The 30 bp duplex containing the central mispair OG:FA or OG A was 32 P-5′-end-labeled on the strand containing the OG, FA or A. MPE-Fe(II) footprinting was performed as described earlier ( 27 ), except using 10 nM DNA. Maxam-Gilbert G+A reactions using piperidine—formate were performed using standard protocols ( 32 ).

The affinity of MutY for R- and FA-containing duplexes is also greatly enhanced if these analogs are base paired with OG rather than G. Furthermore, the relative ability of the substrate analogs R and FA to mimic A is also accentuated by the presence of OG. This is consistent with the trends that we and others have previously observed indicating that MutY specifically recognizes properties of the base pair containing OG ( 20 , 24 , 27 , 28 , 40 ). Distinct differences in the activity of MutY with OG- versus G-containing substrates have also been observed in pre-steady-state kinetics experiments ( 20 ). Indeed, the intrinsic rate of removal of A from OG:A base pairs is considerably faster than from G:A base pairs. Thus both recognition and catalysis depend on the characteristics of both partners of the OG:A base pair. This may be due to the [OG( syn ):A( anti )] conformation of the base pair, which is more stable and adopted preferentially with OG but not with G ( 6 , 41 ). Assuming that the OG:FA/R base pairs adopt a similar conformation, this may position recognition elements of the substrate analog FA or R for facile recognition by MutY. It is also possible that MutY makes specific contacts with OG that are not possible with G, and these contacts may be required to position the ‘A’ or A analog into a base-specific pocket of MutY ( 19 ).

Lens fcanon

MutY cleavage assays were performed as described previously ( 27 ) using the DNA substrate duplexes described above, where Y = FA, R or A and the Y-containing strand was 32 P-5′-end-labeled. Specific modifications included final DNA and enzyme concentrations of 10 and 300 nM in 20 mM Tris—HCl, pH 7.6, 0.1 μg/μl BSA, 10 mM EDTA, 15 mM NaCl and 10% glycerol.

Storage phosphor autoradiogram of N -glycosylase assays with OG:A-, OG:FA- and OG:R-containing 30 bp duplexes. The duplex used in each assay is denoted at the top and the asterisk (*) denotes the 32 P-labeled strand. Aliquots of 10 nM DNA and 300 nM MutY enzyme were incubated at 37°C for various times. Lanes 1, 6 and 11 are control lanes with no enzyme added; lanes 2–5, OG:A-containing duplex at 5, 15, 30 and 60 min; lanes 7–10, OG:FA-containing duplex at 5, 15, 30 and 60 min; lanes 12–15, OG:R-containing duplex at 5, 15, 30 and 60 min.

The lower the f‑number the larger the aperture diameter, meaning more thermal IR energy is reaching the infrared sensor. This increases the detail, contrast, and overall performance of the thermal infrared cameras, especially for surveillance where long-range detection, recognition and identification is desired in even in low contrast scenes. A thermal camera’s sensitivity is measured by its NETD (Noise Equivalent Temperature Difference). Often described in millikelvins (mK, thousandths of a degree), NETD is determined by the thermal camera’s sensor sensitivity and the f‑stop of the lens.

MutY was purified as described previously ( 27 ) with minor modifications (N.H.Chmiel and S.S.David, unpublished results). The enzyme concentration was determined by the method of Bradford (Bio-Rad) using BSA as a standard ( 30 ). The percent active enzyme of the MutY protein used in the gel retardation and MPE-Fe(II) footprinting experiments was determined to be 85% using standard active site titration methods modified for MutY ( 20 ). The MutY protein used for the DMS footprinting was determined to contain 50% active enzyme. All MutY concentrations listed were corrected for the active enzyme concentration.

The catalytic reaction of MutY produces an OG/G:(AP site) that is monitored in vitro by treatment of the reaction mixture with base (sodium hydroxide or piperidine) to produce strand scission at the AP site ( 20 ). Glycosylase assays using duplexes containing a central mispair of either OG:FA or OG:R were performed and compared to the reaction with the corresponding OG:A duplex ( Fig. 2 ). Each reaction contained an ∼40-fold molar excess of enzyme, and under these conditions full conversion of the OG:A substrate to product was observed within 5 min. In contrast, no detectable cleavage at the FA or R was observed in the corresponding reaction with the OG:FA-containing and the OG:R-containing duplexes. Similarly, MutY was unable to exhibit glycosylase activity with duplexes containing a G:FA/R mispair (data not shown). These results show that the FA- and R-containing duplexes are not substrates for MutY.

