Novel Imidazotetrazine Evades Known Resistance Mechanisms and Is Effective against Temozolomide-Resistant Brain Cancer in Cell Culture

Abstract

Glioblastoma (GBM) is the most lethal primary brain tumor. Currently, frontline treatment for primary GBM includes the DNA-methylating drug temozolomide (TMZ, of the imidazotetrazine class), while the optimal treatment for recurrent GBM remains under investigation. Despite its widespread use, a majority of GBM patients do not respond to TMZ therapy; expression of the O⁶-methylguanine DNA methyltransferase (MGMT) enzyme and loss of mismatch repair (MMR) function as the principal clinical modes of resistance to TMZ. Here, we describe a novel imidazotetrazine designed to evade resistance by MGMT while retaining suitable hydrolytic stability, allowing for effective prodrug activation and biodistribution. This dual-substituted compound, called CPZ, exhibits activity against cancer cells irrespective of MGMT expression and MMR status. CPZ has greater blood–brain barrier penetrance and comparable hematological toxicity relative to TMZ, while also matching its maximum tolerated dose in mice when dosed once-per-day over five days. The activity of CPZ is independent of the two principal mechanisms suppressing the effectiveness of TMZ, making it a promising new candidate for the treatment of GBM, especially those that are TMZ-resistant.

Graphical Abstract

INTRODUCTION

The prognosis of patients with glioblastoma (GBM), the most prevalent and malignant type of primary brain tumor, is extremely poor. GBM is incurable, and without treatment, median survival of these patients is less than three months.¹ The standard treatment regimen for GBM patients consists of bulk surgical resection and radiotherapy (RT) with concomitant and adjuvant temozolomide (TMZ), a small molecule DNA alkylating agent.² TMZ is an imidazotetrazine prodrug that undergoes hydrolytic ring opening under physiological conditions (t_1/2 ~ 1.5–2 h in human patients),^3,4 releasing a methyl diazonium ion, the active DNA alkylating species (Figure 1A) in addition to 5-aminoimidazole-4-carboxamide (AIC). The cytotoxicity of the methyl diazonium ion is derived from methylation of the O⁶ position of guanine residues, which triggers futile cycles of mismatch repair (MMR)-mediated DNA repair and subsequent DNA double strand breaks.^5–7 The addition of TMZ to RT improves the overall patient survival to 15 months (up to an 8 month survival benefit vs RT alone),⁸ which has cemented its use in the clinic since it gained frontline approval in 2005 for GBM patients.

Figure 1.

Predominant clinical resistance mechanisms to TMZ and overview of the strategy to develop MGMT and MMR-independent imidazotetrazines. (A) TMZ hydrolyzes spontaneously and releases a methyl diazonium ion, which can alkylate DNA. (B) When expressed, MGMT confers resistance to TMZ by directly repairing O⁶-methylguanine lesions. Additionally, in the absence of a functional MMR system, tumors can tolerate O⁶-methylguanine lesions. Thus, only GBM patients possessing tumors that are MGMT-negative and MMR-proficient derive a survival benefit from TMZ. (C) An imidazotetrazine modified at the N3 position could hydrolyze and release a diazonium ion that installs adduct “R” on DNA that is irremovable by MGMT and does not require recognition by the MMR system to elicit cell death.

