Comparative drug screening in NUT midline carcinoma.

BACKGROUND: The NUT midline carcinoma (NMC) is a rare but fatal cancer for which systematic testing of therapy options has never been performed. METHODS: On the basis of disease biology, we compared the efficacy of the CDK9 inhibitor flavopiridol (FP) with a panel of anticancer agents in NMC cell lines and mouse xenografts. RESULTS: In vitro anthracyclines, topoisomerase inhibitors, and microtubule poisons were among the most cytotoxic drug classes for NMC cells, while efficacy of the bromodomain inhibitor JQ1 varied considerably between lines carrying different BRD4 (bromodomain-containing protein 4)-NUT (nuclear protein in testis) translocations. Efficacy of FP was comparable to vincristine and doxorubicin, drugs that have been previously used in NMC patients. All three compounds showed significantly better activity than etoposide and vorinostat, agents that have also been used in NMC patients. Statins and antimetabolites demonstrated intermediate single-agent efficacy. In vivo, vincristine significantly inhibited tumour growth in two different NMC xenografts. Flavopiridol in vivo was significantly effective in one of the two NMC xenograft lines, demonstrating the biological heterogeneity of this disease. CONCLUSIONS: These results demonstrate that FP may be of benefit to a subset of patients with NMC, and warrant a continued emphasis on microtubule inhibitors, anthracyclines, and topoisomerase inhibitors as effective drug classes in this disease.

Idiotypes on major histocompatibility complex-restricted virus-immune cytotoxic T lymphocytes.

Specificity-associated determinants could be demonstrated on major histocompatibility complex (MHC)-restricted virus-immune cytotoxic T lymphocytes (CTL) using antisera raised in syngeneic mice. This result indicates that the number of clonotypes sufficiently small to allow the detection of such idiotypic determinants. The functional properties of three anti-idiotypic antisera were tested in blocking studies and by idiotypic antisera plus complement. Whereas the former test did not reveal any results obtained from binding studies, i.e., all three anti-idiotypic antisera specifically reacted with CTL of the type used for immunization and had no effect on syngeneic influenza-immune CTL, syngeneic alloreactive CTL, or NDV-immune CTL of an H-2-disparate strain. These results are discussed in terms of current models for MHC-restricted CTL responses.

A single genetic element in H-2K affects mouse T-cell antiviral function in poxvirus infection.

Cell transfer experiments using mice with recombinant H-2 haplotypes were used to map the H-2 regions which must be shared by ectromelia-immune T-cell donors and virus-infected recipients for transfer of virus clearance mechanisms in the spleen. K- or D-region genes were necessary and sufficient; I-region genes were not involved. The remainder of the mouse genome could be varied widely without impairing the efficacy of T-cell antiviral function, provided either a K or a D region was shared in the donor-receipient combination. A mutation in a single genetic element of the K region of the H-2 complex abolished the antiviral effect of immune T-cell transfer in a donor-recipient combination which shared the K end.

Most influenza A virus-specific memory cytotoxic T lymphocytes react with antigenic epitopes associated with internal virus determinants.

This paper shows that most murine (C57BL/6) influenza A virus-specific memory cytotoxic T lymphocyte (CTL) clones tested in limiting dilution did not react with the influenza A virus surface glycoproteins, hemagglutinin (HA) and neuraminidase (NA). This lysis of syngeneic target cells infected with the influenza A virus strains, Aichi (H3N2), PR8 (H1N1), or recombinant strain X31 (H3N2) indicates that most antigenic epitopes recognized are associated with internal virus determinants. X31 and PR8 share the internal, and X31 and Aichi the external, viral determinants. Extensive CTL cross-reactivity was observed in experiments with target cells infected with virus carrying internal determinants homologous with the priming virus. In contrast, when the internal viral determinants differed between the priming virus and the virus used to infect the target cells, and although HA and NA were homologous, we found almost complete CTL-specificity for the priming virus. Thus, the predominant reactivity of influenza A virus-specific CTL differs from that of anti-influenza A antibodies, which are primarily directed towards epitopes on the virus surface glycoproteins. This finding may be relevant for the role of influenza A virus-specific CTL in recurrent infections with different influenza A viruses.

