Search

Increased expression levels of miR-181 family members have been shown to be associated with favorable outcome in patients with cytogenetically normal acute myeloid leukemia. Here we show that increased expression of miR-181a and miR-181b is also significantly (P < .05; Cox regression) associated with favorable overall survival in cytogenetically abnormal AML (CA-AML) patients. We further show that up-regulation of a gene signature composed of 4 potential miR-181 targets (including HOXA7, HOXA9, HOXA11, and PBX3), associated with down-regulation of miR-181 family members, is an independent predictor of adverse overall survival on multivariable testing in analysis of 183 CA-AML patients. The independent prognostic impact of this 4-homeobox-gene signature was confirmed in a validation set of 271 CA-AML patients. Furthermore, our in vitro and in vivo studies indicated that ectopic expression of miR-181b significantly promoted apoptosis and inhibited viability/proliferation of leukemic cells and delayed leukemogenesis; such effects could be reversed by forced expression of PBX3. Thus, the up-regulation of the 4 homeobox genes resulting from the down-regulation of miR-181 family members probably contribute to the poor prognosis of patients with nonfavorable CA-AML. Restoring expression of miR-181b and/or targeting the HOXA/PBX3 pathways may provide new strategies to improve survival substantially.

The human myeloid-lymphoid leukemia gene, MLL (also calledALL-1, Htrx, or HRX ), maps to chromosomal band 11q23. MLL is involved in translocations that result in de novo acute lymphoblastic leukemia (ALL), acute myelogenous leukemia (AML), mixed lineage leukemia, and also in therapy AML (t-AML) and therapy ALL (t-ALL) resulting from treatment with DNA topoisomerase II (topo II) targeting drugs. MLL can recombine with more than 30 other chromosomal bands, of which 16 of the partner genes have been cloned. Breaks in MLL occur in an 8.3-kb breakpoint cluster region (BCR) encompassing exons 5 through 11. We recently demonstrated that 75% of de novo patient breakpoints in MLL mapped in the centromeric half of the BCR between two scaffold-associated regions (SAR), whereas 75% of the t-AML patient breakpoints mapped to the telomeric half of the BCR within a strong SAR. We have mapped additional structural elements in the BCR. An in vivo DNA topo II cleavage site (induced with several different drugs that target topo II) mapped near exon 9 in three leukemia cell lines. A strong DNase I hypersensitive site (HS) also mapped near exon 9 in four leukemia cell lines, including two in which MLL was rearranged [a t(6;11) and a t(9;11)], and in two lymphoblastoid cell lines with normalMLL. Two of the leukemia cell lines also showed an in vivo topo II cleavage site. Our results suggest that the chromatin structure of the MLL BCR may influence the location of DNA breaks in both de novo and therapy-related leukemias. We propose that topo II is enriched in the MLL telomeric SAR and that it cleaves the DNase I HS site after treatment with topo II inhibitors. These events may be involved in recombination associated with t-AML/t-ALL breakpoints mapping in the MLL SAR.

CBFA2(AML1) has emerged as a gene critical in hematopoiesis; its protein product forms the DNA-binding subunit of the heterodimeric core-binding factor (CBF) that binds to the transcriptional regulatory regions of genes, some of which are active specifically in hematopoiesis. CBFA2 forms a fusion gene with ETO andMDS1/EVI1 in translocations in myeloid leukemia and withETV6(TEL) in the t(12;21) common in childhood pre-B acute lymphoblastic leukemia. We have analyzed samples from 30 leukemia patients who had chromosome rearrangements involving 21q22 by using fluorescence in situ hybridization (FISH). Our analysis showed that 7 of them involved CBFA2 and new translocation partners. Two patients had a t(17;21)(q11.2;q22), whereas the other 5 had translocations involving 1p36, 5q13, 12q24, 14q22, or 15q22. Five of these novel breakpoints in CBFA2 occurred in intron 6; this same intron is involved in the t(3;21). One breakpoint mapped to the t(8;21) breakpoint region in intron 5, and 1 mapped 5′ to that region. All 7 CBFA2 rearrangements resulted from balanced translocations. All 7 patients had myeloid disorders (acute myeloid leukemia or myelodysplastic syndrome); 2 were de novo and 5 had treatment histories that included topoisomerase II targeting agents. The association of therapy-related disorders with translocations involving CBFA2 was significant by Fisher’s exact test (P < .003). These results provide further evidence that this region of CBFA2 is susceptible to breakage in cells exposed to topoisomerase II inhibitors. © 1998 by The American Society of Hematology.

BNORMALITIES of the short arm of chromosome 12 A are relatively common in childhood acute lymphoblastic leukemia (ALL). Approximately 5% to 7% of children with ALL have cytogenetic evidence of a translocation involving 12p, whereas 3% to 5% have deletions that suggest the presence of a tumor-suppressor gene at this location.’ Through the use of sensitive molecular techniques, 12p or 12~12-13 loss of heterozygosity (LOH) has been shown in approximately 25% of childhood ALL cases.z4 Fluorescence in situ hybridization (FISH) mapping has detected a minimum region of overlap for the 12p deletions between the TEL. and CDKNIB (KIpl) genes.’ Recently, chromosomal translocations involving the TEL gene at 12~13 have been cloned in several hematopoietic In ALL cells with the t(12;21)(p13;q22), the 5’ part of TEL is fused with the AMLl gene.’.” Abnormalities of 12p, especially the t(12;21), are more reliably detected by FISH than by classical cytogenetics because the translocated portions of 12p and 21q are virtually identical cytogenetically.’* When FISH was combined with Southern blotting and reverse transcriptasepolymerase chain reaction (RT-PCR), the t(12;21) was identified as a recurring chromosomal abnormality in 16% to 25% of childhood B-lineage Despite the identification of the fusion partners in the t(12;21), the actual function of the TEL-AML1 and AMLl-TEL fusion proteins in promoting malignant transformation is unclear. Whether the TEL-AMLl fusion alone is necessary and sufficient for malignant transformation in ALL and whether TEL inactivation has a role in leukemogenesis is currently unknown. Loss of the CDKN2 (p16) gene at 9p21 is a common genetic abnormality in ALL and a variety of other malignancies. Homozygous CDKN2 deletions have been detected in approximately 15% of B-lineage ALL and 75% of T-lineage ALL The p16 protein functions as an inhibitor of cyclin-dependent kinase 4, and normally acts to stop cell cycle progres~ion.’~**~ Whether loss of p16 alone is necessary and sufficient for malignant transformation in childhood ALL has not been determined. In this report, 6 cell lines with the TEL-AMLl fusion transcript are described. Although only 1 of the 6 cell lines translocated allele of EL and 5 cell lines lacked expression of CDKN2. Moreover, in 2 patients (1 with the TEL-AMLI transcript and 1 without), EL exprwsion was lost with disease progression; ie, EL was expressed in the initial cell lines (established at diagnosis or first relapse) whereas 7Ef was not expressed in the cell lines established from these patients in late-stage disease. These data show the coexistence of multiple genetic defects in childhood B-lineage ALL. Cell lines with t(12;21) will faciliate the study of TEL-AML1 and AMLl-TEL fusion proteins as well as EL and CDKM gene inactivation in leukemia transformation and progression.

A major unresolved question for 11q23 translocations involving MLL is the chromosomal mechanism(s) leading to these translocations. We have mapped breakpoints within the 8.3-kb BamHI breakpoint cluster region in 31 patients with acute lymphoblastic leukemia and acute myeloid leukemia (AML) de novo and in 8 t-AML patients. In 23 of 31 leukemia de novo patients, MLL breakpoints mapped to the centromeric half (4.57 kb) of the breakpoint cluster region, whereas those in eight de novo patients mapped to the telomeric half (3.87 kb). In contrast, only two t-AML breakpoints mapped in the centromeric half, whereas six mapped in the telomeric half. The difference in distribution of the leukemia de novo breakpoints is statistically significant (P = .02). A similar difference in distribution of breakpoints between de novo patients and t-AML patients has been reported by others. We identified a low- or weak-affinity scaffold attachment region (SAR) mapping just centromeric to the breakpoint cluster region, and a high-affinity SAR mapping within the telomeric half of the breakpoint cluster region. Using high stringency criteria to define in vitro vertebrate topoisomerase II (topo II) consensus sites, one topo II site mapped adjacent to the telomeric SAR, whereas six mapped within the SAR. Therefore, 74% of leukemia de novo and 25% of t-AML breakpoints map to the centromeric half of the breakpoint cluster region map between the two SARs; in contrast, 26% of the leukemia de novo and 75% of the t-AML patient breakpoints map to the telomeric half of the breakpoint cluster region that contains both the telomeric SAR and the topo II sites. Thus, the chromatin structure of the MLL breakpoint cluster region may be important in determining the distribution of the breakpoints. The data suggest that the mechanism(s) leading to translocations may differ in leukemia de novo and in t-AML.

The inversion of chromosome 16 [inv(16)] in acute myeloid leukemia (AML) is associated with a p-arm deletion in a subset of patients. The inversion results in two fusion genes: 5′-CBFB/MYH11–3′ on 16p and 5′- MYH11/CBFB-3′ on 16q. We have studied cells from 42 patients with inv(16) (38 patients) or t(16;16) (four patients) to define the frequency and characteristics of the deletion further. Using fluorescence in situ hybridization (FISH) with probes from cosmids, cosmid contigs, and yeast artificial chromosomes (YACs), we found that six patients with inv(16) had a deletion of between 150 and 350 kb centromeric to the p-arm inversion breakpoint cluster region (p-ibc). This region was shown to contain the 5′ portion of the myosin heavy chain (MYH11) gene. YACs containing the p-ibc, which had been useful as FISH probes in the diagnosis of inv(16), detected the inversion in deletion as well as nondeletion patient cells. Thus, the deleted region identified in patients is entirely contained within the human genomic content of the YACs. Southern blot experiments using probes flanking the p-ibc indicated that the deletion removes segments within 10 kb centromeric of the p-ibc. Reverse transcription-polymerase chain reaction (RT-PCR) using primers from the 5′ region of CBFB and the 3′ region of MYH11 (distal to the p-ibc) produced the 5′-CBFB/MYH11–3′ chimeric transcript in inv(16)/del patients. These data confirm that the 5′-CBFB/MYH11–3′ chimeric transcript, rather than the reciprocal 5′- MYH11/CBFB-3′, is the critical product for chromosome 16-related leukemogenesis.

