Expression of endogenous oncogenic V600EB-raf induces proliferation and developmental defects in mice and transformation of primary fibroblasts

Abstract

Mutations of the human B-RAF gene are detected in approximately 8% of cancer samples, primarily in cutaneous melanomas (70%). The most common mutation (90%) is a valine-to-glutamic acid mutation at residue 600 (V600E; formerly V599E according to previous nomenclature). Using a Cre/Lox approach, we have generated a conditional knock-in allele of (V600E)B-raf in mice. We show that widespread expression of (V600E)B-Raf cannot be tolerated in embryonic development, with embryos dying approximately 7.5 dpc. Directed expression of mutant (V600E)B-Raf to somatic tissues using the IFN-inducible Mx1-Cre mouse strain induces a proliferative disorder and bone marrow failure with evidence of nonlymphoid neoplasia of the histiocytic type leading to death within 4 weeks of age. However, expression of mutant B-Raf does not alter the proliferation profile of all somatic tissues. In primary mouse embryonic fibroblasts, expression of endogenous (V600E)B-Raf induces morphologic transformation, increased cell proliferation, and loss of contact inhibition. Thus, (V600E)B-Raf is able to induce several hallmarks of transformation in some primary mouse cells without evidence for the involvement of a cooperating oncogene or tumor suppressor gene.

Figure 1

A, generation of LSL-B-raf^V600E and Lox-B-raf^V600E alleles. The targeting vector contains left and right arms with the right arm containing exon 15 with the T1799A mutation (*). These two arms are separated by the LSL cassette. This cassette contains three LoxP sequences (black arrows), a minigene (MG) encoding exons 15 to 18 of wild-type B-Raf with a splice acceptor (SA) sequence at the 5′end. Two STOP sequences, represented by polyadenylation (PA) sequences, are located at the 3′end of the minigene and at the 3′end of the neo^R cassette. Homologous recombination between the targeting vector and the wild-type B-raf gene in embryonic stem cells generated the LSL-B-raf^V600E allele. Expression of the Cre recombinase allows deletion of the LSL cassette and generation of the Lox-B-raf^V600E allele. Location of primers used for PCR genotyping (arrows and A-C). B, PCR genotyping to detect LSL-B-raf^V600E, WT B-raf, and Cre alleles. Primers A-E were used in combination on tail DNA samples from intercrosses between heterozygous LSL-B-raf^V600E mice and Cre mice. In this example, the WT B-raf allele (466 bp), Cre allele (350 bp), and LSL-B-raf^V600E allele (140 bp) are analyzed in tail DNAs from heterozygous LSL-B-raf^V600E mice without Cre (lane 1), wild-type mice with Cre (lane 2), and heterozygous LSL-B-raf^V600E mice with Cre (lane 3). C, PCR genotyping to detect Lox-B-raf^V600E and WT B-raf alleles. Primers A and C were used on the same DNA samples described above in (B). A 518-bp product indicates the presence of the Lox-B-raf^V600E, and a 466-bp product indicates the presence of the WT B-raf allele. D, photographs of E7.5 embryos resulting from LSL-B-raf^V600E × CMV-Cre intercrosses showing LSL-B-raf^V600E − CMV-Cre embryo (left) and LSL-B-raf^V600E + CMV-Cre embryo (right). Bar, 250 μm.

