Development and Characterization of a Novel NEIL1 Nanobody

2.1. Plasmids, Nanobody Expression, and Purification

Hybrigenics Services SAS (Paris, France) delivered the sequences of the top 10 VHHs in a pEBY vector containing a C-terminal 10x histidine (His) and Flag tags for E. coli expression. We further cloned the sequences into a pFUSE-mFc(IL2ss)NN-adapt vector fused to a mouse Fc fragment for minibody generation in mammalian cells. For purification, the C-terminal Flag and 10x His tag of the pEBY vector was reduced to 6x His and a TEV cleavable 6x His tag via standard cloning techniques and digested with NcoI and NotI (NEB) restriction enzymes (Fig. 1A). A5 expression and purification was performed as previously described (Fig. 1B)²⁰.

Briefly, A5–6His cell pellets were resuspended in a buffer containing 50 mM HEPES pH 7.4, 500 mM NaCl, 10% (v/v) glycerol, 1 mM ethylenediaminetetraacetic acid (EDTA), 1 mM dithiothreitol (DTT), 1 mM phenylmethylsulfonyl fluoride (PMSF), and cOmplete (EDTA-free protease inhibitor cocktail; Sigma), sonicated, and centrifuged at 25,000 × g. Clarified cell lysates were incubated with pre-equilibrated TALON Superflow beads (Cytiva) for 2 hours at 4 °C. Beads were washed with 10 column volumes of wash buffer (50 mM HEPES pH 7.4, 200 mM NaCl, 10% (v/v) glycerol, and 1 mM DTT), prior to elution with increasing elution buffer (50 mM HEPES pH 7.4, 200 mM NaCl, 10% (v/v) glycerol, 250 mM imidazole, and 1 mM DTT). The protein elutes in with 50% elution buffer and is subsequently concentrated and subjected to size exclusion chromatography (SEC) using a Superdex 200 column (GE Healthcare) in a buffer containing 50 mM HEPES pH 8.2, 200 mM NaCl, 10% (v/v) glycerol, and 1 mM DTT. Fractions were pooled, concentrated, and flash-frozen aliquots were stored at −80 °C.

NEIL1 was expressed and purified as previously described¹⁹, except for the size exclusion step that was performed in a buffer containing 50 mM HEPES pH 8.2, 200 mM NaCl, 10% (v/v) glycerol, and 1 mM DTT.

2.2. VHH-fc Production and Isolation

10⁶ HEK 293 cells were seeded into a 10 cm dish and incubated until 70% confluent. 6 μg A5 pFuse DNA was diluted with 1.5 mL Opti-MEM. 6 μL lipofectamine was incubated with 1.5 mL Opti-MEM. Both solutions were incubated for 10 minutes, then mixed and incubated for 20 min. The solution was then added to the cells and incubated for 96 hours. Media was collected, flash frozen, and stored at −20°C.

2.3. Glycosylase Activity Assay

The TAMRA-labeled, thymine glycol (Tg)-containing, double stranded, 22-mer oligodeoxynucleotide (Fig. 2A) was generated by mixing 100 μM of each strand in 20 mM Tris-HCl pH 8.0 with 100 mM KCl and was heated to 95°C for 5 min and allowed to slowly cool to room temperature.

The activity assay was performed using 50 nM annealed oligo, 50 nM NEIL1, and varying concentrations of A5 (0, 50, and 500 nM) in a reaction buffer containing 20 mM Tris-HCl pH 8.0, 100 mM KCl, and 0.1% BSA. Reactions were carried out at room temperature until the indicated time point (0, 50, 100, 400, 1600 seconds) and quenched using 1 volume of 98% formamide with 5 mM EDTA. The fluorescence was subsequently analyzed on a TECAN infinite M1000Pro (excitation filter 525 nm, and emission filter 598 nm). The activity assay was also performed for 30 min using 50 nM of DNA substrate and increasing amounts of A5 (0, 25, or 500 nM) to vary the substrate:NEIL1 ratio (1:0, 1:0.5, 1:1, 1:2, 1:4, 1:8, 1:16, and 1:32). As a control, the assay was performed using identical ratios of substrate to BSA.

