Synthesis, Biological Evaluation and Structure-Activity Relationships of a Novel Class of Apurinic/Apyrimidinic Endonuclease 1 Inhibitors

Abstract

APE1 is an essential protein that operates in the base excision repair (BER) pathway and is responsible for ≥95% of the total apurinic/apyrimidinic (AP) endonuclease activity in human cells. BER is a major pathway that copes with DNA damage induced by several anti-cancer agents, including ionizing radiation and temozolomide. Overexpression of APE1 and enhanced AP endonuclease activity has been linked to increased resistance of tumor cells to treatment with monofunctional alkylators, implicating inhibition of APE1 as a valid strategy for cancer therapy. We report herein the results of a focused medicinal chemistry effort around a novel APE1 inhibitor, N-(3-(benzo[d]thiazol-2-yl)-6-isopropyl-4,5,6,7-tetrahydrothieno[2,3-c]pyridin-2-yl)acetamide (3). Compound 3 and related analogs exhibit single-digit µM activity against the purified APE1 enzyme, comparable activity in HeLa whole cell extract assays, and potentiate the cytotoxicity of the alkylating agents methylmethane sulfonate and temozolomide. Moreover, this class of compounds possesses a generally favorable in vitro ADME profile, along with good exposure levels in plasma and brain following intraperitoneal dosing (30 mg/kg body weight) in mice.

Introduction

Apurinic/apyrimidinic (AP) endonuclease 1 (APE1) is the primary mammalian enzyme responsible for the removal of abasic (or AP) sites in DNA.^1,2 Such lesions can arise as products of the repair activity of DNA glycosylases, which carry out the first step of base excision repair (BER), or as products of spontaneous or damage-induced hydrolysis of the N-glycosidic bond that links the base to the sugar moiety of the phosphodiester DNA backbone. APE1 initiates AP site repair by catalyzing a Mg²⁺-facilitated strand incision event immediately 5’ to the lesion, leaving behind a single-strand break with a normal 3’-hydroxyl residue and a 5’-abasic fragment. Other proteins of BER, such as DNA polymerase β and DNA ligase 3, complete the repair response by removing the remaining abasic residue, replacing the missing nucleotide, and sealing the nick.³ Besides being the predominant repair enzyme for AP sites in DNA, APE1 also possesses repair nuclease activities against various non-conventional 3’-blocking terminal groups, such as phosphoglycolates, phosphates, and chain-terminating nucleoside analogs.⁴ In addition, APE1 has roles in regulating gene expression, most notably through its capacity to modulate the DNA binding activity of several transcription factors, e.g. AP-1, p53 and NF-κB, via a less well understood redox mechanism.^5,6,7

APE1 protein levels and intracellular distribution have been correlated with cancer type and stage, as well as responsiveness to clinical DNA-damaging agents, such as the alkylator temozolomide (TMZ).^8,9 TMZ is a promising drug recently added to the arsenal of alkylating agents for the adjuvant chemotherapy of brain cancers due to its ability to readily cross the blood-brain barrier.¹⁰ It has therefore been postulated that APE1 would be an attractive target in anti-cancer treatment paradigms involving co-administration with certain DNA-interactive drugs, where strategic regulation of its repair activity would improve the therapeutic efficacy and clinical outcome.

Targeting DNA repair enzymes as single-agent cancer therapy has been validated as a viable strategy by the discovery and clinical evaluation of poly ADP-ribose polymerase (PARP) inhibitors.¹¹ PARP1 is an enzyme that facilitates efficient repair of single-strand breaks in DNA. Thus, inhibition of PARP1 leads to the accumulation of one-ended double-strand DNA breaks upon replication fork collapse that are ultimately repaired via homologous recombination (HR).¹² BRCA1/2 are proteins involved in the HR pathway, and consequently, treatment of BRCA-deficient cancer cells (e.g. ~10–20% of triple negative breast cancers) with PARP1 inhibitors leads to irreparable DNA damage and ultimately cell death.¹³ This synthetic lethal relationship offers the prospect of selective targeting of cancer cells, since normal cells would maintain the ability to repair DNA double-strand breaks. It should be noted however that despite continued efforts and promising results in this area of research, the use of PARP1 inhibitors has not been without its recent setbacks in the clinic.¹⁴ While synthetically lethal combinations involving APE1 inhibitors have not yet been established, it is not unreasonable to postulate the use of APE1 inhibitors as single agent therapy by such a mechanism.

As a result of the promising therapeutic potential of this target, several reports have described the identification and characterization of small molecules that inhibit APE1 repair endonuclease activity.¹⁵ Kelley and co-workers described the identification of 2,4,9-trimethylbenzo[b][1,8]naphthyridin-5-amine, 1 (AR03) through a fluorescence-based high-throughput screen (HTS) of 60,000 compounds (Figure 1).¹⁶ 1 was found to have low µM in vitro potency against purified human APE1, and inhibited AP site incision activity of whole cell extracts and the repair of AP sites in SF767 glioblastoma cells. Moreover, 1 potentiated the cytotoxicity of methyl methansulfonate (MMS) and TMZ in SF767 cells. 7-nitro-1H-indole-2-carboxylic acid, 2 (CRT0044876) (Figure 1) was identified by Madhusudan et al. in 2005, and they described specific inhibition of the exonuclease III family of AP endonucleases and the induction of AP sites in HT1080 fibrosarcoma cells.¹⁷ While a synergistic cell killing effect was seen with the inhibitor when combined with MMS or TMZ, other subsequent studies have been unable to reproduce the potentiating effect of 2.¹⁸ More recently, Madhusudan and colleagues described the results of a virtual screen of 2.6 million compounds from which several low µM APE1 inhibitors were found.¹⁹ Other reported APE1 inhibitors include the bis-carboxylic acid containing small molecules described by Zawahir et al.,²⁰ lucanthone (also a topoisomerase inhibitor),²¹ methoxyamine,²² and various arylstibonic acids.²³

Figure 1.

Previously reported APE1 endonuclease inhibitors (1 and 2) and the lead chemotype (3).

We recently reported on the development of a 1536-well fluorescence-based, quantitative HTS (qHTS) assay, which was used to screen the Library of Pharmacologically Active Compounds (LOPAC¹²⁸⁰) for novel APE1 endonuclease inhibitors.²⁴ This library is a collection of well characterized bioactive molecules that is often used to help validate an assay platform before screening of an entire small molecule collection. Our initial efforts identified several compounds, including 6-hydroxy-DL-DOPA, reactive blue and myricetin, which inhibit AP site cleavage activity of purified APE1, as well as HEK293T and HeLa whole cell extracts, and exhibit potentiating effects of MMS toxicity. These prior studies established general methods for identification and validation of APE1 inhibitors. While encouraged by these results, we were eager to identify novel small molecule inhibitors that would represent an improved starting point for medicinal chemistry optimization, as the three LOPAC¹²⁸⁰ mentioned above appear to be generally promiscuous given the broad publications on their various modes of action.

