1. Introduction
The DNA damage response (DDR) is a collection of signaling pathways initiated by independent DNA-binding proteins that sense DNA damage and specific DNA structures including DNA double-strand breaks (DSBs). These sensors, namely MRE11, RPA, and Ku then activate the phosphatidylinositol-3-kinase-related kinases (PIKKs) ATM, ATR, and DNA-PK, respectively, to regulate cell cycle, cell death, DNA replication, and DNA repair [
1]. Targeting the DDR is a promising therapeutic strategy, as many therapeutic modalities induce cancer cell death through generation of DNA damage.
DSBs are induced by ionizing radiation (IR), radiomimetic agents, and other chemotherapeutics that directly damage DNA. DSBs can also be induced during repair of DNA damage. A variety of pathways can repair DNA DSBs; however, repair-proficient cells will mostly employ the highly efficient and cell cycle-independent, error-prone non-homologous end-joining (NHEJ) pathway [
2]. NHEJ is initiated by binding of the Ku heterodimer to DNA termini, creating a platform for binding and subsequent activation of DNA-PKcs. The importance of DNA-PK in IR response has led to development of several small-molecule DNA-PK inhibitors that target kinase active sites, and these inhibitors are being pursued both pre-clinically and clinically [
3]. As an alternative to ATP-mimetic kinase inhibitors, we first reported the discovery of Ku–DNA binding inhibitors (Ku-DBis), demonstrating in vitro inhibition of Ku binding and DNA-PK activation [
4]. Through initial optimization of Ku-DBi, we identified a series of compounds with potent in vitro inhibition of DNA-PK- and NHEJ-catalyzed DSB repair [
5]. Although limited cellular uptake of these Ku-DBis necessitated use of serum-free media for cellular Ku-DBi treatment, we were able to demonstrate on-target cellular activity and sensitization to IR and other DSB-inducing treatments [
5,
6].
In this report, we performed further chemical optimization, leading to the discovery of a series of oxindole derivatives based on our earlier Ku-DBi core structure. These novel Ku-DBis exhibited improved cellular uptake and retained potent Ku inhibitory activity. Interestingly, HR-deficient triple-negative breast cancer (TNBC) cells were more resistant to Ku-DBis and displayed antagonism with DSB-inducing agents. On the other hand, the results obtained in NSCLC models were markedly different, with the results in ATM-null DSB repair-deficient lines displaying high sensitivity and synergy with IR. In addition, we demonstrated in vivo activity in a NSCLC xenograft model where IR-induced DNA-PK autophosphorylation was abrogated by pre-treatment with the optimized Ku-DBi. Furthermore, a decreased expression in the proliferation marker Ki-67 was observed in the cells treated with Ku-DBi in combination with IR, as assessed by both Western blotting and immunohistochemistry (IHC). An increase in γ-H2AX was also observed in combination-treated tumors, consistent with ATM activation, as we have previously described in vitro [
6]. These data demonstrate the first in vivo assessment of an optimized novel Ku-DBi and pave the way for further development of this anticancer therapeutic strategy in combination with radiotherapy.
2. Materials and Methods
2.1. Chemistry and Drug Reconstitution
The synthetic schemes, procedures, and characterization of the novel Ku-DBis are provided in the
Supplementary Data. Ku-DBis (10 mM) and NU-7441 (5 mM) (Cat: 3712, TOCRIS, Minneapolis, MN, USA) powder were reconstituted in dimethyl sulfoxide (DMSO) and stored at room temperature and protected from light. Bleomycin stock (2.5 mM) (bleomycin sulfate, cell grade; Cat: J67560-S, Alfa Aesar, Haverhill, MA, USA) was prepared in sterile water and stored at −20 °C.
2.2. Protein Purification and Preparation
Purification of Ku 70/80 heterodimer and DNA-PKcs was prepared from baculovirus-infected Sf9 cells and HeLa or HEK-293 cells, respectively, as previously described [
7].