MutY has a strong affinity for duplexes containing FA or R analogs opposite OG, as illustrated by the subnanomolar dissociation constants, 0.12 ± 0.02 and 0.05 ± 0.03 nM, respectively. These Kd values for the OG:FA and OG:R duplexes indicate an enhanced affinity over a non-specific DNA duplex containing a G:C base pair (153 ± 64 nM) instead of the mispair of 3000-and 1200-fold, respectively. These values are lower than the Kd value reported for the natural substrate OG:A, which was determined under conditions where MutY was bound to a mixture of substrate and product (0.2 ± 0.1 nM). Furthermore, the Kd values for OG:R- and OG:FA-containing duplexes are lower than that for the corresponding OG:F duplex (0.28 ± 0.02 nM).

Cindy Lou Chepanoske, Silvia L. Porello, Tsuyoshi Fujiwara, Hiroshi Sugiyama, Sheila S. David, Substrate recognition by Escherichia coli MutY using substrate analogs , Nucleic Acids Research, Volume 27, Issue 15, 1 August 1999, Pages 3197–3204, https://doi.org/10.1093/nar/27.15.3197

Schematic representation of the results from footprinting experiments. The central portion of the sequence of the 30 bp duplex is indicated (X = FA, F, R or A). The brackets indicate the region of protection in the MPE-Fe(II) footprinting on the OG:FA-containing duplex. The protected (^) and hyperreactive (↑) guanines in the DMS footprinting are shown. The large and the small symbols represent a larger or a smaller effect, respectively.

Substrate and substrate analogs for MutY. 2′-Deoxyadenosine (A) shown in the top panel is the natural substrate for MutY when opposite G or OG. 2′-Deoxy-2′-fluoroadenosine (FA), 2′-deoxyaristeromycin (R) and 2′-deoxy-formycin (F) mimic structural properties of A.

2′-Deoxyribonucleotides of the following sequences were synthesized: 5′-CGATCATGGAGCCAC X AGCTCCCGTTACAG-3′ and 5′-CTGTAACGGGAGCT Y GTGGCTCCATGATCG-3′, where X = G or OG and Y = A, FA, F or R. The 32 P-5′-end-labeled strands were annealed to 1.5 molar excess of the complementary sequence in a buffer containing 150 mM NaCl, 20 mM Tris—HCl, pH 7.6, 10 mM EDTA. Annealing was facilitated by heating the mixture to 90°C and then slowly cooling to 4°C for 6–8 h. DNA concentrations reported represent the upper limit for the concentration based on the estimated amount of DNA recovered after labeling.

f-stop photography

A unique opportunity exists with MutY to map the course of the glycosylase reaction using the substrate and transition state analogs that have been identified. For example, this information may be obtained via structural studies of MutY bound to various substrate analogs, transition state analogs and the product of the glycosylase reaction. Thus, this type of approach is useful in dissecting the molecular details of recognition and processing of DNA damage by DNA repair enzymes.

The affinity of MutY for the DNA duplexes was reduced when OG was replaced with G, as illustrated by the increase in Kd values. Indeed, the measured Kd values for G:R and G:FA are 4.4 ± 0.6 and 5.8 ± 0.6 nM, respectively. These values are much lower than the Kd for a non-specific duplex. These values are also lower than the reported values for the corresponding G:A-containing duplex (21 ± 4 nM) ( 27 ). However, under these conditions MutY converts G:A substrates completely to product and therefore the dissociation constant measured in this case is representative of MutY association with the product G:(AP site). Thus, MutY binds more tightly to these substrate analogs paired with G than both non-specific DNA and the product. This suggests that the FA- and R-containing duplexes mimic properties of the substrate to provide for an increased association with MutY ( 27 ).

f-number formula

The 2′-deoxy-2′-fluoroadenosine phosphoramidite was kindly provided by Gregg Kamilar and Dr Peter Beal (University of Utah). The 2′-deoxyaristeromycin phosphoramidite was synthesized as previously described ( 29 ). Standard 2′-deoxy-nucleotide-β-cyanoethyl phosphoramidites were purchased from Applied Biosystems or Beckman Instruments. The OG phosphoramidite was purchased from Glen Research. All substrate 2′-deoxyoligonucleotides were synthesized on an Applied Biosystems model 392 or a Beckman Oligo automated oligonucleotide synthesizer following the manufacturer's protocol.

