Effect of ATP analogs of DNA synthesis in isolated nuclei

Optimal synthesis of DNA in Ehrlich ascites cell nuclei is shown to be dependent upon the presence of both ATP and ADP. ATP can be replaced only by dATP. An ATP regenerating system is less effective than ATP alone or ATP in combination with ADP. ATP does not stimulate DNA synthesis primarily by maintenance of deoxyribonucleotide triphosphate levels. When the inhibition of DNA synthesis by high ATP levels is taken into account, the ATP analogs adenosine 5'-(alpha,beta-methylene)triphosphate, adenosine 5'-(beta, gamma-methylene)-triphosphate, and adenosine 5'-(beta, gamma-imino)triphosphate can neither substitute for ATP nor inhibit the ATP stimulation of DNA synthesis. Adenosine 5'-(3-thio)triphosphate, however, is a competitive inhibitor of DNA synthesis.     

Participation of ATP in the binding of a yeast replicative complex to DNA.

The activity that replicates yeast DNA in vitro can be isolated from cells of the budding yeast Saccharomyces in a high-Mr (approximately 2 X 10(6] form. Several lines of evidence indicate that this fraction contains a multiprotein replicative complex. A functional assay has been developed for the analysis of the interaction of the replicating activity with DNA. Binding of the activity required Mg2+, but did not require the addition of ATP or the other ribo- or deoxynucleoside triphosphates. However, the ATP analogues adenosine 5'-[gamma-thio]triphosphate and adenosine 5'-[beta gamma-imido]triphosphate blocked the binding, suggesting that ATP participates in the interaction at some stage. The binding was template (origin)-specific in either the presence or the absence of ATP and the other nucleoside triphosphates; however, ATP stabilized the replicating activity. The preferential inhibition of binding that was observed in the presence of the DNA topoisomerase II inhibitor coumermycin suggests that the requirement for ATP may be at least partially accounted for by the involvement of this enzyme in the initial interaction of the replicating activity with DNA. Finally, the binding was rapid. In contrast, DNA synthesis displayed a lag when assayed directly without first allowing a period for the replicating activity to bind to the DNA. In addition, binding was 'tight', as judged by the resistance of the protein--DNA complexes to salt in comparison with the relative sensitivity of binding. The replicating activity was not readily displaced from the complexes by exogenous DNAs, either possessing or lacking yeast origins of replication. The results suggest that the interaction of the replicating activity with the DNA occurs in more than one stage.     

The heat-shock protein ClpB in Escherichia coli is a protein-activated ATPase.

The clpB gene in Escherichia coli encodes a heat-shock protein that is a close homolog of the clpA gene product. The latter is the ATPase subunit of the multimeric ATP-dependent protease Ti (Clp) in E. coli, which also contains the 21-kDa proteolytic subunit (ClpP). The clpB gene product has been purified to near homogeneity by DEAE-Sepharose and heparin-agarose column chromatographies. The purified ClpB consists of a major 93-kDa protein and a minor 79-kDa polypeptide as analyzed by polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate. Upon gel filtration on a Superose-6 column, it behaves as a 350-kDa protein. Thus, ClpB appears to be a tetrameric complex of the 93-kDa subunit. The purified ClpB has ATPase activity which is stimulated 5-10-fold by casein. It is also activated by insulin, but not by other proteins, including globin and denatured bovine serum albumin. ClpB cleaves adenosine 5'-(alpha,beta-methylene)-triphosphate as rapidly as ATP, but not adenosine 5'-(beta,gamma-methylene)-triphosphate. GTP, CTP, and UTP are hydrolyzed 15-25% as well as ATP. ADP strongly inhibits ATP hydrolysis with a Ki of 34 microM. ClpB has a Km for ATP of 1.1 mM, and casein increases its Vmax for ATP without affecting its Km. A Mg2+ concentration of 3 mM is necessary for half-maximal ATP hydrolysis. Mn2+ supports ATPase activity as well as Mg2+, and Ca2+ has about 20% their activity. Anti-ClpB antiserum does not cross-react with ClpA nor does anti-ClpA antiserum react with ClpB. In addition, ClpB cannot replace ClpA in supporting the casein-degrading activity of ClpP. Thus, ClpB is distinct from ClpA in its structural and biochemical properties despite the similarities in their sequences.     

Purification and enzymological characterization of DNA-dependent ATPase IV from the Novikoff hepatoma.