Two general types of clonal chromosome abnormality are observed in de novo acute myeloid leukemia (AML): the unbalanced aberrations with visible gain or loss of chromosome material and the balanced aberrations without such visible gain or loss. AML can be induced by therapy with cytostatic drugs and radiation. The alkylating agents reacting directly with DNA induce AML which often presents as myelodysplasia with unbalanced aberrations, primarily loss of chromosome material. Cytostatic agents targeting DNA-topoisomerase II, frequently administered together with alkylating agents or cisplatin, induce the same type of leukemia. In addition, they often induce another type with a more rapid onset and with specific balanced chromosome aberrations rarely observed after therapy with alkylating agents alone. All of the most important chromosome aberrations found in de novo AML are now also found in therapy-related AML (t-AML); thus, t- AML may serve as a model in the search for mechanisms leading to the development of AML in general. Unbalanced chromosome aberrations with partial deletions or with loss of whole chromosomes may develop as a result of alkylation of DNA or other cellular targets. Balanced chromosome aberrations, on the other hand, may develop as illegitimate recombinations related to the activity of DNA-topoisomerase II. The balanced translocations contribute to malignant transformation by the formation of abnormal chimeric genes, whereas deletions may contribute by the loss of putative tumor suppressor genes. In either situation, the chromosome changes provide the altered cells with a proliferative advantage compared with normal cells.

Pericentric inversion of chromosome 16 [inv(16)(p13q22)] and the related t(16;16)(p13;q22) are seen in a subset of acute myelogenous leukemia (AML) phenotypically and prognostically differing from other cases. We have recently shown that inv(16) results in fusion of CBFB/PEBP2B, a gene encoded at 16q22 to MYH11, a smooth muscle myosin heavy chain gene encoded at 16p13. Chimeric transcripts consisting of upstream CBFB fused to downstream MYH11 coding sequences result from this fusion. In this study we have examined a series of 37 of these cases using reverse transcriptase-polymerase chain reaction (RT-PCR) to detect expression of a hybrid CBFB/MYH11 transcript. Chimeric cDNAs were detected in all but 1 of 37 leukemias with typical inv(16) or t(16;16). Such chimeric products were not seen in a case with inv(16)(p13q24) (ie, a variant q arm breakpoint) or any of 10 cases of AML without these chromosomal changes. Four different chimeric transcripts were found, representing differing fusion points within MYH11 spliced to position 495 of CBFB. Primer sets are described for efficient amplification of these different cDNA forms. Amplification of cDNA showed that all but 17 codons of the CBFB coding sequence are included in the abnormal transcripts. RT-PCR was shown to be highly sensitive and potentially useful for detection of leukemic cells during morphologic remission.

Rearrangements of chromosome band 11q23 are common in infant leukemias, comprising more than 70% of the observed chromosome abnormalities in children less than 1 year of age. The MLL gene, which is located at the 11q23 breakpoint in infant, childhood, and adult acute leukemias, has been cloned and has homology to the Drosophila trithorax gene. The breakpoints in MLL are restricted to an 8.3-kilobase pair (kb) region of the gene that is involved in translocations with as many as 29 other chromosomal regions in a number of phenotypically distinct acute leukemias. We have detected an identical, clonal, nonconstitutional rearrangement of the MLL gene in peripheral blood cells from a pair of female infants twins with acute lymphoblastic leukemia (ALL) and a t(11;19)(q23;p13.3). The detection of nonidentical IGH rearrangements suggests that the MLL rearrangement took place in a B-cell precursor or hematopoietic stem cell in one twin which was transferred in utero to the other fetus resulting in ALL with an identical aneuploid karyotype in both infants. We speculate that the other MLL-related infant leukemias may also develop in utero, and that the rearrangements may occur consistently in stem cells or early precursor cells, accounting for the frequency of mixed-lineage leukemia in infants.

Chromosome band 11q23 is frequently involved in acute myeloid leukemia (AML) and acute lymphoblastic leukemia (ALL) de novo, as well as in myelodysplastic syndromes (MDS) and lymphoma. Five percent to 15% of patients treated with chemotherapy for a primary neoplasm develop therapy-related AML (t-AML) that may show rearrangements, usually translocations involving band 11q23 or, less often, 21q22. These leukemias develop after a relatively short latent period and often follow the use of drugs that inhibit the activity of DNA-topoisomerase II (topo II). We previously identified a gene, MLL (myeloid-lymphoid leukemia or mixed-lineage leukemia), at 11q23 that is involved in the de novo leukemias. We have studied 17 patients with t-MDS/t-AML, 12 of whom had cytogenetically detectable 11q23 rearrangements. Ten of the 12 t-AML patients had received topo II inhibitors and 9 of these, all with balanced translocations of 11q23, had MLL rearrangements on Southern blot analysis. None of the patients who had not received topo II inhibitors showed an MLL rearrangement. Of the 5 patients lacking 11q23 rearrangements, some of whom had monoblastic features, none had an MLL rearrangement, although 4 had received topo II inhibitors. Our study indicates that the MLL gene rearrangements are similar both in AML that develops de novo and in t-AML. The association of exposure to topo II- reactive chemotherapy with 11q23 rearrangements involving the MLL gene in t-AML suggests that topo II may play a role in the aberrant recombination events that occur in this region both in AML de novo and in t-AML.

The translocation between chromosomes 8 and 21, t(8;21) (q22;q22), is the most frequent abnormality in acute myeloid leukemia (AML) with French-American-British type M2 (FAB-M2) morphology. The breakpoints in this translocation have been characterized at the molecular level, and the genes involved are AML1 on chromosome 21 and ETO on chromosome 8. The rearrangement of the two chromosomes results in a fusion gene and in the production of a consistent fusion transcript on the der(8) chromosome. We have used oligonucleotide primers derived from both sides of the fusion cDNA junction and reverse transcription-polymerase chain reaction (RT-PCR) to analyze six AML-M2 patients with a t(8;21) during various stages of their disease. Two patients studied at diagnosis and one studied at first relapse are alive off therapy and in continuous complete remission for 83 to 94 months. We have detected the AML/ETO fusion transcript in recent peripheral blood samples from each of them. Three other patients also had a fusion transcript detected after 1 to 4 months in remission. Two of these patients subsequently relapsed and died whereas the third patient is alive and in continuous complete remission 70 months later. Thus, our preliminary data suggest that cells with the translocation are still circulating in t(8;21) patients in long-term remission. This finding raises serious questions regarding the interpretation of positive results obtained only with this technique that may not be suitable to decide appropriate further treatment for patients in clinical remission.

The (8;21)(q22;q22) translocation is a frequent karyotypic abnormality seen in approximately 40% of patients with acute myeloid leukemia subtype M2 (AML-M2) and an abnormal karyotype. The translocation interrupts two genes, AML1 on chromosome 21 and ETO on chromosome 8, that are consequently fused in the der(8) chromosome to produce a novel chimeric gene and message. Selected genomic DNA probes from chromosome 21 and from chromosome 8 near the breakpoint junction detect rearrangements in the DNA of about 80% of the patients with the rearrangement at diagnosis and in relapse. We analyzed the DNA of 20 patients with t(8;21) AML by standard Southern blot with probes originating from chromosomes 21 and 8 near the breakpoint junction, and we identified rearranged bands in 17 of the 20 patients at diagnosis and in relapse. We also used the polymerase chain reaction (PCR) with appropriate primers from the AML1 and ETO genes to amplify the cDNAs from a cell line with the t(8;21) and from seven AML patients with the t(8;21). We detected a fused transcript in the cell line and in all of the patients analyzed, including three patients who did not show any rearrangement by Southern blot analysis and one patient in hematologic remission, who later relapsed. Combining the results from Southern blot and PCR analysis, we could detect the t(8;21) in all of the patients tested. These results indicate that, whereas several DNA probes used as genetic markers do detect the t(8;21) in most, but not all Southern blots of patients with AML, PCR amplification with primers from AML1 and ETO can be used as a more sensitive and accurate means for detecting this chromosomal abnormality, and for observing the patients' response to therapy.

We have performed a retrospective analysis of the clinical, morphologic, and cytogenetic findings in 26 patients diagnosed between January 1969 and September 1991 with acute erythroblastic leukemia de novo (EL or AML-M6). Clonal chromosomal abnormalities were found in 20 (77%) patients (95% confidence interval [CI], 61% to 93%). Loss of all or part of the long arm (q) of chromosomes 5 and/or 7 was observed in 17 (65%) patients (95% CI, 47% to 83%). In addition, the karyotypes were often complex, with multiple abnormalities and subclones. Among the remaining nine patients, six had a normal karyotype and one each had trisomy 8, t(3;3), or t(3;5). The overall frequency of abnormalities of chromosomes 5 and/or 7 observed in our M6 patients is similar to that observed in our patients with therapy-related acute myeloid leukemia (t-AML; 99 of 129 patients, 77%), but substantially higher than that noted in our other patients with AML de novo (French- American-British [FAB] subtypes M1-M5: 52 of 334 patients, 16%). Our M6 patients with abnormalities of chromosomes 5 and/or 7 were older and had a shorter median survival (16 v 77 weeks [P = .005]) than did the M6 patients without these abnormalities. We found no correlation between morphologic features and either cytogenetic abnormalities or clinical outcome. Of note was the finding that the percentage of myeloblasts, which may account for only a small fraction of the total marrow elements when the revised FAB criteria are applied, had no bearing on prognosis. We conclude that acute erythroblastic leukemia, when defined by morphologic criteria, consists of two distinctive subgroups: one group tends to be older, has complex cytogenetic abnormalities, especially of chromosomes 5 and/or 7, and shares biologic and clinical features with t-AML; the other group, with simple or no detectable cytogenetic abnormalities, has a more favorable prognosis when treated with intensive chemotherapy.

Trisomy 12 is the most common cytogenetic abnormality in chronic lymphocytic leukemia (CLL), and a number of studies have suggested that it may be an adverse prognostic indicator. We have evaluated the usefulness of fluorescence in situ hybridization with a chromosome 12- specific probe as a simple means for detecting trisomy 12 in interphase cells. Forty cases of B-cell CLL previously studied with conventional cytogenetic techniques were analyzed with a biotinylated probe to the centromeric region of chromosome 12. Thirty of these retrospective cases could be reevaluated with in situ hybridization. Our analysis showed three hybridization signals (ie, trisomy 12) in interphase cells from seven of seven cases found previously to have trisomy 12. Trisomy 12 was also detected in five additional cases: in one case thought to have a normal karyotype, in two cases that had been inadequate for routine cytogenetic analysis, and in two cases that had been found to have an abnormal karyotype without trisomy 12. In a prospective series of 20 newly accrued CLL cases, all cases were analyzed successfully by in situ hybridization and six (30%) showed trisomy 12. We were able to perform the analysis on routinely prepared and previously Wright- stained peripheral blood smears. We conclude that fluorescence in situ hybridization is a simple means for the detection of trisomy 12 in CLL. The technique is more sensitive than conventional cytogenetic analysis and would be a useful tool in clinical studies.