2.4. Differential Scanning Fluorimetry

2.5. Microscale Thermophoresis

A Monolith X instrument (NanoTemper Technologies, LLC, South San Francisco, CA) was used to perform the Microscale thermophoresis (MST) experiment. His-Tag containing NEIL1 and cleaved A5 were purified in a buffer containing 50 mM HEPES pH 8.2, 200 mM NaCl, and 5% (v/v) glycerol. This buffer was also used as the assay buffer. NEIL1 was labeled using the His-Tag Labeling Kit RED-tris-NTA 2nd Generation (NanoTemper Technologies) by incubating 100 nM NEIL1 with 50 nM RED-tris-NTA dye (1:1 volume ratio) for 30 min. The affinity of the His-Tag to the dye was 1.5 nM. Under the labeling condition, 100% of the dye would be bound to the protein. Labeled NEIL1 was incubated with serial dilutions of cleaved A5 ranging from 100 μM to 3.05 nM (100 μM, 50 μM, 25 μM, 12.5 μM, 6.3 μM, 3.13 μM, 1.6 μM, 781.3 nM, 390.6 nM, 195.2 nM, 97.7 nM, 48.8 nM, 24.4 nM, 12.2 nM, 6.1 nM, 3.1 nM). The MST measurements were performed in premium capillaries with 100% LED/excitation power. All data analysis was preformed using the Monolith proprietary software.

2.6. Dot Blot Analysis

50 μg purified NEIL1, truncated NEIL1 proteins (Δ40, Δ56, Δ100, GST-290–349, GST-313–349, and GST-313–390), and BSA control (VWR cat# 97061–420) were placed on the nitrocellulose membrane (Bio-Rad) and allowed to dry. The blot was blocked with 3% milk TBST for 1 hour at room temperature prior to incubation with A5-Fc primary antibody (diluted 1:300 in 3% milk TBST) at 4°C overnight. The membrane was then washed with TBST three times for 10 minutes each followed by incubation for 1 hour at room temperature with anti-mouse secondary antibody (Cell Signaling Technologies cat# 7076S) diluted 1:1000 in 3% milk TBST. The membrane was then washed with TBST three times for 10 minutes each prior to imaging with Advansta Substrates WesternBright ECL HRP Substrate Kit. Colorimetric and chemilumnescent images were taken on a Bio-Rad Gel Dox XR+ for further analysis.

2.7. Far Western Analysis

100 pmol of NEIL1-FL, Δ40, Δ56, Δ56Δaf (af comprises amino acids 202–222), Δ78, Δ100, GST-290–390, GST-290–349, GST-313–349, GST, and BSA, as well as 0.2 pmol of A5–10His-Flag were run on a 4–20% polyacrylamide gel (Bio-Rad) and transferred onto a nitrocellulose membrane (Bio-Rad) via a Bio-Rad TurboBlot system. The protein on the membrane was washed in PBS with 1 mM DTT twice for 10 minutes at RT before being refolded with a serial dilution of guanidine hydrochloride (6 M, 3 M, 1.5 M, 0.75 M, 0.375 M, 0.1871 M, 0.09 M, 0.0 M) in PBS with 1 mM DTT. The initial seven incubations were done at RT for 30 minutes and the final step was performed overnight at 4°C.

