The solution structure of the Zα domain of the human RNA editing enzyme ADAR1 reveals a prepositioned binding surface for Z-DNA

Double-stranded RNA deaminase I (ADAR1) contains the Z-DNA binding domain Zα. Here we report the solution structure of free Zα and map the interaction surface with Z-DNA, confirming roles previously assigned to residues by mutagenesis. Comparison with the crystal structure of the (Zα)2/Z-DNA complex shows that most Z-DNA contacting residues in free Zα are prepositioned to bind Z-DNA, thus minimizing the entropic cost of binding. Comparison with homologous (α+β)helix–turn–helix/B-DNA complexes suggests that binding of Zα to B-DNA is disfavored by steric hindrance, but does not eliminate the possibility that related domains may bind to both B- and Z-DNA.

Minimizing a binding domain from protein A.

We present a systematic approach to minimizing the Z-domain of protein A, a three-helix bundle (59 residues total) that binds tightly (Kd = 10 nM) to the Fc portion of an immunoglobin IgG1. Despite the fact that all the contacts seen in the x-ray structure of the complex with the IgG are derived from residues in the first two helices, when helix 3 is deleted, binding affinity is reduced > 10(5)-fold (Kd > 1 mM). By using structure-based design and phage display methods, we have iteratively improved the stability and binding affinity for a two-helix derivative, 33 residues in length, such that it binds IgG1, with a Kd of 43 nM. This was accomplished by stepwise selection of random mutations from three regions of the truncated Z-peptide: the 4 hydrophobic residues from helix 1 and helix 2 that contacted helix 3 (the exoface), followed by 5 residues between helix 1 and helix 2 (the intraface), and lastly by 19 residues at or near the interface that interacts with Fc (the interface). As selected mutations from each region were compiled (12 in total), they led to progressive increases in affinity for IgG, and concomitant increases in alpha-helical content reflecting increased stabilization of the two-helix scaffold. Thus, by sequential increases in the stability of the structure and improvements in the quality of the intermolecular contacts, one can reduce larger binding domains to smaller ones. Such mini-protein binding domains are more amenable to synthetic chemistry and thus may be useful starting points for the design of smaller organic mimics. Smaller binding motifs also provide simplified and more tractable models for understanding determinants of protein function and stability.

A Z-DNA binding domain present in the human editing enzyme, double-stranded RNA adenosine deaminase

Editing of RNA changes the read-out of information from DNA by altering the nucleotide sequence of a transcript. One type of RNA editing found in all metazoans uses double-stranded RNA (dsRNA) as a substrate and results in the deamination of adenosine to give inosine, which is translated as guanosine. Editing thus allows variant proteins to be produced from a single pre-mRNA. A mechanism by which dsRNA substrates form is through pairing of intronic and exonic sequences before the removal of noncoding sequences by splicing. Here we report that the RNA editing enzyme, human dsRNA adenosine deaminase (DRADA1, or ADAR1) contains a domain (Zα) that binds specifically to the left-handed Z-DNA conformation with high affinity (KD = 4 nM). As formation of Z-DNA in vivo occurs 5′ to, or behind, a moving RNA polymerase during transcription, recognition of Z-DNA by DRADA1 provides a plausible mechanism by which DRADA1 can be targeted to a nascent RNA so that editing occurs before splicing. Analysis of sequences related to Zα has allowed identification of motifs common to this class of nucleic acid binding domain.

