APE1 active site residue Asn174 stabilizes the AP site and is essential for catalysis

Abstract

Apurinic/apyrimidinic (AP) sites are common and highly mutagenic DNA lesions that can arise spontaneously or as intermediates during base excision repair. The enzyme apurinic/apyrimidinic endonuclease 1 (APE1) initiates repair of AP sites by cleaving the DNA backbone at the AP site via its endonuclease activity. Here, we investigated the functional role of the APE1 active site residue N174 that contacts the AP site during catalysis. To do so, we rationally designed four APE1 mutants (N174A, N174D, N174Q, and N174K) that altered the hydrogen bonding potential, electronegativity, and side chain length of residue 174. Kinetic analysis determined that AP site cleavage was impaired in our APE1_N174A, APE1_N174D, and APE1_N174K mutants, but not in our APE1_N174Q mutant. Using X-ray crystallography and molecular dynamic simulations of our APE1 mutants, we attributed the poor cleavage of our APE1_N174A and APE1_N174D mutants to lack of hydrogen bonding between residue 174 and the AP site. From these experiments, we also discovered that the altered charge of residue 174 in APE1_N174D and APE1_N174K mutants contributed to impaired cleavage. From analysis of our mutants, we determined that in APE1_WT, N174 stabilizes AP DNA within the APE1 active site via hydrogen bonding promoting cleavage of AP sites. We also found that the neutral charge of N174 in APE1_WT contributes to an optimal electrostatic environment during the APE1 endonuclease reaction resulting in efficient AP site cleavage. Cumulatively, we demonstrate the importance of N174 in APE1’s function and provide new insights into the molecular mechanism by which APE1 processes AP sites during DNA repair.

Oxidative DNA damage can result from exposure to oxidizing agents, either in the environment or generated as byproducts of cellular metabolism (1, 2, 3, 4). Repair of oxidative DNA damage is essential to protect genome stability and prevent deleterious mutations that give rise to multiple human diseases (5, 6, 7). This repair is primarily accomplished via the base excision repair (BER) pathway (8, 9, 10, 11). During BER, damaged DNA nucleobases are identified and removed by a damage-specific DNA glycosylase that generates baseless sugar moieties known as apurinic/apyrimidinic (AP) sites (12, 13, 14, 15). AP sites are then processed by apurinic/apyrimidinic endonuclease 1 (APE1), which cleaves the phosphodiester DNA backbone at the 5′ end of the AP site using its endonuclease activity (16, 17, 18, 19, 20). This substrate is subsequently processed by downstream BER proteins to complete repair of the damaged base (11, 21, 22). APE1 cleavage of AP sites is an essential process to protect genome stability during DNA repair as the knockout of APE1 is embryonic lethal (18). Similarly, expression of inactive APE1 variants sensitizes cells to DNA-damaging agents (23, 24, 25, 26). Because APE1 cleavage of AP sites constitutes an important step in DNA repair, investigation of the APE1 cleavage mechanism and key active site residues is of interest to more broadly understand how DNA damage is processed by this essential enzyme.

Previous high-resolution X-ray crystal structures of APE1 bound to AP site containing DNA, both before and after cleavage, have provided insight into the mechanism of APE1 cleavage and APE1 active site organization (Fig. 1, A and B) (27, 28, 29). Upon binding an AP site, APE1 flips the AP site out of the DNA double helix and into an extrahelical position within the compact APE1 active site (Fig. 1A) (27, 28). The AP site is positioned near active site residues E96, Y171, D308, and H309 along with a single Mg²⁺ ion (Fig. 1B) (30, 31). This places the phosphate backbone of the AP site in position for cleavage through a one metal hydrolysis reaction (28, 31, 32). To coordinate this reaction, a water molecule is positioned for inline nucleophilic attack on the phosphorus atom of the AP site through hydrogen bonding interactions with residues N212 and D210 (Fig. 1B) (28, 33). D210 deprotonates the water in this position, activating the water for nucleophilic attack and initiating hydrolysis (28, 29). After cleavage, the DNA product consists of a 5′-sugar phosphate terminus and a 3′-hydroxyl terminus. In this product state, N212 and D210 are within hydrogen bonding distance of the 5′ phosphate terminus of the cleaved AP site (Fig. 1C) (28). Similarly, E96, Y171, D308, H309, and the Mg²⁺ cofactor maintain coordination of the phosphate group, now the 5′ terminus, as well as the newly formed 3′ hydroxyl terminus (Fig. 1C) (27, 28, 30).

Figure 1.

The structure of APE1_WT and N174 mutations.A, overview of APE1_WT bound to a DNA substrate containing an AP site (PDB ID: 5DFI). B, close up of the APE1 active site substrate complex (PDB ID: 5DFI) bound to a DNA substrate containing an AP site. Waters are shown as blue spheres and a modeled Mg²⁺ ion is shown as a red sphere. Distances are in angstroms. 5DFI lacks a metal in the structure, but Mg²⁺ ions coordinate the APE1_WT active site in both substrate and product complexes. We have estimated the position of Mg²⁺ in the substrate complex by alignment with its position in the product complex 5DFF (RMSD = 0.301 of 3945 atoms). C, close up of the active site of the APE1_WT product complex (PDB ID: 5DFF). Waters are shown as blue spheres; Mg²⁺ is shown as a red sphere. D, side chains of mutations at residue 174: N174A, N174D, N174Q, and N174K. AP, apurinic/apyrimidinic; APE1, apurinic/apyrimidinic endonuclease 1.

While many of the residues in the APE1 active site have been investigated, there is a glaring lack of knowledge about active site residue N174. X-ray crystal structures of APE1 indicate that N174 contacts the AP site (Fig. 1, B and C), but the role of N174 during AP site cleavage is unknown. Interestingly, N174 is highly conserved among chordate APE1 and shares similarity with Q112 in the APE1 homolog and primary AP site endonuclease in Escherichia coli, EXOIII (Fig. S1) (34, 35, 36, 37, 38, 39, 40). This suggests N174 may be required for APE1 cleavage of AP sites. In the human APE1 substrate complex, N174 is within hydrogen bonding distance of N212 and the AP site bridging oxygen (O5) and deoxyribose sugar oxygen (O4) (Fig. 1B) (27, 28). In the product complex, N174 maintains hydrogen bonding distance to O4 and O5 of the AP site after cleavage (Fig. 1C) (27, 28). Because of its unique position near the AP site, as well as its direct interaction with the phosphodiester backbone at this location, we hypothesized that N174 may stabilize substrate DNA during the cleavage reaction. To assess the role of N174 during APE1 catalysis, we rationally designed four mutations of N174: N174A, N174D, N174Q, and N174K. We investigated these APE1 variants using kinetic, structural, and computational approaches. From these studies, we demonstrated that N174 is required to efficiently position the AP site in the APE1 active site during catalysis. In addition, N174 establishes an electrostatic environment required for catalysis. These observations expound upon the mechanism of the APE1 cleavage reaction and provide insight into the role of N174 during AP site cleavage by APE1.

