Transcriptional control analyses of the Xiphophorus melanoma oncogene

Abstract

Melanoma development in interspecific hybrids of Xiphophorus is induced by the overexpression of the mutationally activated receptor tyrosine kinase Xmrk in pigment cells. Based on the melanocyte specificity of the transcriptional upregulation, a pigment cell-specific promoter region was postulated for xmrk, the activity of which is controlled in healthy purebred fish by the molecularly still unidentified regulator locus R. However, as yet the xmrk promoter region is still poorly characterized. In order to contribute to a better understanding of xmrk expression regulation, we performed a functional analysis of the entire putative gene regulatory region of the oncogene using conventional plasmid-based reporter systems as well as a newly established method employing BAC-derived luciferase reporter constructs in melanoma and non-melanoma cell lines. Using the melanocyte-specific mitfa promoter as control, we could demonstrate that our in vitro system is able to reliably monitor regulation of transcription through cell type-specific regulatory sequences. We found that sequences within 200kb flanking the xmrk oncogene do not lead to any specific transcriptional activation in melanoma compared to control cells. Hence, xmrk reporter constructs fail to faithfully reproduce the endogenous transcriptional regulation of the oncogene. Our data therefore strongly indicate that the melanocyte-specific transcription of xmrk is not the consequence of pigment cell-specific cis-regulatory elements in the promoter region. This hints at additional regulatory mechanisms involved in transcriptional control of the oncogene, thereby suggesting a key role for epigenetic mechanisms in oncogenic xmrk overexpression and thereby in tumor development in Xiphophorus.

Keywords: Cis-regulatory element; EGF receptor; Melanoma; Pigment cell; Transcriptional control; Tumor suppressor; Xiphophorus; xmrk oncogene.

Figure 1. Transcriptional activity of mitfa promoter sequences in Xiphophorus melanoma and non-melanoma cell lines

Reporter constructs containing the firefly luciferase gene under control of the mitfa promoter from Oryzias latipes were transiently transfected into the Xiphophorus melanoma cell line PSM (A, C) and the non-melanoma cell line A2 (B, D). Reporter constructs containing the firefly luciferase gene without promoter (empty) or under control of either the Herpes simplex virus thymidine kinase promoter (TK) or the TK promoter derived TATA box element (TATA) were transfected as reference. Activity of firefly luciferase was normalized to activity of a cotransfected vector containing the renilla luciferase gene under control of either the CMV (pRL-CMV) or the HSV TK (pGL4.74) promoter. For (A) and (B) promoter sequences were inserted into pLUC+ and pRL-CMV was used for normalization, for (C) and (D) vectors pGL4.20 (firefly luciferase) and pGL4.74 (renilla luciferase) were used.

Figure 2. Analysis of endogenous expression levels in Xiphophorus cell lines

Quantitative real-time PCR analysis of (A) mitfa, (B) xmrk and (C) egfrb transcript levels in the melanoma cell line PSM and in non-pigment cell lines. For mitfa and egfrb, endogenous expression was determined in the X. hellerii derived embryonic epithelial cell line A2. This cell line was also used for reporter gene assays. However, as A2 cells do not contain the xmrk gene, endogenous expression of xmrk was determined in the X. maculatus derived epithelial-like cell line SdSr24.

Figure 3. Functional analysis of the proximal 9.3 kb of the putative xmrk^B promoter region

Transcriptional activity of luciferase reporter constructs containing a set of 5′ deletions spanning the proximal 9.3 kb of the xmrk^B upstream region (A) was measured following transfection into the Xiphophorus melanoma cell line PSM (B) and the non-melanoma cell line A2 (C). Reporter constructs containing the firefly luciferase gene without promoter (empty) or under control of either the Herpes simplex virus thymidine kinase promoter (TK) or the TK promoter derived TATA box element (TATA) were transfected as reference. All firefly luciferase reporter constructs were generated by inserting promoter sequences into pLUC+. For normalization of firefly luciferase activity, pRL-CMV containing the renilla luciferase gene Rluc under control of the CMV promoter was cotransfected. Data are presented as mean ± standard deviation of two independently performed luciferase experiments.

Figure 4. Transcriptional activity of selected xmrk^B promoter fragments determined using pGL4 luciferase reporter vectors

Firefly luciferase reporter constructs containing selected xmrk^B promoter deletion fragments (A) were transfected into the Xiphophorus melanoma cell PSM (B) and the non-melanoma cell line A2 (C). Reporter constructs containing the luc2 firefly luciferase gene without promoter (empty) or under control of either the TK promoter (TK) or the TK promoter derived TATA box (TATA) were transfected as reference. All firefly luciferase reporter constructs were generated by inserting promoter sequences into pGL4.20. For normalization of firefly luciferase activity, pGL4.74 containing the renilla luciferase gene hRluc under control of the HSV TK promoter was cotransfected.

Figure 5. Transcriptional activity of BAC derived reporter constructs containing up to 200 kb xmrk^B flanking sequences in cells of melanoma and non-melanoma origin

(A) Schematic drawing of the full length xmrk^B BAC reporter construct (BAC-xmrkB-LUC+) and various deletion constructs containing between 67 kb and 9.3 kb xmrk^B 5′ flanking sequences. The full length reporter construct was generated by inserting the firefly luciferase gene luc+ together with a selection marker at the translational start site of xmrk^B. Deletion constructs were generated subsequently by BAC recombineering. Transcriptional activity of full length and deletion constructs was measured in the Xiphophorus melanoma cell line PSM (B, D) and in the X. hellerii embryonic fibroblast cell line A2 (C, E). For normalization, BAC reporter constructs containing the renilla luciferase gene Rluc under control of the CMV promoter were cotransfected. As differences in the transfection rates between BAC constructs of different size can lead to distortion of luciferase data, the cotransfected renilla luciferase BACs had the same size as the corresponding firefly luciferase reporter constructs. BAC-xmrkB-LUC+, BAC-67kbxmrkB-LUC+, BAC-28kbxmrkB-LUC+, BAC-19kbxmrkB-LUC+ and BAC-9.3kbxmrkB-LUC+ were cotransfected with BAC-RLUC, BAC-RLUC-85kb, BAC-LUC-40kb, BAC-RLUC-30kb and BAC-RLUC-20kb, respectively.

Figure 6. Transcriptional activity of xmrk^B, xmrk^A and egfrb BAC reporter constructs in Xiphophorus melanoma and non-melanoma cells

(A) Schematic drawing of the full length xmrk^B, xmrk^A and egfrb BAC reporter constructs (BAC-xmrkB-LUC+, BAC-xmrkA-LUC+ and BAC-egfrb-LUC+). The BAC derived reporter constructs were generated by inserting the firefly luciferase gene luc+ together with a selection marker at the translational start site of xmrk or egfrb. Transcriptional activity of reporter constructs was measured in (B) the Xiphophorus melanoma cell line PSM and (C) the X. hellerii embryonic epithelial cell line A2. For normalization, a BAC reporter construct containing the renilla luciferase gene Rluc under control of the CMV promoter (BAC-RLUC) was cotransfected.