1. Introduction
Prostate cancer (PCa) is the leading cause of cancer incidence in men in the United States. PCa is associated with disparities in patient outcomes wherein African American (AA) men have an earlier onset and poorer clinical outcomes compared to European American (EA) men [
1]. PCa growth is dependent on ligand-associated AR signaling that renders them susceptible to therapies that target androgen signaling, such as androgen-deprivation therapy (ADT), which are widely used in the clinic to treat patients as first-line treatment for metastatic disease. Over time, however, these tumors develop resistance to ADT, and invariably develop to a state known as castration resistant prostate cancer (CRPC). The development of CRPC is thought to occur via a variety of mechanisms that may involve ligand-independent AR signaling inclusive of pathways that utilize AR variants that lack the ligand binding domain (e.g., AR-v7). Our group previously demonstrated metabolic reprogramming as a key component of PCa growth and metastatic progression [
2,
3,
4,
5,
6]. Further, using a combination of targeted mass spectrometry and metabolic phenotyping microarrays, our laboratory previously found that the uridine diphosphate glucuronosyltransferase (UGT) pathway is one of the major altered metabolic pathways in both hormone-sensitive and castration-resistant prostate cancer [
4]. Moreover, in the same study, we found UGT2B28 expression to be associated with PCa progression [
4], which characterized its tumor promoting role.
The UGT family of enzymes is responsible for androgen glucuronidation, which regulates the steady-state levels of androgens in the body [
7]. Decreased UGT2B15 expression and increased UGT2B17 expression were shown to promote PCa progression [
8]. UGT2B28 is a key member of this group of UGTs whose mechanistic role in prostate cancer remains uncharacterized. In this study, we described the clinical relevance of UGT2B28, compared its expression between AA and EA PCa, identified a regulatory role for androgen signaling in UGT2B28 expression, demonstrated a tumor promoting function for the gene independent of the presence of androgens, and provided insights into potential effector pathways associated with its tumor promoting function.
2. Materials and Methods
2.1. Reagents
The UGT2B28 antibody for the tissue microarray analysis shown in
Figure 1A was purchased from Abnova (Cat. #H00054490-B01P). The UGT2B28 antibody (Ab2321) used for tissue microarray data in
Figure 1B–D and
Figure 2D,E was custom made by Dr. Levesque and Guillemette according to the previously published method [
9]. AR and PSA antibodies were purchased from Santa Cruz Biotechnology. Antibodies for beta-actin and GAPDH were purchased from Sigma Aldrich.
2.2. Cell Lines
LNCaP cells were obtained from the Tissue Culture Core at Baylor College of Medicine. LAPC-4 cells were a gift from Dr. Daniel Frigo (MD Anderson) and VCaP cells were a gift from Dr. Sean McGuire (Baylor College of Medicine). All cell lines were short tandem repeat (STR)-types from the MD Anderson Cytogenetics and Cell Authentication Core and were regularly confirmed to be free of mycoplasma contamination using the MycoAlert ™ Mycoplasma Detection Kit (Lonza, Anaheim, CA, USA).
2.3. Cell Culture
LNCaP cells were grown in RPMI-1640 media (Invitrogen Corp., Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (FBS, Hyclone Labs, Thermo Scientific, Rockford, IL, USA) and 1% penicillin-streptomycin (Hyclone Labs, Thermo Scientific, Rockford, IL, USA). LAPC-4 cells were grown in IMDM (Invitrogen Corp., Carlsbad, CA, USA) supplemented with 15% FBS, 1% penicillin-streptomycin, and 1 nM R1881 (Sigma Aldrich, St. Louis, MO, USA). VCaP cells were grown in DMEM (Invitrogen Corp., Carlsbad, CA, USA) and supplemented with 10% FBS and 1% penicillin-streptomycin. All cells were maintained at 37 °C, 5% CO2, and 95% humidity.
