Identification of the Promoter of Human Carbonyl Reductase 3 (CBR3) and Impact of Common Promoter Polymorphisms on Hepatic CBR3 mRNA Expression

Abstract

Purpose

Recent studies suggest that polymorphisms in human carbonyl reductase 3 (CBR3) influence the pharmacodynamics of doxorubicin. First, we sough to identify the promoter of CBR3. Next, we examined whether two CBR3 promoter polymorphisms (CBR3 -725T>C and CBR3 -326T>A) dictate promoter activity and hepatic CBR3 mRNA levels.

Methods

The promoter activities of CBR3 reporter constructs were investigated in HepG2 and MCF-7 cells. CBR3 mRNA levels were documented in 95 liver samples from white (n = 62) and black (n = 33) donors. Genotype-phenotype correlation analyses were used to determine the impact of the CBR3 -725T>C and CBR3 -326T>A polymorphisms on hepatic CBR3 mRNA levels.

Results

We identified the promoter of human CBR3. Liver samples from black donors showed higher relative CBR3 mRNA levels than samples from whites (CBR3 mRNA_blacks = 3.0 ± 3.1 relative fold vs. CBR3 mRNA_whites = 1.6 ± 1.5 relative fold, p = 0.021). The variant -725C and -326A alleles did not modify the gene reporter activities of engineered CBR3 promoter constructs. In line, hepatic CBR3 mRNA levels were not associated with CBR3 -725T>C and CBR3 -326T>A genotype status.

Conclusions

These studies provide the first insights into the regulation and variable hepatic expression of polymorphic CBR3.

INTRODUCTION

In humans, there are three carbonyl reductases: carbonyl reductase 1 (CBR1), carbonyl reductase 3 (CBR3) and the recently identified carbonyl reductase 4 (CBR4) (1). Carbonyl reductase activity reduces the anticancer anthracyclines doxorubicin and daunorubicin into their corresponding C-13 alcohol metabolites doxorubicinol and daunorubicinol (2, 3). Anthracycline alcohol metabolites have diminished tumor cell killing activity and exert cardiotoxicity in some cancer patients (4). New studies suggest that CBR1 and CBR3 single nucleotide polymorphisms (SNPs) influence the variable pharmacokinetics and pharmacodynamics of doxorubicin. For example, Fan et al. pinpointed associations between a CBR3 non-synonymous SNP (CBR3 C4Y, G>A, rs8133052) and the pharmacokinetics of doxorubicin in Southeast Asian breast cancer patients (5). CBR3 C4Y genotype status was associated with doxorubicinol area under the concentration-time curve (AUC; GG vs. AA, p = 0.004), and with the AUC ratio of doxorubicinol/doxorubicin (GG vs. AA, p = 0.009). However, Lal et al., failed to detect associations between CBR3 polymorphisms and the pharmacokinetics of doxorubicin in 101 breast cancer patients; but the authors identified significant associations between CBR1 polymorphisms (e.g. CBR1 627C>T, rs20572 and CBR1 967G>A, rs9024) and doxorubicin’s clearance, exposure, and peak plasma concentration of doxorubicinol (6). Recently, we showed that a CBR3 non-synonymous SNP (CBR3 V244M, G>A, rs1056892) results in CBR3 protein isoforms with distinctive catalytic activities towards doxorubicin. Recombinant CBR3 V244 (G allele) synthesized 2.6-fold more doxorubicinol than CBR3 M244 (A allele; CBR3 V244: 8.3 ± 3.6 nmol doxorubicinol/h.mg vs. CBR3 M244: 3.2 ± 0.7 nmol doxorubicinol/h.mg, p = 0.010). In line, multivariate analyses on 30 cases of anthracycline-related congestive heart failure (CHF) and 115 matched controls pinpointed a trend towards an association between CBR3 V244M genotype status and the risk of anthracycline-related CHF among pediatric cancer survivors (GG vs. AA, odds ratio: 8.16, p = 0.056) (7).

Watanabe et al. mapped CBR3 to chromosome 21q22.1 during the annotation of genes potentially involved in the pathogenesis of Down syndrome (8). There is a paucity of reports focused on the functional characterization of the CBR3 gene promoter. Thus, our first aim was to identify the promoter of human CBR3 by performing gene-reporter experiments with constructs encompassing up to 1.3 kilobases upstream from the translation start site of CBR3. Our second aim was to determine the extent of variability in the hepatic expression of CBR3 mRNA by analyzing 95 liver samples from white (n = 62) and black (n = 33) donors. To further characterize the molecular bases of variable CBR3 expression, we investigated the impact of two common CBR3 promoter polymorphisms (CBR3 -725T>C, rs2239566 and CBR3 -326T>A, rs8132243) by performing gene reporter assays with engineered CBR3 promoter constructs in HepG2 and MCF-7 cells. These studies were complemented by analyzing whether CBR3 -725T>C and CBR3 -326T>A genotype status correlate with hepatic CBR3 mRNA levels.

