A novel microscopy-based high-throughput screening method to identify proteins that regulate global histone modification levels.

Posttranslational modifications of histones play an important role in the regulation of gene expression and chromatin structure in eukaryotes. The balance between chromatin factors depositing (writers) and removing (erasers) histone marks regulates the steady-state levels of chromatin modifications. Here we describe a novel microscopy-based screening method to identify proteins that regulate histone modification levels in a high-throughput fashion. We named our method CROSS, for Chromatin Regulation Ontology SiRNA Screening. CROSS is based on an siRNA library targeting the expression of 529 proteins involved in chromatin regulation. As a proof of principle, we used CROSS to identify chromatin factors involved in histone H3 methylation on either lysine-4 or lysine-27. Furthermore, we show that CROSS can be used to identify chromatin factors that affect growth in cancer cell lines. Taken together, CROSS is a powerful method to identify the writers and erasers of novel and known chromatin marks and facilitates the identification of drugs targeting epigenetic modifications.

Mitogenic signaling of insulin-like growth factor I in MCF-7 human breast cancer cells requires phosphatidylinositol 3-kinase and is independent of mitogen-activated protein kinase.

Addition of insulin-like growth factor I (IGF-I) to quiescent breast tumor-derived MCF-7 cells causes stimulation of cyclin D1 synthesis, hyperphosphorylation of the retinoblastoma protein pRb, DNA synthesis, and cell division. All of these effects are independent of the mitogen-activated protein kinase (MAPK) pathway since none of them is blocked by PD098059, the specific inhibitor of the MAPK activating kinase MEK1. This observation is consistent with the finding that the phorbol ester 12-O-tetradecanoylphorbol-13-acetate (TPA), a strong inducer of MAPK activity in MCF-7 cells, effectively inhibits proliferation. The anti-proliferative effect of TPA in these cells may be accounted for, at least in part, by the MAPK-dependent stimulation of the synthesis of p21(WAF1/CIP1), an inhibitor of cyclin/cyclin-dependent kinase complexes. In contrast, all of the observed stimulatory effects of IGF-I on cell cycle progression, cyclin D1 synthesis, and pRb hyperphosphorylation were blocked by the specific phosphatidylinositol 3-kinase inhibitor LY294002, suggesting that phosphatidylinositol 3-kinase activity but not MAPK activity is required for transduction of the mitogenic IGF-I signal in MCF-7 cells.

The promoter of the salmon insulin-like growth factor I gene is activated by hepatocyte nuclear factor 1.

Hepatocyte nuclear factor 1 (HNF-1) was found to have a potent stimulatory action on the activity of the promoter of the salmon insulin-like growth factor I (IGF-I) gene in transient transfection experiments. This liver-enriched transcription factor was shown to bind to an element in the proximal region of the promoter with distinct nucleotide sequence homology to the HNF-1 consensus binding sequence. Mutating this sequence to a variant no longer capable of HNF-1 binding resulted in the loss of the stimulatory effect. Since the sequence of the HNF-1 binding site is conserved in all mammalian, avian, and amphibian species from which the IGF-I promoter sequences have been derived to date, we propose that HNF-1 may be an important regulator of IGF-I gene expression in all of these species.

Expression of the insulin-like growth factor I gene is stimulated by the liver-enriched transcription factors C/EBP alpha and LAP.

The expression of the human insulin-like growth factor I (hIGF-I) gene is regulated in a developmental stage- and tissue-specific manner. Postnatally, the liver becomes the main endocrine source of this important growth factor. The hIGF-I gene contains two alternatively used leader exons, exon 1 and exon 2. In human adult liver, exon 1 sequences are represented in about 80% of the transcripts. In this study we have investigated the role of promoter 1 (P1), located upstream of leader exon 1, in the tissue-specific expression of the IGF-I gene in human adult liver. Factors involved in this process have not been described to date. In this report we show, employing transient transfection experiments in Hep3B cells, that two liver-enriched transcription factors, CCAAT/enhancer binding protein alpha (C/EBP alpha) and liver-enriched activating protein (LAP), enhance the activity of IGF-I P1. DNase I footprinting experiments demonstrate that a C/EBP-LAP binding site is located 119 base pairs upstream of the major transcription start site in exon 1. Comparison with other C/EBP-LAP binding sites reveals that the binding site in P1 is a high affinity binding site. Mutations of the C/EBP-LAP binding site completely abolished the enhancing effect of C/EBP alpha and LAP, indicating that their activating signal is indeed conferred by this binding site. These results suggest that both C/EBP alpha and LAP play important roles in the liver-specific expression of the hIGF-I gene and provide the first clues in the elucidation of its complicated developmental stage- and tissue-specific expression pattern.

Functional analysis of the human IGF-I gene promoters.

