British Iournaf of Huemutofogy. 1992, 81, 370-373
Detection of allele-specific expression of N-RAS oncogenes in human leukaemia cells MICHAELLUBBERT,ROLAND MERTELSMANN AND FRIEDHELM HERRMANN Department of HaematologylOncology, University of Freiburg Medical Centre, Germany Received 19 September 1991; accepted for publication 24 February 1992
Summary. We report analysis of allele-specific expression of N-KAS transcripts in myeloid leukaemic cells and cell lines. Expression was assessed by an assay of reverse-transcription/ PCR combined with differential hybridization with mutationspecific oligonucleotides. In cells from all patients with acute myeloid leukaemia ( A M ) and in myeloid cell lines HL-60
and THP-1, expression of both the wild-type allele and the abnormal allele altered by a point mutation could be detected, albeit not always at comparable levels. This might be due for instance to allelic exclusion. The assay described provides a means of analysing the degree of expression of dominant oncogenes.
RAS genes code for cytoplasmic 2 1 kD proteins (p2 1) that share several similarities with G-proteins and probably are similarly involved in intracellular signal transduction pathways (reviewed by Barbacid. 1987).Single basepair substitutions at critical codons of H-. K- and N-RAS genes (almost always codons 12, 13 and 61) result in oncogenes with transforming ability in vitro. It has become clear that the transforming effect of RAS oncogenes is dose-dependent (Paterson et al, 1987: Cohen & Levinson. 1988). Furthermore, additional mutations in a RAS gene can result in strong upregulation of its expression (Cohen & Levinson, 1988). Point mutations resulting in an activated p21 have been detected in numerous human malignancies. In almost all instances, these mutations were heterozygous (for a review, see Bos, 1988).Mutations of N-RAS occur predominantly in neoplasias of haematopoietic cells (Janssen et al, 1987: Neri et al, 1988: Liibbert et al. 1990). Most of the more recent studies examining cancer cells for the presence of activating point mutations of RAS protooncogenes have employed amplification of the critical sequences of these genes from DNA template by use of the polymerase chain reaction (PCR) (Saiki er al, 1986) in conjunction with differential oligonucleotide hybridization (Verlaan-de-Vries et al, 1986) or DNA sequencing. When using this assay, one assumes that the mutated RAS allele detected in the malignant cells is also actively transcribed by these cells. However, the assay does not yield any information
about the gene product dosage of the oncogene. Expression of N-RAS mRNA is hardly detectable or is undetectable by Northern blot technique in primary AML cells (Mavilio et al. 1986: Liibbert et al. unpublished results). In addition, previously described approaches for differentiating, at the RNA (Winter et al. 1985) or protein (Shen et al, 1987) level, between gene products of RAS alleles which differ by a single nucleotide are relatively time-consuming, are not specific for defined nucleotide substitutions and are not applicable for small amounts of cells as starting material. In the present study we wished to examine the expression of both normal and mutated N-RAS alleles in myeloid leukaemic cells which in 15-30% have mutations of N-RAS (Bos, 1988). We therefore adapted the PCR of cDNA reverse transcribed from N-RAS mRNA (RT-PCR)and oligonucleotidehybridization so that we were able to differentiate between normal and mutated transcripts. MATERIALS A N D METHODS Cells and cell lines. Leukaemic cells isolated from peripheral blood or bone marrow of 50 patients with AML had been screened for N-RAS point mutations on the DNA level as described below. Peripheral blood mononuclear leukaemic cells from six patients were shown to carry monoclonal, isolated and heterozygous point mutations at either exon I or I1 of N-RAS (Liibbert et al, 1992). Cells from five of these patients as well as myeloid cell lines HL-60 and THP-I (known N-RAS point mutations) were further analysed for allele-specific expression. Cell preparations of primary peripheral blood samples contained >90% leukaemic blasts as
Correspondence: Dr M. Liibbert, Department of Haematology/ Oncology, University of Freiburg Medical Centre, Hugstetter Str. 5 5 , D-7800 Freiburg, Germany
3 70
N-RAS Oncogenes in Human Leukaemia Cells
N-ras Gene
- RS 50
PCR Product from DNA
371
c]
RS61
115 bp
4
ll2bp
RS49
N-ras mRNA PCR Product from cDNA
4
b 210 bp
