The ankyrin repeat: a diversity of interactions on a common structural framework.

The ankyrin repeat is one of the most common protein sequence motifs. Recent X-ray and NMR structures of ankyrin-repeat proteins and their complexes have provided invaluable insights into the molecular basis of the extraordinary variety of biological activities of these molecules. In particular, they have begun to reveal how a large family of structurally related proteins can interact specifically with such a diverse array of macromolecular targets.

X-ray structural analysis of the yeast cell cycle regulator Swi6 reveals variations of the ankyrin fold and has implications for Swi6 function.

Swi6 is a 92,000 Mr protein common to two distinct transcriptional activation complexes (SBF and MBF) that coordinate gene expression at the G1-S boundary of the yeast cell cycle. The X-ray structure of a central 36,000 Mr fragment has been determined and refined at 2.1 A resolution. The structure reveals a basic framework of five ankyrin repeat modules that is elaborated through a series of helical insertions distinguishing it from structures of other ankyrin repeat proteins. A second domain contains an approximately 30-residue region of extended structure that interacts with the ankyrin repeat core over a substantial proportion of its surface. Conservation of residues buried by these interactions indicates that all members of the Swi6/Cdc10 family share a similar architecture. Several temperature-sensitive mutations within Swi6 and Cdc10 appear to disrupt these interdomain contacts rather than destabilize the ankyrin repeat core. The unusual domain arrangement may be crucial for the modulation of interactions with other co-regulatory molecules such as cyclin-CDK complexes, and has implications for the quaternary interactions within the multisubunit SBF and MBF transcription complexes.

An epitope structure for the C-terminal domain of dystrophin and utrophin.

The muscular dystrophy protein, dystrophin, and the closely related protein, utrophin, are large cytoskeletal proteins which link actin microfilaments to the plasma membrane. A panel of 38 monoclonal antibodies (mAbs) has been produced against the C-terminal domains of dystrophin and utrophin. This domain interacts with both dystrobrevins, via their "leucine zipper" coiled-coil helices, and syntrophins, adaptor proteins which also interact with nitric oxide synthetase and transmembrane sodium channels. The amino acid sequences recognized by the mAbs have now been identified using a variety of epitope mapping techniques, including fragmentation by transposon mutagenesis, synthetic peptides, phage-displayed peptide libraries, and mutant dystrophins expressed in transgenic mice. In addition to defining antibody recognition sites, mapping was sufficiently precise to provide structural information, since individual amino acids accessible on the surface of the native protein were identified in many cases. In two regions of the domain, short linear epitopes were found in proline-rich sequences which may form surface loops, turns, or linkers, but these were separated by a third region which contained mainly conformational epitopes. The results are consistent with a loose and flexible structure for much of the C-terminal domain, especially around the highly conserved second leucine zipper or coiled-coil helix (CC-H2), but there is evidence for denaturation-resistant tertiary structure in the syntrophin-binding region and the first coiled-coil helix (CC-H1).

Structural and functional architecture of the yeast cell-cycle transcription factor swi6.

The structural and functional organisation of Swi6, a transcriptional regulator of the budding yeast cell cycle has been analysed by a combination of biochemical, biophysical and genetic methods. Limited proteolysis indicates the presence of a approximately 15 kDa N-terminal domain which is dispensable for Swi6 activity in vivo and which is separated from the rest of the molecule by an extended linker of at least 43 residues. Within the central region, a 141 residue segment that is capable of transcriptional activation encompasses a structural domain of approximately 85 residues. In turn, this is tightly associated with an adjacent 28 kDa domain containing at least four ankyrin-repeat (ANK) motifs. A second protease sensitive region connects the ANK domain to the remaining 30 kDa C-terminal portion of Swi6 which contains a second transcriptional activator and sequences required for heteromerisation with Swi4 or Mbp1. Transactivation by the activating regions of Swi6 is antagonised when either are combined with the central ankyrin repeat motifs. Hydrodynamic measurements indicate that an N-terminal 62 kDa fragment comprising the first three domains is monomeric in solution and exhibits an unusually high frictional coefficient consistent with the extended, multi-domain structure suggested by proteolytic analysis.

