Leukemia Research Vol. 15, No. 10, pp. 935-941, 1991. Printed in Great Britain.

N-RAS

GENE

MUTATIONS

01,I,5-2126/91 $3.00 + .00 Pergamon Press plc

IN CHILDHOOD

LYMPHOBLASTIC

ACUTE

NON-

LEUKEMIA*

NAOHIRO TERADA,~" THOMAS J. SMITH,:~ LINDA C. STORK,:~ LORRIE F. ODOM:~ and ERWIN W. GELFAND§ tDivision of Basic Sciences, and §Raymond and Beverly Sackler Foundation Laboratory, Department of Pediatrics, National Jewish Center for Immunology and Respiratory Medicine, Denver and Department of Pediatric Hematology/Oncology, University of Colorado and $Children's Hospital of Denver, Denver, U.S.A. (Received 3 September 1990. Revision accepted 13 April 1991)

Abstract--In view of the potential role for ras activation in leukemogenesis, we have screened a number of children with acute non-lymphoblastic leukemia (ANLL) for activating point mutations at codons 12, 13 and 61 of the N-ras proto-oncogene using panels of oligonucleotide probes in conjunction with polymerase chain reaction (PCR) gene amplification. In contrast to the frequent occurrence (-30%) of N-ras mutation reported in adult ANLL, 6 of 46 cases (13%) at the time of diagnosis had N-ras mutations involving codons 12 and 13. In these patients we also determine whether presenting clinical symptoms, cellular pathology, karyotype, or eventual outcome distinguished them from the ras-negative group. N-ras activation tended to be associated with a higher white blood cell count at diagnosis (mean of 225 000/gl vs 91 000/~d) and fewer remissions obtained after 28 days of therapy (3/6, 50% vs 24/32, 75%). It is possible that activation of N-ras oncogene may be involved in the progression of some cases of childhood ANLL. Key words: N-ras activation, childhood ANLL.

explaining the disparity between ras activation in myeloid and lymphoblastic leukemia is that the Nras gene in myeloid cells may be more susceptible to mutagenic agents than in lymphoid cells. However the methylation status of the N-ras gene was found to be the same in both cell types [14]. Another possibility for the disparity in ras mutations could be related to the differing age distribution in these two types of leukemia, as A N L L generally occurs in an older population of patients than does ALL. It is unclear how ras activation may be involved in leukemogenesis. Several reports have demonstrated that in the myelodysplastic syndrome (MDS), patients with N-ras activation may progress more frequently into A N L L with a poorer prognosis than those without activation [8, 15, 16]. However, in none of the previous reports of ras activation in A N L L was any correlation observed between Nras activation and clinical features, other than the suggestion that myelomonocytic morphology was more commonly associated with N-ras mutation [79]. In order to address the role of N-ras activation in childhood leukemia, we have determined the frequency of N-ras mutation in 46 children with ANLL. We also examined the correlation of N-ras mutation with clinical, morphologic, and cytogenetic characteristics.

INTRODUCTION ACTIVATED ras oncogenes have been found in 10. 15% of human neoplasms, although the incidence varies greatly among different types of tumors [1]. The highest incidence of ras gene mutation is seen in some solid tumors such as pancreatic cancer (90%, K-ras) [2, 3], colorectal cancer (50%, mainly K-ras) [4,5] or thyroid cancer (50%, mainly K-ras) [6]. Among leukemias, acute non-lymphoblastic leukemia (ANLL) has the highest incidence of ras activation with 30% of cases showing N-ras mutations [7-9]. In contrast, ras gene mutation has been less frequently detected in acute lymphoblastic leukemia (ALL) (6-20%, mainly N-ras) [10-14]. This difference may be relevant to the role of ras activation in hematopoietic malignancy. One hypothesis * Supported in part by Grants AI-26490 and AI-29704 from the NIH and the Colorado Cancer League. Abbreviations: A L L , acute lymphoblastic leukemia; A N L L , acute non-lymphoblastic leukemia; PCR, polymerase chain reaction; MDS, myelodysplastic syndrome; FAB, French-American-British; WBC, white blood cell. Correspondence to: Erwin W. Gelfand, M.D., Department of Pediatrics, National Jewish Center for Immunology and Respiratory Medicine, 1400 Jackson Street, Denver, CO 80206, U.S.A. 935

