Introduction

Cytogenetic studies have been frequently performed in childhood acute lymphoblastic leukemia (ALL), and have identified nonrandom translocations and deletions (Raimondi, 1993; Harrison, 2001). Like other tumors, ALL results from accumulation of genetic alterations that affect normal control of cellular growth. Malignant transformation can arise as a result of the increased activity of growth-promoting genes (oncogenes), and/or the inactivation or loss of growth-constraining genes (tumor-suppressor genes). Inactivation or deletion of tumor-suppressor genes is a pivotal pathway of leukemogenesis in childhood ALL. The paradigm of inactivation of tumor-suppressor genes is mutation of one allele and loss of the second allele. This reduction to homozygosity in the region of the tumor-suppressor gene can be detected as a loss of heterozygosity (LOH) of informative markers in the region of the tumor-suppressor gene. Thus, LOH analysis is an indirect method to search for an inactivated tumor-suppressor gene.

Approximately 70–80% of patients with childhood ALL are cured by chemotherapy. However, if the disease relapses, the outcome is poor, with a 6-year survival rate of 20% (Gaynon et al., 1998). Therefore, defining the biologic and genetic basis for relapse is important. Relapsed patients are often sensitive to the same drug as used in the initial treatment, suggesting that relapsed leukemic cells retain many of the same biologic features as were present at initial presentation. In contrast, some patients have additional chromosomal abnormalities at relapse (Heerema et al., 1992). Cell lines are established more easily from the patients at relapse than at the time of initial diagnosis. Moreover, relapsed ALL cells engraft in scid mice more rapidly than ALL cells obtained at the time of initial diagnosis (Kamel-Reid et al., 1991). These findings suggest that additional genetic changes are associated with the progression of ALL.

A large number of genetic studies has been performed in childhood ALL. We and others have previously done an extensive LOH analysis on childhood ALL using microsatellite markers, and have found that LOH of 6q, 9p, 11q, and 12p are frequent in childhood ALL (Takeuchi et al., 1995b, 1996, 1997a, 1998, 1999). LOH of the short arm of chromosome 9 occurred in 40% of children with ALL, and further analysis showed that deletions of the p16/INK4a gene at 9p21 occurred in about 15% of children with precursor-B ALLs and 75% of those with T-ALLs (Okuda et al., 1995; Takeuchi et al., 1995a, 1997a). LOH of the short arm of chromosome 12 with concomitant TEL-AML1 fusion occurred in 16–33% of childhood B-lineage ALL (Shurtleff et al., 1995; Stegmaier et al., 1995; Mclean et al., 1996; Takeuchi et al., 1996, 1997b). An LOH of 6q was reported in 15% of childhood ALL at diagnosis (Gérard et al., 1997; Takeuchi et al., 1998). Furthermore, LOH of 11q occurred in 16% of childhood ALL (Takeuchi et al., 1999). These findings suggest that LOH is an important event related to the development of childhood ALL.

However, most of these studies have focused on the initiation of childhood ALL. Little is known about the genetic features of relapsed ALL, because of the scarcity of paired specimens. A systematic genomewide screen for the possible location of tumor-suppressor genes in childhood ALL using LOH analysis has not been reported. We designed a comprehensive set of microsatellite markers for allelotype analysis, in order to evaluate the genomic changes in relapsed childhood ALL. Our objective is to identify the genes which might play an important role in the progression of childhood ALL.

Results

Analysis of LOH in 38 relapse childhood ALL samples using microsatellite markers

We screened 38 relapse childhood ALL samples for LOH, with a panel of 71 highly informative microsatellite markers representing every autosomal chromosome. Table 1 shows the markers used, the number of informative ALL samples, and the frequency of LOH at each chromosomal arm. Figure 1 displays representative autoradiographs showing LOH. The allelotype results are presented diagrammatically in Figure 2. Of the 38 patients analysed, 26 of them (68%) showed allelic loss on at least one chromosomal arm, indicating that LOH is a frequent event at relapse of childhood ALL (Table 2).

