Structural insights into abasic site for Fpg specific binding and catalysis: comparative high-resolution crystallographic studies of Fpg bound to various models of abasic site analogues-containing DNA

Fpg is a DNA glycosylase that recognizes and excises the mutagenic 8-oxoguanine (8-oxoG) and the potentially lethal formamidopyrimidic residues (Fapy). Fpg is also associated with an AP lyase activity which successively cleaves the abasic (AP) site at the 3′ and 5′ sides by βδ-elimination. Here, we present the high-resolution crystal structures of the wild-type and the P1G defective mutant of Fpg from Lactococcus lactis bound to 14mer DNA duplexes containing either a tetrahydrofuran (THF) or 1,3-propanediol (Pr) AP site analogues. Structures show that THF is less extrahelical than Pr and its backbone C5′–C4′–C3′ diverges significantly from those of Pr, rAP, 8-oxodG and FapydG. Clearly, the heterocyclic oxygen of THF is pushed back by the carboxylate of the strictly conserved E2 residue. We can propose that the ring-opened form of the damaged deoxyribose is the structure active form of the sugar for Fpg catalysis process. Both structural and functional data suggest that the first step of catalysis mediated by Fpg involves the expulsion of the O4′ leaving group facilitated by general acid catalysis (involving E2), rather than the immediate cleavage of the N-glycosic bond of the damaged nucleoside.     

8-OxoA inhibits the incision of an AP site by the DNA glycosylases Fpg, Nth and the AP endonuclease HAP1

Ionizing radiation induces clustered DNA damage sites, whereby two or more individual DNA lesions are formed within one or two helical turns of DNA by a single radiation track. A subset of DNA clustered damage sites exist in which the lesions are located in tandem on the same DNA strand. Recent studies have established that two closely opposed lesions impair the repair machinery of the cell, but few studies have investigated the processing of tandem lesions. In this study, synthetic double-stranded oligonucleotides were synthesized to contain 8-oxoA and an AP site in tandem, separated by up to four bases in either a 5' or 3' orientation. The influence 8-oxoA has on the incision of the AP site by the E. coli glycosylases Fpg and Nth protein and the human AP endonuclease HAP1 was assessed. 8-OxoA has little or no effect on the efficiency of incision of the AP site by Nth protein; however, the efficiency of incision of the AP site by Fpg protein is reduced in the presence of 8-oxoA even up to a four-base separation in both the 5' and 3' orientations. 8-OxoA influences the efficiency of HAP1 incision of the AP site only when it is 3' to the AP site and separated by up to two bases. This study demonstrates that the initial stages of base excision repair can be impaired by the presence of a second base lesion in proximity to an AP site on the same DNA strand. This impairment could have biological consequences, such as mutation induction, if the AP site is present at replication.     

Sequence and structural determinants of mannose recognition

Mannose, an abundant cell surface monosaccharide binds to a diverse set of receptors, which are involved in a variety of important cellular processes. Structural analysis has been carried out on all the proteins containing non-covalently bound mannose as a monosaccharide in the Protein Data Bank, to identify common recognition principles. Proteins, highly specific to mannose, belonging to the super family of bulb lectins, are found to contain a consensus sequence motif QXDXNXVXY, which has been identified to be essential for mannose binding. Analysis of this motif in the crystal structures of bulb lectins has led to the understanding of the contribution of individual residues in mannose recognition. Comparison with other mannose binding proteins, reveals common hydrogen bonding patterns in all of them, despite differences in sequence, overall fold and the substructures at the binding sites of individual proteins. A database analysis also suggests that although the topology of the backbone, as at the binding site in bulb lectins, can generate mannose binding capability in a few other proteins, sequence and disposition of not only the residues in the motif, but also the residues in the neighborhood play a crucial role in allowing that property to be retained.     