936

N. TERADAet al. MATERIALS AND METHODS

Patients and cell samples Forty-six patients under 18 years of age, diagnosed with ANLL at The Children's Hospital of Denver from 1977 through 1989, were studied. The diagnosis of ANLL was made using the French-American-British (FAB) morphologic classification and the cell surface markers. Twenty cases were males and 26 were females. The age at diagnosis of our study population ranged from 6 weeks to 17 years with a mean of 7.3 + 0.7 years. Follow-up was from 1.5 to 10.5 years. At the time of diagnosis, cells were obtained from bone marrow of the patients and subjected to FicollHypaque centrifugation and cryopreservation in DMSO at either -80°C or in liquid nitrogen. All of the samples contained greater than 90% malignant cells. The patients were advised of procedures and attendant risks, and informed consent was obtained in accordance with institutional guidelines. Analysis of N-ras mutation High molecular weight DNA was extracted from the leukemic cells and analyzed for N-ras codon 12, 13 and 61 mutations using PCR and oligonucleotide hybridization methods [17]. One microgram of each DNA was enzymatically amplified by PCR as described elsewhere [18]. To amplify sequences spanning 109 base pairs across codons 12 and 13 of the N-ras gene, 20-mer oligonucleotide primers were synthesized (5'-GACTGAGTACAAACTGGTGG (sense), 5'-CTCTATGGTGGGATCATATT (antisense)). For 103 base pairs across codon 61, other primers were used (5'-GGTGAAACCTGTI'TGTGGA (sense), 5'-ATACACAGAGGAAGCCTTCG (anti-sense)) [19]. One tenth (10~tl) of the PCR-amplified DNA was then slot-blotted onto nylon membranes and hybridized with mutation-specific oligonucleotide probes of N-ras codon 12 and 13 (wild type probe; 5'GTTGGAGCAGGTGGTGTTG, mutation probes for codon 12; 5'GTTGGAGCAnGTGGTGTTG, 5'-GTTGGAGCAGnTGGTGTTG ( n = A,T,C), mutation probes for codon 13; 5'-GGAGCAGGTnGTGTTGGGA, 5'GGAGCAGGTGnTGTI'GGGA (n = A,T,C)), or codon 61 (wild type probe; 5'-ACAGCTGGACAAGAAGAGT, mutation probes; 5'-ACAGCTGGAnAAGAAGAGT (n = G,A), 5'-ACAGCTGGACnAGAAGAGT (n = G,T,C), 5'-ACAGCTGGACAnGAAGAGT (n = T,C)) [7]. These probes are representative of all possible mutations activating each of the codons as well as the wild type. Probes were labeled with [r-32p]ATP by means of T4 polynucleotide kinase. Prehybridization, hybridization and washing of membranes were in solutions containing 3 M tetramethyl ammonium chloride as described [20].

RESULTS N-ras mutations in childhood A N L L Diagnostic bone marrow samples from 46 cases of childhood A N L L were screened for N-ras mutations in codons 12, 13 and 61, using the mixture of mutation-specific probes (Fig. 1). Six of 46 cases contained N-ras mutations in codon 12 (3 cases) or codon 13 (3 cases). No mutations in codon 61 were

