Molecular Targets for Therapy (MTT)

Hox expression in AML identifies a distinct subset of patients with intermediate cytogenetics


We previously reported that favorable and poor prognostic chromosomal rearrangements in acute myeloid leukemia (AML) were associated with distinct levels of HOX expression. We have now analyzed HOX expression in 50 independent adult AML patients (median age=62 years), together with FLT3 and FLT3-ligand mRNA levels, and FLT3 mutation determination. By cluster analysis, we could divide AMLs into cases with low, intermediate and high HOX expression. Cases with high expression were uniquely restricted to a subset of AMLs with intermediate cytogenetics (P=0.0174). This subset has significantly higher levels of FLT3 expression and appears to have an increase of FLT3 mutations (44%), while CEBPα mutations were infrequent (6%). FLT3 mRNA levels were correlated with the expression of multiple HOX genes, whereas FLT3 mutations were correlated with HOXB3. In some cases, FLT3 was expressed at levels equivalent to GAPDH in the absence of genomic amplification. We propose that high HOX expression may be characteristically associated with a distinct biologic subset of AML. The apparent global upregulation of HOX expression could be due to growth-factor signaling or, alternatively, these patterns may reflect a particular stage of differentiation of the leukemic cells.


In adult acute myeloid leukemia (AML), three common favorable cytogenetic abnormalities, that is, t(8;21), inv(16) and t(15;17), and five adverse rearrangements, that is, −5, −7, del(5q), 3q alterations and a complex karyotype, define favorable and poor prognostic variants, respectively. Additional poor prognostic factors include t(11;19)(q23;p13.1), trisomy 8 together with t(9;11) and CML-blast crisis. Most other abnormalities, as well as a normal karyotype, carry an intermediate prognosis.1,2 Other than chromosomal rearrangements, relatively little is known of the biologic differences that give rise to these diverse prognostic groups.

Homeodomain genes encode transcription factors that act during development on pattern formation, differentiation and proliferation. Humans possess 39 clustered HOX genes located on 7p15 (HOXA), 17q21 (HOXB), 12q13 (HOXC) and 2q31 (HOXD). Paralogs (eg, HOXA9, B9, C9, D9) are more closely related than adjacent genes within the same cluster. In certain cases, these paralogs are functionally equivalent and it is the overall expression level of the group that determines response.3 HOX deregulation has been unequivocally implicated in the pathogenesis of human leukemia, including translocations that fuse NUP98 with HOXA9, A11, A13, PMX1, C11, C13, D11 or D13.4,5,6,7,8,9,10,11 In the BXH-2 mouse model of leukemia, Hoxa7 and a9 are frequent targets of retroviral integration,12 and are necessary for leukemia development in mice carrying a MLL-ENL rearrangement.13

We previously reported that HOX expression was characteristically low in favorable AML subtypes compared to poor prognostic cases.14 This observation has been confirmed by others.15,16 We also observed substantial variations among patients with intermediate cytogenetics, suggesting that the HOX profiles might be associated, or responsible, for some characteristics of the disease. In addition to loss of differentiation by transcriptional means, cancer often involves upregulation of growth-factor signaling. Activating mutations in FLT3, a receptor tyrosine kinase, fit this paradigm. FLT3 mutations are frequent in cases with intermediate cytogenetics and in some forms of good prognostic AML, whereas they are infrequent in poor prognostic cases.17 Since FLT3 mutations occur nonrandomly, we wondered if they might be linked to HOX expression. Here, we show that the highest levels of HOX expression are confined to a subset of patients with intermediate cytogenetics, and that FLT3 levels were significantly correlated with HOX expression. Although confirmatory studies are required, the high HOX expression subset was associated with more FLT3 mutations and, conversely, fewer CEBPα mutations. Thus, high HOX expression might identify a biologic subset of AML.


