Hypoxia-induced down-regulation of MITF is mediated by HIF1. Western blot analysis of MITF and HIF1α levels in (A) human primary melanocytes or (B) UACC62 cells 48 h after infection with GFP-control or HIF1α-GFP adenoviruses at increasing MOIs as indicated. (C) UACC62 cells were transfected with control or anti-HIF1α siRNA and then subjected to 200 μM CoCl₂ treatment for 24 h. M-MITF bands are marked with arrowheads.

DEC1 suppresses M-MITF transcription. (A) qPCR analysis indicates that DEC1 expression is rapidly induced in UACC62 cells treated with 200 μM CoCl₂, preceding M-MITF down-regulation. (B) Transient overexpression of DEC1 or DEC2 in UACC62 cells is sufficient for M-MITF protein down-regulation. (C) DEC1 silencing using two different DEC1 shRNA (sequence 1 in gray bars; sequence 2 in dotted bars) rescues M-MITF mRNA from down-regulation by 200 μM CoCl₂ compared with cells introduced with control shRNA (black bars). (D) UACC62 cells introduced with shRNA against DEC1 (sequence 1, gray bars) or control shRNA (black bars) were assayed for M-MITF promoter activity using the pGL3-hMIP-2.3K luciferase reporter plasmid. HIF1α (P420A) overexpression resulted in dose-dependent inhibition of M-MITF promoter activity, which was rescued by DEC1 silencing. (E) Reporter assays demonstrate that the HIF1α-induced suppression of the M-MITF promoter activity is dependent on an upstream DEC1 binding site. (F) DEC1 is recruited to the M-MITF promoter, but not to the β-actin promoter, upon CoCl₂ treatment of MALME-3M cells. Immunoprecipitation with a control IgG is shown as control for precipitation specificity. (G) Knockdown of DEC1 fully rescues MITF protein levels in CoCl₂-treated UACC62 cells. M-MITF bands are marked with arrowheads.

DMOG suppresses melanoma xenograft tumor growth. (A–C) Nude mice were implanted with 10⁶ UACC62 melanoma cells per site and, after establishment of tumors, treated with DMOG (20 mg per injection, twice per day; n = 10) or vehicle (PBS solution; n = 8). Measured tumor volume (A) and representative images (B) demonstrate DMOG efficacy without affecting mouse weight (C). Data are represented as mean ± SEM. (D) UACC62-Vec or UACC62-MITF polyclonal melanoma lines were injected in nude mice. After establishment of tumors, each group was treated with DMOG (10 mg/d) or PBS solution. Note that DMOG reduced tumor volume in UACC62-Vec–injected mice but not in mice injected with UACC62-MITF harboring hypoxia-insensitive MITF. Data were combined from two independent experiments on day 11 of treatment, and normalized to the vehicle control for each group. Group size was 11 to 17 as indicated. Statistical significance was calculated by Student t test.

MITF levels are down-regulated by hypoxia. (A) H&E and MITF immunostaining of metastatic melanoma shows that MITF expression is limited to a vascularized zone (serial sections). (B) Immunofluorescent staining of human metastatic melanoma indicates reciprocal expression of MITF (green) and HIF1α (red). (C) Expression analysis of MITF and some of its target genes (CDK2, Silver, and Tyrosinase) in melanoma specimens and cell lines (asterisk) shows inverse Pearson correlation relative to hypoxia-induced genes. (D) MITF expression in three melanoma cell lines grown under normal oxygen tension (21%) or subjected to hypoxia (0.5% O₂) or hypoxia mimicry (with 200μM CoCl₂) for 24 h. HIF1α levels are presented as readout of cellular hypoxic response, and α-tubulin (αTub) levels serve as loading control. MALME-3M and UACC257 are lines with amplified MITF [n = 6–8 and n = 4–5 copies, respectively (20)]. (E) MITF levels in primary human melanocytes decrease following CoCl₂ treatment. M-MITF bands are marked with arrowheads.

MITF RNA levels are decreased as a result of reduced transcription rate. (A) MITF RNA levels decrease in a time-dependent manner in UACC62 cells treated with CoCl₂. (B) Decrease in MITF RNA levels following incubation in low oxygen conditions or adeno-HIF1α infection (time 24 h). (C) Nascent and spliced MITF RNA species exhibit the same magnitude of down-regulation under CoCl₂ treatment. VEGF levels are shown as positive control. (D) ChIP showing diminished recruitment of the RNAPII machinery to the MITF locus following CoCl₂ treatment of UACC62 cells. (E) Transfection of pcDNA3-full length human M-MITF mRNA to mouse B16F0 melanoma cells reveals that promoter swap can rescue MITF RNA down-regulation. (F) UACC62 polyclonal stable line expressing HA-MITF off a heterologous promoter and UACC62–empty vector control line were subjected to CoCl₂ treatment. MITF protein level was evaluated by Western blot by using MITF and HA antibodies. M-MITF bands are marked with arrowheads. Note that HA-MITF protein levels are unaffected following CoCl₂ treatment under conditions that reduced the endogenous protein. In all CoCl₂ treatments, a dose of 200 μM was applied. Results are triplicates from one experiment (mean ± SD of fold relative to untreated cells) and representative of three independent experiments. P values were calculated relative to control by using a Student t test.

MITF rescues UACC62 melanoma cells from DMOG-induced loss of viability. (A) Stable UACC62 lines carrying HA-MITF or empty vector express moderate MITF levels compared with other human melanoma lines. (B) UACC62-Vec cells show dose-dependent reduction in MITF levels following DMOG treatment, whereas HA-MITF levels are unaffected. M-MITF bands are marked with arrowheads. (C–E) Ectopic MITF expression rescues UACC62 cells from DMOG-induced loss of viability. Images in C represent random fields and corresponding crystal violet-stained wells after 6 d of DMOG treatment. Growth/survival curve of UACC62-Vec (D) or UACC62-MITF (E) cells treated with 0 to 1 mM DMOG. Results are triplicates from one experiment (mean ± SD) and representative of three independent experiments.

Inhibition of MITF transcription by DEC1. Low oxygen pressure or prolyl-hydroxylase inhibitors stabilize the HIF1 heterodimer by rescuing HIF1α subunit from degradation. Consequently, stabilized and activated HIF1 binds hypoxia response elements (HRE) and induced transcription of DEC1. Acting as a transcriptional repressor, DEC1 binds the MITF promoter and inhibits its transcription.