The membrane was washed three times for 10 minutes with PBS at 4°C and blocked in cold 5% milk PBS for 1 hour at RT. After blocking, the membrane was washed three times with NEIL storage buffer (50 mM HEPES pH 8.2, 200 mM NaCl, 10% (v/v) glycerol, and 1 mM DTT) for 10 minutes each at 4°C prior to incubation with 10 nmol of purified A5–10His-Flag (1 μM) diluted in NEIL storage buffer at 4°C overnight. The membrane was washed three times with cold NEIL storage buffer for 10 minutes each at 4°C before incubating at 4°C overnight with anti-Flag antibody (Abcam; ab205606) diluted 1:5000 with 5% milk in NEIL storage buffer. The membrane was washed three times in cold NEIL storage buffer for 10 minutes at 4°C before incubation with an anti-rabbit secondary (Cytiva, NA934–1ML) diluted 1:5000 with 5% milk in cold NEIL storage buffer for 1 hour at 4°C. Prior to development with the Advansta Substrates WesternBright ECL HRP Substrate Kit, the membrane was washed three times with cold NEIL storage buffer at 4°C for 10 minutes each. Blot images were taken on Bio-Rad Gel Dox XR+.

2.8. Localization and Recruitment via Confocal Microscopy

7.4×10⁴ cells were seeded into 35 mm fluorodishes (WPI). Twenty-four hours after seeding cells were chemically transfected using 250 ng of no MTS NEIL1-GFP and A5-mCherry. A 1:3 ratio of DNA to transfection reagent (XtremeGeneHD) was used and cells were incubated for 48 hours at 37°C and 5% CO₂¹⁸. 40 min prior to imaging, NucBlue Live Cell Stain (Hoechst 33342) was added to the cells at a concentration of 1 drop per mL media and incubated at 37°C for 20 minutes, followed by an additional 20 min incubation with fresh media. All localization and laser micro-irradiation was performed using a Nikon A1r confocal microscope and the MIDAS (Modular Irradiation, Detection, and Analysis System) acquisition program³⁸.

2.9. NEIL1-A5 Complex Formation

The NEIL1-A5 complex was prepared by mixing both proteins or the protein and DNA in a 1:1 molar ratio in a NEIL storage buffer. The complex was incubated on ice for 1 hour prior to SEC analysis. The SEC was performed with a Superdex-200 10/300 Increase column using the NEIL storage buffer. Appropriate fractions were aliquoted, flash cooled in liquid nitrogen, and stored at −80°C. The column was calibrated with blue dextran (to determine the void volume) and three standards of known molecular weights (MW) using Gel Filtration LMW Calibration Kit (Cytiva # 28403841).

2.10. AlphaFold 3 NEIL1-Nanobody Modeling

Structural predictions for the NEIL1-A5-DNA complex were conducted using the AlphaFold 3 Server³⁹. Available structural information for the A5 nanobody and NEIL1-DNA complex were sourced from the PDB database (8FTT and 5ITR, respectively)^15,20. The CIF format results produced by AlphaFold 3 were subsequently imported into PyMOL for visualization. Interface residue mapping in PyMOL identified residues at the interaction interface capable of forming hydrogen and ionic bonds within four angstroms, providing insights into the molecular interactions stabilizing A5-NEIL1 binding.

2.11. Statistical analysis

Average values and standard error of the mean (SEM) were calculated from the means (of technical replicates) of multiple independent experiments (n = number of independent experiments as indicated in figure legends), unless stated otherwise. Student’s t-test was used to test for significant differences as appropriate, with results generally compared to controls and as indicated in the figure legends. P-values are indicated by asterisks (*p<0.05, **p<0.01, ***p<0.001, ****p<0.0001) or are stated in the figure legend. Statistical analyses were performed using GraphPad Prism except those determined using MIDAS³⁸.

3.1. Development of novel nanobodies against the NEIL1 DNA glycosylase

Nanobody libraries have emerged as powerful tools for identifying antigen-specific binders with high affinity and specificity⁴⁰. To generate nanobodies targeting NEIL1, we utilized the Nali-H1 yeast library, a fully synthetic single-domain antibody library with a complexity of 2 × 10⁷ and performed yeast two-hybrid (Y2H) screening in collaboration with Hybrigenics Services SAS (Paris, France). The initial LexA screen, which analyzed 98 million clones, identified three positive colonies, all corresponding to the same nanobody. A subsequent Gal4 screen of 69 million clones yielded 33 positive colonies, representing 17 distinct nanobodies.

