The structure of the Antennapedia homeodomain determined by NMR spectroscopy in solution: comparison with prokaryotic repressors

Crystal structure of an engrailed homeodomain-DNA complex at 2.8 A resolution: a framework for understanding homeodomain-DNA interactions

The crystal structure of a complex containing the engrailed homeodomain and a duplex DNA site has been determined at 2.8 A resolution and refined to a crystallographic R factor of 24.4%. In this complex, two separate regions of the 61 amino acid polypeptide contact a TAAT subsite. An N-terminal arm fits into the minor groove, and the side chains of Arg-3 and Arg-5 make contacts near the 5' end of this "core consensus" binding site. An alpha helix fits into the major groove, and the side chains of IIe-47 and Asn-51 contact base pairs near the 3' end of the TAAT site. This "recognition helix" is part of a structurally conserved helix-turn-helix unit, but these helices are longer than the corresponding helices in the lambda repressor, and the relationship between the helix-turn-helix unit and the DNA is significantly different.

Oct-6: a POU transcription factor expressed in embryonal stem cells and in the developing brain

A family of octamer binding proteins is expressed during mouse development. Oct-4 and Oct-6 have been identified as two octamer binding proteins present in embryonal stem cells. Here we report the complementary DNA cloning and characterization of the mouse Oct-6 gene. The protein of 448 amino acids contains a glycine/alanine-rich amino terminal region, a histidine-rich sequence with homology to a region of kininogen associated with clotting, a POU domain and a short proline/histidine-rich carboxy terminal region. Expression of Oct-6 in HeLa cells is sufficient for transcriptional activation from the octamer motif, identifying Oct-6 as a transcription factor. The Oct-6 expression is downregulated upon embryonic stem cell differentiation increasing again during brain development. Expression in brain is present in certain areas of telencephalon, mesencephalon and brain stem with abundant expression in the cortex anlagen and in the developing colliculi. Thus Oct-6 is a new octamer binding transcription factor specifically regulated during mouse development.

Homeobox containing genes in the nematode Caenorhabditis elegans

We designed a unique 36-mer oligonucleotide probe, based on the most highly conserved amino acid sequences of Antennapedia-like homeodomains and the codon bias of Caenorhabditis elegans. This probe was then used to isolate four classes of genes from a C. elegans genomic library. Sequencing reveals that we have isolated three new homeobox genes, designated ceh-1, ceh-9 and ceh-10. The fourth homeobox gene, ceh-11, has recently been described by Schaller et al (Nucleic Acids Res. 18, 2033-2036). The amino acid sequence of ceh-1 is 87% similar to the honeybee H40 homeodomain, 85% similar to the Drosophila NK-1 homeodomain and 82% similar to the chicken CHox3 homeodomain. The sequence ceh-10 appears to be a member of the paired class of homeodomains. The other two sequences, ceh-9 and ceh-11, remain unclassified. Three of the four sequences have at least one intron within the homeobox region. Transcripts of ceh-10 and ceh-11 are present in embryonic RNA but are greatly diminished in later developmental stages. Three of the four new genes have been placed on the C. elegans genomic map.

A distal dimerization domain is essential for DNA-binding by the atypical HNF1 homeodomain

Hepatic Nuclear Factor 1 (HNF1, also referred to as LFB1, HP1 or APF) is a liver-specific transcription factor required for the expression of many hepatocyte specific genes. We report here the purification of this rat liver nuclear protein and the cloning of its cDNA using a PCR-derived approach. Seven independent clones reveal 3 alternative polyadenylation sites and a unique open reading frame. Both a motif homologous to the homeodomain and a distal dimerization domain are required for specific DNA binding. Sequence comparisons reveal several atypical features at key positions in the segment corresponding to helices III and IV of the Antaennapedia homeodomain as well as a potential 24 amino acid loop in place of the universal turn between helices II and III. Together with its property to dimerize in the presence or absence of DNA, these features place HNF1 as the prototype of a novel subclass of transcription factors distantly related to homeoproteins.