detected. When each mutation-specific probe was used, two of the codon 12 amino acid substitutions demonstrated Gly to Ser ( G G T to A G T ) , while the other substitution was Gly to Asp ( G G T to G A T ) (Fig. 2). Two of the codon 13 substitutions showed Gly to Asp ( G G T to G A T ) , while the other substitution was Gly to Val ( G G T to G T I ' ) . Of the six N-ras mutations, five involved a G to A transition, while one had a G to T transition. The amplified DNA sample of case 30 also formed faint but stable hybrids with a second mutation probe (Asp of codon 13). Cases which showed borderline intensity when a mixture of mutation-specific probes were used (Fig. 1), were also examined with each mutation-specific probe of that codon. No mutations were found in these cases. Correlation with clinical, morphological and cytogenetic features The clinical features of patients with an N-ras mutation were compared to those of patients without a mutation. The factors analyzed at the time of diagnosis included age, sex, presenting symptoms, presence of hepatomegaly or splenomegaly, peripheral white blood cell (WBC) count, hematocrit, platelet count, FAB blast morphology, blast karyotype and presence of blasts in cerebrospinal fluid. We also examined the length of time from diagnosis to remission and ultimate outcome (Tables 1,2). There were no differences in the presenting symptoms with both groups of patients generally having some combination of lethargy, malaise, pain, fever, pallor and easy bruising. Using the Fisher exact test for 2 x 2 table (2 tailed) and the Student's t-test, we found no significant differences between the groups in terms of sex, age, presence of hepatomegaly or splenomegaly, platelet count, hematocrit, blasts in CSF, or ultimate outcome. According to FAB criteria, there were no significant distribution differences between cases with N-ras mutation (M2 : 3, M4 : 1, M5 : 1, and M6 : 1) and those without (M1 : 5, M2 : 12, M3 : 4, M4 : i0, M5:8, and M7: 1). None of the cases with N-ras mutation had evidence of terminal deoxyribonucleotide transferase activity within blasts. Among the 6 positive cases, only one case (case 43) had rearrangement of the immunoglobulin heavy chain gene; the five other cases showed a germline configuration. All six cases showed a germline configuration for the T-cell receptor fi gene (data not shown). Although the small number of patients with N-ras mutation prevented adequate statistical analysis, two trends emerged. First, patients with N-ras mutation tended to have a high WBC count at diagnosis. The mean WBC count in the N-ras mutation group

N - 12, 13 w t

N - 12 m u t

N - 13 m u t

FIG. 1 Screening of N-ras gene mutations at codon 12 and 13 in childhood ANLL. PCR-amplified DNA was screened by oligonucleotide probe hybridization, as described in Materials and Methods. Wild type probe for codon 12, 13 of N-ras gene (N-12, 13wt), or mixture of mutation-specific probes for codon 12 (N-12 mut) or 13 (N-13 mut) were used, respectively. As a negative control, human placenta DNA was used (C). For positive controls, KG-1 cell line (12+) and ALL sample DNA previously reported (13+) were used (12). Specimens 11 and 2 are ANLL cases with N-ras 12 and N-ras 13 mutations respectively.

t~q7

N - 12, 13 w t

N - 12Ser

N - 12Asp

N - 13 A s p

N - 13 V a l

C F1G. 2. Hybridization with each mutation-specific oligonucleotide. PCR-amplified DNA of N-ras mutation-positive cases were hybridized with each mutation-specific oligonucleotide, as described in Materials and Methods.

938

N-ras mutations in childhood ANLL

939

TABLE 1. CLINICAL AND LABORATORY CORRELATES OF N-ras GENE ACTIVATION AT DIAGNOSIS

Patients with N-ras mutation (n = 6) Hepatomegaly (>2 cm below right costal margin) Splenomegaly ( palpable spleen) Mean white blood cell count (~l) >125 000/gl Mean platelet count (gl) Mean hematocrit Cytogenetic abnormality FAB classification

Patients without N-ras mutation (n = 40)

3/6(50%)

15/40(37.5%)

3/6(50% )

21/40( 52.5 % )

225 000 _+ 105 000 4/6(67%) 92 000 -+ 29 000 20.5 -+ 3.7 5/6(83%) (5 structural abnormalities; 1 hyperdiploid) M2 = 3(50%) M4 = 1(17%) M5 = 1(17%) M6 = l (17%)

Positive cerebrospinal fluid with blasts (>5 wbc/~tl) >28 days before bone marrow remission Death

91 000 - 20 000 10/40(25%) 89 000 + 21 000 24.7 -+ 1.0 19/32(59%) (13 structural abnormalities; 9 hyperdiploid) M1 -- 5(13%) M2 = 12(30%) M3 = 4(10%) M4 = 10(25%) M5 = 8(20%) M7 = 1(3%)

2/6(33%) 9/34(27% ) 3/6(50%) 8/32(25%) 26/40(65%)

3/6(50%)

TABLE 2. CLINICAL AND LABORATORY FEATURES OF A N L L CASES DEMONSTRATING N-ras MUTATIONS

Patient

Age (yr)/ sex

FAB subtype*

White blood cell count (~tl)

Karyotype abnormalities

2

3,M

M4

586 000

ins(4q)

11

7,F

M2

144000

t(6;9)(p23;p34)

27

9,M

M5

4000

None Hyperdiploid 48XX t(5;11)(q32;q32)

30

I,F

M2

414000

35

7,M

M2

11 000

del (12p)

43

1,F

M6

192 000

del (6q)

Amino acid substitution 13 Aspartate (GGT to CAT) 12 Serine (GGT to ACT) 13 Aspartate (GGT to GAT) 13 Valine (GGT to GTI') possible 13 Aspartate (GGT to CAT) 12 Aspartate (GGT to GAT) 12 Serine (GGT to ACT)

Remission after 28 days of therapy No No Yes

Yes Yes No

* French-American-British classification. (225 000/dl) was more than twice that of the control group (91000/~tl), sixty-seven percent of patients (4/6) with N-ras mutations had W B C counts greater than 125 000/~tl while only twenty-five percent of patients (10/40) did in the remainder. In addition, patients with N-ras mutation tended to have more difficulty obtaining initial remission. Fifty percent (3/6) of these patients failed to obtain bone marrow

remission ( < 5 % blasts) at 28 days of therapy. In the N-ras mutation-negative group, the majority of patients (24/32) achieved remission after 28 days of therapy. DISCUSSION AND CONCLUSION In the present p a p e r we have demonstrated that

940

N. TERADAet al.

N-ras mutations occurred in 6 of 46 patients (13%) with childhood ANLL. Codons 12 (3/6) and 13 (3/6) were involved, and G to A transitions were the most commonly detected base changes (5/6). This frequency of N - r a s mutation in childhood A N L L is considerably lower than the reported frequencies in adult A N L L (14/52; 27% [7], 7/26; 27% [8], 15/57; 26% [9], 38/148; 26% [1]). Since there are no major clinical or histologic differences between childhood and adult A N L L other than age, this lower incidence of mutations may be age related. Indeed, review of previous reports reveals a similar tendency in childhood vs adult ALL: 2/19; 11% [10], 2/15; 13% [12], 6/100; 6% [13] in childhood A L L , 6/33; 18% [11] in childhood and adult A L L , 13/64; 20% [14] in adult ALL. We also found that in childhood acute undifferentiated leukemia the frequency of N-ras mutation is also low (1/10, unpublished observations). The prevalence of N - r a s mutation might subtly increase in hematopoietic precursor cells with ageing. It is also known that some chemical carcinogens cause a specific G to A base transition [21]. Our data and those of others [7-11] demonstrate that G to A transitions are the most common mutations in hematopoietic malignancies, suggesting that some environmental etiology might be involved in this accumulation. In addition, the ability to repair DNA damage might diminish with age and this could cause a higher prevalence of N - r a s mutations in older patients. Indeed, A N L L and A L L seem to have similar frequencies of N - r a s mutations when age is considered. These observations suggest that N - r a s activation does not preferentially occur in myeloblasts as compared to lymphoblasts. It should also be noted mutations of K - r a s and H-ras have also been associated with different forms of leukemia [1}. In a recent publication, 24% of childhood AML demonstrated N - r a s mutations and in 13% a mutation of K-ras. In this group of patients there was no apparent clinical correlation with ras gene mutations