We have examined a population of patients with acute myeloid leukemia (AML) or myelodysplastic syndrome (MDS) for loss of heterozygosity of polymorphic markers on chromosomes 5 and 7. The rationale for this study was the observation that the majority of patients with therapy- related leukemia (t-AML or t-MDS), resulting from cytotoxic treatment for prior malignancies, have loss of chromosome 5 and/or 7 or deletions involving the long arms of one or both of these chromosomes. This cytogenetic finding suggested that tumor-suppressor genes, important in the development of AML, may be located in these chromosomal regions. We analyzed a total of 60 patients, 43 with primary MDS/AML de novo and 17 with t-MDS/t-AML. Leukemia cells were evaluated for restriction fragment length polymorphisms (RFLPs). Leukemia cell genotypes were compared with lymphoblastoid cell genotypes from the same patients. Two cases of loss of heterozygosity were identified from chromosomes lacking visible deletions: one involving chromosome 5 in a patient with AML de novo who had a visible deletion of 5q at a later stage of the disease, and one involving chromosome 7 in a patient with t-AML. We conclude that allele loss from loci on chromosomes 5 and 7 in MDS/AML, when it occurs, usually results from major deletion or simple chromosome loss, rather than from mitotic recombination or chromosome loss with duplication of the remaining homologue.

The partner gene of MLL was identified in a patient with treatment-related acute myeloid leukemia in which the karyotype suggested t(3;11)(q25;q23). Prior therapy included the DNA topoisomerase II inhibitors, teniposide and doxorubicin. Southern blot analysis indicated that the MLL gene was involved in the translocation. cDNA panhandle polymerase chain reaction (PCR) was used, which does not require partner gene-specific primers, to identify the chimeric transcript. Reverse-transcription of first-strand cDNAs with oligonucleotides containing known MLL sequence at the 5′ ends and random hexamers at the 3′ ends generated templates with an intra-strand loop for PCR. In-frame fusions of either MLLexon 7 or exon 8 with the GMPS (GUANOSINE 5′-MONOPHOSPHATE SYNTHETASE) gene from chromosome band 3q24 were detected. The fusion transcript was alternatively spliced. Guanosine monophosphate synthetase is essential for de novo purine synthesis. GMPS is the first partner gene ofMLL on chromosome 3q and the first gene of this type in leukemia-associated translocations.

The partner gene of MLL was identified in a patient with treatment-related acute myeloid leukemia in which the karyotype suggested t(3;11)(q25;q23). Prior therapy included the DNA topoisomerase II inhibitors, teniposide and doxorubicin. Southern blot analysis indicated that the MLL gene was involved in the translocation. cDNA panhandle polymerase chain reaction (PCR) was used, which does not require partner gene-specific primers, to identify the chimeric transcript. Reverse-transcription of first-strand cDNAs with oligonucleotides containing known MLL sequence at the 5′ ends and random hexamers at the 3′ ends generated templates with an intra-strand loop for PCR. In-frame fusions of either MLLexon 7 or exon 8 with the GMPS (GUANOSINE 5′-MONOPHOSPHATE SYNTHETASE) gene from chromosome band 3q24 were detected. The fusion transcript was alternatively spliced. Guanosine monophosphate synthetase is essential for de novo purine synthesis. GMPS is the first partner gene ofMLL on chromosome 3q and the first gene of this type in leukemia-associated translocations.

It is hard today, in 2008, when genetic changes in cancer and specifically hematologic malignancies are accepted as central players, to look back 50 years and realize how different the landscape was then. In 1957, knowledge of the correct chromosome number of man was 1 year old,[1][1] and that DNA

We have identified an identical reciprocal translocation between the long arms of chromosomes 3 and 21 with breakpoints at bands 3q26 and 21q22, [t(3;21)(q26;q22)], in the malignant cells from five adult patients with therapy-related myelodysplastic syndrome (t-MDS) or acute myeloid leukemia (t-AML). Primary diagnoses were Hodgkin's disease in two patients and ovarian carcinoma, breast cancer, and polycythemia vera in one patient each. Patients had been treated with chemotherapy including an alkylating agent for their primary disease 1 to 18 years before the development of t-MDS or t-AML. We have not observed the t(3;21) in over 1,500 patients with a myelodysplastic syndrome or acute myeloid leukemia arising de novo or in over 1,000 patients with lymphoid malignancies. We have previously reported that the t(3;21) occurs in Philadelphia chromosome-positive chronic myelogenous leukemia (CML). Thus, the t(3;21) appears to be limited to t-MDS/t-AML and CML, both of which represent malignant disorders of an early hematopoietic precursor cell. These results provide a new focus for the study of therapy-related leukemia at the molecular level.

Therapy-related myelodysplasia and myeloid leukemia (t-MDS/t-AML) is a distinctive clinical syndrome occurring after exposure to chemotherapy (CT) or radiotherapy (RT). We report findings on 306 consecutive patients referred to our institution with morphologic review and cytogenetic analyses. Since 1972, 141 males and 165 females with a median age of 51 years (range, 3-83 years) at primary diagnosis and 58 years (range, 6-86 years) at secondary diagnosis were analyzed. Patients had been administered various cytotoxic agents, including alkylating agents (240 patients, 78%) and topoisomerase 2 inhibitors (115 patients, 39%). One hundred twenty-one (40%) had undergone CT alone, 43 (14%) had undergone RT alone, and 139 (45%) had undergone both modalities. At diagnosis of t-MDS/t-AML, 282 (92%) had clonal abnormalities involving chromosome 5 (n = 63), chromosome 7 (n = 85), chromosomes 5 and 7 (n = 66), recurring balanced rearrangements (n = 31), other clonal abnormalities (n = 39), or normal karyotype (n = 24). Abnormalities of chromosome 5, 7, or both accounted for 76% of all cases with an abnormal karyotype. Seventeen patients acquired t-MDS/t-AML after autologous stem cell transplantation, but no unique pattern of cytogenetic abnormalities was observed. Shorter latency was observed for patients with balanced rearrangements (median, 28 vs 67 months; P < .0001). Patients with acute leukemia were more likely to have balanced rearrangement than those with myelodysplasia (28% vs 4%; P < .0001). Median survival time after diagnosis of t-MDS/t-AML was 8 months; survival at 5 years was less than 10%. These data confirm and extend previous associations between clinical, morphologic, and cytogenetic findings in t-MDS/t-AML. (Blood. 2003;102:43-52)

The report by Popovic and colleagues in this issue of Blood , while a milestone on our way to personalizing immunologic cancer therapy, highlights the pitfalls in identifying specific peptides for immunotherapy. 1

A critical area of cancer biology is the study of the deregulation of noncoding RNAs called microRNAs (miRNAs). Acute leukemia represents one of the most deadly cancers in the United States. One subset of leukemia with a poor to intermediate clinical outcome are chromosomal translocations involving Mixed Lineage Leukemia (MLL). As MLL-translocations are sufficient to drive leukemogenesis, and few additional mutations are observed in patients, it is imperative to understand the biology driving leukemogenesis. Previously, we and others have shown that several miRNAs are deregulated in MLL-rearranged Acute Myeloid Leukemia (AML). To identify miRNAs that are driving leukemogenesis we performed messenger RNA and miRNA expression profiling on primary patient samples and identified microRNA-9 (miR-9) as specifically overexpressed in MLL-rearranged AML. We further confirmed this observation using publically available microRNA sequencing data from the Cancer Genome Atlas (TCGA) and several AML cell lines. After showing that MLL directly binds and regulates miR-9 we show that depletion of MLL-fusion expression leads to the loss of miR-9 expression. Using publically available Illumina 450K methylation data from TCGA, we show that there is no significant difference in modified cytosine between miR-9 high and miR-9-low patients, suggesting that expression of miR-9 is likely not be regulated by DNA methylation machinery in AML patients. We show that miR-9 in the presence of MLL-AF9 (a common MLL-fusion) promotes colony growth over multiple passages while blocking miR-9 using a miR-9 sponge remarkably inhibits MLL-fusion-mediated cell transformation. Furthermore, we show that miR-9 increases proliferation and reduces apoptosis of human MLL-rearranged leukemic cells in vitro using MTT and Caspase 3/7 assays. We then show that co-transfection of miR-9 with MLL-AF9 in a bone marrow transplantation assay results in a higher leukemia burden in vivo compared to MLL-AF9 alone and promotes an immature cellular phenotype. Using microarray data we found several putative miR-9 targets by identifying genes that had an inverse correlation to miR-9. Next, we verified several genes were being inhibited by miR-9 such as Ras homology gene family member H (RHOH) and Ring1- and YY1-binding protein (RYBP). To understand the role of miR-9 in context with other miRNAs we did an association analysis of the top 300 differentially expressed miRNAs in the TCGA dataset. We found interestingly, that two of the miR-9 genes, miR-9-1 and miR-9-2, are highly correlated with each other across all the patients although they are located on distinct chromosomes. We also found that several other miRs were either negatively (e.g., miR-130a and miR-221) or positively (e.g., miR-191 and miR-642) associated with miR-9 expression, suggesting that these miRs might be operating either cooperatively or antagonistically in a complex circuitry. To support this hypothesis we found in univariate analysis that miR-9 itself was not a good predictor of patient survival but was a better predictor when combined with other miRs including the miR-181 family. Together this suggests that miR-9 is an important and critical regulator of MLL-rearranged AML and is a very good candidate for potential therapeutic targeting. Disclosures: No relevant conflicts of interest to declare.