Molecular interactions within the ecdysone regulatory hierarchy: DNA binding properties of the Drosophila ecdysone-inducible E74A protein

The E74 early ecdysone-inducible gene plays a key role in the regulatory hierarchy activated by ecdysone at the onset of Drosophila metamorphosis. We show here that E74A protein binds to three adjacent sites in the middle of the E74 gene. The consensus sequence for E74A protein binding, determined by random-sequence oligonucleotide selection, contains an invariant purine-rich core sequence, C/AGGAA. This sequence is also present in the binding sites of two mammalian proteins that, like E74A, are related to the ets oncoprotein. Antibody staining of larval salivary gland polytene chromosomes revealed that E74A protein binds to both early and late ecdysone-inducible puffs. This study supports Ashburner's proposal that the early puffs encode site-specific DNA binding proteins that directly interact with the early and late ecdysone-inducible puffs.

The structure of the homeodomain and its functional implications

The three-dimensional structure of the homeodomain, as determined by nuclear magnetic resonance spectroscopy, reveals the presence of a helix-turn-helix motif, similar to the one found in prokaryotic gene regulatory proteins. Isolated homeodomains bind with high affinity to specific DNA sequences. Thus, the structure-function relationship is highly conserved in evolution.

Expression of HOX homeogenes in human neuroblastoma cell culture lines

Mammalian genes containing a class-I homeobox (HOX genes) are highly expressed in the embryonic nervous system. As a first step towards the molecular analysis of the role these genes play in neural cells, we studied the expression of four human HOX genes in five neuroblastoma (NB) cell lines - SK-N-BE, CHP-134, IMR-32, SK-N-SH and LAN-1 - during the process of differentiation induced by treatment with retinoic acid (RA). The four genes, HOX1D, 2F, 3E and 4B, located at corresponding positions in the four HOX loci, share a high degree of sequence similarity with the Drosophila Deformed homeotic gene and constitute a homology group, group 10. One of these genes, HOX1D, is not expressed in the cells used, whereas the other three are highly expressed in untreated and RA-induced NB cells, even though the expression pattern in the various lines is slightly different for the three genes. Our analysis reveals a complex and specific expression pattern in these lines, paving the way to an identification of different NB-cell populations by means of specific HOX gene expression schemes. On the other hand, in every line studied, morphological maturation toward a neuronal differentiated phenotype appears to be associated with increased HOX gene expression.

The specificities of Sex combs reduced and Antennapedia are defined by a distinct portion of each protein that includes the homeodomain

The sequence requirements for distinguishing the functional specificities of two homeodomain proteins, Antennapedia and Sex combs reduced, involved in the specification of segmental identities in Drosophila, have been determined. A series of deletions and hybrid proteins was generated and assayed for their function in vivo after heat shock-induced ectopic expression during development. A distinct portion of each protein, including the residues within and adjacent to both ends of the homeodomain, has been found to almost entirely determine its functional specificity as measured by diagnostic cuticular transformations of embryonic and adult head structures. The remaining sequences contribute to the potency with which the proteins act in different cells and are to a limited extent functionally transferable from one protein to the other.

Assignment of the side-chain 1H and 13C resonances of interleukin-1 beta using double- and triple-resonance heteronuclear three-dimensional NMR spectroscopy

