B-Raf and Ha-ras mutations in chemically induced mouse liver tumors

Article metrics

Abstract

The mitogen-activated protein kinase signalling pathway is a central regulator of tumor growth, which is constitutively activated in chemically induced mouse liver tumors. In about 30–50% of cases this effect can be related to activation of the Ha-ras gene by point mutations, whereas in the remaining cases mutations may occur in other members within this pathway, such as Raf kinases. Recently, B-raf has been shown to be frequently mutated in human melanomas and certain other cancers, with a V599E amino-acid change representing the most predominant mutation type. We now screened 82 N-nitrosodiethylamine-induced liver tumors from C3H/He mice for mutations within the hotspot positions in the Ha-ras and B-raf genes. About 50% (39/82) of tumors showed Ha-ras codon 61 mutations and 16 tumors (20%) harbored mutations at codon 624 of the B-raf gene, which corresponds to codon 599 in human B-raf. None of the tumors was mutated in both Ha-ras and B-raf. The high prevalence of Ha-ras and B-raf mutations in mouse liver tumors is in striking contrast to human hepatocellular cancers which very infrequently harbor mutations in the two genes. These fundamental differences between the biology of liver tumors in mice and man may be of toxicological relevance.

Main

Chemically induced hepatocarcinogenesis in rodents is a very useful tool to study critical alterations occurring during tumor development. Neoplasias in mouse liver are characterized by a high frequency of activating mutations in proto-oncogenes of the ras family which code for small membrane-bound GTPases that play an essential role in signal transduction from cell surface receptors to the nucleus (for recent reviews, see Hunter, 1997; Downward, 2003; Malumbres and Barbacid, 2003). The frequency at which ras mutations occur in mouse liver tumors varies between different strains of mice and may exceed 50% depending on strain and treatment conditions (Buchmann et al., 1991; Dragani et al., 1991; Bauer-Hofmann et al., 1992; Maronpot et al., 1995; Schwarz et al., 1995). Activating point mutations in ras genes are almost exclusively found at three codons, namely 12, 13 and 61, and Ha-ras is the most frequently affected ras family member in mouse liver tumors (e.g. see Buchmann et al., 1991; Maronpot et al., 1995; Kalkuhl et al., 1998).

Ras-mediated signal transduction pathways are involved in the regulation of cell growth, apoptosis and differentiation, which constitute fundamental processes that are deregulated in tumors carrying mutated ras genes (Hunter, 1997; Downward, 2003; Malumbres and Barbacid, 2003). Most information on Ras-dependent signalling has been accumulated on the activation of the extracellular signal-regulated kinase (ERK) family of mitogen-activated protein kinases (MAPK), which is mediated via Raf kinases and MEK (MAPK/ERK). The activity of Raf-1 kinase, as determined by measurement of phosphorylation of its downstream target MEK, was found to be increased in mouse liver tumors, but did not significantly differ between ras-mutated and wild-type tumors (Kalkuhl et al., 1998). This suggested that mutations in other genes coding for proteins upstream of MEK can substitute for mutated ras in activation of the MEK/ERK signal transduction pathway.

Three mammalian Raf kinases are known to date, that is, Raf-1, A-Raf and B-Raf, all of which are upstream regulators of the MEK/ERK pathway (reviewed in Hagemann and Rapp, 1999; Mercer and Pritchard, 2003). Very recently, point mutations in the B-raf gene have been observed in more than 60% of human melanoma and, at lower frequencies, in several other human cancers (for a recent review, see Mercer and Pritchard, 2003). Most of the B-raf mutations affect the activation segment of the protein and a V599E substitution within this region is the most predominant mutation type observed in human tumors (Mercer and Pritchard, 2003). In our present study, we comparatively analysed the prevalence of Ha-ras and B-raf mutations in liver tumors induced in mice by a single injection of N-nitrosodiethylamine (DEN) at infancy. The tumor material was available to us from a previous experimental study conducted in our laboratory (Jaworski et al., in preparation). In brief, male C3H/He mice were injected i.p. with a single dose of DEN (10 μg/g of body weight) at 2 weeks of age. After weanling, mice were housed individually in macrolon cages and kept on a standard diet without further treatment. All mice were killed 35 weeks after DEN treatment, and larger liver tumors were excised, immediately frozen in liquid nitrogen and stored at −70°C. In total, 82 tumors with diameters between 3 and 10 mm were used.