Pig I , Schematic strategy of in-vitro amplification of genomic sequences (upperpanel)and cDNA synthesized from mRNA (lowerpanel) of N-RAS. Boxes indicate exons: oligonucleotide primers used for PCR are indicated. RS49: CTGGTGTGAAATGACTGAGT:RS50: GGTGGGATCATATTCATCTA: RS 60: GTTATAGATGGTGAAACCTG:RS61: ATACACAGAGGAAGCCTTCCG. assessed by Wright-Giemsa staining and visual inspection. In cells from patients 2 9 and 39. exon I of N-KAS was mutated at nucleotides 1 and 2 of codon 12, respectively. Cells from patients 12, 1 7 and 2 3 carried mutations of N-RAS exon 11. codon 61, nucleotides 1, 2 and 3 , respectively. THP-1 has a mutation at N12p2; HL-60 is altered at N61p2 (Janssen et al, 1987: Bos et a / . 1984). For several of the patients' samples, only up to 2 x 1 0 " cells were available. Reverse-transcriptase polymerase chain reaction. Total RNA was extracted by the method of Chirgwin et a1 (1979). In cases where only small numbers of cells were available, cytoplasmic KNA was extracted by lysis of cells in isotonic high pH buffer (IHB: 140 mM NaCI, 1 0 mM Tris-HCI. pH 8.0. 1.5 mM MgC12)containing 0.5%, (v/v) NP-40 (Sigma) and 0 . 1 %diethyl pyrocarbonate (DEP: Sigma). After vortexing for 30 s. nuclei were spun down by centrifugation at 6000 rpm for 3 0 s. Supernatants were transferred to a second microfuge tube and placed at 37OC for 2 0 min. DEP was degraded by a 10 min incubation at 90°C. precipitates pelleted by centrifugation at 1 3 000 rpm for 2 min. Supernatant was then used for the KT-PCK reaction. I pg of total RNA or 10 pl of cytoplasmic R N A was transcribed into cDNA for 1 h at 42°C in the presence of 2 0 0 IJ MoMuLV reverse transcriptase (BRL, Gaithersburg, Md.), 1 LJ RNAsin (Boehringer Mannheim, Germany), 1 0 0 p~ hexamer (Pharmacia, Stockholm, Sweden), 1 0 nmol of dNTPS (Boehringer Mannheim). 50 mM KCI. 20 mM Tris-HCI, pH 8.4. 2.5 mM MgC12 and 1 0 pg/ml bovine serum albumin. After cDNA synthesis, RT reaction mixture was expanded to 100 pl final volume and in vitro amplification was performed with 20-2 5 cycles using oligonucleotides RS49 and RS61 (PCK strategy depicted in Fig 1 ). PCH amplification from DNA template was carried out essentially as described previously (Lubbert et nl. 1 990). Since point-mutated and wild-type alleles of a RAS gene are not necessarily present in equimolar amounts in a given leukaemic cell population (Farr et al, 1987). samples had first been analysed for mono-, sub- or oligoclonality of all possible NKAS mutations within codons 12 and 13 using differential
hybridizations of DNA oligos to PCR products amplified from DNA template. Monoclonality was defined by three criteria: (i) similar signal intensity of mutation-specific hybridization and the respective wild-type hybridization signal at similar exposure times of the films, (ii) similar hybridization signal with a mutation-specific probe of the sample of interest and a cell line with a known RAS mutation. (iii) lack of signal with other mutation-specific probes, thus ruling out other isolated missense point mutations at codons 12 and 13. Subsequently, only cell samples with apparent monoclonality were chosen for the analysis described. Blotting and hybridization techniques. The 2 70 bp RT-PCR products encompassing N-RAS exons I and I1 were electrophoresed ( 1 0 pl) on agarose gels together with 1 1 5 bp or 1 12 bp PCR product of either exon 1 or 2 of N-RAS. respectively (depending on localization of the activating point mutation). This internal control had been amplified from genomic DNA isolated from the same cells. Nucleic acids were blotted onto nylon-based membrane (Zetaprobe. BioRad Laboratories, Richmond, Calif.) by alkaline transfer and sequentially hybridized to oligonucleotides complementary to either unmutated sequence or to the mutated sequences present in these cells (oligonucleotide sequences available upon request). Hybridizations took place in the presence of 5 x SSPE ( 1 x SSPE= 10 mM sodium phosphate. 180 mM NaCI, 1 mM EDTA, pH 7.0),5 x Denhardt'ssolution, 100pg/mldenatured sonicated salmon sperm DNA, 1 0 mM EDTA and 1% sodium dodecyl sulphate (SDS) at 63°C for probes complementary to N-RAS exon I and 59°C for probes complementary to N-RAS exon 11. Filters were then rinsed for 5 min at the same temperatures in 5 x SSPE, 0.1 % SDS. In order to be able to compare relative transcript levels, exposure of filters to X-AR 5 film was equalized for hybridization signal of the 115 or 1 1 2 bp (DNA template) PCR products. After hybridizations, probes were removed by rinsing of filters a t 65°C in 0.1 x SSPE/l% SDS for 15-30 rnin. Great care was taken to use identical hybridization and washing conditions for probes differing in a single nucleotide.
372
M. Liibbert, R. Mertelsmann and F. Herrmann
Fig 2. Detection of allele-specific mRNA expression in myeloid leukaemic cells carrying isolated heterozygous activating mutations of N-RAS. Abbreviations: nl PBL. peripheral blood lymphocytes from a healthy donor: pt. 12, an AML patient whose cells carry a mutation at the first nucleotide of N-RAS codon 61; pt. 39, mutation at the second nucleotide of N-RAS codon 12; pt. 29, mutation at the first nucleotide of N-RAS codon 12: pt. 17, mutation at the second nucleotide ofcodon 61 of N-RAS. Filters were sequentially hybridized and rinsed under high-stringency conditions, first with the wild-type-specific probe for N-RAS exon I (GGAGCACGTGGTGTTGCCAA) (nl PBL. pt. 29. 39) or exon I1 (TACTC?TC?TGTCCAGCTCT)(pt. 12, 17). then with the mutation-specific probes. Lane ‘nl PBL/M‘: Rlter was hybridized to a probe detecting a point mutation at the second nucleotide of codon 12 of N-RAS.