Real-time quantitative RT-PCR

Preparation of RNA, cDNA and performance of the various quantitative RT-PCR assays using SYBR-green for the indicated homeodomain genes was performed as previously described. 14 Limitations imposed by the patient samples required that these analysis were performed on bulk marrow. All real-time PCR reactions were performed under identical conditions and all PCR products have been extensively verified by DNA sequencing from multiple sources including cell lines and pooled patient samples. To further minimize any nonspecific products, we eliminated any values that appeared after 34 PCR cycles (ie, undetected in Table 2). In contrast to our previous study, we utilized an ABI 7000 instead of the ABI 5700. Some products were more efficiently amplified, by 2–3 PCR cycles, with the ABI 7000. These instrument differences can be eliminated by using the 9600 emulation feature on the ABI 7000, and by increasing the annealing/extension time on the ABI 5700 from 60 to 75 s.


Adult patients with AML underwent bone marrow aspiration for morphology, cytogenetics and immunophenotyping. Samples for molecular analyses were frozen and local IRB approval was obtained. The diagnosis of AML was based on May–Grünwald–Giemsa-stained bone marrow smears and cytochemistry according to the French–American–British (FAB) Group criteria.18 All patients were treated using an anthracycline and cytarabine-based induction chemotherapy.

Amplification of FLT3 and CEBPα

Conditions and primers are provided in supplementary material.

Statistical analysis

Cases were analyzed using clustering algorithms in the SAS PROC CLUSTER procedure and GeneLinkerTM Platinum (Molecular Mining Corporation, 2002). Each produced essentially identical results. The results reported here were based on the method of Ward in SAS PROC CLUSTER. The Spearman correlation coefficients were used to look for associations between levels of HOX expression and other parameters such as WBC, percent blasts, etc. The Wilcoxon rank-sum test was used to compare HOX expression levels with the presence or absence of an activating mutation in FLT3, as well as with treatment response vs nonresponse, relapse vs no relapse, and dead vs alive at follow-up.


Highest levels of HOX expression are confined to the intermediate cytogenetic subset

From 68 recent cases, frozen samples were available from 57 consenting adults with newly diagnosed AML. The median and mean age was 62 and 56 years, respectively. Of these, seven were excluded because of insufficient RNA. The remaining 50 were analyzed for HOX and FLT3 expression using real-time RT-PCR with SYBR-green, and for FLT3 activating mutations by DNA sequencing. The complete karyotypes are shown in Table 1(supplementary material), grouped according to MRC and CALGB classifications,2,19 along with age, WBC, blasts and outcome where known. Four patients had favorable cytogenetics, 33 were intermediate and 13 had poor prognostic features.

The results are shown in Table 2 (supplementary material), grouped by HOX expression levels along with the predominant cytogenetic feature. The choice of HOX genes was based on our previous work, which used a degenerate RT-PCR screen of AML cell lines and patient samples to establish an initial set of quantitative assays.14 HOX expression (× 1000) was normalized to GAPDH as previously described14 and similar results were obtained using 18S RNA as the normalization (not shown). We applied various clustering algorithms in the SAS PROC CLUSTER procedure and in GeneLinkerTM Platinum (Molecular Mining Corporation, 2002), which indicated that the cases could be grouped into clusters marked by distinct patterns of high, intermediate and low HOX expression. Strikingly, of 18 cases in the cluster with high HOX expression, 17 (94%) belonged to the intermediate cytogenetic group, a highly nonrandom observation (χ2=11.99; P=0.0174). The exception (patient 13) was deemed to be poor prognosis group based on monosomy 7 in two of 20 metaphases, while the other 18 metaphases contained a normal karyotype. Thus the majority of cells, based on metaphase analysis, would be considered intermediate. These results indicate that the highest levels of HOX expression uniquely associate with a subset of cases having intermediate cytogenetics. Table 3 (supplementary material) shows the mean and median expression levels for the individual homeodomain genes.

Overall, the intermediate cytogenetic group was heterogeneous containing cases with low (8/35), intermediate (10/35) and high (17/35) HOX expression. Although a diversity of rearrangements are included, the majority have normal cytogenetics. Few other biologic variables have been uniquely associated with this group. Preudhomme et al. reported that C/EBPα mutations were confined to patients with intermediate cytogenetics,20 occurring in 15 of 91 cases (17%). To determine whether C/EBPα mutations and high HOX expression were related, we sequenced C/EBPα from the cDNA of 17 cases with high HOX expression. Although two silent changes were identified, that is, GCG GCA at codon 134 and ACGACT at codon 230 (based on NM_004364), only one mutation with a coding change was detected (6%). This consisted of a 5-bp insertion (GCGCG) at bp 869 of the coding region, predicted to change the reading frame. Thus, we conclude that high HOX expression is not due to C/EBPα mutations. We have not yet tested the alternative possibility that C/EBPα mutations are more frequent in cases with intermediate/low HOX expression (see Discussion).