Unusual structural features of hydantoin lesions translate into efficient recognition by Escherichia coli Fpg

Oxidation of guanine (G) and 8-oxoguanine (OG) with a wide variety of oxidants yields the hydantoin lesions, guanidinohydantoin (Gh) and spiroiminodihydantoin (Sp). These two lesions have garnered much recent attention due to their unusual structures and high mutagenic potential. We have previously shown that duplexes containing Gh and Sp are substrates for the base excision repair glycosylase Escherichia coli Fpg (EcFpg). To evaluate the recognition features of these unusual lesions, binding and footprinting experiments were performed using a glycosylase inactive variant, E3Q EcFpg, and 30 bp duplexes containing the embedded lesions. Surprisingly, E3Q EcFpg was found to bind significantly more tightly ( approximately 1000-fold) to duplexes containing Gh or Sp over the corresponding duplexes containing OG. This may be a consequence of the helix-destabilizing nature of the hydantoin lesions that facilitates their recognition within duplex DNA. Though DNA binding affinities of E3Q EcFpg with Gh- and Sp-containing duplexes were found to be similar to each other, hydroxyl radical footprinting using methidium-propyl-EDTA (MPE)-Fe(II) revealed subtle differences between binding of E3Q EcFpg to the two lesions. Most notably, in the presence of E3Q EcFpg, the Sp nucleotide (nt) is hyperreactive toward cleavage by MPE-Fe(II)-generated hydroxyl radicals, suggestive of the formation of an intercalation site for the MPE-Fe(II) reagent at the Sp nt. Interestingly, increasing the duplex length from 18 to 30 bp enhanced the excision efficiency of Gh and Sp paired with C, G, or T by EcFpg such that these substrates are processed as efficiently as the signature substrate lesion, OG. Moreover, the base removal activity with these two lesions was more efficient than removal of OG when in a base pairing context opposite A. The high affinity and efficient activity of EcFpg toward the hydantoin lesions suggest that EcFpg mediates repair of the lesions in vivo. Notably, the facile activity of EcFpg toward Gh and Sp in base pairing contexts with G and A, which are likely to be present after DNA replication, would be detrimental and enhance mutagenesis.     

APE-type non-LTR retrotransposons: determinants involved in target site recognition

Non-long terminal repeat (Non-LTR) retrotransposons represent a diverse and widely distributed group of transposable elements and an almost ubiquitous component of eukaryotic genomes that has a major impact on evolution. Their copy number can range from a few to several million and they often make up a significant fraction of the genomes. The members of the dominating subtype of non-LTR retrotransposons code for an endonuclease with homology to apurinic/apyrimidinic endonucleases (APE), and are thus termed APE-type non-LTR retrotransposons. In the last decade both the number of identified non-LTR retrotransposons and our knowledge of biology and evolution of APE-type non-LTR retrotransposons has increased tremendously.     

Asymmetrical recognition of the palindromic AP1 binding site (TRE) by Fos protein complexes.

Fos and Jun proteins form a tight complex which binds specifically to the AP1 recognition sequence, a palindromic DNA element also referred to as the TPA responsive element (TRE). To elucidate the mechanism of Fos-Jun interaction with the TRE we have performed UV cross-linking studies using oligonucleotides where thymines were replaced with bromouracil. Our results indicate that both Fos and Jun directly contact the TRE but that the interaction of Fos and Jun with thymines in structurally equivalent positions in the two half sites of the TRE is different. In addition, we have carried out a comprehensive mutagenesis study of the TRE by introducing all possible point mutations plus thymine----uracil substitutions into the palindromic TRE core sequences and the adjacent nucleotides on both sides. The results of this analysis clearly show that the palindromic TRE is asymmetrical with respect to binding of Fos-Jun. We also show that a Fos protein complex with a homodimeric DNA binding site binds considerably less efficiently to TRE mutants with a perfect dyad symmetry compared with the binding to the wild-type TRE. This demonstrates that the asymmetrical recognition of the TRE is not due to the heterodimeric nature of the Fos/Jun complex but directly related to an asymmetry in the TRE sequence. The methyl groups of all four thymine residues within the TRE seem to be functionally crucial since thymine----uracil substitutions strongly reduce or abolish binding to Fos/Jun. The relevance of structurally equivalent methyl groups in the TRE core sequence is different, lending further support to the conclusion that the TRE is asymmetrical.     

Multiple components are required for sequence recognition of the AP1 site in the gibbon ape leukemia virus enhancer.