As such, we conducted a large-scale qHTS in a 1536-well concentration response format on our complete in-house collection, which at the time contained 352,498 small molecules²⁵ as part of the NIH Molecular Libraries Small Molecule Repository (MLSMR). This effort represents the largest reported screening campaign against APE1 thus far, and led to the identification of several novel chemotypes displaying potent inhibition of APE1 activity. In this report, we discuss the synthesis, biological characterization and structure-activity relationship (SAR) of one of the molecules, N-(3-(benzo[d]thiazol-2-yl)-6-isopropyl-4,5,6,7-tetrahydrothieno[2,3-c]pyridin-2-yl)acetamide (3) (Figure 1). This compound, along with related analogs, displayed low µM APE1 inhibition and activity in HeLa whole cell extract assays, potentiated the cytotoxicity of MMS and TMZ, and led to a hyper-accumulation of AP sites in HeLa cells treated with MMS.

Chemistry

The synthesis of the lead chemotype 3 from the qHTS campaign commenced with treatment of commercially available N-(Boc)-4-piperidone with 2-benzothiazoleacetonitrile in the presence of elemental sulfur and morpholine to afford the key intermediate 3a as shown in Scheme 1.²⁶ Acetylation of the 2-amino group, followed by Boc-deprotection with trifluoroacetic acid (TFA) and reductive amination with acetone and NaCNBH₃, gave the desired compound 3 in good yield. Starting with intermediate 3a, several different N-substituted analogs were synthesized including R = Me (6) (via reductive amination with paraformaldehyde), R = Bn (7) (via reductive amination with benzaldehyde), and R = Ac (8) (via acylation with AcCl).

Scheme 1.

Synthesis of qHTS “hit” 3 and related analogs 4–8.a

^aReagents and conditions: (a) S, morpholine, EtOH, reflux, 2 h (87%); (b) AcCl, Hunig’s base, CH₂Cl₂, rt, 1 h (91%); (c) TFA, CH₂Cl₂, rt, 1 h (99%); (d) 3: acetone, NaCNBH₃, MeOH, rt, 6 h (75%); 6: paraformaldehyde (aq.), NaCNBH₃, MeOH/THF, rt, 6 h; 7: PhCHO, NaCNBH₃, MeOH/THF, 6 h; 8: AcCl, Hunig’s base, CH₂Cl₂, rt.

The synthesis of analogs 9–24, which involved modification to the C-3 benzothiazole ring, required the development of a more convergent route. While the original synthesis of lead compound 3 only involved 4-steps, diversity of the benzothiazole ring occurs in the first step and thus requires the availability of the necessary acetonitrile derivative. Moreover, each modified benzothiazole analog would involve a minimum of 4-steps, making library synthesis quite laborious. As such, we envisaged a late-stage Suzuki coupling of commercially available boronic acids to the 3-bromothiophene intermediate (11) allowing for rapid access to a variety of analogs at the 3-position as shown in Scheme 2. Reaction of N-(Boc)-4-piperidone with ethyl 2-cyanoacetate using the conditions described above (S, morpholine), followed by acetylation and subsequent saponification of the formed ethyl ester, gave the desired 3-COOH intermediate 9a. Decarboxylation using dimethylacetamide at 170 °C for two hours in a microwave gave the des-carboxy product 10a after optimization of the reaction conditions. Typically, this transformation is carried out with copper using quinoline as the solvent or oxalic acid in isopropanol, both at elevated temperatures. Our conditions are quite mild (albeit at elevated temperatures) and a wide range of functional groups should be tolerated. Subsequent bromination of 10a using Br₂ in chloroform at −10 °C followed by Boc-deprotection gave 11 in good yield. Importantly, we found that the late-stage Suzuki coupling could be accomplished using the unprotected piperidine moiety via treatment of 11 with various commercially available boronic acids to provide 12–20 in good yields. Reductive amination with acetone, NaCNBH₃ and MeOH/THF as the solvent gave the N-isoproyl analogs 22–24.

Scheme 2.

Synthesis of analogs 9–24.a

^aReagents and conditions: (a) S, morpholine, EtOH, reflux, 2 h (84%); (b) AcCl, Hunig’s base, CH₂Cl₂, rt, 1 h (91%); (c) LiOH, THF/H₂O, reflux, 24 h, (65%); (d) TFA, CH₂Cl₂, rt, 0.5 h; (e) dimethylacetamide, 170 °C, µW, 2 h, (77%); (f) Br₂, CHCl₃, −10 °C, 1 h (70%); (g) R-B(OH)₂, Pd(PPh₃)₄, Na₂CO₃, DME, 150 °C, µW, 0.5–1.5 h; (h) acetone, NaCNBH₃, MeOH, rt, 6–10 h.

Analogs 25–46 were synthesized in an effort to further understand the SAR of the amide moiety as shown in Scheme 3. Analog 25, in which the 2-amino group was left unacetylated, was prepared in 2-steps from the common intermediate 3a via Boc-deprotection and regioselective reductive amination of the piperidine moiety with acetone and NaCNBH₃. Analogs 26–33, involving either reaction with the requisite acid chloride or EDC-mediated peptide coupling conditions (for 31, 33, and 36), were utilized to afford the desired products in good yields. In addition to the various amide analogs, we wanted to investigate amide bioisosteres such as oxadiazoles 43–46. The synthesis of oxadiazole analogs 43 and 44 commenced with conversion of the 2-amino group to the corresponding bromide using tert-butyl nitrite and copper bromide in acetonitrile. Palladium-catalyzed carboxylation was achieved using cat. Pd(OAc)₂, cat. dppp, in DMSO-MeOH under an atmosphere of CO_(g) to afford the desired methyl ester derivative 41a in 69% yield. Saponification of 41a with lithium hydroxide in a THF/MeOH/H₂O mixture gave carboxylic acid 42a in high yield. Formation of the acylhydrazide was accomplished using acetohydrazide and EDC in DMF. Dehydrative cyclization was achieved using methyl N-(triethylammoniumsulfonyl)carbamate (Burgess Reagent) in THF at 100 °C in the microwave to give, after Boc-deprotection (TFA, CH₂Cl₂), the 1,3,4-substituted oxadiazole 43 in good yield. 1,2,4-oxadiazole derivative 44 was prepared in 3-steps from intermediate 42a via treatment with N'-hydroxyacetimidamide, HATU, Hunig’s base at (50 °C→100 °C) to afford the cyclized product, which after Boc-deprotection gave the desired product 44. Access to 2-amino oxadiazole analogs 45 and 46 was achieved via Buchwald-Hartwig-type cross couplings of the requisite commercially available amino-oxadiazole (5-methyl-1,3,4-oxadiazol-2-amine: 45 and 3-methyl-1,2,4-oxadiazol-5-amine: 46) using Pd₂(dba)₃, Xantphos, cesium carbonate at 120 °C in the microwave for two hours followed by Boc-deprotection.