It is well-known that certain GBM patient populations do not respond to TMZ therapy.⁹ Expression of O⁶-methylguanine DNA methyltransferase (MGMT), an alkyltransferase responsible for the direct repair of O⁶-methylguanine DNA adducts, serves as the primary source of intrinsic resistance to TMZ. Silencing of the MGMT gene through promoter methylation, a common epigenetic phenomenon in gliomas, is strongly correlated with a therapeutic response to TMZ and an increased survival rate,¹⁰ but 55–60% of GBMs have MGMT unmethylated promoters.^8,11,12 Patients bearing these MGMT-positive tumors have median survival times 11 months less than those with MGMT-negative tumors when treated with TMZ.⁸ The MGMT promoter methylation status is routinely used as a biomarker for GBM patients but is not implemented to stratify therapeutic decision-making; patients with both MGMT methylated and unmethylated promoters are treated with TMZ, even though there is virtually no benefit in MGMT-unmethylated patients.¹³ The other major clinical mode of resistance to TMZ is loss of MMR function, as proper MMR is required to trigger the futile processing of O⁶-methylguanine lesions. A defective MMR system leads to tumoral tolerance of the O⁶-methylguanine adducts generated by TMZ, global G:C → A:T transition mutations, and more malignant hypermutated tumors.^14–17 Approximately 25% of recurrent GBM tumors harbor inactivating mutations in the MMR pathway (namely, in the MSH6, MSH2, MLH1, or PMS2 genes).¹⁸ Distressingly, complete inactivation of MMR is not required to bestow resistance to TMZ, and even a 15% reduction in MMR protein expression can result in insensitivity to the drug.¹⁹ The frequency of these two resistance mechanisms (MGMT expression and loss of MMR function) precludes a majority of GBM patients, both newly diagnosed and recurrent, from deriving any benefit from TMZ (Figure 1B).

MGMT is a suicide enzyme that loses its enzymatic activity upon methylation and is subsequently targeted for degradation by the proteasome.²⁰ Therefore, several strategies to deplete tumoral MGMT and thus regain sensitivity to TMZ have been explored. One approach is to alter the dosing regimen of TMZ to achieve continuous drug exposure and maintain alkylated (inactive) MGMT to perpetuate chemotherapy sensitization. Prolonged inhibition of MGMT expression or activity has been measured in peripheral blood mononuclear cells taken from patients treated with dose-dense TMZ (e.g. 21/28 days vs 5/28 days),²¹ but ensuing clinical trials have reported no differences in efficacy when compared to the standard 5/28 day protocol.^22,23 Another approach is to co-administer TMZ together with an MGMT inhibitor. The most widely used MGMT inhibitor is pseudosubstrate O⁶-benzylguanine (O⁶BG, reviewed by Rabik et al.),²⁴ which potently inhibits MGMT activity. Despite potentiation of the cytotoxicity of TMZ in MGMT-expressing cell lines and murine tumor models,^25–31 no clinical benefit has been observed from this combination^32,33 or from the combination of TMZ with other MGMT inhibitors.^34,35

Instead of attempting to deplete tumoral MGMT, a distinct strategy would be to create an imidazotetrazine analogue that delivers a moiety other than methyl, which would be irremovable by MGMT (Figure 1C). These “MGMT-independent” compounds would ideally retain the favorable properties of TMZ [aqueous prodrug activation, blood–brain barrier (BBB) permeability, etc.] and serve as a treatment option for all GBM patients irrespective of the MGMT status. Previous attempts to employ this strategy using imidazotetrazines with a variety of alkyl groups at the N3 position presume transfer of the alkyl group to DNA;^36–40 however, obtaining an N3-substituted analogue that retains activity in MGMT-positive cell lines and demonstrates proper hydrolytic stability has proven elusive, so this strategy remains untested in vivo.

One possible obstacle for an MGMT-independent imidazotetrazine is the potential for enhanced in vivo toxicity, as MGMT functions as a systemic protectant against the cytotoxic and mutagenic effects of O⁶-methylguanine and therefore mitigates the toxic effects of TMZ in non-malignant tissues. As a result, when TMZ was co-administered with O6BG in the clinic, the cumulative TMZ dose had to be reduced by 50–75% due to exacerbated myelotoxicity.^41,42 These dose reductions have been blamed for the disappointing clinical results of TMZ/MGMT inhibitor combinations. Herein, we report the discovery of a novel imidazotetrazine, the compound CPZ, that evades both MGMT and MMR-mediated resistance. To minimize on-target but off-tumor myelosuppressive effects, CPZ was designed to have enhanced localization to the central nervous system (CNS). As reported herein, CPZ not only outperforms TMZ in culture with cell lines expressing MGMT and lacking functional MMR, but also shows an increase in BBB penetrance and noninferior hematological toxicity relative to TMZ. CPZ represents a new therapeutic alternative to TMZ, particularly for the large GBM patient population (primary and recurrent) that possesses MGMT-positive and/or MMR-deficient tumors. The compounds detailed herein can now be used to directly test the hypothesis that imidazotetrazines delivering alternative alkyl groups can be effective against TMZ-resistant brain tumors in vivo.