f-number calculator

In addition to the analogs described above, several other substrate and transition state analogs for MutY have been identified ( 14 , 23 ). For example, Bulychev et al. ( 28 ) have reported that MutY binds with nanomolar affinity to a modified substrate containing a methoxy substituent at the 6 position of OG. This modification of the OG provides resistance to the glycosylase activity of MutY; however, the mechanism of this resistance is unclear. An extremely tight binding analog for MutY was prepared by Deng et al. using the concept of transition state mimicry ( 40 ). The pyrrolidine abasic site analog which mimics the accumulation of positive charge at O1′ in the transition state was elaborated to be specific for MutY by appending an adenine base via a methylene linker to the pyrrolidine (referred to as phA). Indeed, a duplex containing a OG:phA base pair exhibited a sub-picomolar dissociation constant for binding to MutY. Bulychev et al. ( 28 ) also reported that OG:R-containing duplexes are non-cleavable substrate analogs for MutY. Surprisingly, the reported Kd of 13 nM is significantly higher than the Kd value reported herein of 50 pM. However, it is not clear that these two values can be directly compared since the two data sets were generated using different DNA duplexes under different conditions. Most notably, the higher Kd value was obtained using a high DNA concentration ([DNA] = 10 nM).

The 5′-end-labeling was performed with T4 polynucleotide kinase purchased from New England Biolabs and [γ- 32 P]ATP from Amersham Life Sciences. Labeled oligonucleotides were purified using mini Quick Spin DNA columns from Boehringer Mannheim. Methidiumpropyl EDTA and dimethyl sulfate were purchased from Sigma and Acros, respectively. Bovine serum albumin (BSA) and Bradford reagents were purchased from Bio-Rad. All other reagents were of analytical purity and purchased from Fisher, US Biochemical or Mallinckrodt Baker. Gel imaging and quantitation were performed on a Molecular Dynamics Storm 840 PhosphorImager using Image-Quant software.

Storage phosphor autoradiogram of MPE-Fe(II) footprinting of MutY with OG:A- and OG:FA-containing 30 bp duplexes. The duplex or single-stranded oligonucleotide assayed in each set of lanes is indicated at the top and the 32 P-labeled strand is marked with an asterisk (*). Lanes 1, 2 and 15, Maxam—Gilbert G+A sequencing reactions; lanes 3, 7 and 11, control reactions without MutY and MPE-Fe(II); lanes 4, 8 and 12, all reagents without MutY; lanes 5, 9 and 13, 250 nM MutY; lanes 6, 10 and 14, 750 nM MutY. Other reaction concentrations are as follows: 10 nM DNA, 25 mM MPE-Fe(II), 500 mM calf thymus DNA.

Lens fvs aperture

A particularly interesting result from the DMS footprinting was the hyper-reactivity of a G residue flanking the F, FA or R in the OG:F/FA/R duplexes. This may be a result of DNA distortion which provides more facile access of DMS to this G residue in the presence of MutY. This hyper-reactivity is observed in the DMS experiments but not in the MPE-Fe(II) experiments. However, a conformational change in the helix that results in disruption of base stacking interactions may not allow the MPE-Fe(II) reagent to intercalate. This will appear as a protected region on the DNA when, in reality, this may be an exposed region without a helical structure. The action of DMS does not depend on the integrity of the DNA helix but on the accessibility of the N-7 of guanine, thus revealing increased exposure of the R/F/FA strand to the solvent upon binding of MutY. This result is in agreement with the nucleotide flipping mechanism proposed for the enzyme based on the crystal structure of the catalytic domain of MutY ( 19 ). Such a mechanism would lead to a conformational change and disruption of the base pairing in the DNA consistent with the described footprinting pattern.

F-stop chart

In the MPE-Fe(II) footprinting with the OG:A mispair-containing duplex in which the A strand was labeled, a band corresponding to the glycosylase activity at the mispaired A was observed ( Fig. 5 , lanes 5 and 6), consistent with our previous MPE-Fe(II) footprinting experiments ( 27 ). The diffuse band on the 3′-side of the A is due to the α,β-unsaturated aldehyde formed via β elimination of the product AP site and has been observed previously in the absence of base treatment ( 34 , 35 ). Addition of 0.1M sodium hydroxide to an MPE-Fe(II) footprinting reaction produced an increase in the band that comigrates with the Maxam—Gilbert A reaction (data not shown) due to β and δ elimination of the AP site to produce the expected phosphate end ( 12 ). Another notable feature of MPE-Fe(II) footprinting reactions with the OGA substrate is the absence of cleavage bands on the 5′-end of the A strand. This result appears to be related to the cleavage reaction and the interaction of MutY with the product of the glycosylase reaction, since under conditions where the duplex has not been completely converted to product (i.e. lower concentrations of MutY) bands on the 5′-end are observed (data not shown). Alteration of the duplex structure due to the cleaved fragment may result in less efficient cleavage by the MPE-Fe(II) reagent. This phenomenon also appears to be a consequence of the intercalating requirement of MPE-Fe(II), since analogous footprinting reactions with non-intercalating agents (e.g. DMS; see Fig. 5 ) produce the expected cleavage bands on the 5′-side of the A.

