Identification and characterization of interactions between the vertebrate polycomb-group protein BMI1 and human homologs of polyhomeotic.

In Drosophila melanogaster, the Polycomb-group (PcG) genes have been identified as repressors of gene expression. They are part of a cellular memory system that is responsible for the stable transmission of gene activity to progeny cells. PcG proteins form a large multimeric, chromatin-associated protein complex, but the identity of its components is largely unknown. Here, we identify two human proteins, HPH1 and HPH2, that are associated with the vertebrate PcG protein BMI1. HPH1 and HPH2 coimmunoprecipitate and cofractionate with each other and with BMI1. They also colocalize with BMI1 in interphase nuclei of U-2 OS human osteosarcoma and SW480 human colorectal adenocarcinoma cells. HPH1 and HPH2 have little sequence homology with each other, except in two highly conserved domains, designated homology domains I and II. They share these homology domains I and II with the Drosophila PcG protein Polyhomeotic (Ph), and we, therefore, have named the novel proteins HPH1 and HPH2. HPH1, HPH2, and BMI1 show distinct, although overlapping expression patterns in different tissues and cell lines. Two-hybrid analysis shows that homology domain II of HPH1 interacts with both homology domains I and II of HPH2. In contrast, homology domain I of HPH1 interacts only with homology domain II of HPH2, but not with homology domain I of HPH2. Furthermore, BMI1 does not interact with the individual homology domains. Instead, both intact homology domains I and II need to be present for interactions with BMI1. These data demonstrate the involvement of homology domains I and II in protein-protein interactions and indicate that HPH1 and HPH2 are able to heterodimerize.

RING1 is associated with the polycomb group protein complex and acts as a transcriptional repressor.

The Polycomb (Pc) protein is a component of a multimeric, chromatin-associated Polycomb group (PcG) protein complex, which is involved in stable repression of gene activity. The identities of components of the PcG protein complex are largely unknown. In a two-hybrid screen with a vertebrate Pc homolog as a target, we identify the human RING1 protein as interacting with Pc. RING1 is a protein that contains the RING finger motif, a specific zinc-binding domain, which is found in many regulatory proteins. So far, the function of the RING1 protein has remained enigmatic. Here, we show that RING1 coimmunoprecipitates with a human Pc homolog, the vertebrate PcG protein BMI1, and HPH1, a human homolog of the PcG protein Polyhomeotic (Ph). Also, RING1 colocalizes with these vertebrate PcG proteins in nuclear domains of SW480 human colorectal adenocarcinoma and Saos-2 human osteosarcoma cells. Finally, we show that RING1, like Pc, is able to repress gene activity when targeted to a reporter gene. Our findings indicate that RING1 is associated with the human PcG protein complex and that RING1, like PcG proteins, can act as a transcriptional repressor.

Pertubation of B and T cell development and predisposition to lymphomagenesis in Emu Bmi1 transgenic mice require the Bmi1 RING finger.

Proviral activation of the Bmi1 gene has implicated Bmi1 as a collaborator of c-Myc in lymphomagenesis. To determine the effect of Bmi1 overexpression on hematopoiesis and lymphomagenesis transgenic mice were generated that overexpress different forms of the Bmi1 protein in their lymphoid compartment. Emu Bmi1 transgenic mice, overexpressing the wild type Bmi1 protein showed a perturbed lymphoid development and were highly susceptible to B and T cell lymphomagenesis. Mutational analysis of the Bmi1 protein demonstrated that the conserved N-terminal RING finger and central part of Bmi1 are essential for its oncogenic potential whereas the C-terminal Pro-Ser rich region is not required. We have used provirus tagging in the Emu Bmi1 mice to identify genes that cooperate with Bmi1 in lymphomagenesis. MoMLV infection in Emu Bmi1 transgenic mice accelerated lymphoma development. Proviral activation of the Pim and Myc genes but not the Gfi1 gene were frequently observed in these tumors. These results demonstrate that Bmi1 is a potent oncogene and suggest that it plays an important role in early lymphoid development.

MPc2, a new murine homolog of the Drosophila polycomb protein is a member of the mouse polycomb transcriptional repressor complex.