RESULTS

Design and Evaluation of N3-Substituted Imidazotetrazines.

The substituents known to be removable from the O⁶ position of guanine by MGMT extend beyond methyl and include ethyl, chloroethyl, and benzyl.⁴³ As such, a panel of imidazotetrazines was synthesized containing diverse functionality at the N3 position beyond simple alkyl groups (Figure 2A). To identify candidate compounds exhibiting anticancer activity independent of MGMT expression, the U87 (no MGMT expression) and T98G (high MGMT expression) GBM cell lines were chosen for an initial screen. TMZ has an IC₅₀ value of 50 μM in U87 cells but exhibits >10-fold loss of activity in the T98G cell line (IC₅₀ > 550 μM) (Figure 2A), attributable to MGMT-mediated resistance.

Figure 2.

Selection and evaluation of MGMT-independent imidazotetrazines. (A) Seven-day IC₅₀ values (μM) of N3-substituted imidazotetrazines in the U87 and T98G GBM cell lines. Error is standard error of the mean (SEM), n ≥ 3. (B) Examples of N3-substituted imidazotetrazines that demonstrate MGMT-independent anticancer activity and their relative BBB penetrance predicted using clogBB values and CNS MPO scores. Molecular properties were calculated using ChemDraw or the Chemicalize platform (ChemAxon). (C) IC₅₀ values (μM, 7 days) of TMZ and compound 10 in an expanded panel of GBM cell lines. Cell viability was assessed via the Alamar Blue assay, n ≥ 3, error is SEM. (D) Hydrolytic stability of TMZ at pH 7, 7.4, and 8. Percentage of prodrug remaining was quantified by HPLC, n ≥ 2. (E) Hydrolytic stability of compound 10 at pH 7, 7.4, and 8. Percentage of prodrug remaining was quantified by HPLC, n ≥ 2. (F) Half-lives of TMZ and 10 calculated from the hydrolytic stability data.

As a control, imidazotetrazine 1, possessing no alkylating moiety at N3,⁴⁴ was constructed and was unable to induce cell death (Figure 2A). Consistent with the literature,⁴⁵ ethyl variant 2 demonstrated differential activity in the presence and absence of MGMT but was much less active than TMZ, presumably due to the competing β-elimination pathway possible for the ethyl but not methyl diazonium ion; strong evidence for this hypothesis comes from NMR studies that suggest the formation of ethene from 2 in aqueous media.⁴⁶ Reduced activity was also observed for higher alkyl analogue 3. First-generation imidazotetrazine mitozolomide (MTZ) possessed robust cytotoxicity against U87 and T98G cells but is a known DNA cross-linker.^47,48 Sterically demanding isopropyl (4), tert-butyl (5), and neopentyl (6) groups were incorporated to potentially disfavor the S_N2-mediated dealkylation pathway utilized by MGMT.⁴⁹ Each of these compounds, however, were found to be inactive in cells (Figure 2A), perhaps due to a challenging DNA alkylation event from bulky diazonium ions. Imidazotetrazines 7 and 8 were designed with electron-rich arenes at the N3 position. Presumably, these compounds are precursors to arene diazonium ions, which could generate DNA species for which S_N2 chemistry is not feasible. Interestingly, both 7 and 8 exhibit a desirable cytotoxicity profile as each inhibits GBM cell growth with IC₅₀ values <100 μM in both cell lines (Figure 2A), suggestive of MGMT-independent activity. Notably, arene diazoniums exhibit a wide array of aryl transfer activity in biological systems including direct protein labeling of electron-rich side chains, particularly tyrosine and cysteine;^50–52 this phenomenon, coupled with their significantly higher clogP compared to TMZ (Figure 2B), prompted us to eliminate these derivatives from further advancement. Allyl derivative 9 is not exceptionally active, but the installation of a propargyl group, employed frequently for labeling studies in biological systems due to its biorthogonality,^53–55 imbued 10 with a promising profile (Figure 2A); activity in the presence of MGMT has been previously observed for compound 10.⁴⁰ Embedded ether-containing imidazotetrazine 11 also exhibits a MGMT-independent phenotype with potent IC₅₀ values <40 μM in each cell line, but a similar compound with an embedded amide (12) does not provide the same degree of activity (Figure 2A).