Results

To determine the role of N174 during APE1 cleavage, we generated four APE1 mutants (APE1_N174A, APE1_N174D, APE1_N174Q, and APE1_N174K) that altered distinct attributes of the amino acid side chain at residue 174: hydrogen bonding potential, charge, and size (Fig. 1D). The N174A mutation has no hydrogen bonding potential, no charge or polarity, and has a smaller side chain length than N174. The N174D mutation has altered hydrogen bonding potential due to lack of hydrogen bonding donors and has a negative charge compared to N174 but is the same size as N174. The N174Q mutation has the same hydrogen bonding potential and charge compared to N174 but has a larger side chain size than N174. And finally, the N174K mutation has altered hydrogen bonding potential due to lack of hydrogen bonding acceptors, a positive charge compared to N174, and a larger side chain size than N174. Using these three APE1 mutants, we analyzed how hydrogen bonding, charge, and size of N174 affect APE1 cleavage of AP sites.

N174A, N174D, and N174K mutations inhibit APE1 cleavage of AP sites

Prior to cleavage, N174 is within hydrogen bonding distance of O4 and O5 of the AP site (Fig. 1B). We hypothesized that mutation of this residue would destabilize the substrate DNA and impair catalysis. To test this, we observed cleavage of DNA containing a tetrahydrofuran (THF) AP site analog by APE1_WT and each of our APE1 mutants using single-turnover kinetic analysis (Table 1, Figs. S2 and S3). Under single-turnover conditions, the rate constant that corresponds to cleavage of AP DNA by APE1_WT, k_obs, is estimated at ≥850 s⁻¹ (41). This k_obs value is provided as an estimate because the rate of cleavage by APE1_WT exceeds measurements quantifiable by rapid quench-flow methodology (41). Unlike APE1_WT, we found all of our APE1 mutants exhibited quantifiable and comparatively lower rates of AP DNA cleavage. The APE1_N174A mutant had a k_obs of 0.039 ± 0.003 s⁻¹, corresponding to a 22,000-fold reduction in k_obs compared to APE1_WT. The APE1_N174D mutant had a k_obs of 0.0026 ± 0.0010 s⁻¹, corresponding to a 330,000-fold reduction in k_obs compared to APE1_WT. The APE1_N174Q mutant had a k_obs of 2.9 ± 0.3 s⁻¹, corresponding to a 290-fold reduction in k_obs compared to APE1_WT. Lastly, the APE1_N174K mutant had a k_obs of 0.11 ± 0.01 s⁻¹, corresponding to a 7700-fold reduction in k_obs compared to APE1_WT. Overall, our single-turnover kinetic experiments indicate that residue N174 is required for cleavage in APE1_WT. The greatest change in cleavage was observed upon N174D mutation followed by the N174A and N174K mutations, respectively. This indicates that the hydrogen bonding potential, charge, and potentially size that is altered in these mutants affected APE1 cleavage. In contrast, the N174Q mutation, which maintains the hydrogen bonding and charge of N174, more moderately impacted the cleavage rate of APE1. These findings are consistent with evidence that the hydrogen bonding and charge of N174 are required for catalysis in APE1_WT.

Table 1.

Summary of kinetic rate constants and binding affinities for APE1_WT, APE1_N174A, APE1_N174D, APE1_N174Q, and APE1_N174K

	APE1WT	APE1N174A	APE1N174D	APE1N174Q	APE1N174K
kobs (s−1)	≥850a	0.039 ± 0.003	0.0026 ± 0.0010	2.9 ± 0.3	0.11 ± 0.01
Fold change in kobs	-	22,000 x ↓	330,000 x ↓	290 x ↓	7,700 x ↓
KD App (nM)	0.9 ± 0.4	44 ± 3	13 ± 5	3.7 ± 1.8	17 ± 7
Fold change in KD App	-	50 x ↑	10 x ↑	4 x ↑	20 x ↑

Rate constants corresponding to cleavage (k_obs) were determined by single-turnover kinetic analysis (mean ± SD, N = 3–4). Apparent binding affinities (K_{D App}) were determined by EMSA (mean ± SD, N = 3–5). Fold changes are relative to APE1_WT values. Italics are used to indicate the appropriate nomenclature for equilibrium dissociation and kinetic rate constants.

APE1, apurinic/apyrimidinic endonuclease 1.

^aAs determined by the following reference: (41). See methodology for additional information.

To determine if N174 contributes to the ability of APE1 to bind DNA, we performed EMSAs. These experiments enabled us to determine the apparent binding affinity (K_{D App}) of APE1 for DNA containing an AP site (Table 1, Figs. S4 and S5). APE1_WT had a K_{D App} of 0.9 ± 0.4 nM. The APE1_N174A mutant had a K_{D App} of 44 ± 3 nM, corresponding to a 50-fold decrease in apparent binding affinity compared to APE1_WT. The APE1_N174D mutant had a K_{D App} of 13 ± 5 nM, corresponding to a 10-fold decrease in apparent binding affinity compared to APE1_WT. The APE1_N174Q mutant had a K_{D App} of 3.7 ± 1.8 nM, corresponding to a 4-fold decrease in apparent binding affinity compared to APE1_WT. And lastly, the APE1_N174K mutant had a K_{D App} of 17 ± 7 nM, corresponding to a 20-fold decrease in apparent binding affinity compared to APE1_WT. This indicates that residue N174 does contribute to AP DNA binding, though the effect on the K_{D App} is modest compared to the effect on k_obs.

N174A mutation destabilizes the AP site within the APE1_N174A active site

Previously solved X-ray crystal structures of the APE1_WT active site revealed that N174 is within hydrogen bonding distance of O4 and O5 of the AP site prior to cleavage (Fig. 1B) (27, 28). Therefore, we predicted that N174 would contribute to APE1 cleavage by stabilizing substrate DNA during catalysis. To test this, we crystallized the APE1_N174A mutant protein bound to a 21-mer dsDNA oligo containing a centrally located THF AP site analog. This substrate complex diffracted to 2.29 Å resolution in the P1 space group (Fig. 2, A and B, Table 2). We have clear density for the N174A mutation as well as the AP site within the APE1_N174A mutant active site (Fig. 2, A and B). Overlay of the APE1_WT and APE1_N174A substrate complexes revealed similar active sites, with no notable differences in the position of key active site residues (RMSD = 0.299 of 2199 atoms, Fig. 2C) (28). In contrast, shifts to the DNA within the active site are observed, with the AP site, phosphate backbone, and surrounding DNA adopting a different position in the APE1_N174A mutant compared to APE1_WT (Fig. 2D). Specifically, the AP site phosphate group is shifted 1.1 Å out of the APE1 active site in the APE1_N174A mutant compared to APE1_WT (Fig. 2D). This shift increases the distance between the phosphate of the AP site and the nucleophilic water, ultimately resulting in misalignment of the phosphate backbone for attack by the nucleophilic water (Fig. 2E). This loss of alignment does not appear to arise from changes in position of active site residues since the positions of both the nucleophilic water and the active site residues which coordinate it, N212 and D210, are not changed in the APE1_N174A mutant (Fig. 2C). Based on these results, we conclude that the N174A mutation compromises hydrogen bonding between residue 174 and O4 and O5 of the AP site. Therefore, in APE1_WT, the hydrogen bonds between N174 and the AP site are necessary to position the phosphate backbone of the AP site within the APE1 active site to support nucleophilic attack during catalysis.

