Selection of reversions and suppressors of a mutation in the CBF binding site of a lymphomagenic retrovirus

Novel principles of gamma-retroviral insertional transcription activation in murine leukemia virus-induced end-stage tumors

BACKGROUND

Insertional mutagenesis screens of retrovirus-induced mouse tumors have proven valuable in human cancer research and for understanding adverse effects of retroviral-based gene therapies. In previous studies, the assignment of mouse genes to individual retroviral integration sites has been based on close proximity and expression patterns of annotated genes at target positions in the genome. We here employed next-generation RNA sequencing to map retroviral-mouse chimeric junctions genome-wide, and to identify local patterns of transcription activation in T-lymphomas induced by the murine leukemia gamma-retrovirus SL3-3. Moreover, to determine epigenetic integration preferences underlying long-range gene activation by retroviruses, the colocalization propensity with common epigenetic enhancer markers (H3K4Me1 and H3K27Ac) of 6,117 integrations derived from end-stage tumors of more than 2,000 mice was examined.

RESULTS

We detected several novel mechanisms of retroviral insertional mutagenesis: bidirectional activation of mouse transcripts on opposite sides of a provirus including transcription of unannotated mouse sequence; sense/antisense-type activation of genes located on opposite DNA strands; tandem-type activation of distal genes that are positioned adjacently on the same DNA strand; activation of genes that are not the direct integration targets; combination-type insertional mutagenesis, in which enhancer activation, alternative chimeric splicing and retroviral promoter insertion are induced by a single retrovirus. We also show that irrespective of the distance to transcription start sites, the far majority of retroviruses in end-stage tumors colocalize with H3K4Me1 and H3K27Ac-enriched regions in murine lymphoid tissues.

CONCLUSIONS

We expose novel retrovirus-induced host transcription activation patterns that reach beyond a single and nearest annotated gene target. Awareness of this previously undescribed layer of complexity may prove important for elucidation of adverse effects in retroviral-based gene therapies. We also show that wild-type gamma-retroviruses are frequently positioned at enhancers, suggesting that integration into regulatory regions is specific and also subject to positive selection for sustaining long-range gene activation in end-stage tumors. Altogether, this study should prove useful for extrapolating adverse outcomes of retroviral vector therapies, and for understanding fundamental cellular regulatory principles and retroviral biology.

The purinergic P2Y14 receptor axis is a molecular determinant for organism survival under in utero radiation toxicity

In utero exposure of the embryo and fetus to radiation has been implicated in malformations or fetal death, and often produces lifelong health consequences such as cancers and mental retardation. Here we demonstrate that deletion of a G-protein-coupled purinergic receptor, P2Y₁₄, confers potent resistance to in utero radiation. Intriguingly, a putative P2Y₁₄ receptor ligand, UDP-glucose, phenocopies the effect of P2Y₁₄ deficiency. These data indicate that P2Y₁₄ is a receptor governing in utero tolerance to genotoxic stress that may be pharmacologically targeted to mitigate radiation toxicity in pregnancy.

Control of pathogenicity and disease specificity of a T-lymphomagenic gammaretrovirus by E-box motifs but not by an overlapping glucocorticoid response element

Although transcription factors of the basic helix-loop-helix family have been shown to regulate enhancers of lymphomagenic gammaretroviruses through E-box motifs, the overlap of an E-box motif (Egre) with the glucocorticoid response element (GRE) has obscured their function in vivo. We report here that Egre, but not the GRE, affects disease induction by the murine T-lymphomagenic SL3-3 virus. Mutating all three copies of Egre prolonged the tumor latency period from 60 to 109 days. Further mutating an E-box motif (Ea/s) outside the enhancer prolonged the latency period to 180 days, suggesting that Ea/s works as a backup site for Egre. While wild-type SL3-3 and GRE and Ea/s mutants exclusively induced T-cell lymphomas with wild-type latencies mainly of the CD4(+) CD8(-) phenotype, Egre as well as the Egre and Ea/s mutants induced B-cell lymphomas and myeloid leukemia in addition to T-cell lymphomas. T-cell lymphomas induced by the two Egre mutants had the same phenotype as those induced by wild-type SL3-3, indicating the incomplete disruption of T-cell lymphomagenesis, which is in contrast to previous findings for a Runx site mutant of SL3-3. Mutating the Egre site or Egre and Ea/s triggered several tumor phenotype-associated secondary enhancer changes encompassing neighboring sites, none of which led to the regeneration of an E-box motif. Taken together, our results demonstrate a role for the E-box but not the GRE in T lymphomagenesis by SL3-3, unveil an inherent broader disease specificity of the virus, and strengthen the notion of selection for more potent enhancer variants of mutated viruses during tumor development.