RESULTS Allele-specific expression of N-RAS was examined in leukaemic cells with monoclonal activation of this gene as described in Materials and Methods, as well as in cell lines with known point mutations of this gene. RT-PCR, size fractionation by gel electrophoresisand Southern blot transfer were combined with differential hybridization. Under the conditions used, only amplification of the target sequence was observed. This may be due to the fact that the PCR product expected to amplify across intron 1 of N-RAS would be relatively much larger (2.5 kb) than the spliced product (0.27 kb) or due to the fact that the RNA template preparations generated by ultracentrifugation or by hypotonic lysis of cells were quite free of contaminating DNA. The PCR strategy is schematized in Fig 1. We found that both wild-type N-RAS transcripts (lanes ‘N’ in Fig 2 ) and mutated transcripts (lanes ‘M’)were expressed in all cases of acute myeloid leukaemia (AML) studied. Similar results were seen in the HL-60 and THP-I cell lines (blotsnot shown). When comparing the amounts of Table I. Relative expression levels of N-RAS oncogenes and proto-oncogenes in myeloid cells. The ratio of hybridizing bands was determined by visual inspection. Abbreviations: p, nucleotide position: mut., mutated allele: wt. wild-type allele. Cells
Point mutation
Ratio mut. :wt
Pt 29 Pt 39 THP- 1 Pt 12 Pt 1 7 HL60 Pt 23
Nl2pl N12p2 N12p2 N61pl N6 1p2 N61p2 N6lp3
3:l 5: 1
3: 1 1:l 1:lO 1:l 1:l
amplified transcripts of either allele after equalization of autoradiographic signal of the internal control (PCR product amplified from genomic DNA), in repeated experiments transcript levels were similar in some samples but variable in others. For instance, in cells from patient 39. levels of transcripts encoded by the mutated allele were about 5-fold higher as judged by visual inspection. In contrast, mutated transcripts were underrepresented about 10-fold in patient 17 in spite of the presence of > 90% leukaemic blasts in the cell preparation. These results are summarized in Table I. Since mixtures of two or three oligonucleotide probes were used to detect nucleotide substitutions at a given position, we cannot completely exclude the possibility of subclonality through different base pair substitutions at the same nucleotide confounding the results.
DISCUSSION Thus far. few studies have addressed the question of relative expression of dominant RAS oncogenes compared to the remaining wild-type allele in primary human malignant cells (Winter et al, 1985: Shen et al, 1987). Our results demonstrate that the mutated, transforming allele of N-RAS present in AMLs with heterozygous mutations of this gene was transcribed in both primary and cultured leukaemic cells. The other, normal allele was never completely silenced in the presence of the altered transcript. However, transcript levels of the two alleles were not equal in all cases. This observation is compatible with the notion of RAS genes being dominant oncogenes. Gene dosage of N-RAS oncogenes has been shown to be of importance for the transforming effect of this gene (Paterson et al, 1987). One might speculate that the oncogenic RAS in some cases down-regulates its normal counterpart by a mechanism similar to that of allelic exclusion described for the altered MYC oncogene in Burkitt’s lymphoma (Ar-Rushdie et al. 1983). There is an important precedent for differential expression of RAS oncogenes in vivo:
N-RAS Oncogenes in Human Leukaemia Cells one H-KAS allele of the T24/EJ bladder carcinoma cells is overexpressed 10-fold compared to the normal allele. This is due to a n additional mutation in intron 4 of the oncogenic allele (H-KAS ValiL)(Cohen & Levinson. 1988) which results in changes of the rate of alternative splicing of this oncogene (Cohen p t a!. 1989). Overexpression of the mutated allele in turn is associated with a higher transforming ability in vitro. The observation that cells from one of our patients (No. 17) apparently underexpressed the mutated allele c. 10-fold is of special interest, since it suggests that the oncogene encoded by the mutated allele might be particularly powerful in cellular transformation. The assay of RT-PCR adapted for a differential hybridization described by us requires only small amounts of RNA and does not depend o n a high degree of intactness of RNA. It should be useful for the semiquantitative detection of single heterozygous point mutations on the mRNA level, for example in studies where active transcription of these mutated genes over the background of the unmutated allele is examined. This is the case for instance in transfection studies using a mutated gene differing only by a single nucleotide from the normal, intrinsic gene expressed by the transfected cells. However, this assay obviously does not yield information about translational or post-translational means of regulation which might also influence levels of RAS proteins. In conclusion, transcripts coding for N-RAS oncogenes are expressed at levels similar to those of the normal allele in some myeloid leukaemic cells but overexpressed or underexpressed in others. This might reflect regulation mechanisms whereby the oncogene and proto-oncogene interact as well as variations in the transforming ability of different N-RAS oncogenes. ACKNOWLEDGMENTS Supported by a grant from Deutsche Krebshilfe (W 56/88/ Os2).