Figure 1.

Representative autoradiographs showing LOH. Child no. 15 showed LOH at D9S1749 and D12S89, with apparent retention of heterozygosity at D6S268 and D11S976 at the time of initial diagnosis. This child showed LOH of D12S89 at relapse. Arrowheads indicate the presence of LOH; T, tumor DNA; and N, normal DNA

Full figure and legend (113K)

Figure 2.

Frequency of allelic loss on individual chromosomal arms in relapsed childhood ALL. Allelotyping was accomplished using polymorphic microsatellite analysis. The probes that were used are listed in Table 1

Full figure and legend (17K)

Table 1 - LOH in relapse ALL.

Full table

Table 2 - Clinical characteristics and loss of alleles in ALL.

Full table

The site of the most frequent LOH was at the p16/INK4a, and/or CDKN2B/INK4B/p15, and/or p14(ARF) locus on chromosomal arm 9p, where 15 of 38 (39%) informative samples showed LOH (Table 1, Figure 2). LOH at the TEL gene locus on chromosomal arm 12p was also frequent (nine of 36; 25%). Frequent allelic loss was seen on chromosome arms 4q (four of 20; 20%), 6q (eight of 38; 21%), and 17q (four of 20; 20%). LOH of chromosomal arm 11q was observed in five of 38 (13%) patients. In total, 34 of 39 (87%) chromosomal arms showed LOH for at least one patient (Table 1). Each of the five chromosomal arms (1q, 7q, 14q, 16q, 20q) did not show any LOH.

Two patients (Nos. 5 and 5590) showed LOH on almost all chromosomal arms (Table 2, Figure 3). This is consistent with a near-haploid chromosomal number. Except for these two individuals, apparent monosomy was identified only on chromosome 9.

Figure 3.

FAL values for 38 relapsed childhood ALL samples. FAL is calculated as the number of chromosome arms with loss/total number of informative chromosome arms. Median FAL value was 0.04

Full figure and legend (14K)

Fractional allelic loss (FAL), calculated for each tumor as the total number of chromosome arms lost/total number of arms with information, showed a median value of 0.04 and a mean of 0.11 (range 0–0.92), indicating that alleles were lost from 11% of evaluable chromosome arms (Table 2). The distribution of FAL is shown in Figure 3.

Sequential LOH analysis from diagnosis to relapse in 38 childhood ALL

Previously, we performed an extensive LOH analysis in childhood ALL, using DNA samples obtained from the same patients as used in this study (Takeuchi et al., 1995b, 1996, 1997a, 1998, 1999), and found that LOH at 6q, 9p, 11q, and 12p is a frequent event at the time of initial diagnosis in childhood ALL We compared those LOH data obtained at initial diagnosis with the data obtained at the time of relapse in 38 sequential samples (Table 3). Of the 38 patients, 24 (63%) showed the identical LOH status of 6q, 9p, 11q, and 12p at initial diagnosis and at relapse, suggesting the presence of the same clone during both phases. However, 14 patients (37%) showed a different LOH pattern at relapse: five patients acquired LOH at relapse, and nine of them lost LOH at relapse, suggesting a clonal diversity in the progression of childhood ALL. By chromosomal arms, three patients (Nos. 6, 18, 5158) acquired LOH of 9p at relapse, four patients (Nos. 23, 117, 177, 5514) lost LOH of 6q at relapse, and three patients (Nos. 2, 15, 5142) lost LOH of 12p at relapse. We compared the region of 9p LOH between initial diagnosis and at the time of relapse. As shown in Figure 4, nine patients had the same regions of LOH at the time of relapse as the initial diagnosis, two patients (Nos. 5346, 5438) showed larger regions of LOH, and no individual displayed a smaller region of LOH. Therefore, two individuals showed clear clonal evolution of LOH at 9p at relapse.

Figure 4.

