An AML-1 consensus sequence binds an osteoblast-specific complex and transcriptionally activates the osteocalcin gene.

Tissue and cell-type specific expression of the rat osteocalcin (rOC) gene involves the interplay of multiple transcriptional regulatory factors. In this report we demonstrate that AML-1 (acute myeloid leukemia-1), a DNA-binding protein whose genes are disrupted by chromosomal translocations in several human leukemias, interacts with a sequence essential for enhancing tissue-restricted expression of the rOC gene. Deletion analysis of rOC promoter-chloramphenicol acetyltransferase constructs demonstrates that an AML-1-binding sequence within the proximal promoter (-138 to -130 nt) contributes to 75% of the level of osteocalcin gene expression. The activation potential of the AML-1-binding sequence has been established by overexpressing AML-1 in osteoblastic as well as in nonosseous cell lines. Overexpression not only enhances rOC promoter activity in osteoblasts but also mediates OC promoter activity in a nonosseous human fibroblastic cell line. A probe containing this site forms a sequence specific protein-DNA complex with nuclear extracts from osteoblastic cells but not from nonosseous cells. Antisera supershift experiments indicate the presence of AML-1 and its partner protein core-binding factor beta in this osteoblast-restricted complex. Mutations of the critical AML-1-binding nucleotides abrogate formation of the complex and strongly diminish promoter activity. These results indicate that an AML-1 related protein is functional in cells of the osteoblastic lineage and that the AML-1-binding site is a regulatory element important for osteoblast-specific transcriptional activation of the rOC gene.

Identification of the promoter of the mouse obese gene.

Primer extension and RACE (rapid amplification of cDNA ends) assays were used to identify and sequence the 5' terminus of mouse ob mRNA. This sequence was used to obtain a recombinant bacteriophage containing the first exon of the encoding gene. DNA sequence analysis of the region immediately upstream of the first exon of the mouse ob gene revealed DNA sequences corresponding to presumptive cis-regulatory elements. A canonical TATA box was observed 30-34 base pairs upstream from the start site of transcription and a putative binding site for members of the C/EBP family of transcription factors was identified immediately upstream from the TATA box. Nuclear extracts prepared from primary adipocytes contained a DNA binding activity capable of avid and specific interaction with the putative C/EBP response element; antibodies to C/EBP alpha neutralized the DNA binding activity present in adipocyte nuclear extracts. When linked to a firefly luciferase reporter and transfected into primary adipocytes, the presumptive promoter of the mouse ob gene facilitated luciferase expression. When transfected into HepG2 cells, which lack C/EBP alpha, the mouse ob promoter was only weakly active. Supplementation of C/EBP alpha by cotransfection with a C/EBP alpha expression vector markedly stimulated luciferase expression. Finally, an ob promoter variant mutated at the C/EBP response element was inactive in both primary adipocytes and HepG2 cells. These observations provide evidence for identification of a functional promoter capable of directing expression of the mouse ob gene.

Pax3 modulates expression of the c-Met receptor during limb muscle development.

Pax3 is a transcription factor whose expression has been used as a marker of myogenic precursor cells arising in the lateral somite destined to migrate to and populate the limb musculature. Accruing evidence indicates that the embryologic origins of axial and appendicular muscles are distinct, and limb muscle abnormalities in both mice and humans harboring Pax3 mutations support this distinction. The mechanisms by which Pax3 affects limb muscle development are unknown. The tyrosine kinase receptor for hepatocyte growth factor/scatter factor encoded by the c-met protooncogene is also expressed in limb muscle progenitors and, like Pax-3, is required in the mouse for limb muscle development. Here, we show that c-met expression is markedly reduced in the lateral dermomyotome of Splotch embryos lacking Pax3. We show that Pax3 can stimulate c-met expression in cultured cells, and we identify a potential Pax3 binding site in the human c-MET promoter that may contribute to direct transcriptional regulation. In addition, we have found that several cell lines derived from patients with rhabdomyosarcomas caused by a t(2;13) chromosomal translocation activating PAX3 express c-MET, whereas those rhabdomyosarcoma cell lines examined without the translocation do not. These results are consistent with a model in which Pax3 modulates c-met expression in the lateral dermomyotome, a function that is required for the appropriate migration of these myogenic precursors to the limb where the ligand for c-met (hepatocyte growth factor/scatter factor) is expressed at high levels.

