditions, or even different portions of the same tissue (Dittmer & Dittmer, 2006; Eaton et al., 2013; Ferguson et al., 2005; Ghosh, Gilda, & Gomes, 2014; J. Lin & Redies, 2012). For example, the expression levels of β-actin vary up to 22-fold between different tissues from the same mouse and expression levels of β-actin and β-tubulin in spinal cord of spinal muscular atrophy mice are 4- to 11- fold lower than in normal mice, respectively (Eaton et al., 2013). Similarly, in human kidneys, GAPDH protein is 2.8-fold higher in tumors than in normal tissues, while β-tubulin protein was undetectable in one out of five tumor samples tested (Ferguson et al., 2005). These observations raise questions on whether a single ‘house-keeping’ gene is indeed an appropriate internal control for normalizing WB signals. Several studies have indicated that total protein normalization appears to be a better approach that more accurately estimates target protein expression (Eaton et al., 2013; Romero-Calvo et al., 2010). Recently, a novel capillary electrophoresis-based WB technology has emerged that automates traditional WB, and enables multiple detection channels including total protein quantification (e.g. the Jess instrument from ProteinSimple). The utility of this technology for protein analyses has been validated in protein lysates from cells and tissues (Chen et al., 2014; Sage et al., 2020). However, in quantitative protein analysis for a large number of samples, the performance of capillary electrophoresis-based WB and total protein normalization has not been evaluated.

Cytochrome P450s metabolize a wide variety of endogenous substrates and xenobiotics including steroid hormones, carcinogens, toxins and drugs. Of these, the cytochrome P450 3A4 (CYP3A4) is the most abundant and is responsible for metabolism of nearly 50% of clinically used drugs (Zanger & Schwab, 2013). There exists large inter-person variability in CYP3A4 expression, the causes of which are still unclear, that greatly affects drug exposure and treatment outcomes. Previous studies have identified cis-acting genetic polymorphisms (Wang, Guo, Wrighton, Cooke, & Sadee, 2011), trans-acting transcription factors (Wang, Lu, Rempala, & Sadee, 2019), and epigenetic factors (Kacevska et al., 2012) that affect CYP3A4 expression, but most of these results are based on CYP3A4 mRNA expression with unknown consequences on its protein levels.

The purpose of this study was to evaluate the utility of the capillary electrophoresis-based WB approach for the quantification of CYP3A4 in a large number of liver samples and to assess whether total protein or an internal standard (β-actin) performed better for protein normalization. Our results show after normalizing protein amount by total detectable protein, β-actin expression is variable across liver samples. Furthermore, mRNA-protein correlation and SNP association results indicate that total protein normalization is a more robust approach for CYP3A4 protein quantification in human liver samples.

Materials and Methods:

Materials:

The protein normalization assay module for Jess that includes all reagents necessary for running capillary WB, as well as secondary antibodies with different protein conjugations: horseradish peroxidase (HRP), near infrared red (NIR) or infrared red (IR) fluorescence, and the chemiluminescent substrate were purchased from ProteinSimple (San Jose, CA, USA). Mouse anti-CYP3A4 antibody (R&D MAB 9079, 1:20 dilution), rabbit anti-β-actin antibody (NB600-503, 1:12.5 dilution), rabbit anti-heat shock protein 60 antibody (anti-HSP60, R&D AF-18000-SP, 1:20 dilution) and mouse anti-β-actin (R&D MAB8929, 1:25 dilution) were purchased from Biotechne (CA, USA). RNA mini prep kit was from Zymos Research (CA, USA). Other reagents used for liver tissue lysate preparation were from Life Technologies (CA, USA). Purified GST-fusion CYP3A4 protein was purchased from Fisher Scientific (CA, USA).

Liver tissue samples:

A total of 179 human liver samples was obtained from The Cooperative Human Tissue Network (CHTN). The demographic information of donors is: age, 58 ± 17 y, 45% African Americans and 54% female. The University of Florida Biosafety Committee and IRB Committee approved the human tissue study.