For biophysical analyses, full-length Ku heterodimer was expressed in and purified from Sf21 insect cells using a multiBac expression system as previously described [
8,
9]. Protein labeling was achieved following buffer exchange using an Amicon 50 kDa cutoff concentrator into PBS (10 mM phosphate buffer, pH 7.4, 2.7 mM KCl, 137 mM NaCl) supplemented with 0.005% Tween-20.
2.3. Biophysical Analysis
2.3.1. Microscale Thermophoresis (MST)
Ku heterodimer was labeled using the Monolith NT™ Protein Labeling Kit RED-NHS (2nd generation amine reactive, NanoTemper Technologies GmbH, MO-L011, München, Germany), following the manufacturer’s protocol. Ku-DBi 3392 powder was resuspended in 100% DMSO, and the concentration was adjusted to 50 mM after UV absorbance measurement (absorbance at 417 nm with an extension coefficient equal to 15.2 mM−1 cm−1). The ligand stock solution was sonicated in an ultrasonic bath before preparations of the dilution series. The MST and Differential Scanning Fluorometry (DSF) assays contain 5% DMSO to fit the DMSO concentration at the highest ligand concentration (200 µM). The labeled protein (5 nM) was incubated with ligand 3392 ranging from 98 nM to 200 µM in a 12-point 1:1 dilution series. Protein–ligand solutions were incubated for 30 min at room temperature before MST measurements, which were performed in triplicate. Proteins were transferred to capillaries (Monolith NT. Automated Premium Capillary Chips), and binding was analyzed with a Monolith NT. Automated pico-RED device using MO. Control Software with nano-red excitation, LED light adjusted to 20% excitation power, and infrared laser (MST power) set to medium. The dissociation constants (KD) were determined with MO. Binding Affinity software ( NanoTemper Technologies GmbH, v1.6) using the single-site fit binding mode (referred to as KD binding mode in the software).
2.3.2. nanoDifferential Scanning Fluorometry (nanoDSF)
The thermostability and aggregation propensity of Ku was assessed using a Prometheus Panta nanoDSF instrument (NanoTemper Technologies GmbH, München, Germany). High-sensitivity capillaries were filled with Ku (10 μL each, experiments in triplicate) at a concentration of 5 µM into the instrument. We titrated against ligand 3392 covering a concentration ranging from 0 to 30 µM with measurements collected at a 1 °C-per-minute scan rate from 25 to 95 °C. Thermal unfolding was assessed by the increase in intrinsic tryptophan fluorescence detected at 350 and 330 nm following excitation at 280 nm. Inflection points in the thermal denaturation curves were identified using the first derivatives of the ratio r = fluorescence (350 nm)/fluorescence (330 nm). We measured dynamic light scattering (DLS) and turbidity throughout the heating ramp using back reflection to characterize protein aggregation. We determined aggregation temperatures using turbidity variations and increases in particle sizes.
2.4. Biochemical Activity Assays
The inhibitory activity of Ku-DBis on Ku-DNA binding was evaluated in vitro using Electrophoretic mobility shift assays (EMSA) with a [
32P]-labeled 30-bp duplex DNA and purified Ku 70/80. Kinase assays were performed to measure the DNA-dependent transfer of [
32P] from γ-[
32P] ATP to a synthetic p53-based peptide substrate, both as previously described [
7]. Both EMSA and DNA-PK inhibitor titration assays were performed in triplicate. Data were fit to standard binding curves using GraphPad Prism to determine IC
50 values.
2.5. Cell Lines and Cell Culture
Human NSCLC cell lines H460 (HTB-177™), A549 (CCL-185™), H1299 (CRL-5803™), and H23 (CRL-5800™) and triple negative breast cancer (TNBC) cell lines MDA-MB-436 (HTB-130™) and MDA-MB-468 (HTB-132™) were obtained from ATCC®.
H460, H1299, and H23 cells were cultured in RPMI-1640 medium (RPMI-1640 with L-glutamine; Cat:10-040-CV, Corning, Corning, NY, USA). A549 cells were cultured in F-12K medium (Ham’s F-12K nutrient mixture, Kaighn’s modified, with L-glutamine; Cat: 10-025-CV, Corning). MDA-MB-436 and MDA-MB-468 cells were cultured in DMEM/F12 mixture (1X Dulbecco’s modification of Eagle’s medium with 4.5 g/L glucose, L-glutamine and sodium pyruvate. Cat: 10-013-CV, Corning/Ham’s F-12, 1X, modified with L-glutamine. Cat: 10-080-CV, Corning). Cell number and seeding details for each experiment are provided in the corresponding assay descriptions. All cell lines were obtained from American Type Culture Collection (ATCC, Manassas, VA, USA).