Abstract 3522 MLL (mixed lineage leukemia) gene rearrangements account for about 10% of human acute leukemias, including ∼80% of infant acute leukemia. At present, the majority of patients die within two years of diagnosis, and a more effective therapeutic strategy is thus urgently needed. MLL fusion proteins induce aberrant expression of a group of homeobox gene super-family members, including HOXA and co-factors such as MEIS1 and PBX3. Overexpression of individual HOXA genes can induce myeloproliferation and block differentiation. Co-expression of Meis1 and Hoxa9 is sufficient to transform normal hematopoietic progenitor cells and to induce a rapidly fatal leukemia in transplanted mice, and their aberrant overexpression is required for the induction and maintenance of MLL-rearranged leukemia. Although Pbx proteins are known to interact with Hox proteins, thereby increasing the DNA-binding affinity of the latter and enhancing the transcription of downstream target genes, little is known about the interaction between Pbx proteins and Hox proteins (e.g., Hoxa9) in cell transformation and leukemogenesis. Recently, we showed that increased expression of a four-homeobox-gene signature (HOXA7, HOXA9, HOXA11 and PBX3) was an independent predictor of shorter overall survival in patients with cytogenetically abnormal acute myeloid leukemia (CA-AML) (Li Z., et al., Blood. 2012). Our analysis of the expression profiles of three independent large-scale patient sets showed that PBX3 was the only member of the PBX family that was consistently co-expressed with HOXA9 in various subtypes of CA-AML, particularly in MLL-rearranged AML; in contrast, both PBX1 and PBX2 tended to exhibit an inverse correlation of expression with HOXA9 in CA-AML. We then investigated the role of PBX3 in CA-AML, because its function in leukemia was unclear. We found a similar pattern of co-expression of Hoxa9 and Pbx3 in MLL fusion-mediated mouse leukemia models. We then showed that depletion of Pbx3 (but not Pbx1 and Pbx2) by Pbx3 shRNA dramatically inhibited MLL-AF9 induced transformation/immortalization of mouse normal bone marrow progenitor cells (about 50% reduction in colony numbers and about 70% reduction of number of cells in each colony). Furthermore, we demonstrated that forced expression of PBX3 exhibited a significantly synergistic effect with HOXA9 in promoting cell transformation/immortalization in vitro. In mouse bone marrow reconstitution/transplantation assay, the PBX3+HOXA9 mice developed leukemia significantly faster than HOXA9 alone with overall median survival of 71 days versus 140 days (P Disclosures: No relevant conflicts of interest to declare.

Abstract 2332 MicroRNA-9 (miR-9) is required for normal neurogenesis and organ development. The expression of miR-9 is altered in several types of solid tumors suggesting that it may have a function in cell transformation. However the role of this miR in normal hematopoiesis and leukemogenesis is unknown. Here we show that miR-9 is expressed at low levels in hematopoietic stem/progenitor cells (HSCs/HPCs), and that it is upregulated during hematopoietic differentiation. Ectopic expression of miR-9 strongly accelerates terminal myelopoiesis, while promoting apoptosis in vitro and in vivo. In addition, the inhibition of miR-9 in HPC with a miRNA sponge blocks myelopoiesis. EVI1, required for normal embryogenesis, and is considered an oncogene because inappropriate upregulation induces malignant transformation in solid and hematopoietic cancers. In vitro, EVI1 severely affects myeloid differentiation. Here we show that EVI1 binds to the promoter of miR-9–3 leading to DNA hypermethylation of the promoter as well as repression of miR-9. We also show that ectopic miR-9 reverses the myeloid differentiation block that is induced by EVI1. Our findings suggest that inappropriately expressed EVI1 delays or blocks myeloid differentiation, at least in part by DNA hypermethylation and downregulation of miR-9. It was previously reported that FoxOs genes inhibit myeloid differentiation and prevent differentiation of leukemia initiating cells. Here we identify FoxO3 and FoxO1 as new direct targets of miR-9 in hematopoietic cells, and we find that upregulation of FoxO3 in miR-9-positive cells reduces the acceleration of myelopoiesis. These results reveal a novel role of miR-9 in myelopoiesis and in the pathogenesis of EVI1-induced myeloid neoplasms. They also provide new insights on the potential chromatin-modifying role of oncogenes in epigenetic changes in cancer cells. Disclosures: No relevant conflicts of interest to declare.

Abstract 3499 MicroRNAs (miRNAs), a class of small, non-coding RNAs, are important for posttranscriptional gene regulation in both health and disease. Expression of miRNAs is under stringent regulation at both transcriptional and post-transcriptional levels. Disturbance at either level could cause dysregulation of miRNAs. However, though altered expression of many miRNAs has been reported in various cancers, including acute myeloid leukemia (AML), their dysregulational mechanisms and pathologic functions remain less well understood. Here we report that mature miR-150 level is significantly downregulated in most AML samples, including those with rearrangements of the mixed lineage leukemia (MLL) gene. Strikingly, we found that despite the over 75% decrease of mature miR-150, its primary and precursor transcript abundance is increased to 2∼4 fold in human MLL-associated AML, relative to normal controls. Interestingly, we show that while MLL fusion proteins can bind to the promotor region of miR-150 and promote its primary transcription, they also negatively regulate the maturation process of miR-150 through the MYC/LIN28 functional axis. MiR-150 has been implicated as either an oncogene or a tumor suppressor in various types of solid tumors. However, its function in the pathogenesis of AML is unknown. Here we showed that ectopic expression of miR-150 dramatically inhibited cell growth and promoted apoptosis of human MLL-associated leukemic cells. Furthermore, using colony-forming/replating assays, we found that co-transduction of miR-150 and MLL-AF9 (a fusion gene resulting from t(9;11)) into mouse bone marrow (BM) progenitor cells, caused a significant reduction in colonies (down to 1∼10%; p Previous studies have shown that there is an autoregulatory feedback loop between FLT3/MYB and HOXA9/MEIS1, and the latter two are critical downstream targets of MLL fusion proteins. In addition, FLT3 has been identified as an upstream regulator of MYC, while MYC is also a downstream target of MLL fusion proteins and an upstream regulator of Lin28. These previous findings together with the data we reported above suggest that there is a critical MLL-fusion/MYC/LIN28-miR-150-FLT3/MYB/HOXA9/MEIS1 regulatory circuit in MLL-associated leukemia (see Fig. 1). In this circuit, MLL fusion proteins function as the driver, and their presence leads to the significant up-regulation of all six downstream genes, MYC, LIN28, FLT3, MYB, HOXA9, and MEIS1, as well as the primary transcription of miR-150. The up-regulation of MYC/LIN28 results in the blockade of the miR-150 maturation process. This in turn leads to the release of miR-150 inhibition on FLT3 and MYB expression, which would enhance the expression of HOXA9, MEIS1, MYC, and LIN28, and further enhance/maintain the blockade of miR-150 maturation. As a result, the cells reach and maintain high levels of MYC/LIN28/FLT3/MYB/HOXA9/MEIS1, and thereby transform the cells and lead to leukemogenesis. Our further systematic studies confirmed the existence/fidelity of this regulatory circuit in MLL-associated leukemia. Taken together, we revealed a previously unappreciated regulatory circuit. Our findings may advance our understanding of the complex molecular mechanisms underlying the development and maintenance of MLL-associated leukemia, and may also provide new strategies to treat MLL-associated leukemia, a disease that is presently treatment resistant, and likely also other subtypes of AML (as miR-150 is down-regulated in all subtypes of AML), or even other types of cancer that also utilize at least part of the signaling circuit we have described herein. Disclosures: No relevant conflicts of interest to declare.

Abstract 525 Acute leukemia represents one of the most deadly cancers in the United States. Clinical treatments in leukemia have progressed significantly through the use of therapies targeted specifically to chromosomal translocations. The success of these therapies has provided a model for future treatment in various cancers. However, there are various subtypes of leukemia where five-year survival and relapse rates have poor clinical outcome, indicating that new therapies are needed. A particular leukemia subtype, namely mixed lineage leukemia (MLL)-rearranged leukemia that is a result of chromosomal rearrangements leading to fusions between MLL and partner genes, is associated with a dismal outcome. Therapeutic targeting of MLL rearrangements has proven challenging as there have been dozens of described rearrangements. An emerging hallmark of cancer is the deregulation of non-coding RNAs called microRNAs (miRs). We hypothesized that MLL-associated leukemias have aberrant microRNA expression. We performed microRNA microarray analyses on leukemia patient samples and showed that microRNA-9 (miR-9) is highly upregulated in MLL-associated leukemias. We discovered that expression of miR-9 enhances cell transformation in vitro and tumorigenesis in vivo, and the opposite is true when miR-9 function is inhibited by anti-miR sponges. Interestingly, we observed that mice transplanted with both MLL-AF9 and miR-9 had a significantly higher amount of blast-like cells and immature cells in the bone marrow and fewer in the peripheral blood compared to mice transplanted with MLL-AF9 alone. Furthermore, inhibition of miR-9 function causes a significant defect in stem cell self-renewal and myeloid differentiation. Thus, this suggests that miR-9 has a critical role in stem cell potentiation and myeloid promotion in MLL-associated leukemias. To investigate target genes of miR-9, we correlated microarray expression of miR-9 and those of its putative target genes predicted with multiple prediction algorithms. We identified six potential target genes that exhibit a significantly inverse correlation of expression with miR-9 and are inhibited in MLL-associated leukemia. These identified genes are Transforming growth factor beta-induced protein (TGFBI), E-cadherin (CDH1), Nuclear factor-kappa-B p105 subunit (NFkB1), Ras homolog gene family member H (RHOH), Ring1 and YY1-binding protein (RYBP) and Foxhead Box O3 (FOXO3), We demonstrated that all these targets genes are inhibited upon expression of miR-9, in vitro and in vivo and blocking miR-9 function rescues their expression. As these target genes affect multiple cell processes such as proliferation, stem cell differentiation, chromatin remodeling, and cellular migration, it appears that miR-9 is a potential master regulator during MLL-associated tumorigenesis. Disclosures: No relevant conflicts of interest to declare.

Abstract 2453 Acute Leukemia represents one of the most deadly cancers in the United States. Clinical treatments in leukemia have progressed significantly through the use of therapies targeted specifically to chromosomal translocations. The success of these therapies has provided a model for future treatment in various cancers. However, there are various subtypes of leukemia where five-year survival and relapse rates have poor clinical outcome, indicating that new therapies are needed. A particular leukemia subtype, namely mixed lineage leukemia (MLL)-rearranged leukemia that is a result of chromosomal rearrangements leading to fusions between MLL and partner genes, is associated with a dismal outcome. Therapeutic targeting of MLL rearrangements has proven challenging as there have been dozens of described rearrangements. We have performed both messenger RNA (mRNA) and microRNA (a class of small non-coding RNA) microarray analyses on over 100 leukemia patient samples and have identified that microRNA-9 (miR-9) is highly upregulated in MLL-associated leukemias. Through correlating expression of miR-9 and those of its predicted target genes, we identified TGFbeta-induced protein (TGFBI) as a potential target of miR-9 that exhibited a significantly (P Disclosures: No relevant conflicts of interest to declare.