2.4. Lentiviral Transduction
To generate a stable knockdown of UGT2B28 in LNCaP, LAPC-4, and VCaP cells, cells were transduced with two independent clones of lentiviral UGT2B28 shRNA (Clone ID: V3LHS_378777) and non-targeting control pGIPZ shRNA purchased from the Cell-Based Assay Screening Service Core (Baylor College of Medicine). Transduction was carried out at an MOI of 5, and cells containing the KD were stably selected with 1 µg/mL (LNCaP, LAPC-4), and 2 µg/mL (VCaP) puromycin from Sigma Aldrich (Cat. #P8833) in their respective media. Similarly, stable overexpression of UGT2B28 in NT and KD cells (NTOE, KD2OE, and KD4OE) was achieved using over-expression lentiviral vector for UGT2B28 (LPP-H0607-Lv157-050) along with a negative control (LPP-NEG-Lv157-050) purchased from Genecopoeia at an MOI of 5 and the overexpressing cells were stably selected using neomycin from Sigma Aldrich (1 mg/mL).
2.5. Organoid Forming Assay
LNCaP and LAPC-4 cells were seeded at a density of 600,000 cells in 300 µL media on 12 mm Millicell™ inserts (Millipore Sigma, St. Louis, MO, USA) with cellulose ester filters. The inserts were placed in 24-well plates with each well containing 600 µL media at the bottom and the cells were allowed to spontaneously form organoids by 16 h (LNCaP) and 24 h (LAPC-4). The organoids were imaged using a fluorescence confocal microscope with brightfield and GFP filters at 4× and 10× magnifications at the Integrated Microscopy Core at Baylor College of Medicine.
2.6. Perturbation of Androgen Signaling in the Organoid Assay
LNCaP cells were seeded and grown in RPMI-1640 media as described above in 6-well plates for 24 h on Day 1. The medium was changed to RPMI-1640 media (Invitrogen Corp., Carlsbad, CA, USA) lacking phenol red and supplemented with 10% charcoal-stripped FBS (CSS, Invitrogen Corp. Carlsbad, CA, USA) and 1% penicillin-streptomycin on Day 2 for another 24 h. On Day 3, cells were treated with CSS, 1 nM 5α-dihydrotestosterone (DHT) (Sigma Aldrich, St. Louis, MO, USA), 10 μM enzalutamide (MDV-3100) (Sigma Aldrich, St. Louis, MO), 300 nM ARCC4 (Sigma Aldrich, St. Louis, MO, USA), 1 nM DHT + 10 μM MDV-3100, or FBS for 24 h. On Day 4, the treated cells were trypsinized and reseeded onto Millicell ™ inserts in their respective media and were allowed to form organoids for another 24 h. The organoids were imaged as described above.
2.7. Immunohistochemistry Analysis
2.7.1. University of Michigan Prostate Cancer Tissue Microarrays
Established clinically annotated tissue microarrays (TMAs) created from tissues obtained during radical prostatectomy (n = 165; benign, n = 95 and localized PCa, n = 70) with progression endpoints, such as biochemical recurrence, were used for immunohistochemistry analysis. TMAs composed of metastatic samples (CRPC) that were also used to measure UGT2B28 protein expression by immunohistochemistry were obtained from the University of Michigan’s warm autopsy program which includes 45 CRPC tissues sampled across multiple sites, with each site replicated into 3 cores (i.e., prostate, liver, lymph node, soft tissue, dura, bladder, adrenal gland, seminal vesicle, diaphragm, and pancreas). UGT2B28 antibody for the staining of these TMAs was purchased from Abnova. Immunohistochemistry staining was performed at the histopathology core at University of Michigan. Immunostaining was categorized into weak, moderate, and strong staining by genitourinary pathologists Dr. Rohit Mehra and Dr. Lisha Wang.
2.7.2. Baylor College of Medicine Prostate Cancer Tissue Microarrays
An analysis of UGT2B28 expression in tissue microarrays containing AA and EA PCa patients (
n = 105 AA patients,
n = 102 EA patients) was conducted with a custom made UGT2B28 antibody as previously described [
9]. Matched tumor-benign pairs of prostate tissues from patients were processed into formalin-fixed paraffin-embedded tissues and organized into tissue microarrays. Custom made UGT2B28 antibody (Ab2321, 1:500 dilution) was used to stain tissue microarrays with Autostainer Link 48 from DAKO using the EnVision FLEX10 protocol (Agilent, Santa Cara, CA, USA). The intensity of staining was scored for both nuclear and cytoplasmic compartments. Both the percentages of negative and positive nuclei were recorded, as well as the percentage of stained cytoplasm.