At least two subunits contributed to the formation in vitro of a specific complex binding to the AP1 consensus sequence (TGAGTCA) in the gibbon ape leukemia virus (GALV) enhancer in MLA144 cells. This complex can be dissociated on a monomeric GALV oligonucleotide affinity column. One protein, termed the core protein, was retained on the oligonucleotide affinity column. The second protein flowed through the oligonucleotide affinity column and, when alone, did not bind to DNA; however, when present with the core protein, it bound strongly and very specifically to the GALV sequence. MLA144 cells contained only trace amounts of c-fos and c-jun by immunoblot analysis, suggesting that the proteins specifically binding to the GALV AP1 site were distinct from c-fos and c-jun. In addition to the major complex that recognized the GALV element, MLA144 cells contained a minor complex that is chromatographically different from and antigenically related to c-fos. The factor in the flowthrough complemented a human T-cell nuclear extract (Jurkat cell line), which, when alone, had no assayable complex that specifically bound to the GALV enhancer; this complementation gave rise to a specific complex similar to that seen in MLA144 cells. Together, these results suggest that the GALV enhancer can interact with multicomponent protein complexes in a cell-line-specific manner.     

Elusive recognition determinants for ubiquitination

How are proteins recognized as substrates for ubiquitination? Here we summarize insights from recent experiments that address this issue. These highlight the diversity and complexity of determinants for substrate recognition, and raise many questions for further investigation. Copyright © 2002 John Wiley & Sons, Ltd.     

The molecular determinants of NEDD8 specific recognition by human SENP8

Although neuronal-precursor-cell-expressed developmentally downregulated protein-8 (NEDD8) and ubiquitin share the highest level of sequence identity and structural similarity among several known ubiquitin-like proteins, their conjugation to a protein leads to distinct biological consequences. In the study, we first identified the NEDD8 protein of Chlamydomonas reinhardtii (CrNEDD8) and discovered that CrNEDD8 is fused at the C-terminus of a ubiquitin moiety (CrUb) in a head-to-tail arrangement. This CrUb-CrNEDD8 protein was termed CrRUB1 (related to ubiquitin 1) by analogy with a similar protein in Arabidopsis thaliana (AtRUB1). Since there is high sequence identity in comparison to the corresponding human proteins (97% for ubiquitin and 84% for NEDD8), a His-CrRUB1-glutathione S-transferase (GST) fusion construct was adopted as the alternative substrate to characterize the specificity of NEDD8-specific peptidase SENP8 for CrNEDD8. The data showed that SENP8 only cleaved the peptide bond beyond the di-glycine motif of CrNEDD8 and His-RUB1 was subsequently generated, confirming that SENP8 has exquisite specificity for CrNEDD8 but not CrUb. To further determine the basis of this specificity, site-directed mutagenesis at earlier reported putative molecular determinants of NEDD8 specific recognition by SENP8 was performed. We found that a single N51E mutation of CrNEDD8 completely inhibited its hydrolysis by SENP8. Conversely, a single E51N mutation of CrUb enabled this ubiquitin mutant to undergo hydrolysis by SENP8, revealing that a single residue difference at the position 51 contributes substantially to the substrate selectivity of SENP8. Moreover, the E51N/R72A double mutant of the CrUb subdomain can further increase the efficiency of cleavage by SENP8, indicating that the residue at position 72 is also important in substrate recognition. The E51N or R72A mutation of CrUb also inhibited the hydrolysis of CrUb by ubiquitin-specific peptidase USP2. However, USP2 cannot cleave the N51E/A72R double mutant of the CrNEDD8 subdomain, suggesting that USP2 requires additional recognition sites.     

Identification of the determinants for the specific recognition of single-strand telomeric DNA by Cdc13

The single-strand overhang present at telomeres plays a critical role in mediating both the capping and telomerase regulation functions of telomeres. The telomere end-binding proteins, Cdc13 in Saccharomyces cerevisiae, Pot1 in higher eukaryotes, and TEBP in the ciliated protozoan Oxytricha nova, exhibit sequence-specific binding to their respective single-strand overhangs. S. cerevisiae telomeres are composed of a heterogeneous mixture of GT-rich telomeric sequence, unlike in higher eukaryotes which have a simple repeat that is maintained with high fidelity. In yeast, the telomeric overhang is recognized by the essential protein Cdc13, which coordinates end-capping and telomerase activities at the telomere. The Cdc13 DNA-binding domain (Cdc13-DBD) binds these telomere sequences with high affinity (3 pM) and sequence specificity. To better understand the basis for this remarkable recognition, we have investigated the binding of the Cdc13-DBD to a series of altered DNA substrates. Although an 11-mer of GT-rich sequence is required for full binding affinity, only three of these 11 bases are recognized with high specificity. This specificity differs from that observed in the other known telomere end-binding proteins, but is well suited to the specific role of Cdc13 at yeast telomeres. These studies expand our understanding of telomere recognition by the Cdc13-DBD and of the unique molecular recognition properties of ssDNA binding.     