Abstract 3462 Acute myeloid leukemia (AML) bearing MLL (mixed lineage leukemia) translocations are associated with poor survival, and only fewer than 50% of the patients survive longer than 5 years. Thus, an improved strategy leading to a higher cure rate is urgently needed to treat MLL-associated AML. MicroRNAs (miRNAs), a class of small non-coding RNAs, have been postulated to be important gene expression regulators in all biology including human leukemia. Through large-scale, genome-wide miRNA expression profiling assays, we determined that miR-495 is significantly down-regulated in the majority of human AML samples, particularly, in those with MLL rearrangements. More interestingly, through correlating the expression signature of miR-495 with clinical outcome of AML patients, we revealed that a low expression level of miR-495 is a predictor of poor prognosis in most AML patients. Our further qPCR assays confirmed that the expression of miR-495 is even more significantly downregulated in MLL-rearranged AML primary patient samples and cell lines. Through in vitro colony-forming/replating assays and in vivo bone marrow transplantation studies, we found that forced expression of miR-495 significantly inhibits the capacity of the MLL-AF9 fusion gene to support colony formation in mouse bone marrow progenitor cells in vitro and to induce leukemia in vivo. In leukemia cell lines, overexpression of miR-495 greatly inhibits the viability of the cells, while increasing apoptosis. Furthermore, by using 3 algorithms for miR-495 3'UTR binding sites, we identified several well-known MLL leukemia-related genes, e. g. BMI1, MEF2C, BID and MEIS1, as potential targets of miR-495. Results of qPCR revealed that forced expression of miR-495 significantly inhibits the expression levels of these genes in leukemia cell lines, mouse bone marrow progenitor cells, as well as mouse peripheral blood cells with MLL fusion genes. Therefore we hypothesize that miR-495 may function as a tumor suppressor in AML with MLL rearrangements by targeting essential tumor-related genes. Further studies will focus on: 1) effects of miR-495 on the functions of target genes studied in vitro and in vivo; 2) the epigenetic mechanisms and the signaling pathways involved in regulating the expression level of miR-495 in human leukemia. Disclosures: No relevant conflicts of interest to declare.

Abstract 236 Altered expression of microRNAs (miRNAs, a class of small regulatory RNAs) is associated with various types of cancers, including acute myeloid leukemia (AML). We showed previously that increased expression of miR-181a with or without miR-181b was associated with a favorable prognosis for patients with cytogenetically normal AML (CN-AML). However, the prognostic value of miR-181 expression in cytogenetically abnormal AML (CA-AML) remains elusive, even though CA-AML represents the majority of human AML. To investigate the association of expression signatures of miR-181 family members and of their potential target genes with outcome in patients with primary CA-AML, we employed two independent sets of 86 CA-AML patients to investigate the association of expression signatures of miR-181 family members with outcome. We also used four independent sets of 454 CA-AML patients to identify and validate a prognostic signature of miR-181 targets. In addition, we investigated the biological functions of miR-181a/b and target(s) in leukemia cell lines and in a leukemia mouse model. As with CN-AML, we found that both miR-181a and miR-181b expression signatures are significantly (P Disclosures: No relevant conflicts of interest to declare.

Chromosome translocations are among the most common genetic abnormalities in human leukemia. The mixed lineage leukemia (MLL) gene was identified as a common target of chromosomal translocations associated with human acute leukemias; it is located on human chromosome 11 band q23 and on mouse chromosome 9. More than 50 different loci are rearranged in11q23 leukemias involving MLL, resulting in either acute myeloid leukemia (AML) or acute lymphoblastic leukemia (ALL). In general, MLL rearrangements are associated with a poor prognosis. MLL-ELL and MLL-ENL resulting from t(11;19)(q23;p13.1) and t(11;19)(q23;p13.3) respectively are two common examples of these rearrangements. These two fusions are frequently involved in human AML, while MLL-ENL is also involved in human ALL. There is a common observation that important biological properties are often conserved across species. Cross-species sequence comparison has been widely used to infer gene function, but it is becoming apparent that sequence similarity is not always proportional to functional similarity. To determine the function of a gene precisely, therefore, we need to investigate not only its sequence characteristics but also its expression characteristics. Model organisms have contributed substantially to our understanding of the etiology of human disease and the development of new treatment methodologies. However, although genetically engineered mouse leukemia models have been established for many years, there are few systematic studies to identify and study the genes that exhibit similar abnormal expression patterns in both human leukemia and mouse leukemia model cells. To perform an interspecies gene expression comparative study in leukemia, we used the serial analysis of gene expression (SAGE) technique to examine gene expression profiles between MLL-ELL or MLL-ENL myeloid leukemia progenitor cells and normal myeloid progenitor cells in both humans and mice. We obtained 484,303 total SAGE tags for the nine samples and a total of 103,899 unique SAGE tags from five human and 60,993 from four mouse samples. We identified 88 genes that appeared to be significantly deregulated (32 up- and 56 down-regulated) in both human and murine MLL-ELL and/or MLL-ENL leukemia. Fifty-seven genes have not been reported previously. A large-scale quantitative real-time PCR (qPCR) assay was performed to validate the candidate genes, and 84% (36/43) of the tested SAGE candidate genes were confirmed. The most up-regulated genes include several HOX genes (e.g., HOX A5, HOXA9 and HOXA10) and a HOX cofactor MEIS1; their overexpression is a hallmark of MLL-rearrangement leukemia. The top down-regulated genes include LTF, LCN2, MMP9, S100A8, S100A9, PADI4, TGFBI and CYBB. Remarkably, up-regulated genes have a much higher percentage of enrichment in Gene Ontology (GO) terms related to gene expression and transcription, whereas down-regulated genes are more enriched in GO terms related to apoptosis, signal transduction and response. Thus, the up-regulation of genes responsible for gene expression and transcription but down-regulation of genes responsible for apoptosis, signal transduction and response, can promote cell proliferation and inhibit apoptosis, and thereby contribute to the development of leukemia. We showed that the CpG islands of several significantly down-regulated genes including LIF, TGFBI and G0S2 are hypermethylated. We also examined the expression of microRNAs from the mir-17–92 cluster, which are overexpressed in human MLL-rearrangement leukemias, and showed that seven individual microRNAs (i.e., miR-17-5p, miR-17-3p, miR-18a, miR-19a, miR-20a, miR-19b and miR-92) within this cluster are also overexpressed in mouse MLL-rearrangement leukemia cells. Nineteen putative targets (i.e., APP, RASSF2, SH3BP5, DBN1, ELK3, FLT1, GNAI1, HIF1A, ITGA6, MN1, POU4F1, RB1, RGL1, RNF167, SASH1, SLC24A3, TNFRSF21, WWP1 and YES1) of these microRNAs were reported and/or confirmed by our qPCR to be down-regulated in MLL-rearrangement leukemias. We further confirmed both APP and RASSF2 as direct targets of miR-17 through luciferase reporter and mutagenesis assay. The identification and validation of gene expression changes in MLL-rearrangement human and murine leukemia provides important insights into the genetic pathways that are important for MLL fusion-induced leukemogenesis.

Acute myeloid leukemia (AML) is the most common type of acute leukemia in adults. It is estimated that 13,410 cases will be diagnosed and 8,990 will die of AML in the United States in 2007 (http://seer.cancer.gov). AML is a genetically diverse hematopoietic malignancy with variable response to treatment. Expression profiling of protein-coding genes using DNA microarray in AML has resulted in inconsistent data from different laboratories. Therefore, further validation of these observations in large cohorts and in independent studies is definitely required before clinical application becomes feasible. Recently, Golub and colleagues described a new, bead-based flow cytometric microRNA (miRNAs, miRs) expression profiling method that could successfully classify tumors. MiRNAs are endogenous ∼22 nucleotide non-coding RNAs, which can function as oncogenes and tumor suppressors. To provide new insights into the complex genetic alterations in leukemogenesis and to identify novel markers for diagnosis and treatment of AML, we performed a genome-wide analysis of miRNA expression profiles using the bead-based method on 54 AML samples with common translocations including t(15;17), t(8;21), inv(16), and 11q23 rearrangement, along with normal controls. In both unsupervised and supervised hierarchical cluster analyses, we observed that t(15;17) samples grouped together as one cluster, as do the 11q23 rearrangement samples. Interestingly, t(8;21) and inv(16), both CBF (core-binding factor) AMLs, grouped together as a unique cluster. Forty-one miRNAs exhibited significantly differential expression between different subtypes of AMLs, and/or between AMLs and normal controls. Notably, expression signature of a minimal number of two, three, and seven miRNAs could be used for class prediction of CBF, t(15;17), and 11q23 rearrangement AMLs, respectively, with an overall diagnostic accuracy of 94–96%. We further showed that overexpression of the two discriminatory miRNAs in CBF AML is associated with epigenetic regulation, rather than DNA copy number amplification. Moreover, several important target genes of these discriminatory miRNAs have also been validated. We are currently exploring the role of these discriminatory miRNAs and their critical target genes in the development of AML using in vitro and in vivo models. This work will enhance our understanding of the biological role of these miRNAs and their targets in leukemogenesis.

In MDS/AML with t(1;3)(p36;q21), overexpression of MEL1 lacking PR domain (MEL1S) driven by the RPN1 promoter at 3q21.3 is reported to be closely associated with the pathogenesis. Previously, we have defined the 1p36.3 breakpoint (BP) in 5 patients and the 3q21.3 BP in 4 patients with BAC/PAC probes. The BPs at 1p36 were clustered within 3 BP regions (BR), spanning an approximately 200-kb area: BR-1, a 70-kb region in the first intron of MEL1 covered by RP1-163G9; BR-2, a 29-kb region in 5′ region of MEL1 covered by RP5-907A6; BR-3, a 68.5-kb region between the BR-1 and BR-2. On the other hand, the BPs at 3q21.3 were clustered within only one BR, a 108.9-kb region belonging to RP11–475N22, centromeric to RPN1, in which GR6 gene and the promoter region of GATA2 gene are included. Among 5 patients, 2 had the 1p36.3 BPs within the first intron of MEL1, suggesting the existence of MEL1 fusion genes. As expected, using 3′- and 5′-RACE, we have identified 10 MEL1 fusion genes including MEL1/HOXA9 among the 4 patients, not only in patients with the 1p36.3 BP within MEL1 but also in patients with the 1p36.3 BP outside MEL1. None of reciprocal fusion transcripts were detected. Using two normal bone marrow samples and HNT-34 cell line as control, no MEL1 fusion transcripts were detected with TA cloning analysis of 50 clones obtained from each sample. Surprisingly, none of these partner genes were located at band 3q21.3, suggesting that different mechanisms from chromosome translocation might be involved in generating these fusion transcripts. This is the first report showing that MEL1 is involved in the (1;3)(p36;q21)-leukemia not only by inappropriate expression but also by generation of fusion transcripts. Taken together, our findings suggest that in addition to overexpression of MEL1S, MEL1 fusion transcripts may play important roles in the pathogenesis of t(1;3)-leukemia.

