—  SOCIETY FOR HEMATOPATHOLOGY   —

The Molecular Pathology of Pediatric Acute Leukemia


James Downing
St. Jude Children's Hospital
Memphis, TN


A major goal of leukemia research has been to gain a better understanding of the genetic changes responsible for the establishment of the leukemic clone. This pursuit is fueled by the hope that the information obtained will not only help us to understand the differences in clinical response that are observed among leukemic patients, but will also lead to the identification of rational molecular targets against which novel therapeutic agents can be developed. One of the major insights into the underlying pathogenesis of acute leukemia has come from the careful analysis of chromosomal changes within the leukemic cells. These studies have revealed the presence of clonal chromosomal translocations and inversions in greater than 50% of cases. Efforts to determine the targets of these abnormalities have identified a series of genes that play critical roles in the normal development of the hematopoietic system [1, 2, 3] .

The genes encoding the AML1/CBFb transcription factor complex are among the most frequent targets of genetic alterations in pediatric and adult acute leukemia [4, 5] . AML1 is targeted by the t(8;21) and t(16;21) in acute myeloid leukemia (AML), the t(3;21) in myelodysplatic syndrome, and the t(12;21) in pediatric acute lymphoblastic leukemia (ALL), where as CBFb is the target of the AML-associated chromosomal rearrangements, inv(16) and t(16;16). In addition, germ-line mutations of AML1 are the underlying abnormality in a familial platelet disorder in which the patients have a high predisposition to develop AML, and somatic mutations in AML1 occur in rare cases of sporadic AML. Collectively these leukemias are referred to as core binding factor (CBF) leukemias.

Work performed in my laboratory, as well as others, demonstrated that AML1/CBFb functions as a master regulatory switch that establishes a transcriptional cascade necessary for the development of the definitive hematopoietic stem cell (HSC) [6]. To explore the leukemia potential of translocation-encoded AML1 chimeric genes, we generated mice with an AML1-ETO fusion gene by knocking ETO into the AML1 genomic locus [7]. Expression of AML1-ETO induced an embryonic lethal phenotype similar to that resulting from the loss of AML1/CBFb. However, although AML1/CBFb-deficient embryos lacked detectable hematopoietic progenitors, fetal livers from AML1-ETO expressing embryos contained dysplastic hematopoietic progenitors that had a high self-renewal capacity. These cells, however, failed to result in overt leukemias either within the developing embryos, or when transplanted into recipient mice. Thus, expression of AML1-ETO was insufficient to induce leukemia, but instead altered the self-renewal capacity of HSC resulting in a preleukemic population.

To further define the leukemic potential of AML1-ETO, we generated mice that contain a conditional AML1-ETO knock-in allele, in which a strong transcriptional stop cassette bracketed by loxP sites was placed 5' to the AML1-ETO fusion [8]. To induce the activation of the allele in vivo, the mice were crossed with a murine line that contains an interferon-inducible Mx1-Cre transgene. Induction of Cre expression in double transgenic mice resulted in the efficient expression of AML1-ETO within HSC and their mature progeny. Interestingly, expression of AML1-ETO in these adult mice resulted in only minor perturbation within the hematopoietic system. Moreover, no overt leukemia developed during the first year. However, when mice were mutagenized with ENU at a dose that induced only rare T-cell leukemias in control animals, 35% of the AML1-ETO expressing mice developed an AML that closely resembled human t(8;21)-containing AML.

An important insight into the nature of the secondary mutation that cooperated with AML1-ETO in this experimental system arose from the observation that the non-leukemic AML1-ETO expressing cell lines were cytokine-dependent, where as cell lines derived from the murine AML1-ETO expressing leukemias were factor-independent. These data strongly suggest that one signaling pathway that collaborates with AML1-ETO is cytokine- or growth factor-mediated proliferation or survival. To further assess this possibility, we analyzed 31 human CBF leukemias and a similar number of control non-CBF leukemias for evidence of activating mutations in N-RAS, K-RAS, and c-KIT. Remarkably, 48% of the CBF leukemias contained activating mutations in N-RAS or K-RAS, and an additional 13% had c-KIT activating mutations, where as the frequency of these mutations in the non-CBF leukemias were 13% and 0%, respectively (unpublished observation, Sun and Downing). Thus, activating mutations in these growth factor signaling molecules appears to be a frequent even in CBF leukemias. These data suggest that inhibitors of either the AML1/CBFb transcriptional cascade or RAS-mediated signaling pathways may be an effective means to inhibit the growth of leukemic cells containing these genetic lesions. These hypotheses are being tested in murine models engineered to contain these specific genetic lesions. In addition, the conditional AML1-ETO mouse is being used in genetic screens to determine the range of mutations that can cooperate with the expression of AML1-ETO to induce leukemia. Any lesions identified through these murine screens will be assessed in human AML1-ETO leukemias to see if they contribute to the molecular pathology of this disease.

