Two functionally distinct RNA-binding motifs in the regulatory domain of the protein kinase DAI.

The RNA-binding domain of the protein kinase DAI, the double-stranded RNA inhibitor of translation, contains two repeats of a motif that is also found in a number of other RNA-binding proteins. This motif consists of 67 amino acid residues and is predicted to contain a positively charged alpha helix at its C terminus. We have analyzed the effects of equivalent single amino acid changes in three conserved residues distributed over each copy of the motif. Mutants in the C-terminal portion of either repeat were severely defective, indicating that both copies of the motif are essential for RNA binding. Changes in the N-terminal and central parts of the motif were more debilitating if they were made in the first motif than in the second, suggesting that the first motif is the more important for RNA binding and that the second motif is structurally more flexible. When the second motif was replaced by a duplicate of the first motif, the ectopic copy retained its greater sensitivity to mutation, implying that the two motifs have distinct functions with respect to the process of RNA binding. Furthermore, the mutations have the same effect on the binding of double-stranded RNA and VA RNA, consistent with the existence of a single RNA-binding domain for both activating and inhibitory RNAs.     

Preferential selection of adenosines for modification by double-stranded RNA adenosine deaminase.

Double-stranded RNA adenosine deaminase (dsRAD), previously called the double-stranded RNA (dsRNA) unwinding/modifying activity, modifies adenosines to inosines within dsRNA. We used ribonuclease U2 and a mutant of ribonuclease T1 to map the sites of modification in several RNA duplexes. We found that dsRAD had a 5' neighbor preference (A = U > C > G) but no apparent 3' neighbor preference. Further, the proximity of the strand termini affected whether an adenosine was modified. Most importantly, dsRAD exhibited selectivity, modifying a minimal number of adenosines in short dsRNAs. Our results suggest that the specific editing of glutamate receptor subunit B mRNA could be performed in vivo by dsRAD without the aid of specificity factors, and support the hypothesis that dsRAD is responsible for hypermutations in certain RNA viruses.     

Substrate-dependent contribution of double-stranded RNA-binding motifs to ADAR2 function

ADAR2 is a double-stranded RNA-specific adenosine deaminase involved in the editing of mammalian RNAs by the site-specific conversion of adenosine to inosine (A-to-I). ADAR2 contains two tandem double-stranded RNA-binding motifs (dsRBMs) that are not only important for efficient editing of RNA substrates but also necessary for localizing ADAR2 to nucleoli. The sequence and structural similarity of these motifs have raised questions regarding the role(s) that each dsRBM plays in ADAR2 function. Here, we demonstrate that the dsRBMs of ADAR2 differ in both their ability to modulate subnuclear localization as well as to promote site-selective A-to-I conversion. Surprisingly, dsRBM1 contributes to editing activity in a substrate-dependent manner, indicating that dsRBMs recognize distinct structural determinants in each RNA substrate. Although dsRBM2 is essential for the editing of all substrates examined, a point mutation in this motif affects editing for only a subset of RNAs, suggesting that dsRBM2 uses unique sets of amino acid(s) for functional interactions with different RNA targets. The dsRBMs of ADAR2 are interchangeable for subnuclear targeting, yet such motif alterations do not support site-selective editing, indicating that the unique binding preferences of each dsRBM differentially contribute to their pleiotropic function.     

Deduced amino acid sequence of Escherichia coli adenosine deaminase reveals evolutionarily conserved amino acid residues: implications for catalytic function

The goal of the research reported here is to identify evolutionarily conserved amino acid residues associated with enzymatic deamination of adenosine. To do this, we isolated molecular clones of the Escherichia coli adenosine deaminase gene by functional complementation of adenosine deaminase deficient bacteria and deduced the amino acid sequence of the enzyme from the nucleotide sequence of the gene. Nucleotide sequence analysis revealed the presence of a 996-nucleotide open reading frame encoding a protein of 332 amino acids having a molecular weight of 36,345. The deduced amino acid sequence of the E. coli enzyme has approximately 33% identity with those of the mammalian adenosine deaminases. With conservative amino acid substitutions the overall sequence homology approaches 50%, suggesting that the structures and functions of the mammalian and bacterial enzymes are similar. Additional amino acid sequence analysis revealed specific residues that are conserved among all three adenosine deaminases and four AMP deaminases for which sequence information is currently available. In view of previously published enzymological data and the conserved amino acid residues identified in this study, we propose a model to account for the enzyme-catalyzed hydrolytic deamination of adenosine. Potential catalytic roles are assigned to the conserved His 214, Cys 262, Asp 295, and Asp 296 residues of mammalian adenosine deaminases and the corresponding conserved amino acid residues in bacterial adenosine deaminase and the eukaryotic AMP deaminases.     