For mutation analysis, genomic DNA was extracted from the tumors and the DNA regions containing the hotspot sites for mutations in the Ha-ras gene (codon 61) and B-raf gene (codon 624, which corresponds to human codon 599; see Figure 1c) were amplified by the polymerase chain reaction (PCR) using standard methods (for PCR primers, see Table 1 ). PCR products of both genes were analysed for mutations by two independent methods, that is, pyrosequencing (see below) and denaturing high-performance liquid chromatography (DHPLC; for details, see Xiao and Oefner, 2001). In addition, Ha-ras mutations were verified by restriction fragment length polymorphism (RFLP) analysis and some of the tumors with codon 624 mutations in the B-raf gene were also analysed by classical dideoxynucleotide sequencing (for typical examples, see Figure 2). Pyrosequencing is a novel technology for rapid determination of short DNA sequences, which is used for high-throughput analysis of single-nucleotide polymorphisms (SNP) but is also applicable for mutation analysis if the location and types of mutations are known (for a recent review, see Berg et al., 2002). In pyrosequencing reactions, a sequencing primer is annealed to the template DNA and nucleotides are added step by step in a defined order that allows detection of the respective mutations (for nucleotide dispensation orders, see Figure 1). Nucleotide incorporations during the reaction generate light signals via a series of enzymatic reactions, which are displayed as sequencing pyrograms (for representative pyrograms of Ha-ras and B-raf wild-type and mutated tumors, see Figure 2a). A summary of our data on Ha-ras and B-raf mutations detected by the various methods is given in Table 2. In accordance with previous observations, 50% of liver tumors (39/82) contained Ha-ras codon 61 mutations, with C → A transversions at the first base (c.181C>A) and A → G transitions at the second base (c.182A>G) being the most predominant mutation types. In addition, 16 tumors harbored a T → A transversion at the second base (c.1871T>A) in codon 624 of the B-raf gene, which leads to a Val → Glu amino-acid change. This mutation corresponds to the V599E substitution in human B-raf which is known to strongly enhance B-Raf kinase activity (Davies et al., 2002; Wan et al., 2004). Thus, Ha-ras and B-raf mutations represent the molecular basis for constitutive activation of the MEK/ERK signal transduction pathway in about two-thirds (55 out of 82) of the liver tumors analysed. The actual frequency may even be higher, because we largely restricted our analyses to the known mutational hotspots within the two genes. Rare mutations may be present at other sites in a small fraction of tumors classified as nonmutated in our study and additional mutations may affect other ras genes such as Ki-ras, although DEN-induced mouse liver tumors very infrequently carry mutations other than in Ha-ras codon 61 (Maronpot et al., 1995; Kalkuhl et al., 1998).

Figure 1
figure1

Pyrosequencing of Ha-ras (a) and B-raf mutations (b). Pyrosequencing of PCR products (see Table 1) was performed on a PSQ™96MA System (Biotage AB, Uppsala, Sweden) according to the manufacturer's protocol. Briefly, the sequencing reaction includes the following essential steps: After annealing of the indicated sequencing primers to their single-stranded template DNA, nucleotides are added one at a time in a mutation-type-specific order. Nucleotides that are complementary to the respective template bases are incorporated and each incorporation event yields an enzyme-catalysed light signal that is proportional to the amount of incorporated nucleotides. The sequence can then be determined from the order and intensity of signals in the resulting pyrograms (for typical examples, see Figure 2a). Nucleotide dispensation orders for the different types of mutations are shown along with the expected relative incorporation signals (figures in boxes). Nucleotides marked by an asterisk served as negative controls. Note that mutations in mouse liver tumors were generally heterozygous, which results in signals for both the wild-type and the mutated bases at the respective positions (marked in bold). The two mutational hotspots, Ha-ras codon 61 and B-raf codon 624, are underlined. Mouse codon 624 in the B-raf gene corresponds to human codon 599 as shown in the partial protein sequence of human and mouse B-Raf (c). V599 and V624, respectively, are marked in bold. Additional positions known to be mutated in human tumors are underlined. #The previously updated human B-Raf sequence (accession number M95712.2) contains an additional amino acid. To facilitate comparison with recent literature data, the nonupdated numbering is used

Table 1 PCR amplification of Ha-ras and B-raf regions containing hotspot mutation sites and methods used for mutation analysisa
Figure 2
figure2

Typical examples of B-raf and Ha-ras mutation analyses with mouse liver tumors. (a) Pyrosequencing of B-raf codon 624 (left) and Ha-ras codon 61 (right) mutated and wild-type tumors. For details on conditions of pyrosequencing, see Figure 1. The pyrograms show the inverted sequences, that is, TTG and TCG at Ha-ras codon 61 correspond to CAA and CGA, respectively, whereas CAC and CTC at B-raf codon 624 correspond to GTG and GAG, respectively. Negative control nucleotides are marked by an asterisk and start of the sequencing reaction is indicated by an arrow (). (b) Mutation analysis by DHPLC. DHPLC analysis was conducted on the automated WAVE™ nucleic acid fragment analysis system (Transgenomic Inc., Omaha, USA). PCR products (see Table 1) were denatured for 2 min at 94°C and allowed to re-anneal slowly, which leads to the formation of homoduplices (wild-type tumors) or heteroduplices (tumors with a heterozygous mutation). PCR products were then injected into a preheated (60.5°C for Ha-ras and 58°C for B-raf) reversed phase column based on nonporous poly (styrene/divinylbenzene) particles (DNASep column; Transgenomic), and eluted from the column by a linear acetonitrile gradient at a flow rate of 1.5 ml/min for 2 min. The linear acetonitrile gradient consisted of 47% buffer A (0.1 M triethylammonium acetate (TEAA); pH 7.0) and 53% buffer B (0.1 M TEAA; pH 7.0, containing 25% acetonitrile), and buffer B increased at 5%/min. The DHPLC elution profiles give a single peak for wild-type tumors (homoduplices) and a double peak for tumors with a heterozygous mutation (heteroduplices). Examples of a Ha-ras-mutated and a wild-type tumor are shown; similar profiles were obtained with B-raf-mutated tumors. (c) Verification of Ha-ras codon 61 mutations by RFLP analysis. Three of the Ha-ras codon 61 mutations analysed generate new restriction enzyme recognition sites which can be detected by the following enzymes (mutated codon 61 sequences are given in parenthesis): TaqI (CGA), XbaI (CTA) and BspHI (CAT; positive control sample from a previous experiment), whereas the mutated sequence AAA leads to the loss of a Hpy188III recognition site. PCR products (see Table 1) were digested with the respective restriction enzymes and separated on 10% polyacrylamide gels. Mutation-specific fragments are marked by arrows, and two mutation-independent fragments (TaqI) are indicated by an asterisk. M, size marker (pBR322, MspI-digest). (d) Dideoxynucleotide sequencing of B-raf-mutated liver tumors. Examples of a B-raf wild-type tumor (left) and a tumor with a codon 624 mutation (right) are shown

Table 2 Prevalence of Ha-ras and B-raf mutations in mouse liver tumors

Deregulation of the Ras signalling pathway is causally linked to tumor development and thus constitutes an interesting target for cancer therapy (Downward, 2003). The transforming effects of Ras are mediated via activation of downstream effector pathways in which Raf proteins are involved differentially. Studies with Raf knockout mice show that Raf-1-deficient mice die during embryogenesis with abnormalities in the placenta and increased apoptosis in various organs including liver, which differs from the phenotypes of A-Raf- and B-Raf-deficient mice (reviewed in O'Neill and Kolch, 2004). Interestingly, cell proliferation was not impaired and MEK/ERK activation was normal in the Raf-1-deficient mice (Hüser et al., 2001; Mikula et al., 2001). The latter effect was probably mediated by B-Raf, indicating that B-Raf can substitute for MEK/ERK-related effects on cell cycle progression, but not for Raf-1-mediated protection from apoptosis, which must be related to effectors other than MEK/ERK (Hüser et al., 2001; Mikula et al., 2001). In this context, it is of interest to note that mouse liver tumors with and without Ha-ras mutations showed no obvious differences in cell proliferation as estimated by BrdU labelling, and phosphorylation of MEK by Raf-1 kinase was equally elevated in both tumor types in comparison to normal liver tissue (Kalkuhl et al., 1998). Since liver tumors in that study were induced under treatment conditions which were almost identical to the present study, a certain fraction of tumors very likely contained activating B-raf mutations. This would explain, at least in part, why ras wild-type tumors showed labelling indices comparable to their mutated cousins. Apoptotic cell death, however, was much lower in Ha-ras-mutated than in Ha-ras wild-type tumors (Frey et al., 2000), indicating that the activated Ras protein was capable to block apoptosis of ras-transformed mouse hepatocytes, whereas the activated B-raf protein may lack this activity.