The assignment of the aliphatic 1H and 13C resonances of IL-1 beta, a protein of 153 residues and molecular mass 17.4 kDa, is presented by use of a number of novel three-dimensional (3D) heteronuclear NMR experiments which rely on large heteronuclear one-bond J couplings to transfer magnetization and establish through-bond connectivities. These 3D NMR experiments circumvent problems traditionally associated with the application of conventional 2D 1H-1H correlation experiments to proteins of this size, in particular the extensive chemical shift overlap which precludes the interpretation of the spectra and the reduced sensitivity arising from 1H line widths that are often significantly larger than the 1H-1H J couplings. The assignment proceeds in two stages. In the first step the 13C alpha chemical shifts are correlated with the NH and 15N chemical shifts by a 3D triple-resonance NH-15N-13C alpha (HNCA) correlation experiment which reveals both intraresidue NH(i)-15N(i)-13C alpha (i) and some weaker interresidue NH(i)-15N(i)-C alpha (i-1) correlations, the former via intraresidue one-bond 1JNC alpha and the latter via interresidue two-bond 2JNC alpha couplings. As the NH, 15N, and C alpha H chemical shifts had previously been sequentially assigned by 3D 1H Hartmann-Hahn 15N-1H multiple quantum coherence (3D HOHAHA-HMQC) and 3D heteronuclear 1H nuclear Overhauser 15N-1H multiple quantum coherence (3D NOESY-HMQC) spectroscopy [Driscoll, P.C., Clore, G.M., Marion, D., Wingfield, P.T., & Gronenborn, A.M. (1990) Biochemistry 29, 3542-3556], the 3D triple-resonance HNCA correlation experiment permits the sequence-specific assignments of 13C alpha chemical shifts in a straightforward manner. The second step involves the identification of side-chain spin systems by 3D 1H-13C-13C-1H correlated (HCCH-COSY) and 3D 1H-13C-13C-1H total correlated (HCCH-TOCSY) spectroscopy, the latter making use of isotropic mixing of 13C magnetization to obtain relayed connectivities along the side chains. Extensive cross-checks are provided in the assignment procedure by examination of the connectivities between 1H resonances at all the corresponding 13C shifts of the directly bonded 13C nuclei. In this manner, we were able to obtain complete 1H and 13C side-chain assignments for all residues, with the exception of 4 (out of a total of 15) lysine residues for which partial assignments were obtained. The 3D heteronuclear correlation experiments described are highly sensitive, and the required set of three 3D spectra was recorded in only 1 week of measurement time on a single uniformly 15N/13C-labeled 1.7 mM sample of interleukin-1 beta.(ABSTRACT TRUNCATED AT 400 WORDS)

Eukaryotic transcription factors

Images

Crystallization and preliminary X-ray diffraction studies of the engrailed homeodomain and of an engrailed homeodomain/DNA complex

The homeodomain from the engrailed protein of Drosophila has been crystallized from ammonium phosphate at pH 6.8. The crystals form in space group P6(1)22 (or P6(5)22), with cell dimensions a = b = 44.8 A and c = 118.2 A. These crystals diffract to 1.8 A resolution. A complex containing the engrailed homeodomain and a duplex DNA site also has been crystallized. The cocrystals form in space group C2 with a = 131.2 A, b = 45.5 A, c = 72.9 A and beta = 119.0 degrees. These crystals diffract to 2.6 A resolution.

A homeo domain protein reveals the metameric nature of the developing chick hindbrain

The segmented embryonic hindbrain of vertebrates develops by sequential constriction of the neural tube into eight metameric units known as rhombomeres. The cellular and molecular basis of this segmentation process is largely unknown. Using an antibody, we analyzed the expression pattern of the chick homeo domain-containing protein Ghox-lab in the developing chick hindbrain. At the neural plate stage, prior to the appearance of rhombomeres, Ghox-lab is expressed within a single domain that extends anteriorly up to the site where rhombomere 4 will later form. After rhombomere 4 has appeared and as hindbrain segmentation progresses, the level of Ghox-lab protein increases significantly within the fourth rhombomere. This intensification, accompanied by the elimination of Ghox-lab protein in rhombomeres 5 and 6, eventually results in the formation of a distinct island of expression in rhombomere 4. All cells in the newly formed rhombomere 4 express Ghox-lab, except for the cells of the floor plate. In addition, neural crest cells migrating from the fourth rhombomere are also Ghox-lab-positive. These data raise the possibility that Ghox-lab protein might be one of the factors involved in the specification of the metameric pattern of the vertebrate hindbrain.