Cytotoxic T-cell responses show more restricted specificity for self than for non-self H-2D-coded antigens.

The specificity of recognition of H-2 antigens by various subsets of Tc cells was investigated with respect to the two separate molecules known to be coded in the H-2D(d) region (a) D which carries the private specificity H-2.4 and (b) D'which carries the public specificity H-2.28. BALB/c.H-2(db) mutant mice express D but not D' on their cell surfaces, whereas wild-type BALB/c mice express both D and D'. H-2 restricted Tc cells specific for viral-plus- H-2D(d)-coded antigens on infected self cells, or minor H-plus-H-2D(d)-coded antigens on H-2-compatible cells apparently recognize D, but do not detectably recognize D. In contrast, BALB/c-H-2(db) anti-BALB/c Tc cell responses do recognize D' (the only known antigen which is not shared by mutant and wild-type); furthermore, D' is also detectably recognized by a significant proportion of the Tc cells that respond in MLR to H-2D(d)-coded antigens. In these latter responses, D' was recognized separately from D, i.e., the response was not "H-2 restricted". These results indicate that H-2 restricted Tc cell responses to modified-self cells are more specific for self H-2D(d)-coded antigens then are allogeneic Tc cell responses directed at the same antigens, in that haplotype-unique (private) specificity recognition (of the D molecule) exclusively occurs only in the former, not the latter case. The implications of this specificity of H-2 restricted responses for possible processes of somatic selection of anti-self recognition structures on progenitor Tc cells are briefly discussed.

Specific adsorption of H-2-restricted cytotoxi T cells to macrophage monolayers.

These experiments tested whether Tc cells specific for foreign antigen (X) plus self H-2 adsorbed to macrophage monolayers displaying (a) X with allogeneic H-2; (B) self H-2 alone; (c) X plus self H-2. Specific adsorption occurred only in case (c), a result compatible with altered self and requiring further operational assumptions in dual recognition models.

Cytotoxic T-cell response to Ectromelia virus-infected cells. Different H-2 requirements for triggering precursor T-cell induction or lysis by effector T cells defined by the BALB/c-H-2db mutation.

The T(c)-cell response to ectromelia virus infection was studied in BALB/c-H-2(db) mice which carry a loss mutation in the H-2D region that results in the absence from cell surfaces of a molecule (D') bearing certain public H-2 specificities. When infected, these mice showed a poor response of T(c) cells that recognize H-2D(d) plus virus-specific determinants on infected macrophage targets, but gave a normal response to H-2K d plus virus-specific antigens. However, their own infected macrophages do display wild-type antigenic patterns involving virus and H-2D(d) since they were killed as efficiently as wild-type (BALB/c,H- 2(d))-infected cells by T(c) cells specific only for H-2D(d) plus viral antigens. When tested in vitro, infected BALB/c-H-2(db) cells stimulated a poor T(c)-cell response to H-2D plus virus-specific antigens, but stimulated a normal response (in comparison with infected BALB/c macrophages) to H-2K(d) plus viral antigens. Uninfected BALB/c-H-2(db) cells stimulated a normal T(c)-cell response to minor H antigens or trinitrophenyl in association with H-2D(d), thus suggesting that the defective response to infection may reside in a failure of the relevant H-2D(d) antigens of mutant cells to physically associate with viral antigens. Close association of viral and H-2D-coded molecules was also suggested by ability of specific anti-H-2K or -H-2D to partially block T(c)-cell-mediated lysis of infected targets. These results were interpreted to mean that H-2Dd-dependent, virus- immune T(c) cells recognized an antigenic pattern consisting of virus- specific and H-2D(d) determinants with the latter borne on an H-2D molecule carrying serologically-defined H-2D(d) private specificities. A second H-2D(d)-coded molecule (D') was not required for recognition and lysis by activated T(c) cells, but was apparently necessary for efficient stimulation of precursor T(c) cells, perhaps by promoting appropriate physical association of viral and H-2D(d) molecules.