Tissue preparation:

Total mRNA was prepared from liver samples using the RNA mini prep kit (Zymos Research, CA, USA) according to the manufacturer’s protocol. Liver tissue samples (~10 mg) were homogenized in 300 μl buffer containing 10 mM HEPES, pH 7.9, 25 mM Tris HCl, pH 8.0, 137 mM NaCl, 10% glycerol, 1 mM EDTA, 1mM PMSF and Roche EasyPack inhibitor cocktail, using a Bead Ruptor 4 (Omni International, GA, USA) (speed 4, 2 × 45 sec, rest 45 sec on ice in between). NP40 was added to the homogenates to a final concentration of 1% and incubated on ice for 10 min, followed by centrifugation at 12,000 × g for 10 min. The clear supernatant was transferred to a clean tube and total protein concentration was measured using the Bradford assay (ThermoFisher, CA, USA). Tissue lysates were aliquoted and stored at −80°C until use.

Capillary western blotting using Jess (ProteinSimple Inc, CA, USA):

Capillary WB analyses were performed using the ProteinSimple Jess system (ProteinSimple, CA, USA) according to the manufacturer’s protocol. Briefly, tissue lysates were diluted with 0.1x sample buffer to concentration of 1 mg/ml. Then, 4-parts of diluted samples were combined with 1-part 5x fluorescent master mix (containing 5x sample buffer, 5x fluorescent standard, and 200 mM DTT) and heated at 95°C for 5 min. The resulting denatured samples, blocking reagents, total protein quantification reagents, mouse anti-CYP3A4 antibody (1:20) and/or rabbit anti-β-actin antibody (1:12.5), HRP-, NIR- or IR-conjugated anti-mouse or anti-rabbit secondary antibody (1:20) and chemiluminescent substrate (for HRP-conjugated secondary antibody only) were dispensed into designated wells in the assay plate. A biotinylated ladder provided the molecular weight standard for each assay. After plate loading, the separation, electrophoresis, loading, total protein quantification, and immunodetection steps were conducted in the fully automated Jess system.

Quantitative mRNA analysis:

mRNA levels of CYP3A4 and seven transcription factors known to regulate CYP3A4 expression were measured with real-time PCR using the TaqMan assay (Thermo Fisher Scientific™) or SYBR Green with gene-specific primers (see Supplemental Table 1 for probe ID and primer sequences). Measurements were conducted on a Quantabio Q real-time PCR instrument (VWR, PA, USA). We also measured the expression levels of GAPDH, β-actin, and 18sRNA as internal controls for RT-PCR. The relative expression of each gene was calculated using the following formula: expression level of tested gene = antilog2(mean Ct value of internal control – Ct value of tested gene)*106. After Log10 transformation, the expression levels of all genes tested followed a normal distribution.

Genotyping:

The single nucleotide polymorphism of CYP3A4 (rs62471956) (Collins & Wang, 2020), which is in complete linkage disequilibrium (LD) with the reduced expression allele CYP3A4*22 (Wang et al., 2011; Wang & Sadee, 2016), was genotyped using the OpenArray genotyping platform (probe#ANH6F7J, QuanStudio 12K Flex System) according to the manufacture’s protocol (Life Technology, CA, USA).

Data analysis:

Data are expressed as mean ± SD. Statistical analysis was performed using Prism 9 (GraphPad Software, San Diego, CA, USA). Association between the CYP3A4 protein levels and genotypes was performed using the Minitab software. We used forward and backward stepwise regression to select the best set of predictors (sex, age, race, and the expression of several transcription factors) in the multiple linear regression models with cutoff p-value ≤0.05. Race, and the expression levels of NR1I3, NR1I2, ESR1 and HNF4A were included as covariates.

Results:

Jess performance.

The Jess instrument has one channel for total protein quantification and three channels for immunodetection: enhanced chemiluminescence (ECL), near infrared red (NIR), and infrared red (IR) fluorescence. To test the reproducibility of Jess, we used a reference sample obtained via pooling several liver tissue lysates to detect CYP3A4 (ECL channel), and a HeLa cell lysate to detect the heat shock protein (NIR channel) and β-actin (IR channel). The coefficient of variation (CV) within the same run (24 repetitions) for each channel was: ECL-16.7%, NIR-12.6%, and IR-7.2%, while the CV between twelve runs for each channel was: ECL-14.4%, NIR-16.3% and IR-17.5%. In addition, the CV between different runs for total protein quantification was 9.5%. These results indicated that the Jess immunoassay and total protein quantification are highly reproducible.

Total protein and β-actin normalization.