REFERENCES Ar-Rushdi. A,. Nishikura. K.. Erickson. J.. Watt. K.. Rovera. G. & Croce. C. (1983) Differential expression of the translocated and untranslocated c-myc oncogene in Burkitt lymphoma. Science. 222, 390-393. Barbacid. M. (1987) ras genes. Annual Review of Biochemistry. 56. 779-827. Bos. J.L. ( 1988) The ras gene family and human carcinogenesis. Mirtatiorr Reseurch. 195. 255-271. Bos. 1.1 ... Verlaan-de Vries. M.. Jansen. A.M., Veeneman. G.H.. van Boom. J.H. & van der Eb, A.J. (1984)Three different mutations in codon 6 1 of the human N-ras gene detected by synthetic oligonucleotide hybridization. Nuclric Acids Research. 12, 9 1 5 59163.
3 73
Chirgwin. J.M.. Przybyla. A.E.. MacDonald. R.J. & Rutter, W.J. (1979) Isolation of biologically active ribonucleic acids from sources rich in ribonuclease. Biochemistry. 18, 5294-5299. Cohen. J.B.. Broz, S.D. & Levinson. A.D. (1989) Expression of the Hras proto-oncogeneis controlled by alternative splicing. Cell. 58, 461-472. Farr. C.J.. Saiki. R.K.. Erlich. H.A., McCormick, F. & Marshall, C.J. ( 1 987) Analysis of ras gene mutations in acute myeloid leukemia by polymerase chain reaction and oligonucleotide probes. Proceedings of the National Academy of Sciences of the United States of America. 85, 1629-1632. Janssen. J.W.G.. Steenvoorden.A.C.M.. Lyons, J.. Anger, B., Bohlke, J.U.. Bos. J.L.. Seliger. H. & Bartram. C.R. (1987) Ras gene mutations in acute and chronic myelocytic leukemias, chronic myeloproliferative disorders, and myelodysplastic syndromes. Proceedings ofthe National Academy ofsciences ofthe United States of America, 84, 9228-923 1 . Lubbert, M., Mirro. J.. Jr. Miller, C.W.. Kahan. J.. Isaac, G.. Kitchingman. G.. Mertelsmann. R.. Herrmann. F.. McCormick. F. & Koeffler, H.P. (1990) N-ras gene point mutations in childhood acute lymphocytic leukemia correlate with a poor prognosis. Blood. 7 5 , 1163-1169. Liibbert. M.. Oster, W.. Knopf. H.-P.. McCormick.F.. Mertelsmann.R. & Herrmann. F. (1992) Clonal analysis of N-ras gene activation and its association with expression of Interleukin-6 in myeloid leukemia. journal of Clinical Investigation, in press. Mavilio, F.. Sposi, N.M.. Petrini. M.. Bottero. L., Marinucci. M.. De Rossi. G.. Amadori. S . . Mandelli. F. & Peschle. C. (1986) Expression of cellular oncogenes in primary cells from human acute leukemias. Proceedings of the National Academy of Sciences of the United States of America. 83. 4394-4398. Neri. A.. Knowles. D.M.. Greco. A.. McCormick.F. & Dalla-Favera. R. (1988)Analysis of RAS oncogene mutations in human lymphoid malignancies. Proceedings ofthe National Academy of Sciences ofthe United States of America, 85, 9268-9271. Paterson, H.. Reeves, B.. Brown, R.. Hall, A., Furth. M., Bos, J., Jones. P. & Marshall, C. (1987) Activated N-ras controls the transformed phenotype of HT1080 human fibrosarcoma cells. Cell, 51, 803-81 2. Saiki. R.K.. Bugawan, T.L.. Horn, G.T.. Mullis. K.B. & Erlich. H.A. ( 1 986) Analysis of enzymatically amplified beta-globin and HLADQ-alpha DNA with allele-specific oligonucleotide probes. Nature. 324, 163-165. Shen. W.P.V.. Aldrich. T.H.. Venta-Perez. G.. Franza. B.R.. Jr& Furth. M.E. (1987) Expression of normal and mutant ras proteins in human acute leukemia. Oncogene. 1 , 157-165. Verlaan-de Vries. M.. Bogaard. M.E.. van den Elst. H.. van Boom. J.H.. van der Eb. A.J. & Bos. J.L. (1986) A dot-blot screening procedure for mutated ras oncogenes using synthetic oligodeoxynucleotides. Gene. 50, 313-320. Winter, E.. Yamamoto, F., Almoguera. C. & Perucho. M. (1985) A method to detect and characterire point mutations in transcribed genes: amplification and overexpression of the mutant c-Ki-ras allele in human tumor cells. Proceedings of the National Academy of Scierires ofthe United States of America. 82, 7575-7579.
