The effects of rifampicin treatment on SULT2A1 mRNA expression were evaluated in 23 preparations of primary cultured human hepatocytes. In contrast to the consistently occurring induction of CYP3A4, a prototypical pregnane X receptor (PXR) target gene, rifampicin treatment increased SULT2A1 mRNA levels in 12 of the hepatocyte preparations, but it produced little change or even suppression in the others. Transient transfection of HepG2 cells with a series of reporter constructs implicated two SULT2A1 5Ј-flanking regions as containing rifampicin-responsive information. Each of these regions contained a hepatocyte nuclear factor 4 (HNF4) binding site (at nucleotide [nt] Ϫ6160 and Ϫ54), as demonstrated by in vitro binding and site-directed mutagenesis. HNF4␣ bound to the HNF4-54 region of the endogenous SULT2A1 gene, as indicated by chromatin immunoprecipitation. Cotransfection of HepG2 cells with pregnane X receptor (PXR) dose-dependently suppressed reporter expression from SULT2A1 constructs containing the HNF4 sites, and rifampicin treatment augmented the suppression. Rifampicin treatment concentration-dependently suppressed SULT2A1 reporter expression at the same concentrations that progressively induced expression from a PXR-responsive CYP3A4 reporter, whereas higher rifampicin concentrations reversed the SULT2A1 suppression. The suppressive effect of rifampicin was diminished, whereas the activating effect was augmented, in HepG2 cells with RNA interference-mediated PXR knockdown. These results suggest that HNF4␣ plays a central role in the control of SULT2A1 transcription and that rifampicin-liganded PXR suppresses SULT2A1 expression by interfering with HNF4␣ activity. By contrast, the rifampicin-inducible SULT2A1 expression that occurs in many human hepatocyte preparations seems to be mediated through a PXR-independent mechanism.

TaqMan Gene Expression assays were used to profile the mRNA expression of estrogen receptor (ERα and ERβ) and estrogen metabolism enzymes including cytosolic sulfotransferases (SULT1E1, SULT1A1, SULT2A1, and SULT2B1), steroid sulfatase (STS), aromatase (CYP19), 17β-hydroxysteroid dehydrogenases (17βHSD1 and 2), CYP1B1, and catechol-O-methyltransferase (COMT) in an MCF10A-derived lineage cell culture model for basal-like human breast cancer progression and in ERα-positive luminal MCF7 breast cancer cells. Low levels of ERα and ERβ mRNA were present in MCF10A-derived cell lines. SULT1E1 mRNA was more abundant in confluent relative to subconfluent MCF10A cells, a non-tumorigenic proliferative breast disease cell line. SULT1E1 was also expressed in preneoplastic MCF10AT1 and MCF10AT1K.cl2 cells, but was markedly repressed in neoplastic MCF10A-derived cell lines as well as in MCF7 cells. Steroid-metabolizing enzymes SULT1A1 and SULT2B1 were only expressed in MCF7 cells. STS and COMT were widely detected across cell lines. Pro-estrogenic 17βHSD1 mRNA was most abundant in neoplastic MCF10CA1a and MCF10DCIS.com cells, while 17βHSD2 mRNA was more prominent in parental MCF10A cells. CYP1B1 mRNA was most abundant in MCF7 cells. Treatment with the histone deacetylase inhibitor trichostatin A (TSA) induced SULT1E1 and CYP19 mRNA but suppressed CYP1B1, STS, COMT, 17βHSD1, and 17βHSD2 mRNA in MCF10A lineage cell lines. In MCF7 cells, TSA treatment suppressed ERα, CYP1B1, STS, COMT, SULT1A1, and SULT2B1 but induced ERβ, CYP19 and SULT2A1 mRNA expression. The results indicate that relative to the MCF7 breast cancer cell line, key determinants of breast estrogen metabolism are differentially regulated in the MCF10A-derived lineage model for breast cancer progression.