The evolutionarily conserved polycomb and trithorax-group genes are required to maintain stable expression patterns of homeotic genes and other target genes throughout development. Here, we report the cloning and characterization of a novel mouse polycomb homolog, MPc2, in addition to the previously described M33 polycomb gene. Co-immunoprecipitations and subnuclear co-localization studies show that MPc2 interacts with the mouse polycomb-group oncoprotein Bmi1 and is a new member of the mouse polycomb multiprotein complex. Gal4DB-MPc2 or -M33 fusion proteins mediate a five- to tenfold repression of stably integrated reporter constructs carrying GAL4 binding sites, demonstrating that these proteins are transcriptional repressors. The MPc2 gene is localized on chromosome 11, in close proximity to the classical mouse mutations tail short (Ts) and rabo torcido (Rbt). Ts and Rbt hemizygous mice display anemia and transformations of the axial skeleton reminiscent of phenotypes observed in mice with mutated polycomb or trithorax-group genes, suggesting that MPc2 is a candidate gene for Ts and Rbt.

Characterization and chromosomal localization of the human proto-oncogene BMI-1.

The proto-oncogene bmi-1 is frequently activated by Moloney murine leukemia proviral insertions in E mu-myc transgenic mice1,2. Using a mouse bmi-1 cDNA probe a transcript of 3.3 kb was detected on Northern blots of human Burkitt's lymphoma cell lines. We have isolated and sequenced cDNA clones from a human erythroleukemia cell line (K562) derived cDNA library, using different mouse bmi-1 cDNA fragments as a probe. Analysis of genomic BMI-1 sequences reveals a gene structure which is very similar to that of the mouse, consisting of at least 10 exons. The human cDNA is 3203 bp in length and shows 86% identity to the mouse nucleotide sequence. The open reading frame encodes a protein of 326 amino acids which shares 98% identity to the amino acid sequence of mouse bmi-1 protein. In vitro translation experiments show that human cDNA derived RNA translates into a protein with a mobility of 44-46 kD on SDS polyacrylamide gels. Fluorescence in situ hybridization (FISH) on metaphase chromosome spreads located the human BMI-1 gene to the short arm of chromosome 10 (10p13), a region known to be involved in translocations in various leukemias.

Transformation of axial skeleton due to overexpression of bmi-1 in transgenic mice.

The oncogene bmi-1, which was originally found to be involved in B- and T-cell lymphoma formation encodes a protein with a domain of homology to the Drosophila protein Posterior sex combs (Psc) and its relative Suppressor 2 of Zeste (Su(z)2) (refs 4 and 5). Psc is a member of the Polycomb-group gene family, which is required to maintain the repression of homeotic genes that regulate the identities of Drosophila segments. The possibility that bmi-1 may play a similar role in vertebrates was suggested by our previous finding that mice lacking the bmi-1 gene show posterior transformations of the axial skeleton. Here we report that transgenic mice overexpressing Bmi-1 protein show the opposite phenotype, namely a dose-dependent anterior transformation of vertebral identity. The anterior expression boundary of the Hoxc-5 gene is shifted in the posterior direction, indicating that Bmi-1 is involved in the repression of Hox genes. We propose that Bmi-1 is a member of a vertebrate Polycomb complex that regulates segmental identity by repressing Hox genes throughout development.

Genetic interactions and dosage effects of Polycomb group genes in mice.

In Drosophila and mouse, Polycomb group genes are involved in the maintenance of homeotic gene expression patterns throughout development. Here we report the skeletal phenotypes of compound mutants for two Polycomb group genes bmi1 and M33. We show that mice deficient for both bmi1 and M33 present stronger homeotic transformations of the axial skeleton as compared to each single Polycomb group mutant, indicating strong dosage interactions between those two genes. These skeletal transformations are accompanied with an enhanced shift of the anterior limit of expression of several Hox genes in the somitic mesoderm. Our results demonstrate that in mice the Polycomb group genes act in synergy to control the nested expression pattern of some Hox genes in somitic mesodermal tissues during development.

Interlocked circle formation by group I introns: structural requirements and mechanism.

Precursor RNA transcribed from the yeast mitochondrial gene coding for the large ribosomal RNA contains a group I intron that can excise itself in vitro. Apart from group I specific sequence elements the intron also contains a gene encoding a DNA endonuclease involved in intron dispersal. A precursor RNA derivative from which this gene has been removed self-splices efficiently, but due to activation of cryptic opening sites located in the 5' exon, the 3' part of this exon is sometimes co-excised with the intron. Upon further reaction, this enlarged intron molecules give rise to interlocked circles, comprising small circles derived from 5' exon parts and large circles of the intron. Sequence comparison between cryptic opening sites and authentic splice sites reveals in most cases homology with the 3' exon part that is capable of interacting with the Internal Guide Sequence. The role of the IGS was further substantiated by replacing the cryptic opening sites with well defined sequences of authentic splice sites: one corresponding to the 3' splice site and its mutant derivatives, the other to a fragment containing the natural 5'-3' exon junction. Precursor RNAs derived from these constructs give rise to interlocked circles, and mutation studies confirm that the 3' exon nucleotides flanking a 3' splice site are essential for their formation. The results underline the crucial role of the IGS in interlocked circle formation which behaves similarly as in the normal self-splicing reactions. It has been proposed that the two short helices formed by basepairing of the IGS with the 5' and 3' exon can co-axially stack on top of each other forming a quasi continuous RNA double helix or pseudoknot. We present a model explaining how transesterification reactions of a mutant precursor RNA in such a pseudoknot can lead to interlocked circles. The experiments support the notion that a similar structure is also operative in splicing of wild type precursor RNA.

The DNA binding specificity of the bipartite POU domain and its subdomains.

The POU domain is a conserved DNA binding region of approximately 160 amino acids present in a family of eukaryotic transcription factors that play regulatory roles in development. The POU domain consists of two subdomains, the POU-specific (POUS) domain and a POU-type homeodomain (POUHD). We show here that, like the POUHD, the Oct-1 POUS domain can bind autonomously to DNA but with low affinity. DNA binding studies and in vitro binding site selection revealed that the POU subdomains each have a different sequence specificity. The binding consensus of the POUS domain [gAATAT(G/T)CA] and POUHD (RTAATNA) respectively overlap the 'left half' and right half' of the POU domain recognition sequence [a(a/t)TATGC(A/T) AAT(t/a)t]. In addition to the core sequence, which is very similar to the octamer motif (ATGCAAAT), the flanking bases make a significant contribution to the binding affinity of the POU domain. Interestingly, at some positions the sequence preferences of the isolated POU subdomains are distinct from those of the POU domain, suggesting that the POU domain binding site is more than a simple juxtaposition of the POUS and POUHD target sequences. In addition, analysis of the binding kinetics of the POU domain and POUHD indicates that the POUS domain enhances the binding affinity by reducing the dissociation rate. Our results show that the POU domain proteins have DNA binding properties distinct from those of classic homeodomain proteins. We suggest a model for the way in which an additional conserved domain adds further specificity to DNA recognition by homeodomain proteins.

Identification of Bmi1-interacting proteins as constituents of a multimeric mammalian polycomb complex.

The Bmi1 gene has been identified as a mouse Polycomb group (Pc-G) gene implicated in the regulation of Hox gene expression. Here we describe the characterization of a Bmi binding protein Mph1, which shares similarity to Drosophila polyhomeotic. Coimmunoprecipitation experiments indicate that Bmi1 and Mph1, as well as the Mel18 and M33 proteins described previously, are constituents of a multimeric protein complex in mouse embryos and human cells. A central domain of Bmi1 interacts with the carboxyl terminus of Mph1, whereas a conserved alpha-helical domain in the Mph1 protein is required for its homodimerization. Transgenic mice overexpressing various mutant Bmi1 proteins demonstrate that the central domain of Bmil is required for the induction of anterior transformations of the axial skeleton. Bmi1, M33, and Mph1 show an overlapping speckled distribution in interphase nuclei. These data provide molecular evidence for the existence of a mammalian Polycomb complex.