Correlation between HOX expression and FLT3 and evidence for FLT3 overexpression

FLT3 receptor mRNA levels were positively correlated with expression of multiple homeodomain genes (supplementary material, Table 4). Among these, MEIS1 and HOXA7 showed the highest correlations (Spearman r=0.56 and 0.53, respectively, P<0.0001). FLT3 activating mutations, on the other hand, were correlated with HOXB3 (Wilcoxon, P=0.018). As shown in Table 2, approximately twice the number of FLT3 mutations were present in cases with high HOX expression (8/18, 44%) compared to those with low (3/12, 25%) or intermediate expression (4/20, 20%). This did not reach significance (P=0.2), but it is likely to be reproducible with the predominance of intermediate cytogenetics and higher frequency of FLT3 mutations in this group.17 If only cases with intermediate cytogenetics are considered, then FLT3 mutations were present in 8/17 (47%) with high HOX expression, 4/10 (40%) with intermediate expression and 2/8 (25%) with low expression. The median/average FLT3 levels, derived from Table 2, were 97/108 in the high HOX cluster vs 44/66 and 31/58 in the intermediate and low HOX clusters, respectively. These differences were significant (P=0.0059) using the nonparametric Kruskal–Wallis test. However, there was no difference in the overall FLT3 levels in cases with and without mutations (ie, median/average=49/77 vs 53/80, respectively).

Among all cases, three (Table 2, Nos. 35, 28 and 39) had levels of FLT3 mRNA that were greater than 2 standard deviations above the mean. A fourth case (not shown) with a complex karyotype had FLT3 levels nearly twice that of GAPDH, although there was insufficient material for HOX analysis. Only one (No. 35) had an activating mutation of FLT3. Each of the first three cases was examined for genomic amplification of FLT3 using real-time PCR along with several loci on different chromosomes for normalization. No amplification was detected (not shown). Thus, very high relative levels of FLT3 can be observed in the absence of genomic amplification.

Correlations with clinical parameters

Although our study was not designed to address such questions, we looked for correlations between clinical parameters, HOX expression and aspects of FLT3. No significant associations were identified between HOX expression and age, sex, response (CR vs. NR), relapse and outcome (dead vs alive at last follow-up). In the overall group, HOXB9 was the most highly overexpressed gene among treatment responders vs nonresponders (ie, median 1.0 (vs GAPDH × 1000) in responders vs 0.1 in nonresponders; Wilcoxon test P-value=0.1132). In those with intermediate cytogenetics, MEIS2 was most highly overexpressed among treatment responders vs nonresponders (ie, median 1.0 (vs GAPDH × 1000) in responders vs 0.3 in nonresponders; Wilcoxon test P-value=0.017). We interpret these results cautiously since neither meets the significance bound after adjustment for multiple testing (ie the Bonferroni bound=0.05/17 (genes)=0.0029). Each individual homeodomain gene, except MEIS2, showed a positive correlation between expression and the percent blasts, although none reached statistical significance (not shown). Results with wild-type FLT3 levels were similar. Only mutated FLT3 was significantly correlated with higher blast percentages (68 vs. 51%, P=0.02) using the Wilcoxon rank-sum test. FLT3 ligand expression was detected in all samples (Table 2). Here, the Spearman correlation between percent blasts and expression levels was negative (r=−0.37, P=0.0086), meaning that lower levels of FLT3 ligand were associated with more blasts. Serum FLT3 ligand has been inversely correlated with BM cellularity in patients with aplastic anemia and chemotherapy-induced aplasia21 and in patients following transplant.22