ALY is a common coactivator of RUNX1 and c-Myb on the type B leukemogenic virus enhancer

Type B leukemogenic virus (TBLV), a mouse mammary tumor virus (MMTV) variant, often induces T-cell leukemias and lymphomas by c-myc activation following viral DNA integration. Transfection assays using a c-myc reporter plasmid indicated that the TBLV long terminal repeat (LTR) enhancer is necessary for T-cell-specific increases in basal reporter activity. The sequence requirements for this effect were studied using mutations of the 62-bp enhancer region in an MMTV LTR reporter vector. Deletion of a nuclear factor A-binding site dramatically reduced reporter activity in Jurkat T cells. However, a 41-bp enhancer missing the RUNX1 site still retained minimal enhancer function. DNA affinity purification using a TBLV enhancer oligomer containing the RUNX1 binding site followed by mass spectrometry resulted in the identification of ALY. Subsequent experiments focused on the reconstitution of enhancer activity in epithelial cells. ALY overexpression synergized with RUNX1B on TBLV enhancer activity, and synergism required the RUNX1B-binding site. A predicted c-Myb binding site in the enhancer was confirmed after c-myb overexpression elevated TBLV LTR reporter activity, and overexpression of c-Myb and RUNX1B together showed additive effects on reporter gene levels. ALY also synergized with c-Myb, and coimmunoprecipitation experiments demonstrated an interaction between ALY and c-Myb. These experiments suggest a central role for ALY in T-cell enhancer function and oncogene activation.

Sequence analysis of murine leukemia virus envelope gene from inoculated mice

Introduction of amino acids substitutions in murine leukemia virus genome is a powerful method to determine the relative importance of various viral factors in pathogenesis. However, introduction of such amino acids substitution could result in viruses at a selective disadvantage, and eventual selection of revertants. It is thus essential to verify if the mutation is maintained stably in replicating virus and in infected tumor cells. In the present study, viral nucleic acid sequences from diseased animals were determined using different approaches. Small blood samples were found adequate for direct RNA extraction and reverse transcriptase-PCR amplification followed by automated DNA sequencing. Alternatively, replication-competent viruses were recovered specifically by applying blood samples onto permissive cells; viral RNA is then extracted from tissue culture medium and similarly sequenced. Tissue samples were also used to amplify viral sequences from tumors DNA while small pieces of tumors tissues were applied onto permissive cells to isolate replicating viruses. The combined experimental approach was used to show sequence conservation using a mutant altered in the intracytoplasmic region of viral envelope glycoprotein. No difference was observed between viruses recovered directly from the animal and those amplified onto cultured cells.

Mutation of all Runx (AML1/core) sites in the enhancer of T-lymphomagenic SL3-3 murine leukemia virus unmasks a significant potential for myeloid leukemia induction and favors enhancer evolution toward induction of other disease patterns

SL3-3 murine leukemia virus is a potent inducer of T-lymphomas in mice. Using inbred NMRI mice, it was previously reported that a mutant of SL3-3 with all enhancer Runx (AML1/core) sites disrupted by 3-bp mutations (SL3-3dm) induces predominantly non-T-cell tumors with severely extended latency (S. Ethelberg, J. Lovmand, J. Schmidt, A. Luz, and F. S. Pedersen, J. Virol. 71:7273-7280, 1997). By use of three-color flow cytometry and molecular and histopathological analyses, we have now performed a detailed phenotypic characterization of SL3-3- and SL3-3dm-induced tumors in this mouse strain. All wild-type induced tumors had clonal T-cell receptor beta rearrangements, and the vast majority were CD3(+) CD4(+) CD8(-) T-lymphomas. Such a consistent phenotypic pattern is unusual for murine leukemia virus-induced T-lymphomas. The mutant virus induced malignancies of four distinct hematopoietic lineages: myeloid, T lymphoid, B lymphoid, and erythroid. The most common disease was myeloid leukemia with maturation. Thus, mutation of all Runx motifs in the enhancer of SL3-3 severely impedes viral T-lymphomagenicity and thereby discloses a considerable and formerly unappreciated potential of this virus for myeloid leukemia induction. Proviral enhancers with complex structural alterations (deletions, insertions, and/or duplications) were found in most SL3-3dm-induced T-lymphoid tumors and immature myeloid leukemias but not in any cases of myeloid leukemia with maturation, mature B-lymphoma, or erythroleukemia. Altogether, our results indicate that the SL3-3dm enhancer in itself promotes induction of myeloid leukemia with maturation but that structural changes may arise in vivo and redirect viral disease specificity to induction of T-lymphoid or immature myeloid leukemias, which typically develop with moderately shorter latencies.