N3-propargyl compound 10 was selected for further development because of its small size and comparable clogP to TMZ (Figure 2B). Compound 10 also possesses equivalent clogBB and CNS MPO scores to TMZ, suggestive of analogous BBB permeability (Figure 2B). The anticancer activity of 10 was assessed against an expanded panel of GBM cell lines with or without MGMT expression (Figure 2C). Cellular resistance conferred by MGMT is apparent for TMZ as there is >6-fold difference in IC₅₀ values between MGMT-negative and positive cell lines. Conversely, no such differential profile is observed for compound 10 as all IC₅₀ values are <100 μM, suggesting that compound cytotoxicity is not affected by MGMT expression. Despite promising activity against GBM cells in culture, 10 is known to have limited efficacy in an intracranial efficacy model.⁵⁶ This discrepancy prompted a closer examination of the most critical feature of the imidazotetrazine class of molecules, hydrolytic stability. TMZ spontaneously hydrolyzes in aqueous solution with a half-life of ~1.5–2 h in vitro (PBS, pH 7.4, 37 °C)⁵⁷ and in vivo (human pharmacokinetics).^3,4 An assay was recently reported that permits the comparison of C8-substituted imidazotetrazines by quantifying the percentage of prodrug remaining after a 2 h timepoint.⁵⁸ As expected, the hydrolytic stability of TMZ was strongly dependent on pH, and TMZ has a stability of 40% remaining after 2 h at pH 7.4 (Figure 2D). Compared to TMZ, compound 10 exhibits a much faster (but still pH-dependent) rate of hydrolysis with only 8% remaining after 2 h at pH 7.4 (Figure 2E). This finding was reinforced when the half-life of 10 measured in PBS (pH 7.4, 37 °C) was found to be ~30 min compared to ~90 min for TMZ (Figure 2F). Thus, the observed inability of 10 to elicit an anticancer effect and prolong survival in an intracranial murine model of GBM may be attributed to rapid compound hydrolysis prior to reaching the brain and alkylating tumoral DNA.

Tuning Hydrolytic Stability and the Identification of CPZ.

Fortunately, a recent report revealed that the hydrolytic stability of imidazotetrazines can be tuned by modifying the substituent at the C8 position, and these modifications do not disturb the alkylating capacity of the resultant prodrugs but may enhance drug pharmacokinetics and/or BBB penetrance.⁵⁸ According to the trend derived between the electronics of the C8 substituent (measured by the σ_p value) and prodrug stability, replacing the primary amide at the C8 position (of either TMZ or 10) with a less electron-withdrawing substituent than the primary amide (σ_p = 0.36), should confer greater aqueous stability; indeed, replacement of the amide of TMZ with a variety of functional groups imparted greater hydrolytic stability due to this electronic effect, as seen with the secondary amide (13, σ_p = 0.36), tertiary amide (14, σ_p = 0.28), and chloro (15, σ_p = 0.23) derivatives (Figure 3A).⁵⁸

Figure 3.