Human acute leukemias include acute lymphoblastic leukemia (ALL) and acute myeloid leukemia (AML). It is estimated that 5,200 and 13,410 cases will be diagnosed with and 1,420 and 8,990 will die of ALL and AML, respectively, in the United States in 2007. Although remarkable progress has been made in the past decades in the treatment and in the understanding of the biology of acute leukemias, the 5-year overall survival rate of patients with AML is only approximately 22%, which is much lower than that of ALL patients (65%; http://seer.cancer.gov). One of the most exciting recent findings is the discovery of an abundant class of small (∼22 nt), non-(protein-)coding RNAs, called microRNAs (miRNAs, miRs), which can function as oncogenes and tumor suppressors, whose deregulation is clearly associated with the development of cancer. To understand the distinct mechanisms in leukemogenesis between ALL and AML and to identify novel markers for diagnosis and treatment of acute leukemia, we have performed a large-scale miRNA expression profiling assay with a bead-based flow cytometric method and identified 27 differentially expressed miRNAs. Among them, miR-128a and b are significantly overexpressed while let-7b and miR-223 are significantly down-regulated in ALL compared to AML. They are the most discriminatory miRNAs between ALL and AML. Using the expression signatures of any two of the four most significantly discriminatory miRNAs in diagnosis of ALL and AML resulted in an accuracy rate of 97–100%. The differential expression patterns of these four miRNAs were validated further through quantitative real-time PCR on 98 acute leukemia samples covering most of the common cytogenetic subtypes of AML and B- and T-cell ALL, along with 10 normal controls. Furthermore, we found that overexpression of miR-128a and b in ALL was at least partly associated with hypomethylation, rather than amplification of DNA locus copy. Moreover, several important target genes of these four miRNAs have also been validated. We are currently exploring the role of these four miRNAs and their critical target genes in leukemogenesis and in the determination of lineage fate during leukemia development using in vitro and in vivo models. This work will enhance our understanding of the biological role of these miRNAs and their targets in leukemogenesis, and in determining the lineage fate of acute leukemia.

The t(4;11)(q21;q23) translocation is a hallmark of infant acute lymphoblastic leukemia (ALL), which results in the fusion of the MLL gene on chromosome 11 and the AF4 gene on chromosome 4. MLL-AF4 fusion is the most common consequence of chromosomal translocations in infant leukemia and is associated with a poor prognosis. To identify leukemia-related genes, we used the SAGE technique to compare gene expression profiles between two MLL-AF4 patient samples and one normal sample (CD19+ progenitor B cells; 216,464 tags in total). We identified 61 candidate genes that appear to be abnormally expressed in the leukemia samples (29 up- and 32 down-regulated). Remarkably, we found that many candidate genes appear to play important role in the development of B cells. In addition, many candidate genes can bind with and/or regulate other candidates in the candidate gene list. For example, SYK, BTK and BLNK can bind directly and regulate each other. SYK can also bind directly with TNFRSF1B. In addition, EBF may positively regulate BLK, while BLK can bind directly with BTK. All six of these genes are significantly down-regulated in MLL-AF4 leukemia samples. BTK, SYK and BLK are tyrosine kinases. BTK (B-cell progenitor tyrosine kinase) is a key regulator in B-lymphocyte differentiation and activation. BLK (B-lymphocyte-specific tyrosine kinase) is expressed only in B lymphocytes, and controls pre-B cell development. SYK (spleen tyrosine kinase) is widely expressed in hematopoietic cells, which can phosphorylate BLNK (B-cell linker protein). BLNK represents a central linker protein that bridges the B-cell receptor-associated kinases and may regulate B-cell function and development. EBF (early B-cell factor) is a tissue-specific and differentiation stage-specific DNA-binding protein, and mice lacking Ebf are unable develop B lymphoid cells. TNFRSF1B is strongly expressed on stimulated T and B lymphocytes. Moreover, previous studies indicate that BTK, BLK, SYK, BLNK and TNFRSF1B can positively regulate apoptosis, while BTK can also positively regulate differentiation. Thus, their down-regulation may inhibit apoptosis and differentiation, and thereby contribute to leukemogenesis. In contrast, GNA12, a transforming oncogene which can enhance proliferation and transformation and can bind directly with BTK, is significantly up-regulated in MLL-AF4 leukemia cells. Its up-regulation may also be important to leukemogenesis. Taken together, the deregulation of the important candidate genes may contribute to leukemogenesis through inhibiting apoptosis and differentiation while promoting proliferation of hematopoietic cells. We have validated the expression patterns of the candidate genes with real-time quantitative RT-PCR and are studying the functions and pathways of the validated candidate genes using RNAi and retrovirus transduction over-expression methods. In addition, we will also establish knock-in or knock-out mouse models for the most promising functional candidate genes to see the effect on the development of leukemia. Our studies will provide important insights into the complex functional pathways related to MLL rearrangements in the development of acute lymphoblastic leukemia, which may lead to more effective therapy for these leukemias.

Acute lymphoblastic leukemia (ALL) cells are derived from B and T cell precursors and typically carry rearranged immunglobulin (Ig) or T cell receptor (TCR) variable (V) region genes devoid of somatic mutations. The Philadelphia chromsome (Ph) encoding the oncogenic BCR-ABL1 kinase defines a subset of ALL with a particularly unfavorable prognosis. Here we show that oncogenic BCR-ABL1 kinase activity induces aberrant somatic hypermutation in Ph-positive ALL cells. Under physiological conditions, somatic hypermutation is restricted to mature germinal center B cells and depends on expression of the DNA-deaminating enzyme AID. Comparing Ph-positive and Ph-negative ALL cells, AID expression was found in 24 of 28 Ph-positive but only 3 of 80 Ph-negative ALLs. As shown by RT-PCR and Western blot, expression of AID in Ph-positive ALL cells reached similar levels as in germinal center B cells. Forced expression of BCR-ABL1 in Ph-negative ALL cells and usage of the BCR-ABL1-kinase inhibitor STI571 revealed that BCR-ABL1 kinase activity is required and sufficient to induce aberrant expression of AID in Ph-positive ALL. Consistent with aberrant AID expression in Ph-positive ALL, Ig VH region genes were mutated in most Ph-positive but unmutated in Ph-negative cases. Of note, also non-Ig genes including BCL6 and MYC harbored somatic mutations in Ph-positive but not Ph-negative ALL cells. Likewise, Ph-positive T cell lineage ALL cells express AID and carry somatically mutated TCRβ V region genes. As demonstrated by ligation-mediated PCR, AID introduced DNA-single-strand breaks also within the tumor suppressor gene CDKN2B in Ph-positive ALL cells, which was sensitive to both inhibition of BCR-ABL1 kinase activity and silencing of AID expression by RNA interference. These findings identify AID as a BCR-ABL1-induced mutator in Ph-positive ALL cells, which may be relevant with respect to the particularly unfavorable prognosis of this leukemia subset.

Chromosome translocations are among the most common genetic abnormalities in human leukemia. Their abnormally expressed genes identify specific markers for their clinical diagnosis. Important biological properties are often conserved across species. However, although genetically engineered mouse leukemia models are well-established, few systematic studies have validated the genes that exhibit similar abnormal expression patterns in both human and mouse leukemia models. MLL-ELL and MLL-ENL fusion genes resulting from t(11;19)(q23;p13.1) and t(11;19)(q23;p13.3), respectively, are frequently involved in human acute leukemia, and in retrovirus-mediated mouse leukemia models. We used the SAGE technique to compare gene expression profiles between MLL-ELL or MLL-ENL myeloid leukemia progenitor cells and normal myeloid progenitor cells in both human and mouse. We analyzed four patient samples (two with each fusion) and two retrovirally-induced mouse leukemias containing either MLL-ELL or MLL-ENL fusions, and a leukemia cell line with an MLL-ELL fusion. 484,303 SAGE tags were identified from the nine samples, yielding 103,899 unique tags in human and 60,993 in mouse samples. We identified 40 candidate genes that appear to be abnormally expressed in both human and murine MLL-ELL leukemias (2 up- and 38 down-regulated), and 72 in both human and murine MLL-ENL leukemias (23 up and 49 down). 25 candidate genes are down-regulated in both types of leukemias, and many of them can bind with and/or regulate other candidate genes in the candidate list. For example, LCN2 can bind directly with and positively regulate MMP9; MMP9 and TMSB4X may positively regulate FOS; FOS and JUNB can bind directly and positively regulate each other. JUNB may inhibit proliferation and promote apoptosis, and it was reported that inactivation of JunB in LT-HSC leads to MPD while its inactivation in committed myeloid progenitors also predisposes to leukemia evolution. LCN2 may also positively regulate apoptosis. Meanwhile, some important candidate genes are observed only in one type of leukemia. For example, both PXN and ARHGEF1 are down-regulated only in MLL-ELL leukemias. PXN can bind directly with ARHGEF1, and the latter may inhibit proliferation. Similarly, MYB is significantly upregulated only in MLL-ENL leukemias, which was reported to play a role in MLL-ENL-mediated transformation. Taken together, some common pathways may exist in the development of both types of leukemias, whereas each may also have their own pathway. The deregulation of the important candidate genes may contribute to leukemogenesis through inhibiting apoptosis while promoting proliferation of hematopoietic cells. We have validated the expression patterns of the candidate genes, and are studying the functions and pathways of the validated candidate genes. Our studies will provide important insights into the complex functional pathways related to MLL rearrangements in the development of acute myeloid leukemia, which may lead to more effective therapy for these leukemias.