Membrane-bound state of the colicin E1 channel domain as an extended two-dimensional helical array

Atomic level structures have been determined for the soluble forms of several colicins and toxins, but the structural changes that occur after membrane binding have not been well characterized. Changes occurring in the transition from the soluble to membrane-bound state of the C-terminal 190-residue channel polypeptide of colicin E1 (P190) bound to anionic membranes are described. In the membrane-bound state, the α-helical content increases from 60–64% to 80–90%, with a concomitant increase in the average length of the helical segments from 12 to 16 or 17 residues, close to the length required to span the membrane bilayer in the open channel state. The average distance between helical segments is increased and interhelix interactions are weakened, as shown by a major loss of tertiary structure interactions, decreased efficiency of fluorescence resonance energy transfer from an energy donor on helix V of P190 to an acceptor on helix IX, and decreased resonance energy transfer at higher temperatures, not observed in soluble P190, implying freedom of motion of helical segments. Weaker interactions are also shown by a calorimetric thermal transition of low cooperativity, and the extended nature of the helical array is shown by a 3- to 4-fold increase in the average area subtended per molecule to 4,200 Å2 on the membrane surface. The latter, with analysis of the heat capacity changes, implies the absence of a developed hydrophobic core in the membrane-bound P190. The membrane interfacial layer thus serves to promote formation of a highly helical extended two-dimensional flexible net. The properties of the membrane-bound state of the colicin channel domain (i.e., hydrophobic anchor, lengthened and loosely coupled α-helices, and close association with the membrane interfacial layer) are plausible structural features for the state that is a prerequisite for voltage gating, formation of transmembrane helices, and channel opening.

Genetic definition of a protein-splicing domain: Functional mini-inteins support structure predictions and a model for intein evolution

Inteins are protein-splicing elements, most of which contain conserved sequence blocks that define a family of homing endonucleases. Like group I introns that encode such endonucleases, inteins are mobile genetic elements. Recent crystallography and computer modeling studies suggest that inteins consist of two structural domains that correspond to the endonuclease and the protein-splicing elements. To determine whether the bipartite structure of inteins is mirrored by the functional independence of the protein-splicing domain, the entire endonuclease component was deleted from the Mycobacterium tuberculosis recA intein. Guided by computer modeling studies, and taking advantage of genetic systems designed to monitor intein function, the 440-aa Mtu recA intein was reduced to a functional mini-intein of 137 aa. The accuracy of splicing of several mini-inteins was verified. This work not only substantiates structure predictions for intein function but also supports the hypothesis that, like group I introns, mobile inteins arose by an endonuclease gene invading a sequence encoding a small, functional splicing element.

Construction of a Z-DNA-specific restriction endonuclease

Novel restriction enzymes can be created by fusing the nuclease domain of FokI endonuclease with defined DNA binding domains. Recently, we have characterized a domain (Zα) from the N-terminal region of human double-stranded RNA adenosine deaminase (hADAR1), which binds the Z-conformation with high specificity. Here we report creation of a conformation-specific endonuclease, Zα nuclease, which is a chimera of Zα and FokI nuclease. Purified Zα nuclease cleaves negatively supercoiled plasmids only when they contain a Z-DNA forming insert, such as (dC-dG)13. The precise location of the cleavage sites was determined by primer extension. Cutting has been mapped to the edge of the B-Z junction, suggesting that Zα nuclease binds within the Z-DNA insert, but cleaves in the nearby B-DNA, by using a mechanism similar to type IIs restriction enzymes. These data show that Zα binds Z-DNA in an environment similar to that in a cell. Zα nuclease, a structure-specific restriction enzyme, may be a useful tool for further study of the biological role of Z-DNA.

The PDZ Domain of the LIM Protein Enigma Binds to β-Tropomyosin

PDZ and LIM domains are modular protein interaction motifs present in proteins with diverse functions. Enigma is representative of a family of proteins composed of a series of conserved PDZ and LIM domains. The LIM domains of Enigma and its most related family member, Enigma homology protein, bind to protein kinases, whereas the PDZ domains of Enigma and family member actin-associated LIM protein bind to actin filaments. Enigma localizes to actin filaments in fibroblasts via its PDZ domain, and actin-associated LIM protein binds to and colocalizes with the actin-binding protein α-actinin-2 at Z lines in skeletal muscle. We show that Enigma is present at the Z line in skeletal muscle and that the PDZ domain of Enigma binds to a skeletal muscle target, the actin-binding protein tropomyosin (skeletal β-TM). The interaction between Enigma and skeletal β-TM was specific for the PDZ domain of Enigma, was abolished by mutations in the PDZ domain, and required the PDZ-binding consensus sequence (Thr-Ser-Leu) at the extreme carboxyl terminus of skeletal β-TM. Enigma interacted with isoforms of tropomyosin expressed in C2C12 myotubes and formed an immunoprecipitable complex with skeletal β-TM in transfected cells. The association of Enigma with skeletal β-TM suggests a role for Enigma as an adapter protein that directs LIM-binding proteins to actin filaments of muscle cells.