All media were supplemented with 10% fetal bovine serum and 1X penicillin–streptomycin [Penicillin Streptomycin solution, 100×; penicillin (10,000 IU/mL) and streptomycin (10,000 μg/mL) Cat: 30-002-CI, Corning].
For treatments, the vehicle controls correspond to 1% DMSO. Ku-DBi treatments were prepared in supplemented media at a 1% DMSO final concentration.
2.6. Determination of Cellular Drug Uptake
Cellular uptake was evaluated in two NSCLC cell lines, H460 and A549. Cells were seeded into 6-well plates at a density of 1 × 106 cells per well and cultured as monolayers at 37 °C and 5% CO2 for 18–24 h. Ku-DBis were added to the media at a concentration of 10 µM and incubation continued for the times indicated. Media were removed, cells were washed 3 times with PBS, 1 mL of methanol was added to each well, and cells were agitated overnight at 4 °C. The methanol was collected, and wells were washed with an additional 1 mL of methanol that was then pooled with the original. The methanol extracts were then dried under vacuum, resuspended in methanol, and analyzed by HPLC using a 250 × 4.6 and 5 μm C-18 reverse phase column with 1% TFA mobile phase in acetonitrile (ACN) gradient elution. Absorption was monitored at 425 nm, HPLC peaks were quantified using a standard curve, and picomoles of compound per million cells was calculated.
2.7. Cell Viability Assay Assessment
Cell metabolism/viability was assessed using a mitochondrial metabolism assay (CCK-8) kit (Cell Counting Kit-8, Cat: CK04, Dojindo Laboratories, Rockville, MD, USA) as previously described [
6]. Briefly, cells were seeded at a density of 3 × 10
3 cells per well in 96-well plates and allowed to adhere and stretch for 18–24 h prior to treatment. Cells were pre-treated with Ku-DBi for 24 h before bleomycin treatments. Subsequently, cells were incubated for an additional 48 h at 37 °C after which CCK-8 assay was performed. The absorbance of the formed formazan product in each well was measured at 450 nm and compared to vehicle-treated controls to determine percent cell viability. The results represent the average and SEM of triplicate determinations.
2.8. Cell Irradiation and Clonogenic Survival Assays
NSCLC cells were seeded at 2 × 105 cells per well into 24-well plates and grown as monolayers at 37 °C and 5% CO2 in media one day before Ku-DBi or vehicle treatments. After 2 h, cells were placed on ice for 10 min and kept on ice during irradiation with either 2 or 5 Gy of 160 kVp X-rays using a Precision X-ray machine (North Branford, CT, USA) at a dose rate of 0.687 Gy/min. Radiation dosimetry measurements were performed using a Farmer-type ionization chamber (PTW Model N30013, Freiburg, Germany) in conjunction with a Keithley electrometer (Model K602, Cleveland, OH, USA). After irradiation, cells were incubated for 1 h at 37 °C, trypsinized, replated in 100 mm dishes, and incubated at 37 °C in 5% CO2. After 11 days, cells were carefully washed in PBS before fixing and staining in 0.5% crystal violet (Cat: C581-100, Fisher Scientific, Hampton, NH, USA) with 6% glutaraldehyde (Glutaraldehyde solution, 25%. Cat: O2957-1, Fisher Scientific) in PBS for 1 h at RT. The staining solution was removed, and cells were washed with water and air-dried. For clonogenic survival analysis, colonies containing more than 50 cells were manually counted and normalized to vehicle control conditions to determine surviving fractions.