Relative age of proviral porcine endogenous retrovirus sequences in Sus scrofa based on the molecular clock hypothesis

Porcine endogenous retroviruses (PERV) are discussed as putative infectious agents in xenotransplantation. PERV classes A, B, and C harbor different envelope proteins. Two different types of long terminal repeat (LTR) structures exist, of which both are present only in PERV-A. One type of LTR contains a distinct repeat structure in U3, while the other is repeatless, conferring a lower level of transcriptional activity. Since the different LTR structures are distributed unequally among the proviruses and, apparently, PERV is the only virus harboring two different LTR structures, we were interested in determining which LTR is the ancestor. Replication-competent viruses can still be found today, suggesting an evolutionary recent origin. Our studies revealed that the age of PERV is at most 7.6 x 10(6) years, whereas the repeatless LTR type evolved approximately 3.4 x 10(6) years ago, being the phylogenetically younger structure. The age determined for PERV correlates with the time of separation between pigs (Suidae, Sus scrofa) and their closest relatives, American-born peccaries (Tayassuidae, Pecari tajacu), 7.4 x 10(6) years ago.

Molecular cloning and functional characterization of infectious PERV and development of diagnostic tests

Pigs are the donor animals of choice for xenotransplantation (XTx) and xenogeneic cell therapy measurements. Most known porcine pathogens can be controlled by conventional means like vaccination, medication or specific pathogen-free breeding conditions. As pigs have co-evolved very closely with humans for a few millennia it is not very likely that even asymptomatic pathogens have escaped attention. Porcine endogenous retroviruses (PERV) are different from conventional pathogens as they are chromosomally fixed in every cell of the animal, hence PERV cannot be easily controlled. While PERV show no phenotype in the porcine host, recent data demonstrate that some polytropic proviruses can be activated by external stimuli and that those can productively infect human cells in vitro. In evaluation of the retrovirological safety of XTx, we determined the number of replication-competent PERV to be limited and to exhibit a heterogeneous distribution, therefore suggesting that they could be removed by conventional breeding. The transcriptional regulation of some PERV due to repetitive elements in their long terminal repeats enables their adaptation to new host cells. The diagnostic tools available, based on immunological and polymerase chain reaction techniques, were shown to be sensitive in both the animal and in vitro, but must still show their potential in human XTx recipients, where they are confronted with very low antigen expression and the phenomenon of microchimerism.

Duplication of U3 sequences in the long terminal repeat of mink cell focus-inducing viruses generates redundancies of transcription factor binding sites important for the induction of thymomas

The ability of mink cell focus-inducing (MCF) viruses to induce thymomas is determined, in part, by transcriptional enhancers in the U3 region of their long terminal repeats (LTRs). To elucidate sequence motifs important for enhancer function in vivo, we injected newborn mice with MCF 1dr (supF), a weakly pathogenic, molecularly tagged (supF) MCF virus containing only one copy of a sequence that is present as two copies (known as the directly repeated [DR] sequence) in the U3 region of MCF 247 and analyzed LTRs from supF-tagged proviruses in two resulting thymomas. Tagged proviruses integrated upstream and in the reverse transcriptional orientation relative to c-myc provided the focus of our studies. These proviruses are thought to contribute to thymoma induction by enhancer-mediated deregulation of c-myc expression. The U3 region in a tagged LTR in one thymoma was cloned and sequenced. Relative to MCF 1dr (supF), the cloned U3 region contained an insertion of 140 bp derived predominantly from the DR sequence of the injected virus. The inserted sequence contains predicted binding sites for transcription factors known to regulate the U3 regions of various murine leukemia viruses. Similar constellations of binding sites were duplicated in two proviral LTRs integrated upstream from c-myc in a second thymoma. We replaced the U3 sequences in an infectious molecular clone of MCF 247 with the cloned proviral U3 sequences from the first thymoma and generated an infectious chimeric virus, MCF ProEn. When injected into neonatal AKR mice, MCF ProEn was more pathogenic than the parental virus, MCF 1dr (supF), as evidenced by the more rapid onset and higher incidence of thymomas. Molecular analyses of the resultant thymomas indicated that the U3 region of MCF ProEn was genetically stable. These data suggest that the arrangement and/or redundancy of transcription factor binding sites generated by specific U3 sequence duplications are important to the biological events mediated by MCF proviruses integrated near c-myc that contribute to transformation.