DNA-dependent ATPase IV has been purified to near homogeneity from the Novikoff rat hepatoma. The enzyme is devoid of DNA polymerase, RNA polymerase, exonuclease, endonuclease, phosphomonoesterase, 3'- or 5'-phosphodiesterase, polynucleotide kinase, protein kinase, topoisomerase, helicase or DNA reannealing activities at a detection level of 10(-5) to 10(-7) relative to the ATPase activity. The enzyme is a monomer of Mr 110,000, has a sedimentation coefficient of 5.9 S, a Stokes radius of 40 A and a frictional coefficient of 1.32. In the presence of Mg2+ ion and a polynucleotide effector, ATPase IV hydrolyzes either ATP or dATP to the nucleoside diphosphate plus Pi. Other ribo- or deoxyribonucleoside triphosphates are not substrates. ATPase IV utilizes double-stranded DNA and single-stranded DNA as effector; however, it does not utilize poly(dT). The Km for dsDNA or ssDNA is 2.2 microM (nucleotide). A variety of ATP analogues were found to be competitive inhibitors of ATPase IV.     

Nucleoside 5'-(beta, gamma-peroxytriphosphates)

Procedures for the synthesis, purification, and characterization of beta, gamma-peroxy analogues of the eight common ribo- and deoxyribonucleoside triphosphates have been developed. Although adenosine 5'-(beta, gamma-peroxytriphosphate) was stable to conditions in most biochemical systems, incubation of a solution of the analogue at 100 degrees C led to formation of AMP and ATP, as well as ADP. NAD+ pyrophosphorylase was the only enzyme among 13 tested for which adenosine 5'-(beta, gamma-peroxytriphosphate) was a good substrate, but the analogue was an effective inhibitor for a number of kinases. The peroxy compounds tested inactive with Escherichia coli RNA polymerase and DNA polymerase I, as well as with wheat germ RNA polymerase II.     

Stimulation of Ca2+ efflux from sarcoplasmic reticulum by preincubation with ATP and inorganic phosphate.

Preincubation of sarcoplasmic reticulum with 1 mM-ATP completely inhibits Ca2+ accumulation and stimulates ATPase activity by over 2-fold. This effect of ATP is obtained only when the preincubation is carried out in the presence of Pi, but not with arsenate, chloride or sulphate. The inhibition by ATP of Ca2+ accumulation is pH-dependent, increasing as the pH is increased above 7.5. Inhibition of Ca2+ accumulation is observed on preincubation with ATP, but not with CTP, UTP, GTP, ADP, adenosine 5'-[beta gamma-methylene]triphosphate or adenosine 5'-[beta gamma-imido]triphosphate. The presence of Ca2+, but not Mg2+, during the preincubation, prevents the effect of ATP + Pi on Ca2+ accumulation. The ATP + Pi inhibition of Ca2+ accumulation is not due to modification of the ATPase catalytic cycle, but rather to stimulation of a rapid Ca2+ efflux from actively or passively loaded vesicles. This Ca2+ efflux is inhibited by dicyclohexylcarbodi-imide. Photoaffinity labelling of sarcoplasmic-reticulum membranes with 8-azido-[alpha-32P]ATP resulted in specific labelling of two proteins, of approx. 160 and 44 kDa. These proteins were labelled in the presence of Pi, but not other anions.     

Regulatory properties of magnesium-dependent guanylate cyclase in Dictyostelium discoideum membranes

We have characterized a magnesium-dependent guanylate cyclase in homogenates of Dictyostelium discoideum cells. 1) The enzyme shows an up to 4-fold higher cGMP synthesis in the presence of GTP analogues with half-maximal activation at about 1 microM guanosine 5'-O-(3-thio)triphosphate (GTP gamma S) or 100 microM guanosine 5'-(beta, gamma-imido)triphosphate; little or no stimulation was observed with GTP, guanosine mono- and diphosphates or with adenine nucleotides, with the exception of the ATP analogue adenosine 5'-(beta, gamma-imido)triphosphate. 2) Both basal and GTP gamma S-stimulated guanylate cyclase activity were rapidly lost from homogenates as was the ability of GTP gamma S to stimulate the enzyme after cell lysis. 3) Inclusion of 25 microM GTP gamma S during cell lysis reduced the KM for GTP from 340 to 85 microM and increased the Vmax from 120 to 255 pmol/min.mg protein, as assayed in homogenates 90 s after cell lysis. 4) Besides acting as an activator, GTP gamma S was also a substrate for the enzyme with a KM = 120 microM and a Vmax = 115 pmol/min.mg protein. 5) GTP gamma S-stimulated, Mg2+-dependent guanylate cyclase was inhibited by submicromolar concentrations of Ca2+ ions, and by inositol 1,4,5-trisphosphate in the absence of Ca2+ chelators. 6) Guanylate cyclase activity was detected in both supernatant and pellet fractions after 1 min centrifugation at 10,000 x g; however, only sedimentable enzyme was stimulated by GTP gamma S. We suggest that the Mg2+-dependent guanylate cyclase identified represents the enzyme that in intact cells is regulated via cell surface receptors, and we propose that guanine nucleotides are allosteric activators of this enzyme and that Ca2+ ions play a role in the maintenance of the enzyme in its basal state.     