F-number oflens

The crystal structure of MutY with bound adenine suggests that the mispaired adenine in the corresponding substrate duplex is ‘flipped-out’ of the helix in order to be accommodated into the identified adenine-specific pocket ( 19 ). The hyper-reactivity of G residues on the substrate analog-containing strand toward DMS suggests that MutY may minimally contact the rest of the substrate A-containing strand. However, the protection of the OG-containing strand from DMS by MutY suggests that there may be contacts between MutY and this strand. Similar conclusions may be gleaned from interference footprinting experiments reported by Lu et al. ( 42 ) that showed that ethylation or methylation of a large number of sites on the G-containing strand disrupted binding of MutY while ethylation of only two phosphates positioned directly adjacent to the mismatch on the A-containing strand interfered with binding of MutY. Thus, both the interference footprinting and DMS footprinting results indicate that MutY makes contacts with the OG- or G-containing strand of the duplex, which may be important for promoting extrusion of the misincor-porated A into an adenine-specific pocket of MutY.

In this work, we have shown that MutY binds with high affinity to both R- and FA-containing duplexes using gel retardation assays and footprinting experiments. Indeed, these substrate analogs opposite OG and G bind with higher affinity to MutY than the corresponding F analogs. The tighter association of MutY with duplexes containing R and FA over F may be a consequence of the lack of modification of the adenine hetero-cycle. This idea is supported by the recent crystallographic characterization of the catalytic domain of MutY in the presence of the adenine base ( 19 ). In fact, the structure reveals specific hydrogen bonding interactions between Glu37 and the N-7 of the adenine base. In the C -nucleoside F, the unusual pyrazole ring places a hydrogen bond donor (N1-H) at the corresponding position of the hydrogen bond acceptor N-7 of A, precluding the analogous hydrogen bonding interaction. However, in previous work, we also observed a reduced affinity of MutY for duplexes containing OG/G:Z (where Z = 7-deaza-2′-deoxy-adenosine) base pairs compared to the analogous F-containing duplexes. This is notable, since the Z analog lacks a hydrogen bond donor or acceptor (C7-H) at the analogous position. The features of substrate recognition by MutY may favor F with respect to Z but not with respect to R and FA. Thus, the R and FA analogs will provide information on the MutY—DNA complex that will be complementary to that obtained with other noncleavable analogs such as F and Z.

An f‑number (ƒ/#) or f‑stop refers to the ratio of a lens’s focal length to its aperture’s diameter (lens opening vs focal length) and indicates the amount of IR energy (heat) coming through the lens to the infrared sensor. An f/1.0 lens means the aperture diameter is equal to the focal length, whereas an f/2.0 lens would mean the aperture diameter is half of the focal length. Lenses are typically specified with their maximum aperture (some lens apertures are internally adjustable) and max focal length/zoom power.

In order to determine whether FA and R analogs in mispairs with G and OG are efficiently recognized by MutY, experiments were performed to quantify the binding of MutY to duplexes containing OG/G:FA and OG/G:R mispairs. Non-denaturing gel retardation experiments under conditions where [DNA] > Kd were used to monitor the presence of a specific MutY-DNA complex as a function of increasing [MutY] and therefore determine relative equilibrium dissociation constants ( Kd ) ( 31 ). A plot of the percent OG:FA- and G:FA-containing duplexes bound as a function of the MutY concentration that was used to determine Kd values is shown in Figure 3 . The Kd values determined for duplexes containing FA or R opposite OG or G are listed in Table 1 . These values are compared to Kd values for the corresponding duplexes containing C and A reported previously under similar conditions ( 27 ).