The neck region of the myosin motor domain acts as a lever arm to generate movement.

The myosin head consists of a globular catalytic domain that binds actin and hydrolyzes ATP and a neck domain that consists of essential and regulatory light chains bound to a long alpha-helical portion of the heavy chain. The swinging neck-level model assumes that a swinging motion of the neck relative to the catalytic domain is the origin of movement. This model predicts that the step size, and consequently the sliding velocity, are linearly related to the length of the neck. We have tested this point by characterizing a series of mutant Dictyostelium myosins that have different neck lengths. The 2xELCBS mutant has an extra binding site for essential light chain. The delta RLCBS mutant myosin has an internal deletion that removes the regulatory light chain binding site. The delta BLCBS mutant lacks both light chain binding sites. Wild-type myosin and these mutant myosins were subjected to the sliding filament in vitro motility assay. As expected, mutants with shorter necks move slower than wild-type myosin in vitro. Most significantly, a mutant with a longer neck moves faster than the wild type, and the sliding velocities of these myosins are linearly related to the neck length, as predicted by the swinging neck-lever model. A simple extrapolation to zero speed predicts that the fulcrum point is in the vicinity of the SH1-SH2 region in the catalytic domain.

Crystal structure of the T state of allosteric yeast chorismate mutase and comparison with the R state.

The crystal structure of the tyrosine-bound T state of allosteric yeast Saccharomyces cerevisiae chorismate mutase was solved by molecular replacement at a resolution of 2.8 angstroms using a monomer of the R-state structure as the search model. The allosteric inhibitor tyrosine was found to bind in the T state at the same binding site as the allosteric activator tryptophan binds in the R state, thus defining one regulatory binding site for each monomer. Activation by tryptophan is caused by the larger steric size of its side chain, thereby pushing apart the allosteric domain of one monomer and helix H8 of the catalytic domain of the other monomer. Inhibition is caused by polar contacts of tyrosine with Arg-75 and Arg-76 of one monomer and with Gly-141, Ser-142, and Thr-145 of the other monomer, thereby bringing the allosteric and catalytic domains closer together. The allosteric transition includes an 8 degree rotation of each of the two catalytic domains relative to the allosteric domains of each monomer (domain closure). Alternatively, this transition can be described as a 15 degree rotation of the catalytic domains of the dimer relative to each other.

Creation of drug-specific herpes simplex virus type 1 thymidine kinase mutants for gene therapy.

Herpes simplex virus type 1 (HSV-1) thymidine kinase is currently used as a suicide agent in the gene therapy of cancer. This therapy is based on the preferential phosphorylation of nucleoside analogs by tumor cells expressing HSV-1 thymidine kinase. However, the use of HSV-1 thymidine kinase is limited in part by the toxicity of the nucleoside analogs. We have used random sequence mutagenesis to create new HSV-1 thymidine kinases that, compared with wild-type thymidine kinase, render cells much more sensitive to specific nucleoside analogs. A segment of the HSV-1 thymidine kinase gene at the putative nucleoside binding site was substituted with random nucleotide sequences. Mutant enzymes that demonstrate preferential phosphorylation of the nucleoside analogs, ganciclovir or acyclovir, were selected from more than one million Escherichia coli transformants. Among the 426 active mutants we have isolated, 26 demonstrated enhanced sensitivity to ganciclovir, and 54 were more sensitive to acyclovir. Only 6 mutant enzymes displayed sensitivity to both ganciclovir and acyclovir when expressed in E. coli. Analysis of 3 drug-sensitive enzymes demonstrated that 1 produced stable mammalian cell transfectants that are 43-fold more sensitive to ganciclovir and 20-fold more sensitive to acyclovir.

Recognition of diverse sequences by class I zinc fingers: asymmetries and indirect effects on specificity in the interaction between CF2II and A+T-rich elements.