Specificity of S-adenosylmethionine synthetase for ATP analogues mono- and disubstituted in bridging positions of the polyphosphate chain

The entire family of ATP analogues that are either mono- or disubstituted with imido and methylene bridges in the polyphosphate chain of ATP have been investigated as substrates and inhibitors of S-adenosylmethionine synthetase (ATP:L-methionine S-adenosyltransferase). The disubstituted analogues adenosine 5'-(alpha,beta:beta,gamma-diimidotriphosphate) (AMPNPNP) and adenosine 5'-(alpha,beta:alpha,beta'-diimidotriphosphate) [AMP(NP)2] have been synthesized for the first time, and a new route to adenosine 5'-(alpha,beta:beta,gamma-dimethylenetriphosphate) (AMPCPCP) has been developed. S-Adenosylmethionine synthetase catalyzes a two-step reaction: the intact polyphosphate chain is displaced from ATP, yielding AdoMet and tripolyphosphate, followed normally, but not obligatorily, by the hydrolysis of the tripolyphosphate to pyrophosphate and orthophosphate. Uniformly, the imido mono- or disubstituted derivatives are both better substrates and better inhibitors than their methylene counterparts. AMPNPNP reacts rapidly to give a single equivalent of product per active site, but subsequent turnovers are at least 1000-fold slower, enabling it to be used to quantify enzyme active site concentrations. In contrast, AMPCPCP is not detectably a substrate (less than 10(-5)% of ATP). AMP(NP)2, a branched isomer of linear AMPNPNP, was not a substrate but was a linear competitive inhibitor, greater than 100 fold more potent than ADP, indicating a reasonable degree of bulk tolerance at the alpha-phosphoryl group binding site.(ABSTRACT TRUNCATED AT 250 WORDS)     

DNA helicase from calf thymus. Purification to apparent homogeneity and biochemical characterization of the enzyme

We have purified a DNA helicase from calf thymus to apparent homogeneity by monitoring the activity with a strand displacement assay. DNA helicase followed the DNA polymerase alpha-primase complex through chromatography on phosphocellulose and hydroxylapatite. Separation from DNA polymerase alpha-primase complex as well as from the bulk of another DNA-dependent ATPase was achieved on heparin-Sepharose. Further purification steps included ATP-agarose and fast protein liquid chromatography-Mono S. A 47-kDa polypeptide cosedimented with the DNA helicase activity in a glycerol gradient as well as in gel filtration on Superose 6. The calf thymus DNA helicase had a sedimentation coefficient of 4-7 S and Stokes radius of about 45 A suggesting that the enzyme might be monomer in its functional form. DNA helicase activity requires a divalent cation with Mg2+ being more efficient than Mn2+ or Ca2+. Hydrolysis of ATP is required since the two nonhydrolyzable ATP analogs adenosine 5'-O-(3-thiotriphosphate) and adenylyl (beta, gamma-methylene)diphosphonate cannot substitute for ATP or dATP in the displacement reaction. Calf thymus DNA helicase is able to use ATP, dATP, dideoxy-ATP, CTP, and dCTP with Km for ATP and dATP of 0.2 and 0.25 mM, respectively. The enzyme can displace a fragment of 24 bases completely in an enzyme concentration- and time-dependent manner. The DNA helicase appears to bind to single-stranded DNA and to move to single-strand double-strand transition. The directionality of unwinding is 3'----5' with respect to the single-stranded DNA to which the enzyme is bound.     

Characterisation of Bacillus stearothermophilus PcrA helicase: evidence against an active rolling mechanism.