Figure 2.

The APE1_N174A mutant substrate complex.A–B, active site of the APE1_N174A substrate complex crystal structure is shown in green. Density is shown as gray mesh. Waters are shown as blue spheres. C, active site residues of the APE1_N174A substrate complex crystal structure compared to active site residues of the APE1_WT substrate complex is shown in gray (5DFI). D, the DNA backbone in the APE1_N174A mutant compared to APE1_WT. E, distance between the nucleophilic water and DNA backbone increases in the APE1_N174A substrate complex crystal structure compared to APE1_WT. The nucleophilic water is shown as a blue sphere in the APE1_N174A complex and as a gray sphere in the APE1_WT complex. APE1, apurinic/apyrimidinic endonuclease 1.

Table 2.

X-ray structure data collection and atomic model refinement statistics

Data collection	APE1N174A: Substrate DNA complex	APE1N174A: Product DNA complex	APE1N174D: Product DNA complex	APE1N174Q: Product DNA complex
Space group	P1	P1	P1	P43
Cell dimensions
a, b, c (Å)	43.9, 60.5, 73.1	44.3, 61.8, 72.6	44.2, 60.8, 73.6	153.4, 153.4, 45.4
α, β, γ (°)	81.8, 75.6, 88.7	83.0, 78.3, 87.2	82.6, 77.2, 85.6	90.0, 90.0, 90.0
Resolution (Å)	25.06–2.29	34.03–1.97	24.95–2.10	48.52–2.25
Rmeas (%)	0.133 (0.215)	0.053 (0.334)	0.145 (0.759)	0.100 (0.656)
Rmerge	0.096 (0.157)	0.053 (0.291)	0.128 (0.581)	0.102 (0.527)
I/σI	7.61 (4.39)	27.3 (4.36)	7.52 (0.563)	13.7 (0.846)
cc1/2	(0.915)	(0.934)	(0.646)	(0.616)
Completeness (%)	96.6 (97.3)	99.7 (98.2)	98.2 (91.2)	98.5 (86.7)
Redundancy	2.5 (2.6)	3.9 (3.6)	3.3 (1.8)	5.6 (2.4)
Refinement
Resolution (Å)	25.06–2.29	34.03–1.97	24.95–2.10	48.52–2.25
No. reflections	29,791	76,015	35,000	63,992
Rwork/Rfree (%)	22.96/25.08	18.52/20.33	25.34/27.44	21.49/24.79
No. atoms
Protein	4156	4307	4294	4388
DNA	839	836	836	1674
Water	274	515	298	252
B factors (Å2)
Protein	35.79	27.84	34.39	29.26
DNA	52.26	45.55	52.53	59.12
Water	37.99	34.64	34.04	31.26
RMSDs
Bond length (Å)	0.012	0.011	0.012	0.013
Bond angles (°)	1.705	1.390	1.754	1.816
PDB ID	9DP1	9DP2	9DP3	9DP4

Parentheses indicate value for highest resolution shell.

APE1, apurinic/apyrimidinic endonuclease 1.

We next crystallized APE1_N174A with the same AP DNA in the presence of MnCl₂, promoting APE1 cleavage of the AP site (31, 42). From these crystals, we obtained a 1.97 Å structure of the APE1_N174A mutant product complex in the P1 space group (Table 2). Within the active site of the APE1_N174A product complex, there is clear density for the cleaved DNA backbone and bound Mn²⁺ (Fig. 3, A and B). We overlayed the APE1_N174A mutant product complex with the APE1_WT product complex and observed similar active site conformations (RMSD = 0.208 of 2449 atoms, Fig. 3C) (28). An exception was active site residue N212, which was positioned differently in the APE1_N174A mutant compared to APE1_WT. Prior to cleavage in APE1_WT, the carbonyl group of N212 orients the nucleophilic water through a hydrogen bonding interaction (Figure 1, Figure 3D and 1B). Following cleavage, N212 undergoes a rotameric shift, positioning its amine group within hydrogen bonding distance of the 5′-phosphate terminus (Figure 1, Figure 3D and 1C). In the APE1_N174A mutant, N212 fails to make this rotameric shift and remains in the same position as in the substrate structure (Fig. 3, B and D). We expect that N174A mutation causes an unfavorable environment for the N212 rotameric shift, but that the N212 rotameric shift is not essential to cleavage.

Figure 3.

The APE1_N174A mutant product complex.A-B, active site of the APE1_N174A product complex crystal structure is shown in cyan. Density is shown as gray mesh. Waters are shown as blue spheres and Mn²⁺ is shown as a cyan sphere. C, active site residues of the APE1_N174A product complex crystal structure in cyan compared to the active site of APE1_WT product complex shown in gray (PDB ID: 5DFF). D, position of N212 in the APE1_WT substrate complex crystal structure in light gray (PDB ID: 5DFI) and the APE1_WT product complex in dark gray (PDB ID: 5DFF). The rotameric position of N212 in the APE1_N174A product complex resembles the APE1_WT substrate complex. E, comparison of product DNA in the APE1_N174A product complex crystal structure in cyan compared to the APE1_WT product complex in gray. Mn²⁺ is shown as a cyan sphere and waters are blue spheres for APE1_N174A; Mg²⁺ is shown as a dark gray sphere and waters are light gray spheres for APE1_WT. F, O4 of the AP site shifts 1.6 Å away from N174A in the APE1_N174A mutant compared to APE1_WT. G, N174A mutation increases the distance between Mn²⁺ ion and the 3′ hydroxyl terminus compared to APE1_WT. Mn²⁺ in the APE1_N174A complex is shown as a cyan sphere and Mg²⁺ in the APE1_WT complex is shown as a dark gray sphere. AP, apurinic/apyrimidinic; APE1, apurinic/apyrimidinic endonuclease 1.

Other than N212, the remaining active site residues in the APE1_N174A mutant do not differ from APE1_WT (Fig. 3C). In the APE1_N174A mutant product complex, we did however observe changes in the positioning of the cleaved termini of the AP DNA (Fig. 3E). In APE1_WT, N174 is within hydrogen bonding distance of the 5′ phosphate terminus at O4 and O5 (Fig. 1C). Hydrogen bonding is absent in the APE1_N174A mutant, resulting in a 1.6 Å shift of the 5′ phosphate terminus away from residue N174A compared to WT (Fig. 3F). We also observed a substantial shift of the 3′ hydroxyl terminus of the cleaved DNA in the APE1_N174A mutant (Fig. 3G). In APE1_WT, N174 acts as a steric barrier that flanks the 3′ hydroxyl terminus, keeping it in position to coordinate with the metal cofactor (Fig. 3G). In the APE1_N174A mutant, residue 174 cannot act as a steric barrier due to its reduced side chain length, resulting in a shift of over 2 Å away from the active site metal ion (Fig. 3G). In APE1_WT, the 3′ hydroxyl is 2.1 Å from its metal cofactor, but this distance increases to 4.4 Å in the APE1_N174A mutant (Fig. 3G). This finding indicates that in APE1_WT, the side chain size of N174 acts as a steric barrier that, along with hydrogen bonding, properly positions the AP site within the APE1 active site for cleavage.