Quantitative gel retardation assays ( 31 ) were performed using the DNA substrate duplexes described above, where Y = FA, F or R, and the Y-containing strand was 32 P-5′-end-labeled. Conditions were modified slightly from the previous work ( 27 ). Reactions contained 10 pM duplex, 20 mM Tris—HCl pH 7.5,100 mM NaCl, 1 mM EDTA, 1 mM DTT, 10% glycerol, 0.1 μg/μl BSA and varying amounts of MutY in a reaction volume of 100 μl. MutY solutions of varying concentrations were freshly prepared by diluting aliquots of MutY at 4°C with dilution buffer containing 20 mM Tris—HCl pH 7.5, 10 mM EDTA and 20% glycerol. The samples containing DNA and MutY were incubated at 25°C for 20–35 min. After addition of 1 μl of non-denaturing loading dye (0.25% bromophenol blue, 0.25% xylene cylanol, 30% glycerol in 1× TBE), the samples were electrophoresed at 4°C on a 6% non-denaturing poly-acrylamide gel (17 × 14 × 0.3 cm) with 0.5× TBE buffer at 200 V for 25 min followed by 100 V for 90 min. The gel was dried and exposed to a Molecular Dynamics phosphorimager screen for at least 18 h. Kd values were determined by fitting the data (percent bound substrate versus log[MutY]) to the equation for one-site ligand binding using the program Grafit v.4.09 (Erithacus). Kd values were determined from four to six separate experiments.

The 30 bp duplex containing a central OG opposite A, R, FA or F was 32 P-5′-end-labeled on the appropriate strand. Reaction mixtures (10 μl) containing 5 nM duplex and 0,100 or 200 nM MutY were incubated for 15 min at room temperature in 50 mM sodium cacodylate and 1 mM EDTA. A solution containing 10% DMS in ethanol was added to provide a final DMS concentration of 1%, and the mixture was incubated for 10 min at room temperature. The reaction was quenched by the addition of 10 μl of a DMS stop solution (1.5 M sodium acetate and 1 mM β-mercaptoethanol, pH 7.0) and 200 μl of ethanol and 2 μl of calf thymus DNA (25 mM) were also added to the reaction mixture. The samples were placed in dry ice for 15 min, centrifuged and the resalting pellet washed with 70% ethanol and dried under vacuum. The DNA was then treated with 50 μl of 1M piperidine for 30 min at 90°C and dried in vacuo . Denaturing loading buffer was added to the dried samples which were then electrophoresed on a 20% acrylamide (8 M urea and 1× TBE) gel at 1600 V for 3 h and exposed to X-ray film overnight. The entire procedure was repeated at least three times using freshly labeled DNA duplexes. The protection versus hyper-reactivity was confirmed by quantitation of the storage phosphor auto-radiograms using ImageQuant software. The reported quantitation represents the percent cleavage of a given band compared to the total cleaved DNA in each lane.

We thank Ms Angela Vetsch and Mr Chad Jariangprasert for initial contributions to this research. We also thank Mr Mike Langer for purifying and determining the active site concentration of MutY used in this work. S.S.D. is an A.P. Sloan Research Fellow (1998–2000). This work was supported by National Institutes of Health grant CA67985 (S.S.D.) and the University of Utah.

Plot of percent bound DNA determined from quantitation of gel retardation experiments as a function of MutY concentration. The MutY concentration is shown on a logarithmic scale. The open circle represents the percent bound for OG:FA-containing duplexes and the closed circle represents the percent bound for G:FA-containing duplexes. The determined Kd values are 0.12 ± 0.02 and 5.8 ± 0.6 nM for the OG:FA- and G:FA-containing duplexes, respectively.

The Escherichia coli adenine glycosylase MutY is involved in the repair of 7,8-dihydro-8-oxo-2′-deoxy-guanosine (OG):A and G:A mispairs in DNA. Our approach toward understanding recognition and processing of DNA damage by MutY has been to use substrate analogs that retain the recognition properties of the substrate mispair but are resistant to the glycosylase activity of MutY. This approach provides stable MutY—DNA complexes that are amenable to structural and biochemical characterization. In this work, the interaction of MutY with the 2′-deoxyadeno-sine analogs 2′-deoxy-2′-fluoroadenosine (FA), 2′-deoxyaristeromycin (R) and 2′-deoxyformycin A (F) was investigated. MutY binds to duplexes containing the FA, R or F analogs opposite G and OG within DNA with high affinity; however, no enzymatic processing of these duplexes is observed. The specific nature of the interaction of MutY with an OG:FA duplex was demonstrated by MPE-Fe(II) hydroxyl radical foot-printing experiments which showed a nine base pair region of protection by MutY surrounding the mispair. DMS footprinting experiments with an OG:A duplex revealed that a specific G residue located on the OG-containing strand was protected from DMS in the presence of MutY. In contrast, a G residue flanking the substrate analogs R, F or FA was observed to be hypersensitive to DMS in the presence of MutY. These results suggest a major conformational change in the DNA helix upon binding of MutY that exposes the substrate analog-containing strand. This finding is consistent with a nucleotide flipping mechanism for damage recognition by MutY. This work demonstrates that duplex substrates for MutY containing FA, R or F instead of A are excellent substrate mimics that may be used to provide insight into the recognition by MutY of damaged and mismatched base pairs within DNA.