We used a dilution series of the CYP3A4 reference sample to test the relationship between the amount of sample loaded and immunodetection. CYP3A4, β-actin, and total protein were detected simultaneously using ECL, NIR, and total protein channels, respectively. We started with 2 μg liver tissue lysate, which had been predetermined to be within the CYP3A4 linear detection range. The pooled lysate was diluted with sample buffer to achieve 1x, 0.5x, 0.25x, 0.125x, and 0.0625x concentrations. There was a linear relationship between sample dilution and total protein measurement (r=0.984), as well as immunodetection for CYP3A4 (r=0.9254) and β-actin (r=0.9752) (Figure 1a & 1b). Furthermore, when sample dilution was corrected via relative total protein levels (% of 0.5x sample) in each capillary, the correlation coefficient improved (Figure 1c & 1d). As expected, after normalizing by total protein, the estimated amount of CYP3A4 protein in the reference sample was consistent across the dilution series (Figure 2a). Similarly, after total protein normalization, the β-actin signal was mostly consistent, with an exceptional decreased detection at the 1:16 dilution (Figure 2b), possibly due to β-actin approaching its detection limit at this dilution in NIR channel. Consequently, when CYP3A4 was normalized by β-actin instead of total protein, the decreased β-actin signal in the 1:16 dilution caused a two-fold increase in the estimated CYP3A4 amount (Figure 2c). This result indicated that different linear detection ranges between a protein of interest and an internal control, either due to different protein amounts or immunoassay module detection sensitivities, may cause artifacts in estimated protein expression. In contrast, measurement of total protein was more resistant to such variation in detection, indicating that total protein normalization can effectively correct for differences in sample handling and electrophoresis.

Figure 1.

Figure 1.

Open in a new tab

CYP3A4 and β-actin dilution series. A pooled liver lysate was serially diluted (1x, 0.5x, 0.25x, 0.125x, and 0.0625x) and signals for CYP3A4, β-actin, and total protein measured simultaneously using the ECL, NIR, and protein channels, respectively. The measured CYP3A4 (a &c) and β-actin (b & d) signals are plotted against their dilution factor (a & b) or measured relative total protein levels (%) (protein level in 0.5x sample set at 1) (c & d). Each point is the average of duplicates.

Figure 2.

Figure 2.

Open in a new tab

Corrected protein signal after normalization. Panels a & b. Total protein normalized CYP3A4 (a) or β-actin (b) protein levels in the pooled liver sample across various dilutions. Panel c. CYP3A4 protein levels after β-actin normalization. Mean ± SD, n=4.

Comparison of total protein and β-actin normalized CYP3A4 protein levels.

We measured CYP3A4, β-actin, and total protein in 179 liver samples using the NIR, IR and total protein channels, respectively. When normalized by β-actin, CYP3A4 protein signals followed a normal distribution with 89-fold inter-person variability and 2.7-fold interquartile range (IQR) (Figure 3a). In contrast, when normalized by total protein, inter-person variability of CYP3A4 reduced to 48-fold and IQR to 1.7-fold (Figure 3b). Comparison of these two normalization techniques showed only a moderate correlation in estimated CYP3A4 protein amount (r=0.659, P<0.0001) (Figure 4).

Figure 3.

Figure 3.

Open in a new tab

Histograms showing distribution of CYP3A4 protein levels in 179 human liver lysates after being normalized by β-actin (a) or total protein (b). Data were log10 transformed.

Figure 4.

Figure 4.

Open in a new tab

The correlation between β-actin and total protein normalized CYP3A4 protein levels across 179 human liver lysates.

Because the β-actin signal is presumed to be directly related to the total amount of protein, we also compared its correlation with the measured total protein. However, β-actin was only moderately correlated with total protein level (r=0.480, p<0.0001), indicating potential inter-person variability in β-actin levels. In support of this, after being normalized by total protein, β-actin levels varied 23-fold between samples (Figure 5a). To determine whether this variation was due to detection limitations with the IR channel, we repeated measurement of β-actin using the NIR channel. Again, we found moderate correlation between β-actin and total protein measurements (r=0.392, p<0.0001), and after total protein normalization, the β-actin levels showed normal distribution with 18-fold inter-person variability (Figure 5b). To validate this result, we re-extracted proteins from ten samples that had the highest and lowest levels of β-actin and measured β-actin using two different antibodies: one from rabbit (as above) and a different mouse monoclonal antibody. We found similar fold differences in total protein normalized β-actin levels in these samples using either the rabbit (21-fold) or mouse antibody (25-fold). Furthermore, the results are highly correlated with the previous measurements: rabbit antibody (r=0.979, P<0.0001), mouse antibody (r=0.916, P<0.0001) (Supplemental Figure 1a&b). These results indicate that our measurements of β-actin are robust and that there are large differences β-actin protein amounts between human liver samples.