What retroviruses teach us about the involvement of c-Myc in leukemias and lymphomas

Overexpression of the cellular oncogene c-Myc frequently occurs during induction of leukemias and lymphomas in many species. Retroviruses have enhanced our understanding of the role of c-Myc in such tumors. Leukemias and lymphomas induced by retroviruses activate c-Myc by: (1) use of virally specified proteins that increase c-Myc transcription, (2) transduction and modification of c-Myc to generate a virally encoded form of the gene, v-Myc, and (3) proviral integration in or near c-Myc. Proviral integrations elevate transcription by insertion of retroviral enhancers found in the long terminal repeat (LTR). Studies of the LTR enhancer elements from these retroviruses have revealed the importance of these elements for c-Mycactivation in several cell types. Retroviruses also have been used to identify genes that collaborate with c-Myc during development and progression of leukemias and lymphomas. In these experiments, animals that are transgenic for c-Mycoverexpression (often in combination with the overexpression or deletion of known proto-oncogenes) have been infected with retroviruses that then insertionally activate novel co-operating cellular genes. The retrovirus then acts as a molecular 'tag' for cloning of these genes. This review covers several aspects of c-Myc involvement in retrovirally induced leukemias and lymphomas.

Disruption of hematopoiesis and thymopoiesis in the early premalignant stages of infection with SL3-3 murine leukemia virus

A time course analysis of SL3-3 murine leukemia virus (SL3) infection in thymus and bone marrow of NIH/Swiss mice was performed to assess changes that occur during the early stages of progression to lymphoma. Virus was detectable in thymocytes, bone marrow, and spleen as early as 1 to 2 weeks postinoculation (p.i.). In bone marrow, virus infection was detected predominantly in immature myeloid or granulocytic cells. Flow cytometry revealed significant reductions of the Ter-119(+) and Mac-1(+) populations, and significant expansions of the Gr-1(+) and CD34(+) populations, between 2 and 4 weeks p.i. Analysis of colony-forming potential confirmed these findings. In the thymus, SL3 replication was associated with significant disruption in thymocyte subpopulation distribution between 4 and 7 weeks p.i. A significant thymic regression was observed just prior to the clonal outgrowth of tumor cells. Proviral long terminal repeats (LTRs) with increasing numbers of enhancer repeats were observed to accumulate exclusively in the thymus during the first 8 weeks p.i. Observations were compared to the early stages of infection with a virtually nonpathogenic SL3 mutant, termed SL3DeltaMyb5, which was shown by real-time PCR to be replication competent. Comparison of SL3 with SL3DeltaMyb5 implicated certain premalignant changes in tumorigenesis, including (i) increased proportions of Gr-1(+) and CD34(+) bone marrow progenitors, (ii) a significant increase in the proportion of CD4(-) CD8(-) thymocytes, (iii) thymic regression prior to tumor outgrowth, and (iv) accumulation of LTR enhancer variants. A model in which disrupted bone marrow hematopoiesis and thymopoiesis contribute to the development of lymphoma in the SL3-infected animal is discussed.