H-2 restriction of cell-mediated immunity to an intracellular bacterium: effector T cells are specific for Listeria antigen in association with H-21 region-coded self-markers.

The protective activity of anti-Listeria-immune T cells assayed in an adoptive transfer system in H-2 restricted. As shown in the present studies, the demonstration of the restriction is directly dependent on the dose and the relative protective activity of spleen cells. In addition, some H-2-unrestricted protection is conferred predominantly by other than immunoglobulin-negative spleen cells. Thus, the activity of Listeria-immune T cells appears to be 'absolutely' restricted and is in this respect comparable to in vivo T-cell-mediated anti-viral protection. The predominant genetic region of H-2 coding for the structures which are mainly involved in this restriction in T-cell immunity to this prototype intracellular bacterium is the I region. The specificity of Listeria-immune T cells is determined by the H-2 haplotype of the donor. Thus, F1 hybrids seem to possess at least two separable sets of T cells, each specific for one parental haplotype. As is true in the virus model, the results cannot distinguish between an altered-self or a dual recognition model of T-cell recognition to explain H-2 restriction. They are, however, compatible with the idea and I-coded cell surface structures may serve as receptors for cell-specific differentiation signals, which trigger direct or lymphokin-mediated activation of macrophages to manifest increased bactericidal capacity. The interesting parallels in self-marker recognition of T cells in the virus and intracellular bacterium systems, respectively, appear to be reasonably explained by the different types of signals transmitted by T cells to various target cells via the distinctly different self-markers employed (i.e., K or D vs I).

Glucocorticoid resistance in T-lineage acute lymphoblastic leukaemia is associated with a proliferative metabolism.

Glucocorticoids (GCs) are among the most important drugs for acute lymphoblastic leukaemia (ALL), yet despite their clinical importance, the exact mechanisms involved in GC cytotoxicity and the development of resistance remain uncertain. We examined the baseline profile of a panel of T-ALL cell lines to determine factors that contribute to GC resistance without prior drug selection. Transcriptional profiling indicated GC resistance in T-ALL is associated with a proliferative phenotype involving upregulation of glycolysis, oxidative phosphorylation, cholesterol biosynthesis and glutamate metabolism, increased growth rates and activation of PI3K/AKT/mTOR and MYC signalling pathways. Importantly, the presence of these transcriptional signatures in primary ALL specimens significantly predicted patient outcome. We conclude that in lymphocytes the activation of bioenergetic pathways required for proliferation may suppress the apoptotic potential and offset the metabolic crisis initiated by GC signalling. It is likely that the link between GC resistance and proliferation in T-ALL has not been fully appreciated to date because such effects would be masked in the context of current multiagent therapies. The data also provide the first evidence that altered expression of wild-type MLL may contribute to GC-resistant phenotypes. Our findings warrant the continued development of selective metabolic inhibitors for the treatment of ALL.

The nuclear oncoprotein TLX1/HOX11 associates with pericentromeric satellite 2 DNA in leukemic T-cells.

TLX1/HOX11, a DNA-binding homeodomain protein, was originally identified by virtue of its aberrant expression in T-cell leukemia and subsequently found to be crucial for normal spleen development. The precise mechanism of TLX1 function remains poorly understood, although it is known that it can act as both a transcriptional activator and repressor and can downregulate the Aldh1a1 gene in embryonic mouse spleen. Using a whole-genome PCR approach, we show here that TLX1 protein directly interacts with pericentromeric human satellite 2 DNA sequences. Such DNA is known to localize to heterochromatin, which among other roles has been implicated in gene silencing. The interaction was confirmed in vitro and in vivo by gel retardation and chromatin immunoprecipitation assays involving satellite 2 DNA, which contained sequences resembling TLX1 binding sites. Using immunofluorescence microscopy, TLX1 demonstrated a punctate pattern of staining in the nuclei of leukemic T-cells (ALL-SIL). Double labelling indicated that TLX1 colocalized with the centromeric protein CENP-B, demonstrating that the TLX1 foci corresponded to clusters of centromeric DNA. The novel interaction of TLX1 with constitutive heterochromatin adds an additional level of complexity to the intracellular functions of this transcriptional regulator and may have relevance to its roles in transcriptional repression and T-cell immortalization.