Inhibition of Xbra transcription activation causes defects in mesodermal patterning and reveals autoregulation of Xbra in dorsal mesoderm.

The Brachyury (T) gene is required for formation of posterior mesoderm and for axial development in both mouse and zebrafish embryos. In this paper, we first show that the Xenopus homologue of Brachyury, Xbra, and the zebrafish homologue, no tail (ntl), both function as transcription activators. The activation domains of both proteins map to their carboxy terminal regions, and we note that the activation domain is absent in two zebrafish Brachyury mutations, suggesting that it is required for gene function. A dominant-interfering Xbra construct was generated by replacing the activation domain of Xbra with the repressor domain of the Drosophila engrailed protein. Microinjection of RNA encoding this fusion protein allowed us to generate Xenopus and zebrafish embryos which show striking similarities to genetic mutants in mouse and fish. These results indicate that the function of Brachyury during vertebrate gastrulation is to activate transcription of mesoderm-specific genes. Additional experiments show that Xbra transcription activation is required for regulation of Xbra itself in dorsal, but not ventral, mesoderm. The approach described in this paper, in which the DNA-binding domain of a transcription activator is fused to the engrailed repressor domain, should assist in the analysis of other Xenopus and zebrafish transcription factors.

Transposon tools for recombinant DNA manipulation: characterization of transcriptional regulators from yeast, Xenopus, and mouse.

Transposon Tn1000 has been adapted to deliver novel DNA sequences for manipulating recombinant DNA. The transposition procedure for these "tagged" Tn1000s is simple and applicable to most plasmids in current use. For yeast molecular biology, tagged Tn1000s introduce a variety of yeast selective markers and replication origins into plasmids and cosmids. In addition, the beta-globin minimal promoter and lacZ gene of Tn(beta)lac serve as a mobile reporter of eukaryotic enhancer activity. In this paper, Tn(beta)lac was used to localize a mouse HoxB-complex enhancer in transgenic mice. Other tagged transposons create Gal4 DNA-binding-domain fusions, in either Escherichia coli or yeast plasmids, for use in one- and two-hybrid tests of transcriptional activation and protein-protein interaction, respectively. With such fusions, the Saccharomyces cerevisiae Swi6 G1/S-phase transcription factor and the Xenopus laevis Pintallavis developmental regulator are shown to activate transcription. Furthermore, the same transposon insertions also facilitated mapping of the Swi6 and Pintallavis domains responsible for transcriptional activation. Thus, as well as introducing novel sequences, tagged transposons share the numerous other applications of transposition such as producing insertional mutations, creating deletion series, or serving as mobile primer sites for DNA sequencing.

Epitope mapping of recombinant antigens by transposon mutagenesis.

We describe a method for generating a plasmid library expressing random truncations of a recombinant protein and for epitope mapping by screening the library with monoclonal antibodies. The key step is the random introduction of the transposon, Tn1000, which carries stop codons in all three reading frames, into a bacterial expression plasmid by using a simple bacterial mating procedure. Antibody-positive clones are then selected and the point of protein truncation is determined by sequencing the plasmid DNA at the point of transposon insertion. One advantage of the method is that no subcloning or in vitro manipulation of DNA is necessary.

Protein splicing in the maturation of M. tuberculosis recA protein: a mechanism for tolerating a novel class of intervening sequence.

The M. tuberculosis recA locus comprises an 85 kd open reading frame but produced 38 kd RecA and 47 kd products in E. coli. No RNA processing was detected; rather, an 85 kd precursor protein was spliced, releasing a 47 kd spacer protein, and joining its terminal fragments to form mature RecA protein. "Spacer" protein was also produced in M. tuberculosis and from a hybrid spacer-LacZ alpha fusion molecule. Mutagenesis at codon wobble positions at one splice junction showed that protein rather than nucleotide sequence determined splicing activity. Other mutants defined additional regions needed for splicing and allowed processing to be followed. Splicing was essential for RecA activity in E. coli. The possibility that splicing is a manifestation of a novel class of genetic element is discussed.