2.8. Immunofluorescence
Immunofluorescence experiments for UGT2B28 were performed using custom made UGT2B28 antibody. Briefly, cells were plated on poly-d-lysine coated slides then fixed in 4% paraformaldehyde and permeabilized with 1× PBS–0.2% Triton X-100. After blocking with 1× PBS-5%-BSA-5% goat serum, cells were incubated with anti-UGT2B28 (1:200) and anti-Androgen Receptor antibody (1:200, 441 from Santa Cruz, Dallas, TX, USA) and incubated with secondary antibodies, including Alexa Fluor 488-labeled goat anti-rabbit antibody and Alexa Fluor 555-labeled goat anti-mouse antibody (1:1000) (Invitrogen, Burlington, ON, Canada). Cells were stained with DRAQ5 (1:2000) (Abcam, Branford, CT, USA). Images were captured using a LSM510 META NLO laser scanning confocal microscope (Zeiss, Toronto, ON, Canada). Zen 2009 software version 5.5 SP1 (Zeiss, Toronto, ON, Canada) was used for image acquisition.
2.9. Mouse Xenograft Studies
Animal well-being and animal experiments were performed and monitored in accordance with a specific animal protocol approved by the Institutional Animal Care and Use Committee (IACUC) of Baylor College of Medicine. Six-week-old athymic nude mice (Strain #553), both intact and castrated, were purchased from Charles River, Frederick, MD. The 60-day release testosterone pellets for immuno-deficient mice (Cat. #SA-151) were purchased from Innovative Research of America, Sarasota, FL. Mice were subcutaneously implanted with pellets (12.5 mg/kg body weight) and administered subcutaneous injections of a mixture of 2,000,000 tumor cells (LNCaP, LAPC-4, or VCaP), 50,000 HPS-19I prostate stromal cells [
10], and Matrigel (Corning, Tewksbury, MA, USA). Tumors were measured three times a week. LNCaP tumors were resected at ~500 mm
3 to observe KD growth and LAPC-4 and VCaP tumors were resected when the control tumors reached ~600 mm
3 to obtain relative tumor take curves. Similar subcutaneous injections were performed in castrated mice for VCaP tumors.
2.10. Quantitative Real-Time Polymerase Chain Reaction
RNA extraction from cells or tissues was performed using the RNeasy Mini Kit from QIAGEN. RNA was reverse transcribed into cDNA using the cDNA Superscript Mix (Quanta Biosciences, Cat. #95048-500) and RT-qPCR was performed using SYBR green (Life Technologies, Cat. #4385614). Either 18S or β-actin were used as appropriate house-keeping controls. The primers used in this study are listed in
Supplementary Table S1.
2.11. Sample Preparation for Mass Spectrometry
LNCaP cells were starved in RPMI without glucose (Invitrogen Corp., Carlsbad, CA, USA) for 24 h and treated with 12 mM 13C-glucose (Sigma Aldrich, St. Louis, MO, USA) containing RPMI media for an additional 24 h. Cells were washed three times with ice-cold PBS and pelleted. Frozen pellets were thawed on ice and subjected to lysis by repeated cycles of freeze–thaw in liquid nitrogen and at room temperature. A total of 750 µL of ice-cold methanol:water (4:1) with 20 µL of spiked internal standard was added to each sample and homogenized for 30 s pulses of 1 min each. A total of 450 µL of ice-cold chloroform and 150 µL of water were sequentially added to each sample. Both the organic and aqueous layers were combined and filtered via a 3 kDa Amicon Ultracel molecular filter (Millipore, Billerica, MA, USA). The filtrates were then vacuum-dried using a Genevac EZ-2 plus (Gardinier, NY, USA) and resuspended in 100 µL of 1:1 methanol:water with 0.1% formic acid before injecting the samples into the mass spectrometer.