Altered glucose metabolism in childhood pre-B acute lymphoblastic leukaemia.

The cells of solid tumours are known to have an altered metabolism, with high rates of glucose uptake and glycolysis, which results in the excessive production of lactate. To date there has been no definitive research documenting metabolic changes in acute lymphoblastic leukaemia (ALL) cells. In order to investigate whether ALL cells have an altered metabolism, we initially compared the transcriptional profiles of 22 specimens from paediatric patients diagnosed with ALL to five CD34+ specimens isolated from bone marrow, which was verified in an independent cohort of 101 specimens. Profiling revealed the upregulation of genes facilitating glycolysis in the ALL specimens compared to the CD34+ specimens, while those involved in the tricarboxylic acid cycle were downregulated. Functional studies supported the microarray findings threefold: (1) higher expression of the glucose transport protein glucose transporter 1 in ALL compared to CD34+ specimens, (2) the excessive production of lactate in ALL cell lines and (3) sensitivity of ALL cell lines to the glycolysis inhibitor 2-deoxy-D-glucose. While metabolic alterations have been well documented in solid tumours, this is the first study to provide direct evidence for the existence of metabolic changes in the leukaemic cells of ALL patients. The finding offers new options for targeted therapy for ALL patients.

Systematic chemical and molecular profiling of MLL-rearranged infant acute lymphoblastic leukemia reveals efficacy of romidepsin.

To address the poor prognosis of mixed lineage leukemia (MLL)-rearranged infant acute lymphoblastic leukemia (iALL), we generated a panel of cell lines from primary patient samples and investigated cytotoxic responses to contemporary and novel Food and Drug Administration-approved chemotherapeutics. To characterize representation of primary disease within cell lines, molecular features were compared using RNA-sequencing and cytogenetics. High-throughput screening revealed variable efficacy of currently used drugs, however identified consistent efficacy of three novel drug classes: proteasome inhibitors, histone deacetylase inhibitors and cyclin-dependent kinase inhibitors. Gene expression of drug targets was highly reproducible comparing iALL cell lines to matched primary specimens. Histone deacetylase inhibitors, including romidepsin (ROM), enhanced the activity of a key component of iALL therapy, cytarabine (ARAC) in vitro and combined administration of ROM and ARAC to xenografted mice further reduced leukemia burden. Molecular studies showed that ROM reduces expression of cytidine deaminase, an enzyme involved in ARAC deactivation, and enhances the DNA damage-response to ARAC. In conclusion, we present a valuable resource for drug discovery, including the first systematic analysis of transcriptome reproducibility in vitro, and have identified ROM as a promising therapeutic for MLL-rearranged iALL.

Resistance to 1-beta-D-arabinofuranosylcytosine after high-dose treatment childhood lymphoblastic leukemia: isolation of a drug resistant and a sensitive cell line.