Detection of allele-specific expression of N-RAS oncogenes in human leukaemia cells MICHAELLUBBERT,ROLAND MERTELSMANN AND FRIEDHELM HERRMANN Department of HaematologylOncology, University of Freiburg Medical Centre, Germany Received 19 September 1991; accepted for publication 24 February 1992
Summary. We report analysis of allele-specific expression of N-KAS transcripts in myeloid leukaemic cells and cell lines. Expression was assessed by an assay of reverse-transcription/ PCR combined with differential hybridization with mutationspecific oligonucleotides. In cells from all patients with acute myeloid leukaemia ( A M ) and in myeloid cell lines HL-60
and THP-1, expression of both the wild-type allele and the abnormal allele altered by a point mutation could be detected, albeit not always at comparable levels. This might be due for instance to allelic exclusion. The assay described provides a means of analysing the degree of expression of dominant oncogenes.
RAS genes code for cytoplasmic 2 1 kD proteins (p2 1) that share several similarities with G-proteins and probably are similarly involved in intracellular signal transduction pathways (reviewed by Barbacid. 1987).Single basepair substitutions at critical codons of H-. K- and N-RAS genes (almost always codons 12, 13 and 61) result in oncogenes with transforming ability in vitro. It has become clear that the transforming effect of RAS oncogenes is dose-dependent (Paterson et al, 1987: Cohen & Levinson. 1988). Furthermore, additional mutations in a RAS gene can result in strong upregulation of its expression (Cohen & Levinson, 1988). Point mutations resulting in an activated p21 have been detected in numerous human malignancies. In almost all instances, these mutations were heterozygous (for a review, see Bos, 1988).Mutations of N-RAS occur predominantly in neoplasias of haematopoietic cells (Janssen et al, 1987: Neri et al, 1988: Liibbert et al. 1990). Most of the more recent studies examining cancer cells for the presence of activating point mutations of RAS protooncogenes have employed amplification of the critical sequences of these genes from DNA template by use of the polymerase chain reaction (PCR) (Saiki er al, 1986) in conjunction with differential oligonucleotide hybridization (Verlaan-de-Vries et al, 1986) or DNA sequencing. When using this assay, one assumes that the mutated RAS allele detected in the malignant cells is also actively transcribed by these cells. However, the assay does not yield any information
about the gene product dosage of the oncogene. Expression of N-RAS mRNA is hardly detectable or is undetectable by Northern blot technique in primary AML cells (Mavilio et al. 1986: Liibbert et al. unpublished results). In addition, previously described approaches for differentiating, at the RNA (Winter et al. 1985) or protein (Shen et al, 1987) level, between gene products of RAS alleles which differ by a single nucleotide are relatively time-consuming, are not specific for defined nucleotide substitutions and are not applicable for small amounts of cells as starting material. In the present study we wished to examine the expression of both normal and mutated N-RAS alleles in myeloid leukaemic cells which in 15-30% have mutations of N-RAS (Bos, 1988). We therefore adapted the PCR of cDNA reverse transcribed from N-RAS mRNA (RT-PCR)and oligonucleotidehybridization so that we were able to differentiate between normal and mutated transcripts. MATERIALS A N D METHODS Cells and cell lines. Leukaemic cells isolated from peripheral blood or bone marrow of 50 patients with AML had been screened for N-RAS point mutations on the DNA level as described below. Peripheral blood mononuclear leukaemic cells from six patients were shown to carry monoclonal, isolated and heterozygous point mutations at either exon I or I1 of N-RAS (Liibbert et al, 1992). Cells from five of these patients as well as myeloid cell lines HL-60 and THP-I (known N-RAS point mutations) were further analysed for allele-specific expression. Cell preparations of primary peripheral blood samples contained >90% leukaemic blasts as
Correspondence: Dr M. Liibbert, Department of Haematology/ Oncology, University of Freiburg Medical Centre, Hugstetter Str. 5 5 , D-7800 Freiburg, Germany
3 70
N-RAS Oncogenes in Human Leukaemia Cells
N-ras Gene
- RS 50
PCR Product from DNA
371
c]
RS61
115 bp
4
ll2bp
RS49
N-ras mRNA PCR Product from cDNA
4
b 210 bp