Negative charge in the APE1_N174D mutant inhibits AP site cleavage

Of our N174 APE1 mutants, APE1_N174D exhibited the greatest decrease in the rate of AP site cleavage. To determine which features of APE1_N174D contribute to this decrease, we sought to investigate the active site organization of the APE1_N174D mutant, both before and after cleavage. Unfortunately, we were not able to obtain APE1_N174D substrate crystals and only obtained crystals of a product complex (described below). To provide insight into the APE1_N174D substrate complex, we performed computational mutagenesis to generate an APE1_N174D substrate model using the high-resolution X-ray crystal structure of the APE1_WT substrate complex (42). Using this mutant model, we performed molecular dynamic simulations of the APE1_N174D substrate complex to investigate the conformation of N174D. Analysis of the simulations found that the N174D rotamer fails to stabilize at a favored conformation in the APE1_N174D mutant active site (Fig. 4A). N174D is not capable of hydrogen bonding to the AP site since it lacks the hydrogen bonding donor present in N174. While we would not expect hydrogen bonding between N174D and the AP site for this reason, our simulations confirm that N174D is also never within hydrogen bonding distance of the AP site (Fig. 4B). This is also expected given the clash between the side chain oxygens of N174D and the highly electronegative DNA backbone. To compare the computationally generated APE1_N174D substrate complex to APE1_WT, we performed molecular dynamics (MD) simulations with the APE1_WT substrate complex. Analysis of the APE1_WT substrate complex simulations showed that unlike N174D, N174 of APE1_WT stabilizes at favored rotamer conformations (Fig. S6A) and was within hydrogen bonding distance of the AP site for the entire simulation (Fig. S6B). This analysis confirms that our MD simulations of the APE1_WT substrate complex is consistent with N174 hydrogen bonding to the phosphate backbone and demonstrates that the N174D mutation does not hydrogen bond to the phosphate backbone. These findings and our kinetics data demonstrate that the hydrogen bonding potential and charge of N174 in APE1_WT are necessary for catalysis.

Figure 4.

APE1_N174D substrate simulations and X-ray product complex.A, rotamer states of N174D in the APE1_N174D substrate complex determined by computational modeling. N174D rotamers fail to assume a primary rotameric state. B, distances determined between O2 of N174D and O4 and O5 of the AP site in the APE1_N174D substrate complex determined by computational modeling were plotted to determine if N174D is within hydrogen bonding distance of the AP site. Hydrogen bonding range is marked by the green box on distance plots as distances between 2.2 and 3.5 Å. C–D, active site of the APE1_N174D product complex crystal structure in magenta. Density is shown as gray mesh. Waters are shown as blue spheres. E, active site residues of the APE1_N174D product complex crystal structure in magenta compared to the active site of the APE1_WT product complex in gray (PDB ID: 5DFF). F, the position of DNA termini within the APE1_N174D product complex crystal structure compared to the APE1_WT product complex in gray. Waters are blue spheres for APE1_N174D; Mg²⁺ is shown as a dark gray sphere and waters are light gray for APE1_WT. APE1, apurinic/apyrimidinic endonuclease 1.

Next, we crystallized the APE1_N174D mutant protein bound to a 21-mer dsDNA oligo containing a centrally located THF AP site analog. In these crystals, the AP site has been cleaved by APE1 resulting in a product complex that diffracted to 2.10 Å resolution in the P1 space group (Table 2). Within this structure, both N174D and the AP site are positioned in the active site with clear density for the cleaved DNA backbone, however there is no metal bound (Fig. 4, C and D). We overlayed the APE1_N174D mutant product complex with the APE1_WT product complex and observed similar active site conformations (RMSD = 0.335 of 2321 atoms, Fig. 4E) (28). Consistent with what we observed for the APE1_N174A mutant, residue N212 in the APE1_N174D mutant failed to undergo the rotameric shift observed in APE1_WT following product formation (Fig. 4E). The 5′ phosphate and 3′ hydroxyl DNA termini in the APE1_N174D mutant product complex are only moderately shifted compared to their positions in the APE1_WT product complex (Fig. 4F). These shifts are considerably smaller than those observed in the APE1_N174A mutant product complex, supporting the idea that the N174 side chain acts as a steric barrier to keep the cleaved 5′ and 3′ DNA termini within the APE1 active site.

Overall, the structure and simulations of our APE1_N174D mutant indicate that hydrogen bonding and the side chain size of residue 174 stabilize DNA within the APE1 active site. As is the case for APE1_N174A, we partially attribute the reduced cleavage rate of the APE1_N174D mutant to the lack of hydrogen bonding interactions observed in our simulations and structure. However, given that the rate of cleavage for the APE1_N174D mutant is even lower than in the APE1_N174A mutant, we hypothesize the charge of the N174D mutation may account for the remaining difference in cleavage rates between our mutants. To understand whether the charge of N174 mutation could impair cleavage, we visualized the averaged electrostatic potential of APE1_WT and each of our mutants prior to cleavage (Fig. 5A). We found that a larger area of the APE1_N174D mutant active site surface was more negatively charged than APE1_WT. This indicates that the N174D mutation more broadly impacts the electrostatic environment of the APE1_N174D mutant active site and could impact catalysis.

Figure 5.

Effect of N174 mutation on the averaged electrostatic potential of the APE1 active site.A, the averaged electrostatic potential of APE1_WT, APE1_N174A, APE1_N174D, APE1_N174Q, and APE1_N174K substrate complexes determined by computational simulations. Intercalating residues R177 and M270 of all models have been removed for clarity of the active site cavity. The active site of the APE1_N174D substrate complex is more negatively charged than that of APE1_WT and the active site of the APE1_N174K substrate complex is slightly less negatively charged than APE1_WT. APE1, apurinic/apyrimidinic endonuclease 1.

Glutamine stabilizes AP DNA in the APE1_N174Q mutant and mimics the WT asparagine during AP site cleavage

APE1_N174Q exhibited a decreased rate of AP site cleavage, but to a lesser extent than some of our other mutants. To understand why APE1_N174Q was able to partially restore cleavage activity, we attempted to investigate the active site organization of the APE1_N174Q mutant via X-ray crystallography. Unfortunately, we were not able to obtain suitable APE1_N174Q substrate complex crystals and only obtained crystals of a product complex (described below). Therefore, to provide insight into the APE1_N174Q substrate complex, we performed computational mutagenesis to generate an APE1_N174Q substrate model using the high-resolution APE1_WT substrate complex (42). Analysis of MD simulations of this complex showed that N174Q primarily adopted rotameric conformations most similar to a single standard rotameric state, tp-100 (Χ₁ = 180°, Χ₂ = 65°, Χ₃ = −100°, Fig. 6A) (https://www2.mrc-lmb.cam.ac.uk/personal/pemsley/coot/web/docs/coot.html, 43). Additionally, distances of the amine group of N174Q were within hydrogen bonding distance of the AP site at O4 or O5 throughout the course of the simulation (Fig. 6B). This indicates the added side chain length of N174Q is accommodated in the APE1_N174Q active site and that N174Q can still hydrogen bond to the DNA backbone. Therefore, the hydrogen bonding potential and charge at residue 174 are essential for proper catalysis in APE1.