A patient with acute lymphoblastic leukemia refractory to conventional combination chemotherapy schedules was treated with high-dose 1-beta-D-arabinofuranosylcytosine (ara-C) (8 doses of 3 g/m2). The patient achieved complete remission but relapsed 6 weeks later. Two cell lines, PER-145 and PER-163, were established from bone marrow samples obtained before and after treatment, respectively. Both cells represent common acute lymphoblastic leukemia cells (CALLA+, HLA-DR+, sIg-, cIg-). Exposure of the two cell lines to ara-C in vitro revealed that the primary line PER-145 is susceptible to ara-C, while cell line PER-163 is more than 1000-fold more resistant (based on 50% inhibitory doses for growth in culture). Moreover, it was observed that ara-C concentrations from 1 to 33 micrograms/ml resulted in the stimulation of this cell line. Five weeks after his relapse the patient was given another high-dose ara-C course during which the blasts increased by a factor of 10, thus showing a response in vivo similar to that of the PER-163 cells in vitro. This pair of human acute lymphoblastic leukemia cells provides a unique opportunity to investigate the mechanisms underlying the acquired resistance and proliferative response to ara-C.

Development of resistance to 1-beta-D-arabinofuranosylcytosine after high-dose treatment in childhood lymphoblastic leukemia: analysis of resistance mechanism in established cell lines.

Cell lines PER-163 and PER-164 are derived from a patient with acute lymphoblastic leukemia who developed resistance to 1-beta-D-arabinofuranosylcytosine (ara-C) after high-dose (HD) therapy. Both lines are highly resistant to ara-C and have maintained stable resistance for more than 18 mo. The resistance in PER-164 cells is the result of a selection process in vivo only, while PER-163 cells have in addition been exposed to ara-C in culture. Comparison with cell line PER-145, which is sensitive to ara-C and was established from the same patient before HDara-C therapy, revealed no differences with respect to surface markers, morphology, cytochemical stains, or requirements for growth in vitro. The leukemic origin of the three cell lines is indicated by the close similarities of all three cell lines to the patient's fresh cells. The analysis of the two resistant cell lines shows that resistance to ara-C is not due to lower ara-C transport capacity nor to cytokinetic reasons, since the percentage of cells in S-phase is similar in all three cell lines. In addition, the resistant cell lines do not show any increased cytidine deaminase activity. PER-164 cells show a markedly reduced deoxycytidine kinase activity, 4.8 nmol/h/mg of protein, compared to PER-145 cells with an enzyme activity of 21.48 nmol/h/mg of protein. In PER-163 cells, no deoxycytidine kinase activity could be detected. Furthermore, the two resistant cell lines show significantly different dCTP levels. The sensitive PER-145 cells generated 97.9 pmol of 1-beta-D-arabinofuranosylcytosine triphosphate (ara-CTP)/10(7) cells during a 45-min incubation period in the presence of 10(-6) M ara-C. This contrasts with 0.16 and 12 pmol of ara-CTP/10(7) cells for PER-163 and PER-164 cells, respectively. These investigations suggest that cell phenotypes with distinct features can be generated after HDara-C treatment and that decreased deoxycytidine kinase activity appears to be one of the major mechanisms of resistance.

The partial nontandem duplication of the MLL (ALL1) gene is a novel rearrangement that generates three distinct fusion transcripts in B-cell acute lymphoblastic leukemia.

A partial nontandem duplication (PNTD) of mixed lineage leukemia (MLL) gene is described in B-cell acute lymphoid leukemia without structural cytogenetic abnormalities at 11q23 and 9p22. A duplicated portion of MLL is interrupted by the insertion of a region of 9p22 that includes the 3'-end of the AF9 gene. The PNTD encodes: (a) a PNTD transcript; (b) a partial tandem duplication of MLL; and (c) a chimeric transcript fusing MLL to the 3'-end of AF9, mimicking the t(9;11)(p22;q23) and expressed 1024-fold higher than the other two. The MLL PNTD, therefore, contributes toward leukemogenesis through simultaneous production of fusion transcripts that are otherwise encoded by three distinct genetic defects.

Silencing of GATA3 defines a novel stem cell-like subgroup of ETP-ALL.