Insight into substrate recognition by DNA repair enzymes has been facilitated by the use of site-directed mutagenesis to alter catalytically relevant residues followed by co-crystallization of the modified protein bound to its substrate DNA ( 21 , 22 ). Recently, designed synthetic DNA analogs or the corresponding 2′-deoxynucleotide derivatives of natural products have been incorporated into DNA to study the damage recognition properties of many BER enzymes ( 14 , 23 ). The resulting substrate and transition state analogs are excellent tools for biochemical and structural studies since they are specifically recognized but not cleaved by BER glycosylases. Indeed, Verdine and coworkers have synthesized a pyrrolidine abasic site analog that mimics the positive charge distribution of the transition state species and showed that it can be used as a general inhibitor to characterize many BER glycosylases ( 24 ). In contrast, designing analogs that destabilize the transition state of the catalytic reaction can be employed for examination of the enzyme—substrate complex. For instance, substrates that are fluorinated at the 2′-position of the deoxyribose ring use the electron-withdrawing nature of the fluorine atom to destabilize the positive charge of the transition state ( 25 , 26 ). This type of analog abolishes or significantly reduces the catalytic action of DNA glycosylases. Previously, we have characterized the interaction of MutY with natural product analogs as structural mimics for 2′-deoxy-adenosine incorporated into a DNA duplex opposite G or OG ( 27 ). These analogs, 2′-deoxytubercidin (7-deaza-2′-deoxyadeno-sine, Z) and 2′-deoxyformycin A (F) opposite OG, bind to MutY with high affinity ( Kd < 5 nM) and contain structural properties which provide resistance to the glycosylase activity of MutY.

8-Oxo-7,8-dihydro-2μ-deoxyguanosine (OG) is the most stable product formed in vivo by reactive oxygen species ( 4 ). This damaged nucleotide arises primarily via direct oxidation of a guanine base in a DNA strand ( 5 ). Replication events can lead to the incorporation of an adenine (A) opposite OG, forming a relatively stable base pair ( 6 ). If this OG:A base pair is not repaired, a permanent G:C→T:A transversion mutation is created. Escherichia coli MutY is part of an intricate pathway dedicated to the repair of oxidative damage to DNA that includes MutM (Fpg) and MutT ( 7 , 8 ). MutY is a base excision repair (BER) adenine glycosylase specific for OG:A mispairs that removes the undamaged adenine by hydrolysis of the N -glycosidic bond to form an apurinic (AP) site ( 9 , 10 ). The resulting AP site is further processed and repaired by additional enzymes in the BER pathway, including AP endo-nucleases, DNA polymerase and DNA ligase ( 7 , 9 , 11 , 12 ). MutY also has activity toward G:A and C:A mispairs ( 10 , 13 , 14 ).

The footprinting reagent DMS reacts with the N-7 of guanines and the N-3 of adenines to produce piperidine-labile sites ( 32 ). DMS experiments were performed on OG:A, OG:F, OG:FA and OG:R duplexes under conditions where only the N-7 of G was subject to methylation. In these experiments with the OG:A-containing substrate, protection by MutY from reaction with DMS of a G located 2 bp on the 3′-side to the OG was observed ( Fig. 5 , lanes 1–4). Using the same duplex in which the complementary A-containing strand is labeled, a band corresponding to cleavage at the A was observed resulting from base-sensitive cleavage of the AP site generated by the glycosylase reaction of MutY ( Fig. 5 , lanes 5–8). When adenine analogs F, FA or R are incorporated opposite OG in the DNA duplex, only a weak protection of the G on the 3′-side of the OG was observed (data not shown). In contrast, in the corresponding region on the F-containing strand where a footprint was observed with the MPE-Fe(II) reagent, two G residues exhibit an increased reaction with the DMS reagent in the presence of MutY ( Fig. 5 , lanes 9–12). This hyper-reactivity toward DMS is particularly strong at the G flanking the F on the 3′-side. Quantitation of this hyper-reactive band from five separate experiments showed that this band was 1.6- to 2-fold more intense than the corresponding band in lanes without MutY. Similar results were observed with the corresponding duplexes containing the R and FA analogs. These results imply that the strand containing the substrate analog becomes more accessible to DMS upon MutY binding. This suggests that a conformational change in the DNA helix accompanies binding of MutY, thereby exposing this strand. The protection and hyper-reactivity patterns from both DMS and MPE-Fe(II) footprinting on the 30 bp duplex are summarized in Figure 6 .