Figure 6.

APE1_N174Q substrate simulations and X-ray product complex.A, rotamer states of N174Q in the APE1_N174Q substrate complex determined by computational modeling. Rotamers cluster near the standard rotamer conformation tp-100 (Χ₁ = 180°, Χ₂ = 65°, Χ₃ = −100°). B, distances determined between N2 of N174Q and O4 and O5 of the AP site in the APE1_N174Q substrate complex determined by computational modeling were plotted to determine if N174Q is within hydrogen bonding distance of the AP site. Hydrogen bonding range is marked by the green box on distance plots as distances between 2.2 and 3.5 Å. C-D, active site of the APE1_N174Q product complex crystal structure in yellow. Density is shown as gray mesh. Waters are shown as blue spheres and Mg²⁺ is shown as a yellow sphere. E, active site residues of the APE1_N174Q product complex crystal structure in yellow compared to the active site of the APE1_WT product complex in gray (PDB ID: 5DFF). F, DNA termini in the APE1_N174Q product complex in yellow compared to APE1_WT product complex in gray. Mg²⁺ is shown as a yellow sphere and waters are blue spheres for APE1_N174Q; Mg²⁺ is shown as a dark gray sphere and waters are light gray spheres for APE1_WT. G, the longer side chain length of glutamine in the APE1_N174Q mutant than asparagine in APE1_WT displaces the 3′ hydroxyl terminus. Mg²⁺ is shown as a yellow sphere and waters are blue spheres for APE1_N174Q; Mg²⁺ is shown as a dark gray sphere and waters are light gray spheres for APE1_WT. AP, apurinic/apyrimidinic; APE1, apurinic/apyrimidinic endonuclease 1.

We crystallized the APE1_N174Q mutant bound to a 21-mer dsDNA oligo containing a centrally located AP site analog, THF. In these crystals, the backbone has been cleaved by APE1 resulting in a product complex that diffracted to 2.25 Å resolution in the P43 space group (Table 2) (31). The crystal form of the APE1_N174Q mutant product complex contains two nearly identical APE1_N174Q:DNA complexes (RMSD = 0.452 of 2329 atoms). We utilized the APE1_N174Q:DNA complex in chain A for our structural analysis since it contained better electron density for the active site residues. The AP site is positioned in the APE1_N174Q mutant active site with clear density for the cleaved DNA backbone and a bound Mg²⁺ ion (Fig. 6, C and D). We overlayed the APE1_N174Q mutant product complex with the APE1_WT product complex and observed similar active site conformations (RMSD = 0.510 of 2236 atoms, Fig. 6E) (28). The active site residues of APE1_N174Q align well with their positions in the APE1_WT active site (Fig. 6E). As in both the APE1_N174A and APE1_N174D mutant product complexes, residue N212 failed to undergo the rotameric shift observed in the APE1_WT product complex structure and remains in the substrate conformation (Fig. 6E). We did observe one additional change not present in either the APE1_N174A or APE1_N174D mutant product complexes at residue D308, which adopts a different rotameric confirmation than in APE1_WT. This rotamer shift likely arises from minor changes in the position of the cleaved 3′ hydroxyl terminus and the coordinating metal ion (Fig. 6D). In the APE1_N174Q mutant, the 3′ hydroxyl terminus shifts 2.6 Å away from the mutant active site (Fig. 6F). Comparing APE1_WT to the APE1_N174Q mutant indicates that this shift in the 3′ hydroxyl terminus arises to allow the mutant active site to accommodate the extra side chain length of N174Q and prevent a clash between N174Q and the sugar moiety of the 3′ terminus (Fig. 6G).

Overall, the structure of our APE1_N174Q product complex is consistent with simulations of our APE1_N174Q substrate complex. Both the structure and simulations demonstrate that the hydrogen bonding potential and charge of N174Q assist the APE1_N174Q mutant in cleaving AP sites. While the APE1_N174Q mutant retains better catalytic capabilities than other mutants, its activity is still less than APE1_WT. We suggest that this difference likely results from a subtle displacement of the DNA within the APE1_N174Q mutant active site to accommodate the added side chain length of N174Q as observed in our product complex. Because the decrease in APE1_N174Q cleavage rate is moderate, N174Q may still be able to partially stabilize the APE1 cleavage reaction despite any potential DNA displacement. Therefore, in APE1_WT, we expect the side chain length of N174 does contribute to cleavage, but that ultimately the hydrogen bonding potential and charge of N174 are most important for cleavage.

Similarly to N174Q, N174K can also form a hydrogen bond to the phosphate backbone of the AP site and has a longer side chain than N174. However, APE1_N174K exhibited a decreased rate of AP site cleavage that was worse than APE1_N174Q but better than APE1_N174A or APE1_N174D. To understand why APE1_N174K exhibited impaired cleavage, we attempted to investigate the active site organization of the APE1_N174K mutant via X-ray crystallography. Unfortunately, we were not able to obtain suitable crystals of either the APE1_N174K substrate or product complexes. To provide some insight into the APE1_N174K mutant active site, we performed computational mutagenesis to generate an APE1_N174K substrate complex model using the high-resolution APE1_WT substrate complex (42). These MD simulations revealed that N174K adopted two primary nonstandard p rotamer conformations (Χ₁ = ∼60°) (https://www2.mrc-lmb.cam.ac.uk/personal/pemsley/coot/web/docs/coot.html, 43). Additionally, distances of the amine group of N174K were within hydrogen bonding distance of the AP site at O4 for a portion of the simulation (Fig. 7B). Similarly to APE1_N174Q, the added side chain length of N174K appears to be accommodated in the mutant active site in our simulations. Because N174K can hydrogen bond to the AP site and the side chain length may be accommodated in the mutant active site, we believe the charge of N174K is interfering with cleavage in this mutant (Fig. 5A). APE1 therefore seems to require the neutral charge of N174 for cleavage, since the addition of both negative and positive charge within the APE1 active site were detrimental to cleavage.

Figure 7.

APE1_N174K substrate simulations.A, rotamer states of N174K in the APE1_N174K substrate complex determined by computational modeling. Rotamers cluster near two different p rotamer conformations (Χ₁ = 60°). B, distances determined between the terminal nitrogen of lysine, NZ, of N174K and either O4 or O5 of the AP site in the APE1_N174K substrate complex as determined by computational modeling were plotted to determine if N174K is within hydrogen bonding distance of the AP site. Hydrogen bonding range is marked by the green box on distance plots as distances between 2.2 and 3.5 Å. AP, apurinic/apyrimidinic; APE1, apurinic/apyrimidinic endonuclease 1.