Design and evaluation of the dual-substituted imidazotetrazine CPZ. (A) Strategy used to generate a hydrolytically-stable, MGMT-independent imidazotetrazine. Tuning the stability of compound 10 via the C8 position led to novel dual-substituted compounds, including CPZ. Stability values represent the percentage of parent prodrug remaining after 2 h incubation in PBS (pH 7.4, 37 °C). (B) 7 day IC₅₀ values (μM) of compound 10 and CPZ in an expanded panel of GBM cell lines. Cell viability was assessed via the Alamar Blue assay, n ≥ 3, error is SEM. Data for compound 10 is the same as shown in Figure 2C. (C) 7 day dose–response curves of TMZ and CPZ in cell lines with variable MGMT and MMR status. Error is SEM, n ≥ 3. (D) Killing kinetics of CPZ compared to TMZ in the U87 cell line. Cells were treated with compound once and viability was assessed every 24 h via Alamar Blue assay. Error is SEM, n ≥ 2.

As shown in Figure 3A, the stabilizing substitutions at C8 were combined with the MGMT-evading substituent at N3 to give novel dual-substituted imidazotetrazines. All three compounds (16, 17, 18, Figure 3A) appear very promising; chloro-substituted compound 18 (dubbed CPZ) was prioritized for further assessment due to its low molecular weight and the stability of the C8-chloro to oxidative metabolism. Tuning the hydrolytic stability with a C8-chloro modification significantly improves the 2-h stability from 8% (for 10) to 30% (for CPZ) at pH 7.4, 37 °C (Figure 3A). The hydrolytic stability remained pH-dependent and CPZ demonstrated a half-life of 72 min at pH 7.4 (Figure S1). The other two dual-substituted compounds, 16 and 17, also demonstrated an improved two-hour stability (18 and 27%, respectively, at pH7.4, 37 °C). All three derivatives demonstrated anticancer activity in both MGMT− and MGMT+ cell lines, with CPZ being the most potent (Figure 3A).

Mode of Action Studies of CPZ.

CPZ exhibits superior anticancer efficacy in each cell line tested compared to its precursor 10, and most importantly, it retains activity in MGMT-expressing cell lines with an average IC₅₀ value of 16 μM in MGMT (−) cell lines versus 22 μM in MGMT (+) cell lines (Figure 3B). Beyond MGMT, loss of a functional MMR system is another primary resistance mechanism to TMZ, both in cell culture and in human GBM patients.^{5,12,18,19,27,59–61} Therefore, GBM tumors that express MGMT and/or have MMR deficiencies suppress the activity of TMZ. These clinical scenarios were recapitulated in culture by employing cell lines with different MGMT/MMR status; when possible, cell lines derived from human brain tumors were used (Figure 3C). As stylized in Figure 1B and shown in Figure 3C, the MGMT(−)/MMR(+) U87 cell line was responsive to TMZ (IC₅₀ ~ 50 μM), while the MGMT(+)/MMR(+) T98G cell line was not (IC₅₀ > 550 μM), consistent with clinical results; CPZ was effective against both of these cell lines (Figure 3C).

To assess the necessity of the MMR system, the activity of TMZ and CPZ was evaluated against MGMT(−)/MMR(−) HCT116 cells, which have a mutated MLH1 gene. As expected, these cells were insensitive to TMZ with an IC₅₀ value > 800 μM (Figure 3C). Importantly, CPZ was still able to potently elicit cell death with an IC₅₀ value of 14 μM. Perhaps most striking, CPZ has efficacy in two different MGMT(+)/MMR(−) cell lines. The D341 Med medullo-blastoma cell line is highly resistant to TMZ (IC₅₀ = 460 μM), while CPZ is able to induce cell death with low micromolar potency (IC₅₀ = 8 μM, Figure 3C). Similarly, the RKO colon cancer cell line, also MGMT(+)/MMR(−) and exceptionally resistant to TMZ (IC₅₀ > 1000 μM), is potently killed by CPZ (IC₅₀ = 10 μM, Figure S2). The convincing cell death elicited by CPZ in these cell lines suggests complete avoidance of the two primary resistance mechanisms to TMZ.