In this work, the interaction of MutY with analogs based on transition state destabilization via designed inhibitors and natural product 2′-deoxynucleotides is characterized. We report 2′-deoxy-2′-fluoroadenosine (FA) as a new analog that is specifically recognized opposite G or OG within DNA by MutY. Furthermore, the binding of this new analog to MutY is compared to the substrate analogs 2′-deoxyaristeromycin (R) ( 28 ) and F ( Fig. 1 ). We show that MutY binds with high affinity ( Kd in the picomolar range) to duplexes containing FA and R opposite OG. The interaction of MutY with these duplexes is further characterized using MPE-Fe(II) and DMS footprinting experiments.

Lens fcamera

Autoradiogram of DMS footprinting of MutY on OG:A- and OG:F-containing 30 bp duplexes. At the top of each set of four lanes, the central base pair and the base in the 32 P-labeled strand (*) is designated. Lanes 1, 5 and 9, control reaction with no MutY and no DMS; lanes 2, 6 and 10, DMS reaction in the absence of MutY; lanes 3, 7 and 11, DMS reaction in the presence of 100 nM MutY; lanes 4, 8 and 12, DMS reaction in the presence of 200 nM MutY. The concentrations of DNA and DMS were 5 nM and 1%, respectively. In this particular experiment, the percent protection (lanes 3 and 4) was 30%. The percent intensity of the hyper-reactive bands are 1.6× (lane 7) and 1.9× (lane 8) greater than the control (lane 6). This autoradiogram was representative of five separate DMS experiments.

An ƒ/1.0 Ge lens allows for 250% more infrared thermal energy to be transferred to the infrared sensor than an ƒ/1.6 lens. This means the f‑number can be even more important than the sensor as it relates to a camera’s overall NETD sensitivity, particularly for uncooled LWIR cameras which are the most common type of thermal imaging. See the following image for a visual example of how much the area of the aperture changes with different f‑stop values.

Biochemical and structural studies of BER enzymes are beginning to provide insight into the remarkable ability of these enzymes to efficiently locate and remove aberrant bases from DNA. A relatively new approach to examining BER enzymes is the use of synthetic or natural product, substrate and transition state analogs to provide stable BER glycosylase—DNA complexes for structural and biochemical studies. This approach has recently been exploited to provide a structural ‘snap-shot’ of the human 3-methyladenine-DNA glycosylase complexed to DNA containing the transition state mimic of the glycosylase reaction, the pyrrolidine abasic nucleotide ( 36 ). In addition, 2′-fluoro-2′-deoxyuridine substrate analogs have been used in fluorescence experiments to provide a detailed description of the kinetic properties of base flipping with uracil-DNA glyco-sylase ( 37 ). Not only are these analogs useful for characterizing the BER enzyme—DNA complex, they can also be useful tools in screening and isolating BER enzymes. For example, the yeast OG glycosylase (yOGG1) was isolated from yeast extracts using an oligonucleotide duplex containing a reduced abasic site as an affinity matrix ( 38 ). In our laboratory, we have capitalized on the high affinity and specificity of MutY for OG:F-containing duplexes to screen the DNA-binding properties of mutated forms of MutY in cellular lysates ( 39 ).

The use of the substrate analogs provides stable MutY—DNA complexes that are amenable to further characterization using biochemical footprinting techniques ( 12 ). We have previously shown that duplexes containing an OG:F central mispair are excellent templates for hydroxyl radical footprinting using MPE-Fe(II) ( 27 ). In these previous experiments, we observed a region of ∼ 10–12 bp surrounding the mispair that was protected from hydroxyl radical cleavage in the presence of MutY.

Enzymes involved in the BER pathway are highly conserved between species, and human homologs of well-characterized repair proteins from E.coli are rapidly being discovered ( 12 ). Eukaryotic enzymes homologous to MutY have been identified, for example, in Schizosaccharomyces pombe (SPMYH) and in humans (MYH), that exhibit similar enzymatic properties ( 15–17 ). Recent structural and biochemical data are providing important insights into factors affecting recognition and repair by MutY. Importantly, the structure of a truncated form of MutY representative of the catalytic core of the enzyme ( 18 ) was solved by X-ray crystallography ( 19 ). In addition, pre-steady-state kinetics of MutY showed significant differences in the rates of the glycosylase reaction depending on whether the substrate contained an OG:A or G:A base pair. ( 20 ). The kinetic experiments also showed that MutY remains tightly bound to the product of its glycosylase activity with OG:A substrates, which is likely an important feature of the in vivo processing of OG:A mismatches by MutY. The properties of substrate recognition and repair by MutY are likely to be shared with the entire class of BER glycosylases, including those in higher organisms.