Thermodynamic,kinetic, and structural basis for recognition and repair of 8-oxoguanine in DNA by Fpg protein from Escherichia coli

X-ray analysis does not provide quantitative estimates of the relative importance of the molecular contacts it reveals or of the relative contributions of specific and nonspecific interactions to the total affinity of specific DNA to enzymes. Stepwise increase of DNA ligand complexity has been used to estimate the relative contributions of virtually every nucleotide unit of 8-oxoguanine-containing DNA to its total affinity for Escherichia coli 8-oxoguanine DNA glycosylase (Fpg protein). Fpg protein can interact with up to 13 nucleotide units or base pairs of single- and double-stranded ribo- and deoxyribo-oligonucleotides of different lengths and sequences through weak additive contacts with their internucleotide phosphate groups. Bindings of both single-stranded and double-stranded oligonucleotides follow similar algorithms, with additive contributions to the free energy of binding of the structural components (phosphate, sugar, and base). Thermodynamic models are provided for both specific and nonspecific DNA sequences with Fpg protein. Fpg protein interacts nonspecifically with virtually all of the base-pair units within its DNA-binding cleft: this provides approximately 7 orders of magnitude of affinity (Delta G degrees approximately equal to -9.8 kcal/mol) for DNA. In contrast, the relative contribution of the 8-oxoguanine unit of the substrate (Delta G degrees approximately equal to -0.90 kcal/mol) together with other specific interactions is <2 orders of magnitude (Delta G degrees approximately equal to -2.8 kcal/mol). Michaelis complex formation of Fpg protein with DNA containing 8-oxoguanine cannot of itself provide the major part of the enzyme specificity, which lies in the k(cat) term; the rate is increased by 6-8 orders of magnitude on going from nonspecific to specific oligodeoxynucleotides.     

Immunological recognition of specific antigenic determinants in UV- and gamma-irradiated phage DNA

Thermally denatured DNA of coliphage T1 after treatment with uv-light (2537 A) and 60Co-gamma rays acts as a hapten with antigenic determinant groups specific for radiation-induced alterations of the macromolecule. After conjugation to methylated bovine serum albumin the DNA becomes immunogenic in rabbits. Antibodies against irradiated DNA do not react with unirradiated single-stranded DNA. Antigen-antibody complexes were demonstrated by CsCl-density gradient centrifugation. The decrease in buoyant density of the DNA is proportional to the amount of antibody protein bound to the antigen. By this means photoproducts as well as alterations due to ionizing radiation in DNA were detected independent of the type of antigen-antibody complex, i.e. precipitating or soluble aggregate.     

Delineation of the structural determinants of the N-methyl-D-aspartate receptor glycine binding site

In this study, we have further delineated the domains of the N-methyl-D-aspartate receptor NR1 subunit that contribute to the glycine co-agonist binding site. Taking an iterative approach, we have constructed truncation mutants of the NR1 subunit, transiently expressed them in HEK-293 cells, and determined the binding of the glycine site antagonist [3H]L-689,560. Amino acids 380-811 were sufficient to form a glycine binding site with affinities for [3H]L-689,560 and glycine that were not significantly different from wild-type NR1. More extensive deletions, from either the amino- or the carboxy-terminal end, resulted in loss of ligand binding. Additional constructs were made starting from amino acids 380-843 of NR1, replacing the transmembrane (TMI-TMIII) domain with intervening linker sequences while retaining the TMIV domain so as to anchor the polypeptide to the membrane. Although robust amounts of polypeptides were synthesised by transfected cells, only low levels of [3H]L-689,560 binding sites could be detected. This suggests that only a small proportion of the synthesised polypeptide folds in the appropriate manner so as to form a ligand binding site. These data indicate that although it is possible to reduce the glycine binding site to minimal so-called S1 and S2 domains, efficient folding of the polypeptide so as to form a ligand binding site may require sequences within the TMI-TMIII domain.     