Figure 2

Analysis of mice resulting from Lox-B-raf^V600E × Mx1-Cre intercrosses. A, % survival of LSL-B-raf^V600E + Mx1-Cre animals versus LSL-B-raf^V600E − Mx1-Cre animals at various ages after birth. B, PCR genotyping to detect Lox-B-raf^V600E allele in DNA from various tissues derived from LSL-B-raf^V600E + Mx1-Cre mice. C, splenomegaly in LSL-B-raf^V600E + Mx1-Cre mice. Left, increase in size of the spleen of P21 animals. Right, pooled weight data from four spleens of LSL-B-raf^V600E ± Mx1-Cre mice at P21. LSL-B-raf^V600E + Mx1-Cre mice have a mean spleen weight of 0.2 g compared with 0.04 g for the control LSL-B-raf^V600E − Mx1-Cre mice. Columns, means; bars, SD. D, histologic analysis of spleen sections from LSL-B-raf^V600E ± Mx1-Cre mice. Spleen sections were either stained with H&E, with antibodies for Ki67 or phospho-ERK, or processed for TUNEL analysis. Bar, 50 μm (the same for all sections). E, proliferation analysis of spleens from LSL-B-raf^V600E ± Mx1-Cre mice. Mice were injected with BrdUrd, splenocytes were isolated, and the % incorporation of BrdUrd was assessed by FACS analysis. Columns, means; bars, SD. F, Northern blot analysis of spleen RNA isolated from LSL-B-raf^V600E ± Mx1-Cre mice. Northern blots were probed with the markers indicated. G, histologic analysis of spleen sections from LSL-B-raf^V600E ± Mx1-Cre mice with a histiocyte marker. Spleen sections were stained with a Mac-2 antibody to detect histiocyte amplification. Bar, 50 μm.

Figure 3

Analysis of tissues from mice resulting from LSL-B-raf^V600E × Mx1-Cre intercrosses. A, histologic analysis of liver sections from LSL-B-raf^V600E ± Mx1-Cre mice. Liver sections were either stained with H&E, with a Ki67 antibody, with a phospho-ERK antibody or processed for TUNEL analysis. Right, H&E section was photographed at higher magnification (×400) to indicate site of hemopoiesis (black arrow) in the mutant liver. Bar, 50 μm (the same for all sections) except for 25 μm (enlarged H&E section, right). B, analysis of bone marrow from LSL-B-raf^V600E ± Mx1-Cre mice. Bone marrow cells were isolated from the femurs of LSL-B-raf^V600E ± Mx1-Cre mice and analyzed by the Coulter counter. There are far fewer RBC and WBC in the bone marrow of LSL-B-raf^V600E + Mx1-Cre mice (shaded area) than LSL-B-raf^V600E − Mx1-Cre mice (clear area). PCR analysis of DNA from bone marrow of LSL-B-raf^V600E + Mx1-Cre animal, indicating the presence of the Lox-B-raf^V600E allele. C, photograph of histiocyte arising on skin of LSL-B-raf^V600E + Mx1-Cre animal. D, histologic analysis of histiocyte sections from LSL-B-raf^V600E ± Mx1-Cre mice. Sections were either stained with H&E, with antibodies for Ki67, phospho-ERK, or Mac-2, or processed for TUNEL analysis. Bar, 50 μm.

Figure 4

Analysis of B-RAF/MEK/ERK activation in tissues derived from LSL-B-raf^V600E ± Mx1-Cre mice. Protein lysates were harvested from various tissues and subjected to (A) B-Raf kinase assays and (B) Western blot analysis with antibodies for phospho-ERK, phospho-MEK, and cyclin D1. Immunoblotting with a total ERK antibody was used to confirm protein loading.

Figure 5

Analysis of primary MEFs derived from LSL-B-raf^V600E mice. A, morphologic transformation. Primary LSL-B-raf^V600E MEFs were either mock-transfected (−Cre) or transfected with the pCrePac plasmid (+Cre). After 48 hours, the cells were either photographed (top) or subjected to immunofluorescence analysis with FITC-phalloidin (bottom). B, proliferation analysis. Mock- and Cre-transfected cells were serum starved for 20 hours (open columns) and then stimulated with serum for 16 hours (gray columns). % BrdUrd incorporation was assessed under each condition. Columns, means; bars, SD. The cells transfected with Cre undergo a high level of proliferation even in the absence of serum compared with mock-transfected controls. C, Western blot analysis of lysates prepared from Cre- and mock-transfected MEFs with an antibody for phospho-ERK. Immunoblotting with a total ERK antibody was used to confirm protein loading. D, Western blot analysis of lysates prepared from Cre- and mock-transfected MEFs with antibodies for cyclin D1 and p21^CIP1. Immunoblotting with an antibody for actin was used to confirm protein loading. E, focus formation assay of Cre- and mock-transfected LSL-B-raf^V600E primary MEFs. MEFs were grown to confluency and the media was changed every 2 to 3 days until foci formed for 18 days. F, soft agar assays of Cre- and mock-transfected LSL-B-raf^V600E primary MEFs.