2.9. Cancer Cell Line-Derived Xenograft (CDX) Model Studies
The in vivo studies were conducted as approved by the Institutional Animal Care and Use Committee at Indiana University School of Medicine. A549 cells (~2.5 × 10
6) in 50% Matrigel were injected into the hind flanks of 8–10-week-old NOD.Cg-Rag1tm1Mom Il2rgtm1Wjl/SzJ (NRG) mice (IVT, In vivo Therapeutics Core, Indiana University Melvin and Bren Simon Comprehensive Cancer Center, Indianapolis, IN, USA) as previously described [
10]. Tumor volumes were monitored by electronic caliper measurement. Mice were randomized into groups of 3 or 4 using a random group generator; 6 different scenarios were generated, and the scenario with the most similar tumor volume averages across groups was selected. Treatments were initiated when tumors were ~600 mm
3. Tumors were administered Ku-DBi (30 µL of 10 mM) or vehicle (30 µL of DMSO) via intratumoral injection 4 h before 5 Gy irradiation of the tumor with a single dose of X-rays (250 kVp; dose rate = 1.4 Gy/min; 2 cm × 2 cm field) under isoflurane anesthesia. Mice that did not receive the irradiation treatment were placed under isoflurane anesthesia for the same length of time as the irradiated mice (~8 min). Mice were sacrificed 2 h post-irradiation (or 2 h post-sham irradiation); tumors were excised and processed for immunohistochemistry (IHC) assays. Cell-free protein extracts were prepared to assess DNA-PKcs autophosphorylation and γ-H2AX levels.
2.10. Protein Extraction and Western Blotting
Cell cultures were kept on ice for 15 min before protein extraction. Media was removed, and cells were washed in cold PBS and lysed in RIPA buffer (50 mM Tris, pH 8.0, 150 mM sodium chloride, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM EDTA). A protease- and phosphatase-inhibitors cocktail [Halt™ Protease and phosphatase inhibitor, single-use cocktail (Thermo Fisher Scientific, Waltham, MA, USA)] was added to lysis buffer before use. Bath sonication on ice was used for disrupting cellular membranes and to release the cells contents. Lysates were centrifuged at 4 °C for 20 min at 14,000 rpm, and the supernatants were collected. Pellets were saved and processed for γ-H2AX immunodetection from nuclear extracts.
2.11. Preparation of Nuclear Extracts from Cellular Lysates or Tissue
Nuclear extracts for γ-H2AX immunodetection were prepared following a protocol modified from Abmayr et al. [
11]. Briefly, cell pellets were lysed in 200–300 µL nuclear extraction buffer (50 mM Tris, pH 8.0, 300 mM sodium chloride, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM EDTA, 10% glycerol). Protease and phosphatase inhibitors were added before use, and DNA in the samples was sheared using a 21-gauge needle, vortexed every few minutes for 10 min, and sonicated on ice using a Qsonica Sonicator (Q125 Sonicator, Newtown, CT, USA) for 16 s on 0.2 s on/off pulse cycles at 20% amplitude. Samples were centrifuged at 14,000 rpm at 4 °C for 20 min, and supernatant containing the nuclear fraction was collected. Protein content was quantified using a BCA protein assay kit (Pierce™ BCA protein assay kit; Cat: 23227, Thermo Scientific, Waltham, MA, USA).
2.12. Electrophoresis and Western Blotting
Protein lysates, ranging from 20 to 30 μg, were separated by SDS-PAGE (4–20% Mini Protean TGX Gels, BioRad, Hercules, CA, USA) and transferred onto PVDF membranes (Bio-Rad) through wet transfer. Membranes blocked in 0.5% Tween and 3% BSA in 1× TBS for 1 h were probed with the following primary antibodies: Phospho-DNA PKcs (S2056) (Cat: ab124918, Abcam, Cambridge, UK), total DNA-PKcs (Cat: sc-5282, Santa Cruz Biotechnology, Dallas, TX, USA), and γ-H2AX pS139 (Cat: 613401, BioLegend, San Diego, CA, USA). Membranes blocked in 0.5% Tween and 5% non-fat dried milk in 1× TBS for 1 h were probed with the following primary antibodies: Ki-67 (Cat: ab16667, Abcam) and GAPDH (Cat: MA-15738, Thermo Fisher, Waltham, MA, USA). All primary antibodies were prepared with 3% BSA and 0.02% NaN
3 in TBS with 0.5% Tween. Goat anti-rabbit or anti-mouse IgG (H + L)–HRP conjugates (Cat: 170-6515 or 170-6516, Bio-Rad) were used as secondary antibodies to detect these proteins. Imaging, densitometric analysis, and quantification of protein expression were performed as previously described [
5,
6].
