Differentiation‐associated urothelial cytochrome P450 oxidoreductase predicates the xenobiotic‐metabolizing activity of “luminal” muscle‐invasive bladder cancers

Extra‐hepatic metabolism of xenobiotics by epithelial tissues has evolved as a self‐defence mechanism but has potential to contribute to the local activation of carcinogens. Bladder epithelium (urothelium) is bathed in excreted urinary toxicants and pro‐carcinogens. This study reveals how differentiation affects cytochrome P450 (CYP) activity and the role of NADPH:P450 oxidoreductase (POR). CYP1A1 and CYP1B1 transcripts were inducible in normal human urothelial (NHU) cells maintained in both undifferentiated and functional barrier‐forming differentiated states in vitro. However, ethoxyresorufin O‐deethylation (EROD) activity, the generation of reactive BaP metabolites and BaP‐DNA adducts, were predominantly detected in differentiated NHU cell cultures. This gain‐of‐function was attributable to the expression of POR, an essential electron donor for all CYPs, which was significantly upregulated as part of urothelial differentiation. Immunohistology of muscle‐invasive bladder cancer (MIBC) revealed significant overall suppression of POR expression. Stratification of MIBC biopsies into “luminal” and “basal” groups, based on GATA3 and cytokeratin 5/6 labeling, showed POR over‐expression by a subgroup of the differentiated luminal tumors. In bladder cancer cell lines, CYP1‐activity was undetectable/low in basal PORlo T24 and SCaBER cells and higher in the luminal POR over‐expressing RT4 and RT112 cells than in differentiated NHU cells, indicating that CYP‐function is related to differentiation status in bladder cancers. This study establishes POR as a predictive biomarker of metabolic potential. This has implications in bladder carcinogenesis for the hepatic versus local activation of carcinogens and as a functional predictor of the potential for MIBC to respond to prodrug therapies.


Western Blotting
Whole cell lysates were obtained from cultures by in situ lysis in sodium dodecyl sulphate Detection of human CYP1A1 and GAPDH protein was performed as described [49] and not according to the method above. CYP1A1 antibody raised in rabbits against purified human recombinant CYP1A1 was a generous gift from Prof F. Peter Guengerich (Vanderbilt University, USA) and used at a dilution of 1:4000.
All Densitometry was performed using Image Studio Lite version 5.0 (LI-COR Biosciences).

Reverse Transcribed -quantitative Polymerase Chain Reaction (RT-qPCR)
The relative abundance of selected transcripts was assessed using the following forward and reverse primers (all 5'-3') to amplify CYP1A1 (ACCAAGAACTGCTTAGCCTAGTCAA and GGTTGGGTAGGTAGCGAAGAATAG), CYP1B1 (CCAGCTTTGTGCCTGTCACTAT and GGGAATGTGGTAGCCCAAGA), POR (AGCATGACGGACATGATTCTGT and TCAATGTCTGAATTTTGGTGAACTC) and GAPDH (CAAGGTCATCCATGACAACTTTG and GGGCCATCCACAGTCTTCTG). Primers were optimised to give a linear response over a 1,000 fold dilution range and a single product as characterised by a single peak in the dissociation curve. Amplification was monitored using SYBR Green dye on an ABI7300 machine (Applied Biosystems). All measurements were performed in triplicate and calculated using the ΔΔct method relative to GAPDH expression. Total human liver mRNA from pooled donors (Agilent) was used as a reference to normalise gene expression.

Immunoperoxidase labelling
Bladders from Ahr-knockout and wild-type C57BL/6 male mice aged 6-12 weeks were a kind gift from Prof Roland Wolf [50]. Tissues were fixed in 10% formalin and processed into paraffin wax. Sections (5 µm) were collected onto Superfrost TM Plus slides (VWR), dewaxed in xylene and rehydrated through ethanol to water. Antigen retrieval was performed by boiling slides in 10 mM citric acid buffer (pH 6) for 10 min (for Ahr/AHR labelling 0.1% (w/v) NP40 was added to the buffer). Endogenous avidin and biotin binding sites were blocked (Vector) and non-specific antibody binding was prevented by 10% (v/v) rabbit or goat serum (Dako), depending on the host of the secondary antibody. Tissues were incubated with primary antibodies raised against AhR (details as per western blotting, 1:2,000 dilution), Claudin 5 (Invitrogen, Clone 4C3C2, Mouse, 10 µg/mL) and Uroplakin 3a (Progen, Clone AU1, Mouse, 1:40 dilution) at 4°C overnight. Appropriate biotinylated secondary anti-immunoglobulin antibody (Dako) was applied for 30 min at ambient temperature, then streptavidin-biotinhorseradish peroxidase complex (Vector), with washing between each step. Bound antibody was detected using diaminobenzidine (Sigma) and slides were counterstained with Mayer's haematoxylin before mounting in DPX.

Trans-Epithelial Electrical Resistance (TEER) Monitoring Method
Following preconditioning for 5 days in 5% adult bovine serum (ABS) cells were harvested and reseeded onto 1.13 cm 2 permeable Snapwell TM supports (Corning) at 5x10 5 cells per membrane. After 24 h, the exogeneous calcium concentration was increased to 2 mM and the cultures maintained for 7 days before treatment and TEER was monitored using a portable Epithelial Volt-ohmmeter (World Precision Instruments). 1 µM ITE treatment was applied throughout the differentiation process, including preconditioning.

Analysis of The Cancer Genome Atlas (TCGA) cohort of MIBC
RNA sequencing data for TCGA cohort of MIBC [51] was downloaded from the NIH Genome Data Commons as normalised counts data and log transformed. MIBC was classified as either luminal or basal as previously described [52].

Basal/Luminal Classification of Cancer Cell Line Models
From previously published gene array data for a collection of bladder cancer cell lines [53], the data for RT4, RT112, T24 and SCaBER cells was extracted. The Choi et al. molecular classification of luminal and basal muscle invasive bladder cancers describes two gene set "classifiers" [52] and those genes were extracted and displayed as a heatmap.