Fine structural recognition specificities of IgE antibodies distinguishing amoxicilloyl and amoxicillanyl determinants in allergic subjects.

IgE antibodies in the sera of subjects allergic to beta-lactam antibiotics detect a spectrum of specificities ranging from side-chain groups to an entire penicillin or cephalosporin molecule. In addition to such structural heterogeneity of allergenic determinants, IgE antibodies in the sera of different allergic subjects show heterogeneous recognition responses. Detailed immunochemical studies were carried out on the sera of penicillin-allergic subjects that showed selective and unexpected reactions with the frequently prescribed penicillin, amoxicillin. Antibodies from one subject reacted only with the amoxicilloyl determinant while IgE from another subject showed multiple reactivity with penicilloyl and penicillanyl determinants of different penicillins but not with the amoxicilloyl determinant. Quantitative hapten inhibition studies revealed that the combining sites of the former antibodies were complementary to amoxicillin in a form that permits binding to the hydroxyaminobenzyl side-chain and the thiazolidine ring carboxyl. These conditions are satisfied with the drug in the '-oyl' but not in the '-anyl' form which involves linkage through the 2-carboxyl of the thiazolidine ring. With the second serum, adsorption studies showed that the wide-ranging reactivity of IgE was due to a single population of antibodies that detected a common specificity on the different penicillins. Combining site studies revealed clear recognition of the benzyl portion of the side-chain of benzylpenicilloyl, benzylpenicillanyl, ampicilloyl, ampicillanyl and amoxicillanyl determinants when free antibody access to the side-chain was possible but little or no recognition of the ring hydroxyl of amoxicillin. Such uninhibited access may not occur, however, when amoxicillin is conjugated in the '-oyl' form since opening the beta-lactam ring allows increased flexibility and rotation of the molecule and the possibility of close association of the hydroxyaminobenzyl side-chain of amoxicillin with the linked peptide carrier. In such close steric association, H-bonding involving the ring hydroxyl and amino acids of the carrier may prevent antibody access to the side-chain region of the amoxicilloyl determinant.     

Characterization of the AP endonucleases from Thermoplasma volcanium and Lactobacillus plantarum: Contributions of two important tryptophan residues to AP site recognition

Escherichia coli AP endonuclease (ExoIII) and its human homolog (APE1) have the sole tryptophan residue for AP site recognition (AP site recognizer) but these residues are at different positions near the catalytic sites. On the other hand, many bacterial AP endonucleases have two tryptophan residues at the same positions of both ExoIII and APE1. To elucidate whether these residues are involved in AP site recognition, the ExoIII homologs of Thermoplasma volcanium and Lactobacillus plantarum were characterized. These proteins showed AP endonuclease and 3'-5'exonculease activities. In each enzyme, the mutations of the tryptophan residues corresponding to Trp-280 of APE1 caused more significant reductions in activities and binding abilities to the oligonucleotide containing an AP site (AP-DNA) than those corresponding to Trp-212 of ExoIII. These results suggest that the tryptophan residue corresponding to Trp-280 of APE1 is the predominant AP site recognizer, and that corresponding to Trp-212 of ExoIII is the auxiliary recognizer.     

HIV‐1 ENV gp120 structural determinants for peptide triazole dual receptor site antagonism

Despite advances in HIV therapy, viral resistance and side‐effects with current drug regimens require targeting new components of the virus. Dual antagonist peptide triazoles (PT) are a novel class of HIV‐1 inhibitors that specifically target the gp120 component of the viral spike and inhibit its interaction with both of its cell surface protein ligands, namely the initial receptor CD4 and the co‐receptor (CCR5/CXCR4), thus preventing viral entry. Following an initial survey of 19 gp120 alanine mutants by ELISA, we screened 11 mutants for their importance in binding to, and inhibition by the PT KR21 using surface plasmon resonance. Key mutants were purified and tested for their effects on the peptide's affinity and its ability to inhibit binding of CD4 and the co‐receptor surrogate mAb 17b. Effects of the mutations on KR21 viral neutralization were measured by single‐round cell infection assays. Two mutations, D474A and T257A, caused large‐scale loss of KR21 binding, as well as losses in both CD4/17b and viral inhibition by KR21. A set of other Ala mutants revealed more moderate losses in direct binding affinity and inhibition sensitivity to KR21. The cluster of sensitive residues defines a PT functional epitope. This site is in a conserved region of gp120 that overlaps the CD4 binding site and is distant from the co‐receptor/17b binding site, suggesting an allosteric mode of inhibition for the latter. The arrangement and sequence conservation of the residues in the functional epitope explain the breadth of antiviral activity, and improve the potential for rational inhibitor development. Proteins 2013. © 2012 Wiley Periodicals, Inc.     