The striking anticancer activity even in an MMR-deficient cell line indicates that CPZ may be operating, at least in part, through a unique mode of cell death. Indeed, when the timing of cell death was studied, CPZ was close to reaching its peak threshold of cell death within 72 h in several cancer cell lines (Figures 3D and S3). In contrast, TMZ (at 100 μM) takes between 5 and 10 days to induce cell death in culture (Figure 3D), known to require >2 futile cycles of DNA replication before apoptosis is initiated.⁷

Importantly, because CPZ was designed to release a different active species than TMZ, the alkylation pattern in a cellular context is likely different. Therefore, to assess if alkylation still occurs with CPZ, a DNA alkylation assay was performed. GL261 cells were treated with 100, 300, or 500 μM TMZ or CPZ for 8 h after which the cells were harvested and genomic DNA was extracted and subsequently hydrolyzed to constituent deoxyribonucleosides. LC–MS/MS analysis was employed to quantify the amount of O⁶-methyl-2′-deoxyguanosine (O⁶-Me-dG) or O⁶-propargylated-2′-O⁶-deoxyguanosine (O⁶-Prop-dG) in each sample. A clear dose-dependent increase in the concentration of O⁶-Me-dG was observed in the GL261 cell line after treatment with TMZ. Strikingly, there was no detection of O⁶-Prop-dG adducts for CPZ (Figure 4A). To evaluate if this result was unique to GL261, O⁶-dG alkylation adducts of TMZ and CPZ were quantified in two other MGMT-negative GBM cell lines, A172 and D54. Consistent with the results observed in GL261 cells, TMZ-treated cells showed a dose-dependent increase in methyl adducts whereas CPZ-treated cells did not harbor measurable levels of the propargyl adduct (Figure S4). The undetectable concentrations of O⁶-Prop-dG adducts versus the clear detection and dose-dependent increase of O⁶-Me-dG adducts suggests a different alkylation pattern for CPZ or an alternate fate for the propargyl active species.

Figure 4.

Investigation of the mechanism of action and efficacy of CPZ. (A) Detection of O⁶-methyl-2′-deoxyguanosine or O⁶-propargyl-2′-deoxyguanosine in GL261 cells after treatment with TMZ or CPZ, respectively. GL261 cells were treated with the indicated concentration of compound for 8 h before they were harvested and genomic DNA was extracted. DNA (10 μg) was hydrolyzed and submitted to LC–MS/MS for quantitation. Both O⁶-Me-dG and O⁶-proparyl-dG were below the limit of detection in the DMSO control and CPZ samples. Error is SEM, n ≥ 3.(B) Cytotoxicity of TMZ (left) or CPZ (right) in T98G cells pretreated (3 h) with MGMT inhibitor O6BG. Error is SEM, n ≥ 3. (C) Western blot of GL261, GL261 MGMT+, and T98G (10 μg protein loading, 1:1000 anti-MGMT antibody). (D) 7 day dose–response curves of TMZ and CPZ in GL261 cells. Error is SEM, n ≥ 3. (E) 7 day dose–response curves of TMZ and CPZ in GL261 MGMT+ cells. Error is SEM, n ≥ 3.

To probe if CPZ releases the propargyl diazonium ion, CPZ was hydrolyzed under various alkaline conditions in an attempt to recover the corresponding 4-chloro-5-aminoimidazole byproduct (Figure S5A). Recovery of this species would permit the reasonable assumption that the putative propargyl diazonium ion was released. However, attempts to identify 4-chloro-5-aminoimidazole were not fruitful, likely a consequence of its instability, unsurprising given the absence of this compound in the literature. Instead, when 10, which bears the same propargyl N3 group, was hydrolyzed under alkaline aqueous conditions, the presence of AIC was detected via LC/MS and aligned with both a synthetic standard and AIC recovered from degraded TMZ (Figure S5B,C). This result suggests that propargyl-bearing imidazotetrazines can hydrolyze in a manner akin to TMZ.