2.13. Tissue Sections and Immunohistochemistry (IHC) Staining
Excised tumors were fixed in 4% paraformaldehyde (PFA) solution in PBS for 24 h at 4 °C. After fixation, tumors were transferred to 70% ethanol and paraffin embedding. Sectioning was completed by the Indiana University Histology Core at the IU School of Medicine. Representative 3 μm-thick sections for each treatment were stained with H and E (HE; Harris hematoxylin, regressive method) for histologic examination or were processed for Ki-67 IHC immunodetection. Briefly, paraffin-embedded sections were deparaffinized in a xylene series (4 × 3 min) and rehydrated in a series of graded alcohol dilutions (100%, 95%, 70%, and ddH2O, 2 × 2 min each). Ki-67 antigens were retrieved using a low-pH antigen-retrieval solution (10 mM sodium citrate, pH 6.0, Cat: 00-4955-58, Thermo Fisher Scientific) and subsequently cooled on ice for 15 min. Hydrogen peroxide 3% in water reagent was used to block endogenous peroxidase activity (Hydrogen Peroxide 30% v/v, Cat: H1009 7722-84-1, Sigma, St. Louis, MO, USA), and unspecific binding was blocked in blocking buffer (normal goat serum, Cat: PK-6101 Vectastain ® Elite ABC + HRP, Vector Laboratories, Newark, CA, USA) for 1 h at RT. CDX sections were then incubated with primary antibody against Ki-67 (Cat: ab16667, Abcam) prior to incubating in anti-rabbit biotinylated secondary antibody. A rabbit-specific ABC + HRP Detection Kit (Cat: PK-6101 Rabbit IgG, Vectastain ® Elite ABC + HRP, Vector Laboratories) was used, and Mayer’s hematoxylin was used as a counterstain (Hematoxylin Solution, Mayer’s modified, Cat: ab220365, Abcam). Diaminobenzidine substrate [DAB (3,3′-diaminobenzidine) Cat: SK-4100, Vector laboratories] was used as a stain to detect IHC reactivity. Stained tissue sections were scanned using an Aperio ScanScope AT Slide Imager (Leica Biosystems; Buffalo Grove, IL, USA) at 20× magnification. The resulting images were then viewed using Aperio ImageScope (v12.3.2).
2.14. Statistical Analysis
Statistical analyses were performed using GraphPad Prism software (v.10.2.3). The specific statistical tests are indicated in the figure legends with a p < 0.05 indicative of statistical significance.
4. Discussion
Targeting the DDR to treat cancer has gained considerable traction following the successful deployment of PARP inhibitors (PARPi) in the treatment of breast, ovarian, and prostate cancers that harbor mutations in BRCA1 and 2. While the exact PARPi mechanism of action and the relevant DNA “signal” that PARPi exploits is debated [
17,
18,
19,
20], there is unequivocal data that these inactivating mutations render cells hyper-sensitive to PARPi [
21]. As the DDR is potentiated with the activation of PI3KK, a wide array of small-molecular inhibitors targeting numerous DDR kinases have been reported, with DNA-PK being no exception [
3]. The initial development of active site-binding ATP mimetics has resulted in high in vitro potency under assay conditions for screening and selection [
22,
23]. However, DNA-PK has a
KM for ATP of 25 µM, and intracellular ATP concentration can be as high as 5 mM [
24], which makes ATP-competitive reagents less effective in cells and tissues despite impressive in vitro inhibitory constants. Consistently, recent reports from clinical trials assessing DNA-PKcs inhibitors in combination with IR and other chemotherapeutics have revealed inadequacy of target inactivation, with insufficient inactivation at tolerable doses [
25,
26,
27], highlighting the limitations of ATP mimetics: specificity and selectivity. While in vitro IC
50s are often in the nM range, higher clinical doses are required to overcome ATP concentrations, which can result in high off-target activity and toxicity. It has been estimated that off-target activity can extend to dozens of kinases [
28,
29], which does not account for numerous other ATP-binding proteins that could be affected. Therefore, development of allosteric inhibitors represents one avenue to circumvent the ATP-competitive mechanism, and the size and complexity of the DNA-PKcs structure provides ample opportunity toward this end [
30].