Chromosome rearrangements and viral integrations affecting HOX genes result in AML in humans and mice. In humans, the majority of rearrangements involve 5′-HOX genes.4,5,6,7,8,9,10,11 However, other homeodomain genes, that is, HOXA5,23 HoxB324,25 and HoxB426,27 affect hematopoietic proliferation and differentiation. We previously reported that multiple HOX genes were overexpressed in poor prognostic AML compared to cases with favorable cytogenetics.14 Thompson et al15 also observed downregulation of multiple HOX genes in APL. This dichotomy of HOX expression may provide a common distinguishing feature between favorable and poor prognostic groups. Experimentally, overexpression of various HOX genes has profound effects on myeloid development ranging from increased self-renewal to overt leukemia. Thus, it seems reasonable to suspect that multiple HOX genes may be contributing to the pathogenesis of AML when jointly overexpressed.

The present study demonstrates that highest levels of HOX expression are confined to a subset of AML with intermediate cytogenetics. This was highly significant (P=0.0174), but should be independently confirmed. In agreement with these results, we previously noted that several normal cytogenetic samples contained the highest levels of HOX expression, although the overall number of cases was small and a cluster analysis had not been applied. Debernardi et al,16 using microarrays (see Figure 4, p. 156), noted that five HOX genes (HOXB2, B3, B5, B6 and B7) were highly expressed in AMLs with normal karyotypes compared to 14 patients with favorable cytogenetics, and three of four patients with 11q23 alterations. Thus, these results and our previous and current findings are in agreement.

Our studies and those of Thompson et al15 and Debernardi et al16 reach the common conclusion that low HOX expression is associated with favorable cytogenetics. However, we have now identified two cases (Nos. 16 and 32) where HOX levels reached intermediate values. Patient (Pt.) 16 was a 55-year-old with t(8;21) AML who died after transplant, and Pt. 32 (age 56 years) had an inv(16) and relapsed. Although speculative, higher HOX expression might be associated with older patients or a worse outcome. A recent study by Weltermann et al28 concluded that response duration and cytogenetics were independent factors. Thus, it should be possible to identify markers associated with response duration. Conceivable, HOX genes might be among these. We also observed isolated overexpression of individual HOX genes. For instance, patient 51 had the highest levels of HOXA13 despite low expression of other HOXA genes. We have noted similar results from a chromosomal translocation involving the HOXA cluster (unpublished data). This patient, as well as patient 2, was placed in the low expression group by cluster analysis, although both had levels of MEIS1 that fit better in the intermediate category. It is also possible that HOX expression could be affected by chromosomal deletions containing HOX clusters. We had three patients with deletion of No. 7 in the majority of metaphases, which would affect the HOXA cluster. Patient 2, with low HOX expression, was discussed above. However, Nos 15 and 33 had intermediate HOX levels. Thus, deleting one copy of chromosome 7 does not necessarily result in low HOXA gene expression.

What might be responsible for high levels of HOX expression in the subset of patients with intermediate cytogenetics? Rearrangements of MLL, which functions during normal development to maintain HOX expression,29 appear not to be responsible. Neither the t(9;11) or t(11;19) translocations were associated with the highest levels of HOX expression (Table 2); similar results were previously reported by us14 and by Debernardi et al.16 Alterations in Polycomb genes, which normally repress HOX expression,29,30 have not been sufficiently explored. In a preliminary analysis of AML cell lines, we have not observed an association between HOX levels and BMI-1 or EZH2. However, there are several additional Polycomb candidates.

We excluded the possibility that CEBPα mutations, reportedly confined to patients with intermediate cytogenetics,20 were responsible for high HOX expression. Only one mutation (6%) was detected among 17 cases of high HOX expression. However, we did not have enough cases to statistically test the alternate possibility that CEBPα mutations and high HOX expression might be exclusive. Recently, Frohling et al (ASH abstract, #337, 2003) found that C/EBPα mutations were less frequent in cases with normal cytogenetics and FLT3 mutations (28 vs 49%, P=0.01). Debernardi et al (ASH abstract, #1322, 2003) found that C/EBPα mutations were higher in patients with low HOX expression. These observations support our contention that there is a subset of patients with intermediate cytogenetics having high HOX expression, elevated levels of FLT3 and FLT3 mutations, together with a low incidence of C/EBPα mutations. The analysis of FLT3 mRNA levels led to the identification of cases with high expression in the absence of mutations or gene amplification. Although elevated FLT3 expression has been reported to occur in childhood leukemia with MLL rearrangements,31 our observations suggest that this occurrence is more widespread though not random. Among patients with the highest FLT3 levels, case 35 had normal cytogenetics, case 28 had rearrangements of both 6q and 9q, which may have been an unrecognized t(6;9) translocation, and case 39 had a t(9;11) translocation that involves MLL.