BACKGROUND: GATA3 is pivotal for the development of T lymphocytes. While its effects in later stages of T cell differentiation are well recognized, the role of GATA3 in the generation of early T cell precursors (ETP) has only recently been explored. As aberrant GATA3 mRNA expression has been linked to cancerogenesis, we investigated the role of GATA3 in early T cell precursor acute lymphoblastic leukemia (ETP-ALL). METHODS: We analyzed GATA3 mRNA expression by RT-PCR (n = 182) in adult patients with T-ALL. Of these, we identified 70 of 182 patients with ETP-ALL by immunophenotyping. DNA methylation was assessed genome wide (Illumina Infinium® HumanMethylation450 BeadChip platform) in 12 patients and GATA3-specifically by pyrosequencing in 70 patients with ETP-ALL. The mutational landscape of ETP-ALL with respect to GATA3 expression was investigated in 18 patients and validated by Sanger sequencing in 65 patients with ETP-ALL. Gene expression profiles (Affymetrix Human genome U133 Plus 2.0) of an independent cohort of adult T-ALL (n = 83) were used to identify ETP-ALL and investigate GATA3low and GATA3high expressing T-ALL patients. In addition, the ETP-ALL cell line PER-117 was investigated for cytotoxicity, apoptosis, GATA3 mRNA expression, DNA methylation, and global gene expression before and after treatment with decitabine. RESULTS: In our cohort of 70 ETP-ALL patients, 33 % (23/70) lacked GATA3 expression and were thus defined as GATA3low. DNA methylation analysis revealed a high degree of GATA3 CpG island methylation in GATA3low compared with GATA3high ETP-ALL patients (mean 46 vs. 21 %, p < 0.0001). Genome-wide expression profiling of GATA3low ETP-ALL exhibited enrichment of myeloid/lymphoid progenitor (MLP) and granulocyte/monocyte progenitor (GMP) genes, while T cell-specific signatures were downregulated compared to GATA3high ETP-ALL. Among others, FLT3 expression was upregulated and mutational analyses demonstrated a high rate (79 %) of FLT3 mutations. Hypomethylating agents induced reversal of GATA3 silencing, and gene expression profiling revealed downregulation of hematopoietic stem cell genes and upregulation of T cell differentiation. CONCLUSIONS: We propose GATA3low ETP-ALL as a novel stem cell-like leukemia with implications for the use of myeloid-derived therapies.

High expression of connective tissue growth factor accelerates dissemination of leukaemia.

To improve treatment of acute lymphoblastic leukaemia (ALL), a better understanding of disease development is needed to tailor new therapies. Connective tissue growth factor (CTGF/CCN2) is highly expressed in leukaemia cells from the majority of paediatric patients with B-lineage ALL (pre-B ALL). CTGF is a matricellular protein and plays a role in aggressive cancers. Here we have genetically engineered leukaemia cells to modulate CTGF expression levels. Elevated CTGF levels accelerated disease dissemination and reduced survival in NOD/SCID mice. In vitro studies showed that CTGF protein induces stromal cell proliferation, promotes adhesion of leukaemia cells to stromal cells and leads to overexpression of genes associated with cell cycle and synthesis of extracellular matrix (ECM). Corresponding data from our leukaemia xenograft models demonstrated that CTGF leads to increased proliferation of non-leukaemia cells and deposition of ECM in the bone marrow. We document for the first time a functional role of CTGF in altering disease progression in a lymphoid malignancy. The findings provide support for targeting the bone marrow microenvironment in aggressive forms of leukaemia.

Deletions of the p16 gene in pediatric leukemia and corresponding cell lines.