Scheme 3.

Synthesis of analogs 25–46.a

^aReagents and conditions: (a) TFA, CH₂Cl₂, rt, 0.5 h; (b) acetone, NaCNBH₃, MeOH, rt, 6 h; (c) RC(O)Cl, (iPr)₂NEt, CH₂Cl₂, rt; 31,33,36: RC(O)OH, EDC, DMAP, DMF, rt, 1 h, (85–90%); (d) NaNO₂, HCl-H₂SO₄, 0 °C, 2 h; (e) H₃PO₂, rt, 1 h; (f) t-BuNO₂, CuBr₂, MeCN, 0 °C→rt, 4 h; (g) Pd(OAc)₂, dppp, CO_(g) (1 atm), NEt₃, DMSO-MeOH (1:1), 60 °C, 24 h, (69%); (h) LiOH, THF/MeOH/H₂O (6:2:1), rt, 3 h, (91%); (i) acetohydrazide, EDC, DMF, rt, 2 h (97%); (j) Burgess reagent, THF, 100 °C, µW, 0.5 h, (73%), then TFA, CH₂Cl₂, rt, 0.5 h; (k) N'-hydroxyacetimidamide, HATU, (iPr)₂NEt, 50 °C-15 min. then 150 °C-15 min. µW, 1 h, then TFA, CH₂Cl₂, rt, 0.5 h; (l) 5-methyl-1,3,4-oxadiazol-2-amine, Pd₂(dba)₃, Xantphos, Cs₂CO₃, dioxane, 125 °C, µW, 2 h, then TFA, CH₂Cl₂, rt, 0.5 h; (m) 3-methyl-1,2,4-oxadiazol-5-amine, Pd₂(dba)₃, Xantphos, Cs₂CO₃, dioxane, 125 °C, µW, 2 h, then TFA, CH₂Cl₂, rt, 0.5 h.

As shown in Scheme 4, we were eager to investigate various heteroatoms in place of the piperidine nitrogen of lead compound 3 (e.g. O:47, S:48 and SO₂:49). Accordingly, the synthesis of 47 and 48 were carried out in a manner similar to that shown in Scheme 1, except tetrahydropyran-4-one and tetrahydro-thiopyran-4-one were used for the preparation of 47 and 48, respectively. Synthesis of sulfone derivative 49 was accomplished via treatment of 48 with m-CPBA in methylene chloride at 0 °C, and pyridine analog 50 was obtained via MnO₂ oxidation of intermediate 4. 5-Membered ring analogs 51 and 52 were prepared using the route described in scheme 1 except N-Boc-3-pyrrolidinone was used as the starting material whereas for 53, N-(Boc)-3-piperidone was utilized. Analogs 54 and 55 were synthesized from 1-Boc-4-azepanone, which upon cyclization (S, morpholine), gave two separable regioisomers that were easily distinguishable by ¹H NMR; both reactions were carried through the described sequence to provide the desired products.

Scheme 4.

Synthesis of analogs 47–55.

^aReagents and conditions: (a) m-CPBA, CH₂Cl₂, 0 °C, 1 h, (87%); (b) MnO₂, toluene, 120 °C, µW, 10 min., (49%).

Results and Discussion

SAR analysis of compound 3 using the fluorescence-based HTS assay

As a benchmark, we included some of the previously reported APE1 inhibitors. Specifically, 1 was already part of the MLSMR collection and 2 was purchased from Sigma Chemical Company. As shown in Table 1, compound 2 showed only modest inhibition (40%) at concentrations up to 57 µM, in agreement with previous reports that showed an inability to reproduce the activity of this small molecule.¹⁸ 1 did show APE1 inhibition, albeit lower than the reported IC₅₀ value (1.5 µM, reported¹⁶ vs. 32 µM, observed). Since assay variability is to be expected, this observation emphasizes the need to include prior art as internal comparative controls. It should also be noted that this result is from the primary screen and not from a synthesized or purified commercial powder source, so the source and effective concentration of the sample could also contribute to the discrepancy. Our lead compound 3 exhibited an IC₅₀ value of 2 µM in the qHTS assay (Table 1) and an IC₅₀ value of 12 µM in a radiotracer incision assay (RIA). The confirmed activity, chemical tractability and preliminary cell-based data, which showed that compound 3 potentiated the cytotoxicity of MMS in HeLa cells (vide infra), led us to proceed with optimization and SAR exploration of this chemotype.

Table 1.

Activity of previously reported APE1 inhibitors (1 and 2) and SAR of analogs 3–8.

			IC50a (µM)
	Compound	R	[± SD (µM)]	Inh. @ 57 µMb
	1	NA	32c
	2	NA	—	40%
	3	CH(CH3)2	2.0 [0.1]
	4	H	2.9 [0.1]
	5	Boc	>57
	6	Me	3.8 [0.3]
	7	Bn	—	41%
	8	Ac	—	59%

^aIC₅₀ values were determined using the qHTS protocol and represent the average of three separate experiments.

^bFor compounds that did not achieve full inhibition (>85%), the % inhibition at the highest tested concentration (57 µM) is noted. Compounds noted as >57 µM are considered inactive.

^cIC₅₀ value of this compound reflects data from primary screen; follow-up studies with a powder sample was not further pursued.

Our first SAR investigations involved modification of the piperidine nitrogen moiety, which revealed that acylation was not well tolerated, as introduction of both a Boc group (5) and acetyl group (8) led to a significant loss of activity. Moreover, the benzyl substituted analog (7) showed greatly reduced activity with only 41% inhibition at the highest concentration tested (57 µM). In contrast, removal of the isopropyl group (4) and introduction of the smaller alkyl group (R = Me, 6) had little effect on potency with IC₅₀ values of 2.9 and 3.8 µM respectively. These data suggest that potency is greatly affected by the absence of a basic nitrogen (analogs 5 and 8). However, there is also not much tolerance for larger hydrophobic groups given that the benzyl substituted analog, which maintains the basic character of the nitrogen, is weakly active.