The Drosophila CF2II protein, which contains zinc fingers of the Cys2His2 type and recognizes an A+T-rich sequence, behaves in cell culture as an activator of a reporter chloramphenicol acetyltransferase gene. This activity depends on C-terminal but not N-terminal zinc fingers, as does in vitro DNA binding. By site-specific mutagenesis and binding site selection, we define the critical amino acid-base interactions. Mutations of single amino acid residues at the leading edge of the recognition helix are rarely neutral: many result in a slight change in affinity for the ideal DNA target site; some cause major loss of affinity; and others change specificity for as many as two bases in the target site. Compared to zinc fingers that recognize G+C-rich DNA, CF2II fingers appear to bind to A+T-rich DNA in a generally similar manner, but with additional flexibility and amino acid-base interactions. The results illustrate how zinc fingers may be evolving to recognize an unusually diverse set of DNA sequences.

The chimeric genes AML1/MDS1 and AML1/EAP inhibit AML1B activation at the CSF1R promoter, but only AML1/MDS1 has tumor-promoter properties.

The (3;21)(q26;q22) translocation associated with treatment-related myelodysplastic syndrome, treatment-related acute myeloid leukemia, and blast crisis of chronic myeloid leukemia results in the expression of the chimeric genes AML1/EAP, AML1/MDS1, and AML1/EVI1. AML1 (CBFA2), which codes for the alpha subunit of the heterodimeric transcription factor CBF, is also involved in the t(8;21), and the gene coding for the beta subunit (CBFB) is involved in the inv(16). These are two of the most common recurring chromosomal rearrangements in acute myeloid leukemia. CBF corresponds to the murine Pebp2 factor, and CBF binding sites are found in a number of eukaryotic and viral enhancers and promoters. We studied the effects of AML1/EAP and AML1/MDS1 at the AML1 binding site of the CSF1R (macrophage-colony-stimulating factor receptor gene) promoter by using reporter gene assays, and we analyzed the consequences of the expression of both chimeric proteins in an embryonic rat fibroblast cell line (Rat1A) in culture and after injection into athymic nude mice. Unlike AML1, which is an activator of the CSF1R promoter, the chimeric proteins did not transactivate the CSF1R promoter site but acted as inhibitors of AML1 (CBFA2). AML1/EAP and AML1/MDS1 expressed in adherent Rat1A cells decreased contact inhibition of growth, and expression of AML1/MDS1 was associated with acquisition of the ability to grow in suspension culture. Expression of AML1/MDS1 increased the tumorigenicity of Rat1A cells injected into athymic nude mice, whereas AML1/EAP expression prevented tumor growth. These results suggest that expression of AML1/EAP and AML1/MDS1 can interfere with normal AML1 function, and that AML1/MDS1 has tumor-promoting properties in an embryonic rat fibroblast cell line.

Transcriptional activation by protein-induced DNA bending: evidence for a DNA structural transmission model.

Integration host factor (IHF) is a DNA-bending protein that binds to an upstream activating sequence (UAS1) and, on a negatively supercoiled DNA template, activates transcription from the ilvPG promoter of the ilvG-MEDA operon of Escherichia coli. The transcriptional initiation site of the ilvGMEDA operon is located 92 bp downstream of UAS1. Activation is still observed when the orientation of the upstream IHF binding site is reversed. This manipulation places the IHF binding site on the opposite face of the DNA helix, directs the IHF-induced DNA bend in the opposite direction, and presents the opposite face of the nonsymmetrical, heterodimeric, IHF molecule to the downstream RNA polymerase. Lymphoid enhancer-binding factor, LEF-1, is a DNA-bending, lymphoid-specific, mammalian transcription factor that shares no amino acid sequence similarity with IHF. When the IHF site in UAS1 is replaced with a LEF-1 site, LEF-1 activates transcription from the downstream ilvPG promoter in E. coli as well as it is activated by its natural activator, IHF. These results suggest that specific interactions between IHF and RNA polymerase are not required for activation. The results of DNA structural studies show that IHF forms a protein-DNA complex in the UAS1 region that, in the absence of RNA polymerase, alters the structure of the DNA helix in the -10 hexanucleotide region of the downstream ilvPG promoter. The results of in vitro abortive transcription assays show that IIIF also increases the apparent rate of RNA polymerase isomerization from a closed to an open complex. We suggest, therefore, that IHF activates transcription by forming a higher-order protein-DNA complex in the UAS1 region that structurally alters the DNA helix in a way that facilitates open complex formation at the downstream ilvPG promoter site.