Although more than 50 different loci translocate to the MLL gene at chromosome band 11q23, resulting in either acute myeloid leukemia (AML) or acute lymphoblastic leukemia (ALL), no unifying property is shared by all partner genes. The translocations result in a functional fusion of the N-terminal part of MLL gene and the C-terminal part of each partner gene, presumably leading to changes in the expression of the normal target genes, most of which have not been identified. Although genetically engineered mouse leukemia models have been widely used, few systematic studies have evaluated whether such models are valid equivalents of human leukemia. We used serial analysis of gene expression (SAGE) to obtain genome-wide gene expression profiles in normal myeloid progenitor cells from human CD15+ and mouse Gr-1+ cells. We also analyzed four patient samples (two with each fusion) and two retrovirally-induced mouse leukemias containing either MLL-ELL [t(11;19)(q23;p13.1)] or MLL-ENL [t(11;19)(q23;p13.3)] fusions, and a cell line from a leukemia mouse transduced with an MLL-ELL fusion. MLL-ELL and MLL-ENL fusions are frequently involved in human AML, while MLL-ENL is also seen in human ALL. 484,303 SAGE tags were identified from the nine samples (40,000 to 100,000 tags per sample), yielding 103,899 unique SAGE tags in the human and 60,993 in the mouse samples. Analysis of the SAGE data identified 43 candidate genes that appear to be abnormally expressed in both human and mouse myeloid leukemia progenitor cells with either MLL-ELL or MLL-ENL fusions (9 up-regulated and 34 down-regulated; Table 1). Increasing evidence suggests that endogenous antisense RNAs may play critical roles in gene regulation and cancer. Natural antisense RNAs include cis-encoded antisense RNAs transcribed from the opposite strand of the same genomic locus as the sense target genes, and trans-encoded antisense RNAs such as microRNAs (miRNAs) transcribed from a genomic locus different from the sense target genes. 26 of the 43 candidate genes have antisense partners (with a total of 7 cis-encoded antisense RNAs and 36 trans-encoded miRNAs) and thereby might be regulated by endogenous antisense RNAs. We are currently validating the expression pattern of the 43 candidate genes in at least 30 different human and mouse leukemia and normal control samples with quantitative RT-PCR, and measuring the level of expression of all known miRNAs via microarray in these samples. Our studies on the abnormally expressed genes and their potential antisense partners will provide important insights into the complex functional pathways related to MLL rearrangements in the development of acute leukemia, which may lead to more effective therapy for these leukemias. Table 1. Genes deregulated in both human and mouse leukemiasa Total number Up-regulated genes Down-regulated genes Genes with antisense partner(s) aThe genes have at least 3 fold difference in expression with a significance P < 0.05 between each leukemia sample and the normal control sample. In MLL-ELL fusions 21 1 20 12 In MLL-ENL fusions 33 8 25 21 In both types of fusions 11 0 11 7 Total unique genes 43 9 34 26

The tumor suppressor INK4B(p15) gene is silenced by CpG island hypermethylation in a majority of acute myeloid leukemias (AML). This silencing can be reversed by the treatment with hypomethylating agents, and these agents are currently being tested for therapeutic intervention. So far, it was not investigated whether or not the INK4B is hypermethylated in all cytogenetic subtypes of AML. Our experiments, which compare levels of INK4B methylation in AML with inv(16), t(8:21) and t(15;17) reveal a strikingly low level of methylation in all leukemias with inv(16). This contrasts with significant levels of DNA methylation in a high proportion of the AML from the other two groups. Surprisingly, even though there is a lack of INK4B methylation in samples from patients with inv(16), expression of the gene is very low when compared to that of PBL from healthy individuals or HL60 cells. Subsequent experiments uncovered a novel mechanism to explain the low level of INK4B expression in the inv(16) AMLs. Overexpression of the aberrant chromosome 16-associated gene CBFb-MYH11 in U937 cells results in failure to induce INK4B in response to vitamin D3. Furthermore, CBFb-SMMHC, encoded by CBFb-MYH11 directly represses transcription from an INK4B promoter in a reporter assay. Electromobility shift assays in the AML-derived cell line ME-1 and U937 cells expressing the fusion gene demonstrate that the repression is due to a change in the composition of the complexes recognizing the binding sites for the transcription regulator CBF. In conclusion, we have found that methylation is not the only way to induce silencing of the tumor suppressor p15INK4B in AML. In inv(16)-containing AML, loss of gene expression is accomplished by the direct transcriptional repressor activity of CBFβ-SMMHC.

Chromosome translocations are among the most common genetic abnormalities in human leukemia. The abnormally expressed genes from each translocation can be used to identify specific markers for clinical diagnosis of each translocation. Microarrays have identified genes differentially expressed in different translocations but the results between laboratories are not always compatible. We used SAGE to quantitate gene expression in bone marrow(BM) samples from 22 patients with four types of AML, [de novo AML M2 with t(8;21), AML M3 or M3V with t(15;17), AML M4Eo with inv(16), AML M5 with t(9;11) or secondary t(9;11)].We made SAGE libraries from CD15+ leukemic myeloid progenitor cells, collecting over 106 SAGE tags, of which 209,486 were unique tags; 136,010 were known genes and ESTs, and 73,476 were novel transcripts. SAGE tags for further analysis were selected based on a 5-fold difference between patient’s samples and normal CD15+ BM; they were also statistically significantly different at the 5% level. Using these strict criteria, we identified 2,381 unique tags, of which 2,053 were known genes and ESTs, and 328 were novel transcripts that were either specific for each translocation or were common(55) SAGE tags for all 4 translocations. The major change in all translocations was a decrease in expression in leukemia cells compared with normal cells; the decrease was least in the t(8;21) cells. Changes in expression of these known genes, which fall into different gene ontology functional categories, varied by translocation. Those associated with macromolecular biosynthesis, transport and transcription were most altered in the t(8;21); those related to defense response and apoptosis were altered in the t(15;17); cell proliferation genes were most affected by the t(9;11). From this analysis, we identified the functional molecular signature of each translocation. We designed a custom microarray to validate our SAGE data analysis. Our initial microarray contained 349 probes including 212 known genes, 61 ESTs, 28 novel sequences based on our data and 48 genes reported by others. We have now included 65 additional probes that appeared to be correlated with survival. Using 63 samples with the four translocations [16 inv(16), 4 t(9;11), 20 t(15;17), 4 t(8;21) and 19 other translocations], we are validating which genes provide a robust, reproducible “fingerprint” for each translocation, for all translocations, and which ones provide reliable information related to prognosis and survival. Our results will provide new insights into genes that collaborate with each translocation to lead to a fully leukemic phenotype as well as which genes appear to provide valid prognostic information.

Reciprocal translocation t(1;3)(p36;q21) is found in a subset of MDS/AML characterized by trilineage dysplasia, especially dysmegakaryopoiesis and a poor prognosis. In this leukemia, ectopic expression of MEL1 at 1p36.3 driven by the promoter of RPN1 at 3q21 is thought to be associated with the pathogenesis. Recently, we reported the identification of 2 kinds of the 1p36 breakpoint (BP) region in 3 patients with t(1;3), 1) a 70.7-kb region within the first intron of MEL1 in one patient, and 2) a 29.2-kb region in the 5′ region of MEL1 in the remaining two. However, the precise BP at 3q21 has not yet been determined. Here, we report that we have defined the 3q21 BP in 4 patients, and the 1p36.3 BP in 2 additional patients with FISH using bone marrow cells obtained from 4 patients (4 with MDS-related AML-M4 and one with CMML). To identify the BPs and deletions at 1p36.3 or at 3q21.3, we used 18 non-contig BAC/PAC probes and 7 contig BAC/PAC probes, respectively. FISH study revealed that the 3q21 BPs were clustered within RP11-475N22 probe in the 4 patients, in the 5′ region of GATA2. The 1p36.3 BP was found within the first intron of MEL1 in one patient, but in a 68.5-kb region which included a part of the first intron and a region of 5′ site of MEL1 in the other. Therefore, 3 BP regions at 1p36 were found among the 5 patients. No deletions were detected in the vicinity of the 1p36 BP or 3q21 BP in any of our 5 patients, although Lahortiga et al., detected a large deletion at the both BPs in their patients with MDS (RAEB) and t(1;3) (Oncogene 23:311–316, 2004). In Patient 1 (the 1p36 BP was in the first intron of MEL1) and Patient 3 (the 1p36 BP was in 5′ site of MEL1), the expression patterns of MEL1, ARPM2, and GATA2 which are located near the BPs were analyzed with RT-PCR. ARPM2 was not overexpressed in the two patients, while MEL1 was overexpressed in Patient 1, but not in Patient 3 when the expression level was compared with that of normal bone marrow cells. It is reported that GATA2, one of the zinc finger transcription factors, serves as a regulator of genes controlling proliferative capacity of early hematopoietic cells, and that two zinc finger domains of GATA2 are closely involved in inducing megakaryocytic differentiation. Interestingly, a deleted GATA2 transcript was detected in Patient 3 together with a wild type GATA2 transcript. The short GATA2 transcript contained 1009 nucleotides deletion from exon 3 to exon 6 with a frameshift, resulting in lacking the two zinc finger domains. Thus, there is the possibility that dysmegakaryopoiesis in this patient is associated with the expression of GATA2 without the zinc finger domains.

The 8;21(q22;q22) translocation is found in 12% of all AML cases, including 40–50% of AML M2 subtype and a small portion of M0, M1, and M4 subtypes. The translocation brings together the DNA sequence encoding the N-terminus of the AML1 (RUNX1) with nearly all of ETO (MTG8) to generate the AML1-ETO fusion protein. The most commonly known AML1-ETO fusion protein (full length AML1-ETO) is encoded by AML1 exons 1 to 4 and by ETO exons 2 to 11 and has 752 a.a. However, a recently discovered ETO exon 9a can provide a protein translation stop codon after the last amino acid encoded by exon 8. The predicted AML1-ETO protein using exon 9a (AML1-ETO9a) has 575 a.a. and does not include the NHR3 and NHR4 domains at the C-terminus of full length AML1-ETO. To demonstrate the presence of the AML1-ETO9a transcript, we performed RT-PCR using RNA prepared from t(8;21) patient samples. Exon 9a transcript was detected in all patient samples when PCR generated a 595 bp fragment with ETO exon 6 and exon 9 primers and was detected in seven of the 14 samples when PCR generated a 1381 bp fragment with AML1 exon 4 and ETO exon 9a primers, indicating the presence of AML1-ETO9a mRNA in patient samples. The predicted molecular weight of AML1-ETO9a protein with 575 a.a. is 63.8 kDa and a 64 kDa AML1-ETO fusion protein has been reported in both Kasumi-1 cells and t(8;21) leukemic blasts. We compared the effects of AML1-ETO and AML1-ETO9a on proliferation and differentiation of hematopoietic cells using FDCPmix cell line, which is an IL-3 dependent multipotent hematopoietic stem cell line. Both forms of fusion proteins inhibited the myeloid lineage differentiation when cells were cultured in the presence of GM-CSF and G-CSF. However, these fusion proteins showed opposite effects on cell cycle progression. AML1-ETO suppressed cell cycle with prolonged G0/G1; AML1-ETO9a enhanced cell cycle with shortened G0/G1. Expression of these fusion proteins in primary bone marrow cells showed a similar result. Most importantly, only AML1-ETO9a efficiently induced acute myeloid leukemia in mice. Thus, the observation that a shorter form of AML1-ETO is a dominant factor in t(8;21) related leukemogenesis exhibits a new paradigm in the understanding of t(8;21) tumorigenesis.