In model organisms, signaling between Wnt and Hox pathways has been demonstrated.32,33,34 Other growth factors, such as fibroblast growth factors, have also been tied to Hox expression.35 Our findings that FLT3 levels correlate with expression of multiple HOX genes, and that FLT3 activating mutations correlate with HOXB3, suggest an additional HOX-interacting pathway. Overexpression of Hoxb3 in mice leads to myeloid proliferation, although the cells retain capacity to differentiate.25 This may resemble aspects of FLT3 mutation in human AML.17 In murine thymocytes, Flt3 signaling may induce myeloid differentiation,36 consistent with changes in Hox expression. Mizuki et al37 reported that FLT3 activation is associated with downregulation of Pu.1 and CEBPα in 32Dcl3 cells. Thus, FLT3 may affect AML progression in part through effects on HOX genes and CEBPα, although the hierarchy of these events is unknown. An alternative hypothesis to growth-factor driven HOX expression is that elevated levels of HOX genes and FLT3 reflect a specific stage of differentiation. This is supported by studies in murine hematopoiesis38 and in humans.39 These alternate hypotheses, of course, do not need to be mutually exclusive.

Lastly, is there a role for HOX expression as a prognostic marker in AML? Golub et al40 reported that HOXA9 was among a set of genes that distinguished AML from ALL. Although no multigene expression signature correlated with outcome, HOXA9 (from an array of 6817 genes) showed the best correlation. However, no statistical significance was reported and the analysis involved only 15 patients, seven with long-term survival and eight who failed to achieve remission. Our patients were older and only two were in remission for 36+ months. Thus, our observations and those of Golub et al cannot be directly compared. The nonrandom association of AMLs with high HOX expression and intermediate cytogenetics suggests that this subset may be biologic distinct. If confirmed, it will be important to determine whether there are any distinguishing clinical features. Given the importance of HOX genes in the pathogenesis of acute leukemia, it seems important to better understand these expression phenomena.


  1. 1

    Grimwade D, Walker H, Oliver F, Wheatley K, Harrison C, Harrison G et al. The importance of diagnostic cytogenetics on outcome in AML: analysis of 1,612 patients entered into the MRC AML 10 trial. The Medical Research Council Adult and Children's Leukaemia Working Parties. Blood 1998; 92: 2322–2333.

    CAS  PubMed  PubMed Central  Google Scholar 

  2. 2

    Grimwade D, Walker H, Harrison G, Oliver F, Chatters S, Harrison CJ et al. The predictive value of hierarchical cytogenetic classification in older adults with acute myeloid leukemia (AML): analysis of 1065 patients entered into the United Kingdom Medical Research Council AML11 trial. Blood 2001; 98: 1312–1320.

    CAS  Article  Google Scholar 

  3. 3

    Greer JM, Puetz J, Thomas KR, Capecchi MR . Maintenance of functional equivalence during paralogous Hox gene evolution. Nature 2000; 403: 661–665.

    CAS  Article  Google Scholar 

  4. 4

    Borrow J, Shearman AM, Stanton Jr VP, Becher R, Collins T, Williams AJ et al. The t(7;11)(p15;p15) translocation in acute myeloid leukaemia fuses the genes for nucleoporin NUP98 and class I homeoprotein HOXA9. Nat Genet 1996; 12: 159–167.

    CAS  Article  Google Scholar 

  5. 5

    Fujino T, Suzuki A, Ito Y, Ohyashiki K, Hatano Y, Miura I et al. Single-translocation and double-chimeric transcripts: detection of NUP98-HOXA9 in myeloid leukemias with HOXA11 or HOXA13 breaks of the chromosomal translocation t(7;11)(p15;p15). Blood 2002; 99: 1428–1433.