Genetic analysis of the LexA repressor: isolation and characterization of LexA(Def) mutant proteins

We report the isolation of LexA mutant proteins with impaired repressor function. These mutant proteins were obtained by transforming a LexA-deficient recA-lacZ indicator strain with a randomly mutagenized plasmid harbouring the lexA gene and subsequent selection on MacConkey-lactose indicator plates. A total of 24 different lexA(Def) missense mutations were identified. All except three mutant proteins are produced in near-normal amounts suggesting that they are fairly resistant to intracellular proteases. All lexA(Def) missense mutations are situated within the first 67 amino acids of the amino-terminal DNA binding domain. The properties of an intragenic deletion mutant suggest that the part of the amino-terminal domain important for DNA recognition or domain folding should extent at least to amino acids 69 or 70. A recent 2D-NMR study (Lamerichs et al. 1989) has identified three alpha helices in the DNA binding domain of LexA. The relative orientation of two of them (helices 2 and 3) is reminiscent of, but not identical to, the canonical helix-turn-helix motif suggesting nevertheless that helix 3 might be involved in DNA recognition. The distribution of the lexA(Def) missense mutations along the first 67 amino-terminal amino acids indeed shows some clustering within helix 3, since 8 out of the 24 different missense mutations are found in this helix. However one mutation in front of helix 1 and five mutations between amino acids 61 and 67 suggest that elements other than helices 2 and 3 may be important for DNA binding.

Structural studies of protein-nucleic acid interaction: the sources of sequence-specific binding

Structural studies of DNA-binding proteins and their complexes with DNA have proceeded at an accelerating pace in recent years due to important technical advances in molecular genetics, DNA synthesis, protein crystallography and nuclear magnetic resonance. The last major review on this subject by Pabo & Sauer (1984) summarized the structural and functional studies of the three sequence-specific DNA-binding proteins whose crystal structures were then known, theE. colicatabolite gene activator protein (CAP) (McKay & Steitz, 1981; McKayet al.1982; Weber & Steitz, 1987), acrorepressor from phage λ (Andersonet al.1981), and the DNA-binding proteolytic fragment ofλcIrepressor protein (Pabo & Lewis, 1982) Although crystallographic studies of theE. coli lacrepressor protein were initiated as early as 1971 when it was the only regulatory protein available in sufficient quantities for structural studies (Steitzet al.1974), little was established about the structural aspects of DNA-binding proteins until the structure of CAP was determined in 1980 followed shortly thereafter by the structure ofλcrorepressor and subsequently that of the λ repressor fragment. There are now determined at high resolution the crystal structures of seven prokaryotic gene regulatory proteins or fragments [CAP,λcro,λcIrepressor fragment, 434 repressor fragment (Andersonet al.1987), 434crorepressor (Wolbergeret al.1988),E. coli trprepressor (Schevitzet al.1985),E. coli metrepressor (Raffertyet al.1989)],EcoRI restriction endonuclease (McClarinet al.1986), DNAse I (Suck & Ofner, 1986), the catalytic domain of γδ resolvase (Hatfullet al.1989) and two sequence-independent double-stranded DNA-binding proteins [the Klenow fragment ofE. coliDNA polymerase I (Olliset al.1985) and theE. coliHu protein (Tanakaet al., 1984)].

Four-dimensional heteronuclear triple-resonance NMR spectroscopy of interleukin-1 beta in solution

A method is presented that dramatically improves the resolution of protein nuclear magnetic resonance (NMR) spectra by increasing their dimensionality to four. The power of this technique is demonstrated by the application of four-dimensional carbon-13--nitrogen-15 (13C-15N)--edited nuclear Overhauser effect (NOE) spectroscopy to interleukin-1 beta, a protein of 153 residues. The NOEs between NH and aliphatic protons are first spread out into a third dimension by the 15N chemical shift of the amide 15N atom and subsequently into a fourth dimension by the 13C chemical shift of the directly bonded 13C atoms. By this means ambiguities in the assignment of NOEs between NH and aliphatic protons that are still present in the three-dimensional 15N-edited NOE spectrum due to extensive chemical shift overlap and degeneracy of aliphatic resonances are completely removed. Consequently, many more approximate interproton distance restraints can be obtained from the NOE data than was heretofore possible, thereby expanding the horizons of three-dimensional structure determination by NMR to larger proteins.

Murine developmental control genes

Various strategies have been used to isolate genes that participate in the regulation of mouse development. Gene families that have been identified on the basis of their homology to motifs within Drosophila control genes or human transcription factor genes, namely homeobox (Hox), paired-box (Pax), and POU genes, can be compared with respect to gene organization, structure, and expression patterns. The functions of these genes can be analyzed molecularly in vitro and in vivo with the use of available mouse mutants or transgenic mice. In addition, it has been possible to generate gain- or loss-of-function mutations by random or targeted introduction of transgenes. Models derived from these studies can reveal the successive steps of developmental control on a genetic level.