MATERIALS AND METHODS

Human liver DNA and RNA samples

The Institutional Review Board of the State University of New York at Buffalo approved this research. Human liver tissue from cadaveric donors (whites, n = 62; and black, n = 33) was processed through Dr. Mary Relling’s laboratory at St. Jude Children’s Research Hospital, and was provided by the liver Tissue Cell Distribution System, which was funded by NIH Contract #N01-DK-7-0004/HHSN267200700004C. Liver DNA, RNA and cytosols were isolated by following standardized procedures as described (9, 10). In most cases, no information was available regarding the exact geographical origin of the donors, and no information was available on the ethnicity of the donor's parents. Correlation between self-reported ethnicity (i.e. black and white) and geographical ancestry (i.e. African and European) was determined by genotyping 176 ancestry informative markers as described (10).

CBR3 reporter gene studies

DNA sequence upstream (1297 bp) from the translation start codon of CBR3 (ATG) was amplified by PCR from human DNA sample HD17249 (Coriell Institute for Medical Research, Camden, NJ). The amplicon was cloned into a pGL3 basic firefly luciferase vector (Promega, Madison, WI), and six 5’ progressive deletion constructs were made by PCR. The _-1297CBR3 reporter construct was modified with the QuickChange site-directed mutagenesis kit (Stratagene, La Jolla, CA) to generate two single nucleotide variants (-725C/-326T and -725T/-326A), as well as the double nucleotide variant -725C/-326A. The identities of each construct and the absence of cloning artifacts were verified by direct sequencing.

Cultures of HepG2 and MCF-7 cells were co-transfected with CBR3 reporter constructs (2.0 μg) and the pRLTK plasmid (0.3 μg for HepG2 cells and 20 ng for MCF-7 cells; Promega). The pRLTK renilla luciferase plasmid was used to correct for differences in transfection efficiencies. The pGL3-basic empty vector was transfected into control cultures to correct for background luciferase activity. Luciferase reporter gene activities were determined 48 h after co-transfections with the Dual-Luciferase Reporter Assay System (Promega). Light intensities were measured in a Synergy HT luminometer (BioTek, Winooski, VT). Corrected luciferase activity values were expressed relative to the activity of construct _-1297CBR3 which was set arbitrarily as 100%.

CBR3 -725T>C and CBR3 -326T>A genotyping

CBR3 -725T>C (rs2239566) and CBR3 -326T>A (rs8132243) were genotyped by direct sequencing after PCR amplification with the following primers: 5’-ATGTCATGGTTCTCCCCAAA-3’ (forward) and 5’-CTGTGCCCTCTGTTCCAACT-3’ (reverse). Trace files were analyzed with the PolyPhred package (http://lpgws.nci.nih.gov/perl/snp/snp_cgi.pl).

Hepatic CBR3 mRNA expression

Total RNA from liver (100 ng) was reverse-transcribed and amplified by using one-step QuantiTect SYBR Green RT-PCR kits (Qiagen, Valencia, CA) with the following primers: 5’-GCTTCCACCAACTGGACATC-3’ (forward) and 5’-GGGCATTGGATCATCACTCT-3’ (reverse). RT-PCR reaction mixtures were incubated in a MX3050P thermal cycler equipped with proprietary software for data analysis (MxPro v3.00, Stratagene, La Jolla, CA). Relative CBR3 mRNA levels were determined by the comparative quantitation method with individual β-actin mRNA levels as normalizers (11, 12). Cycling parameters for the amplifications in parallel of CBR3 and β-actin mRNAs were: 50°C for 30 min (reversetranscription), 95°C for 10 min (Taq polymerase activation);40 cycles of 95°C for 15 s (denaturation), 54°C for30 s (annealing), 72°C for 30 s (extension), and 84°Cfor 30 s (fluorescence collection). Melting curve analyses demonstrated a unique PCR amplification product. Amplification efficiencies for CBR3 and β-actin mRNAs were comparable. Standard curves for both mRNAs were run in parallel to ensure accurate quantifications (20-fold dynamic range, r² ≥0.96). Experimental samples and standards for calibration curves were analyzed in quadruplicate. The relative amount of CBR3 mRNA in each liver sample was automatically calculated with the comparative quantitation algorithm by using individual β-actin mRNA levels as normalizers. CBR3 mRNA values were expressed relative to the normalized CBR3 mRNA content of liver sample 237. Liver sample 237 was randomly selected for data normalization (11–13).