PcrA from Bacillus stearothermophilus is a DNA helicase for which, despite the availability of a crystal structure, there is very little biochemical information. We show that the enzyme has a broad nucleotide specificity, even being able to hydrolyse ethenonucleotides, and is able to couple the hydrolysis to unwinding of DNA substrates. In common with the Escherichia coli helicases Rep and UvrD, PcrA is a 3'-5' helicase but at high protein concentrations it can also displace a substrate with a 5' tail. However, in contrast to Rep and UvrD, we do not see any evidence for dimerisation of the protein even in the presence of DNA. The enzyme shows a specificity for the DNA substrate in gel mobility assays, with the preferred substrate being one with both single and double stranded regions of DNA. We propose that these data, together with existing structural evidence, support an inchworm rather than a rolling model for 3'-5' helicase activity.     

The dnaB-dnaC replication protein complex of Escherichia coli. II. Role of the complex in mobilizing dnaB functions

The dnaC protein of Escherichia coli, by forming a complex with the dnaB protein, facilitates the interactions with single-stranded DNA that enable dnaB to perform its ATPase, helicase, and priming functions. Within the dnaB-dnaC complex, dnaB appears to be inactive but becomes active upon the ATP-dependent release of dnaC from the complex. With adenosine 5'-(gamma-thio)triphosphate substituted for ATP, the dnaB-dnaC complex does not direct dnaB to its targeted actions. Excess dnaC inhibits dna beta actions and augments the ATP gamma S effects. In the dnaA protein-driven initiation of duplex chromosome replication, dnaB is introduced for its essential helicase role via the dnaB-dnaC complex. Similarly, when the dnaA protein interacts nonspecifically with single-stranded DNA, the dnaB-dnaC complex is essential to introduce dnaB for its role in primer formation by primase.     

Dynamic conformational changes of 21 S dynein ATPase coupled with ATP hydrolysis revealed by proteolytic digestion

Conformational changes of 21 S dynein ATPase from sea urchin sperm flagella were examined by tryptic digestion under physiological conditions. In the presence of 2 mM ATP or ADP plus 100 microM inorganic vanadate (Vi), dynein heavy chains were digested by trypsin into quite different polypeptides from those obtained in other cases (no addition, 2 mM ATP, 4 mM adenosine 5'-(beta,gamma-imido)triphosphate, 4 mM adenosine 5'-(beta,gamma-methylene)triphosphate, 2 mM ADP, 100 microM Vi). In the presence of 4 mM adenosine 5'-O-(3-thiotriphosphate), however, the digestion pattern was similar to that in the presence of ATP (ADP) and Vi, to a certain extent. In all conditions other than the presence of ATP (ADP) and Vi, 165- and 135-kDa polypeptides were the main products, whereas in the presence of ATP (ADP) and Vi, 200-, 150/148-, and 105/96-kDa peptides were produced and 320-kDa peptide became rather inaccessible to trypsin. The latter digestion pattern was not observed in the absence of divalent cations. These results suggest that, in the ATP hydrolysis cycle, dynein changes its conformation remarkably in the dynein-ADP-Pi state, which is presumably responsible for force generation.     

Characterisation of an ATP diphosphohydrolase (Apyrase, EC 3.6.1.5) activity in Trichomonas vaginalis

In the present report the enzymatic properties of an ATP diphosphohydrolase (apyrase, EC 3.6.1.5) in Trichomonas vaginalis were determined. The enzyme hydrolyses purine and pyrimidine nucleoside 5'-di- and 5'-triphosphates in an optimum pH range of 6.0--8.0. It is Ca(2+)-dependent and is insensitive to classical ATPase inhibitors, such as ouabain (1 mM), N-ethylmaleimide (0.1 mM), orthovanadate (0.1 mM) and sodium azide (5 mM). A significant inhibition of ADP hydrolysis (37%) was observed in the presence of 20 mM sodium azide, an inhibitor of ATP diphosphohydrolase. Levamisole, a specific inhibitor of alkaline phosphatase, and P(1), P(5)-di (adenosine 5'-) pentaphosphate, a specific inhibitor of adenylate kinase, did not inhibit the enzyme activity. The enzyme has apparent K(m) (Michaelis Constant) values of 49.2+/-2.8 and 49.9+/-10.4 microM and V(max) (maximum velocity) values of 49.4+/-7.1 and 48.3+/-6.9 nmol of inorganic phosphate x min(-1) x mg of protein(-1) for ATP and ADP, respectively. The parallel behaviour of ATPase and ADPase activities and the competition plot suggest that ATP and ADP hydrolysis occur at the same active site. The presence of an ATP diphosphohydrolase activity in T. vaginalis may be important for the modulation of nucleotide concentration in the extracellular space, protecting the parasite from the cytolytic effects of the nucleotides, mainly ATP.     