Evidence for auto-inhibition by the N terminus of hADAR2 and activation by dsRNA binding

Adenosine deaminases that act on RNA (ADARs) catalyze adenosine to inosine conversion in RNA that is largely double stranded. Human ADAR2 (hADAR2) contains two double-stranded RNA binding motifs (dsRBMs), separated by a 90-amino acid linker, and these are followed by the C-terminal catalytic domain. We assayed enzymatic activity of N-terminal deletion constructs of hADAR2 to determine the role of the dsRBMs and the intervening linker peptide. We found that a truncated protein consisting of one dsRBM and the deaminase domain was capable of deaminating a short 15-bp substrate. In contrast, full-length hADAR2 was inactive on this short substrate. In addition, we observed that the N terminus, which was deleted from the truncated protein, inhibits editing activity when added in trans. We propose that the N-terminal domain of hADAR2 contains sequences that cause auto-inhibition of the enzyme. Our results suggest activation requires binding to an RNA substrate long enough to accommodate interactions with both dsRBMs.     

Mutation of a highly conserved aspartate residue in subdomain IX abolishes Fer protein-tyrosine kinase activity

Before the structure of cAMP-dependent protein kinase had been solved, sequence alignments had already suggested that several highly conserved peptide motifs described as kinase subdomains I through XI might play some functional role in catalysis. Crystal structures of several members of the protein kinase superfamily have suggested that the nearly invariant aspartate residue within subdomain IX contributes to the conformational stability of the catalytic loop by forming hydrogen bonds with backbone amides within subdomain VI. However, substitution of this aspartate with alanine or threonine in some protein kinases have indicated that these interactions are not essential for activity. In contrast, we show here that conversion of this aspartate to arginine abolished the catalytic activity of the Fer protein-tyrosine kinase when expressed either in mammalian cells or in bacteria. Structural modeling predicted that the catalytic loop of the FerD743R mutant was disrupted by van der Waal's repulsion between the side chains of the substituted arginine residue in subdomain IX and histidine-683 in subdomain VI. The FerD743R mutant model predicted a shift in the peptide backbone of the catalytic loop, and an outward rotation of histidine-683 and arginine-684 side chains. However, the position and orientation of the presumptive catalytic base, aspartate-685, was not substantially changed. The proposed model explains how substitutions of some, but not all residues could be tolerated at this nearly invariant aspartate in kinase subdomain IX.     

Analysis of Xenopus dsRNA adenosine deaminase cDNAs reveals similarities to DNA methyltransferases.

We isolated two similar, but distinct, cDNA classes that encode Xenopus double-stranded RNA (dsRNA) adenosine deaminase. The longest, full-length open reading frame (ORF) predicts a 1,270-amino acid protein of 138,754 Da that is similar in size and about 50% identical to proteins encoded by mammalian cDNAs, yet larger than the 120-kDa protein purified from Xenopus eggs. Alignments of the Xenopus and mammalian ORFs show N-terminal heterogeneity, three conserved dsRNA binding motifs (dsRBMs), and strongly conserved carboxyl termini. Consistent with the observation of two cDNA classes, northern analyses of Xenopus oocyte poly A+ RNA show at least three mRNA species. Multiple nuclear polyadenylation hexamers and putative cytoplasmic polyadenylation elements were found in the 3'' UTRs of cDNAs corresponding to the largest mRNA. In vitro translation experiments show that the cDNAs encode active deaminases and that the entire N-terminus and first dsRBM are dispensable for deaminase activity. Importantly, an analysis of the C-termini of five known dsRNA adenosine deaminases, and two putative deaminases, reveals motifs that are strikingly similar to the conserved motifs of the DNA-(adenine-N6alpha)-aminomethyltransferases and the DNA-(cytosine-5)-methyltransferases.     