Figure 5.

Figure 5.

Open in a new tab

Histograms showing the distribution of total protein normalized β-actin levels in 179 human liver lysates detected either with the IR (a) or NIR (b) channels.

Correlation between CYP3A4 mRNA and protein expression levels.

We next tested the correlation between CYP3A4 mRNA and estimated protein amounts to determine whether these two techniques for measuring expression agreed. CYP3A4 mRNA levels were normalized by β-actin, GAPDH, or 18sRNA. As shown in Table 1, the correlation between CYP3A4 mRNA and protein expression is highest when using β-actin for mRNA normalization and total protein for protein normalization (r=0.778, p<0.0001). Conversely, the correlation is lowest when using 18sRNA for mRNA normalization and β-actin for protein normalization (r=0.507, P<0.0001). Comparing the different mRNA normalization methods revealed that β-actin and GAPDH are similar (Table 1) and highly corelated (r=0.903, p<0.0001). However, the correlation of β-actin/18SRNA (r=0.580, p<0.0001) or GAPDH/18SRNA (r=0.509, p<0.0001) is lower. These results highlight that estimation of both CYP3A4 mRNA and protein levels are susceptible to the choice of normalization method, and that total protein is a better internal control for CYP3A4 protein quantification using capillary-based WB, while β-actin or GAPDH appear to be appropriate controls for RT-PCR.

Table 1.

Correlation between levels of CYP3A4 protein and mRNA using different normalization methods. Table shows the correlation coefficient r. The correlation between all pairs are significant with p<0.0001.

Protein normalization
mRNA normalization Total protein β-actin
β-actin 0.778 0.574
GAPDH 0.753 0.552
18sRNA 0.622 0.507
Open in a new tab

Association between CYP3A4 protein level and CYP3A4 polymorphisms.

The CYP3A4*22 (rs35599367) and r62471956 polymorphisms are known to cause reduced expression of CYP3A4 [12, 15, 16–19], and are in complete linkage disequilibrium (LD) in all populations in the 1000 genomes project. We therefore tested for the association between rs62471956 and CYP3A4 protein levels as a practical application of the two normalization methods. In agreement with previous results using CYP3A4 mRNA (Wang et al., 2011), after normalizing by total protein, CYP3A4 was significantly associated with the rs62471956 genotype after adjusting for covariates (mRNA expression levels of NR1I3, NR1I2, ESR1, HNF4A and race), with each variant allele A associating with 0.69-fold reduced protein level (p=0.028) (Figure 6a). However, when β-actin was instead used an internal control, the association between CYP3A4 protein amount and the rs62471956 genotype became insignificant (p=0.773) (Figure 6b). The loss of association was likely the result of incorrect CYP3A4 normalization due to inter-individual variability in β-actin, which ultimately masked the effect of the genotypes.

Figure 6.

Figure 6.

Open in a new tab

Association between the rs62471956 genotypes and CYP3A4 protein levels normalized with total protein (a) or β-actin (b). CYP3A4 protein levels were adjusted for race, and the mRNA expression levels of NR1I3, HNF4A, and ESR1. The box and horizontal lines show the 25% and 75% percentiles, and the whiskers show the minimum and maximum values. Stars indicate the outliers.

Estimation of the absolute amount of CYP3A4 protein in human liver.

To estimate the absolute amount of CYP3A4 protein in human liver samples, we generated a standard curve using a purified GST-fusion CYP3A4 protein (Figure 7a) and calculated CYP3A4 protein per mg total protein. Figure 7b shows the distribution of CYP3A4 protein levels across 179 liver samples. The average CYP3A4 protein concentration is 101 ± 45 pmol/mg (± SD), with a large range spanning from 6.9 pmol/mg to 257 pmol/mg. These values are similar to liver CYP3A4 protein levels measured using LC/MS/MC methods (Ohtsuki et al., 2012) and are consistent with results from a meta-analysis (Achour, Barber, & Rostami-Hodjegan, 2014), supporting that our capillary WB and total protein normalization approach accurately quantified CYP3A4 protein in human liver samples.

Figure 7.

Figure 7.