[22]. Activation of ras oncogenes may serve as one of the important steps in carcinogenesis since it can dramatically transform immortalized fibroblasts [23] and initiate chemical mutagen-induced carcinogenesis [24, 25]. However, although ras mutations are frequently found in some tumors, the incidence is always below fifty percent with the exception of adenocarcinoma of the pancreas (90%) [1]. In addition, no reported clinical differences in any tumor type have been found between those with or without a ras mutation. This observation has lead to the suggestion that ras activation must not be an essential event in tumor generation. Although previous reports of A N L L were unable to show any

correlation between ras activation and clinical features, several reports of MDS and A L L suggest that N - r a s activation may be involved in the progression of leukemogenesis [8, 12, 13, 15, 16]. DNA sequencing of ras genes in malignant cells shows a significantly higher frequency of mutations in codons 12, 13 or 61 than in other codons not known to cause cell transformation [11,26, 27]. This suggests that mutation of codons 12, 13 or 61 of ras genes in malignant cells could impart a growth advantage and allow the possible selection of these cells resulting in tumor progression, Our data of high WBC counts at diagnosis and the greater difficulty of obtaining remission, may implicate ras-gene activation in the progression of ANLL. A c k n o w l e d g e m e n t s - - W e thank Drs H. Wilson and L. McGavran for morphologic and cytogenetic review, Lynn Barczuk and Irene Young for technical assistance and Jane Watkins for preparation of the manuscript.

REFERENCES 1. Bos J. L. (1989) Ras oncogenes in human cancer: Review. Cancer Res, 49, 4682. 2. Almoquera C., Shibata D., Forrester K., Martin J., Arnheim N. & Perucho M. (1988) Most human carcinomas of the exocrine pancreas contain mutant c-Kras genes. Cell 53, 549. 3. Smit V. T. H. B. M., Boot A. J. M., Smits A. M. M., Fleuren G. J., Connelisse C. J. & Bos J. L. (1988) K-ras codon 12 mutations occur very frequently in pancreatic adenocarcinomas. Nucleic Acids Res. 16, 7773. 4. Bos J. L., Fearon E. R., Hamilton S. R., Verlaan-de Vries M., van Boom J. H., van tier Eb A. J. & Vogelstein B. (1987) Prevalence of ras gene mutations in human colorectal cancers. Nature 327, 293. 5. Forrester K., Almoquera C., Han K., Grizzle W. E. & Perucho M. (1987) Detection of high incidence of K-ras oncogenes during human colon tumorigenesis. Nature 327, 298. 6. Lemoine N. R., Mayall E. S., Wyllie F. W., Williams E. D., Goyns M., Stringer B. & Wynford-Thomas D. (1989) High frequency of ras oncogene activation in all stages of human thyroid tumorigenesis. Oncogene 4, 159. 7. Farr C. J., Saiki R. K., Ehrlich H. A., McCormick F. & Marshall C. J. (1988) Analysis of ras gene mutations in acute myeloid leukemia using the polymerase chain reaction and oligonucleotide probes. Proc. natn. Acad. Sci. U.S.A. 85, 1629. 8. Yunis J. J., Boot A. J. M., Mayer M. G, & Bos J. L. (1988) Mechanism of ras mutation in myelodysplastic syndrome. Oncogene 4, 609. 9. Bartram C. R., Ludwig W. D., Hiddemann W., Lyons J., Buschle M., Ritter J., Harbott J., Frohlich A. & Janssen J. W. G. (1989) Acute myeloid leukemia: analysis of ras gene mutations and clonality defined by polymorphic X-linked loci. Leukemia 3, 247. 10. Rodenhuis S., Bos J. L., Slater R. M., Behrendt H.,