Identification of domains in apobec-1 complementation factor required for RNA binding and apolipoprotein-B mRNA editing

The C-to-U editing of apolipoprotein-B (apo-B) mRNA is catalyzed by an enzyme complex that recognizes an 11-nt mooring sequence downstream of the editing site. A minimal holoenzyme that edits apo-B mRNA in vitro has been defined. This complex contains apobec-1, the catalytic subunit, and apobec-1 complementation factor (ACF), the RNA-binding subunit that binds to the mooring sequence. Here, we show that ACF binds with high affinity to single-stranded but not double-stranded apo-B mRNA. ACF contains three nonidentical RNA recognition motifs (RRM) and a unique C-terminal auxiliary domain. In many multi-RRM proteins, the RRMs mediate RNA binding and an auxiliary domain functions in protein-protein interactions. Here we show that ACF does not fit this simple model. Based on deletion mutagenesis, the RRMs in ACF are necessary but not sufficient for binding to apo-B mRNA. Amino acids in the pre-RRM region are required for complementing activity and RNA binding, but not for interaction with apobec-1. The C-terminal 196 amino acids are not absolutely essential for function. However, further deletion of an RG-rich region from the auxiliary domain abolished complementing activity, RNA binding, and apobec-1 interaction. The auxiliary domain alone did not bind apobec-1. Although all three RRMs are required for complementing activity and apobec-1 interaction, the individual motifs contribute differently to RNA binding. Point mutations in RRM1 or RRM2 decreased the Kd for apo-B mRNA by two orders of magnitude whereas mutations in RRM3 reduced binding affinity 13-fold. The pairwise expression of RRM1 with RRM2 or RRM3 resulted in moderate affinity binding.     

Functional importance of the conserved N-terminal domain of the mitochondrial replicative DNA helicase

The mitochondrial replicative DNA helicase is an essential cellular protein that shows high similarity with the bifunctional primase-helicase of bacteriophage T7, the gene 4 protein (T7 gp4). The N-terminal primase domain of T7 gp4 comprises seven conserved sequence motifs, I, II, III, IV, V, VI, and an RNA polymerase basic domain. The putative primase domain of metazoan mitochondrial DNA helicases has diverged from T7 gp4 and in particular, the primase domain of vertebrates lacks motif I, which comprises a zinc binding domain. Interestingly, motif I is conserved in insect mtDNA helicases. Here, we evaluate the effects of overexpression in Drosophila cell culture of variants carrying mutations in conserved amino acids in the N-terminal region, including the zinc binding domain. Overexpression of alanine substitution mutants of conserved amino acids in motifs I, IV, V and VI and the RNA polymerase basic domain results in increased mtDNA copy number as is observed with overexpression of the wild type enzyme. In contrast, overexpression of three N-terminal mutants W282L, R301Q and P302L that are analogous to human autosomal dominant progressive external ophthalmoplegia mutations results in mitochondrial DNA depletion, and in the case of R301Q, a dominant negative cellular phenotype. Thus whereas our data suggest lack of a DNA primase activity in Drosophila mitochondrial DNA helicase, they show that specific N-terminal amino acid residues that map close to the central linker region likely play a physiological role in the C-terminal helicase function of the protein.     

The interaction between Z-DNA and the Zab domain of double-stranded RNA adenosine deaminase characterized using fusion nucleases.