B-Raf mutants can be subdivided into two groups, those with high/intermediate and those with low/impaired kinase activity, both of which are able to activate Raf-1 and the MEK/ERK pathway. Whereas low-activity mutants require Raf-1 to phosphorylate MEK, Raf-1 is dispensable in high-activity mutants which are able to do so directly (Wan et al., 2004). Interestingly, high-activity B-raf mutations such as the one leading to the V599E substitution usually do not coincide with ras mutations in human tumors, indicating that they can fully substitute each other in terms of their oncogenic potential (Davies et al., 2002). Other B-raf mutations, however, may be complementary to ras mutations in certain tumor types, for example, colorectal cancers (Davies et al., 2002; Yuen et al., 2002). In accordance with these observations, none of the B-raf codon-624-mutated mouse liver tumors contained Ha-ras mutations.

In synthesis, our results show that the previously observed constitutive activation of the Raf-1/MEK/ERK signal transduction pathway in mouse liver tumors (Kalkuhl et al., 1998) can be attributed to mutations in the Ha-ras or B-raf genes in about two-thirds of cases. This pathway is mainly involved in the regulation of transcription of genes coding for proteins involved in cell cycle progression (Chang et al., 2003). Enhanced proliferation of mouse hepatocytes with either of the two mutations is therefore reasonably explained by activation of this pathway. It remains puzzling, however, which pathways confer protection from apoptosis selectively to Ha-ras-mutated tumor cells that have much lower apoptotic rates than Ha-ras wild-type cells (Frey et al., 2000). There are several reports indicating that Ras may mediate antiapoptotic effects via activation of the phosphatidyl-inositol-3-kinase (PI3K) pathway, activation of nuclear factor kappaB (Downward, 1998) or stimulation of translocation of Raf-1 to the mitochondrial membrane, where the kinase may interact with Bcl-2 family proteins (Wang et al., 1994, 1996). Comparatively little, however, is known about the role of B-Raf in apoptosis regulation. In one study, overexpression of B-Raf has been shown to protect fibroblasts against cell death induced by growth factor withdrawal through MEK/ERK-dependent inhibition of caspase activation, an effect which was independent of PI3K (Erhardt et al., 1999). Whether similar effects are also relevant for hepatocyte apoptosis is presently unknown. It will therefore be interesting to determine in future studies how mutated B-Raf affects apoptosis in mouse liver tumors and other cancers.

In our present study, Ha-ras and B-raf mutations were analysed in liver tumors from C3H/He mice, which are highly susceptible to liver carcinogenesis. Previous studies from our and other laboratories have indicated that the frequency of ras mutations varies between different strains of mice, and liver tumors from genetically resistant mice or rats very infrequently carry Ha-ras codon 61 mutations (Buchmann et al., 1991; Dragani et al., 1991; Maronpot et al., 1995; Schwarz et al., 1995). It is not known at present whether the same variability also applies to the occurrence of B-raf mutations in liver tumors. Apart from these strain-dependent differences, several fundamental discrepancies at the molecular level are also present with regard to hepatocellular tumors in humans. Ras mutations, for example, are almost completely absent in human hepatocellular cancers, whereas mutations in the p53 tumor suppressor gene are frequently observed in human hepatocellular cancers but not in mouse liver tumors (for a summary, see Schwarz et al., 1995). Our present results add a further divergence at the molecular level, since B-raf mutations which were present in 20% of liver tumors from C3H/He mice were undetectable in human hepatocellular carcinoma (Tannapfel et al., 2003). These differences are of toxicological relevance with regard to interspecies extrapolation of carcinogenicity data, raising the question of whether or not rodent carcinogenicity tests with highly susceptible strains are suitable surrogates for risk extrapolation to humans.

References

  1. Bauer-Hofmann R, Klimek F, Buchmann A, Müller O, Bannasch P and Schwarz M . (1992). Mol. Carcinogen., 6, 60–67.

  2. Berg LM, Sanders R and Alderborn A . (2002). Expert Rev. Mol. Diagn., 2, 361–369.

  3. Buchmann A, Bauer-Hofmann R, Mahr J, Drinkwater NR, Luz A and Schwarz M . (1991). Proc. Natl. Acad. Sci. USA, 88, 911–915.

  4. Chang F, Steelman LS, Shelton JG, Lee JT, Navolanic PM, Blalock WL, Franklin R and McCubrey JA . (2003). Int. J. Oncol., 22, 469–480.