Statistical analysis

Data are presented as mean ± standard deviation. Student’s T test and one-way ANOVA were used to compare relative luciferase activities of reporter gene constructs. The normality of data sets was analyzed with the Kolmogorov-Smirnov test. The Mann-Whitney and Kruskal-Wallis tests were used to compare means from data sets with non-normal distributions. The Chi-square test was used to compare genotype distributions. Correlation analyses were performed with Spearman’s test. Differences were considered to be significant at p<0.05. Analyses were performed with SPSS version 15.0 (SPSS, Inc., Chicago, IL), and GraphPad Prism version 5.00 (GraphPad Software Inc, La Jolla, CA). Linkage disequilibrium analyses were performed with Haploview (version 4.1).

RESULTS

Annotation and functional analysis of the human CBR3 promoter

A fragment of 1297 bp upstream the translation initiation codon (ATG) of CBR3 was cloned into a pGL3 basic firefly luciferase vector to perform gene-reporter studies. Direct sequencing of the insert revealed 100% of identity with a segment of nucleotide sequence from locus NT_011512 (Homo sapiens genomic DNA, chromosome 21q, Entrez Nucleotide Database). The proximal promoter of CBR3 has the configuration of a typical CpG island promoter with an initiator element TTAGTC [Inr, Py-Py-A+1-N-T/A-Py-Py], several conserved motifs for the Sp1 transcription factor (SP1), a proximal GC-box, no TATA box and no core promoter element (DPE, A/G₊28-G-A/T-C/T-G/A/C; Figure 1). The CpG island encompasses 361 bp of 5’-flanking region sequence and extends 839 bp downstream of the ATG codon (14, 15). The Database of Transcriptional Start Sites (DBTSS, release 6.0.1) reports several transcription start sites (TSS) for CBR3 (e.g. -95, -107, -108, -228 and -281 bp). The TSS at -107 and -108 bp contain the Inr element. Further screening of the -1297 bp region with the web tools TESS and AliBaba2.1 revealed potential consensusmotifs for a number of transcription factors including octamer binding transcription factor 1 (OCT-1 or POU2F1), activator protein 1 (AP1 or JUN), and additional SP1 sites (Figure 1).

Figure 1.

Annotated sequence from the 5’-flanking region of human CBR3. The initiator element (Inr) is highlighted. Putative transcription factor binding sites are underlined. The location of primers is enclosed in boxes. NF-KappaB, nuclear factor kappa B; SP1, Sp1 transcription factor; HNF-3, hepatic nuclear factor 3; YY1, Ying Yang 1; GATA1, GATA binding protein 1 (globin transcription factor); HNF-1, hepatic nuclear factor 1; ER, estrogen receptor; OCT-1, octamer binding transcription factor; AP1, activator protein 1.

Reporter gene experiments demonstrated that the -1297 bp region exerts significant promoter activity in HepG2 and MCF-7 cells. Similar experiments with 5’ progressive deletion constructs pinpointed areas on the CBR3 promoter that may regulate transcription in a cell-dependent manner (Figure 2). For example, deletion of the -1297/-1148 bp region resulted in a construct (_-1148CBR3) with relatively diminished gene reporter activity in HepG2 cells (_-1297CBR3 vs. _-1148CBR3 p<0.001). In contrast, _-1148CBR3 showed 49% higher promoter activity in MCF-7 cells than the reference _-1297CBR3 construct (_-1148CBR3 vs. _-1297CBR3, p<0.001). Similarly, deletion of the -935/-726 bp fragment (_-726CBR3) did not modify luciferase activity in HepG2 cells, but reduced gene reporter activity by 70% in MCF-7 cells (p<0.001). Furthermore, the _-313CBR3 construct exhibited the highest and lowest promoter activities in HepG2 and MCF-7 cells, respectively. Experiments also showed that the -313/-172 bp segment is essential to sustain CBR3 promoter activity in both cell lines because of the negligible gene reporter activity of the _-172CBR3 construct.

Figure 2.

Functional analysis of human CBR3 promoter constructs in HepG2 cells (A) and MCF-7 cells (B). Panels show schematic representations of each CBR3 promoter construct (left) and its corresponding luciferase activity from gene reporter experiments (right). Corrected luciferase activity values were expressed relative to the activity of construct _-1297CBR3 which was set arbitrarily as 100%. Each value represents the mean ± SD of six independent experiments (*** p<0.001, ** p<0.01).

