Inhibition of androgen/AR signaling inhibits diethylnitrosamine (DEN) induced tumour initiation and remodels liver immune cell networks - PubMed (original) (raw)
. 2021 Feb 11;11(1):3646.
doi: 10.1038/s41598-021-82252-x.
Riley D Mullins # 3, Jennifer M Thomas-Ahner 2, Samuel K Kulp 2 3, Moray J Campbell 2 3 4, Fabienne Lucas 5, Nathan Schmidt 2, Dana M LeMoine 6, Surafel Getaneh 3, Zhiliang Xie 3, Mitch A Phelps 3, Steven K Clinton 2 7, Christopher C Coss 8 9
Affiliations
- PMID: 33574348
- PMCID: PMC7878907
- DOI: 10.1038/s41598-021-82252-x
Inhibition of androgen/AR signaling inhibits diethylnitrosamine (DEN) induced tumour initiation and remodels liver immune cell networks
Timothy H Helms et al. Sci Rep. 2021.
Abstract
A promotional role for androgen receptor (AR) signaling in hepatocellular carcinogenesis is emerging. In pre-clinical models, including diethylnitrosamine- (DEN-) induced hepatocellular carcinoma (HCC), anti-androgen therapies delay hepatocarcinogenesis. However, pharmacologic anti-androgen therapy in advanced HCC patients fails, suggesting that AR plays a role in HCC onset. This study aims to characterize AR expression and function throughout DEN-induced liver inflammation and carcinogenesis and evaluate the efficacy of prophylactic AR antagonism to prevent hepatocarcinogenesis. We demonstrate that pharmacologic AR antagonism with enzalutamide inhibits hepatocellular carcinogenesis. With enzalutamide treatment, we observe decreased CYP2E1 expression, reducing DEN-induced hepatocyte death and DNA ethyl-adducts. AR protein expression analyses show that DEN causes an initial upregulation of AR in portal fibroblasts and leukocytes, but not hepatocytes, suggesting that hepatocyte-autonomous AR signaling is not essential for DEN-induced carcinogenesis. Ablating androgen signaling by surgical castration reduced pre-carcinogen Kupffer cell populations but did not alter DEN-mediated immune cell recruitment nor AR expression. In this study, we identified that anti-androgen interventions modulate mutagenic DNA adducts, tumour initiation, and immune cell composition. Additionally, we find that AR expression in hepatocytes is not present during nor required for early DEN-mediated carcinogenesis.
Conflict of interest statement
The authors declare no competing interests.
Figures
Figure 1
Low dose, preventive ENZ inhibits DEN-induced hepatocellular carcinogenesis. (a) Study Design—simplified timeline highlighting preventive anti-androgen interventions. Carcinogen challenge consisting of the described DEN/PH/DEN procedure is represented by carcinogen pictogram. SHAM, sham castration; ORX, castration. Enzalutamide (ENZ) groups received 10, 30, or 100 mg/kg and control groups received vehicle (VEH), all once daily by oral gavage. Carcinogen-challenged, preventive ORX served as anti-androgen phenotype positive controls; carcinogen-challenged, SHAM rats served as anti-androgen phenotype negative controls. (b) Low-dose ENZ maintained an anti-androgen effect on bodyweight-normalized urogenital tract weights. *** p < 0.0005; ****p < 0.0001 (vs. VEH), one-way ANOVA followed by Dunnett’s multiple comparisons. (c) ENZ at 10 mg/kg reduced liver damage marker, serum ALT(U/L). *p = 0.0298 (10 mg/kg ENZ vs. VEH), one-way ANOVA followed by Dunnett’s multiple comparisons. (d) ENZ at 30 mg/kg reduced the number of pGST-positive, preneoplastic foci. Left, Representative photomicrographs (40 × total magnification) of IHC for pGST for VEH (left) and 30 mg/kg ENZ (right) treatment groups. Right, Comparison of the total number of pGST-positive foci between treatment groups. *p = 0.033 (30 mg/kg vs. VEH), one-way ANOVA followed by Dunnett’s Multiple Comparisons. All data represented as mean ± standard deviation.