The two putative promoter regions of the human insulin-like growth factor-I (IGF-I) gene, located upstream of the two alternatively used leader exons 1 and 2, were tested for promoter activity in transient transfection assays. Both regions were shown to possess promoter activity. The results further indicate cell-type-specific regulation of the activities of the two promoters in endogenously IGF-I producing human cell lines of different origin.

The human IGF-I gene contains two cell type-specifically regulated promoters.

The human insulin-like growth factor-I (IGF-I) gene contains two alternative leader exons: exons 1 and 2. We have identified, by transient transfection experiments, the putative promoters P1 and P2 upstream of these leader exons. The promoter regions were cloned in front of the luciferase reporter gene and their promoter activities were measured in transfected SK-N-MC (human neuroepithelioma) and OVCAR-3 (human ovarian carcinoma) cells. Both of these cell lines express the IGF-I gene endogenously, resulting in normally sized IGF-I mRNAs of 7.6, 1.3 and 1.1 kb. In SK-N-MC cells, in which P1 is the most active IGF-I promoter, P2 displayed a three times lower promoter activity than P1. However, in OVCAR-3 cells, P2 is four times more active than P1, resulting in an overall 12-fold difference in the relative promoter activities of the two IGF-I gene promoters in these two cell types. This indicates that the IGF-I promoters show a cell type-specific expression pattern.

Initial characterization of the four promoters of the human insulin-like growth factor II gene.

The human insulin-like growth factor II (IGF-II) gene contains four promoters (P1-P4), which are expressed in a tissue-specific and development-dependent way. Analysis of IGF-II mRNAs in different tissues has revealed that promoters P3 and P4 are expressed in all fetal and in nonhepatic adult tissues. In adult liver, however, the promoters P2, P3 and P4 are completely shut off and another promoter, P1, is activated. To obtain more insight in the mechanisms involved in the regulation of IGF-II gene expression we have performed an initial characterization of the IGF-II promoters employing transient expression of IGF-II promoter constructs in Hep3B and HeLa cells. These studies have revealed that promoters P1, P3 and P4 are active in both cell lines tested, while no activity of promoter P2 could be detected. Employing gel retardation and DNaseI footprint analysis we have identified in the three IGF-II promoters a number of elements which are bound by nuclear proteins.

Differential expression of the human insulin-like growth factor II gene. Characterization of the IGF-II mRNAs and an mRNA encoding a putative IGF-II-associated protein.

Insulin-like growth factor II (IGF-II) is a polypeptide of 67 amino acids which is thought to play an important role in fetal growth and development. The human IGF-II gene is situated on chromosome 11, very close to the insulin gene. It extends over 30 kb of chromosomal DNA and consists of five noncoding exons (exons 1-4 and 4B) followed by three protein encoding exons (exons 5-7), one of which (exon 7) contains a long 3'-untranslated region. Here we show that differential initiation of transcription can occur at three distinct promoter sites, resulting in the appearance of mRNA species of different lengths. These promoters show a tissue-specific and a development-specific regulation of expression. Furthermore, we have determined the entire nucleotide sequence of the 3'-terminal exon, exon 7, which is about 4 kb long and contains 3.8 kb of 3'-untranslated sequences. This completes the elucidation of the human IGF-II gene structure. Surprisingly, Northern blot analysis of fetal and adult RNA with a probe derived from the 3'-nontranslated region of exon 7 detects a novel 1.8 kb mRNA which appears to be coordinately expressed with the IGF-II mRNAs. In vitro translation of this 1.8 kb mRNA results in the formation of a translation product of 8.3 kDa, which compares well with the size of a predicted translation product from a 252-nucleotides-long open reading frame.

The human insulin-like growth factor II gene contains two development-specific promoters.

The insulin-like growth factors (IGF) play an important role in fetal and postnatal development. Recently, the nucleotide sequences of the cDNAs encoding IGF-I and IGF-II and part of the human IGF genes were reported. In this communication we describe two distinct IGF-II cDNAs isolated from a human adult liver and a human hepatoma cDNA library, respectively. Using these two cDNAs, we have established that the human IGF-II gene contains at least 7 exons. Two different IGF-II promoters have been identified, 19 kilobases (kb) apart, which are active in a development-specific manner. The promoter, active in the adult stage, is located only 1.4 kb downstream from the insulin gene.

Organization of the human genes for insulin-like growth factors I and II.

Recently, we have reported the isolation of cDNAs encoding the precursors of insulin-like growth factors I and II (IGF-I and II) [(1983) Nature 306, 609-611; (1985) FEBS Lett. 179, 243-246. These cDNAs were employed as specific probes to detect and isolate the corresponding genes from human cosmid DNA libraries. Three cosmids were detected, together containing the entire cDNA sequence of IGF-I, and one cosmid containing the sequence of IGF-II cDNA. Southern blot hybridization, physical mapping and nucleotide sequence analysis of these cosmids revealed that the IGF-I and -II genes have a discontinous structure. The IGF-I gene contains at least four exons spanning a region of probably more that 45 kilobasepairs (kb), while the IGF-II gene consists of at least five exons, spanning a region of 16 kb.