Comparison of the region of 9p LOH between initial diagnosis and relapse in 12 childhood ALL. In all, 12 samples which showed LOH both at initial diagnosis and at relapse are presented. The status of each chromosomal locus is indicated by shading as LOH (black), retention of heterozygosity (white), and not informative (gray). The upper columns show the LOH results obtained at initial diagnosis (I), and lower columns show the LOH results obtained at relapse (R). Patient nos. are listed at the top of each column

Full figure and legend (77K)

Table 3 - Sequential analysis of LOH at 6q, 9p, 11q, and 12p, comparing samples at diagnosis and relapse in 38 childhood ALL.

Full table

Complete allelotype analysis was performed in seven of these samples at the time of initial diagnosis (Takeuchi et al., 1995b). We compared these allelotype results obtained at initial diagnosis with the data obtained at the time of relapse (Table 4). Of the seven patients, one individual (No. 3) had an identical LOH pattern at initial diagnosis and at relapse, indicating the presence of the identical clone at both phase. Two patients (Nos. 5, 15) showed both identical and unique sites of LOH, indicating the presence of clonally related subclones at relapse. Two children (Nos. 2, 18) showed distinct LOH results at relapse, indicating that a clonal change occurred at relapse. Patient No. 6 acquired 9p LOH at relapse, consistent with clonal evolution. Patient No. 23 lost 6q LOH at relapse, indicating genetic convergence.

Table 4 - Sequential allelotype analysis comparing samples at diagnosis and relapse in seven childhood ALL.

Full table

Discussion

The most frequent LOH (39%) was found at the p16/INK4a locus on chromosomal arm 9p. Sequential analysis revealed that three of 38 patients acquired LOH at 9p at relapse, and two children showed larger regions of LOH at 9p at relapse than at the initial diagnosis. Moreover, no individual had a smaller LOH region at relapse. This is consistent with the previous finding that p16/INK4a deletions can be acquired at relapse (Maloney et al., 1999; Carter et al., 2001). Cytogenetic data support our findings that deletions of 9p can occur between diagnosis and relapse (Shikano et al., 1990; Heerema et al., 1992). Two different explanations are possible for the appearance of 9p LOH at relapse. First, 9p LOH was not present in any of the leukemic cells in the patient at diagnosis, but it developed during progression of the disease. Second, a small subclone that contained 9p LOH was present at diagnosis, and this clone become dominant because of its proliferative advantage. Indeed, patients with 9p abnormalities have an increased incidence of relapse, suggesting a growth advantage of these abnormalities (Heerema et al., 1999). Homozygous deletion of p16/INK4a is identified by an apparent retention of heterozygosity on one or more closely located markers within the region of LOH (Cairns et al., 1995). If a homozygous deletion is present at the locus to be amplified, the leukemic cells do not provide a template for amplification. Therefore, amplification of a small amount of contaminated normal cells results in apparent retention of heterozygosity. Homozygous deletion of the p16/INK4a gene occurred in three patients (Nos. 15, 66, 5590) at the time of initial diagnosis, and was acquired in two patients (Nos. 5346, 5438) at relapse.

LOH at the TEL gene locus on chromosomal arm 12p was also found frequently (25%). This is much higher than the cytogenetically reported frequency of 12p deletions (Shikano et al., 1990; Heerema et al., 1992). Thus, cytogenetic studies have probably missed most cases of small interstitial deletions on 12p, showing the extreme power of LOH analysis using microsatellite markers. All the patients with LOH of 12p, for whom clinical information was available, were diagnosed as B-cell lineage ALL, consistent with the finding that LOH and concomitant TEL-AML1 fusion were found exclusively in childhood B-lineage ALL (Shurtleff et al., 1995; Stegmaier et al., 1995; Mclean et al., 1996; Takeuchi et al., 1996, 1997b). Sequential analysis revealed that three of 11 patients lost their 12p LOH between diagnosis and relapse. The loss of 12p LOH at relapse could have two explanations. First, all the leukemic cells with 12p LOH that were present at diagnosis were eradicated by intensive chemotherapy, but a new clone without 12p LOH developed during disease progression. Second, a subclone without 12p LOH was present at diagnosis, and this clone became dominant because of its proliferative advantage, compared to those with 12p LOH. Indeed, children with ALL, whose cells have a loss of 12p, have a favorable prognosis, suggesting that this clone of leukemic cells could have a proliferative disadvantage compared to another clone of leukemic cells (Heerema et al., 2000).