Chromosome translocations are among the most common genetic abnormalities in human leukemia. Each translocation may affect a different pair of genes. The abnormally expressed genes that result from the different translocations provide a rich source for identifying specific markers for clinical diagnosis of each translocation. Microarrays have identified genes differentially expressed in different translocations but the results between laboratories are not always compatible. We used SAGE to quantitate gene expression in bone marrow (BM) samples from 22 patients with four types of AML, namely de novo AML M2 with t(8;21), AML M3 or M3V with t(15;17), AML M4Eo with inv(16), AML M5 with t(9;11) or secondary t(9;11).We generated SAGE libraries from CD15+ leukemic myeloid progenitor cells, collecting over 106 SAGE tags, of which 209,486 were unique tags; 136,010 were known genes and ESTs, and 73,476 were novel transcripts. SAGE tags for further analysis were selected based on a 5-fold difference between patients’ samples and normal CD15+ BM; they were also statistically significantly different at the 5 % level. Using these strict criteria, we identified 1,571 unique tags, of which 1,405 were known genes and ESTs, and 166 were novel transcripts that were either specific for each translocation or were common for all four translocations. Changes in expression of these known genes which fall into different gene ontogeny functional categories varied by translocation. For example, those associated with macromolecular biosynthesis, transport and transcription were most altered in the t(8;21); those related to defense response and apoptosis were altered in the t(15;17); cell proliferation genes were most affected by the t(9;11). Cell surface receptor signaling, intracellular signaling and RNA processing were altered in treatment related but not in de novo t(9;11). From this analysis, we identified the functional molecular signature of each translocation. We designed a custom microarray to validate our SAGE data analysis. Our initial pilot microarray experiment with 96 genes that were specific for each translocation or common for all translocations used mononuclear cells from normal and patient BM and translocation cell lines, ME-1, THP-1, Mono Mac-6, Kasumi 1, NB-4; the array data from BM matched the SAGE data for 48-75 % of genes and the majority of cell lines, except ME-1, matched at least 70 % with the SAGE results for the appropriate translocation. We have now designed a full-scale microarray that contains over 400 probes including 250 known genes, 61 ESTs, 45 novel sequences and 48 genes reported by others. We will test at least 100 patients’ samples with the four translocations to validate which genes provide a robust, reproducible “fingerprint” for each translocation and for all translocations. We will correlate our microarray data with age, sex, race, response to treatment, survival and other mutations (FLT3, MLL ITD, etc) to identify any transcripts that might reliably define these categories. Our results will provide new insights into genes that collaborate with each translocation to lead to a fully leukemic phenotype.

The t(8;21)(q22;q22) that results in an AML1-ETO fusion gene is frequently detected in de novo and in therapy-related acute myeloid leukemia (AML) patients. Most t-AML patients with t(8;21) have received treatment with topoisomerase (topo) II inhibitors for primary tumors. In our previous study in 31 de novo leukemia patients with t(8;21) genomic breakpoints clustered in each three breakpoint cluster regions (BCRs) in intron 5 in AML1 and in intron 1b in ETO which tended to colocalize with DNA topo II cleavage and DNase I hypersensitive sites, implicating these chromatin structural elements in the mechanisms leading to the t(8;21). In this study, we cloned genomic breakpoints in AML1 and ETO in both derivative chromosomes in six t-AML leukemia patients with t(8;21), all of whom were treated with topo II inhibitors for previous tumors. Genomic breakpoints in AML1 and ETO in t-AML patients cluster in the same BCRs previously identified in de novo patients with t(8;21). Our results are unexpected because in MLL translocations involving 11q23, the location of breakpoints in MLL in de novo and in t-AML patients is different. There were deletions and duplications at the breakpoints in both AML1 and ETO, and microhomologies or non-templated nucleotides occurred at most of the genomic fusion junctions. No specific recombination motifs were identified at or near the breakpoint junctions. Both deletions and duplications were larger in de novo leukemia patients than in t-AML patients. In each 10 de novo patients, the deletions ranged from 5 to 556 bp (251 bp in average) in AML1 and from 6 to 225 bp (88 bp) in ETO, whereas in t-AML patients the deletions in AML1 ranged from 1-111bp (39 bp) in three and from 2-125 bp (32 bp) in ETO in five patients. Duplications in AML1 in seven de novo patients ranged from 1-141 bp (76 bp) and in ETO in eight patients from 51 to 355 bp (185 bp); in contrast, duplications in AML1 were 1, 2 and 1085 bp in three t-AML patients, and only one t-AML patient had a duplication (1 bp) in ETO. Similarly, only very small (2-7 bp) deletions or duplications were detected in MLL and CBP in t-AML patients with t(11;16). Thus, there are some unexplained differences in non-homologous end joining repair in de novo and therapy-related leukemia. We mapped a scaffold attachment region (SAR) in intron 4 of AML1, and two SARs in intron 1b of ETO that are near the BCRs. Moreover, we analyzed the free energy (delta G) that is needed to unwind double strand DNA of the BCR-bearing introns in AML1 and ETO and found that the topo II cleavage sites we identified in both genes have the lowest delta G value suggesting that topo II cleaves DNA at the point of the lowest free energy. Our study suggests that illegitimate recombination between AML1 and ETO in the t(8;21) in both de novo and therapy-related leukemia patients is similar, and that topo II cleavage sites with the lowest free energy provide vulnerable sites for breakage. Non-homologous end joining repair is a likely mechanism involved in the formation of the t(8;21).

The Philadelphia chromosome (Ph1) of chronic myelogenous leukemia (CML) contains sequences from chromosome 9, including the ABL protooncogene, that have been translocated to the breakpoint cluster region (bcr) of chromosome 22, giving rise to a bcr-ABL fusion gene, whose product has been implicated in the genesis of CML. Although chromosome 22 translocation breakpoints in CML virtually always occur within the 5.8- kilobase (kb) bcr, chromosome 9 breakpoints have been identified within the known limits of ABL in only a few instances. For a better understanding of the variability of the breakpoints on chromosome 9, we studied the CML cell line BV173. Using pulsed-field gel electrophoresis (PFGE), large-scale maps of the t(9;22) junctions were constructed. The chromosome 9 breakpoint was shown to have occurred within an ABL intron, 160 kb upstream of the v-abl homologous sequences, but still 35 kb downstream of the 5′-most ABL exon. bcr-ABL and ABL-bcr fusion genes were demonstrated on the Ph1 and the 9q+ chromosomes, respectively; both of these genes are expressed. These results suggest that the 9;22 translocation breakpoints in CML consistently occur within the limits of the large ABL gene. RNA splicing, sometimes of very large regions, appears to compensate for the variability in breakpoint location. These studies show that PFGE is a powerful new tool for the analysis of chromosomal translocations in human malignancies.

The aggressive clinical course and the distinctive histologic, cytochemical, and cytogenetic features of an adult non-Sezary T-cell lymphoma with suppressor activity have been investigated. Morphological and ultrastructural analysis of neoplastic cells from peripheral blood and involved lymph nodes revealed cells with convoluted nuclei, prominent cytoplasmic azurophilic granules, well developed Golgi apparatus, short strands of endoplasmic reticulum, and moderate numbers of ribosomes and mitochondria. Cytochemical reactions showed acid phosphatase (ACP) positivity in virtually all of the neoplastic cells; and a substantial percentage of the cells, the tartrate-resistant acid phosphatase (T-ACP) isoenzyme was observed. Granular naphthyl acetate esterase (A-EST) reactivity was not present. The histological and cytochemical features of these neoplastic suppressor cells were compared with those recently described for the suppressor T-cell fraction isolated from normal peripheral blood T-cell by Fc gamma- rosette formation. The aneuploid clone had 47 chromosomes with multiple complex abnormalities, including a 14q + chromosome formed by the tandem translocation of two no. 14 chromosomes and translocations involving the long arms of no. 2 and no. 9 at band 9q34. These latter changes are particularly common in T-cell disorders. The extensive analysis of this histologic, cytochemical, and cytogenetic features of this adult T-cell suppressor lymphoma should help to clarify the criteria for distinguishing among the subsets of T-cell neoplasms with definable immunologic function.

Three patients with acute leukemia, disseminated intravascular coagulation, and a specific acquired chromosome abnormality [t (15;17)] were found by transmission electron microscopy to have the typical distribution of granules seen in promyelocytes. However, the average granule sizes were 120, 170, and 180 nm, respectively, for the three patients, significantly less than the 250-nm resolution of light microscopy. We regard the leukemia in these three patients as comprising a distinct clinical, ultrastructural, and cytogenetic entity that we have chosen to call “microgranular” acute promyelocytic leukemia.

Cytogenetic studies were performed on 26 patients who developed acute nonlymphocytic leukemia (ANLL) or a dysmyelopoietic syndrome after treatment of a primary malignancy. Fifteen patients had radiotherapy and chemotherapy, seven had only chemotherapy, and four had only radiotherapy. The median times from diagnosis of the initial disease to the development of bone marrow dysfunction for these treatment groups were 50, 46, and 49 mo, respectively. Twenty-five patients had an abnormal karyotype in myeloid cells. Loss of part or all of no. 5 and/or no. 7 was noted in 23 of 25 patients with aneuploidy. Loss of no. 5 was noted only in patients who previously had malignant lymphoma, whereas loss of no. 7 was seen in these patients as well as in those who had other malignancies. Abnormalities of both nos. 5 and 7 occurred in 53% of the patients treated with combined therapy and in only 27% of patients treated with either modality alone. Although these changes are distinctly different from those noted in lymphomas, they are similar to those seen in 25% of aneuploid patients with ANLL de novo.

Seventeen patients with therapy-related myelodysplastic syndrome (t- MDS) or therapy-related acute nonlymphocytic leukemia (t-ANLL) were treated with single-agent high-dose cytarabine (HDAC; 1 to 3 g/m2 every 12 hours for 12 doses). The initial neoplasm was still present in eight patients when t-MDS/t-ANLL developed. Fifteen of the 16 patients with chromosomal abnormalities in bone marrow cells had loss or rearrangement of chromosomes 5 and/or 7. One patient had a t(15;17), and one had inadequate material for cytogenetic analysis. Twelve patients had normal metaphase cells (3% to 71%). Indications for HDAC therapy were progressive pancytopenia in 13 patients or rising blast count in four. Five patients died of marrow hypoplasia following therapy. Four others had refractory t-ANLL and died within the subsequent 5 months. Only one of ten patients with a poor performance status (PS greater than or equal to 2 using the ECOG scale) achieved a complete remission, but all seven patients with a good performance status (PS less than or equal to 1) had a complete remission. Hematologic remissions were achieved in 8 patients (47%) after one (6 patients) or two (2 patients) induction courses and were confirmed by recovery of a 100% normal marrow karyotype in six of the seven patients who were retested. Patients in remission received one to four consolidation courses with HDAC alternating with cytarabine/doxorubicin, but seven relapsed within 8 months (median remission duration, 5 months). In every case, the original chromosomal abnormality reappeared at relapse. HDAC has a high response rate for good-performance patients with t-MDS/t-ANLL, but complete remissions are short even when confirmed cytogenetically and consolidated intensively.