Pig I , Schematic strategy of in-vitro amplification of genomic sequences (upperpanel)and cDNA synthesized from mRNA (lowerpanel) of N-RAS. Boxes indicate exons: oligonucleotide primers used for PCR are indicated. RS49: CTGGTGTGAAATGACTGAGT:RS50: GGTGGGATCATATTCATCTA: RS 60: GTTATAGATGGTGAAACCTG:RS61: ATACACAGAGGAAGCCTTCCG. assessed by Wright-Giemsa staining and visual inspection. In cells from patients 2 9 and 39. exon I of N-KAS was mutated at nucleotides 1 and 2 of codon 12, respectively. Cells from patients 12, 1 7 and 2 3 carried mutations of N-RAS exon 11. codon 61, nucleotides 1, 2 and 3 , respectively. THP-1 has a mutation at N12p2; HL-60 is altered at N61p2 (Janssen et al, 1987: Bos et a / . 1984). For several of the patients' samples, only up to 2 x 1 0 " cells were available. Reverse-transcriptase polymerase chain reaction. Total RNA was extracted by the method of Chirgwin et a1 (1979). In cases where only small numbers of cells were available, cytoplasmic KNA was extracted by lysis of cells in isotonic high pH buffer (IHB: 140 mM NaCI, 1 0 mM Tris-HCI. pH 8.0. 1.5 mM MgC12)containing 0.5%, (v/v) NP-40 (Sigma) and 0 . 1 %diethyl pyrocarbonate (DEP: Sigma). After vortexing for 30 s. nuclei were spun down by centrifugation at 6000 rpm for 3 0 s. Supernatants were transferred to a second microfuge tube and placed at 37OC for 2 0 min. DEP was degraded by a 10 min incubation at 90°C. precipitates pelleted by centrifugation at 1 3 000 rpm for 2 min. Supernatant was then used for the KT-PCK reaction. I pg of total RNA or 10 pl of cytoplasmic R N A was transcribed into cDNA for 1 h at 42°C in the presence of 2 0 0 IJ MoMuLV reverse transcriptase (BRL, Gaithersburg, Md.), 1 LJ RNAsin (Boehringer Mannheim, Germany), 1 0 0 p~ hexamer (Pharmacia, Stockholm, Sweden), 1 0 nmol of dNTPS (Boehringer Mannheim). 50 mM KCI. 20 mM Tris-HCI, pH 8.4. 2.5 mM MgC12 and 1 0 pg/ml bovine serum albumin. After cDNA synthesis, RT reaction mixture was expanded to 100 pl final volume and in vitro amplification was performed with 20-2 5 cycles using oligonucleotides RS49 and RS61 (PCK strategy depicted in Fig 1 ). PCH amplification from DNA template was carried out essentially as described previously (Lubbert et nl. 1 990). Since point-mutated and wild-type alleles of a RAS gene are not necessarily present in equimolar amounts in a given leukaemic cell population (Farr et al, 1987). samples had first been analysed for mono-, sub- or oligoclonality of all possible NKAS mutations within codons 12 and 13 using differential
hybridizations of DNA oligos to PCR products amplified from DNA template. Monoclonality was defined by three criteria: (i) similar signal intensity of mutation-specific hybridization and the respective wild-type hybridization signal at similar exposure times of the films, (ii) similar hybridization signal with a mutation-specific probe of the sample of interest and a cell line with a known RAS mutation. (iii) lack of signal with other mutation-specific probes, thus ruling out other isolated missense point mutations at codons 12 and 13. Subsequently, only cell samples with apparent monoclonality were chosen for the analysis described. Blotting and hybridization techniques. The 2 70 bp RT-PCR products encompassing N-RAS exons I and I1 were electrophoresed ( 1 0 pl) on agarose gels together with 1 1 5 bp or 1 12 bp PCR product of either exon 1 or 2 of N-RAS. respectively (depending on localization of the activating point mutation). This internal control had been amplified from genomic DNA isolated from the same cells. Nucleic acids were blotted onto nylon-based membrane (Zetaprobe. BioRad Laboratories, Richmond, Calif.) by alkaline transfer and sequentially hybridized to oligonucleotides complementary to either unmutated sequence or to the mutated sequences present in these cells (oligonucleotide sequences available upon request). Hybridizations took place in the presence of 5 x SSPE ( 1 x SSPE= 10 mM sodium phosphate. 180 mM NaCI, 1 mM EDTA, pH 7.0),5 x Denhardt'ssolution, 100pg/mldenatured sonicated salmon sperm DNA, 1 0 mM EDTA and 1% sodium dodecyl sulphate (SDS) at 63°C for probes complementary to N-RAS exon I and 59°C for probes complementary to N-RAS exon 11. Filters were then rinsed for 5 min at the same temperatures in 5 x SSPE, 0.1 % SDS. In order to be able to compare relative transcript levels, exposure of filters to X-AR 5 film was equalized for hybridization signal of the 115 or 1 1 2 bp (DNA template) PCR products. After hybridizations, probes were removed by rinsing of filters a t 65°C in 0.1 x SSPE/l% SDS for 15-30 rnin. Great care was taken to use identical hybridization and washing conditions for probes differing in a single nucleotide.
372
M. Liibbert, R. Mertelsmann and F. Herrmann
Fig 2. Detection of allele-specific mRNA expression in myeloid leukaemic cells carrying isolated heterozygous activating mutations of N-RAS. Abbreviations: nl PBL. peripheral blood lymphocytes from a healthy donor: pt. 12, an AML patient whose cells carry a mutation at the first nucleotide of N-RAS codon 61; pt. 39, mutation at the second nucleotide of N-RAS codon 12; pt. 29, mutation at the first nucleotide of N-RAS codon 12: pt. 17, mutation at the second nucleotide ofcodon 61 of N-RAS. Filters were sequentially hybridized and rinsed under high-stringency conditions, first with the wild-type-specific probe for N-RAS exon I (GGAGCACGTGGTGTTGCCAA) (nl PBL. pt. 29. 39) or exon I1 (TACTC?TC?TGTCCAGCTCT)(pt. 12, 17). then with the mutation-specific probes. Lane ‘nl PBL/M‘: Rlter was hybridized to a probe detecting a point mutation at the second nucleotide of codon 12 of N-RAS.
RESULTS Allele-specific expression of N-RAS was examined in leukaemic cells with monoclonal activation of this gene as described in Materials and Methods, as well as in cell lines with known point mutations of this gene. RT-PCR, size fractionation by gel electrophoresisand Southern blot transfer were combined with differential hybridization. Under the conditions used, only amplification of the target sequence was observed. This may be due to the fact that the PCR product expected to amplify across intron 1 of N-RAS would be relatively much larger (2.5 kb) than the spliced product (0.27 kb) or due to the fact that the RNA template preparations generated by ultracentrifugation or by hypotonic lysis of cells were quite free of contaminating DNA. The PCR strategy is schematized in Fig 1. We found that both wild-type N-RAS transcripts (lanes ‘N’ in Fig 2 ) and mutated transcripts (lanes ‘M’)were expressed in all cases of acute myeloid leukaemia (AML) studied. Similar results were seen in the HL-60 and THP-I cell lines (blotsnot shown). When comparing the amounts of Table I. Relative expression levels of N-RAS oncogenes and proto-oncogenes in myeloid cells. The ratio of hybridizing bands was determined by visual inspection. Abbreviations: p, nucleotide position: mut., mutated allele: wt. wild-type allele. Cells
Point mutation
Ratio mut. :wt
Pt 29 Pt 39 THP- 1 Pt 12 Pt 1 7 HL60 Pt 23
Nl2pl N12p2 N12p2 N61pl N6 1p2 N61p2 N6lp3
3:l 5: 1
3: 1 1:l 1:lO 1:l 1:l
amplified transcripts of either allele after equalization of autoradiographic signal of the internal control (PCR product amplified from genomic DNA), in repeated experiments transcript levels were similar in some samples but variable in others. For instance, in cells from patient 39. levels of transcripts encoded by the mutated allele were about 5-fold higher as judged by visual inspection. In contrast, mutated transcripts were underrepresented about 10-fold in patient 17 in spite of the presence of > 90% leukaemic blasts in the cell preparation. These results are summarized in Table I. Since mixtures of two or three oligonucleotide probes were used to detect nucleotide substitutions at a given position, we cannot completely exclude the possibility of subclonality through different base pair substitutions at the same nucleotide confounding the results.
DISCUSSION Thus far. few studies have addressed the question of relative expression of dominant RAS oncogenes compared to the remaining wild-type allele in primary human malignant cells (Winter et al, 1985: Shen et al, 1987). Our results demonstrate that the mutated, transforming allele of N-RAS present in AMLs with heterozygous mutations of this gene was transcribed in both primary and cultured leukaemic cells. The other, normal allele was never completely silenced in the presence of the altered transcript. However, transcript levels of the two alleles were not equal in all cases. This observation is compatible with the notion of RAS genes being dominant oncogenes. Gene dosage of N-RAS oncogenes has been shown to be of importance for the transforming effect of this gene (Paterson et al, 1987). One might speculate that the oncogenic RAS in some cases down-regulates its normal counterpart by a mechanism similar to that of allelic exclusion described for the altered MYC oncogene in Burkitt’s lymphoma (Ar-Rushdie et al. 1983). There is an important precedent for differential expression of RAS oncogenes in vivo:
N-RAS Oncogenes in Human Leukaemia Cells one H-KAS allele of the T24/EJ bladder carcinoma cells is overexpressed 10-fold compared to the normal allele. This is due to a n additional mutation in intron 4 of the oncogenic allele (H-KAS ValiL)(Cohen & Levinson. 1988) which results in changes of the rate of alternative splicing of this oncogene (Cohen p t a!. 1989). Overexpression of the mutated allele in turn is associated with a higher transforming ability in vitro. The observation that cells from one of our patients (No. 17) apparently underexpressed the mutated allele c. 10-fold is of special interest, since it suggests that the oncogene encoded by the mutated allele might be particularly powerful in cellular transformation. The assay of RT-PCR adapted for a differential hybridization described by us requires only small amounts of RNA and does not depend o n a high degree of intactness of RNA. It should be useful for the semiquantitative detection of single heterozygous point mutations on the mRNA level, for example in studies where active transcription of these mutated genes over the background of the unmutated allele is examined. This is the case for instance in transfection studies using a mutated gene differing only by a single nucleotide from the normal, intrinsic gene expressed by the transfected cells. However, this assay obviously does not yield information about translational or post-translational means of regulation which might also influence levels of RAS proteins. In conclusion, transcripts coding for N-RAS oncogenes are expressed at levels similar to those of the normal allele in some myeloid leukaemic cells but overexpressed or underexpressed in others. This might reflect regulation mechanisms whereby the oncogene and proto-oncogene interact as well as variations in the transforming ability of different N-RAS oncogenes. ACKNOWLEDGMENTS Supported by a grant from Deutsche Krebshilfe (W 56/88/ Os2).