    CAS  Article  Google Scholar 

  6. 6

    La Starza R, Trubia M, Crescenzi B, Matteucci C, Negrini M, Martelli MF et al. Human homeobox gene HOXC13 is the partner of NUP98 in adult acute myeloid leukemia with t(11;12)(p15;q13). Genes Chromosomes Cancer 2003; 36: 420–423.

    CAS  Article  Google Scholar 

  7. 7

    Panagopoulos I, Isaksson M, Billstrom R, Strombeck B, Mitelman F, Johansson B . Fusion of the NUP98 gene and the homeobox gene HOXC13 in acute myeloid leukemia with t(11;12)(p15;q13). Genes Chromosomes Cancer 2003; 36: 107–112.

    CAS  Article  Google Scholar 

  8. 8

    Taketani T, Taki T, Ono R, Kobayashi Y, Ida K, Hayashi Y . The chromosome translocation t(7;11)(p15;p15) in acute myeloid leukemia results in fusion of the NUP98 gene with a HOXA cluster gene, HOXA13, but not HOXA9. Genes Chromosomes Cancer 2002; 34: 437–443.

    CAS  Article  Google Scholar 

  9. 9

    Taketani T, Taki T, Shibuya N, Kikuchi A, Hanada R, Hayashi Y . Novel NUP98-HOXC11 fusion gene resulted from a chromosomal break within exon 1 of HOXC11 in acute myeloid leukemia with t(11;12)(p15;q13). Cancer Res 2002; 62: 4571–4574.

    CAS  PubMed  Google Scholar 

  10. 10

    Taketani T, Taki T, Shibuya N, Ito E, Kitazawa J, Terui K et al. The HOXD11 gene is fused to the NUP98 gene in acute myeloid leukemia with t(2;11)(q31;p15). Cancer Res 2002; 62: 33–37.

    CAS  PubMed  Google Scholar 

  11. 11

    Raza-Egilmez SZ, Jani-Sait SN, Grossi M, Higgins MJ, Shows TB, Aplan PD . NUP98-HOXD13 gene fusion in therapy-related acute myelogenous leukemia. Cancer Res 1998; 58: 4269–4273.

    CAS  PubMed  Google Scholar 

  12. 12

    Nakamura T, Largaespada DA, Shaughnessy Jr JD, Jenkins NA, Copeland NG . Cooperative activation of Hoxa and Pbx1-related genes in murine myeloid leukaemias. Nat Genet 1996; 12: 149–153.

    CAS  Article  Google Scholar 

  13. 13

    Ayton PM, Cleary ML . Transformation of myeloid progenitors by MLL oncoproteins is dependent on Hoxa7 and Hoxa9. Genes Dev 2003; 17: 2298–2307.

    CAS  Article  Google Scholar 

  14. 14

    Drabkin HA, Parsy C, Ferguson K, Guilhot F, Lacotte L, Roy L et al. Quantitative HOX expression in chromosomally defined subsets of acute myelogenous leukemia. Leukemia 2002; 16: 186–195.

    CAS  Article  Google Scholar 

  15. 15

    Thompson A, Quinn MF, Grimwade D, O'Neill CM, Ahmed MR, Grimes S et al. Global down-regulation of HOX gene expression in PML-RARalpha + acute promyelocytic leukemia identified by small-array real-time PCR. Blood 2003; 101: 1558–1565.

    CAS  Article  Google Scholar 

  16. 16

    Debernardi S, Lillington DM, Chaplin T, Tomlinson S, Amess J, Rohatiner A et al. Genome-wide analysis of acute myeloid leukemia with normal karyotype reveals a unique pattern of homeobox gene expression distinct from those with translocation-mediated fusion events. Genes Chromosomes Cancer 2003; 37: 149–158.

    CAS  Article  Google Scholar 

  17. 17

    Kottaridis PD, Gale RE, Frew ME, Harrison G, Langabeer SE, Belton AA et al. The presence of a FLT3 internal tandem duplication in patients with acute myeloid leukemia (AML) adds important prognostic information to cytogenetic risk group and response to the first cycle of chemotherapy: analysis of 854 patients from the United Kingdom Medical Research Council AML 10 and 12 trials. Blood 2001; 98: 1752–1759.