The next area of SAR exploration involved modification of the benzothiazole moiety at the 3-position of the thiophene. Given the results obtained from our initial round of SAR (Table 1), where we found analog 4 (lacking isopropyl group) to have comparable potency to derivative 3, most analogs were submitted for testing as the free amine to eliminate one step in the analog synthesis. As shown in Table 2, many of the changes were not productive, with many compounds being inactive or only having modest activity at the highest concentration. Replacing the benzotriazole motif with other heterocyles, including furan (13), thiophene (14), benzothiophene (15), benzofuran (17), oxazole (19) and thiazole (20), resulted in greatly diminished activity. Similar results were obtained when that position was modified with a simple phenyl group (22) or numerous other substituted phenyl derivatives (data not shown). Interestingly, even the structurally similar benzoxazole analog (23) had a 10-fold loss in activity; however, the 2-(4-phenylthiazole) analog (24) displayed potent inhibition with activity comparable to the lead compound. The combination of the reduced activity for the benzoxazole derivative (23) and the maintenance of activity for compound 24 suggests that the thiazole motif is involved in an essential interaction with APE1 and should be maintained in some capacity in future analogs. However, additional substitution is required since the thiazole alone (compound 20) displayed minimal activity. As such, additional modifications of the pendant aryl ring in analog 24 may prove fruitful.

Table 2.

SAR of analogs 9–24.

				IC50a (µM)
	Compound	R1	R2	[± SD (µM)]	Inh. @ 57 µMb
	9	H	COOH	—	29%
	10	H	H	—	47%
	11	H	Br	—	30%
	12	H	1-naphthalene	>57
	13	H	2-furan	—	41%
	14	H	2-thiophene	—	34%
	15	H	2-benzothiophene	—	23%
	16	H	3-benzothiophene	>57
	17	H	2-benzofuran	—	66%
	18	H	2-indole	>57
	19	H	2-oxazole	—	45%
	20	H	2-thiazole	—	55%
	21	CH(CH3)2	CN	—	31%
	22	H	Ph	—	42%
	23	CH(CH3)2	benzoxazole	20.4 [0.1]
	24	CH(CH3)2	NA	2.9 [0.1]

^aIC₅₀ values were determined using the qHTS protocol and represent the average of three separate experiments.

^bFor compounds that did not achieve full inhibition (>85%), the % inhibition at the highest tested concentration (57 µM) is noted. Compounds noted as ^>57 µM are considered inactive.

The goal of analogs 25–46 (Table 3) was to gain a better understanding of the SAR associated with the N-acetyl moiety of the lead compound 3. Thus, a variety of different compounds were prepared, including a des-acetyl, numerous amides, carbamates, carboxylic acids, esters and amide bioisosteres (1,2,4-oxadiazoles and 1,3,4-oxadiazoles). Our first interest was to look at the des-acetyl analog 25; this modification was clearly unfavorable with only 40% inhibition at 57 µM being observed. Interestingly, compound 25 was less potent than analog 40, which has the amino group removed all together. Many of the amide analogs resulted in a decrease in potency [e.g. t-Butyl (26) = inactive; cyclopentane (28) = 55% at 57 µM; Bn (29) = 53% at 57 µM; and (CH₂)₂Ph (30) = inactive]. However, other amide analogs exhibited a less pronounced drop in activity [e.g. cyclopropane (27) = 6 µM; cyclohexane (37) = 8.3 µM, CHF₂ (33 and 36) = 3.8 and 5.3 µM, respectively]. Replacing the amide with other functional groups, such as carboxylic acid (42) or ester (41), resulted in complete loss of activity; however, methylcarbamate (32) was marginally active with an IC₅₀ of 11 µM. Given the possibility of hydrolysis of the amide by intracellular amidases, we were eager to explore the tolerance of amide bioisosteres such as 1,2,4-oxadiazoles and 1,3,4-oxadiazoles. To this end, both analogs that lack the 2-amino group (43 and 44) and those that maintained this moiety (45 and 46) were investigated. The studies revealed the necessity of the 2-amino group for maintaining activity for these bioisosteres, as compounds 43 and 44 were completely inactive, while 45 and 46 had modest activity. While these latter two compounds have approximately 8-fold less activity than the corresponding amide analog (4), it does suggest the possibility of utilizing such modifications on this class of compounds in later rounds of optimization if deemed necessary by pharmacokinetic (PK) studies.

Table 3.

SAR of analogs 25–46.

				IC50a (µM)
	Compound	R1	R2	[± SD (µM)]	Inh. @ 57 nMb
	25	CH(CH3)2	NH2	—	40%
	26	H	C(CH3)3	>57
	27	H	cyclopropane	6.0 [0.4]
	28	H	cyclopentane	—	55%
	29	CH(CH3)2	Bn	—	53%
	30	H	(CH2)2Ph	>57
	31	H	CF3	17.6 [2.4]
	32	H	OMe	11.1 [0.7]
	33	H	CHF2	3.8 [0.3]
	34	H	(CH2)2C≡CH	12.0 [0.8]
	35	CH(CH3)2	(CH2)2C≡CH	—	57%
	36	CH(CH3)2	CHF2	5.3 [0.4]
	37	CH(CH3)2	cyclohexane	8.3 [0.1]
	38	CH(CH3)2	(CH2)3CH3	9.1 [0.1]
	39	CH(CH3)2	Ph	—	57%
	40	H	H	25.2 [1.9]
	41	H	C(O)OMe	>57
	42	H	COOH	>57
	43	H		>57
	44	H		>57
	45	H		19.0 [1.3]
	46	H		18.2 [0.1]

^aIC₅₀ values were determined using the qHTS protocol and represent the average of three separate experiments.

^bFor compounds that did not achieve full inhibition (>85%), the % inhibition at the highest tested concentration (57 µM) is noted. Compounds noted as >57 µM are considered inactive.

Our last foray into the SAR characterization described herein involved modification of the piperidine ring, where we replaced the nitrogen with other heteroatoms. In the absence of structural information to guide our medicinal chemistry efforts, we aimed to explore the effect (i.e. altering key hydrogen bond contacts) of moving the nitrogen to different positions on the ring and varying the ring size (e.g. n = 5, 6 and 7). As shown in Table 4, most of these changes were met with limited success; however, a few interesting lessons were learned. Replacing the nitrogen with oxygen (analog 47) or sulfur (analog 48) led to compounds that were still active against APE1, albeit only at the highest concentration. Sulfone derivative 49 had very minimal efficacy (29%) and pyridine analog 50 was completely inactive. Varying the ring size with pyrrolidine analogs 51 and 52 had a limited effect on activity, with IC₅₀ values of 3 µM and 3.1 µM, respectively. Differentially substituted piperidine analog 53 had only modest efficacy at 57 µM, while the seven-membered ring derivatives 54 and 55 displayed activities of 12.9 and 17.6 µM, respectively.