Molecular aspects of the human somatomedins.

The primary structure of the somatomedins (SM) IGF-I and IGF-II has been known for some years. Both from isolation of the SMs from plasma and from the study of cDNAs it has become evident that there are more SMs, needing further study. In plasma the SMs are present in larger molecules of two size-classes: 150 kD and 40 kD, behaving as specific SM-binding proteins. The possibility that such protein classes are heterogeneous and also contain precursor molecules undergoing proteolytic processing is considered. The somatomedin genes are discontinuous and span large regions of the human genome.

Nucleotide sequences of cDNAs encoding precursors of human insulin-like growth factor II (IGF-II) and an IGF-II variant.

We have isolated 3 cDNA clones encoding human IGF-II and a variant of IGF-II. The amino acid sequence encoded by the IGF-II cDNA is identical to the sequence previously described [(1978) FEBS Lett. 89, 283-286]. In the amino acid sequence predicted by the IGF-II variant cDNA, the Ser residue 29 in the B-domain has been replaced by an Arg-Leu-Pro-Gly sequence. The corresponding mRNAs probably arise by alternative splicing of a common RNA precursor. The IGF coding region of the cDNA inserts is flanked by sequences encoding a signal peptide and a carboxy-terminal peptide indicating that both human IGF-II and its variant are synthesized as precursors.

Sequence of cDNA encoding human insulin-like growth factor I precursor.

Somatomedins (SM) or insulin-like growth factors (IGF) constitute a heterogeneous group of peptides with important growth-promoting effects in vitro as well as in vivo. Amino acid sequences have been determined for only two of them, IGF-I and IGF-II, which are highly homologous. IGF-I, which is identical with SM-C, is composed of 70 amino acid residues and IGF-II contains 73 amino acids and may be identical with SM-A. Other peptides with different charge properties but with similar SM-like or insulin-like behaviour in biological and receptor assays, have been described but have not yet been fully characterized. The liver is known to be a major site of production of these peptides, but many other tissues--especially in the fetus--may synthesize them as well. We report here the nucleotide sequence of a human liver cDNA encoding the complete amino acid sequence of IGF-I. The IGF-I coding region is flanked by sequences encoding an amino-terminal peptide of at least 25 amino acid residues and a carboxyl-terminal peptide of 35 amino acids. This provides evidence that IGF-I is synthesized as a precursor protein and that formation of IGF-I from this precursor requires proteolytic processing at both ends.

Structure and function of adenovirus DNA binding protein: comparison of the amino acid sequences of the Ad5 and Ad12 proteins derived from the nucleotide sequence of the corresponding genes.

The adenoviral DNA binding protein (DBP) is a multifunctional protein involved in DNA replication and gene expression. In order to investigate the relation between structure and function of DBP, the amino acid sequences of the serotypes 5 and 12 (Ad5 and Ad12) have been compared. The amino acid sequence of Ad5 DBP was previously established by nucleotide sequence analysis of the Ad5 DBP gene (W. Kruijer, F. M. A. Van Schaik, and J. S. Sussenbach, Nucl. Acids Res. 9, 4439-4457, 1981). In this study the analysis of the Ad5 DBP gene and adjacent regions by determination of the sequence of the first leader in late DBP mRNA's and the splice point between the tripartite leader and the main body of the mRNA encoding the 100-kDa protein has been extended. The nucleotide sequence of the Ad12 DBP gene is also described. From the nucleotide sequence and RNA mapping data of Ad12 DBP mRNA's (I. Saito, J. Sato H. Handa, K. Shiraki, and H. Shimojo, Virology 114, 379-398, 1981) the complete Ad12 DBP amino acid sequence could be deduced. Ad12 DBP contains 484 amino acids and has an actual Mr of 54,992. It is 45 amino acids shorter than Ad5 DBP. Comparison of the Ad12 and Ad5 DBP amino acid sequences shows that several longer deletions are present in the N-terminal 125 amino acid residues of Ad12 DBP. In contrast, only a single amino acid deletion and insertion is found in the C-terminal 359 amino acids of Ad12 DBP. The N- and C-terminal domains of Ad12 and Ad5 DBP are 45 and 80% homologous, respectively. This suggests that both domains of DBP are subjected to different evolutionary pressures. Analysis of various Ad5 mutants with an altered DBP gene, has indicated that the C-terminal domain is involved in DNA replication and early gene expression, while the N-terminal domain has a role in late gene expression in monkey cells. These results are discussed in relation to the structure and function of adenovirus DBP.