    CAS  Article  Google Scholar 

  18. 18

    Bennett JM, Cato D, Daniel MT, Flandrin G, Galton DA, Gralnick HR et al. Proposed revised criteria for the classification of acute myeloid leukemia. A report of the French-American-British Cooperative Group. Ann Intern Med 1985; 103: 620–625.

    CAS  Article  Google Scholar 

  19. 19

    Byrd JC, Mrozek K, Dodge RK, Carroll AJ, Edwards CG, Arthur DC et al. Pretreatment cytogenetic abnormalities are predictive of induction success, cumulative incidence of relapse, and overall survival in adult patients with de novo acute myeloid leukemia: results from Cancer and Leukemia Group B (CALGB 8461). Blood 2002; 100: 4325–4336.

    CAS  Article  Google Scholar 

  20. 20

    Preudhomme C, Sagot C, Boissel N, Cayuela JM, Tigaud I, de Botton S et al. Favorable prognostic significance of CEBPA mutations in patients with de novo acute myeloid leukemia: a study from the Acute Leukemia French Association (ALFA). Blood 2002; 100: 2717–2723.

    CAS  Article  Google Scholar 

  21. 21

    Wodnar-Filipowicz A, Lyman SD, Gratwohl A, Tichelli A, Speck B, Nissen C . Flt3 ligand level reflects hematopoietic progenitor cell function in aplastic anemia and chemotherapy-induced bone marrow aplasia. Blood 1996; 88: 4493–4499.

    CAS  PubMed  Google Scholar 

  22. 22

    Haidar JH, Bazarbachi A, Mahfouz R, Haidar HA, Jaafar H, Daher R . Serum Flt3 ligand variation as a predictive indicator of hematopoietic stem cell mobilization. J Hematother Stem Cell Res 2002; 11: 533–538.

    CAS  Article  Google Scholar 

  23. 23

    Crooks GM, Fuller J, Petersen D, Izadi P, Malik P, Pattengale PK et al. Constitutive HOXA5 expression inhibits erythropoiesis and increases myelopoiesis from human hematopoietic progenitors. Blood 1999; 94: 519–528.

    CAS  PubMed  Google Scholar 

  24. 24

    Thorsteinsdottir U, Kroon E, Jerome L, Blasi F, Sauvageau G . Defining roles for HOX and MEIS1 genes in induction of acute myeloid leukemia. Mol Cell Biol 2001; 21: 224–234.

    CAS  Article  Google Scholar 

  25. 25

    Sauvageau G, Thorsteinsdottir U, Hough MR, Hugo P, Lawrence HJ, Largman C et al. Overexpression of HOXB3 in hematopoietic cells causes defective lymphoid development and progressive myeloproliferation. Immunity 1997; 6: 13–22.

    CAS  Article  Google Scholar 

  26. 26

    Antonchuk J, Sauvageau G, Humphries RK . HOXB4 overexpression mediates very rapid stem cell regeneration and competitive hematopoietic repopulation. Exp Hematol 2001; 29: 1125–1134.

    CAS  Article  Google Scholar 

  27. 27

    Kyba M, Perlingeiro RC, Daley GQ . HoxB4 confers definitive lymphoid-myeloid engraftment potential on embryonic stem cell and yolk sac hematopoietic progenitors. Cell 2002; 109: 29–37.

    CAS  Article  Google Scholar 

  28. 28

    Weltermann A, Fonatsch C, Haas OA, Greinix HT, Kahls P, Mitterbauer G et al. Impact of cytogenetics on the prognosis of adults with de novo AML in first relapse. Leukemia 2004; 18: 293–302.

    CAS  Article  Google Scholar 

  29. 29

    Hanson RD, Hess JL, Yu BD, Ernst P, van Lohuizen M, Berns A et al. Mammalian Trithorax and polycomb-group homologues are antagonistic regulators of homeotic development. Proc Natl Acad Sci USA 1999; 96: 14372–14377.