Zab is a structurally defined protein domain that binds specifically to DNA in the Z conformation. It consists of amino acids 133-368 from the N terminus of human double-stranded RNA adenosine deaminase, which is implicated in RNA editing. Zab contains two motifs with related sequence, Zalpha and Zbeta. Zalpha alone is capable of binding Z-DNA with high affinity, whereas Zbeta alone has little DNA binding activity. Instead, Zbeta modulates Zalpha binding, resulting in increased sequence specificity for alternating (dCdG)n as compared with (dCdA/dTdG)n. This relative specificity has previously been demonstrated with short oligonucleotides. Here we demonstrate that Zab can also bind tightly to (dCdG)n stabilized in the Z form in supercoiled plasmids. Binding was assayed by monitoring cleavage of the plasmids using fusion nucleases, in which Z-DNA-binding peptides from the N terminus of double-stranded RNA adenosine deaminase are linked to the nuclease domain of FokI. A fusion nuclease containing Zalpha shows less sequence specificity, as well as less conformation specificity, than one containing Zab. Further, a construct in which Zbeta has been replaced in Zab with Zalpha, cleaves Z-DNA regions in supercoiled plasmids more efficiently than the wild type but with little sequence specificity. We conclude that in the Zab domain, both Zalpha and Zbeta contact DNA. Zalpha contributes contacts that produce conformation specificity but not sequence specificity. In contrast, Zbeta contributes weakly to binding affinity but discriminates between sequences of Z-DNAs.     

Characterization of the functional domains of Escherichia coli RNase II

RNase II is a single-stranded-specific 3'-exoribonuclease that degrades RNA generating 5'-mononucleotides. This enzyme is the prototype of an ubiquitous family of enzymes that are crucial in RNA metabolism and share a similar domain organization. By sequence prediction, three different domains have been assigned to the Escherichia coli RNase II: two RNA-binding domains at each end of the protein (CSD and S1), and a central RNB catalytic domain. In this work we have performed a functional characterization of these domains in order to address their role in the activity of RNase II. We have constructed a large set of RNase II truncated proteins and compared them to the wild-type regarding their exoribonucleolytic activity and RNA-binding ability. The dissociation constants were determined using different single- or double-stranded substrates. The results obtained revealed that S1 is the most important domain in the establishment of stable RNA-protein complexes, and its elimination results in a drastic reduction on RNA-binding ability. In addition, we also demonstrate that the N-terminal CSD plays a very specific role in RNase II, preventing a tight binding of the enzyme to single-stranded poly(A) chains. Moreover, the biochemical results obtained with RNB mutant that lacks both putative RNA-binding domains, revealed the presence of an additional region involved in RNA binding. Such region, was identified by sequence analysis and secondary structure prediction as a third putative RNA-binding domain located at the N-terminal part of RNB catalytic domain.     

Biological Activities of Murine Low-Affinity Fc Receptors for IgG

Murine low-affinity Fc receptors for IgG (FcγRIIbl, FcγRIIb2, and FcγRIII) bind the same IgG subclasses and are not distinguished by available anti-FcγRII/III mAbs (2.4G2). They trigger various biological activities, among which are the internalization of soluble and particulate immune complexes, cell activation, and its regulation. To determine the biological properties of the three murine receptors, each was expressed by stable transfection of corresponding cDNAs in two model cells: the murine lymphoma B cell IIA1.6 and the rat basophilic leukemia cell RBL-2H3. Biological activities of recombinant receptors were triggered with soluble immune complexes or 2.4G2 IgG in IIA1.6 cells, which express no FcγR, and with 2.4G2 Fab or F(ab′)₂, cross-linked with mouse anti-rat F(ab′)₂ in RBL, which express rat FcγR. Conditions for studying cell activation and endocytosis in both cell models are described, as are conditions for studying phagocytosis in RBL cells and antigen presentation or regulation of cell activation in IIA1.6 cells. Internalization of immune complexes was triggered by FcγRIIb2 and FcγRIII, but not by FcγRIIb1. Intracytoplasmic sequences required for phagocytosis and endocytosis could be distinguished in FcγRIIb2, but not in FcγRIII. Cell activation was restricted to FcγRIII. FcγRIII-mediated endocytosis, phagocytosis, and cell activation involved the consensus tyrosine-containing activation motif found in the intracytoplasmic domain of the γ subunit. Regulation of cell activation was induced by both FcγRII isoforms and depended on the same sequence as endocytosis. As a consequence, a single motif can determine more than one biological response of the cell, and a given response may be triggered by several motifs, borne by different FcγR.     