Mutagenesis induced by bacterial UmuDC proteins and their plasmid homologues.

The popular image of a world full of pollutants mutating DNA is only partly true since there are relatively few agents which can subtly and directly change base coding; for example, some alkylating agents alter guanine so that it pairs like adenine. Many more mutagens are less subtle and simply destroy coding altogether rather than changing it. Such mutagens include ultraviolet light, X-rays, DNA cross-linkers and other agents which make DNA breaks or large adducts. In Escherichia coli, mutagenesis by these agents occurs during a DNA repair process which increases cell survival but with an inherent possibility of changing the original sequence. Such mutagenic DNA repair is, in part, encoded by the E. coli umuDC operon. This article reviews the structure, function, regulation and evolution of the umuDC operon and similar genes found both in other species and on naturally occurring plasmids.

Alleviation of EcoK DNA restriction in Escherichia coli and involvement of umuDC activity.

The activity of the EcoK DNA restriction system of Escherichia coli reduces both the plating efficiency of unmodified phage lambda and the transforming ability of unmodified pBR322 plasmid DNA. However, restriction can be alleviated in wild-type cells, by UV irradiation and expression of the SOS response, so that 10(3)- to 10(4)-fold increases in phage growth and fourfold increases in plasmid transformation occurred with unmodified DNA. Restriction alleviation was found to be a transient effect because induced cells, which initially failed to restrict unmodified plasmid DNA, later restricted unmodified phage lambda. Although the SOS response was needed for restriction alleviation, constitutive SOS induction, elicited genetically with a recA730 mutation, did not alleviate restriction and UV irradiation was still needed. A hitherto unsuspected involvement of the umuDC operon in this alleviation of restriction is characterized and, by differential complementation, was separated from the better known role of umuDC in mutagenic DNA repair. The need for cleavage of UmuD for restriction alleviation was shown with plasmids encoding cleavable, cleaved, and non-cleavable forms of UmuD. However, UV irradiation was still needed even when cleaved UmuD was provided. The possibility that restriction alleviation occurs by a general inhibition of the EcoK restriction/modification complex was tested and discounted because modification of lambda was not reduced by UV irradiation. An alternative idea, that restriction activity was competitively reduced by an increase in EcoK modification, was also discounted by the lack of any increase in the modification of lambda Ral-, a naturally undermodified phage. Other possible mechanisms for restriction alleviation are discussed.

Rapid mapping by transposon mutagenesis of epitopes on the muscular dystrophy protein, dystrophin.

Antibody-binding epitopes in the central helical region of the muscular dystrophy protein, dystrophin, have been mapped using a new strategy of transposon mutagenesis. Tn1000 transposons carrying translation termination codons were introduced randomly by bacterial mating into a large fragment of dystrophin cDNA in a pEX2 plasmid to produce a library of transformants expressing truncated dystrophin fusion proteins. Epitopes were progressively lost as the expressed sequences were shortened, enabling the epitopes recognised by 22 monoclonal antibodies to be placed in order along the dystrophin molecule without in vitro manipulation of DNA. The C-terminus of each truncated fusion protein was precisely located within the dystrophin sequence by direct sequencing of pEX2 transformants using transposon-specific primers. Sequences as short as 7 and 17 amino-acids have been identified as essential for antibody binding in this way. Nineteen of the 22 monoclonal antibodies had been selected for their ability to bind both native and SDS-denatured dystrophin and 15 of these bind to one sequence of 74 amino-acids (residues 1431-1505 of the 3684 residue sequence). This may be an area of high immunogenicity or of close structural similarity between native dystrophin and the SDS-treated recombinant fragment used for immunization.

Cloning and structural characterization of the mcrA locus of Escherichia coli.