Structure and function of DNA binding proteins from revertants of adenovirus type 5 mutants with a temperature-sensitive DNA replication.

H5ts107 and H5ts125 are two adenoviruses type 5 (Ad5) mutants with a temperature-sensitive DNA replication. Both mutants contain an altered gene encoding the DNA binding protein (DBP). We have established by nucleotide sequence analysis that both mutants carry exactly the same mutation in the DBP gene resulting in the substitution of a proline residue at position 413 in the wild-type DBP amino acid sequence (529 amino acid residues long) by a serine residue. Revertants of H5ts107 and H5ts125, which are temperature independent in plaque efficiency and growth in HeLa cells at 32 degrees and 39 degrees, were characterized by nucleotide sequence analysis of their DBP genes. Four types of revertants could be distinguished: revertants with the wild-type DBP amino acid sequence (type I) and, revertants carrying, in addition to the original H5ts107/H5ts125 mutation at position 413, intragenic second site mutations at position 508 histidine leads to tyrosine (type II), at position 352 glycine leads to aspartic acid (type III), and at position 347 alanine leads to proline (type IV), respectively. All intragenic second site mutations are located, together with the H5ts107/H5ts125 mutation, in the C-terminal 45-kD fragment of the adenovirus DBP molecule. This provides further evidence that this part of the DBP molecule plays an important role in viral DNA replication. Phenotypic characterization of the revertants (J.C. Nicolas, F. Suarez, A.J. Levine, and M. Girard (1981), Virology 108, 521-524; (J.C. Nicolas, D. Ingrand, P. Sarnow, and A.J. Levine (1982), Virology 122, 481-485) has shown that the second site mutations reveal additional functional domains in the DBP molecule.

Nucleotide sequence of the gene encoding adenovirus type 2 DNA binding protein.

We have determined the nucleotide sequence of the gene encoding adenovirus type 2 (Ad2) DNA binding protein (DBP). From the nucleotide sequence the complete amino acid sequence of Ad2 DBP has been deduced. A comparison of the amino acid sequences of Ad2 and Ad5 DBP, both 529 residues long, reveals that the C-terminal 354 residues of both sequences are identical. Within the N-terminal 175 amino acid residues Ad2 and Ad5 show nine differences. The site of mutation in Ad2 ND1ts23, a mutant with a temperature-sensitive DNA replication, was mapped at the nucleotide level. A single nucleotide alteration in the DBP gene, resulting in a leucine leads to phenylalanine substitution at position 282 in the amino acid sequence is responsible for the temperature-sensitive character of this mutant. Previously, we localized the mutation of another DBP mutant with a temperature-sensitive DNA replication (H5ts125) at position 413 in the amino acid sequence of the DBP molecule (Nucleic Acids Res. 9 (1981) 4439-4457). These mapping data are discussed in relation to the structure and function of the DBP molecule.

Structure and organization of the gene coding for the DNA binding protein of adenovirus type 5.

We have determined the nucleotide sequence of a region of adenovirus type 5 (Ad5) DNA located between map positions 61.7 and 71.4, which covers the gene form the 72 kD DNA binding protein (DBP) and the sequence encoding the amino-terminal part of the 100 kD protein. Sequence analysis of cDNA copies of DBP mRNA revealed the existence of two abundant species of spliced mRNA molecules. One species consists of two short leader sequences from positions 75.2 (67 and 68 nucleotides long) and 68.8 (77 nucleotides long), respectively, and the main body of the RNA molecules. The other species contains only the leader sequence from position 75.2 and the main body. The amino acid sequence of DBP is encoded entirely by a long open reading frame of 1587 nucleotides in the main body of DBP mRNA. From the nucleotide sequence of the DBP gene it can be derived that DBP contains 529 amino acid residues and has an actual molecular weight of 59,049 daltons. The sites of mutation in the mutants H5hr404 and H5ts125 were determined at the nucleotide level. Single nucleotide alterations were detected in H5hr404 and H5ts125 in the sequences corresponding to the amino-terminal part and the carboxy-terminal part of DBP, respectively. The implications of these mutations are discussed.

Nucleotide sequence analysis of a region of adenovirus 5 DNA encoding a hitherto unidentified gene.

The nucleotide sequence of a region located on the adenovirus type 5 genome between coordinates 59.9 and 62.5 has been determined. This region comprises the 3'-termini of mRNAs coding for the hexon and the DNA binding protein. The sequence reveals the presence of an open reading frame in r-strand transcripts downstream from the hexon termination codon. The protein predicted from this region has a MW of 23,068 daltons. Based on the characteristics of the mutant Ad2ts1, which maps between coordinates 59.9 and 69.9, it is probable that the predicted protein is involved in processing of late proteins.