We describe a novel mammalian DNA binding activity that requires at least two symmetrically methylated CpG dinucleotides in its recognition sequence, preferably within the sequence 5CGCG. A key component of the activity is Kaiso, a protein with POZ and zincfinger domains that is known to associate with p120 catenin. We find that Kaiso behaves as a methylationdependent transcriptional repressor in transient transfection assays. Kaiso is a constituent of one of two methylCpG binding complexes originally designated as MeCP1. The data suggest that zinc-finger motifs are responsible for DNA binding, and may therefore target repression to specific methylated regions of the genome. As Kaiso associates with p120 catenin, Kaiso may link events at the cell surface with DNA methylation-dependent gene silencing.

The involvement of Mts1(S100A4), a small Ca 2+ -binding protein in tumor progression and metastasis had been demonstrated. However, the mechanism by which mts1(S100A4) promoted metastasis had not been identi®ed. Here we demonstrated that Mts1(S100A4) had signi®cant stimulatory eect on the angiogenesis. We detected high incidence of hemangiomas ± benign tumors of vascular origin in aged transgenic mice ubiquitously expressing the mts1(S100A4) gene. Furthermore, the serum level of the Mts1(S100A4) protein increased with ageing. Tumors developed in Mts1-transgenic mice revealed an enhanced vascular density. We showed that an oligomeric, but not a dimeric form of the Mts1(S100A4) protein was capable of enhancing the endothelial cell motility in vitro and stimulate the corneal neovascularization in vivo. An oligomeric fraction of the protein was detected in the conditioned media as well as in human serum. The data obtained allowed us to conclude that mts1(S100A4) might induce tumor progression via stimulation of angiogenesis. Oncogene (2001) 20, 4685 ± 4695.

Mouse genome contains two major families of short interspersed repeats in more than 10(5) copies scattered throughout the whole genome. They are referred to as B1 and B2 sequences since they were first isolated from the genome library by means of a dsRNA-B probe /1/. In this work, two copies of the B2 family were sequenced and compared with the previously sequenced B1 repeat /2/. A B2 ubiquitous repeat is ca. 190 bp long. The members of the family deviate in 3-5% of nucleotides from the consensus sequence. B2 contains regions of homology to the RNA polymerase III split promoter and to 4.5S snRNA I. Both B1 and B2 contain regions which resemble junctions between exons and introns. In contrast to B1, B2 does not contain apparent homologies to papova viral replication origins and a human Alu sequence. One side of the B2 repeat is represented by a very AT-rich sequence (ca. 30 bp long) followed with an oligo (dA) stretch 10-15 nucleotides long. This region of the repeat is the most variable one. The whole unit is flanked with 15-16 bp direct repeats different in sequenced copies of B2. The same is true of some copies of the B1 family. The properties of B1 and B2 repeats suggest that they may represent a novel class of transposon-like elements in eukaryotic genome. A possible role of B-type repeats in genome reorganization, DNA replication and pre-mRNA processing is discussed.

Three copies of a highly repetitive DNA sequence B1 which is complementary to the most abundant class of mouse fold-back RNA have been cloned in pBR322 plasmid and sequenced by the method of Maxam and Gilbert. All the three have a length of about 130 base pairs and are very similar in their base sequence. The deviation from the average sequence is equal to 4% and the overall mismatch between each two is not higher than 8%. One of the recombinant clones used contained two copies of B1 oriented in the same direction. All of the B1 copies are flanked with sequences which possess nonidentical but very similar structure. They consist of a number of AmCn blocks (where m varies from 2 to 8 and n equals 1-2). These peculiar sequences in all cases are separated from B1 by non-homologous DNA stretches of 2-8 residues. In one case, a long polypurine stretch is located next to such a block. It consists of 74 residues most of which represent a reiteration of the basic sequence AAAAG. We have found two regions within the B1 sequence which are homologous to the intron-exon junctions, especially to those present in the large intron of the mouse beta-globin gene. It may indicate the involvement of the B1 sequence in pre-mRNA splicing.

Chromatin subunits ("nucleosomes") which were purified by sucrose gradient centrifugation of a staphylococcal nuclease digest of chromatin have been studied. We found that such a preparation contains nucleosomes of two discrete types which can be separated from each other by polyacrylamide gel electrophoresis. Nucleosome of the first type contains all five histones and a DNA segment of approximately 200 base pairs long, whereas nucleosome of the second type lacks histone H1 and its DNA segment is approximately 170 base pairs long, i.e., about 30 base pairs shorter than the DNA segment of the nucleosome of the first type. Purified dimer of the nucleosome also can be fractionated by gel electrophoresis into three discrete bands which correspond to dinucleosomes containing two molecules of histone H1, one and no H1. These and related findings strongly suggest that the H1 molecule is bound to a short (approximately 30 base pairs) terminal stretch of the nucleosomal DNA segment which can be removed by nuclease (possibly in the form of H1-DNA complex) without any significant disturbance of main structural features of the nucleosome.