A new alpha-helical extension promotes RNA binding by the dsRBD of Rnt1p RNAse III

Rnt1 endoribonuclease, the yeast homolog of RNAse III, plays an important role in the maturation of a diverse set of RNAs. The enzymatic activity requires a conserved catalytic domain, while RNA binding requires the double-stranded RNA-binding domain (dsRBD) at the C-terminus of the protein. While bacterial RNAse III enzymes cleave double-stranded RNA, Rnt1p specifically cleaves RNAs that possess short irregular stem-loops containing 12–14 base pairs interrupted by internal loops and bulges and capped by conserved AGNN tetraloops. Consistent with this substrate specificity, the isolated Rnt1p dsRBD and the 30–40 amino acids that follow bind to AGNN-containing stem-loops preferentially in vitro. In order to understand how Rnt1p recognizes its cognate processing sites, we have defined its minimal RNA-binding domain and determined its structure by solution NMR spectroscopy and X-ray crystallography. We observe a new carboxy-terminal helix following a canonical dsRBD structure. Removal of this helix reduces binding to Rnt1p substrates. The results suggest that this helix allows the Rnt1p dsRBD to bind to short RNA stem-loops by modulating the conformation of helix α1, a key RNA-recognition element of the dsRBD.     

An ADAR that edits transcripts encoding ion channel subunits functions as a dimer

In this report, we establish that Drosophila ADAR (adenosine deaminase acting on RNA) forms a dimer on double-stranded (ds) RNA, a process essential for editing activity. The minimum region required for dimerization is the N-terminus and dsRNA-binding domain 1 (dsRBD1). Single point mutations within dsRBD1 abolish RNA-binding activity and dimer formation. These mutations and glycerol gradient analysis indicate that binding to dsRNA is important for dimerization. However, dimerization can be uncoupled from dsRNA-binding activity, as a deletion of the N-terminus (amino acids 1-46) yields a monomeric ADAR that retains the ability to bind dsRNA but is inactive in an editing assay, demonstrating that ADAR is only active as a dimer. Different isoforms of ADAR with different editing activities can form heterodimers and this can have a significant effect on editing in vitro as well as in vivo. We propose a model for ADAR dimerization whereby ADAR monomers first contact dsRNA; however, it is only when the second monomer binds and a dimer is formed that deamination occurs.     

DNA stabilization at the Bacillus subtilis PolX core��a binding model to coordinate polymerase, AP-endonuclease and 3��-5�� exonuclease activities

Family X DNA polymerases (PolXs) are involved in DNA repair. Their binding to gapped DNAs relies on two conserved helix-hairpin-helix motifs, one located at the 8-kDa domain and the other at the fingers subdomain. Bacterial/archaeal PolXs have a specifically conserved third helix-hairpin-helix motif (GFGxK) at the fingers subdomain whose putative role in DNA binding had not been established. Here, mutagenesis at the corresponding residues of Bacillus subtilis PolX (PolXBs), Gly130, Gly132 and Lys134 produced enzymes with altered DNA binding properties affecting the three enzymatic activities of the protein: polymerization, located at the PolX core, 3′-5′ exonucleolysis and apurinic/apyrimidinic (AP)-endonucleolysis, placed at the so-called polymerase and histidinol phosphatase domain. Furthermore, we have changed Lys192 of PolXBs, a residue moderately conserved in the palm subdomain of bacterial PolXs and immediately preceding two catalytic aspartates of the polymerization reaction. The results point to a function of residue Lys192 in guaranteeing the right orientation of the DNA substrates at the polymerization and histidinol phosphatase active sites. The results presented here and the recently solved structures of other bacterial PolX ternary complexes lead us to propose a structural model to account for the appropriate coordination of the different catalytic activities of bacterial PolXs.     

Differential binding of cAMP-dependent protein kinase regulatory subunit isoforms Ialpha and IIbeta to the catalytic subunit

Limited trypsin digestion of type I cAMP-dependent protein kinase holoenzyme results in a proteolytic-resistant Delta(1-72) regulatory subunit core, indicating that interaction between the regulatory and catalytic subunits extends beyond the autoinhibitory site in the R subunit at the NH(2) terminus. Sequence alignment of the two R subunit isoforms, RI and RII, reveals a significantly sequence diversity at this specific region. To determine whether this sequence diversity is functionally important for interaction with the catalytic subunit, specific mutations, R133A and D328A, are introduced into sites adjacent to the active site cleft in the catalytic subunit. While replacing Arg(133) with Ala decreases binding affinity for RII, interaction between the catalytic subunit and RI is not affected. In contrast, mutant C(D328A) showed a decrease in affinity for binding RI while maintaining similar affinities for RII as compared with the wild-type catalytic subunit. These results suggest that sequence immediately NH(2)-terminal to the consensus inhibition site in RI and RII interacts with different sites at the proximal region of the active site cleft in the catalytic subunit. These isoform-specific differences would dictate a significantly different domain organization in the type I and type II holoenzymes.     

Identification of potential substrate-binding sites in yeast and human acyl-CoA sterol acyltransferases by mutagenesis of conserved sequences

In mammals, the esterification of sterols by ACAT plays a critical role in eukaryotic lipid homeostasis. Using the predominant isoform of the yeast ACAT-related enzyme family, Are2p, as a model, we targeted phylogenetically conserved sequences for mutagenesis in order to identify functionally important motifs. Deletion, truncation, and missense mutations implicate a regulatory role for the amino-terminal domain of Are2p and identified two carboxyl-terminal motifs as required for catalytic activity. A serine-to-leucine mutation in the (H/Y)SF motif (residues 338-340), unique to sterol esterification enzymes, nullified the activity and stability of yeast Are2p. Similarly, a tyrosine-to-alanine change in the FYxDWWN motif of Are2p (residues 523-529) produced an enzyme with decreased activity and apparent affinity for oleoyl-CoA. Mutagenesis of the tryptophan residues in this motif completely abolished activity. In human ACAT1, mutagenesis of the corresponding motifs (residues 268-270, and 403-409, respectively) also nullified enzymatic activity. On the basis of their critical roles in enzymatic activity and their sequence conservation, we propose that these motifs mediate sterol and acyl-CoA binding by this class of enzymes.     

Structural requirements for double-stranded RNA binding, dimerization, and activation of the human eIF-2 alpha kinase DAI in Saccharomyces cerevisiae.

The protein kinase DAI is activated upon viral infection of mammalian cells and inhibits protein synthesis by phosphorylation of the alpha subunit of translation initiation factor 2 (eIF-2 alpha). DAI is activated in vitro by double-stranded RNAs (dsRNAs), and binding of dsRNA is dependent on two copies of a conserved sequence motif located N terminal to the kinase domain in DAI. High-level expression of DAI in Saccharomyces cerevisiae cells is lethal because of hyperphosphorylation of eIF-2 alpha; at lower levels, DAI can functionally replace the protein kinase GCN2 and stimulate translation of GCN4 mRNA. These two phenotypes were used to characterize structural requirements for DAI function in vivo, by examining the effects of amino acid substitutions at matching positions in the two dsRNA-binding motifs and of replacing one copy of the motif with the other. We found that both copies of the dsRNA-binding motif are required for high-level kinase function and that the N-terminal copy is more important than the C-terminal copy for activation of DAI in S. cerevisiae. On the basis of these findings, we conclude that the requirements for dsRNA binding in vitro and for activation of DAI kinase function in vivo closely coincide. Two mutant alleles containing deletions of the first or second binding motif functionally complemented when coexpressed in yeast cells, strongly suggesting that the active form of DAI is a dimer. In accord with this conclusion, overexpression of four catalytically inactive alleles containing different deletions in the protein kinase domain interfered with wild-type DAI produced in the same cells. Interestingly, three inactivating point mutations in the kinase domain were all recessive, suggesting that dominant interference involves the formation of defective heterodimers rather than sequestration of dsRNA activators by mutant enzymes. We suggest that large structural alterations in the kinase domain impair an interaction between the two protomers in a DAI dimer that is necessary for activation by dsRNA or for catalysis of eIF-2 alpha phosphorylation.