CBR3 mRNA expression in livers from black and white donors

The hepatic expression of CBR3 mRNA varied widely in samples from blacks and whites (CBR3 mRNA_blacks range: 0.4 – 12.2 relative fold, 95% CI of mean: 2.0 – 4.1 relative fold; CBR3 mRNA_whites range: 0.1 – 7.6 relative fold, 95% CI of mean: 1.2 – 1.9 relative fold). In both groups, relative CBR3 mRNA levels did not correlate with the donors’ age (blacks: r_s = −0.106, p = 0.605; whites: r_s = −0.147, p = 0.259). On average, samples from blacks showed 1.9-fold higher relative CBR3 mRNA levels than samples from whites (CBR3 mRNA_blacks: 3.0 ± 3.1 relative fold vs. CBR3 mRNA_whites: 1.6± 1.5 relative fold; Mann-Whitney test, p = 0.021; Figure 3). Differences between group means remained significant after the elimination of outlier values located above and/or below the 5^th and 95^th percentiles (blacks, n = 2 outliers, and whites, n = 6 outliers; Mann-Whitney test, p = 0.010; Figure 3).

Figure 3.

Hepatic CBR3 mRNA expression in samples from black (n = 33) and white (n = 62) donors. Relative CBR3 mRNA levels were determined with the comparative quantitation method (see text for details). The whiskers indicate the 5^th and 95^th percentiles. Horizontal lines indicate the medians. Black circles indicate outliers. Samples and standards for calibration curves were analyzed in quadruplicate. (CBR3 mRNA_blacks vs. CBR3 mRNA_whites, Mann-Whitney test, p = 0.021).

Common single nucleotide polymorphisms on the CBR3 promoter

The functional impact of the CBR3 -326T>A and CBR3 -725T>C polymorphisms was first investigated by introducing the variant -326A and -725C alleles into the _-1297CBR3 reporter construct. In HepG2 cells, the variant -326A and -725C alleles did not modify gene reporter activities (Figure 4). In MCF-7 cells, introduction of the -725C variant resulted in a modest 20% decrease in the gene reporter activity of _-1297CBR3 (p<0.001), whereas the double -326A/-725C variant increased gene reporter activity by 36% (p<0.001; Figure 4).

Figure 4.

Impact of CBR3 -725T>C and CBR3 -326T>A on the gene reporter activity of _{- 1297}CBR3. In both cell lines, luciferase activities are expressed relative to the activity of the “wild type” _-1297CBR3 construct (-725T/-326T), which was set arbitrarily as 100%. Each value represents the mean ± SD of six independent experiments (*** p<0.001).

We also examined whether CBR3 -326T>A and CBR3 -725T>C genotype status dictate hepatic CBR3 mRNA levels. The variant CBR3 -326T>A and CBR3 -725T>C alleles were relatively common in liver DNA samples from blacks and whites, and genotype distributions were in Hardy-Weinberg equilibrium in both groups (Table 1). CBR3 -725T>C and CBR3 -326T>A genotypes were in linkage disequilibrium (LD) in whites (D’ = 1.0, r² = 0.707) and blacks (D’ = 1.0, r² = 0.512). Correlation analysis failed to demonstrate significant associations between CBR3 -326T>A and CBR3 -725T>C genotype status and relative hepatic CBR3 mRNA levels in samples from blacks and whites (Figure 5).

Table 1.

CBR3 -326T>A and -725T>C genotype distributions in DNA samples from white and black liver donors.

SNP	Whites (n= 62)		Blacks (n=33)
	Genotypes	p and q*	Genotypes	p and q*
-725T>C	TT:TC:CC = 50:9:3 (80.6%:14.5%:4.8%)	0.879: 0.121	TT:TC:CC = 20:12:1 (60.6%:36.4%:3.0%)	0.788: 0.212
-326T>A	TT:TA:AA = 53:7:2 (85.5%:11.3%:3.2%)	0.911: 0.089	TT:TA = 25:8 (75.8%:24.2%)	0.879: 0.121

^*p denotes the major allele and q denotes the minor allele

Figure 5.

Impact of CBR3 -725T>C and CBR3 -326T>A genotypes on the expression of CBR3 mRNA in liver samples from white (top) and black (bottom) donors. Each circle represents individual liver samples (quadruplicate measurements). Horizontal lines indicate group means.