Discussion

In this study, we demonstrate that the residue N174 is necessary for cleavage of AP sites by APE1. Our study points to three key factors of residue N174 affecting cleavage of an AP site: hydrogen bonding potential, charge, and size. Hydrogen bonding between N174 and the AP site is required to stabilize AP DNA during APE1 cleavage. When hydrogen bonding is lost, as in our APE1_N174A and APE1_N174D mutants, we observed both decreased rates of cleavage and structural differences in the position of AP DNA (Table 1). When hydrogen bonding is maintained, as in our APE1_N174Q and APE1_N174K mutants, cleavage is less affected (Table 1). Therefore, we conclude that hydrogen bonding between N174 and the AP site is necessary for efficient cleavage of AP sites by APE1 during its endonuclease reaction. Of note, hydrogen bonding between N174 and the AP site has been observed in structures of APE1_WT bound to endonuclease substrates as well as structures of APE1_WT bound to exonuclease substrates. N174 could therefore also be implicated in APE1’s exonuclease activities.

While hydrogen bonding explains how N174 stabilizes substrate DNA, we also expect the charge of N174 to contribute to the electrostatic environment of the APE1 active site enabling cleavage. Hydrogen bonding alone cannot account for differences observed in the cleavage rates between our APE1_N174A and APE1_N174D mutants since neither N174A nor N174D are capable of hydrogen bonding to the AP site. Yet our APE1_N174D mutant had a lower cleavage rate than our APE1_N174A mutant (Table 1). Visualization of the averaged electrostatic potential of our APE1 mutant active sites revealed that the APE1_N174D mutant active site was more negatively charged than any of the other APE1 mutants (Fig. 5A). From this information, we hypothesize that additional negative charge within the APE1_N174D active site inhibits formation of the negatively charged phosphorane intermediate of the APE1 endonuclease reaction (Fig. 8, A and B). Interestingly, addition of positive charge as in our APE1_N174K mutant, resulted in better cleavage than our APE1_N174D but poorer cleavage than either our APE1_N174Q mutant or APE1_WT. Therefore, we suggest a model in which residue 174 in APE1 must contribute a favorable combination of neutral charge and hydrogen bonding potential to promote formation of the phosphorane intermediate and allow the endonuclease reaction to proceed (Fig. 8B). We conclude that like hydrogen bonding, the charge of N174 is necessary for efficient cleavage of AP sites by APE1 and may be important in APE1’s exonuclease function.

Figure 8.

The effect of N174 and N174 mutation on the APE1 endonuclease reaction.A, a simplified scheme of the APE1 endonuclease reaction showing coordination of cleavage by D210 and N212 and stabilization of the reaction by N174. Hydrogen bonds are shown as blue dashed lines and forming/breaking covalent bonds are shown as black dashed lines. See the following references for a more complete scheme (28, 29). B, effect of N174 mutation on the phosphorane intermediate of the APE1 endonuclease reaction. APE1, apurinic/apyrimidinic endonuclease 1.

Hydrogen bonding and charge are not the only evidence that N174 is essential to APE1 endonuclease activity. Analysis of our APE1 mutants also indicates that the size of N174 in APE_WT contributes to APE1 cleavage of AP sites. The smaller size of N174A led to considerable shifts in the position of the AP DNA within the APE1_N174A mutant active site, potentially impeding catalysis (Figure 2, Figure 3D and 3G). This suggests N174 in APE1_WT may behave as a steric barrier keeping substrate DNA within the APE1 active site during cleavage. However, the size of residue 174 may be less important than the charge and hydrogen bonding potential since our APE1_N174D mutant exhibited the lowest cleavage rate despite maintaining side chain size (Table 1). Similarly, despite having larger side chain lengths, the APE1_N174Q and APE1_N174K mutants exhibited less impaired catalysis than the other mutants (Table 1) and added side chain size was accommodated within these mutant active sites (Figs. 6 and 7). Cumulatively, this suggests that the size of N174 in APE1_WT is likely not as important for catalysis as hydrogen bonding or charge but does point to a moderate role of the side chain size in properly organizing the active site during catalysis. It also raises an interesting hypothesis that N174 side chain size may play a role in APE1’s exonuclease activity.

As mentioned previously, N174 is highly conserved among the primary AP endonucleases from humans to zebrafish (Fig. S1). However, sequence alignment of human APE1 with the primary AP endonuclease in E. coli, EXOIII (37, 38, 39, 40), reveals alignment of N174 in APE1 with Q112 in EXOIII (Fig. S1). While EXOIII is the primary AP endonuclease in E. coli, EXOIII is considered to have a more robust exonuclease than endonuclease activity (39, 40). Inversely, APE1 is considered to have a more robust endonuclease activity when compared to its exonuclease activity (39, 44). The active sites of APE1 and EXOIII are conserved with the exception of N174/Q112, implicating the difference in the size of N174 and Q112 in the variability of endonuclease versus exonuclease activities across these enzymes. So far, differences in the exonuclease function between APE1 and EXOIII have been linked to bulky aromatic residues expected to stabilize the 3′ terminal base to be excised during exonuclease activity (45). But interestingly, mutation of Q112 does appear to affect EXOIII exonuclease activity on some substrates (46) suggesting that N174 could also affect the exonuclease activity of APE1 along with the suspected aromatic residues. We hope to investigate this possibility in future studies.

Finally, we note that use of THF AP DNA in our experiments and simulations does not perfectly mimic a true AP site. True AP sites contain an additional hydroxyl group at the location of the glycosidic bond of DNA nucleotides (47, 48, 49). Because our structural and MD simulations lack this hydroxyl group, the exact active site organization and electrostatic surface potential may be slightly different than those represented by our results. All structures of APE1 bound to AP DNA in the Protein Data Bank (PDB) and Electron Microscopy Data Bank up to this point in time have used a THF AP site analog, and further analysis on structural differences of APE1 bound to true AP sites versus THF analogs requires additional investigation. Unlike structural comparison, pre-steady-state kinetic characterization of APE1 excision of true AP sites and THF AP site analogs has been previously described by Schermerhorn and Delaney (47). In the presence of Mg²⁺, the authors found a very limited difference in rates of chemistry between excised substrates (47). When other metals were present, APE1 exhibited slightly lower rates of excision of THF substrates than true AP sites (47). Therefore, in our experiments, we expect the overall trends we observed for THF cleavage by our mutants is comparable with a true AP site.

In summary, analysis of N174 in APE1 indicates that AP site cleavage by APE1 is dependent upon the hydrogen bonding, charge, and size of N174. We found that N174 stabilizes AP DNA during endonuclease cleavage via hydrogen bonding and side chain size and that it promotes the endonuclease reaction via charge. Furthermore, we speculate that the size of N174 may be of unexplored interest concerning the differences in APE1’s endonuclease and exonuclease functions when compared to other AP endonuclease homologs, in particular, EXOIII. These findings, taken together, provide novel characterization of an essential residue of the APE1 active site, and further our understanding of how APE1 coordinates its active site and DNA substrates for cleavage.