Selection for c-myc integration sites in polyclonal T-cell lymphomas

Type B leukemogenic virus (TBLV) is highly related to mouse mammary tumor virus but induces rapidly appearing T-cell lymphomas in mice. Unlike other T-cell tumors induced by retroviruses, only 5 to 10% of TBLV-induced lymphomas have detectable viral integrations near c-myc by Southern blotting, whereas Northern blotting has shown that most tumors have two- to sixfold overexpression of c-myc RNA. In this report, PCR was used to demonstrate that at least 30% of these lymphomas have TBLV insertions near c-myc. Some tumors contained multiple TBLV proviruses in different locations and orientations, suggesting that the tumors are polyclonal. The integrated proviruses near c-myc had different numbers (two to four) of long terminal repeat (LTR) enhancer repeats, although LTRs with three-repeat enhancers dominated the proviral population. Passage of polyclonal tumors in immunocompetent mice and semiquantitative PCR revealed that only cells with particular integrations were selected for growth. In three of six tumors tested, proviruses containing four-repeat enhancers near c-myc were selected during tumor passage. Since tumor cell selection may be accomplished by overexpression of c-myc RNA due to proximity to the unique TBLV LTR enhancer, we inserted LTRs at various locations within a plasmid containing the entire c-myc locus and cellular flanking sequences. To quantitatively measure effects on transcription, the Renilla luciferase gene was substituted for most of c-myc exon 2, and transient transfections were performed with c-myc reporter constructs in two different T-cell lines. As expected, insertion of a TBLV LTR with three-repeat enhancers in either orientation, 5" and 3", of the myc gene elevated reporter activity from 2- to 160-fold, consistent with enhancer function, but four-repeat LTRs had lower levels of expression compared to three-repeat LTRs. Surprisingly, LTR insertions that gave maximal c-myc expression in transient-transfection assays declined in tumor cells selected for growth in vivo. Selection for clonal growth may occur in tumor cells that have modest c-myc overexpression after proviral insertion to prevent apoptosis.

Ikaros, a lymphoid-cell-specific transcription factor, contributes to the leukemogenic phenotype of a mink cell focus-inducing murine leukemia virus

Mink cell focus-inducing (MCF) viruses induce T-cell lymphomas in AKR/J strain mice. MCF 247, the prototype of this group of nonacute murine leukemia viruses, transforms thymocytes, in part, by insertional mutagenesis and enhancer-mediated dysregulation of cellular proto-oncogenes. The unique 3' (U3) regions in the long terminal repeats of other murine leukemia viruses contain transcription factor binding sites known to be important for enhancer function and for the induction of T-cell lymphomas. Although transcription factor binding sites important for the biological properties of MCF 247 have not been identified, pathogenesis studies from our laboratory suggested to us that binding sites for Ikaros, a lymphoid-cell-restricted transcriptional regulator, affect the biological properties of MCF 247. In this report, we demonstrate that Ikaros binds to predicted sites in U3 sequences of MCF 247 and that site-directed mutations in these sites greatly diminish this binding in vitro. Consistent with these findings, ectopic expression of Ikaros in murine cells that do not normally express this protein significantly increases transcription from the viral promoter in transient gene expression assays. Moreover, site-directed mutations in specific Ikaros-binding sites reduce this activity in T-cell lines that express Ikaros endogenously. To determine whether the Ikaros-binding sites are functional in vivo, we inoculated newborn mice with a variant MCF virus containing a mutant Ikaros-binding site. The variant virus replicated in thymocytes less efficiently and induced lymphomas with a delayed onset compared to the wild-type virus. These data are consistent with the hypothesis that the Ikaros-binding sites in the U3 region of MCF 247 are functional and cooperate with other DNA elements for optimal enhancer function in vivo.

The number of a U3 repeat box acting as an enhancer in long terminal repeats of polytropic replication-competent porcine endogenous retroviruses dynamically fluctuates during serial virus passages in human cells

The organization and transcriptional regulation of porcine endogenous retrovirus (PERV) long terminal repeats (LTRs) are unknown. We have studied the activity of LTRs from replication-competent molecular clones by performing luciferase reporter assays. The LTRs differ in the presence and number of 39-bp repeats located in U3 that confer strong promoter activity in human, simian, canine, feline, and porcine cell lines, whereas for LTRs devoid of the repeats, the promoter strength was significantly reduced. As the activity of a heterologous simian virus 40 promoter and a homologous repeat-deficient LTR was elevated by four 39-bp repeats independently of its orientation and location, the repeat box complies with the definition of an enhancer. During serial virus passaging of molecular PERV clones on human 293 cells, proviral LTRs demonstrated adaptation of transcriptional activity by dynamic changes of the number of 39-bp repeats in the course of up to 12 passaging cycles.