Frequent allelic loss was observed on chromosome arms 4q (four of 20; 20%) and 17q (four of 20; 20%). Cytogenetic studies have not revealed frequent loss at these sites, again demonstrating the robustness of the LOH analysis using microsatellite markers (Shikano et al., 1990; Heerema et al., 1992). Of interest, our allelotype analysis identified frequent LOH of 4q (17%) and 17q (23%) in the ALL samples obtained at initial diagnosis (Takeuchi et al., 1995b). Frequent LOH of 4q and 17q both at initial diagnosis and relapse implies the existence of a tumor-suppressor gene at both of these sites, which is probably often mutated during the development of childhood ALL. However, no disease-related genes have yet been associated with these LOH sites. Further fine mapping of these chromosomal regions should eventually help clone these candidate tumor-suppressor genes.

FAL, calculated for each tumor as the total number of chromosome arms lost/total number of arms with information, showed a median value of 0.04 and a mean of 0.11 (Table 2). This FAL at relapse is similar to FAL at diagnosis (mean, 0.04; median, 0.12), suggesting that the total numbers of involved sites are similar (Takeuchi et al., 1995b). Interestingly, only chromosome 14q was retained both at diagnosis and relapse in all informative cases, suggesting that this chromosomal arm was not likely to harbor important tumor-suppressor genes for the development of childhood ALL. These FAL values are lower than those reported in solid tumors, suggesting that deletions are less common in ALL and probably fewer putative tumor-suppressor genes are altered in ALL as compared to the majority of solid tumors. We may have underestimated the true frequency of LOH, since not all of the chromosomal sites are informative for all of the samples.

We compared this allelotype data at relapse with our allelotype data at initial diagnosis (Takeuchi et al., 1995b, 2003). The most frequent LOH was 9p at both time points (40% at diagnosis and 39% at relapse). The second most frequent LOH was 12p, which had the same frequency at both phases of the disease (26% at diagnosis and 25% at relapse). Also, chromosomal arms 4q, 6q, and 17q were frequently deleted at both times. Moreover, 14q LOH was not detected at either diagnosis or relapse. Despite the heterogeneous and complex changes, some shared alterations occurred in these matched samples, suggesting that many of the same tumor-suppressor genes are aberrant at both initial diagnosis and relapse in childhood ALL.

The LOH results from paired specimens of the same patient allowed us to define a hypothetical model of clonal progression. Sequential LOH analysis at 6q, 9p, 11q, and 12p at diagnosis, remission, and relapse shows that 63% of relapsed patients have the identical LOH pattern as their initial diagnosis. In about 70% of patients, Ig and T-cell receptor gene rearrangements are largely identical at relapse (Szczepanski et al., 2002). Taken together, about 60–70% of relapse childhood ALL arise from the identical clone as their first presentation. However, five patients acquired an additional LOH at relapse, and three of these five children developed 9p LOH, suggesting that a clonal evolution occurred. In contrast, nine patients lost their LOH at relapse, four children lost 6q LOH, and three of them lost 12p LOH. A possible explanation is that the primary leukemia was heterogeneous and only a small subpopulation of the primary leukemic cells relapsed. If the selective pressure confronting a leukemic cell population exceeds the cells' genetic instability, net reduction of clonal complexity will follow, and the population may become more homogeneous during progression (Heim et al., 1988). Furthermore, sequential allelotype analysis from diagnosis to relapse shows the presence of either identical or clonally related subclones in some cases. However, other children progress with clonal changes, clonal evolutions, or genetic convergence, suggesting that clonal diversity is common during progression of childhood ALL.