Open in a new tab

Panel a. Standard curve of GST CYP3A4 fusion protein. GST CYP3A4 fusion protein was detected with a mouse monoclonal antibody using capillary WB across a serial dilution. The measured CYP3A4 signal in the NIR channel are plotted against the amount of loaded GST-fusion CYP3A4 protein. Each point is the average of duplicates. Panel b. Histogram showing the distribution of CYP3A4 protein levels (pmol/mg protein) in 179 human liver lysates detected with NIR channel.

Discussion:

Traditional WB has been referred to as a semi-quantitative technique due to the multiple assay steps, the lack of linearity, and limited quantitative reproducibility, preventing accurate protein quantification. The new capillary WB technology has greater quantifiable linear range, sensitivity and stability; thus, it can be considered as a quantitative approach. In this study, we evaluated the performance of the automatic capillary WB instrument, Jess, in CYP3A4 protein quantification of 179 liver samples. Our results showed that Jess is highly reproducible and enables accurate and consistent CYP3A4 protein quantification for genetic association studies. Also, our results showed that the commonly used loading control β-actin displayed 18–23-fold inter-person variability in human liver samples. Overall, our results indicate that total protein normalization is the better approach to correct for variations in sample loading and inter-capillary differences in WB-based protein analysis.

Previous studies demonstrated the variable expression of β-actin in different tissues (Eaton et al., 2013; Ferguson et al., 2005) after drug treatment or in disease (Liu & Xu, 2006; Ruan & Lai, 2007). However, it is unknown to what extent its expression differs between different individuals in the liver. Taking advantage of the multiple channel detection capability of Jess, we measured CYP3A4, β-actin, and total protein simultaneously within the same sample. Our results show that different normalization methods can have great impacts on estimated CYP3A4 protein measurements. In particular, total protein normalization can accurately correct for unequal protein loading and avoid artifacts caused by different detection ranges between an internal control (β-actin) and a protein of interest (CYP3A4) (Figure 1). More importantly, the amount of β-actin protein in liver samples is not constant across individuals, showing 18–23-fold variability. As a result, when used to normalize CYP3A4 protein levels, β-actin inflated estimated inter-person variability compared to total protein normalization (89-fold vs 48-fold).

We also compared the CYP3A4 protein amount determined by the two protein normalization methods to the amount of CYP3A4 mRNA. Since previous studies have shown that the levels of CYP3A4 mRNA and protein are highly correlated (Y. S. Lin et al., 2002; Ohtsuki et al., 2012; Watanabe et al., 2004), we assumed that the more accurate protein normalization approach would also result in better correlation of CYP3A4 mRNA and protein. We found that total protein normalization results in better CYP3A4 mRNA protein correlation than β-actin, regardless of the chosen internal standard for mRNA normalization.

Quantification of mRNA via RT-PCR is commonly normalized by β-actin, GAPDH, or 18sRNA to account for total RNA input and reverse transcription efficiency. 18sRNA was reported to be a better internal control than β-actin or GAPDH for RT-PCR, because β-actin and GAPDH are subject to change under treatment or disease, while 18sRNA appears to be more stable (Bas, Forsberg, Hammarström, & Hammarström, 2004; Selvey et al., 2001). However, our result showed CYP3A4 mRNA-protein correlation is stronger when β-actin or GAPDH are used to normalize the RT-PCR data compared to 18sRNA (Table 1). This indicates that β-actin is an appropriate internal control for CYP3A4 mRNA normalization in liver samples. One of the key criteria for selection of an internal control gene in RT-PCR is that its expression levels are similar to the gene of interest (Kozera & Rapacz, 2013). In human liver samples, we found that the expression level of 18sRNA is much higher than CYP3A4 (>6770-fold, mean Ct value 11.9 vs 24.6), while the expression levels of both β-actin and GAPDH are more similar to CYP3A4 (< 10-fold, mean Ct values 21.1 and 21.3 for β-actin and GAPDH, respectively). Moreover, our results showed strong correlation between mRNA levels of β-actin and GAPDH, while there was only moderate correlation between 18sRNA and β-actin or GAPDH. Thus, 18sRNA may not be an appropriate internal control for CYP3A4 mRNA normalization in liver samples. It is worth mentioning that because β-actin is a good internal control for normalization of mRNA, but not for protein, that the observed variation in β-actin protein levels may be occurring either during or after translation, or that β-actin isoforms with different epitope sequences due to alternative splicing or mutations are not effectively measured with a canonical antibody.