Gβ5 prevents the RGS7-Gαo interaction through binding to a distinct Gγ-like domain found in RGS7 and other RGS proteins

The G protein β subunit Gβ5 deviates significantly from the other four members of Gβ-subunit family in amino acid sequence and subcellular localization. To detect the protein targets of Gβ5 in vivo, we have isolated a native Gβ5 protein complex from the retinal cytosolic fraction and identified the protein tightly associated with Gβ5 as the regulator of G protein signaling (RGS) protein, RGS7. Here we show that complexes of Gβ5 with RGS proteins can be formed in vitro from the recombinant proteins. The reconstituted Gβ5-RGS dimers are similar to the native retinal complex in their behavior on gel-filtration and cation-exchange chromatographies and can be immunoprecipitated with either anti-Gβ5 or anti-RGS7 antibodies. The specific Gβ5-RGS7 interaction is determined by a distinct domain in RGS that has a striking homology to Gγ subunits. Deletion of this domain prevents the RGS7-Gβ5 binding, although the interaction with Gα is retained. Substitution of the Gγ-like domain of RGS7 with a portion of Gγ1 changes its binding specificity from Gβ5 to Gβ1. The interaction of Gβ5 with RGS7 blocked the binding of RGS7 to the Gα subunit Gαo, indicating that Gβ5 is a specific RGS inhibitor.

Myozenin: An α-actinin- and γ-filamin-binding protein of skeletal muscle Z lines

To better understand the structure and function of Z lines, we used sarcomeric isoforms of α-actinin and γ-filamin to screen a human skeletal muscle cDNA library for interacting proteins by using the yeast two-hybrid system. Here we describe myozenin (MYOZ), an α-actinin- and γ-filamin-binding Z line protein expressed predominantly in skeletal muscle. Myozenin is predicted to be a 32-kDa, globular protein with a central glycine-rich domain flanked by α-helical regions with no strong homologies to any known genes. The MYOZ gene has six exons and maps to human chromosome 10q22.1-q22.2. Northern blot analysis demonstrated that this transcript is expressed primarily in skeletal muscle with significantly lower levels of expression in several other tissues. Antimyozenin antisera stain skeletal muscle in a sarcomeric pattern indistinguishable from that seen by using antibodies for α-actinin, and immunogold electron microscopy confirms localization specifically to Z lines. Thus, myozenin is a skeletal muscle Z line protein that may be a good candidate gene for limb-girdle muscular dystrophy or other neuromuscular disorders.

A rapid classification protocol for the CATH Domain Database to support structural genomics

In order to support the structural genomic initiatives, both by rapidly classifying newly determined structures and by suggesting suitable targets for structure determination, we have recently developed several new protocols for classifying structures in the CATH domain database (http://www.biochem.ucl.ac.uk/bsm/cath). These aim to increase the speed of classification of new structures using fast algorithms for structure comparison (GRATH) and to improve the sensitivity in recognising distant structural relatives by incorporating sequence information from relatives in the genomes (DomainFinder). In order to ensure the integrity of the database given the expected increase in data, the CATH Protein Family Database (CATH-PFDB), which currently includes 25 320 structural domains and a further 160 000 sequence relatives has now been installed in a relational ORACLE database. This was essential for developing more rigorous validation procedures and for allowing efficient querying of the database, particularly for genome analysis. The associated Dictionary of Homologous Superfamilies [Bray,J.E., Todd,A.E., Pearl,F.M.G., Thornton,J.M. and Orengo,C.A. (2000) Protein Eng., 13, 153–165], which provides multiple structural alignments and functional information to assist in assigning new relatives, has also been expanded recently and now includes information for 903 homo­logous superfamilies. In order to improve coverage of known structures, preliminary classification levels are now provided for new structures at interim stages in the classification protocol. Since a large proportion of new structures can be rapidly classified using profile-based sequence analysis [e.g. PSI-BLAST: Altschul,S.F., Madden,T.L., Schaffer,A.A., Zhang,J., Zhang,Z., Miller,W. and Lipman,D.J. (1997) Nucleic Acids Res., 25, 3389–3402], this provides preliminary classification for easily recognisable homologues, which in the latest release of CATH (version 1.7) represented nearly three-quarters of the non-identical structures.

The extreme C terminus of progesterone receptor contains a transcriptional repressor domain that functions through a putative corepressor.

Binding of a hormone agonist to a steroid receptor leads to the dissociation of heat shock proteins, dimerization, specific DNA binding, and target gene activation. Although the progesterone antagonist RU486 can induce most of these events, it fails to activate human progesterone receptor (hPR)-dependent transcription. We have previously demonstrated that a conformational change is a key event leading to receptor activation. The major conformational distinction between hormone- and antihormone-bound receptors occurs within the C-terminal portion of the molecule. Furthermore, hPR mutants lacking the C terminus become transcriptionally active in the presence of RU486. These results suggest that the C terminus contains a repressor domain that inhibits the transcriptional activity of the RU486-bound hPR. In this study, we have defined a 12 amino acid (12AA) region in the C terminus of hPR that is necessary and sufficient for the repressor function when fused to the C-terminal truncated hPR or to the GAL4 DNA-binding domain. Mutations in the 12AA domain (aa 917-928) generate an hPR that is active in the presence of RU486. Furthermore, overexpression of the 12AA peptide activates the RU486-bound wild-type hPR without affecting progesterone-dependent activation. These results suggest that association of the 12AA repressor region with a corepressor might inactivate hPR activity when it is bound to RU486. We propose that binding of a hormone agonist to the receptor changes its conformation in the ligand-binding domain so that association with coactivator is promoted and activation of target gene occurs.

RIP and FADD: two "death domain"-containing proteins can induce apoptosis by convergent, but dissociable, pathways.

With use of the yeast two-hybrid system, the proteins RIP and FADD/MORT1 have been shown to interact with the "death domain" of the Fas receptor. Both of these proteins induce apoptosis in mammalian cells. Using receptor fusion constructs, we provide evidence that the self-association of the death domain of RIP by itself is sufficient to elicit apoptosis. However, both the death domain and the adjacent alpha-helical region of RIP are required for the optimal cell killing induced by the overexpression of this gene. By contrast, FADD's ability to induce cell death does not depend on crosslinking. Furthermore, RIP and FADD appear to activate different apoptotic pathways since RIP is able to induce cell death in a cell line that is resistant to the apoptotic effects of Fas, tumor necrosis factor, and FADD. Consistent with this, a dominant negative mutant of FADD, lacking its N-terminal domain, blocks apoptosis induced by RIP but not by FADD. Since both pathways are blocked by CrmA, the interleukin 1 beta converting enzyme family protease inhibitor, these results suggest that FADD and RIP can act along separable pathways that nonetheless converge on a member of the interleukin 1 beta converting enzyme family of cysteine proteases.

Functional analysis of retinoid Z receptor beta, a brain-specific nuclear orphan receptor.