Protein-bound adenosine 5'-triphosphate: properties of a key intermediate of the magnesium-dependent subfragment 1 adenosinetriphosphatase from rabbit skeletal muscle

The time course of formation and decay of protein-bound adenosine 5'-triphosphate (ATP) has been monitored during single turnovers of the myosin subfragment 1 ATPase with nonspectrophotometric techniques. The rate constant controlling the ATP cleavage step increases markedly with ionic strength, so that in low salt the protein--ATP complex is observed transiently at higher concentration than the protein-products complex. The kinetics of the ATP cleavage step in a single turnover of the actosubfragment 1 ATPase indicates that under appropriate conditions this step is partially rate limiting during overall steady-state ATPase activity. It follows that a binary subfragment 1-ATP complex is a significant component of the steady-state intermediate of the actosubfragment 1 ATPase. Transient kinetic studies of ATP and adenosine 5'-(3-thiotriphosphate) [ATP (gamma S)] binding show directly that a substrate-induced protein isomerization accompanies ligand binding. The rate constant of the isomerization is 170 s-1 at pH 7.0, 15 degrees C, and 0.01 M ionic strength. Under these conditions nucleotide binding appears to be accompanied by a protein fluorescence increase that is 50% of the increase associated with magnesium-dependent steady-state ATPase activity.     

Kinetic mechanism and ATP-binding site reactivity of p38gamma MAP kinase

Activated p38gamma MAP kinase exhibited significant basal ATPase activity in the absence of a kinase substrate, and addition of a phosphoacceptor substrate increased k(cat)/K(m)20-fold. AMP-PCP was competitive with ATP binding and non-competitive with phosphoacceptor substrate binding. The nucleotide binding site affinity label 5'-(p-fluorosulfonylbenzoyl)adenosine (FSBA) bound stoichiometrically at Lys-56 in the ATP site of both unphosphorylated and activated p38gamma. AMP-PCP only protected the activated enzyme from FSBA inactivation, implying that AMP-PCP does not bind unphosphorylated p38gamma. Basal ATPase activities were also observed for activated p38alpha, ERK2 and JNK3 suggesting that the enzymatic mechanism may be similar for all classes of MAP kinases.     

Purification and characterization of delta helicase from fetal calf thymus.

A DNA helicase (delta helicase) which partially copurifies with DNA polymerase delta has been highly purified from fetal calf thymus. delta helicase differs in physical and enzymatic properties from other eukaryotic DNA helicases described thus far. The enzyme has an apparent mass of 57 kDa by gel filtration and is associated with polypeptides of 56 and 52 kDa by SDS-polyacrylamide gel electrophoresis. Photo-cross-linking of the purified enzyme with [alpha-32P]ATP resulted in labeling of a polypeptide of approximately 58 kDa, suggesting that the active site is present on the larger polypeptide. Unwinding of a partial duplex requires a nucleoside triphosphate which can be either ATP or dATP but not a nonhydrolyzable analogue of ATP. Other ribo- and deoxyribonucleoside triphosphates have little or no activity as cofactors. delta helicase also has DNA-dependent ATPase activity which has a relatively low Km for ATP (40 microM). delta helicase binds to single-stranded DNA but has little or no affinity for double-stranded DNA or single-stranded RNA. Similar to replicative DNA helicases from prokaryotes and the herpes simplex virus type 1 helicase-primase, delta helicase translocates in the 5'-3' direction along the strand to which it is bound and preferentially unwinds DNA substrates with a forklike structure.     

Substrate specificity and adenosine triphosphatase activity of the ATP-dependent deoxyribonuclease of Bacillus subtilis

Studies on the specificity of the ATP-dependent DNase of Bacillus subtilis 168, carried out with pure enzyme at the optimal conditions for its action, have shown that the substrate is double-stranded linear DNA. Linear single-stranded DNA (separated strands of B. subtilis DNA and linear phage fd DNA) is not attacked, neither are there any circular forms (supercoiled or nicked simian virus 40 and circular single-stranded fd DNAs). The double-stranded DNA can be completely hydrolysed, the limit products being, almost exclusively, mononucleotides. The presence of terminal phosphate residues in the substrate (either at the 3' or the 5' end) is not necessary for enzyme action. This DNase appears therefore to be an exonuclease processively liberating mononucleotides from both strands of the native linear DNA. ATP (indispensable for the DNase reaction) is also hydrolysed by the enzyme, to ADP and inorganic orthophosphate (Pi) in the presence of DNA. The apparent Km for ATP, in the ATPase reaction, is 0.15 mM. At high ATP concentrations, which inhibit the DNase activity, there is activation of the ATPase reaction. Three molecules of ATP are consumed for each DNA phosphodiester bond split, at optimal conditions for DNase activity.     