Determination of the three-dimensional structure of the Antennapedia homeodomain from Drosophila in solution by 1H nuclear magnetic resonance spectroscopy

The determination of the three-dimensional structure of the Antennapedia homeodomain from Drosophila in solution is described. The techniques used are 1H nuclear magnetic resonance spectroscopy for the data collection, and calculation of the protein structure with the program DISMAN followed by restrained energy minimization with a modified version of the program AMBER. A group of 19 conformers characterizes a well-defined structure for residues 7 to 59, with an average root-mean-square distance from the backbone atoms of 0.6 A relative to the mean of the 19 structures. The structure contains a helix from residues 10 to 21, a helix-turn-helix motif from residues 28 to 52, which is similar to those reported for several prokaryotic repressor proteins, and a somewhat flexible fourth helix from residues 53 to 59, which essentially forms an extension of the presumed recognition helix, residues 42 to 52. The helices enclose a structurally well-defined molecular core of hydrophobic amino acid side-chains.

A myosin-like dimerization helix and an extra-large homeodomain are essential elements of the tripartite DNA binding structure of LFB1

The transcription activator LFB1 is a major determinant of hepatocyte-specific expression of many genes. To study the mechanisms underlying LFB1 transcriptional selectivity, we have initiated its biochemical characterization. By in vitro complementation assays we have defined two distinct regions required for high levels of transcription, which resemble previously described activation domains. In contrast, the region of LFB1 necessary for DNA binding displays several novel features. The DNA binding domain is tripartite, including a homeodomain of unusual length (81 amino acids) and an N-terminal helix similar to part of myosin. This helical region mediates dimerization, which is shown to be essential for DNA binding.

DNA binding properties of the purified Antennapedia homeodomain

The in vitro DNA binding properties of a purified 68-amino acid Antennapedia homeodomain (Antp HD) peptide have been analyzed. Equilibrium and kinetic binding studies showed that stable DNA-protein complexes are formed with a Kd of 1.6 x 10(-9) M and 1.8 x 10(-10) M, respectively. Heterodimer analysis led to the conclusion that Antp HD interacts in vitro as a monomer with the DNA target sites used in our study. The results of methylation and ethylation interference studies indicated that the Antp HD closely approaches the target DNA primarily from one side in a region extending across three phosphate backbones. The DNA binding properties of the Antp HD and prokaryotic DNA binding domains that share a helix-turn-helix motif are compared.

Eukaryotic transcription factors

Many nuclear proteins have been found recently to interact with short conserved sequences which are involved in regulating the transcription of various genes. Nuclear transcription factors may be arbitrarily subdivided into two groups, ubiquitous and tissue-specific. The transcription of one gene is usually regulated by several factors which interact with different sequences located either in the promoter region of the gene or outside it. The appearance or disappearance of transcription factors for some genes corresponds to certain phases of cell differentiation or dedifferentiation and even to individual stages of the cell cycle.

A DNA-binding homeodomain in histone H1

The structure of the globular domain of chicken histone H1 was compared here with that of the DNA-binding homeodomain in the Drosophila Antp protein, and they were observed to display considerable similarity. Both of them consist of three or four alpha-helices separated by well-defined turns. Charged residues in the aminoterminal end of alpha 3 are therefore suggested to be responsible for sequence-specific recognition of DNA by the histone. In addition, alpha 2 of H1, with a short leucine zipper in it, may be capable of protein-protein interaction in a similar manner to the other homeodomains.