Experimental procedures

DNA substrates

All APE1 DNA substrates were prepared by annealing oligonucleotides purchased and purified by Integrated DNA Technologies. DNA substrates for kinetic analysis and EMSAs were prepared by annealing a 5′ fluorescein amidite (FAM)-labeled oligonucleotide template containing a centrally located THF-type AP site with an unlabeled oligonucleotide complement (42). Oligonucleotides were annealed as follows: FAM-labeled and unlabeled oligonucleotides were mixed in water to a final concentration of 10 μM and 12 μM, respectively, heated to 95 °C, and cooled at a rate of 1 °C per minute until reaching 4 °C. Oligonucleotides for X-ray crystallography structure determination were annealed as follows: oligonucleotides were mixed in buffer containing 100 mM Tris, pH 7.5, and 20 mM MgCl₂ to a final concentration of 2 mM, heated to 95 °C, and cooled at a rate of 1 °C per minute until reaching 4 °C. The 30-bp oligonucleotide sequences used for kinetic analysis and EMSAs were 5′∗-CGT-TCG-CTG-ATG-CGC-XCG-ACG-GAT-CCG-CAT-3′ and 5′-ATG-CGG-ATC-CGT-CGA-GCG-CAT-CAG-CGA-ACG-3′ (where ∗ denotes FAM label and X denotes the THF AP site) (42). The 21-bp oligonucleotide sequences used for X-ray crystallography structure determination for product structures were 5′-GGA-TCC-GTC-GGG-CGC-ATC-AGC-3′ and 5′-GCT-GAT-GCG-CXC-GAC-GGA-TCC-3′ (42).

Protein expression and purification

Human full-length WT and N174 mutant APE1 was expressed and purified as previously described (50). Briefly, APE1 was expressed in NEB LysY E. coli using an optimized PET28a vector. Cells were lysed via sonication and lysate was filtered with a 0.45 μm filter then purified using an ATKA-Pure FPLC with the following columns: Cytiva 5 ml Heparin HP affinity column, Cytiva 5 ml Resource S cation exchange column, and Cytiva HiPrep 16/60 Sephacryl S-200 HR gel filtration column. Proteins were concentrated to stock concentrations of approximately 5 mg/ml for kinetics and EMSA experiments and 20 mg/ml for X-ray crystallography experiments. Final protein concentration was determined using absorbance at 280 nm. Full-length WT and N174 mutant APE1 proteins were utilized for kinetic experiments and EMSAs. Truncated (Δ1-42)/C138A variants, which have been used previously and shown not to alter APE1 function, were utilized for crystallography studies (27, 28, 31).

Single-turnover kinetic analysis

Cleaved DNA formed by full-length WT and N174 mutant APE1 formed over time was measured under single-turnover conditions. Product formation was measured by visualizing FAM-labeled substrate on a 21% denaturing National Diagnostics Ultra-Pure SequaGEL UreaGel gel. Single-turnover conditions were utilized to approximate the cleavage rate constant, k_obs. Cleavage reactions were initiated by mixing substrate DNA with APE1 and allowing the reaction to proceed until quenching at predetermined timepoints. Single-turnover reactions occurred under the following conditions: 50 mM Hepes, pH 7.5, 100 mM KCl, 5 mM MgCl₂, 0.1 mg/ml bovine serum albumin, 50 nM DNA, and 500 nM APE1. Reactions were quenched with 300 mM EDTA, pH 8.0 on a Kintec Rapid Quench Flow, or with formamide loading dye containing 78% w/v formamide, 100 mM EDTA, 0.5 mg/ml bromophenol blue, and 0.25 mg/ml xylene cyanol. Quenched reactions were incubated with formamide loading dye at 95 °C for 5 min then run on a 21% denaturing National Diagnostics Ultra-Pure SequaGEL UreaGel gel to resolve the fraction of cleaved DNA product from substrate and visualized using a Typhoon imager. Product and substrate bands in the gel were quantified by densitometry with ImageQuantTL (GE, v 8.1) or ImageJ (https://imagej.net/ij/) to calculate the amount of product formation over time (51). The resulting curve was fit to Equation 1 for single-turnover reactions using Prism v9.1 (https://www.graphpad.com/features) (41).

$$y=A\left(1-e^{-k_{obs}t}\right)$$ (Eq 1)

In Equation 1, A corresponds to the active fraction of enzyme and is equal to the amplitude of product formation. There were between three and four replicate single-turnover experiments conducted per APE1 variant. All replicates were fit individually to determine individual k_obs values. The final k_obs value as reported in Table 1 corresponds to the mean k_obs from all replicates ± SD. Fig S2 shows the fits of cleaved product formed over time by each variant when all replicates are included.

As reported previously, the k_obs of APE1_WT under single-turnover conditions exceeds quantification using rapid quench flow instrumentation (41). Using this instrumentation, quenching APE1_WT cleavage of AP DNA under single-turnover conditions after 0.0026 s resulted in approximately greater than 45% product. Therefore, we chose to present the k_obs of APE1_WT under single-turnover conditions as the previously determined minimum estimated rate constant necessary to exceed quenching under these conditions of at least 850 s⁻¹ (41).

Electrophoretic mobility shift assay

WT and mutant APE1 affinity for a 30 bp DNA substrate containing an AP site was determined by mixing APE1 with 2 nM DNA and allowing samples to incubate for at least 20 min to approach equilibrium. Concentrations of APE1 varied between 0 and 100 nM for APE1_WT and the APE1_N174Q mutant, between 0 and 300 nM for the APE1_N174A and APE1_N174D mutants, and between 0 and 1000 nM for the APE1_N174K mutant. Experiments were conducted under the following conditions: 50 mM Tris, pH 8.0, 1 mM EDTA, 0.2 mg/ml bovine serum albumin, 50 mg/ml sucrose, 0.5 mg/ml bromophenol blue, and 1 mM DTT. Samples were run on a 10% 59:1 polyacrylamide native gel to determine APE1 bound and unbound fractions of DNA. FAM-labeled DNA was visualized using a Typhoon Imager. Bound and unbound fractions were quantified by densitometry with ImageQuant TL (https://www.cytivalifesciences.com/en/us/shop/protein-analysis/molecular-imaging-for-proteins/imaging-software/imagequant-tl-analysis-software-p-28619?psmenu=2; GE, v 8.1). The resulting curve of bound DNA determined by ligand depletion versus APE1 concentration was fit to Equation 2 using Prism v9.1 to determine the apparent binding affinity, K_{D App} (41).

$$y=\frac{\left(K_{DApp}+E+D\right)-\sqrt{{\left(K_{DApp}+E+D\right)}^2-4ED}}{2}$$ (Eq 2)

In Equation 2, E corresponds to the concentration of APE1 and D corresponds to the concentration of DNA. There were between three and five replicate EMSA experiments conducted per APE1 variant. All replicates were fit individually to determine individual K_{D App} values. The final K_{D App} as reported in Table 1 corresponds to the mean K_{D App} from all replicates ± SD. Fig. S4 shows the fits of DNA bound by each variant when all replicates are included.