The retinoid Z receptor beta (RZR beta), an orphan receptor, is a member of the retinoic acid receptor (RAR)/thyroid hormone receptor (TR) subfamily of nuclear receptors. RZR beta exhibits a highly restricted brain-specific expression pattern. So far, no natural RZR beta target gene has been identified and the physiological role of the receptor in transcriptional regulation remains to be elucidated. Electrophoretic mobility shift assays reveal binding of RZR beta to monomeric response elements containing the sequence AnnTAGGTCA, but RZR beta-mediated transactivation of reporter genes is only achieved with two property spaced binding sites. We present evidence that RZR beta can function as a cell-type-specific transactivator. In neuronal cells, GaI-RZR beta fusion proteins function as potent transcriptional activators, whereas no transactivation can be observed in nonneuronal cells. Mutational analyses demonstrate that the activation domain (AF-2) of RZR beta and RAR alpha are functionally interchangeable. However, in contrast to RAR and TR, the RZR beta AF-2 cannot function autonomously as a transactivation domain. Furthermore, our data define a novel repressor function for the C-terminal part of the putative ligand binding domain. We propose that the transcriptional activity of RZR beta is regulated by an interplay of different receptor domains with coactivators and corepressors.

Amino-terminal protein-protein interaction motif (POZ-domain) is responsible for activities of the promyelocytic leukemia zinc finger-retinoic acid receptor-alpha fusion protein.

Promyelocytic leukemia zinc finger-retinoic acid receptor a (PLZF-RARalpha), a fusion receptor generated as a result of a variant t(11;17) chromosomal translocation that occurs in a small subset of acute promyelocytic leukemia (APL) patients, has been shown to display a dominant-negative effect against the wild-type RARalpha/retinoid X receptor alpha (RXRalpha). We now show that its N-terminal region (called the POZ-domain), which mediates protein-protein interaction as well as specific nuclear localization of the wild-type PLZF and chimeric PLZF-RARalpha proteins, is primarily responsible for this activity. To further investigate the mechanisms of PLZF-RARalpha action, we have also studied its ligand-receptor, protein-protein, and protein-DNA interaction properties and compared them with those of the promyelocytic leukemia gene (PML)-RARalpha, which is expressed in the majority of APLs as a result of t(15;17) translocation. PLZF-RARalpha and PML-RARalpha have essentially the same ligand-binding affinities and can bind in vitro to retinoic acid response elements (RAREs) as homodimers or heterodimers with RXRalpha. PLZF-RARalpha homodimerization and heterodimerization with RXRalpha were primarily mediated by the POZ-domain and RARalpha sequence, respectively. Despite having identical RARalpha sequences, PLZF-RARalpha and PML-RARalpha homodimers recognized with different affinities distinct RAREs. Furthermore, PLZF-RARalpha could heterodimerize in vitro with the wild-type PLZF, suggesting that it may play a role in leukemogenesis by antagonizing actions of not only the retinoid receptors but also the wild-type PLZF and possibly other POZ-domain-containing regulators. These different protein-protein interactions and the target gene specificities of PLZF-RARalpha and PML-RARalpha may underlie, at least in part, the apparent resistance of APL with t(11;17) to differentiation effects of all-trans-retinoic acid.

Z-membranes: artificial organelles for overexpressing recombinant integral membrane proteins.

We have expressed a fusion protein formed between the avian infectious bronchitis virus M protein and the bacterial enzyme beta-glucuronidase in transgenic tobacco cells. Electron microscope images of such cells demonstrate that overexpression of this fusion protein gives rise to a type of endoplasmic reticulum membrane domain in which adjacent membranes become zippered together apparently as a consequence of the oligomerizing action of beta-glucuronidase. These zippered (Z-) membranes lack markers of the endoplasmic reticulum (NADH cytochrome c reductase and ribosomes) and accumulate in the cells in the form of multilayered scroll-like structures (up to 2 micrometers in diameter; 20-50 per cell) without affecting plant growth. The discovery of Z-membranes has broad implications for biology and biotechnology in that they provide a means for accumulating large quantities of recombinant membrane proteins within discrete domains of native membranes.