Estrogen sulfotransferase (SULT1E1) catalyzes the sulfonation of estrogens, which limits estrogen mitogenicity. We recently reported that SULT1E1 expression is low in preconfluent MCF10A human breast epithelial cells but increases when the cells become confluent. Pulse-chase labeling experiments with 5-bromouridine demonstrated that the confluence-mediated increase in SULT1E1 expression was due to increased mRNA synthesis. Because aryl hydrocarbon receptor (AhR) activation has been shown to suppress SULT1E1 expression and loss of cell-cell contact has been shown to activate the AhR in other cell types, we tested whether the confluence-associated changes in SULT1E1 expression were mediated by the AhR. Relative to confluent MCF10A cells, preconfluent cells had higher levels of CYP1A1 mRNA and greater activation of an AhR-responsive luciferase reporter, demonstrating that the AhR was active in the preconfluent cells. AhR and aryl hydrocarbon receptor nuclear translocator mRNA and protein levels were also higher in preconfluent than in confluent cultures. Treatment of preconfluent cells with the AhR antagonist, 3Ј-methoxy-4Ј-nitroflavone (MNF), or AhR knockdown significantly increased SULT1E1 expression. MCF10A cells stably transfected with a luciferase reporter containing ϳ7 kilobases of the SULT1E1 5Ј-flanking region showed both MNF-and confluence-inducible luciferase expression. Preconfluent cells transiently transfected with the reporter showed both MNF treatment-and AhR knockdown-mediated luciferase induction, but mutation of a computationally predicted dioxin response element (DRE) at nucleotide (nt) Ϫ3476 did not attenuate these effects. These results demonstrate that SULT1E1 expression in MCF10A cells is transcriptionally regulated by confluence through a suppressive action of the AhR, which is not mediated through a DRE at nt Ϫ3476.

Hydroxysteroid sulfotransferase (SULT2A) catalyzes the transfer of sulfate from the physiological sulfate donor 3′‐phosphoadenosine‐5′‐phosphosulfate (PAPS) to endogenous and xenobiotic substrates. In human hepatocytes, PAPS synthesis is catalyzed by PAPS synthase 2 (PAPSS2). HepG2 cells were stably transduced with lentivirus expressing either non‐targeting small interfering RNA (siRNA) (siNT‐HepG2) or siRNA directed against PAPS synthase 2 (siPAPSS2‐HepG2). Relative to siNT‐HepG2, siPAPSS2‐HepG2 cells transiently transfected with a reporter plasmid containing 1.5 kb of a murine SULT2A 5′‐flanking region demonstrated a significant (~2‐fold) increase in reporter expression. Disruption of a putative LXR motif in the murine SULT2A 5′‐flanking region suppressed the induction of SULT2A reporter expression produced by PAPSS2 knock‐down. Real‐time RT‐PCR analysis also demonstrated ~3‐fold induction of endogenous human SULT2A1 mRNA expression by PAPSS2 knock‐down. These results indicate that both human and murine SULT2A transcription are up‐regulated in response to PAPSS2 knock‐down, and implicate a role for LXR as a sulfate sensor. Supported by ES058223 (M.R.M.), HL50710 (T.A.K) and ES06636.

No abstract

The MCF10A cell line represents a model of the normal human breast epithelial cell. Estrogen sulfotransferase (SULT1E1), a major estrogen‐inactivating enzyme, is up‐regulated in confluent relative to preconfluent MCF10A cells. The mechanism responsible for SULT1E1 up‐regulation in confluent MCF10A cells was investigated. Transient or stable tranfection of reporter constructs containing up to 7 kb of the SULT1E1 5′‐flanking region failed to recapitulate confluency‐induced SULT1E1 up‐regulation. The half‐life of SULT1E1 mRNA in preconfluent MCF10A cells was estimated to be ~4h based on analysis of SULT1E1 mRNA levels in MCF10A cells that were replated after reaching confluency. Microarray analysis revealed that 85 microRNAs were significantly up‐regulated and 24 microRNAs were down‐regulated in confluent relative to preconfluent MCF10A cells. Computational analysis of the SULT1E1 3′‐untranslated region identified 4 putative candidate binding sites for confluency‐regulated microRNAs. The confluency‐related decrease in two of these microRNAs (miR‐221 and miR‐100*) corresponded with the confluency‐induced increase in SULT1E1 mRNA observed in MCF10A cells. These results support a role for the post‐transcriptional up‐regulation of SULT1E1 mRNA expression in confluent relative to preconfluent MCF10A cells. Supported by NIH grant (ARRA) ES016373 (MRM) and EHS Center grant ES06636.