Furthermore, using total protein normalization, we revealed significant association between the known genetic variant rs62471956/CYP3A4*22 and CYP3A4 protein level, which was missed with β-actin normalization. CYP3A4*22 (rs35599367) is a known functional variant that affects splicing of CYP3A4, thereby reducing expression of the full-length transcript and overall enzymatic activity [12,16], and has been associated with numerous traits related to CYP3A4 metabolism in clinical association studies (Elens, van Gelder, Hesselink, Haufroid, & van Schaik, 2013). CYP3A4*22 is in complete LD with rs62471956, a distal enhancer SNP that regulates CYP3A4 transcription [15]. Our results are consistent with a previous study showing association between CYP3A4*22 and CYP3A4 protein expression in liver microsomes using absolute protein quantification (Okubo et al., 2013). Altogether, our results indicate that total protein normalization is the better approach for CYP3A4 protein quantification, consistent with recent reports for many other proteins (Eaton et al., 2013; Ferguson et al., 2005). In addition, our results show capillary WB with total protein normalization allow absolute CYP3A4 protein quantification in human liver samples using a purified GST CYP3A4 fusion protein as a reference. Since the enzyme activities of cytochrome P450 enzymes are often measured as activity per mg protein, absolute protein quantification will enable comparison between CYP3A4 protein abundance and CYP3A4 enzyme activity.

In conclusion, our results indicate that the automatic capillary WB instrument Jess combined with total protein normalization provides a high throughput and robust approach for protein quantification in large number of samples. We found the approach to be timesaving, and both highly sensitive and reproducible.

Supplementary Material

1
Supp.Table 1

Funding:

This work was supported by National Institute of Health (R01 GM120396 and R35 GM140845 to DW). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Footnotes