Materials and methods

Samples

A total of 28 relapsed DNA samples of childhood ALL were obtained from individuals who participated in the ongoing Multicenter Trial (ALL-BFM 90) of the German Berlin–Frankfurt–Münster (BFM) group (Schrappe et al., 2000). The other 10 relapsed samples were obtained from the Dutch Childhood Leukemia Study Group (Kamps et al., 2002). Informed consent was obtained from the patients, their parents, or both, as appropriate. These patients were selected by the availability of cryopreserved cell samples. The percentage of blast cells in the samples was at least 80%, and usually more than 90%; DNA was extracted from them. The corresponding normal DNAs from the same individuals were obtained from the bone marrows after complete remission (CR) was achieved (at least 12 months after initial diagnosis).

Analysis of LOH using microsatellite markers

The LOH analysis was performed by polymerase chain reaction (PCR) amplification of microsatellite sequences. Primers for microsatellite sequences were obtained from Research Genetics (Huntsville, AL, USA) (Gyapay et al., 1994), and are listed in Table 1. Previously, we performed LOH analysis in childhood ALL, using DNA samples obtained at the time of initial diagnosis from the same patients as used in this study, and identified several distinct regions of deletion in chromosome arms 6q, 9p, 11q, and 12p (Takeuchi et al., 1995b, 1996, 1997a, 1998, 1999). In this study, we used more markers in chromosome arms 6q, 9p, and 11q than other arms, in order to examine all of the previously identified regions of deletion. Each PCR reaction contained 5–25 ng of DNA, 10 pmol of each of the primers, 2 nmol of each of the four deoxyribonucleotide triphosphates (Pharmacia, Stockholm, Sweden), 0.5 U of Taq DNA polymerase (Boehringer-Mannheim, Indianapolis, IN, USA), and 2 Ci dCTP (ICN, Irvine, CA, USA) in 20 l of the specified buffer with 1.5 mM MgCl₂. Samples were amplified using 30–35 cycles of denaturing for 40 s at 94°C, annealing for 30 s at 55°C, and extending for 1 min at 72°C in a Programmable Thermal Controller (MJ Research Inc, Watertown, MA, USA). After amplification, PCR samples were diluted fivefold in loading buffer containing 20 mM EDTA, 96% formamide, and 0.05% of both bromophenol blue and xylene cyanol. The products were heated to 95°C for 5 min, and 1.5 l of each sample was electrophoresed through a 6% polyacrylamide gel containing 8.3 M urea for 3–4 h at 85 W. Subsequently, the gels were dried and subjected to autoradiography using Kodak XAR-5 film (Eastman Kodak, Rochester, NY, USA) at -80°C. LOH was determined by densitometry. LOH was inferred only when substantial reduction (>50%) was measured in the ratio of radiographic signal intensities of an allele in the tumor sample, relative to that in the corresponding normal sample.

Sequential LOH analysis from diagnosis to relapse

Previously, we performed LOH analysis in childhood ALL using DNA samples obtained from the same patients as used in this study (Takeuchi et al., 1995b, 1996, 1997a, 1998, 1999), and found that LOH at 6q, 9p, 11q, and 12p is a frequent event at the time of initial diagnosis. Therefore, chromosome arms 6q, 9p, 11q, and 12p were selected for comparison between diagnosis and relapse in all 38 samples. Complete allelotype analysis was performed in seven of 38 relapsed samples at the time of initial diagnosis (Takeuchi et al., 1995b). We compared these allelotype results obtained at initial diagnosis with the data obtained at the time of relapse.