A second area of investigation that will provide important insights into the molecular mechanisms of leukemogenesis is the analysis of the expression profiles of primary human leukemia samples [8, 9, 10, 11, 12, 13, 14, 15, 16, 17] . In studies performed in my laboratory, expression profiles have been obtained from a large number of pediatric ALLs [12, 13] and AMLs. Specifically, three human leukemia gene expression datasets have been generated: (i) An ALL expression database containing the expression profiles of over 327 diagnostic bone marrow samples using the Affymetrix U95Av2 microarray (http://www.stjuderesearch.org/data/ALL1), (ii) An ALL expression database containing the expression profiles of 132 diagnostic bone marrow samples obtained using the Affymetrix U133A and B microarrays (http://www.stjuderesearch.org/data/ALL3), and (iii) an AML expression database containing the expression profiles from 130 pediatric and 20 adult diagnostic bone marrow samples obtained using the Affymetrix U133A microarray (in press). The analysis of these databases have demonstrate specific gene expression profiles for each of the known prognostically and therapeutically relevant subgroups of childhood ALL and AML. Distinct gene expression profiles were identified for ALL blasts with T-lineage, hyperdiploid >50 chromosomes, BCR-ABL, E2A-PBX1, TEL-AML1, and MLL gene rearrangement, where as in AML, distinct expression profiles were identified for AML blasts with t(15;17)[PML-RARa], t(8;21)[AML1-ETO], inv(16)[CBFb-MYH11], and MLL gene rearrangement. Moreover, using a variety of computer-assisted supervised learning algorithms, the single platform of expression profiling was shown to accurately diagnosis the various leukemia subtypes with an overall accuracy of 96%. Remarkably, this level of accuracy exceeds that typically achieved using a combination of contemporary diagnostic approaches, suggesting that microarray–based gene expression profiling may provide a viable approach for the front-line diagnostic work up of patients with acute leukemia.

The identified expression profiles also suggested new insights into the pathogenesis of these diseases. However, sorting through this data to identify those genes whose altered expression is mechanistically involved in disease pathogenesis will be a significant challenge. This analysis is likely to benefit from cross-species comparisons between the human leukemias and murine models of the same disease. The ability to specifically program into mice the identical genetic lesions found in the human disease, coupled with our ability to follow the murine disease from a preleukemic phase to overt leukemia, should help to identify the alterations in gene expression that are involved in disease pathogenesis.

References

  1. Downing JR, Shannon KM. Acute leukemia: a pediatric perspective. Cancer Cell. 2002;2:437-445.
  2. Pui CH, Campana D, Evans WE. Childhood acute lymphoblastic leukemia—current status and future perspectives. Lancet. 2001;2:597-607.
  3. Pui CH, Relling MV, Downing JR. Acute Lymphoblastic Leukemia. (NEJM, In Press).
  4. Lorsbach R, Downing JR. The Role of the AML1 Transcription Factor in Leukemogenesis. Int'l Journal of Hematology: 258-265, 2001.
  5. Downing JR. The core-binding factor leukemias: lessons learned from murine models. Current Opinions in Genetics Dev. Current Opinion in Genetics & Development 13: 48-54, 2003.
  6. Okuda T, Deursen JV, Hiebert SW, Grosveld G, Downing JR. AML-1, the target of multiple chromosomal translocations in human leukemia, is essential for normal fetal liver hematopoiesis. Cell 84: 321-330, 1996.
  7. Okuda T, Cai Z, Yang S, Lenny N, Lyu C-j, van Deursen J, Harada H, Downing JR. Expression of a knocked-in AML1-ETO leukemia gene inhibits the establishment of normal definitive hematopoiesis and directly generates dysplastic hematopoietic progenitors. Blood 91: 3134-3143, 1998.
  8. Higuchi H, O'Brien D, Kumaravelu P, Lenny N, Yeoh E-J, Downing JR. Expression of a conditional AML1-ETO oncogene bypasses embryonic lethality and establishes a murine model of human t(8;21) acute myeloid leukemia. Cancer Cell 1: 63-74, 2002.
  9. Golub TR, Slonim DK, Tamayo P, et al. Molecular classification of cancer: class discovery and class prediction by gene expression monitoring. Science. 1999;286:531-537.
  10. Armstrong SA, Staunton JE, Silverman LB, et al. MLL translocations specify a distinct gene expression profile that distinguishes a unique leukemia. Nat Genet. 2002;30:41-47.
  11. Ferrando AA, Neuberg DS, Staunton J, et al. Gene expression signatures define novel oncogenic pathways in T cell acute lymphoblastic leukemia. Cancer Cell. 2002;1:75-87.
  12. Yeoh EJ, Ross ME, Shurtleff SA, et al. Classification, subtype discovery, and prediction of outcome in pediatric acute lymphoblastic leukemia by gene expression profiling. Cancer Cell. 2002;1:133-143.
  13. Ross ME, Xiaodong Z, Song G, et al. Classification of pediatric acute lymphoblastic leukemia by gene expression profiling. Blood. 2003;102:2951-2959.
  14. Virtaneva K, Wright FA, Tanner SM, et al. Expression profiling reveals fundamental biological differences in acute myeloid leukemia with isolated trisomy 8 and normal cytogenetics. Proc Natl Acad Sci U S A. 2001;98:1124-1129.
  15. Schoch C, Kohlmann A, Schnittger S, et al. Acute myeloid leukemias with reciprocal rearrangements can be distinguished by specific gene expression profiles. Proc Natl Acad Sci U S A. 2002;99:10008-10013.
  16. Kohlmann A, Schoch C, Schnittgere S, et al. Molecular characterization of acute leukemias by use of microarray technology. Gene Chrom & Cancer. 2003;37:396-405.
  17. Armstrong SA, Kung AL, Mabon ME, et al. Inhibition of FLT3 in MLL: validation of a therapeutic target identified by gene expression based classification. Cancer Cell. 2003;3:173-183.