X-ray crystallography structure determination

APE1 structures were crystallized utilizing constructs with a 42 amino acid N-terminal truncation, a C138A mutation, and N174A, D, or Q mutations (31). To crystallize the APE1_N174A substrate complex, 0.56 mM DNA containing an AP site was mixed with 16 mg/ml APE1_N174A. Crystals were generated via sitting drop vapor diffusion using 2 μl of the above protein/DNA mix combined with 2 μl reservoir solution (100 mM sodium citrate, pH 5.0, 200 mM MgCl₂, 16%–21% PEG 20,000). Resultant crystals were transferred to a cryoprotectant solution containing 75% reservoir solution supplemented with 20% ethylene glycol and 5% MgCl₂, flash-frozen, and subjected to X-ray diffraction. To crystalize the APE1_N174A product complex, the protein/DNA mix was first incubated with 50 mM MnCl₂ for 100 min and then crystalized as above. Resultant crystals were transferred to a cryoprotectant solution containing 75% reservoir solution supplemented with 20% ethylene glycol and 5% MnCl₂, flash-frozen, and subjected to X-ray diffraction. To crystallize the APE1_N174D product complex, 0.56 mM DNA containing an AP site was mixed with 10 mg/ml APE1_N174D. Crystals were generated via sitting drop vapor diffusion using 2 μl of the above protein/DNA mix combined with 2 μl reservoir solution (200 mM lithium sulfate and 15%–25% PEG 3350). Resultant crystals were transferred to a cryoprotectant solution containing 75% reservoir solution supplemented with 20% ethylene glycol and 5% MgCl₂, flash-frozen, and subjected to X-ray diffraction. To crystallize the APE1_N174Q product complex, 0.56 mM DNA containing an AP site was mixed with 10 mg/ml APE1_N174Q. Crystals were generated via sitting drop vapor diffusion using 2 μl of the above protein/DNA mix combined with 2 μl reservoir solution (100 mM Hepes free acid, 200 mM ammonium acetate, and 25% PEG 3350). Resultant crystals were transferred to a cryoprotectant solution containing 75% reservoir solution supplemented with 20% ethylene glycol and 5% MgCl₂, flash-frozen, and subjected to X-ray diffraction. APE1_N174D and APE1_N174Q with a phosphorothioate-containing AP substrate resisted crystallization. The APE1_N174A substrate structure was collected at the Lawrence Berkley National Laboratory Advanced Light Source using Beamline 4.2.2. The APE1_N174A, APE1_N174D, and APE1_N174Q product complex structures were collected at 100 K on a Rigaku MicroMax-007 HF rotating anode diffractometer system at a wavelength of 1.54 Å utilizing a Dectris Pilatus3R 200K-A detector. The software HKL-3000 (https://www.hkl-xray.com/hkl-3000; v 705c, HKL Research Inc.) was used to process and scale diffraction data following collection from all X-ray diffractions sources. Initial models were determined using molecular replacement in PHENIX (https://phenix-online.org/; v 1.19.2-4158-000) using a previously determined APE1:DNA product complex structure (PDB: 5DFF). Refinement and model building were done with PHENIX and Coot (https://www2.mrc-lmb.cam.ac.uk/personal/pemsley/coot/; v 0.9), respectively, and figures were made using PyMOL (https://www.pymol.org/; v 2.5.1, Schrödinger LLC) (52, 53).

Computational mutagenesis and simulations

Computational mutagenesis was performed using leap modular of AMBER v16 suite of programs for biomolecular simulations (54). The side chain was deleted in the input file provided to the leap module, which generated the side chain conformation for mutant residue based on internal template. MD simulations were performed for APE1:AP DNA complexes in explicit water solvent. Computational modeling was performed to determine rotamer states of residue 174 and resulting distance to the AP site in the APE1_WT, APE1_N174D, APE1_N174Q, and APE1_N174K substrate complexes, using an approach similar to those described in previous studies (55, 56). Modeling of the APE1_N174A complex was not performed since alanine is unable to assume more than a single rotamer state. Model preparation and simulations were performed using the AMBER v16 suite of programs for biomolecular simulations (54). AMBER’s ff14SB (57) force fields were used for all simulations. MD simulations were performed using NVIDIA graphical processing units and AMBER's pmemd.cuda simulation engine using the Agarwal lab’s previously published protocols (58, 59).

A total of 4 separate simulations were performed for APE1_WT, APE1_N174D, APE1_N174Q, and APE1_N174K based on the X-ray crystal structure of APE1_WT in complex with substrate DNA (PDB ID: 5DGI). The DNA sequence and APE1 constructs used in this study were similar to the study that produced the referenced structure (28). Any missing hydrogen atoms were added by AMBER’s tleap program. The N174D, N174Q, and N174K mutations were created by AMBER’s tleap program. After processing the coordinates of the protein and substrate, all systems were neutralized by addition of counter ions and the resulting system was solvated in a rectangular box of SPC/E water with a 10 Å minimum distance between the protein and the edge of the periodic box. The prepared systems were equilibrated using a protocol described previously (60). The equilibrated systems were then used to run 100 ns of production MD under constant energy conditions (NVE ensemble). The use of NVE ensemble is preferred as it offers better computational stability and performance (61). The production simulations were performed at a temperature of 300 K. As NVE ensemble was used for production runs, these values correspond to the initial temperature at the start of simulations. A temperature adjusting thermostat was not used in simulations. Over the course of the 100 ns simulation, the temperature fluctuated ± 5 K around the 300 K target, which is typical for well equilibrated systems.

As the goal of the computer simulations was to investigate the rotameric states of the protein residues, the DNA structure was restrained using positional restraints (1 kcal/mol/Ų on all the DNA atoms). This allowed minimal movement in the DNA structure, while allowing protein residues to explore different possible conformations. A total of 1000 conformational snapshots (stored every 100 picoseconds) collected for each system was used for analysis. The conformational snapshots were used for the rotameric side chain evaluations and electrostatic potential calculations.

Electrostatic potential calculations

For the 1000 structures extracted above the electrostatic surfaces were computed using the Adaptive Poisson-Boltzmann Solver software (62). The surfaces for the 1000 structures were averaged to correspond to the representative electrostatic surface for the entire MD simulation trajectory.

Multiple sequence alignment

Full amino acid sequences for human APE1 (UniProt ID: P27695), chimpanzee APE1 (UniProt ID: A2T6Y4), mouse APE1 (UniProt ID: P28352), zebrafish APE1 (UniProt ID: A0MTA1), and E. coli exonuclease III (UniProt ID:P09030) were obtained from UniProtKB/Swiss-Prot and aligned using EMBL-EBI ClustalOmega (CLUSTAL O(1.2.4)) (63, 64). Alignment output was formatted as ClustalW with character counts using default options for this output format. The alignment generated by ClustalOmega was visualized for publication using JalView (v 2.11.4.1) (65) and colored by consensus sequence where sequence identity was greater than or equal to 60% across the five species.

Data availability

Additional information on X-ray crystal structure determination and structure coordinates is available from the Protein Data Bank (PDB codes: 9DP1, 9DP2, 9DP3, and 9DP4).

Supporting information

This article contains supporting information.

Conflict of interest

The authors declare that they have no conflicts of interest with the contents of this article.

Reviewed by members of the JBC Editorial Board. Edited by Patrick J. O'Brien

Supporting information

Figure S1.

Figure S2.

Figure S3.

Figure S4.

Figure S5.

Figure S6.