DISCUSSION

Recent studies suggest that CBR1 and CBR3 polymorphisms impact on the pharmacodynamics of doxorubicin in various cancer settings (5–7). The biological role of CBR3 remains to be elucidated, and the functional annotation of the promoter is a necessary step towards the understanding of the bases that govern variable CBR3 expression. Here, we identified the promoter of CBR3, and examined the impact of two common promoter polymorphisms on the expression of hepatic CBR3 mRNA. First, our results indicate that the region encompassing up to 1.3 kb upstream from the translation start site of CBR3 exerts significant gene promoter activity in two different cellular contexts. Experiments with progressive deletion constructs pinpointed regions on the CBR3 promoter that may regulate transcription in a cell specific manner. For example, the -539/-313 bp segment appears to bear regulatory element/s with contrasting actions in HepG2 and MCF-7 cells (Figure 2). The experiments also showed that the -313/-172 bp segment is essential for promoter activity in both cell lines. The lack of activity of the _-172CBR3 construct is consistent with the location of two major TSS at -228 and -281 bp, respectively. The -313/-172 bp region contains a GC box at -217 bp with an embedded consensus sequence for the binding of SP1 (Figure 1). SP1 activates the transcription of several TATA-less genes, and it will be of interest to test whether the SP1 consensus is key to the constitutive activity of the CBR3 promoter (16). Recently, we have shown that two functional xenobiotic response elements (XRE) on the promoter of the homologous CBR1 activate transcription in response to ligands of the aryl hydrocarbon receptor (AHR)(15). DNA sequence comparisons of up to 1.1 kb by global alignment positioning showed limited identity and similarity (~40%) between the proximal promoter regions of CBR1 and CBR3. In addition, computer assisted searches failed to identify conserved XREs on the CBR3 promoter (6 kb screened). This finding suggests a potential functional distinction between the mechanisms that control the transcription of human CBR1 and CBR3, respectively.

Our second aim was to document CBR3 mRNA expression in liver samples from black and white donors. Our data suggest that hepatic CBR3 mRNA levels are higher in samples from black donors (Figure 3). Recent analysis of ancestry informative markers (AIM) on this set of liver samples demonstrated excellent concordance between the donors’ ethnicity (i.e. black and white) and geographical ancestry (i.e. African and European) (10). Thus, the observed differences may in fact reflect distinctive patterns of CBR3 mRNA expression among individuals with different geographical ancestries, and further confirmatory studies are warranted. It will be interesting to examine whether differences in CBR3 mRNA expression translate into distinct CBR3 protein levels in livers from black and white donors. Biochemical studies have shown that the intracardiac formation of anthracycline alcohol metabolites is relevant to the pathogenesis of anthracycline-related cardiotoxicity (17–19). On the other hand, epidemiological studies in pediatric cancer survivors showed that black ethnicity constitutes a risk factor for the development of anthracycline related cardiotoxicity (20, 21). Thus, it is possible to speculate that potential interethnic differences in the cardiac expression of CBR3 may influence the risk of anthracycline-related cardiotoxicity.

The experiments with engineered reporter constructs in breast cancer-derived MCF-7 cells showed that the double CBR3 -326A/-725C promoter variant resulted in a modest increase (36%) in gene reporter activity (Figure 4). The study by Fan et al. reported that CBR3 C4Y (CBR3 11 G>A) genotype status was associated with basal CBR3 mRNA levels in 66 breast tumor samples (GG vs. AA, p = 0.001) (5). Thus, it will be of interest to test whether specific CBR3 genotype combinations (i.e. -326A/-725C/11G) are associated with distinctive CBR3 mRNA expression levels in breast tumors. On the other hand, we found no associations between CBR3 -326T>A and CBR3 -725T>C genotype status and CBR3 mRNA expression in liver samples from black and white donors (Figure 5). In agreement, the variant CBR3 -326A and CBR3 -725C alleles did not significantly modify the gene reporter activities of engineered CBR3 promoter constructs in HepG2 cells (Figure 4). Thus, our data do not support a significant role for the common CBR3 -326T>A and CBR3 -725T>C polymorphisms in dictating hepatic CBR3 mRNA levels.

This study describes the first insights into the transcriptional regulation of CBR3 and provides a platform to further elucidate the molecular bases that govern variable CBR3 expression in humans.

Abbreviations

CBR3

Carbonyl Reductase 3

CBR1

Carbonyl reductase 1

CBR4

Carbonyl reductase 4

PCR

polymerase chain reaction

SNP

single nucleotide polymorphism

CHF

congestive heart failure

CI

confidence interval

SP1

Sp1 transcription factor

OCT-1 or POU2F1

octamer binding transcription factor 1

AP1 or JUN

activator protein 1

XRE

xenobiotic response elements

AHR

aryl hydrocarbon receptor

AIM

ancestry informative markers