From these, Hybrigenics selected the ten most promising candidates and delivered them to us in the pB27 yeast expression vector. Notably, the nanobody identified most frequently in both the LexA and Gal4 screens, A5, was further subcloned into three additional expression vectors: pEBY (containing a 10×His-tag and Flag-tag for bacterial purification), pFuse (for minibody production), and pmCherry (for mammalian expression) (Fig. 1A). To optimize A5 for biochemical interaction studies, we refined the pEBY construct by reducing the 10×His-tag and Flag tag to either a cleavable or non-cleavable 6× His tag. Expression and purification of A5–10xHis-Flag, A5–6xHis, and A5-tev-6xHis in bacterial systems all yielded proteins of >95% purity (Fig. 1B), demonstrating efficient production and suitability for downstream characterization.

3.2. NEIL1 Glycosylase activity is preserved in the presence of the NEIL1 nanobody

To assess whether A5 affects NEIL1’s enzymatic function, we performed a glycosylase assay (Fig. 2A) using a 1:1 molar ratio of NEIL1 to substrate (50 nM each) in the presence of increasing concentrations of A5 (0 nM, 50 nM, and 500 nM). Fluorescence measurements indicate an initial increase from 0 to 100 seconds, followed by a plateau between 100 and 1600 seconds (Fig. 2B). These data suggest that even in excess amounts, A5 does not impair NEIL1’s glycosylase activity. Additionally, when NEIL1 is present in excess (ratios of DNA:NEIL1 varied between 1:0.5 to 1:32), no significant change in relative fluorescence is observed in the presence or absence of A5 (Fig. 2C). These data confirm that A5 does not interfere with the enzymatic function of NEIL1.

3.3. Physicochemical characterization of selected NEIL1 nanobodies

To assess the binding affinity and stability of the NEIL1-A5 interaction, we employed a combination of biochemical techniques, including differential scanning fluorimetry (DSF), nanoDSF, microscale thermophoresis (MST), and isothermal titration calorimetry (ITC). Using traditional DSF, we determined the individual melting temperatures (T_m) of NEIL1 and A5–10His-Flag at 41°C and 63°C, respectively (Fig. 3A). When we analyzed the NEIL1-A5–10His-Flag complex, we observed two distinct unfolding transitions at 41°C and 57°C, suggesting that A5 binding partially stabilizes NEIL1 while maintaining its native thermal properties. The shifted folding transition remained present in various buffering conditions (Supplementary Fig 1A-B). To further confirm these findings, we performed nanoDSF, which showed that NEIL1 alone displayed a single unfolding transition around 40–41°C (Fig. 3B, Supplementary Fig. 2A), consistent with DSF results. We then examined A5-tev-6His at different concentrations and found that at 9 μM, it unfolded at 60.0°C, while at higher concentrations, an unfolding transition was detected at 57.8°C with an elongated trailing peak (Fig. 3B, Supplementary Fig. 2B), possibly indicating an altered conformational state. When we analyzed the NEIL1-A5-tev-6His complex, we observed two unfolding transitions at 38.5°C and 54.1°C, reinforcing the idea that A5 interacts with NEIL1 and affects its thermal stability (Fig. 3B).

Next, we used MST analysis, which yielded a dissociation constant (K_d) of 23.8 μM. Given that some nanobodies can bind to their target proteins in the low micromolar to nanomolar range this result would indicate a moderate binding affinity between NEIL1 and A5 (Fig. 3C)^41,42. To further validate this interaction, we conducted ITC measurements, which produced a similar K_d of approximately 30 μM (Supplementary Fig. 3), confirming that A5 binds NEIL1 with a moderate binding affinity. Together, these findings establish A5 as a binder of NEIL1, providing a foundation for its use in cellular studies.