Comparison of replication-competent molecular clones of porcine endogenous retrovirus class A and class B derived from pig and human cells

Vertically transmitted endogenous retroviruses pose an infectious risk in the course of pig-to-human transplantation of cells, tissues, and organs. Two classes of polytropic type C porcine endogenous retroviruses (PERV) which are infectious for human cells in vitro are known. Recently, we described the cloning and characterization of replication-competent PERV-B sequences from productively infected human cells (F. Czauderna, N. Fischer, K. Boller, R. Kurth, and R. R. Tönjes, J. Virol. 74:4028-4038, 2000). Here, we report the isolation of infectious molecular PERV-A and PERV-B clones from pig cells and compare these proviruses with clones derived from infected human 293 cells. In addition to clone PERV-A(42) derived from 293 cells, four "native" full-length proviral PERV sequences derived from a genomic library of the porcine cell line PK15 were isolated. Three identical class A clones, designated PK15-PERV-A(42), PK15-PERV-A(45), and PK15-PERV-A(58), and one class B clone, PK15-PERV-B(213), were characterized. PK15-PERV-B(213) is highly homologous but distinct from the previously described clone PERV-B(43). PK15-PERV-A(58) demonstrates close homology to PERV-A(42) in env and to PERV-C in long terminal repeat, gag, and pro/pol sequences. All three PERV clones described here were replication competent upon infection of susceptible cell lines. The findings suggest that the pig genome harbors a limited number of infectious PERV-A and -B sequences.

Type B leukemogenic virus has a T-cell-specific enhancer that binds AML-1

Type B leukemogenic virus (TBLV) induces rapidly appearing T-cell tumors in mice. TBLV is highly related to mouse mammary tumor virus (MMTV) except that TBLV long terminal repeats (LTRs) have a deletion of negative regulatory elements and a triplication of sequences flanking the deletion. To determine if the LTR triplication represents a viral enhancer element, we inserted the triplication upstream and downstream in either orientation relative to the thymidine kinase promoter linked to the luciferase gene. These experiments showed that upregulation of reporter gene activity by the TBLV triplication was relatively orientation independent, consistent with the activity of eukaryotic enhancer elements. TBLV enhancer activity was observed in T-cell lines but not in fibroblasts, B cells, or mammary cells, suggesting that enhancer function is cell type dependent. To analyze the transcription factor binding sites that are important for TBLV enhancer function, we prepared substitution mutations in a reconstituted C3H MMTV LTR that recapitulates the deletion observed in the TBLV LTR. Transient transfections showed that a single mutation (556M) decreased TBLV enhancer activity at least 20-fold in two different T-cell lines. This mutation greatly diminished AML-1 (recently renamed RUNX1) binding in gel shift assays with a mutant oligonucleotide, whereas AML-1 binding to a wild-type TBLV oligomer was specific, as judged by competition and supershift experiments. The 556 mutation also reduced TBLV enhancer binding of two other protein complexes, called NF-A and NF-B, that did not appear to be related to c-Myb or Ets. AML-1 overexpression in a mammary cell line enhanced expression from the TBLV LTR approximately 30-fold. These data suggest that binding of AML-1 to the TBLV enhancer, likely in combination with other factors, is necessary for optimal enhancer function.

Groucho/TLE/R-esp proteins associate with the nuclear matrix and repress RUNX (CBF(alpha)/AML/PEBP2(alpha)) dependent activation of tissue-specific gene transcription

The Runt related transcription factors RUNX (AML/CBF(alpha)/PEBP2(alpha)) are key regulators of hematopoiesis and osteogenesis. Using co-transfection experiments with four natural promoters, including those of the osteocalcin (OC), multi drug resistance (MDR), Rous Sarcoma Virus long terminal repeat (LTR), and bone sialoprotein (BSP) genes, we show that each of these promoters responds differently to the forced expression of RUNX proteins. However, the three RUNX subtypes (i.e. AML1, AML2, and AML3) regulate each promoter in a similar manner. Although the OC promoter is activated in a C terminus dependent manner, the MDR, LTR and BSP promoters are repressed by three distinct mechanisms, either independent of or involving the AML C terminus, or requiring only the conserved C-terminal pentapeptide VWRPY. Using yeast two hybrid assays we find that the C terminus of AML1 interacts with a Groucho/TLE/R-esp repressor protein. Co-expression assays reveal that TLE proteins repress AML dependent activation of OC gene transcription. Western and northern blot analyses suggest that TLE expression is regulated reciprocally with the levels of OC gene expression during osteoblast differentiation. Digital immunofluorescence microscopy results show that TLE1 and TLE2 are both associated with the nuclear matrix, and that a significant subset of each colocalizes with AML transcription factors. This co-localization of TLE and AML proteins is lost upon removing the C terminus of AML family members. Our findings indicate that suppression of AML-dependent gene activation by TLE proteins involves functional interactions with the C terminus of AML at the nuclear matrix in situ. Our data are consistent with the concept that the C termini of AML proteins support activation or repression of cell-type specific genes depending on the regulatory organization of the target promoter and subnuclear localization.