We have exploited a novel approach of targeting the requirement for Ku–DNA binding for DNA-PK activation and developed Ku-DBis capable of blocking the protein–DNA interaction [
4,
5,
6,
31]. While the development of protein–DNA interaction inhibitors has a long history of failure, many of these attempts were made against sequence-specific transcription factors (TF) [
32]. Importantly, the Ku–DNA interaction is sequence independent and is dictated by DNA structure [
33]. This allows for greater flexibility in chemical efforts to perturb the Ku–DNA interaction compared to a TF, where specificity for a short sequence of nucleotides is needed for interacting with the relatively small binding cleft. The Ku bridge-and-pillar structure is a collection of domains that come together to form a unique structure capable of encircling double-stranded DNA [
33]. While other proteins possess the ability to thread onto DNA, including PCNA and hexameric DNA helicases, these proteins do not employ the same structural motifs to support DNA binding [
34,
35]. This unique aspect of the Ku–DNA interaction leads to the potential to identify unique chemical inhibitors that display high specificity and selectivity. Considering that DNA-PK is the only kinase in the human proteome that requires Ku bound to DNA for activation, this strategy portends the potential for high selectivity and specificity.
In this report, we provide evidence for the feasibility of developing small-molecule Ku inhibitors that possess enhanced cellular uptake and membrane permeability to support cellular target inactivation. We believe that the increased lipophilicity of 3392, as determined by total polar surface area (TPSA) and cLOGP determinations, contributes to the improved cellular uptake and potency (
Supplementary Information, Table S1). We demonstrate that Ku-DBis efficacy varied across different cell lines and tumor types that differ from the variations observed with DNA-PK direct active site inhibition, suggesting that DNA-PK-independent roles of Ku may define novel genetic predictors of sensitivity. Combinatorial activity was observed with a series of DSB-inducing treatments. The results spanned additive, antagonistic, and synergistic interactions, depending on the specific agent and cell line. These data suggest the potential for combination cancer treatment, but a detailed analysis of the predicting factors will be necessary to ensure optimal combinatorial activity. We also report the first demonstration of in vivo activity and on-target inactivation of DNA-PK by a Ku-DBi that reduced tumor cell proliferation in combination with IR (
Figure 6). These data provide the impetus for further developments in medicinal chemistry efforts to optimize Ku-DBis for better systemic delivery and favorable pharmacokinetics.
5. Conclusions
The discovery of Ku–DNA binding inhibitors as a new and innovative strategy for DNA-PK inhibition has enabled us to explore an alternative approach for cellular sensitization to current cancer therapies, making Ku–DNA binding inhibitors potentially valuable for use in cancer therapeutics. The development of a new series of derivatives through chemical optimization resulted in enhanced cellular uptake with retained potent Ku-inhibitory activity and cellular response. This improvement allowed for better Ku-DBis assessment in cell culture conditions, where a potent single-agent activity in an ATM-deficient NSCLC cell line was observed, and a differential interaction of the Ku-DBi with DSB-inducing agents as a function of BRCA1 status was discovered. This suggests that different genetic backgrounds may impact combination therapy potential in diverse patient populations.
Finally, the optimization of Ku-DBis enabled evaluation in live organisms, supporting the potential application of chemical Ku–DNA binding inhibition to disrupt DNA-PKcs autophosphorylation and to modulate the DDR. This ultimately led to a reduction in tumor cell proliferation when used in combination with ionizing radiation (IR).
This study is the first to show in vivo activity and on-target inactivation mediated by a Ku-DBi and provides the foundation for further medicinal chemistry development of Ku-DBis as a potential anticancer therapeutic strategy in combination with radiotherapy and other DNA-damaging agents.