REFERENCES Ar-Rushdi. A,. Nishikura. K.. Erickson. J.. Watt. K.. Rovera. G. & Croce. C. (1983) Differential expression of the translocated and untranslocated c-myc oncogene in Burkitt lymphoma. Science. 222, 390-393. Barbacid. M. (1987) ras genes. Annual Review of Biochemistry. 56. 779-827. Bos. J.L. ( 1988) The ras gene family and human carcinogenesis. Mirtatiorr Reseurch. 195. 255-271. Bos. 1.1 ... Verlaan-de Vries. M.. Jansen. A.M., Veeneman. G.H.. van Boom. J.H. & van der Eb, A.J. (1984)Three different mutations in codon 6 1 of the human N-ras gene detected by synthetic oligonucleotide hybridization. Nuclric Acids Research. 12, 9 1 5 59163.
3 73
Chirgwin. J.M.. Przybyla. A.E.. MacDonald. R.J. & Rutter, W.J. (1979) Isolation of biologically active ribonucleic acids from sources rich in ribonuclease. Biochemistry. 18, 5294-5299. Cohen. J.B.. Broz, S.D. & Levinson. A.D. (1989) Expression of the Hras proto-oncogeneis controlled by alternative splicing. Cell. 58, 461-472. Farr. C.J.. Saiki. R.K.. Erlich. H.A., McCormick, F. & Marshall, C.J. ( 1 987) Analysis of ras gene mutations in acute myeloid leukemia by polymerase chain reaction and oligonucleotide probes. Proceedings of the National Academy of Sciences of the United States of America. 85, 1629-1632. Janssen. J.W.G.. Steenvoorden.A.C.M.. Lyons, J.. Anger, B., Bohlke, J.U.. Bos. J.L.. Seliger. H. & Bartram. C.R. (1987) Ras gene mutations in acute and chronic myelocytic leukemias, chronic myeloproliferative disorders, and myelodysplastic syndromes. Proceedings ofthe National Academy ofsciences ofthe United States of America, 84, 9228-923 1 . Lubbert, M., Mirro. J.. Jr. Miller, C.W.. Kahan. J.. Isaac, G.. Kitchingman. G.. Mertelsmann. R.. Herrmann. F.. McCormick. F. & Koeffler, H.P. (1990) N-ras gene point mutations in childhood acute lymphocytic leukemia correlate with a poor prognosis. Blood. 7 5 , 1163-1169. Liibbert. M.. Oster, W.. Knopf. H.-P.. McCormick.F.. Mertelsmann.R. & Herrmann. F. (1992) Clonal analysis of N-ras gene activation and its association with expression of Interleukin-6 in myeloid leukemia. journal of Clinical Investigation, in press. Mavilio, F.. Sposi, N.M.. Petrini. M.. Bottero. L., Marinucci. M.. De Rossi. G.. Amadori. S . . Mandelli. F. & Peschle. C. (1986) Expression of cellular oncogenes in primary cells from human acute leukemias. Proceedings of the National Academy of Sciences of the United States of America. 83. 4394-4398. Neri. A.. Knowles. D.M.. Greco. A.. McCormick.F. & Dalla-Favera. R. (1988)Analysis of RAS oncogene mutations in human lymphoid malignancies. Proceedings ofthe National Academy of Sciences ofthe United States of America, 85, 9268-9271. Paterson, H.. Reeves, B.. Brown, R.. Hall, A., Furth. M., Bos, J., Jones. P. & Marshall, C. (1987) Activated N-ras controls the transformed phenotype of HT1080 human fibrosarcoma cells. Cell, 51, 803-81 2. Saiki. R.K.. Bugawan, T.L.. Horn, G.T.. Mullis. K.B. & Erlich. H.A. ( 1 986) Analysis of enzymatically amplified beta-globin and HLADQ-alpha DNA with allele-specific oligonucleotide probes. Nature. 324, 163-165. Shen. W.P.V.. Aldrich. T.H.. Venta-Perez. G.. Franza. B.R.. Jr& Furth. M.E. (1987) Expression of normal and mutant ras proteins in human acute leukemia. Oncogene. 1 , 157-165. Verlaan-de Vries. M.. Bogaard. M.E.. van den Elst. H.. van Boom. J.H.. van der Eb. A.J. & Bos. J.L. (1986) A dot-blot screening procedure for mutated ras oncogenes using synthetic oligodeoxynucleotides. Gene. 50, 313-320. Winter, E.. Yamamoto, F., Almoguera. C. & Perucho. M. (1985) A method to detect and characterire point mutations in transcribed genes: amplification and overexpression of the mutant c-Ki-ras allele in human tumor cells. Proceedings of the National Academy of Scierires ofthe United States of America. 82, 7575-7579.