Table 4.

SAR of analogs 47–55.

				IC50a (µM)
	Compounds	R	Heterocycle	[± SD (µM)]	Inh. @ 57 µMb
	47	NA		—	41%
	48	NA		—	56%
	49	NA		—	29%
	50	NA		>57
	51	CH(CH3)2		3.1 [0.2]
	52	H		3.3 [0.2]
	53	H		—	31%
	54	CH(CH3)2		12.9 [0.1]
	55	CH(CH3)2		17.6 [1.1]

^aIC₅₀ values were determined using the qHTS protocol and represent the average of three separate experiments.

^bFor compounds that did not achieve full inhibition (>85%), the % inhibition at the highest tested concentration (57 µM) is noted. Compounds noted as >57 µM are considered inactive.

APE1 inhibition in radiotracer incision and HeLa whole cell extract assays

Having tested all of the synthesized analogs using the fluorescence-based HTS assay to establish a rank order of potency, we next aimed to validate their activity by testing in the RIA. This assay is considered the “golden standard”, but due to its low throughput nature, only our top analogs were analyzed via this method (Table 5). Most analogs had comparable activity across both experimental platforms with 52 emerging as the most potent compound in the RIA.

Table 5.

APE1 inhibition of select compounds in radiotracer incision, HTS, and HeLa whole cell extracts assays.

			HeLa cell extract
Compound	RIA IC50a	HTS IC50	% incision activityb
3	12 µM	2.0 µM	0.6 (±0.11)
4	5 µM	2.9 µM	1.4 (±1.1)
6	5 µM	3.8 µM	26 (± 3.7)
23	13 nM	20.4 nM	12 (±1.3)
24	3 µM	2.9 µM	15 (± 0.5)
33	4 µM	3.8 µM	38 (± 7.7)
51	4 µM	3.1 µM	57 (±14)
52	1 µM	3.3 µM	0.1 (± 0.03)
54	14 µM	12.9 µM	14 (±2.8)
55	13 µM	17.6 µM	1.1 (± 0.3)

^aIC₅₀ values were determined for select compounds using the RIA (see Experimental Methods for details); assays were run in duplicate at 0, 1, 3, 10, 30, 50 and 100 µM inhibitor concentration.

^b100 µM of inhibitor was used in these experiments, and the values represent the average and standard deviation of three independent measurements.

Since APE1 comprises ≥95% of the total AP endonuclease activity in mammalian cells, most, if not all, of the AP site incision activity of human whole cell extracts is APE1-dependent.²⁷ Thus, as a means of assessing the specificity of candidate APE1 inhibitors and their potential biological potency, we determined the effect of the most promising actives on total AP site cleavage activity of HeLa whole cell extracts. This experiment was initially done using a single concentration of inhibitor (100 µM) to quickly establish relative activities of the selected analogs. The results showed that several compounds exhibited near 100% inhibition of the incision activity of HeLa extracts (analogs 3, 4, 52 and 55; Table 5). Taking into account the activity observed in the HTS, radiotracer and whole cell extract assays (Table 5), we decided to pursue compounds 52 and 3 further. As such, these two compounds were tested again in the HeLa whole cell extract assay at 5 different concentrations (0, 1, 3, 10, 30 and 100 µM; Figure 2). Notably, each compound was found to inhibit total AP site incision activity with IC₅₀ values comparable to the IC₅₀ obtained with the purified enzyme. In particular, compound 52 showed almost complete inhibition at 10 µM, suggesting selectivity for APE1 even amidst the other non-specific proteins of the extract.

Figure 2.

Inhibition of HeLa whole cell extract AP site incision activity with 3 or 52. HeLa whole cell extract (300 ng) was incubated with 0, 1, 3, 10, 30 or 100 µM of the indicated inhibitor concentration at room temperature for 15 min, prior to the addition of 0.5 pmol radiolabeled AP-DNA substrate. The reaction was then carried out at 37 °C and analyzed as described in the Experimental Methods section. Shown is a bar graph reporting the relative percent incision activity in comparison to the no inhibitor control, arbitrarily set at 100 percent. The values represent the averages and standard deviation of three independent experimental data points.

Mechanism of APE1 inhibition by compounds 3 and 52

Encouraged by the potency of compound 52 in the radiotracer incision and whole cell extract assays, we pursued mechanism of action studies of this compound and our initial lead molecule 3. To reveal the mode of inhibition, we determined the kinetic parameters for APE1 incision in the absence or presence of 5, 10 or 20 µM inhibitor at varying concentrations of DNA substrate. Figure 3 shows the Michaelis-Menten plots, as well as the K_M and k_cat values (determined from Lineweaver – Burk plots), of these experiments. In both cases, the k_cat decreases less than 6-fold, and only at the high inhibitor concentration. In the case of 52, K_M increases substantially (up to 10-fold), while the K_M increases less dramatically with 3 (up to ~4-fold). Such trends would suggest that these compounds act as competitive inhibitors of APE1 activity, and thus, presumably bind the active site of the enzyme.

Figure 3.

Kinetic analysis for compounds 3 and 52. APE1 (~28 pM) was incubated without or with 5, 10 or 20 µM of the indicated inhibitor at room temperature in RIA buffer for 15 min. Varying concentrations of ³²P radiolabeled AP-DNA substrate (i.e. 5, 10, 25, 50, or 100 nM) were then added, and the reactions were performed and analyzed as described in the Experimental Methods section. Shown are the Michaelis-Menten plots of V vs [S], reporting the averages and standard deviations of at least seven data points for each value.

To further explore the potential mechanism of action of the inhibitor compounds, a previously established electrophoretic mobility shift assay (EMSA) was employed to determine the consequence of 3 and 52 on APE1 and ³²P radiolabeled AP-DNA complex formation and stability. The results with both inhibitors which in the initial experiments were pre-incubated with the protein prior to substrate addition, showed that the percentage of APE1-DNA complex (C) decreased in a compound dose-dependent manner (Figure 4). In particular, the binary complex was essentially absent when the protein was pre-treated with 30 µM inhibitor, with compound 52 exhibiting a slightly more reproducible, greater effect. If we pre-incubated either of the inhibitors with the radiolabeled DNA substrate prior to the addition of APE1, we similarly saw a significant reduction in protein-DNA complex formation (Supplementary Figure 1). Not surprisingly, a >30-fold excess of cold (unlabeled) AP-DNA essentially abolished the formation of radiolabeled substrate complex (Supplementary Figure 1). These data suggest that the small molecules bind the same site on APE1 as the DNA substrate (albeit apparently with lesser affinity than AP-DNA itself), thereby acting as competitive inhibitors, although an allosteric effect cannot be ruled out in the absence of high resolution complex structural information. Moreover, it is currently unknown whether these inhibitors affect other activities of APE1, such as its redox function.