Figure 2
Preventive anti-androgen treatments inhibit DEN carcinogenesis in a model specific manner. (a) Study Design—simplified timeline highlighting preventive anti-androgen interventions for follow-up study. ENZ = 30 mg/kg daily oral gavage doses. Note addition of non-DEN, VEH-treated, PH only control group. (b) Heatmap highlighting gene expression (row) by animal (column) within treatment group (column color). Red = downregulation; Blue = upregulation. (c) qRT-PCR confirmation of expression of Wnt/β-catenin targets CYP1A2, CYP2E1, CYP7A1, Glul. Preventive ENZ suppresses CYP2E1 and CYP7A1 expression. (d) Zonal distribution of CYP2E1 (200x) and Glul (400x) is reduced by preventive ENZ; ENZ and ORX suppress DEN induction of CCIII (400x). (e) ENZ reduces CYP2E1 positive staining area (quantification of 2D CYP2E1). (f) ENZ reduces Glul expression (quantification of 2D Glul) Quantitative, positive pixel analysis supports this conclusion. All analyses are One-Way ANOVA, Dunnett’s Multiple comparisons (vs. DEN/VEH): *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. All data represented as mean ± standard deviation.
Figure 3
Characterization of the hepatic AR in DEN carcinogenesis. (a) Left: qRT-PCR comparison of AR expression in non-DEN challenged liver collected at partial hepatectomy (PH) versus sacrifice (SAC). AR transcript expression at SAC is higher (Paired T-test, ***p = 0.0005) implicating a PH-dependent increase in AR signaling. Middle: qRT-PCR comparison of hepatic AR expression at PH and SAC based on DEN challenge and intervention. Significant sources of variation include PH (p = 0.0002) and within subject (p = 0.0034) effects. Data represented as mean ± standard deviation. Right: Western blot for AR at PH and SAC within subjects (Samples from the same animals at PH and SAC are represented in the same lane). GAPDH = reference protein. LNCaP cell lysate was used as positive control. Individual lanes are labelled with each sample’s identification number, three samples per experimental group. Consistent with transcript data, AR is upregulated with DEN challenge at SAC. Full-length blots are presented in Supplementary Fig. S5 online. (b) AR expression is limited to the portal triad in early DEN carcinogenesis. Top Left—positive control, rat prostate, note strong positive nuclear immunoperoxidase staining for AR in glandular epithelium and stromal cells. Remaining Panels—Liver: AR expression is limited to the portal triad/periportal area. Strong AR staining is in the nuclei of spindle to round-shaped nuclei lining bile ducts (BD) and portal vein (PV) and hepatic arteriole (HA). Parenchymal hepatocytes (H) lack AR staining. There is rare, light nuclear staining of smooth muscle walls of the HA and limiting plate hepatocytes. There is no change in the distribution of AR expression with DEN or intervention. (c) AR expression is scarce in DEN-induced adenomas. Top Left—H&E, photomicrograph of AR-positive adenoma. Top Right—IHC for AR of same adenoma. Bottom Left—200 × total magnification inset of dotted region from top left panel. Bottom Right—400 × total magnification inset of dashed region from top left panel. AR immunoperoxidase-positive nuclei of neoplastic hepatocytes are highlighted with arrows. (d) AR expression is scarce in DEN-induced carcinomas. Left—Photomicrographs of representative H&E-stained sections of two AR-positive carcinomas. Right—Photomicrographs of AR IHC in the same carcinomas which again highlight rare to occasional nuclear AR expression in neoplastic hepatocytes (arrows). Cytoplasmic staining is non-specific edge artefact as determined by rabbit IgG isotype, and no primary antibody controls (data not shown). (e) Left—AR-positive nuclei (brown) do not co-localize with cytoplasmic CD31 (blue) but are often seen around CD31-positive vascular structures (arrowhead; 600 × total magnification). Middle—AR-positive nuclei are frequently surrounded by SMA-positive cytoplasmic staining (blue, arrows; 600 × total magnification). Right—AR-positive nuclei are often surrounded by CD45-positive staining cytoplasm (blue, arrows; 600 × total magnification).