  5. Davies H, Bignell GR, Cox C, Stephens P, Edkins S, Clegg S, Teague J, Woffendin H, Garnett MJ, Bottomley W, Davis N, Dicks E, Ewing R, Floyd Y, Gray K, Hall S, Hawes R, Hughes J, Kosmidou V, Menzies A, Mould C, Parker A, Stevens C, Watt S, Hooper S, Wilson R, Jayatilake H, Gusterson BA, Cooper C, Shipley J, Hargrave D, Pritchard Jones K, Maitland N, Chenevix Trench G, Riggins GJ, Bigner DD, Palmieri G, Cossu A, Flanagan A, Nicholson A, Ho JW, Leung SY, Yuen ST, Weber BL, Seigler HF, Darrow TL, Paterson H, Marais R, Marshall CJ, Wooster R, Stratton MR and Futreal PA . (2002). Nature, 417, 949–954.

  6. Downward J . (1998). Curr. Opin. Genet. Dev., 8, 49–54.

  7. Downward J . (2003). Nat. Rev. Cancer, 3, 11–22.

  8. Dragani TA, Manenti G, Colombo BM, Falvella FS, Gariboldi M, Pierotti MA and Della Porta G . (1991). Oncogene, 6, 333–338.

  9. Erhardt P, Schremser EJ and Cooper GM . (1999). Mol. Cell. Biol., 19, 5308–5315.

  10. Frey S, Buchmann A, Bursch W, Schulte-Hermann R and Schwarz M . (2000). Carcinogenesis, 21, 161–166.

  11. Hagemann C and Rapp UR . (1999). Exp. Cell Res., 253, 34–46.

  12. Hunter T . (1997). Cell, 88, 333–346.

  13. Hüser M, Luckett J, Chiloeches A, Mercer K, Iwobi M, Giblett S, Sun XM, Brown J, Marais R and Pritchard C . (2001). EMBO J., 20, 1940–1951.

  14. Kalkuhl A, Troppmair J, Buchmann A, Stinchcombe S, Buenemann CL, Rapp UR, Kaestner K and Schwarz M . (1998). Hepatology, 27, 1081–1088.

  15. Malumbres M and Barbacid M . (2003). Nat. Rev. Cancer, 3, 459–465.

  16. Maronpot RR, Fox T, Malarkey DE and Goldsworthy TL . (1995). Toxicology, 101, 125–156.

  17. Mercer KE and Pritchard CA . (2003). Biochim. Biophys. Acta, 1653, 25–40.

  18. Mikula M, Schreiber M, Husak Z, Kucerova L, Rüth J, Wieser R, Zatloukal K, Beug H, Wagner EF and Baccarini M . (2001). EMBO J., 20, 1952–1962.

  19. O'Neill E and Kolch W . (2004). Br. J. Cancer, 90, 283–288.

  20. Schwarz M, Buchmann A, Stinchcombe S, Luebeck G, Moolgavkar S and Bock KW . (1995). Mutat. Res., 333, 69–79.

  21. Tannapfel A, Sommerer F, Benicke M, Katalinic A, Uhlmann D, Witzigmann H, Hauss J and Wittekind C . (2003). Gut, 52, 706–712.

  22. Wan PT, Garnett MJ, Roe SM, Lee S, Niculescu Duvaz D, Good VM, Jones CM, Marshall CJ, Springer CJ, Barford D and Marais R . (2004). Cell, 116, 855–867.

  23. Wang H-G, Miyashita T, Takayama S, Sato T, Torigoe T, Krajewski S, Tanaka S, Hovey III L, Troppmair J, Rapp UR and Reed JC . (1994). Oncogene, 9, 2751–2756.

  24. Wang HG, Rapp UR and Reed JC . (1996). Cell, 87, 629–638.

  25. Xiao W and Oefner PJ . (2001). Hum. Mutat., 17, 439–474.

  26. Yuen ST, Davies H, Chan TL, Ho JW, Bignell GR, Cox C, Stephens P, Edkins S, Tsui WW, Chan AS, Futreal PA, Stratton MR, Wooster R and Leung SY . (2002). Cancer Res., 62, 6451–6455.

Download references

Acknowledgements

The excellent technical assistance of Mrs E Zabinsky and Mr M Niwar is acknowledged. We also thank Thomas Franck for his support in dHPLC analysis. This study was supported by the Deutsche Forschungsgemeinschaft (SCHW 329/3-1) to MS.

Author information

Correspondence to Michael Schwarz.

Rights and permissions

Reprints and Permissions

About this article

Keywords

  • mouse hepatocarcinogenesis
  • B-Raf
  • Ha-ras
  • mutation analysis
  • pyrosequencing

Further reading