Figure 4.

Stability of the APE1-DNA substrate complex in the presence of 52 or 3. (A) Representative EMSA. APE1 (~28 nM) was incubated without inhibitor (“-”) or with the indicated inhibitor concentration (1, 3, 10, 30 or 100 µM) for 10 min on ice. One hundred fmol of abasic DNA substrate (10 nM) was then added, and the binding reaction was incubated on ice for an additional 5 min. Samples were subjected to non-denaturing polyacrylamide gel electrophoresis to separate the APE1-DNA complex (C) from unbound radiolabeled DNA (DNA). Inh = inhibitor. (B) Relative complex formation without (“0”) or with the indicated inhibitor (in µM). Shown is the average and standard deviation of three independent experimental data points, all relative to the APE1 control, without inhibitor.

Cellular effects of APE1 inhibitors 3 and 52

To better assess the biological potency of 3 and 52, we measured the level of genomic AP sites in MMS-treated HeLa cells with or without exposure to these compounds (Figure 5). As one might predict, treatment of cells with the alkylator MMS or either of the APE1 inhibitors alone resulted in a statistically significant increase in the total number of AP sites relative to the DMSO control. More notably, combined treatment of MMS with either of the inhibitors resulted in a greater than additive increase (~3-fold) in AP sites, providing evidence that these compounds do indeed inhibit APE1 catalyzed repair of abasic damage in cells. Next, we assessed the ability of these compounds to potentiate the cytotoxic effects of MMS and TMZ in a dose-response experiment using HeLa cells and monitoring cell viability using a high-throughput luminescence-based detection of the cells’ ATP content, as shown in Figure 6. Both compounds greatly potentiated the activity of these agents with optimal synergy occurring at a concentration between (5–10 µM) of 3 and (10–15 µM) of 52. However, it should be noted that compound 3 appears more cytotoxic against HeLa cells as a single agent than analog 52, with a 50% reduction in cell viability occurring at ~15 µM and ~30 µM, respectively. (see Supplemental Figure 3 for a semi-log plot of this data). Similar potentiation trends were observed when cell viability was monitored through a more traditional method using staining for live cells (Supplemental Figure 2).

Figure 5.

AP site accumulation assay. AP site counts were measured in HeLa cells after treatment with MMS and compounds alone and in combination. To evaluate inhibition of APE1 by selected compounds in cells, AP site formation was measured using the ARP assay in triplicate (see Experimental Methods). The p values were determined for an n using the Student’s t test, where compound 3, 52 and MMS is compared with 3 and 52 alone (** p≤0.001) and MMS alone (* p≤0.01)

Figure 6.

HeLa cell viability assay with compounds 3 or 52 alone, in combination with either MMS (0.4 mM) or TMZ (1 mM). (A) Combined treatment of the HeLa cells with compounds 3 at concentrations shown in the plot in the presence of 0.4 mM MMS and 1 mM TMZ (B) Combined treatment of the HeLa cells with compound 52 in the presence of 0.4 mM MMS and 1 mM TMZ (see Experimental Methods). The viability values are normalized to 100% for each alkylating agent in the absence of compound. The assay was performed in triplicate.

PK properties of 3 and 52

Given that 3 and 52 exhibit a tractable SAR, on-target APE1 inhibition, and potentiation of the toxicity of MMS and TMZ in cell culture experiments, we were eager to assess the pharmacokinetic (PK) properties of these top compounds (Table 6). As shown in the in vitro absorption, distribution, metabolism and excretion (ADME) data, both compound 3 and 52 have many desirable attributes, yet a few liabilities. Specifically, analog 52 exhibits improved kinetic solubility and Caco-2 permeability relative to the original lead 3. Both compounds possess favorable cell permeability and do not appear to be susceptible to active transport as shown by the efflux ratios of ~1. However, compound 52 was found to be rapidly metabolized by mouse liver microsomes (T_1/2 = 7.8 min), whereas compound 3 shows favorable stability (T_1/2 = 80 min). These results suggest that the general core scaffold is metabolically stable, and through additional structural modifications, the potential metabolic liability of 52 could be addressed while maintaining potency. Surprisingly, as described below, we found 52 had a reasonable PK profile despite the apparent metabolic liability. Finally, both compounds exhibited some inhibition of CYP isozymes (CYP2D6 and/or CYP3A4) at 10 µM. Compound 3, which possesses substituted nitrogen (isopropyl group) and is a 6-membered ring rather than 5-membered, shows no inhibition of CYP2D6 at 10 µM. This result suggests that a more expansive look at all the analogs described herein may uncover structure property relationships that obviate CYP inhibition.

Table 6.

In vitro ADME properties of lead molecules (3 and 52).^a

Compound	aq. kinetic sol. (PBS @ pH 7.4)	CYP2D6 (inh. @ 10 µM)	CYP3A4 (inh. @ 10 µM)	Caco-2 (Papp 10−6 m/s @ pH 7.4)	efflux ratio (B→A)/(A→B)	mouse liver microsome stability (T1/2)	PBS-pH 7.4 Stability: % remaining after 48h	Mouse plasma stability (T1/2)
52	51.6 µM	39%	50%	6.8	0.8	7.8 min.	100%	213 min
3	20.4 µM	0%	53%	5	1.1	80 min.	100%	>250 min

^aAll experiments were conducted by Shanghai Chempartner Co. Ltd and are the average of two experiments.

The two lead compounds (52 and 3) were analyzed for their in vivo PK properties (Table 7) via intraperitoneal (IP) administration at 30 mg/kg body weight in 6–8 week old CD1 mice. Both compounds were well tolerated by the animals, with no adverse effects noted after a 24 hour observation. Compound 52, the more hydrophilic analog (CLogP ~ 1), had a favorable plasma t_(1/2) of 5 hours and a drug concentration (ng/mL) that exceeded the IC₅₀ value for over 12 hours. This compound also had reasonable blood brain barrier (BBB) penetration, with a high initial concentration that quickly tailed off, resulting in a brain/plasma (B/P) ratio of 1.4. In contrast, analog 3, which is more lipophilic (CLogP = 2.8), crosses the BBB quite readily, giving rise to a (B/P) ratio of 21. This result correlates with expectations, as reduction of hydrogen bond donors and increased lipophilicity often leads to improved BBB penetration. The ability to modulate the BBB penetration capacity of these molecules through structural modifications could prove useful depending on the cancer one is targeting.

Table 7.

