The protective role of DOT1L in UV-induced melanomagenesis

Abstract

The DOT1L histone H3 lysine 79 (H3K79) methyltransferase plays an oncogenic role in MLL-rearranged leukemogenesis. Here, we demonstrate that, in contrast to MLL-rearranged leukemia, DOT1L plays a protective role in ultraviolet radiation (UVR)-induced melanoma development. Specifically, the DOT1L gene is located in a frequently deleted region and undergoes somatic mutation in human melanoma. Specific mutations functionally compromise DOT1L methyltransferase enzyme activity leading to reduced H3K79 methylation. Importantly, in the absence of DOT1L, UVR-induced DNA damage is inefficiently repaired, so that DOT1L loss promotes melanoma development in mice after exposure to UVR. Mechanistically, DOT1L facilitates DNA damage repair, with DOT1L-methylated H3K79 involvement in binding and recruiting XPC to the DNA damage site for nucleotide excision repair (NER). This study indicates that DOT1L plays a protective role in UVR-induced melanomagenesis.

Fig. 1

Identification of DOT1L mutations in human melanoma. a Somatic CNVs identified by GISTIC analysis of SNP6 data. Significantly deleted chromosome regions are shown. The deletion in red contains the DOT1L gene. b Gene-based burden test by ANNOVAR. c Percentages of melanoma samples with DOT1L mutations identified from the TCGA data portal, Queensland, and two published papers from the Broad Institute. DOT1L mutations from TCGA data portal and the published papers from the Broad Institute were identified by GATK and MuTect/VarScan using data from both melanoma tissues and paired germline DNA samples. DOT1L mutations from Queensland were analyzed according to the Illumina protocols

Fig. 2

DOT1L mutants in human melanoma are loss of function. a In vitro H3K79me2 levels from melanoma cells with known DOT1L mutation status from Queensland, Australia. b DOT1L-depleted HPMs were infected with lentivirus encoding WT or mutant DOT1L as indicated. MV411 cells were transiently transfected with WT or mutant DOT1L as indicated. c, d MV411 cells or DOT1L-depleted HPMs with WT or mutant DOT1L overexpression were treated with EPZ-5676 (5 µM). Cell growth was detected using cell number counting (c) or MTT assay (d). e H3K79me2 signal intensities in the gene body regions, under the condition of C021 shControl, C021 shDOT1L, and C025 shControl. *p < 0.05, **p < 0.01, ***p < 0.001, unpaired Student’s t-test. Error bars represent ± s.d.

Fig. 3

DOT1L mutations cooperate with BRAF^V600E to promote melanomagenesis. a Engineered human melanocytes were infected with the shDOT1L lentiviral construct to deplete endogenous DOT1L, with control or BRAF^V600E expression. b The cells generated in a were subjected to clonogenic survival assays 14 days after UVR. Crystal violet was used to stain the colonies and the relative colony formation was calculated as mean ± s.d. c The cells generated in a were seeded (10,000 cells per well) for soft agar assay and cultured for 30 days. The relative colony numbers were calculated as mean ± s.d. d DOT1L-depleted hTERT/p53DD/CDK4(R24C)/BRAF^V600E melanocytes expressing empty vector (EV), DOT1L wild-type (WT), M55L, P271L, and P505L mutants were subjected to western blot. e The cells generated in d were subjected to clonogenic survival assays 14 days after UVR. Crystal violet was used to stain the colonies and the relative colony formation was calculated as mean ± s.d. f The cells generated in d were seeded (10,000 cells per well) for soft agar assay and cultured for 30 days. The relative colony numbers were calculated as mean ± s.d. g–i Growth curves for the xenograft experiments. The indicated 3 × 10⁶ tumor cells were inoculated subcutaneously into each flank of the nude mice. The visible tumors were measured at the indicated days, error bars represent ± s.e.m, n = 7 (g). Dissected tumors (h) and tumor weight (h) were displayed. **p < 0.01, ***p < 0.001, unpaired Student’s t-test

Fig. 4

UVR-induced melanoma development in melanocyte-specific DOT1L conditional knockout mice. a Experimental design for UVR-induced melanoma development in the DOT1L^flox/flox/Tyr-CreERT2 mice. b Melanoma incidence in wild-type and DOT1L^flox/flox/Tyr-CreERT2 mice with or without UVB irradiation. c Image of a representative DOT1L^flox/flox/Tyr-CreERT2 cutaneous melanoma. d H&E staining of a representative cutaneous melanoma. Scale bar, 500 μm (upper panel) and 50 μm (lower panel). e Immunohistochemistry staining of S100, tyrosinase, Pax3, and Mart-1 of the representative cutaneous melanoma. Scale bar, 100 μm. f The relative mRNA expression of Pax3 and tyrosinase in mouse melanomas comparing to those of normal skin tissues. ***p < 0.001, unpaired Student’s t-test. Error bars represent ± s.d.

Fig. 5

DOT1L/methylated H3K79 is involved in NER on UVB-induced DNA damage through interacting with XPC. a Melanomas from Queensland, Australia were irradiated with different doses of UVB. Cell viability was measured by MTT assay 24 h after UVB irradiation. b HPMs with stable shDOT1L or shControl were irradiated with different doses of UVB as indicated. Cell viability was measured by MTT assay 24 h after UVB irradiation. c WT or different DOT1L mutants as indicated was introduced into HPMs with stable shDOT1L expression. The resulting cells were irradiated with different doses of UVB as indicated. Cell viability was measured by MTT assay 24 h after UVB irradiation. d–f Melanomas from Queensland, Australia (d), HPMs with stable shDOT1L expression (e), or HPMs with stable shDOT1L expression and WT or different mutant DOT1L reintroduction (f) were irradiated with 100 J m⁻² UVB and then collected at different time points as indicated after UVB irradiation. Genomic DNA was extracted and photoproducts were detected. g, h HPM cells treated with EPZ-5676 or vehicle control (g) or DOT1L depletion or shControl (h) were subjected to 100 J m⁻² UVB under 5 µm micropore filter and were co-stained for CPD and XPC after 0.5 h. Scale bar, 20 μm. i The whole-cell extracts from HPM subjected to 100 J m⁻² UVB were prepared for Co-IP assay to test the interaction of DOT1L with XPC. j The whole-cell extracts from HPM subjected to 100 J m⁻² UVB were prepared for Co-IP assay to test the interaction of XPC with H3K79me2. k HPMs with DOT1L depletion or shControl were exposed to 100 J m⁻² UVB. After 0.5 h, the chromatin fraction was prepared for western blot. *p < 0.05, **p < 0.01, ***p < 0.001, unpaired Student’s t-test. Error bars represent ± s.d.