    CAS  Article  Google Scholar 

  30. 30

    Beuchle D, Struhl G, Muller J . Polycomb group proteins and heritable silencing of Drosophila Hox genes. Development 2001; 128: 993–1004.

    CAS  PubMed  Google Scholar 

  31. 31

    Armstrong SA, Kung AL, Mabon ME, Silverman LB, Stam RW, Den Boer ML et al. Inhibition of FLT3 in MLL. Validation of a therapeutic target identified by gene expression based classification. Cancer Cell 2003; 3: 173–183.

    CAS  Article  Google Scholar 

  32. 32

    Bilder D, Graba Y, Scott MP . Wnt and TGFbeta signals subdivide the AbdA Hox domain during Drosophila mesoderm patterning. Development 1998; 125: 1781–1790.

    CAS  PubMed  Google Scholar 

  33. 33

    Maloof JN, Whangbo J, Harris JM, Jongeward GD, Kenyon CA . Wnt signaling pathway controls hox gene expression and neuroblast migration in C elegans. Development 1999; 126: 37–49.

    CAS  PubMed  Google Scholar 

  34. 34

    McWhirter JR, Neuteboom ST, Wancewicz EV, Monia BP, Downing JR, Murre C . Oncogenic homeodomain transcription factor E2A-Pbx1 activates a novel WNT gene in pre-B acute lymphoblastoid leukemia. Proc Natl Acad Sci USA 1999; 96: 11464–11469.

    CAS  Article  Google Scholar 

  35. 35

    McWhirter JR, Goulding M, Weiner JA, Chun J, Murre C . A novel fibroblast growth factor gene expressed in the developing nervous system is a downstream target of the chimeric homeodomain oncoprotein E2A-Pbx1. Development 1997; 124: 3221–3232.

    CAS  PubMed  Google Scholar 

  36. 36

    Bertho JM, Chapel A, Loilleux S, Frick J, Aigueperse J, Gorin NC et al. CD135 (Flk2/Flt3) expression by human thymocytes delineates a possible role of FLT3-ligand in T-cell precursor proliferation and differentiation. Scand J Immunol 2000; 52: 53–61.

    CAS  Article  Google Scholar 

  37. 37

    Mizuki M, Schwable J, Steur C, Choudhary C, Agrawal S, Sargin B et al. Suppression of myeloid transcription factors and induction of STAT response genes by AML-specific Flt3 mutations. Blood 2003; 101: 3164–3173.

    CAS  Article  Google Scholar 

  38. 38

    Pineault N, Helgason CD, Lawrence HJ, Humphries RK . Differential expression of Hox, Meis1, and Pbx1 genes in primitive cells throughout murine hematopoietic ontogeny. Exp Hematol 2002; 30: 49–57.

    CAS  Article  Google Scholar 

  39. 39

    Kawagoe H, Humphries RK, Blair A, Sutherland HJ, Hogge DE . Expression of HOX genes, HOX cofactors, and MLL in phenotypically and functionally defined subpopulations of leukemic and normal human hematopoietic cells. Leukemia 1999; 13: 687–698.

    CAS  Article  Google Scholar 

  40. 40

    Golub TR, Slonim DK, Tamayo P, Huard C, Gaasenbeek M, Mesirov JP et al. Molecular classification of cancer: class discovery and class prediction by gene expression monitoring. Science 1999; 286: 531–537.

    CAS  Article  Google Scholar 

Download references


We gratefully acknowledge support from L'Association pour la Recherche sur le Cancer (JR), La Ligue Nationale Contre le Cancer, Comité de la Vienne et de Charente-Maritime (JR) and the NIH National Cancer Institute (HD) CA97710-01. We thank those individuals who provided several helpful comments in the preparation of this manuscript.

Author information



Corresponding author

Correspondence to H A Drabkin.

Additional information

Supplementary Information

Supplementary Information accompanies the paper on the Leukemia website (

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Roche, J., Zeng, C., Barón, A. et al. Hox expression in AML identifies a distinct subset of patients with intermediate cytogenetics. Leukemia 18, 1059–1063 (2004).

Download citation


  • acute myeloid leukemia
  • homeodomain
  • CEBPα
  • FLT3

Further reading