In vivo PK data^a

compound	t1/2 (h) [plasma]	t1/2 (h) [brain]	[Brain/Plasma]b	Cmax (µM) [Plasma]	Cmax (µM) [Brain]	tmax (h) [Plasma]	tmax (h) [Brain]	CLogPc
52	5	2.5	1.35	36	92	0.25	0.25	1.02
3	2.1	1	21	16	217	0.25	0.25	2.83

^aExperiments were conducted at Pharmaron Inc. IP administration (30 mpk) of CD1 male mice (n = 3) was monitored at 8 time points (0.25 h, 0.5, 1, 2, 4, 8, 12, 24). Compound 52 was formulated as 50% PEG 200 and 10% Cremophor EL in saline solution, and compound 3 was formulated as 50% PEG 400 and 10% Cremophor EL in saline suspension.

^bcalculated based on the average [b/p] ratio over 8 time points (24 h period).

^cCLogP values were obtained using ChemDraw Ultra^® 10.0.

Conclusion

While the work described herein did not lead to a tremendous improvement in the potency of the initial “hit” molecule 3, it does provide valuable insights into the SAR profile of this chemotype and represents the first reported medicinal chemistry optimization campaign towards the establishment of a novel APE1 inhibitor. In particular, this effort led to the development of compounds with low single-digit µM potency against the purified enzyme (noticeably more potent than prior reported inhibitors), desirable in vitro and in vivo ADME properties, and the capacity to potentiate the cytotoxicity of relevant DNA-damaging agents, namely MMS and TMZ. On-target evidence was supported by the comparable IC₅₀ values against the purified recombinant APE1 protein and human whole cell extracts, as well as by increased genomic AP site accumulation in HeLa cells treated with inhibitors alone. Analysis of the in vivo PK properties of 52 and 3 in mice revealed that analog 52 has a better general cytotoxicity profile, higher exposure levels and a more favorable t_(1/2) in the plasma, whereas compound 3 crosses the BBB more efficiently. Thus, for tumors outside the brain cavity, a compound like 52, which does not efficiently cross the BBB, would be useful in avoiding potential complications associated with this vital organ. However, APE1 has been found to be overexpressed in adult and pediatric gliomas, with an increase in AP endonuclease activity of between 5 and 10-fold.²⁸ This observation indicates the need for the development of APE1 inhibitors, such as 3 or other lipophilic analogs of 52, that efficiently cross the BBB and can potentially be used in combination with a drug like TMZ. Our current efforts are focused on defining the efficacy of this class of compounds in vivo using mouse xenograft models in combination therapy with TMZ and other relevant DNA-damaging cancer chemotherapeutics.

Experimental Methods

General Chemistry

Unless otherwise stated, all reactions were carried out under an atmosphere of dry argon or nitrogen in dried glassware. Indicated reaction temperatures refer to those of the reaction bath, while room temperature (rt) is noted as 25 °C. All solvents were of anhydrous quality purchased from Aldrich Chemical Co. and used as received. Commercially available starting materials and reagents were purchased from Aldrich and were used as received.

Analytical thin layer chromatography (TLC) was performed with Sigma Aldrich TLC plates (5 × 20 cm, 60 Å, 250 µm). Visualization was accomplished by irradiation under a 254 nm UV lamp. Chromatography on silica gel was performed using forced flow (liquid) of the indicated solvent system on Biotage KP-Sil pre-packed cartridges and using the Biotage SP-1 automated chromatography system. ¹H- and ¹³C NMR spectra were recorded on a Varian Inova 400 MHz spectrometer. Chemical shifts are reported in ppm with the solvent resonance as the internal standard (CDCl₃ 7.26 ppm, 77.00 ppm, DMSO-d₆ 2.49 ppm, 39.51 ppm for ¹H, ¹³C respectively). Data are reported as follows: chemical shift, multiplicity (s = singlet, d = doublet, t = triplet, q = quartet, br = broad, m = multiplet), coupling constants, and number of protons. Low resolution mass spectra (electrospray ionization) were acquired on an Agilent Technologies 6130 quadrupole spectrometer coupled to the HPLC system. High resolution mass spectral data was collected in-house using and Agilent 6210 time-of-flight mass spectrometer, also coupled to an Agilent Technologies 1200 series HPLC system. If needed, products were purified via a Waters semi-preparative HPLC equipped with a Phenomenex Luna^® C18 reverse phase (5 micron, 30×75 mm) column having a flow rate of 45 mL/min. The mobile phase was a mixture of acetonitrile (0.025% TFA) and H₂O (0.05% TFA), and the temperature was maintained at 50 °C.

Samples were analyzed for purity on an Agilent 1200 series LC/MS equipped with a Luna® C18 reverse phase (3 micron, 3×75 mm) column having a flow rate of 0.8–1.0 mL/min over a 7-minute gradient and a 8.5 minute run time. Purity of final compounds was determined to be >95%, using a 3 µL injection with quantitation by AUC at 220 nm and 254 nm (Agilent Diode Array Detector).

RIA

Recombinant wild type APE1 protein was purified as previously described.²⁹ Fifty pg of APE1 (~140 pM) was incubated without (positive control containing 1% DMSO) or with the indicated inhibitor at the indicated concentration at room temperature in RIA buffer (50 mM Tris pH 7.5, 25 mM NaCl, 1 mM MgCl₂, 1 mM DTT, 0.01% Tween-20) for 15 min. One-half pmol of ³²P 5’-radiolabeled AP-DNA substrate (18 mer) was added to a 10 µL final volume,³⁰ and the reactions were incubated at 37 °C for 5 min, and stopped by adding stop buffer (0.05% bromophenol blue/xylene cyanol dissolved in 95% formamide, 20 mM EDTA) and heating at 95 °C for 10 min. Intact substrate was separated from incised product on a 15% polyacrylamide denaturing gel in a tris/boric acid/EDTA buffer. Following electrophoresis, the gel was subjected to standard phosphoimager analysis using the ImageQuant 5.2 software, and the percent incision activity (amount of substrate converted to product) was calculated. IC₅₀ values (i.e. the concentration of inhibitor at which 50% inactivation was observed) were extrapolated after plotting the results of duplicate incision sets (at inhibitor concentrations ranging from 1 µM to 100 µM) and fitting data using the Hill equation in GraphPad Prism version 4.0 software.