N-ras mutations in childhood ANLL

van't Veer M, & Smets L. A. (1986) Absence of oncogene amplification and occasional activation of Nras in lymphoblastic leukemia of childhood. Blood 67, 1698--1704. 11. Neri A., Knowles D, M., Greco A., McCormick F. & Dalla-Favera R. (1988) Analysis of ras oncogene mutations in human lymphoid malignancies, Proc. nam. Acad. Sci. U.S.A. 85, 9268. 12. Terada N., Miyoshi J., Kawa-Ha K., Sasai H., Orita S. ,Yumura-Yagi K., Hara J., Fujinami A. & Kakunaga T. (1990) Alteration of N-ras gene mutation after relapse in acute lymphoblastic leukemia. Blood 75, 453. 13. Lubbert M., Mirro J., Miller J. C. W., Kahan J., Isaac G., Kitchingman G., Mertelsman R., Herrmann F., McCormick F. & Koeffler H. P. (1990) N-ras gene point mutations in childhood acute lymphoblastic leukemia correlate with a poor prognosis. Blood 75, 1163. 14. Browett P. J. & Norton J. D. (1989) Analysis of ras gene mutations and methylation state in human leukemias. Oncogene 4, 1029. 15. Hirai H., Kobayashi Y., Mano H., Hagiwara K., Yoshiro M., Omine M., Mizoguchi H., Nishida J. & Takaku F. (1987) A point mutation at codon 13 of the N-ras oncogene in myelodysplastic syndrome. Nature 327, 430. 16. Padua R. A. , Carter G., Hughes D., Gow J., Farr C., Oscier D., McCormick F. & Jacobs A. (1988) Ras mutations in myelodysplasia detected by amplification, oligonucleotide hybridization, and transformation. Leukemia 2, 503. 17. 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 oligonucleotides. Gene 50, 313. 18. Saiki R. K., Gelfand D. H., Stoffel S., Scharf S. V., Higuchi R., Horn G. T., Mullis K. & Erlich H. A. (1988) Primer-directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science 239, 487.

941

19. Taparowsky E,, Shimizu K., Gqldfarb M. & Wigler M. (1983) Structure and activation of the human N-ras gene. Cell 34, 581. 20. Wood W. I., Gitschier J., Lasky L. A. & Lawn R. M. (1985) Base composition-independent hybridization in tetramethylammonium chloride: A method for oligonucleotide screening of highly complex gene libraries. Proc. hath. Acad. Sci. U.S.A. 82, 1585. 21. Zarbl H., Sukumar S., Arthur A. V., Martin-Zanca D. & Barbacid M. (1985) Direct mutagenesis of Haras-1 orlcogenes by N-nitroso-N-methylurea during initiation of mammary carcinogenesis in rats. Nature 315, 382. 22. Vogelstcin B., Civin C. I., Preisinger A. C., Krischer J. P., Steuber P., Ravindranath Y., Weinstein H., Elfferich P. & Bos J. (1990) R A S gene mutations in childhood acute myeloid leukemia: A pediatric oncology group study. Genes Chromosomes Cancer 2, 159. 23. Perucho M., Goldfarb M., Shimizu K., Lama C., Fogh J. & Wigler M. (1981) Human tumor-derived cell lines contain common and different transforming genes. Cell 27, 467. 24. Sukamar S., Notario V., Martin-Zanca D. & Barbacid M. (1983) Induction of mammary carcinomas in rats by nitroso-methylurea involves malignant activation of H-ras-1 locus by single point mutations. Nature 306, 658. 25. Balmain A., Ramsden M., Bowden G. T. & Smith J. (1984) Activation of the mouse cellular Harvey-ras gene in chemically induced benign skin papillomas. Nature 307, 658. 26, Bar-Eli M., Ahuja H., Foti A. & Cline M. J. (1989) N - R A S mutations in T-cell acute lymphoid leukemia: analysis by direct sequencing detects a novel mutation. Br. J. Haemat, 72, 36, 27. Bar-Eli M., Ahuja H., Gonzalez-Cadavid N., Foti A. & Cline M. J. (1989) Analysis of N - R A S Exon-1 mutations in myelodysplastic syndromes by polymerase chain reaction and direct sequencing. Blood 73, 281.