3.4. Specificity of the NEIL1 Nanobody in live cells

To assess the specificity of A5 for NEIL1 in live cells, we examined its subcellular localization and recruitment dynamics using fluorescence imaging. When we co-transfected A5-mCherry with NEIL1-GFP, we observed clear nuclear co-localization, with a Pearson’s correlation coefficient (PCC) of 0.924 (Fig. 4A, Supplemental Figure 4A). In contrast, when A5-mCherry was co-expressed with NEIL2-GFP, A5-mCherry displayed an even distribution between the nucleus and cytoplasm, yielding a lower PCC of 0.532 (Fig. 4B, Supplemental Figure 4A). When co-expressed with NEIL2-GFP, A5 shows a similar cellular distribution as when expressed alone (Supplementary Figure 4B). This supports A5’s specificity for NEIL1 over NEIL2.

We next tested the extent to which A5 tracks NEIL1’s recruitment to sites of DNA damage. Using live-cell imaging following laser-induced DNA damage, we found that both NEIL1-GFP and A5-mCherry were rapidly recruited to damage sites, reaching peak fluorescence at 9.90 and 7.15 seconds, respectively (Fig. 4C–E). NEIL1-GFP fluorescence increased 1.57-fold, while A5-mCherry fluorescence increased 1.38-fold at their respective peaks (Fig. 4F). While A5-mCherry displayed slightly reduced recruitment relative to NEIL1, their recruitment dynamics remained strongly correlated, with a correlation plot slope of 1.23 (Fig. 4G). The slight delay in A5-mCherry recruitment may potentially reflect subtle conformational changes in NEIL1 during its engagement with DNA lesions.

To further evaluate A5’s specificity for NEIL1, we examined its recruitment dynamics in the presence of NEIL2. When we performed the experiment with NEIL2-GFP, we observed rapid recruitment of NEIL2-GFP to DNA damage sites (Fig. 4H–I), with an average peak time of 5.60 seconds and a 1.31-fold increase in fluorescence intensity (Fig. 4J–K). However, A5-mCherry was not recruited to the sites of damage when co-expressed with NEIL2-GFP (Fig. 4H–I). The distribution of peak recruitment times for A5-mCherry in Figure 4J indicates the absence of a distinct peak, as the data points do not cluster near the mean, unlike the recruitment pattern observed for NEIL1-A5 (Fig. 4E). Furthermore, the peak fold-change distribution (Fig. 4K) for A5-mCherry is centered around 1, suggesting no significant change in local fluorescence intensity. The absence of nuclear enrichment of A5-mCherry in the presence of NEIL2-GFP, coupled with the lack of co-recruitment to DNA damage sites, as further illustrated by the correlation plot (Fig. 4L), indicates that A5 does not interact significantly with NEIL2 in cells. These findings reinforce A5’s selective specificity for NEIL1, underscoring its utility as a probe for tracking NEIL1 dynamics in live-cell studies.

3.5. Epitope mapping

To determine the specific region of NEIL1 recognized by A5, we performed dot blot and far western analyses using NEIL1 constructs either lacking portions of the C-terminal domain (FL, Δ40, Δ56, Δ56Δaf, and Δ100; where 40, 56, or 100 amino acids from the C-terminal domain are removed, respectively), containing only segments of the C-terminal domain (290–349, 313–349, 313–390), or lacking the C-terminal 56 amino acids in addition to lacking residues 202–222 (Δ56Δaf) (Fig. 5A). It is this C-terminal domain comprising 100 amino acids that is predicted to be disordered and serves as a protein-interaction domain^12,13,21. The dot blot revealed that A5-Fc preferentially interacted with N-terminal constructs (i.e. lacking portions of the C-terminal domain), indicating a strong preference for the glycosylase domain over the C-terminal disordered tail (Fig. 5B). Similarly, far western analysis using A5-Flag-10His exhibited selective binding to the N-terminus of NEIL1. However, while A5-Flag-10His showed mild interaction with NEIL1 290–390, it did not bind the shorter C-terminal constructs 290–349 or 313–349 (Fig. 5C–D). These findings confirm that A5 primarily recognizes the N-terminal glycosylase domain, reinforcing its specificity and utility for studying NEIL1 interactions.