Major groove recognition by three-stranded beta-sheets: affinity determinants and conserved structural features

We present the results of a rational mutagenesis and binding-affinity study of the three-stranded beta-sheet-DNA interface in the complex formed by the amino-terminal DNA-binding domain of the Tn916 integrase protein and its cognate binding site. The relative importance of interfacial contacts present in its NMR-derived solution structure have been tested through mutagenesis, fluorescence anisotropy, and intrinsic quenching DNA-binding assays. We find that seven protein-DNA hydrogen bonds (two base-specific and five to phosphate groups) significantly contribute to the level of affinity. These interactions span the entire DNA-binding surface on the protein, but primarily originate from residues in only two strands of the sheet and loop L2. Interestingly, we show that highly populated, precisely defined intermolecular hydrogen bonds in the ensemble of conformers are invariably important for DNA-binding, implying that NMR-derived solution structures provide direct insight into the energetics of recognition. Unusual three-stranded beta-sheet-DNA interfaces have recently been discovered in three unrelated protein-DNA complexes. A comparative analysis of these structures reveals similar sheet positioning, the presence of two invariant interfacial contacts to the phosphodiester backbone, and two semi-conserved base-specific hydrogen bonds. Two of these conserved contacts significantly contribute to the affinity of the integrase-DNA complex, suggesting that the three-stranded beta-sheet DNA-binding motif exhibits conserved principles of recognition.     

Role of the tryptophan residue in the vicinity of the catalytic center of exonuclease III family AP endonucleases: AP site recognition mechanism

The mechanisms by which AP endonucleases recognize AP sites have not yet been determined. Based on our previous study with Escherichia coli exonuclease III (ExoIII), the ExoIII family AP endonucleases probably recognize the DNA-pocket formed at an AP site. The indole ring of a conserved tryptophan residue in the vicinity of the catalytic site presumably intercalates into this pocket. To test this hypothesis, we constructed a series of mutants of ExoIII and human APE1. Trp-212 of ExoIII and Trp-280 of APE1 were critical to the AP endonuclease activity and binding to DNA containing an AP site. To confirm the ability of the tryptophan residue to intercalate with the AP site, we examined the interaction between an oligopeptide containing a tryptophan residue and an oligonucleotide containing AP sites, using spectrofluorimetry and surface plasmon resonance (SPR) technology. The tryptophan residue of the oligopeptide specifically intercalated into an AP site of DNA. The tryptophan residue in the vicinity of the catalytic site of the ExoIII family AP endonucleases plays a key role in the recognition of AP sites.     

A structural motif is the recognition site for a new family of bacterial protein O-glycosyltransferases

The Escherichia coli Adhesin Involved in Diffuse Adherence (AIDA‐I) is a multifunctional protein that belongs to the family of monomeric autotransporters. This adhesin can be glycosylated by the AIDA‐associated heptosyltransferase (Aah). Glycosylation appears to be restricted to the extracellular domain of AIDA‐I, which comprises imperfect repeats of a 19‐amino‐acid consensus sequence and is predicted to form a β‐helix. Here, we show that Aah homologues can be found in many Gram‐negative bacteria, including Citrobacter rodentium. We demonstrated that an AIDA‐like protein is glycosylated in this species by the Aah homologue. We then investigated the substrate recognition mechanism of the E. coli Aah heptosyltransferase. We found that a peptide corresponding to one repeat of the 19‐amino‐acid consensus is sufficient for recognition and glycosylation by Aah. Mutagenesis studies suggested that, unexpectedly, Aah recognizes a structural motif typical of β‐helices, but not a specific sequence. In agreement with this finding, we observed that the extracellular domain of the Bordetella pertussis pertactin, a β‐helical polypeptide lacking the 19‐amino‐acid consensus sequence, could be glycosylated by Aah. Overall, our findings suggest that Aah represents the prototype of a new large family of bacterial protein O‐glycosyltransferases that modify various substrates recognized through a structural motif.