Escherichia coli has DNA restriction systems which are able to recognize and attack modified cytosine residues in the DNA of incoming bacteriophages and plasmids. The locus for the McrA/RglA system of modified cytosine restriction was located near the pin gene of the defective element, e14. Hence, loss of the e14 element through abortive induction after UV irradiation caused a permanent loss of McrA restriction activity. e14 DNA encoding McrA restriction was cloned and sequenced to reveal a single open reading frame of 831 bp with a predicted gene product of 31 kDa. Clones expressing the complete open reading frame conferred both McrA and RglA phenotypes; however, a deletion derivative was found which complemented RglA restriction against nonglucosylated T6gt phage but did not complement for McrA restriction of methylated plasmid DNA. Possible explanations for this activity and a comparison with the different organization of the McrB/RglB restriction system are discussed.

Functional complementation between chromosomal and plasmid mutagenic DNA repair genes in bacteria.

The umuDC operons of Escherichia coli and Salmonella typhimurium and the analogous plasmid operons mucAB and impCAB have been previously characterized in terms of their roles in DNA repair and induced mutagenesis by radiation and many chemicals. The interrelationships of these mutagenic DNA repair operons were examined in vivo in functional tests of interchangeability of operon subunits in conferring UV resistance and UV mutability phenotypes to wild-type S. typhimurium and umu mutants of E. coli. This approach was combined with DNA and protein sequence comparisons between the four operons and a fifth operon, samAB, from the S. typhimurium LT2 cryptic plasmid. Components of the E. coli and S. typhimurium umu operons were reciprocally interchangeable whereas impCA and mucA could not function with umuC in either of these species. mucA and impB could also combine to give a mutagenic response to UV. These active combinations were associated with higher degrees of conservation of protein sequence than in other heterologous gene combinations and related to specific regions of sequence that may specify subunit interactions. The dominance of the E. coli umuD44 mutation over umuD was revealed in both wild-type E. coli and S. typhimurium and also demonstrated against impCAB. Finally interspecies transfer showed that the apparently poor activity of the S. typhimurium umuD gene in situ is not the result of an inherent defect in umuD but is due to the simultaneous presence of the S. typhimurium umuC sequence. It is suggested that the limitation of umuD activity by umuC in S. typhimurium is the basis of the poor induced mutability of this organism.

Mutagenic DNA repair in enterobacteria.

Sixteen species of enterobacteria have been screened for mutagenic DNA repair activity. In Escherichia coli, mutagenic DNA repair is encoded by the umuDC operon. Synthesis of UmuD and UmuC proteins is induced as part of the SOS response to DNA damage, and after induction, the UmuD protein undergoes an autocatalytic cleavage to produce the carboxy-terminal UmuD' fragment needed for induced mutagenesis. The presence of a similar system in other species was examined by using a combined approach of inducible-mutagenesis assays, cross-reactivity to E. coli UmuD and UmuD' antibodies to test for induction and cleavage of UmuD-like proteins, and hybridization with E. coli and Salmonella typhimurium umu DNA probes to map umu-like genes. The results indicate a more widespread distribution of mutagenic DNA repair in other species than was previously thought. They also show that umu loci can be more complex in other species than in E. coli. Differences in UV-induced mutability of more than 200-fold were seen between different species of enteric bacteria and even between multiple natural isolates of E. coli, and yet some of the species which display a poorly mutable phenotype still have umu-like genes and proteins. It is suggested that umDC genes can be curtailed in their mutagenic activities but that they may still participate in some other, unknown process which provides the continued stimulus for their retention.

Novel structure of the recA locus of Mycobacterium tuberculosis implies processing of the gene product.

A fragment of Mycobacterium tuberculosis DNA containing recA-like sequences was identified by hybridization with the Escherichia coli recA gene and cloned. Although no expression was detected from its own promoter in E. coli, expression from a vector promoter partially complemented E. coli recA mutants for recombination, DNA repair, and mutagenesis, but not for induction of phage lambda. This clone produced a protein which cross-reacts with antisera raised against the E. coli RecA protein and was approximately the same size. However, the nucleotide sequence of the cloned fragment revealed the presence of an open reading frame for a protein about twice the size of other RecA proteins and the cloned product detected by Western blotting (immunoblotting). The predicted M. tuberculosis RecA protein sequence was homologous with RecA sequences from other bacteria, but this homology was not dispersed; rather it was localized to the first 254 and the last 96 amino acids, with the intervening 440 amino acids being unrelated. Furthermore, the junctions of homology were in register with the uninterrupted sequence of the E. coli RecA protein. Identical restriction fragments were found in the genomic DNAs of M. tuberculosis H37Rv and H37Ra and of M. bovis BCG. It is concluded that the ancestral recA gene of these species diversified via an insertional mutation of at least 1,320 bp of DNA. Possible processing mechanisms for synthesizing a normal-size RecA protein from this elongated sequence are discussed.