Chemical modification of sarcoplasmic reticulum. Stimulation of Ca2+ release.

Pretreatment of sarcoplasmic membranes with acetic or maleic anhydrides, which interact principally with amino groups, resulted in an inhibition of Ca2+ accumulation and ATPase activity. The presence of ATP, ADP or adenosine 5'-[beta, gamma-imido]triphosphate in the modification medium selectively protected against the inactivation of ATPase activity by the anhydride but did not protect against the inhibition of Ca2+ accumulation. Acetic anhydride modification in the presence of ATP appeared to increase specifically the permeability of the sarcoplasmic reticulum membrane to Ca2+ but not to sucrose, Tris, Na+ or Pi. The chemical modification stimulated a rapid release of Ca2+ from sarcoplasmic reticulum vesicles passively or actively loaded with calcium, from liposomes reconstituted with the partially purified ATPase fraction but not from those reconstituted with the purified ATPase. The inactivation of Ca2+ accumulation by acetic anhydride (in the presence of ATP) was rapid and strongly pH-dependent with an estimated pK value above 8.3 for the reactive group(s). The negatively charged reagents pyridoxal 5-phosphate and trinitrobenzene-sulphonate, which also interact with amino groups, did not stimulate Ca2+ release. Since these reagents do not penetrate the sarcoplasmic reticulum membranes, it is proposed that Ca2+ release is promoted by modification of internally located, positively charged amino group(s).     

Structures of the Ca2+-ATPase complexes with ATP, AMPPCP and AMPPNP. An FTIR study

We studied binding of ATP and of the ATP analogs adenosine 5'-(beta,gamma-methylene)triphosphate (AMPCP) and beta,gamma-imidoadenosine 5'-triphosphate (AMPPNP) to the Ca(2+)-ATPase of the sarcoplasmic reticulum membrane (SERCA1a) with time-resolved infrared spectroscopy. In our experiments, ATP reacted with ATPase which had AMPPCP or AMPPNP bound. These experiments monitored exchange of ATP analog by ATP and phosphorylation to the first phosphoenzyme intermediate Ca(2)E1P. These reactions were triggered by the release of ATP from caged ATP. Only small differences in infrared absorption were observed between the ATP complex and the complexes with AMPPCP and AMPPNP indicating that overall the interactions between nucleotide and ATPase are similar and that all complexes adopt a closed conformation. The spectral differences between ATP and AMPPCP complex were more pronounced at high Ca(2+) concentration (10 mM). They are likely due to a different position of the gamma-phosphate which affects the beta-sheet in the P domain.     

Distinct MutS DNA-binding modes that are differentially modulated by ATP binding and hydrolysis.

The role of MutS ATPase in mismatch repair is controversial. To clarify further the function of this activity, we have examined adenine nucleotide effects on interactions of Escherichia coli MutS with homoduplex and heteroduplex DNAs. In contrast to previous results with human MutS alpha, we find that a physical block at one end of a linear heteroduplex is sufficient to support stable MutS complex formation in the presence of ATP.Mg(2+). Surface plasmon resonance analysis at low ionic strength indicates that the lifetime of MutS complexes with heteroduplex DNA depends on the nature of the nucleotide present when MutS binds. Whereas complexes prepared in the absence of nucleotide or in the presence of ADP undergo rapid dissociation upon challenge with ATP x Mg(2+), complexes produced in the presence of ATP x Mg(2+), adenosine 5'-(beta,gamma-imino)triphosphate (AMPPNP) x Mg(2+), or ATP (no Mg(2+)) are resistant to dissociation upon ATP challenge. AMPPNP x Mg(2+) and ATP (no Mg(2+)) reduce MutS affinity for heteroduplex but have little effect on homoduplex affinity, resulting in abolition of specificity for mispaired DNA at physiological salt concentrations. Conversely, the highest mismatch specificity is observed in the absence of nucleotide or in the presence of ADP. ADP has only a limited effect on heteroduplex affinity but reduces MutS affinity for homoduplex DNA.