To further explore the ability of CPZ to evade MGMT-mediated resistance, MGMT-expressing T98G cells were pretreated with MGMT inhibitor O6BG followed by treatment with TMZ or CPZ. A dramatic potentiation of cell death was observed for TMZ relative to non-pretreated cells confirming that MGMT is responsible for the ineffectiveness of TMZ in this cell line (Figure 4B). No such enhancement in activity was observed for CPZ (Figure 4B), suggesting that MGMT expression does not confer resistance to CPZ in these cells. This cell death potentiation by TMZ upon treatment with an MGMT inhibitor was also observed in MGMT-expressing cell lines U118MG and RKO, whereas the efficacy of CPZ was again unaffected (Figure S6). For MGMT(−)/MMR(−) HCT116 cells, negligible potentiation of TMZ cytotoxicity was observed upon pretreatment with O6BG (Figure S7), confirming that loss of MMR capacity, not MGMT expression, was responsible for resistance to TMZ in HCT116 cells. Additionally, human MGMT was knocked into the MGMT-negative mouse glioma cell line, GL261, establishing the novel cell line GL261 MGMT+ (Figure 4C). The IC₅₀ values for TMZ and CPZ in the parental cell line are 120 and 28 μM, respectively (Figure 4D). The isogenic knock-in abolished all TMZ activity (IC₅₀ > 1000 μM), whereas CPZ retains its efficacy (IC₅₀ = 29 μM) (Figure 4E). Similarly, pre-treatment of the knock-in cells with MGMT inhibitor lomeguatrib shows potentiation of cell death for TMZ but not for CPZ (Figure S8).

Next, we measured the mutagenicity of CPZ versus TMZ. TMZ and other DNA methylating chemotherapies are known mutagens,^62,63 inducing G:C → A:T mutations due to erroneous recognition of O⁶-methylguanine. This was reflected by a positive result in the Ames test; in the Salmonella typhimurium TA100 strain, 35/96 colonies were mutated 5 days after treatment with 30 μM of TMZ whereas no revertant colonies were observed when treated with the same concentration of CPZ (Figure S9). CPZ and TMZ (at 30 μM) were also assessed in the Escherichia coli WP2 uvrA pKM101 strain, which is considered one of the more sensitive Ames tester strains available.^64,65 Indeed, TMZ was found to be highly mutagenic in this strain, with 95/96 revertant colonies (Figure S10). In contrast, CPZ again showed no revertant colonies, indicating that at this concentration and in this strain, CPZ is non-mutagenic (Figure S10).

To further probe the molecular basis of CPZ-induced cell death, CPZ was evaluated in a biochemical DNA alkylation assay that has been used to elucidate the mechanism of DNA damaging small molecules.⁶⁶ In the experiment, purified linearized DNA was incubated with compound for 15 h at 37 °C. The DNA was subsequently denatured and the sample was eluted on a 1% agarose gel. Interstrand DNA cross-linkers like cisplatin (Figure 5A) prevent the full denaturation of double-stranded DNA, while DNA alkylating agents like MMS (Figure 5A) or TMZ cause DNA streaking due to shorter DNA fragments that are formed upon alkaline denaturation of abundantly alkylated DNA (Figure 5B). Interestingly, CPZ exhibited clear evidence of both DNA cross-linking and alkylation (Figure 5C). The cross-linking and alkylation are dose-dependent (Figure 5D,E) and occur quickly, observable within minutes after dosing (Figure S11). To probe the effect of the alkylating substituent on DNA alkylation and cross-linking in this assay, 3, 9, and 2-butyne imidazotetrazine 19 (Figure 5A) were also evaluated. No evidence of DNA alkylation was observed for 3 or 9, perhaps due to more unstable diazonium ion species (Figure 5F). Trace DNA alkylation was observed after treatment with 500 μM of 19, but evidence of cross-linking was definitively absent suggesting cross-linking with CPZ was a product of S_N2′ reactivity via attack at the terminal carbon of the alkyne. Finally, to ensure that the cross-linking phenotype exhibited by CPZ was not a product of the assay conditions, unbuffered water (pH 7) and less alkaline denaturation conditions (0.2% NaOH vs 1% NaOH) were implemented (Figure 5G). Both conditions revealed results similar to that observed previously, suggesting that the cross-linking activity of CPZ is independent of the assay conditions. These results show that in a non-cellular context CPZ can alkylate DNA (in addition to cross-linking); taken together with undetectable level of the O⁶-dG adducts in cells treated with CPZ, further studies are warranted to elucidate the major adducts formed and their contribution to the cytotoxicity of CPZ.