The p16 gene (MTS1 or CDK4I) encoding an inhibitor of cyclin-dependent kinase 4 (cdk4), has been reported to be deleted in various tumor cell lines, including lines derived from leukemic cells. The reported frequency of p16 gene loss is much higher in established cell lines than in primary tumor specimens. We investigated the status of the p16 gene in pediatric leukemias using 12 established cell lines of differing phenotypes and their corresponding primary leukemic cells. Six of 12 cell lines, including acute lymphoblastic leukemia (ALL) lines of T-cell (three of four), of precursor-B cell (two of four) and of mixed phenotype (one of four), showed homozygous deletion of the p16 gene using PCR and Southern blotting. Comparison of the cell lines with their corresponding primary leukemic cells clearly showed that in all 12 paired samples there were identical findings with respect to the presence or absence of the p16 gene, demonstrating that loss of the gene was a feature of the primary leukemic cells. This is the first study to show this correlation using a panel of paired samples, indicating that p16 gene deletions were not an artifact of in vitro cell culture. Furthermore, the survival of ALL patients with p16 gene deletions was significantly inferior to those without deletions, suggesting that this genetic alteration may be a clinical prognostic factor.

HOX11 expression in pediatric acute lymphoblastic leukemia is associated with T-cell phenotype.

Based on cytogenetic analysis, chromosomal translocations involving band 10q24 occur in 4-7% of T-cell acute lymphoblastic leukemia (T-ALL). The HOX11 gene is located in this chromosomal band and is activated by translocations t(10;14) (q24;q11) and t(7;10) (q35;q24). Ectopic expression of the HOX11 gene appears to be involved in the development of T-cell tumors. The aim of this study was to determine the frequency of HOX11 activation in pediatric ALL patients and to correlate gene expression with ALL immunophenotype. None of 53 B-lineage ALLs was positive for HOX11 expression, however, Northern blot and RT-PCR analysis revealed that four of 12 T-ALLs (33%) showed expression of the gene. In order to assess whether HOX11 expression is present in other pediatric malignancies we examined a panel of 20 tumor cell lines established from solid tumors and leukemias, but none of them showed expression of HOX11. Using our RT-PCR method we confirmed that HOX11 expression is not detectable in normal T-cells. These findings indicate that HOX11 expression in pediatric ALL is exclusive to T-ALL and does not occur in B-lineage ALL. The frequency detected by molecular techniques was significantly higher than the frequency reported in the literature based on cytogenetic analysis. These results support the notion that ectopic expression of the HOX11 homeobox gene is a crucial step in T-cell tumorigenesis.

Multiple negative elements contribute to repression of the HOX11 proto-oncogene.

The HOX11 proto-oncogene is normally expressed in embryogenesis where it directs the synthesis of the spleen. In adult tissues, HOX11 expression is silenced by an unknown mechanism. Aberrant expression of HOX11 occurs in T-cell acute lymphoblastic leukaemia (T-ALL), where it is thought to be involved in T-cell immortalization. The deregulated expression of HOX11 is frequently associated with chromosomal translocations which juxtapose a T-cell receptor (TCR) gene upstream of the HOX11 gene. In these cases, it is presumed that the activation of HOX11 expression results from bringing the gene under the control of TCR enhancer elements. However, activation of HOX11 also occurs in the absence of an associated translocation in both T-ALL and erythroleukaemia cells, implying that an alternative activation mechanism may exist. We hypothesized that HOX11 may be repressed in normal T-cells and erythroid cells by the action of negative elements which may be deleted or mutated in leukaemia. We therefore conducted a search for negative elements in the human HOX11 promoter which may function to silence its expression in normal cells of the haematopoietic lineages. Since little sequence of the HOX11 promoter was available, we began our investigation by sequencing over 4.5 kilobases of untranslated DNA from upstream of HOX11. The human sequence that overlaps with the 2.1 kb of murine Hox11 is highly conserved, suggesting that a large region of DNA upstream of HOX11 may have a regulatory function. We then used transfection assays to test the ability of portions of the promoter to drive transcription of a reporter gene. These studies identified four negative elements. Two of them (NRE2 and NRE4) function in all cell lines tested, while the remaining two (NRE1 and NRE3) appear to be cell-type specific. The DNA sequences of three elements are conserved between the human and mouse HOX11/Hox11 promoters. We propose a model in which the combined action of these negative elements contributes to the overall repression of HOX11 expression in normal blood cells.