References

1. Cairns P, Polascik J, Eby Y, Tokino K, Califano J, Merlo A, Mao L, Herath J, Jenkins R, Westra W, Rutter JL, Buckler A, Gabrielson E, Tockman M, Cho KR, Hedrick L, Bova GS, Isaacs W, Koch W, Schwab D and Sidransky D. (1995). Nat. Genet., 11, 210–212. | Article | PubMed | ISI | ChemPort |
2. Carter TL, Reaman GH and Kees UR. (2001). Br. J. Haematol., 113, 323–328. | Article | PubMed | ChemPort |
3. Gaynon PS, Qu RP, Chappell RJ, Willoughby ML, Tubergen DG, Steinherz PG and Trigg ME. (1998). Cancer, 82, 1387–1395. | Article | PubMed | ISI | ChemPort |
4. Gérard B, Cavé H, Guidal C, Dastugue N, Vilmer E and Grandchamp B. (1997). Leukemia, 11, 228–232. | Article | PubMed | ISI | ChemPort |
5. Gyapay G, Morissette J, Vignal A, Dib C, Fizames C, Millasseau P, Marc S, Bernardi G, Lathrop M and Weissenbach J. (1994). Nat. Genet., 7, 246–339. | Article | PubMed | ISI | ChemPort |
6. Harrison CJ. (2001). Blood Rev., 15, 49–59. | PubMed |
7. Heerema NA, Palmer CG, Weetman R and Bertolone S. (1992). Leukemia, 6, 185–192. | PubMed | ISI | ChemPort |
8. Heerema NA, Sather HN, Sensel MG, Lee MK, Hutchinson R, Nachman JB, Lange BJ, Steinhertz PG, Bostrom B, Gaynon PG and Uckun FM. (2000). Cancer, 88, 1945–1954. | Article | PubMed | ISI | ChemPort |
9. Heerema NA, Sather HN, Sensel MG, Liu-Mares W, Lange BL, Bostrom BC, Nachman JB, Steinherz PG, Hutchinson R, Gaynon PS, Arthur DC and Uckun FM. (1999). Blood, 94, 1537–1544. | PubMed |
10. Heim S, Mandahl N and Mitelman F. (1988). Cancer Res., 48, 5911–5916. | PubMed | ISI | ChemPort |
11. Kamel-Reid S, Letarte M, Doedens M, Greaves A, Murdoch B, Grunberger T, Lapidot T, Thorner P, Freedman MH and Phillips RA. (1991). Blood, 78, 2973–2981. | PubMed | ChemPort |
12. Kamps WA, Bokkerink JP, Hakvoort-Cammel FG, Veerman AJ, Weening RS, van Wering ER, van Weerden JF, Hermans J, Slater R, van den Berg E, Kroes WG and van der Does-van den Berg A. (2002). Leukemia, 16, 1099–1111. | Article | PubMed | ISI | ChemPort |
13. Maloney KW, McGavran L, Odom LF and Hunger SP. (1999). Blood, 93, 2380–2385. | PubMed | ISI | ChemPort |
14. Mclean TW, Ringold S, Neuberg D, Stegmaier K, Tantravahi R, Ritz J, Koeffler HP, Takeuchi S, Janssen JWG, Seriu T, Bartram CR, Sallan SE, Gilliland DG and Golub TR. (1996). Blood, 88, 4252–4258. | PubMed | ISI | ChemPort |
15. Okuda T, Shurtleff SA, Valentine MB, Raimondi SC, Head DR, Behm F, Curcio-Brint AM, Liu Q, Pui CH, Sherr CJ, Beach D, Look AT and Downing JR. (1995). Blood, 85, 2321–2330. | PubMed | ISI | ChemPort |
16. Raimondi SC. (1993). Blood, 81, 2237–2251. | PubMed | ISI | ChemPort |
17. Schrappe M, Reiter A, Ludwig W-D, Harbott J, Zimmermann M, Hiddemann W, Niemeyer C, Henze G, Feldges A, Zintl F, Kornhuber B, Ritter J, Welte K, Gadner H and Riehm H. (2000). Blood, 95, 3310–3322. | PubMed | ISI | ChemPort |