New type of POU domain in germ line-specific protein Oct-4

Members of a family of murine octamer-binding proteins interact specifically with the octamer motif, a transcription regulatory element found in the promoter and enhancer regions of many genes. Oct-4 is a maternally expressed protein that is also present in the pre-implantation mouse embryo. Although many regulatory proteins are expressed in post-implantation embryos, transcription factors regulating pre-implantation processes have remained elusive. The Oct-4 gene is therefore a prime candidate for an early developmental control gene. Here we report the complementary DNA cloning of the mouse Oct-4 gene, and the characterization of the encoded protein(s) by sequential in vitro transcription, translation, DNA-binding and protease-clipping analysis. Deletion analysis shows that the DNA-binding activity is mediated by a POU domain encoded in an open reading frame corresponding to a 324-amino-acid protein. Sequence comparison with known POU domains reveals that Oct-4 is a novel member of the POU-family.

Conserved residues make similar contacts in two repressor-operator complexes

Comparison of a lambda repressor-operator complex and a 434 repressor-operator complex reveals that three conserved residues in the helix-turn-helix (HTH) region make similar contacts in each of the crystallographically determined structures. These conserved residues and their interactions with phosphodiester oxygens help establish a frame of reference within which other HTH residues make contacts that are critical for site-specific recognition. Such "positioning contacts" may be important conserved features within families of HTH proteins. In contrast, the structural comparisons appear to rule out any simple "recognition code" at the level of detailed side chain-base pair interactions.

Engrailed, a homeodomain protein, can repress in vitro transcription by competition with the TATA box-binding protein transcription factor IID

Engrailed (En) is a homeodomain protein that binds to a consensus sequence (NP) and plays an important role during Drosophila development. Purified En, which is produced in Escherichia coli, binds not only to this consensus sequence but also to the TATA box of the Drosophila Hsp70 promoter and of other eukaryotic promoters. Interestingly, En represses transcription of these promoters in an in vitro-reconstituted mammalian transcription system and footprint analyses show that En competes with the TATA box-binding protein transcription factor IID for binding to the TATA box. In contrast, a stable template-committed complex formed by preincubation of transcription factor IID with the promoter is not disrupted by addition of En, and in this case transcription is not repressed. These in vitro studies suggest a transcriptional repression mechanism, involving competition between En and transcription factor IID for TATA box binding, that may be involved in En-mediated repression in vivo.

A new homeobox gene contributes the DNA binding domain of the t(1;19) translocation protein in pre-B ALL

It was previously shown that the chromosome 19 breakpoint of the t(1;19)(q23;p13.3) translocation, found in human pre-B cell acute lymphoblastic leukemias, is within the E2A transcription factor gene on chromosome 19. A cell line with this translocation contains two novel chimeric mRNAs, both with the same 5'E2A sequences but with different lengths of 3' sequence from a previously unrecognized gene dubbed prl, located on chromosome 1. The chimeric RNAs encode a protein that lacks 171 amino acids of E2A, including its DNA binding and dimerization motifs, but have instead a homeobox-related sequence from prl. Therefore, the production of a chimeric E2A-Prl protein may contribute to the acute lymphoblastic phenotype by directly altering the expression of genes normally responsive to the Prl homeoprotein.

A molecular view of the Ultrabithorax homeotic gene of Drosophila

The pioneering work of E.B. Lewis drew attention early on to the developmental role of Ultrabithorax (Ubx), one of a group of homeotic genes that specify the distinguishing features of Drosophila segments. Recent molecular work sheds light on the genetic complexity discovered by Lewis and demonstrates that Ubx executes its developmental role by producing a family of closely related DNA-binding proteins that can function as transcription factors.

The b alleles of U. maydis, whose combinations program pathogenic development, code for polypeptides containing a homeodomain-related motif

U. maydis is a fungal pathogen of corn with two forms: one is yeast-like and nonpathogenic; the other is filamentous and pathogenic. The b locus, with 25 different alleles, regulates this dimorphism: any combination of two different alleles triggers pathogenic development, whereas the presence of identical alleles results in the yeast-like form. We have cloned four b alleles (b1, b2, b3, and b4) and show that the b locus contains a single open reading frame (ORF) of 410 amino acids with a variable N-terminal region and a highly conserved C-terminal region (60% and 93% identity, respectively). Mutational analysis confirms that this ORF is responsible for b activity. The b polypeptides appear to be DNA binding proteins because they contain a motif related to the homeodomain in their constant region. We propose that combinatorial interactions between b polypeptides generate regulatory proteins that determine the developmental program of the fungus.