Figure 4
DEN increases proportion of AR+ lymphocytes and monocytes in SHAM and ORX. Liver immune cell populations were analyzed by flow cytometry as described in Materials and Methods and represented as percent of parent population and pre-gated on CD45+ and living cells. Positive cells were defined by fluorescence minus one (FMO) controls (See Supplementary Fig. S4 online for gating strategy). (a) Simplified experimental timeline. Surgery was performed on 8 mice per group 19 days prior to treatment with IP VEH or DEN (50 mg/kg). DEN mice were sacrificed at 2 and 7 days following treatment, and VEH mice were sacrificed at 7 days following treatment. (b) ORX increased the percentage of total CD45+ cells (*p < 0.05, Sidak's multiple comparisons). (c) DEN challenge in SHAM mice increased the proportions of granulocytes, AR+ lymphocytes and monocytes and decreased that of lymphocytes at 1 week after DEN challenge (*p < 0.05, Sidak's multiple comparisons). (d) The changes in ORX mice were similar to those in SHAM mice in response to DEN challenge (*p < 0.05, Sidak's multiple comparisons). All data represented as mean ± standard deviation.
Figure 5
ORX- and DEN-mediated changes in immune cell recruitment to the liver. Liver immune cell populations were analyzed by flow cytometry as described in Materials and Methods and represented as percent of parent population and pre-gated on CD45+ and living cells. Positive cells were defined by fluorescence minus one (FMO) controls (See Supplementary Fig. S4 online for gating strategy). (a) ORX increased the proportions of CD3+ T cells and CD19+ B cells and decreased that of F4/80+ CD11b− Kupffer cells in VEH-treated mice (*p < 0.05, Tukey’s multiple comparisons). (b) DEN challenge in ORX mice did not alter immune cell proportions in comparison to SHAM mice. (c) DEN challenge in SHAM mice increased the percentage of CD3+ T cells at 1 week and decreased that of CD19+ B cells at 48 h and 1 week. The proportion of F4/80+ CD11b+ monocyte-derived resident liver macrophages increased at 48 h and subsequently decreased at 1 week following DEN, whereas that of F4/80+ CD11b− Kupffer cells decreased at 1 week after DEN. The percentage of F4/80− CD11b+ myeloid lineage cells increased at 48hrs and remained elevated at 1 week (*p < 0.05, Tukey’s multiple comparisons). (d) DEN challenge in ORX mice caused similar changes as DEN challenge in SHAM mice, except that the percentage of CD3+ T cells decreased at 48 h and that of F4/80+ CD11b− Kupffer cells did not change relative to baseline (*p < 0.05, Tukey’s multiple comparisons). All data represented as mean ± standard deviation.
Figure 6
Graphical hypothesis of the impact of AR-signaling on DEN-mediated carcinogenesis. DEN is metabolized by CYP2E1 in hepatocytes into reactive species that create DNA ethyl adducts and induce liver damage. In the acute immune response to DEN, the proportions of CD3+ T cells, F4/80− CD11b+ myeloid lineage cells, and F4/80+ CD11b+ monocyte-derived resident liver macrophages are increased coinciding with a decreased proportion F4/80+ CD11b− Kupffer cells and CD19+ B cells. Inhibition of the AR/androgen signaling axis reduces DNA adduct formation in hepatocytes, decreases liver damage, and decreases baseline F4/80+ CD11b− Kupffer cells, yet does not alter relative immune cell recruitment. The image of the liver is from TogoTV (2016 DBCLS TogoTV / CC-BY-4.0),
https://doi.org/10.7875/togopic.2014.7
.
References
- Theise, N. D. Liver and Gallbladder. In Robbins and Cotran Pathologic Basis of Disease, 9th edn. (eds. Kumar, V., Abbas, A. K., & Aster, J. C.) 867–875 (2015).
Publication types
MeSH terms
Substances
LinkOut - more resources
Full Text Sources
Other Literature Sources
Medical
Research Materials