Crystallographic analysis of endogenous peptides associated with HLA-DR1 suggests a common, polyproline II-like conformation for bound peptides.

The structure of the human major histocompatibility complex (MHC) class II molecule HLA-DR1 derived from the human lymphoblastoid cell line LG-2 has been determined in a complex with the Staphylococcus aureus enterotoxin B superantigen. The HLA-DR1 molecule contains a mixture of endogenous peptides derived from cellular or serum proteins bound in the antigen-binding site, which copurify with the class II molecule. Continuous electron density for 13 amino acid residues is observed in the MHC peptide-binding site, suggesting that this is the core length of peptide that forms common interactions with the MHC molecule. Electron density is also observed for side chains of the endogenous peptides. The electron density corresponding to peptide side chains that interact with the DR1-binding site is more clearly defined than the electron density that extends out of the binding site. The regions of the endogenous peptides that interact with DRI are therefore either more restricted in conformation or sequence than the peptide side chains or amino acids that project out of the peptide-binding site. The hydrogen-bond interactions and conformation of a peptide model built into the electron density are similar to other HLA-DR-peptide structures. The bound peptides assume a regular conformation that is similar to a polyproline type II helix. The side-chain pockets and conserved asparagine residues of the DR1 molecule are well-positioned to interact with peptides in the polyproline type II conformation and may restrict the range of acceptable peptide conformations.

A combined kinetic and modeling study of the catalytic center subsites of human angiogenin.

Kinetic analysis and molecular modeling have been used to map the ribonucleolytic center of angiogenin (Ang). Pyrimidine nucleotides were found to interact very weakly with Ang, consistent with the inaccessible B1 pyrimidine binding site revealed by x-ray crystallography. Ang also lacks an effective phosphate binding site on the 5' side of B1. Although the B2 site that preferentially binds purines on the 3' side of B1 is also weak, its associated phosphate subsites make substantial contributions: both 3',5'-ADP and 5'-ADP have Ki values 6-fold lower than for 5'-AMP, and adding a 3'-phosphate to the substrate CpA increases Kcat/Km by 9-fold. Thus Ang has a functional P2 site on the 3' side of B2 and a site for a second phosphate on the 5' side of B2. Modeling of an Ang-d(ApTpApA) complex suggested that Arg-5 forms part of the P2 site and that a 2'-phosphate might bind more tightly than a 3'-phosphate. Both predictions were confirmed kinetically. The subsite map obtained by this combined approach indicated that 5'-diphosphoadenosine 2'-phosphate might be a more potent inhibitor than any of the nucleotides tested thus far. Indeed, its Ki value of 150 microM is 50-fold lower than that for the best nucleotide previously reported and 400-fold lower than the Km for the best dinucleotide substrate. This compound may serve as a suitable starting point for the eventual design of tight-binding inhibitors of Ang as antiangiogenic agents for human therapy.

Transcriptional activation of the human proliferating-cell nuclear antigen promoter by p53.

Proliferating-cell nuclear antigen (PCNA) is a DNA damage-inducible protein that performs an essential function in DNA replication and repair as an auxiliary factor for DNA polymerases delta and epsilon. Examination of the human PCNA promoter DNA sequence revealed a site with homology to the consensus DNA sequence bound by p53. PCNA promoter fragments with this site intact bound p53 in vitro and were transcriptionally activated by wild-type p53 in transient expression assays in SAOS-2 cells. The resident p53-binding site could be functionally substituted by a previously described p53-binding site from the ribosomal gene cluster. A plasmid expressing a mutated version of p53 derived from a patient with Li-Fraumeni syndrome failed to activate the PCNA promoter in the cotransfection assay. In different cell types, activation of the PCNA promoter by the p53-binding sequence correlated with the status of p53. Activation of the PCNA promoter by wild-type p53 depends upon the level of p53 expression. This concentration dependence and cell type specificity reconciles the observations presented here with prior results indicating that wild-type p53 represses the PCNA promoter. These findings provide a mechanism whereby p53 modulates activation of PCNA expression as a cellular response to DNA damage.