The trends accompanying the replacement of the substrate analogs FA and R by C in OG- versus G-containing duplexes are also notable. The binding of OG:R-containing duplexes by MutY represents an ∼500-fold greater affinity versus the OG:C duplex, while the binding of G:R-containing duplexes by MutY represents an ∼35-fold greater affinity over the G:C duplex. Thus, tight association of MutY with the OG:R duplex results from the presence of both the substrate analog and OG. Similarly, the binding of OG:FA-containing duplexes by MutY represents an ∼200-fold greater affinity relative to the OG:C duplex, and the binding of G:FA-containing duplexes by MutY represents an ∼25-fold greater affinity over the G:C duplex. These results indicate that the R and FA analogs effectively mimic the recognition properties of 2′-deoxyadenosine, and this ability can be enhanced by pairing these analogs with the more relevant in vivo partner, OG.

The MPE-Fe(II) footprinting experiments provide additional evidence that the OG:FA base pair is specifically recognized by MutY in a manner similar to that observed for the OG:F base pair. However, the size and asymmetry of the protected region surrounding the OG:FA base pair is slightly different than in the OG:F duplex. This indicates that there may be subtle differences in the binding of MutY to duplexes containing different substrate analogs. DMS footprinting provides additional insight into recognition of the substrate and product by MutY. The observation of protection from DMS of a G residue near the OG indicates a close interaction of MutY with the OG strand of the duplex. DMS methylates the N-7 of G, which is located in the major groove of duplex DNA, and therefore, the protection suggests that at least a portion of MutY recognition of its substrate is directed from the major groove side of the duplex. However, if there is considerable DNA distortion and duplex melting, this could significantly alter the orientation of the bases proximal to the mismatch. Notably, the relative amount of protection of this G within the OG strand provided by MutY is greater with the substrate OG:A duplex compared to the substrate analog-containing duplexes. The binding of MutY to the OG:A-containing duplex results in adenine removal, as evidenced by complete cleavage at the ‘A’ in the A-containing strand. Therefore, the footprint observed in this case results from MutY bound to the OG:(AP site) product , while the footprint resulting when using non-cleavable substrate analogs is more representative of binding of MutY to the substrate . The product release in the glycosylase reaction of MutY is extremely slow ( 20 ), suggesting a very tight interaction between the enzyme and its product. Thus, it is possible that the nature of the interaction of MutY with the OG strand in the product could be significantly different compared to the binding to the substrate (analog) and therefore, produce different footprinting patterns.

The cell must counter the onslaught of cytotoxic and mutagenic agents targeted at DNA to ensure survival. DNA modifications can arise from a variety of sources, including spontaneous hydrolysis, deamination, oxidation and alkylation ( 1 ). Furthermore, errors in DNA replication may inaccurately preserve genomic information ( 1 ). Chances for normal cell survival are increased by utilizing pathways that can efficiently recognize and repair DNA damage ( 2 , 3 ).

MPE-Fe(II) footprinting experiments ( 33 ) were performed with duplexes containing OG:FA and OG:A in the presence of increasing amounts of MutY ( Fig. 4 ). In each case, the OG-containing strand or the A/FA-containing strand was 32 P-5′-end-labeled. Maxam-Gilbert G+A reactions ( 32 ) were performed on the OG-, A- and FA-containing strands to determine the location of the central mispair and adjacent bases in the sequences ( Fig. 4 , lanes 1, 2 and 15). Notably, the FA nucleotide does not react in the sequencing reactions, which is consistent with the anticipated resistance of this nucleotide to acid-catalyzed depurination. In the MPE-Fe(II) reaction with the OG:FA duplex, a region of ∼8 nt on both the OG- and FA-containing strands was protected from hydroxyl radical cleavage in the presence of MutY ( Fig. 4 , lanes 9, 10, 13 and 14). Thus, MutY provided protection of a 9 bp stretch surrounding the mispair from hydroxyl radical cleavage. The region of protection surrounding the OG:FA base pair is illustrated in Figure 6 . The observed region of protection with the OG:FA duplex was similar, but not identical, to that previously determined from analogous experiments using the corresponding OG:F-containing duplex. With the OG:FA duplex, the stretch of protected nucleotides on the two strands of the duplex was offset by only 1 bp on each strand, in contrast to the corresponding experiments with the OG:F duplex, in which the observed offset was 2 bp. This results in a slightly smaller duplex region of protection of the OG:FA duplex by MutY. As expected, no protection by MutY was offered in analogous MPE-Fe(II) footprinting experiments with a duplex containing a G:C base pair in the position of the OG:FA base pair (data not shown).

Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide