HYL1 is a double-stranded RNA binding protein involved in microRNA processing in plants. HYL1 enhances the efficiency and precision of the RNase III protein DCL1 and participates in microRNA strand selection. In this work, we dissect the contributions of the domains of HYL1 to the binding of RNA targets. We found that the first domain is the main contributor to RNA binding. Mapping of the interaction regions by nuclear magnetic resonance on the structure of HYL1 RNA-binding domains showed that the difference in binding capabilities can be traced to sequence divergence in β2−β3 loop. The possible role of each domain is discussed in light of previous experimental data.

The regulated activation of NF-κB by antigen receptor signaling is required for normal B and T lymphocyte activation during the adaptive immune response. Dysregulated NF-κB activation is associated with several types of lymphoma, including diffuse large B cell lymphoma (DLBCL). During normal antigen receptor signaling, the multidomain scaffold protein CARD11 undergoes a transition from a closed, inactive state to an open, active conformation that recruits several signaling proteins into a complex, leading to IKK kinase activation. This transition is regulated by the CARD11 inhibitory domain (ID), which participates in intramolecular interactions that prevent cofactor binding to CARD11 prior to signaling, but which is neutralized after receptor engagement by phosphorylation. Several oncogenic CARD11 mutations have been identified in DLBCL that enhance activity and that are mostly found in the coiled-coil domain. However, the mechanisms by which these mutations cause CARD11 hyperactivity and spontaneous NF-κB activation are poorly understood. In this report, we provide several lines of evidence that oncogenic mutations F123I and L225LI induce CARD11 hyperactivity by disrupting autoinhibition by the CARD11 ID. These mutations disrupt ID-mediated intramolecular interactions and ID-dependent inhibition and bypass the requirement for ID phosphorylation during T cell receptor signaling. Intriguingly, these mutations selectively enhance the apparent affinity of CARD11 for Bcl10, but not for other signaling proteins that are recruited to CARD11 in an ID-dependent manner during normal antigen receptor signaling. Our results establish a mechanism that explains how DLBCL-associated mutations in CARD11 can initiate spontaneous, receptor-independent activation of NF-κB.

Eukaryotic translation initiation factor 4G (eIF4G) functions to organize the assembly of initiation factors required for the recruitment of a 40S ribosomal subunit to an mRNA and for interacting with the poly(A) binding protein (PABP). Many eukaryotes express two highly similar eIF4G isoforms. eIFiso4G, one of two isoforms in plants, is highly divergent and unusually small in size. Unlike animal and yeast eIF4G, the domain organization of plant eIF4G proteins is largely unknown. Consequently, little is known about the conservation of plant eIF4G with those in other eukaryotes. In this study, we show that eIFiso4G is similar to other eIF4G proteins in that there are interaction domains for eIF4A and PABP and we identify, for the first time, the interaction domain for eIF4B. In contrast to previous reports, two eIF4A binding domains in eIFiso4G were identified, similar in number and organization to those of animal eIF4G. The eIFiso4G domain organization does differ, however, in that the N-terminal eIF4A binding domain overlaps with the eIF4B and PABP binding domains. Moreover, the eIF4B and PABP binding domains overlap. PABP and eIF4B compete with eIF4A for binding eIFiso4G in the absence of the C-terminal eIF4A binding domain but not when both eIF4A binding domains are present, suggesting that the C-terminal eIF4A interaction domain functions to stabilize the association of eIF4A with eIFiso4G in the presence of eIF4B or PABP. Competitive binding to eIFiso4G was also observed between eIF4B and PABP. These observations reveal an important function of the C-terminal eIF4A binding domain in maintaining the interaction of multiple partner proteins with eIFiso4G despite the substantial divergence in its size and domain organization.

Deinococcus radiodurans single-stranded (ss) DNA binding protein (DrSSB) originates from a radiation-resistant bacterium and participates in DNA recombination, replication, and repair. Although it functions as a homodimer, it contains four DNA binding domains (OB-folds) and thus is structurally similar to the Escherichia coli SSB (EcoSSB) homotetramer. We examined the equilibrium binding of DrSSB to ssDNA for comparison with that of EcoSSB. We find that the occluded site size of DrSSB on poly(dT) is ∼45 nucleotides under low-salt conditions (<0.02 M NaCl) but increases to 50−55 nucleotides at ≥0.2 M NaCl. This suggests that DrSSB undergoes a transition between ssDNA binding modes, which is observed for EcoSSB, although the site size difference between modes is not as large as for EcoSSB, suggesting that the pathways of ssDNA wrapping differ for these two proteins. The occluded site size corresponds well to the contact site size (52 nucleotides) determined by isothermal titration calorimetry (ITC). Electrophoretic studies of complexes of DrSSB with phage M13 ssDNA indicate the formation of stable, highly cooperative complexes under low-salt conditions. Using ITC, we find that DrSSB binding to oligo(dT)s with lengths close to the determined site size (50−55 nucleotides) is stoichiometric with a ΔHobs of approximately −94 ± 4 kcal/mol, somewhat smaller than that for EcoSSB (approximately −130 kcal/mol) under the same conditions. The observed binding enthalpy shows a large sensitivity to NaCl concentration, similar to that observed for EcoSSB. With the exception of the less dramatic change in occluded site size, the behavior of DrSSB is similar to that of EcoSSB protein (although clear quantitative differences exist). These common features for SSB proteins having multiple DNA binding domains enable versatility of SSB function in vivo.

β-Amyloid (Aβ) is the main protein component of the amyloid plaques associated with Alzheimer’s disease. Transthyretin (TTR) is a homotetramer that circulates in both blood and cerebrospinal fluid. Wild-type (wt) TTR amyloid deposits are linked to senile systemic amyloidosis, a common disease of aging, while several TTR mutants are linked to familial amyloid polyneuropathy. Several recent studies provide support for the hypothesis that these two amyloidogenic proteins interact, and that this interaction is biologically relevant. For example, upregulation of TTR expression in Tg2576 mice was linked to protection from the toxic effects of Aβ deposition [Stein, T. D., and Johnson, J. A. (2002) J. Neurosci. 22, 7380−7388]. We examined the interaction of Aβ with wt TTR as well as two mutants: F87M/L110M, engineered to be a stable monomer, and T119M, a naturally occurring mutant with a tetrameric stability higher than that of the wild type. On the basis of enzyme-linked immunoassays as well as cross-linking experiments, we conclude that Aβ monomers bind more to TTR monomers than to TTR tetramers. The data further suggest that TTR tetramers interact preferably with Aβ aggregates rather than Aβ monomers. Through tandem mass spectrometry analysis of cross-linked TTR−Aβ fragments, we identified the A strand, in the inner β-sheet of TTR, as well as the EF helix, as regions of TTR that are involved with Aβ association. Light scattering and electron microscopy studies demonstrate that the outcome of the TTR−Aβ interaction strongly depends on TTR quaternary structure. While TTR tetramers may modestly enhance aggregation, TTR monomers decidedly arrest Aβ aggregate growth. These data provide important new insights into the nature of TTR−Aβ interactions. Such interactions may regulate TTR-mediated protection against Aβ toxicity.

The DsrMKJOP transmembrane complex has a most important function in dissimilatory sulfur metabolism, not only in many sulfur-oxidizing organisms but also in sulfate-reducing prokaryotes. Here, we focused on an individual component of this complex, the triheme cytochrome c DsrJ from the purple sulfur bacterium Allochromatium vinosum. In A. vinosum, the signal peptide of DsrJ is not cleaved off but serves as a membrane anchor. Sequence analysis suggested the presence of three heme c species with bis-His, His/Met, and possibly a very unusual His/Cys ligation. A. vinosum DsrJ produced as a recombinant protein in Escherichia coli indeed contained three hemes, and electron paramagnetic resonance (EPR) spectroscopy provided evidence of possible, but only partial, His/Cys heme ligation in one of the hemes. This heme shows heterogeneous coordination, with Met being another candidate ligand. Cysteine 46 was replaced with serine using site-directed mutagenesis, with the mutant protein showing a small decrease in the magnitude of the EPR signal attributed to His/Cys coordination, but identical UV−vis and RR spectra. The redox potentials of the hemes in the wild-type protein were determined to be −20, −200, and −220 mV and were found to be virtually identical in the mutant protein. However, in vivo the same ligand exchange led to a dramatically altered phenotype, highlighting the importance of Cys46. Our results suggest that Cys46 may be involved in catalytic sulfur chemistry rather than electron transfer. Additional in vivo experiments showed that DsrJ can be functionally replaced in A. vinosum by the homologous protein from the sulfate reducer Desulfovibrio vulgaris.

The BCL-2 gene encodes a 25 kDa membrane protein that plays critical roles in the control of apoptosis. The regulation of BCL-2 gene expression is highly complex and occurs both transcriptionally and posttranscriptionally. In particular, the 5′ upstream region of BCL-2 contains a number of elements that control its expression. We have identified a highly conserved 25-nucleotide G-rich sequence (BCL2Q), with potential to fold into a RNA G-quadruplex structure, located 42 nucleotides upstream of the translation start site of human BCL-2. In this study, we used a series of biophysical experiments to show that the BCL2Q sequence folds into a stable RNA G-quadruplex in vitro, and we conducted functional luciferase reporter-based assays, in a cell-free lysate and in three types of human cell lines, to demonstrate that the BCL2Q sequence modulates protein expression in the context of the 493-nucleotide native 5′ untranslated region of BCL-2.

Two apical caspases, caspase-8 and -10, are involved in the extrinsic death receptor pathway in humans, but it is mainly caspase-8 in its apoptotic and nonapoptotic functions that has been an intense research focus. In this study we concentrate on caspase-10, its mechanism of activation, and the role of the intersubunit cleavage. Our data obtained through in vitro dimerization assays strongly suggest that caspase-10 follows the proximity-induced dimerization model for apical caspases. Furthermore, we compare the specificity and activity of the wild-type protease with a mutant incapable of autoprocessing by using positional scanning substrate analysis and cleavage of natural protein substrates. These experiments reveal a striking difference between the wild type and the mutant, leading us to hypothesize that the single chain enzyme has restricted activity on most proteins but high activity on the proapoptotic protein Bid, potentially supporting a prodeath role for both cleaved and uncleaved caspase-10.

Copper-containing amine oxidases (CAOs) require a protein-derived topaquinone cofactor (TPQ) for activity. TPQ biogenesis is a self-processing reaction requiring the presence of copper and molecular oxygen. Recombinant human diamine oxidase (hDAO) was heterologously expressed in Drosophila S2 cells, and analysis indicates that the purified hDAO contains substoichiometric amounts of copper and TPQ. The crystal structure of a complex of an inhibitor, aminoguanidine, and hDAO at 2.05 Å resolution shows that the aminoguanidine forms a covalent adduct with the TPQ and that the site is ∼75% occupied. Aminoguanidine is a potent inhibitor of hDAO with an IC50 of 153 ± 9 nM. The structure indicates that the catalytic metal site, normally occupied by copper, is fully occupied. X-ray diffraction data recorded below the copper edge, between the copper and zinc edges, and above the zinc edge have been used to show that the metal site is occupied approximately 75% by copper and 25% by zinc and the formation of the TPQ cofactor is correlated with copper occupancy.

The lipid A moiety of Escherichia coli lipopolysaccharide is a hexaacylated disaccharide of glucosamine that is phosphorylated at the 1 and 4′ positions. Expression of the Francisella novicida lipid A 1-phosphatase FnLpxE in E. coli results in dephosphorylation of the lipid A proximal unit. Coexpression of FnLpxE and the Rhizobium leguminosarum lipid A oxidase RlLpxQ in E. coli converts much of the proximal glucosamine to 2-amino-2-deoxygluconate. Expression of the F. novicida lipid A 4′-phosphatase FnLpxF in wild-type E. coli has no effect because FnLpxF cannot dephosphorylate hexaacylated lipid A. However, expression of FnLpxF in E. coli lpxM mutants, which synthesize pentaacylated lipid A lacking the secondary 3′-myristate chain, causes extensive 4′-dephosphorylation. Coexpression of FnLpxE and FnLpxF in lpxM mutants results in massive accumulation of lipid A species lacking both phosphate groups, and introduction of RlLpxQ generates phosphate-free lipid A variants containing 2-amino-2-deoxygluconate. The proposed lipid A structures were confirmed by electrospray ionization mass spectrometry. Strains with 4′-dephosphorylated lipid A display increased polymyxin resistance. Heptose-deficient mutants of E. coli lacking both the 1- and 4′-phosphate moieties are viable on plates but sensitive to CaCl2. Our methods for reengineering lipid A structure may be useful for generating novel vaccines and adjuvants.

HIV-1 infection and antiretroviral therapy are associated with a dyslipidemia marked by low levels of high-density lipoprotein and increased cardiovascular disease, but it is unclear whether virion replication plays a causative role in these changes. The HIV-1 Nef protein can impair ATP cassette binding transporter A1 (ABCA1) cholesterol efflux from macrophages, a potentially pro-atherosclerotic effect. This viral inhibition of efflux was correlated with a direct interaction between ABCA1 and Nef. Here, we defined the ABCA1 domain required for the Nef−ABCA1 protein−protein interaction and determined whether this interaction mediates the ability of Nef to downregulate ABCA1. Nef expressed in HEK 293 cells strongly inhibited ABCA1 efflux and protein levels but did not alter levels of cMIR, another transmembrane protein. Analysis of a panel of ABCA1 C-terminal mutants showed Nef binding required the ABCA1 C-terminal amino acids between positions 2225 and 2231. However, the binding of Nef to ABCA1 was not required for inhibition because the C-terminal ABCA1 mutants that did not bind Nef were still downregulated by Nef. Given this discordance, the mechanism of downregulation was investigated and was found to involve the acceleration of ABCA1 protein degradation but did not to depend upon the ABCA1 PEST sequence, which mediates the calpain proteolysis of ABCA1. Furthermore, it did not associate with a Nef-dependent induction of signaling through the unfolded protein response but was significantly dependent upon proteasomal function and could act on an ABCA1 mutant that fails to exit the endoplasmic reticulum. In summary, we show that Nef downregulates ABCA1 function by a post-translational mechanism that stimulates ABCA1 degradation but does not require the ability of Nef to bind ABCA1.

Affinity selection−mass spectrometry (AS-MS) screening of kinesin spindle protein (KSP) followed by enzyme inhibition studies and temperature-dependent circular dichroism (TdCD) characterization was utilized to identify a series of benzimidazole compounds. This series also binds in the presence of Ispinesib, a known anticancer KSP inhibitor in phase I/II clinical trials for breast cancer. TdCD and AS-MS analyses support simultaneous binding implying existence of a novel non-Ispinesib binding pocket within KSP. Additional TdCD analyses demonstrate direct binding of these compounds to Ispinesib-resistant mutants (D130V, A133D, and A133D + D130V double mutant), further strengthening the hypothesis that the compounds bind to a distinct binding pocket. Also importantly, binding to this pocket causes uncompetitive inhibition of KSP ATPase activity. The uncompetitive inhibition with respect to ATP is also confirmed by the requirement of nucleotide for binding of the compounds. After preliminary affinity optimization, the benzimidazole series exhibited distinctive antimitotic activity as evidenced by blockade of bipolar spindle formation and appearance of monoasters. Cancer cell growth inhibition was also demonstrated either as a single agent or in combination with Ispinesib. The combination was additive as predicted by the binding studies using TdCD and AS-MS analyses. The available data support the existence of a KSP inhibitory site hitherto unknown in the literature. The data also suggest that targeting this novel site could be a productive strategy for eluding Ispinesib-resistant tumors. Finally, AS-MS and TdCD techniques are general in scope and may enable screening other targets in the presence of known drugs, clinical candidates, or tool compounds that bind to the protein of interest in an effort to identify potency-enhancing small molecules that increase efficacy and impede resistance in combination therapy.

In-depth mass spectrometric analysis of disulfide bond patterns in recombinant mouse laminin β1 and γ1 chain N-terminal fragments comprising the laminin N-terminal (LN) domain and the first four laminin epidermal growth factor-like (LE) domains revealed a novel disulfide pattern for LE domains. This showed a (2−3, 4−5, 6−7, 8−1) connectivity with the last cysteine of one LE domain being connected to the first cysteine of the following LE domain. The same pattern was also found in E4, the N-terminal β1 chain fragment derived by elastase digestion of mouse EHS tumor laminin-111, showing that this pattern occurs in native laminin. The strictly linear pattern with an interdomain disulfide has not been described previously for EGF domains. The N-terminal portions of laminin short arms, consisting of the LN domain and LE domains 1−4, are essential for laminin−laminin self-interactions, whereas the internal LE domains 7−9 in the laminin γ1 chain harbor the nidogen binding site and have a conventional disulfide pattern. This suggests that LE domains differing in function also differ in their disulfide patterns.

The STPR motif is composed of 23-amino acid repeats aligned contiguously. STPR was originally reported as the DNA-binding domain of the silkworm protein FMBP-1. ZNF821, the human protein that contains the STPR domain, is a zinc finger protein of unknown function. In this study, we prepared peptides of silkworm FMBP-1 STPR (sSTPR) and human ZNF821 STPR (hSTPR) and compared their DNA binding behaviors. This revealed that hSTPR, like sSTPR, is a double-stranded DNA-binding domain. Sequence-independent DNA binding affinities and α-helix-rich DNA-bound structures were comparable between the two STPRs, although the specific DNA sequence of hSTPR is still unclear. In addition, a subcellular expression experiment showed that the hSTPR domain is responsible for the nuclear localization of ZNF821. ZNF821 showed a much slower diffusion rate in the nucleus, suggesting the possibility of interaction with chromosomal DNA. STPR sequences are found in many proteins from vertebrates, insects, and nematodes. Some of the consensus amino acid residues would be responsible for DNA binding and concomitant increases in α-helix structure content.

Raltegravir is an FDA approved inhibitor directed against human immunodeficiency virus type 1 (HIV-1) integrase (IN). In this study, we investigated the mechanisms associated with multiple strand transfer inhibitors capable of inhibiting concerted integration by HIV-1 IN. The results show raltegravir, elvitegravir, MK-2048, RDS 1997, and RDS 2197 all appear to encompass a common inhibitory mechanism by modifying IN−viral DNA interactions. These structurally different inhibitors bind to and inactivate the synaptic complex, an intermediate in the concerted integration pathway in vitro. The inhibitors physically trap the synaptic complex, thereby preventing target DNA binding and thus concerted integration. The efficiency of a particular inhibitor to trap the synaptic complex observed on native agarose gels correlated with its potency for inhibiting the concerted integration reaction, defined by IC50 values for each inhibitor. At low nanomolar concentrations (<50 nM), raltegravir displayed a time-dependent inhibition of concerted integration, a property associated with slow-binding inhibitors. Studies of raltegravir-resistant IN mutants N155H and Q148H without inhibitors demonstrated that their capacity to assemble the synaptic complex and promote concerted integration was similar to their reported virus replication capacities. The concerted integration activity of Q148H showed a higher cross-resistance to raltegravir than observed with N155H, providing evidence as to why the Q148H pathway with secondary mutations is the predominant pathway upon prolonged treatment. Notably, MK-2048 is equally potent against wild-type IN and raltegravir-resistant IN mutant N155H, suggesting this inhibitor may bind similarly within their drug-binding pockets.

Quadruplex-specific molecules can serve as suitable drugs in cancer therapy. We have synthesized a pair of furan-based cyclic homooligopeptides, ligand 1 and ligand 2, to specifically target G-quadruplexes. We have shown by CD spectroscopy and UV melting that these ligands can effectively induce G-quadruplex structures in the G-rich 22-mer c-MYC DNA sequence and further stabilize the structure. Equilibrium binding constants measured by isothermal titration calorimeter methods indicate a high affinity of the ligands for the quadruplex structures (K ∼ 107 M−1) and no affinity for the duplex DNA, demonstrating that these ligands are selective for G-quadruplex structures. Surface plasmon resonance was also used to compute the binding while fluorescence resonance energy transfer-based assay was additionally used to confirm the selectivity. Moreover, using real time PCR we observed up to 90% downregulation of c-MYC transcripts after 24 h of ligand treatment in HeLa cells. Using a luciferase assay we show the downregulation of the protein levels. Fluorescent-assisted cell sorter-based cell cycle analysis showed a prominent arrest of cells in the sub-G1 stage upon treatment of ligands that leads toward apoptosis. Altogether, these experiments support the hypothesis that the present molecules are effective in specifically binding and stabilizing quadruplexes and provide a suitable scaffold to develop into a quadruplex-targeting therapeutic agent.

Bacillithiol (Cys-GlcN-malate, BSH) has recently been identified as a novel low-molecular weight thiol in Bacillus anthracis, Staphylococcus aureus, and several other Gram-positive bacteria lacking glutathione and mycothiol. We have now characterized the first two enzymes for the BSH biosynthetic pathway in B. anthracis, which combine to produce α-d-glucosaminyl l-malate (GlcN-malate) from UDP-GlcNAc and l-malate. The structure of the GlcNAc-malate intermediate has been determined, as have the kinetic parameters for the BaBshA glycosyltransferase (→GlcNAc-malate) and the BaBshB deacetylase (→GlcN-malate). BSH is one of only two natural products reported to contain a malyl glycoside, and the crystal structure of the BaBshA−UDP−malate ternary complex, determined in this work at 3.3 Å resolution, identifies several active-site interactions important for the specific recognition of l-malate, but not other α-hydroxy acids, as the acceptor substrate. In sharp contrast to the structures reported for the GlcNAc−1-d-myo-inositol-3-phosphate synthase (MshA) apo and ternary complex forms, there is no major conformational change observed in the structures of the corresponding BaBshA forms. A mutant strain of B. anthracis deficient in the BshA glycosyltransferase fails to produce BSH, as predicted. This B. anthracis bshA locus (BA1558) has been identified in a transposon-site hybridization study as required for growth, sporulation, or germination [Day, W. A., Jr., Rasmussen, S. L., Carpenter, B. M., Peterson, S. N., and Friedlander, A. M. (2007) J. Bacteriol. 189, 3296−3301], suggesting that the biosynthesis of BSH could represent a target for the development of novel antimicrobials with broad-spectrum activity against Gram-positive pathogens like B. anthracis. The metabolites that function in thiol redox buffering and homeostasis in Bacillus are not well understood, and we present a composite picture based on this and other recent work.

3-(2′-Deoxy-β-d-erythro-pentofuranosyl)pyrimido-[1,2-a]purin-10(3H)-one (M1dG) is the major adduct derived from the reaction of DNA with the lipid peroxidation product malondialdehyde and the DNA peroxidation product base propenal. M1dG is mutagenic in Escherichia coli and mammalian cells, inducing base-pair substitutions (M1dG → A and M1dG → T) and frameshift mutations. Y-family polymerases may contribute to the mutations induced by M1dG in vivo. Previous reports described the bypass of M1dG by DNA polymerases η and Dpo4. The present experiments were conducted to evaluate bypass of M1dG by the human Y-family DNA polymerases κ, ι, and Rev1. M1dG was incorporated into template-primers containing either dC or dT residues 5′ to the adduct, and the template-primers were subjected to in vitro replication by the individual DNA polymerases. Steady-state kinetic analysis of single nucleotide incorporation indicates that dCMP is most frequently inserted by hPol κ opposite the adduct in both sequence contexts, followed by dTMP and dGMP. dCMP and dTMP were most frequently inserted by hPol ι, and only dCMP was inserted by Rev1. hPol κ extended template-primers in the order M1dG:dC > M1dG:dG > M1dG:dT ∼ M1dG:dA, but neither hPol ι nor Rev1 extended M1dG-containing template-primers. Liquid chromatography−mass spectrometry analysis of the products of hPol κ-catalyzed extension verified this preference in the 3′-GXC-5′ template sequence but revealed the generation of a series of complex products in which dAMP is incorporated opposite M1dG in the 3′-GXT-5′ template sequence. The results indicate that DNA hPol κ or the combined action of hPol ι or Rev1 and hPol κ bypass M1dG residues in DNA and generate products that are consistent with some of the mutations induced by M1dG in mammalian cells.

A typical feature of type II restriction endonucleases (REases) is their obligate sequence specificity and requirement for Mg2+ during catalysis. R.KpnI is an exception. Unlike most other type II REases, the active site of this enzyme can accommodate Mg2+, Mn2+, Ca2+, or Zn2+ and cleave DNA. The enzyme belongs to the HNH superfamily of nucleases and is characterized by the presence of a ββα-Me finger motif. Residues D148, H149, and Q175 together form the HNH active site and are essential for Mg2+ binding and catalysis. The unique ability of the enzyme to cleave DNA in the presence of different metal ions is exploited to generate mutants that are specific to one particular metal ion. We describe the generation of a Mn2+-dependent sequence specific endonuclease, defective in DNA cleavage with Mg2+ and other divalent metal ions. In the engineered mutant, only Mn2+ is selectively bound at the active site, imparting Mn2+-mediated cleavage. The mutant is impaired in concerted double-stranded DNA cleavage, leading to accumulation of nicked intermediates. The nicking activity of the mutant enzyme is further enhanced by altered reaction conditions. The active site fluidity of REases allowing flexible accommodation of catalytic cofactors thus forms a basis for engineering selective metal ion-dependent REase additionally possessing nicking activity.

On the basis of the available X-ray structures of S-adenosylhomocysteine hydrolases (SAHHs), free energy simulations employing the MM-GBSA approach were applied to predict residues important to the differential cofactor binding properties of human and trypanosomal SAHHs (Hs-SAHH and Tc-SAHH), within 5 Å of the cofactor NAD+/NADH binding site. Among the 38 residues in this region, only four are different between the two enzymes. Surprisingly, the four nonidentical residues make no major contribution to differential cofactor binding between Hs-SAHH and Tc-SAHH. On the other hand, four pairs of identical residues are shown by free energy simulations to differentiate cofactor binding between Hs-SAHH and Tc-SAHH. Experimental mutagenesis was performed to test these predictions for a lysine residue and a tyrosine residue of the C-terminal extension that penetrates a partner subunit to form part of the cofactor binding site. The K431A mutant of Tc-SAHH (TcK431A) loses its cofactor binding affinity but retains the wild type’s tetrameric structure, while the corresponding mutant of Hs-SAHH (HsK426A) loses both cofactor affinity and tetrameric structure [Ault-Riche, D. B., et al. (1994) J. Biol. Chem. 269, 31472−31478]. The tyrosine mutants HsY430A and TcY435A alter the NAD+ association and dissociation kinetics, with HsY430A increasing the cofactor equilibrium dissociation constant from approximately 10 nM (Hs-SAHH) to ∼800 nM and TcY435A increasing the cofactor equilibrium dissociation constant from approximately 100 nM (Tc-SAHH) to ∼1 mM. Both changes result from larger increases in the off rate combined with smaller decreases in the on rate. These investigations demonstrate that computational free energy decomposition may be used to guide experimental studies by suggesting sensitive sites for mutagenesis. Our finding that identical residues in two orthologous proteins may give significantly different binding free energy contributions strongly suggests that comparative studies of homologous proteins should investigate not only different residues but also identical residues in these proteins.