HeLa Whole Cell Extract Incision Assays

To prepare protein extracts, HeLa cells maintained in DMEM with 10% fetal bovine serum and 1% penicillin-streptomycin were harvested, washed with 1X PBS, and re-suspended in ice cold hypotonic lysis buffer (50 mM Tris pH 7.4, 1 mM EDTA, 1 mM DTT, 10% glycerol, 0.5 mM PMSF). The suspension was frozen at −80 °C for at least 30 min and then slowly thawed at 4 °C for ~1 h. KCl was added to the cell suspension to a final concentration of 222 mM, followed by incubation on ice for 30 min and clarification by centrifugation at 12,000 xg for 15 min at 4 °C. The supernatant (whole cell extract) was retained, the protein concentration determined using the Bio-Rad Bradford reagent, and aliquots were stored until needed at −80 °C. For the incision assays, 300 ng of HeLa whole cell extract was incubated with 0, 1, 3, 10, 30 or 100 µM of the indicated inhibitor at room temperature for 15 min prior to the addition of 0.5 pmol of ³²P radiolabeled AP-DNA substrate (final volume of 10 µL). The reaction mix was then transferred to 37 µC for 5 min to allow for incision. Following addition of stop buffer and heat denaturation, the reaction products were analyzed as above.

Enzyme Kinetic Studies

Ten pg of APE1 (~28 pM) was incubated without (positive control, 1% DMSO) or with 5, 10 or 20 µM of the indicated inhibitor at room temperature in RIA buffer (see above) for 15 min. Varying concentrations of ³²P radiolabeled AP-DNA substrate (i.e. 5, 10, 25, 50, or 100 nM) were then added to a 10 µL final volume, and the reactions were incubated at 37 °C for 5 min, and stopped by adding stop buffer and heating at 95 °C for 10 min. The reaction velocity (nmolar substrate incised per min) at each substrate concentration was calculated as described above. Lineweaver – Burk plots of 1/V versus 1/[S] were used to determine K_M and k_cat, and the mode of inhibition.

EMSA

Ten ng of APE1 (~28 nM) was incubated without inhibitor (positive control, 1% DMSO) or with increasing concentrations of inhibitor (1, 3, 10, 30 and 100 µM) in binding buffer (50 mM Tris pH 7.5, 25 mM NaCl, 1 mM EDTA, 1 mM DTT, 10% glycerol, 0.01% Tween 20) for 10 min on ice, and then radiolabeled ³²P AP-DNA substrate (100 fmol) was added to a 10 µL final volume. Following incubation on ice for 5 min, samples were subjected to non-denaturing polyacrylamide gel electrophoresis (20 mM Tris pH 7.5, 10 mM sodium acetate, 0.5 mM EDTA, 8% polyacrylamide, 2.5% glycerol) for 2 h at 120 V in electrophoresis buffer (20 mM Tris pH 7.5, 10 mM sodium acetate, 0.5 mM EDTA) to separate the APE1-DNA complex from unbound radiolabeled DNA.³¹ After electrophoresis, the gel was subjected to standard phosphoimager analysis as above, and the percentage of substrate DNA in complex with APE1 was determined.

Genomic AP site accumulation in cells

HeLa cells with 80% confluency in a 25 cm² flask were treated with DMSO, 275 µM MMS, or 7.5 µM APE1 inhibitor alone, or with a combination of 275 µM MMS and 7.5 µM inhibitor for 24 h at 37 °C. Cells were then harvested, and genomic DNA of each sample was isolated according to Qiagen Genomic DNA isolation kit. The concentration of genomic DNA was measured and adjusted to 100 ng/µL. Ten µL of purified DNA was further labeled with an Aldehyde Reactive Probe (ARP) reagent (N’ aminooxymethylcarbonylhydrazino-D-biotin), and AP sites were measured using the DNA Damage Quantification kit from Dojindo Molecular Technologies.

MMS and TMZ potentiation assay

HeLa cells were plated by multichannel pipette or Multidrop Combi dispenser (Thermo) at 6K/ 25 µL/well in DMEM culture medium with 10% FBS into white solid bottom 384-well cell culture plates. Cells were cultured at 37 µC overnight to allow for cell attachment. The following day, the entire cell medium in the well was replaced with fresh medium containing serial dilutions of the compounds of interest (5–30 µM) in the presence or absence of MMS (0.4 mM) or TMZ (1 mM). The plates were incubated for 24 h at 37 µC. Cell viability was then evaluated via luminescence detection by adding 15 µL of CellTiter Glo reagent (Promega, Madison, WI) to each well and incubating at room temperature for 30 min, and subsequently measuring the luminescence using a ViewLux reader. Percent viability was calculated for each concentration of the tested compounds in duplicate relative to the luminescence of the negative DMSO control.

In vivo PK analysis

Compound 3 was dissolved in PEG 400 and cremophor with vortexing and sonification, then saline was gradually added with vortexing and sonification to obtain a final concentration of 3 mg/mL of 3 in 50% PEG 400 and 10% cremophor. Compound 52 was dissolved in PEG 200 cremophor with vortexing and sonification, then saline was gradually added as above to obtain a final concentration of 3 mg/mL of 52 in 50% PEG 200 and 10% cremophor. The dose for both compounds was administered via IP. All blood samples were collected through a cardiac puncture per sampling time point (0.25, 0.5 h, 1 h, 2 h, 4 h, 8 h, 12 h, and 24 h post dose). Approximately 0.12 mL blood was collected at each time point. All blood samples were transferred into plastic micro centrifuge tubes containing Heparin and placed at −80 °C until processed (see below). At each time point (see above), the brain was harvested immediately after euthanasia by carbon dioxide. The brain was rinsed with saline and wiped clean and then weighed in a sterilized plastic tube. The tissue sample was then homogenized in water with a brain weight (g):water (mL) ratio of 1:4 (g:mL). The detected values were then multiplied by 5 to achieve the final concentration of the compound in the brain. Blood samples were processed for plasma by centrifugation at 4 °C at 4000 g for 5 min. Plasma samples were then stored in tubes, quickly frozen in a freezer and kept at −80 °C until LC/MS/MS analysis. Plasma concentration of compound 3 or 52 at the various time points (data obtained from the LC/MS/MS studies) was analyzed using the WinNonlin software program.

Abbreviations

APE1

apurinic/apyrimidinic endonuclease 1

HTS

high-throughput screening

qHTS

quantitative HTS

MLSMR

Molecular Libraries Small Molecule Repository

BER

base excision repair

ML199

N-(3-(benzo[d]thiazol-2-yl)-5,6-dihydro-4H-thieno[2,3-c]pyrrol-2-yl)acetamide

TMZ

temozolomide

RIA

radiotracer incision assay

ADME

absorption, distribution, metabolism and excretion

dppp

1,3-Bis(diphenylphosphino)propane

Xantphos

4,5-Bis(diphenylphosphino)-9,9-dimethylxantene

dba

dibenzylideneacetone

EDC

1-ethyl-3-(3-dimethylaminopropyl)carbodiimide

HATU

2-(1H-7-Azabenzotriazol-1-yl)-1,1,3,3-tetramethyluronoium hexafluorphosphate

MMS

methylmethane sulfonate