Different mechanisms for SOS induced alleviation of DNA restriction in Escherichia coli.

The alleviation of DNA restriction during the SOS response in Escherichia coli has been further investigated. With the EcoK DNA restriction system UV irradiated wild-type cells show a 10(4)-fold increase in ability to plate non-modified lambda phage and a 3-4 fold increase in transformation by non-modified plasmid DNA. A role for the umuDC genes of E coli in the process of SOS-induced restriction alleviation was identified by showing that a umuC122::Tn5 mutant could alleviate EcoK restriction to only 5% that of wild-type levels. Although umuDC are better characterized for their pivotal role in SOS induced mutagenesis, it is demonstrated here that umu-dependent alleviation of EcoK restriction is a transient process in which umu-dependent mutagenesis plays little part. A second form of SOS induced alleviation of DNA restriction is described in this paper involving the McrA restriction system. The mcrA gene is shown to be encoded within a defective prophage called e14 situated at the 25 min region on the Escherichia coli genetic map. e14 is known to abortively excise from the chromosome after SOS induction and it is demonstrated in this report that mcrA is lost from the genome after SOS induction as part of e14. This results in co-ordinate decrease in the level of McrA restriction within a population of cells.

Structural characterization of the Salmonella typhimurium LT2 umu operon.

The umuDC operon of Escherichia coli encodes functions required for mutagenesis induced by radiation and a wide variety of chemicals. The closely related organism Salmonella typhimurium is markedly less mutable than E. coli, but a umu homolog has recently been identified and cloned from the LT2 subline. In this study the nucleotide sequence and structure of the S. typhimurium LT2 umu operon have been determined and its gene products have been identified so that the molecular basis of umu activity might be understood more fully. S. typhimurium LT2 umu consists of a smaller 417-base-pair (bp) umuD gene ending 2 bp upstream of a larger 1,266-bp umuC gene. The only apparent structural difference between the two operons is the lack of gene overlap. An SOS box identical to that found in E. coli is present in the promoter region upstream of umuD. The calculated molecular masses of the umuD and umuC gene products were 15.3 and 47.8 kilodaltons, respectively, which agree with figures determined by transpositional disruption and maxicell analysis. The S. typhimurium and E. coli umuD sequences were 68% homologous and encoded products with 71% amino acid identity; the umuC sequences were 71% homologous and encoded products with 83% amino acid identity. Furthermore, the potential UmuD cleavage site and associated catalytic sites could be identified. Thus the very different mutagenic responses of S. typhimurium LT2 and E. coli cannot be accounted for by gross differences in operon structure or gene products. Rather, the ability of the cloned S. typhimurium umuD gene to give stronger complementation of E. coli umuD77 mutants in the absence of a functional umuC gene suggests that Salmonella UmuC protein normally constrains UmuD protein activity.

Cloning of Salmonella typhimurium DNA encoding mutagenic DNA repair.

Mutagenic DNA repair in Escherichia coli is encoded by the umuDC operon. Salmonella typhimurium DNA which has homology with E. coli umuC and is able to complement E. coli umuC122::Tn5 and umuC36 mutations has been cloned. Complementation of umuD44 mutants and hybridization with E. coli umuD also occurred, but these activities were much weaker than with umuC. Restriction enzyme mapping indicated that the composition of the cloned fragment is different from the E. coli umuDC operon. Therefore, a umu-like function of S. typhimurium has been found; the phenotype of this function is weaker than that of its E. coli counterpart, which is consistent with the weak mutagenic response of S. typhimurium to UV compared with the response in E. coli.