Conflict of interest statement: Authors declare no conflict of interests.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Achour B, Barber J, & Rostami-Hodjegan A (2014). Expression of hepatic drug-metabolizing cytochrome p450 enzymes and their intercorrelations: a meta-analysis. Drug Metab Dispos, 42(8), 1349–1356. [DOI] [PubMed] [Google Scholar]
  2. Bas A, Forsberg G, Hammarström S, & Hammarström ML (2004). 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, 59(6), 566–573. [DOI] [PubMed] [Google Scholar]
  3. Chen H, Shen ZY, Xu W, Fan TY, Li J, Lu YF, et al. (2014). Expression of P450 and nuclear receptors in normal and end-stage Chinese livers. World J Gastroenterol, 20(26), 8681–8690. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Collins JM, & Wang D (2020). Cis-acting regulatory elements regulating CYP3A4 transcription in human liver. Pharmacogenet Genomics. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Dittmer A, & Dittmer J (2006). Beta-actin is not a reliable loading control in Western blot analysis. Electrophoresis, 27(14), 2844–2845. [DOI] [PubMed] [Google Scholar]
  6. Eaton SL, Roche SL, Llavero Hurtado M, Oldknow KJ, Farquharson C, Gillingwater TH, et al. (2013). Total protein analysis as a reliable loading control for quantitative fluorescent Western blotting. PLoS One, 8(8), e72457. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Elens L, van Gelder T, Hesselink DA, Haufroid V, & van Schaik RH (2013). CYP3A4*22: promising newly identified CYP3A4 variant allele for personalizing pharmacotherapy. Pharmacogenomics, 14(1), 47–62. [DOI] [PubMed] [Google Scholar]
  8. Ferguson RE, Carroll HP, Harris A, Maher ER, Selby PJ, & Banks RE (2005). Housekeeping proteins: a preliminary study illustrating some limitations as useful references in protein expression studies. Proteomics, 5(2), 566–571. [DOI] [PubMed] [Google Scholar]
  9. Ghosh R, Gilda JE, & Gomes AV (2014). The necessity of and strategies for improving confidence in the accuracy of western blots. Expert Rev Proteomics, 11(5), 549–560. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Kacevska M, Ivanov M, Wyss A, Kasela S, Milani L, Rane A, et al. (2012). DNA methylation dynamics in the hepatic CYP3A4 gene promoter. Biochimie, 94(11), 2338–2344. [DOI] [PubMed] [Google Scholar]
  11. Kozera B, & Rapacz M (2013). Reference genes in real-time PCR. J Appl Genet, 54(4), 391–406. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Lin J, & Redies C (2012). Histological evidence: housekeeping genes beta-actin and GAPDH are of limited value for normalization of gene expression. Dev Genes Evol, 222(6), 369–376. [DOI] [PubMed] [Google Scholar]
  13. Lin YS, Dowling AL, Quigley SD, Farin FM, Zhang J, Lamba J, et al. (2002). Co-regulation of CYP3A4 and CYP3A5 and contribution to hepatic and intestinal midazolam metabolism. Mol Pharmacol, 62(1), 162–172. [DOI] [PubMed] [Google Scholar]
  14. Liu NK, & Xu XM (2006). beta-tubulin is a more suitable internal control than beta-actin in western blot analysis of spinal cord tissues after traumatic injury. J Neurotrauma, 23(12), 1794–1801. [DOI] [PubMed] [Google Scholar]
  15. Michaels S, & Wang MZ (2014). The revised human liver cytochrome P450 “Pie”: absolute protein quantification of CYP4F and CYP3A enzymes using targeted quantitative proteomics. Drug Metab Dispos, 42(8), 1241–1251. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Ohtsuki S, Schaefer O, Kawakami H, Inoue T, Liehner S, Saito A, et al. (2012). Simultaneous absolute protein quantification of transporters, cytochromes P450, and UDP-glucuronosyltransferases as a novel approach for the characterization of individual human liver: comparison with mRNA levels and activities. Drug Metab Dispos, 40(1), 83–92. [DOI] [PubMed] [Google Scholar]
  17. Okubo M, Murayama N, Shimizu M, Shimada T, Guengerich FP, & Yamazaki H (2013). CYP3A4 intron 6 C>T polymorphism (CYP3A4*22) is associated with reduced CYP3A4 protein level and function in human liver microsomes. J Toxicol Sci, 38(3), 349–354. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Romero-Calvo I, Ocón B, Martínez-Moya P, Suárez MD, Zarzuelo A, Martínez-Augustin O, et al. (2010). Reversible Ponceau staining as a loading control alternative to actin in Western blots. Anal Biochem, 401(2), 318–320. [DOI] [PubMed] [Google Scholar]
  19. Ruan W, & Lai M (2007). Actin, a reliable marker of internal control? Clin Chim Acta, 385(1–2), 1–5. [DOI] [PubMed] [Google Scholar]
  20. Sage JM, Hall L, McVey A, Barohn RJ, Statland JM, Jawdat O, et al. (2020). Use of Capillary Electrophoresis Immunoassay to Search for Potential Biomarkers of Amyotrophic Lateral Sclerosis in Human Platelets. J Vis Exp(156). [DOI] [PubMed] [Google Scholar]
  21. Selvey S, Thompson EW, Matthaei K, Lea RA, Irving MG, & Griffiths LR (2001). Beta-actin--an unsuitable internal control for RT-PCR. Mol Cell Probes, 15(5), 307–311. [DOI] [PubMed] [Google Scholar]
  22. Vogel C, & Marcotte EM (2012). Insights into the regulation of protein abundance from proteomic and transcriptomic analyses. Nat Rev Genet, 13(4), 227–232. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Wang D, Guo Y, Wrighton SA, Cooke GE, & Sadee W (2011). Intronic polymorphism in CYP3A4 affects hepatic expression and response to statin drugs. Pharmacogenomics J, 11(4), 274–286. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Wang D, Lu R, Rempala G, & Sadee W (2019). Ligand-Free Estrogen Receptor alpha (ESR1) as master regulator for the expression of CYP3A4 and other cytochrome P450 enzymes in the human liver. Mol Pharmacol, 96(4), 430–440. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Wang D, & Sadee W (2016). CYP3A4 intronic SNP rs35599367 (CYP3A4*22) alters RNA splicing. Pharmacogenet Genomics, 26(1), 40–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Watanabe M, Kumai T, Matsumoto N, Tanaka M, Suzuki S, Satoh T, et al. (2004). Expression of CYP3A4 mRNA is correlated with CYP3A4 protein level and metabolic activity in human liver. J Pharmacol Sci, 94(4), 459–462. [DOI] [PubMed] [Google Scholar]
  27. Zanger UM, & Schwab M (2013). Cytochrome P450 enzymes in drug metabolism: regulation of gene expression, enzyme activities, and impact of genetic variation. Pharmacol Ther, 138(1), 103–141. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1739658-supplement-1.pdf (112.7KB, pdf)
Supp.Table 1
NIHMS1739658-supplement-Supp_Table_1.pdf (64.4KB, pdf)

RESOURCES