18. Shikano T, Ishikawa Y, Ohkawa M, Hatayama Y, Nakadate H, Hatae Y and Takeda T. (1990). Leukemia, 4, 419–422. | PubMed |
19. Shurtleff SA, Buijs A, Behm FG, Rubnitz JE, Raimondi SC, Hancock ML, Chan GC, Pui CH, Grosveld G and Downing JR. (1995). Leukemia, 9, 1985–1989. | PubMed | ISI | ChemPort |
20. Stegmaier K, Pendse S, Barker GF, Bray-Ward P, Ward DC, Montgomery KT, Krauter KS, Reynolds C, Sklar J, Donnelly M, Bohlander SK, Rowley JD, Sallan SE, Gilliland DG and Golub TD. (1995). Blood, 86, 38–44. | PubMed | ISI | ChemPort |
21. Szczepañski T, Willemse MJ, Brinkhof B, van Wering ER, van der Burg M and van Dongen JJM. (2002). Blood, 99, 2315–2323. | Article | PubMed | ISI | ChemPort |
22. Takeuchi S, Bartram CR, Miller CW, Reiter A, Seriu T, Zimmermann M, Schrappe M, Mori N, Slater J, Miyoshi I and Koeffler HP. (1996). Blood, 87, 3368–3374. | PubMed | ISI | ChemPort |
23. Takeuchi S, Bartram CR, Seriu T, Miller CR, Tobler A, Janssen JWG, Reiter A, Ludwig W-D, Zimmermann M, Schwaller J, Lee E, Miyoshi I and Koeffler HP. (1995a). Blood, 86, 755–760. | PubMed | ISI | ChemPort |
24. Takeuchi S, Bartram CR, Wada M, Reiter A, Hatta Y, Seriu T, Lee E, Miller CW, Miyoshi I and Koeffler HP. (1995b). Cancer Res., 55, 5377–5382. | PubMed | ISI | ChemPort |
25. Takeuchi S, Cho SK, Seriu T, Koike M, Bartram CR, Reiter A, Schrappe M, Takeuchi C, Taguchi H and Koeffler HP. (1999). Oncogene, 18, 7387–7388. | Article | PubMed | ChemPort |
26. Takeuchi S, Koike M, Seriu T, Bartram CR, Schrappe M, Reiter A, Park S, Taub HE, Kubonishi I, Miyoshi I and Koeffler HP. (1998). Cancer Res., 58, 2618–2623. | PubMed | ISI | ChemPort |
27. Takeuchi S, Koike M, Seriu T, Bartram CR, Slater J, Park S, Miyoshi I and Koeffler HP. (1997a). Leukemia, 11, 1636–1640. | Article | PubMed |
28. Takeuchi S, Seriu T, Bartram CR, Golub TR, Miyoshi I, Gilliland DG and Koeffler HP. (1997b). Leukemia, 11, 1220–1223. | Article | PubMed | ISI | ChemPort |
29. Takeuchi S, Tsukasaki K, Bartram CR, Seriu T, Zimmermann M, Schrappe M, Takeuchi N, Park S, Taguchi H and Koeffler HP. (2003). Leukemia, 17, 149–154. | Article | PubMed | ISI | ChemPort |

Acknowledgements

We are grateful to Wolf Dieter Ludwig for providing the immunophenotypic data. We thank Ian K Williamson, Jeffrey Grewal, and Harry E Taub for technical help. We also thank the Board and the Clinicians of the Dutch Childhood Oncology Group for kindly providing ALL cell samples. This work was supported in part by grants from National Institutes of Health, C and H Koeffler Fund, Parker Hughes Trust, Deutsche Forschungsgemeinschaft, Deutsche Krebshilfe, Uehara Memorial Foundation, and Grant-in-Aid from the Ministry of Education, Culture Sports, Science, and Technology of Japan. HPK is a member of the Jonsson Comprehensive Cancer Center and the Molecular Biology Institute, and holds the endowed Mark Goodson Chair of Oncology Research at Cedars-Sinai Medical Center/UCLA School of Medicine.