2.12. Flux Measurement and Analysis of UGT Co-Substrates
The method for the flux measurement of intracellular UGT metabolites was generated using uridine 5′-diphosphoglucuronic acid as our reference standard (Cat. #U6751-100MG), and the transitions that were monitored are tabulated in
Supplementary Table S2.
2.13. Statistical Analysis
Unless otherwise stated, all samples were assayed in triplicate. All in vitro experiments were repeated for a minimum of three independent runs. Unless otherwise indicated, data are represented as mean ± standard deviation (SD), and significance was calculated using Student’s unpaired two-tailed t-test.
4. Discussion
Metabolic reprogramming has been established as a hallmark of cancer and has become a new focus in understanding the molecular underpinnings that drive tumorigenesis [
12]. The goal of our laboratory was to define these dysregulated metabolic pathways that contribute to cancer cell growth and can be exploited for possible cancer therapeutics. Our laboratory delineated the altered metabolites and associated pathway changes associated with AD and CRPC cell lines, among which the UGT pathway emerged as one of the most significant [
4]. UGT2B28, a previously uncharacterized UGT family member, was among the chief altered enzymes associated with this pathway in our profiling results. We demonstrated that UGT2B28 expression is strongly correlated with disease aggressiveness, as previously reported [
9]. Furthermore, we identified elevated levels of UGT2B28 in AA PCa patients compared to their EA PCa counterparts, with the former known to have worse PCa-associated clinical outcomes [
1].
It is particularly interesting that our results show that UGT2B28 is regulated by AR and AR-v7 at the genetic level, and its tumor promoting role is not fully compromised by castrated mouse models or the ablation of AR or androgens. We found that MYC is a remarkable example of another AR target gene with a tumorigenic function in PCa that is independent of androgen stimulation [
13].
UGT2B28 has been described to promote PCa progression and was linked with alteration in circulating androgen levels in men with localized PCa [
9]. However, our in vitro and in vivo functional data revealed similar effects on 3D organoid formation and a reduction in tumor growth upon UGT2B28 KD, both in a ligand-driven and castrated AR signaling environment, which suggests that UGT2B28 may have non-canonical roles in prostate tumorigenesis. An insight into this non-canonical function was obtained by examining the flux measurements of
13C-UDP-glucuronic acid (
Supplementary Figure S3G). We previously determined that UDP-glucuronic acid, the co-substrate of UGT2B28, is a precursor for hyaluronic acid and the depletion of this precursor reduces tumor growth [
14]. Hyaluronan synthase 2 (HAS2) is the enzyme responsible for synthesizing hyaluronic acid, and hyaluronidase-1 (HYAL1) is the enzyme that cleaves high molecular weight hyaluronic acid, which is anti-tumorigenic, into smaller molecular weight tumor promoting fragments known to promote angiogenic, tumorigenic, and migratory phenotypes in many different cancers [
15,
16]. It is possible that an altered balance between HAS2 and HYAL1 upon UGT2B28 KD may contribute to the delay in tumor take and tumor growth, which would need to be verified in the future using additional experiments. We investigated signaling pathways that may be altered upon UGT2B28 expression using mass spectrometry-based proteomics experiments conducted on both VCaP tumors (in intact mice) and LNCaP cells (
Supplementary Figure S6).
One of the challenges in studying UGT2B family members for their role in PCa is their extensive sequence overlap. This has led to a current lack of commercial mono-specific antibodies to UGT2B28. We, along with our collaborators, addressed this by synthesizing a custom-made UGT2B28 antibody and we performed mass spectrometry experiments to further confirm the specificity of our UGT2B28 rescue. We also demonstrated that the UGT2B28 R setting results in the rescue of only UGT2B28 and none of the other family members. Furthermore, in all our studies, we mandated that the interpretation of the results was dictated by the ability of UGT2B28 R model to completely reverse the phenotype resulting from UGT2B28 KD. Finally, to ensure the robustness of our findings, we ensured that the phenotype of the 3D organoids and xenograft tumors resulting from the perturbation of UGT2B28 was identical in three independent PCa models. In light of all of the above, our data provide support for a non-canonical function of UGT2B28 in PCa.