Protein-protein interactions in the rigor actomyosin complex.

Since it has not been possible to crystallize the actomyosin complex, the x-ray structures of the individual proteins together with data obtained by fiber diffraction and electron microscopy have been used to build detailed models of filamentous actin (f-actin) and the actomyosin rigor complex. In the f-actin model, a single monomer uses 10 surface loops and two alpha-helices to make sometimes complicated interactions with its four neighbors. In the myosin molecule, both the essential and regulatory light chains show considerable structural homology to calmodulin. General principles are evident in their mode of attachment to the target alpha-helix of the myosin heavy chain. The essential light chain also makes contacts with other parts of the heavy chain and with the regulatory light chain. The actomyosin rigor interface is extensive, involving interaction of a single myosin head with regions on two adjacent actin monomers. A number of hydrophobic residues on the apposing faces of actin and myosin contribute to the main binding site. This site is flanked on three sides by charged myosin surface loops that form predominantly ionic interactions with adjacent regions of actin. Hydrogen bonding is likely to play a significant role in actin-actin and actin-myosin interactions since many of the contacts involve loops. The model building approach used with actomyosin is applicable to other multicomponent assemblies of biological interest and is a powerful method for revealing molecular interactions and providing insights into the mode of action of the assemblies.

X-ray structure of a vanadium-containing enzyme: chloroperoxidase from the fungus Curvularia inaequalis.

The chloroperoxidase (EC 1.11.1.-) from the fungus Curvularia inaequalis belongs to a class of vanadium enzymes that oxidize halides in the presence of hydrogen peroxide to the corresponding hypohalous acids. The 2.1 A crystal structure (R = 20%) of an azide chloroperoxidase complex reveals the geometry of the catalytic vanadium center. Azide coordinates directly to the metal center, resulting in a structure with azide, three nonprotein oxygens, and a histidine as ligands. In the native state vanadium will be bound as hydrogen vanadate(V) in a trigonal bipyramidal coordination with the metal coordinated to three oxygens in the equatorial plane, to the OH group at one apical position, and to the epsilon 2 nitrogen of a histidine at the other apical position. The protein fold is mainly alpha-helical with two four-helix bundles as main structural motifs and an overall structure different from other structures. The helices pack together to a compact molecule, which explains the high stability of the protein. An amino acid sequence comparison with vanadium-containing bromoperoxidase from the seaweed Ascophyllum nodosum shows high similarities in the regions of the metal binding site, with all hydrogen vanadate(V) interacting residues conserved except for lysine-353, which is an asparagine.

Contrasting histories of avian and mammalian Mhc genes revealed by class II B sequences from songbirds.

To explore the evolutionary dynamics of genes in the major histocompatibility complex (Mhc) in nonmammalian vertebrates, we have amplified complete sequences of the polymorphic second (beta1) and third (beta2) exons of class II beta chain genes of songbirds. The pattern of nucleotide substitution in the antigen-binding site of sequences cloned from three behaviorally and phylogenetically divergent songbirds [scrub jays Aphelocoma coerulescens), red-winged blackbirds (Agelaius phoeniceus), and house finches (Carpodacus mexicanus) reveals that class II B genes of songbirds are subject to the same types of diversifying forces as those observed at mammalian class II loci. By contrast, the tree of avian class II B genes reveals that orthologous relationships have not been retained as in placental mammals and that, unlike class II genes in mammals, genes in songbirds and chickens have had very recent common ancestors within their respective groups. Thus, whereas the selective forces diversifying class II B genes of birds are likely similar to those in mammals, their long-term evolutionary dynamics appear to be characterized by much higher rates of concerted evolution.