Figure 5.

DNA alkylation assay employing linearized pBR322 DNA and relevant imidazotetrazines. (A) Structures of compounds assessed in this study. (B) DNA treatment with TMZ. Conditions: linearized pBR322 DNA (150 ng), 37 °C, Tris buffer (pH 8.5), 15 h. Cisplatin (Cis) and methylmethanesulfonate (MMS) used as positive controls for DNA cross-linking and alkylation, respectively. DNA was visualized using ethidium bromide. (C) DNA treated with CPZ using identical conditions to those described in (B). (D) Concentration dependence of TMZ. Conditions: linearized pBR322 DNA (150 ng), 37 °C, Tris buffer (pH 8.5), 15 h. (E) Concentration dependence of CPZ using identical conditions to those described in (D). (F) Structural variants of TMZ. Conditions: linearized pBR322 DNA (150 ng), 37 °C, Tris buffer (pH 8.5), 15 h. (G) CPZ (500 μM) assayed under different experimental conditions. Tris = Tris buffer, pH 8.5; H₂O = unbuffered water, pH 7; 1% = 1% NaOH used to denature DNA; 0.2% = 0.2% NaOH used to denature DNA. (H) Two potential mechanisms for DNA cross-link formation from propargyl diazonium ion. Mechanism 1 proceeds through an allenyl DNA adduct. Mechanism 2 evokes acrolein as the active cross-linking agent. (I) DNA treatment with acrolein. Conditions: linearized pBR322 DNA (150 ng), 37 °C, Tris buffer (pH 8.5), 15 h. DNA was visualized with ethidium bromide.

Two putative interstrand cross-linking mechanisms can be envisaged for CPZ, both a result of S_N2′ attack on the intermediate propargyl diazonium species 20 (Figure 5H). The first would generate DNA cross-links via an allenyl-DNA intermediate formed after S_N2′ attack on 20 by a nucleophilic site on DNA (e.g. N⁷ guanine). A second plausible mechanism involves an initial S_N2′ addition by water, forming 21 (Figure 5H). Compound 21 can readily tautomerize to 22 (acrolein), which is a known DNA interstrand cross-linker via a multistep mechanism;⁶⁷ indeed, subjecting reagent-grade acrolein to the DNA alkylation assay under identical conditions led to interstrand cross-linking but no evidence of alkylation (Figure 5I). A trapping experiment was performed with acrolein to assess the feasibility of mechanism 2. Upon incubation of acrolein with 2′-deoxyguanosine in PBS, a new peak formed by LC–MS with a mass matching the expected adduct according to literature^68,69 (M + 1 m/z = 324; loss of deoxyribose m/z = 208; Figure S12). However, incubation of CPZ with 2′-deoxyguanosine under identical conditions did not produce any trace of the same adduct (Figure S12). This result preliminarily suggests that significant amounts acrolein are not generated when the propargyl diazonium ion is released, and thus favors mechanism 1 (Figure 5H) as the source of CPZ’s crosslinking.

Biodistribution and Toxicity Studies.

In humans, the cerebral spinal fluid concentration of TMZ averages about 20% of the concentration in the plasma.^70,71 Accumulating even more compound in the brain may be a beneficial strategy to increase imidazotetrazine efficacy against CNS tumors. Of equal importance, targeting more drug to the brain and less to the plasma may offset some of the expected hematological toxicity of an MGMT-independent imidazotetrazine. To investigate the BBB penetrance of TMZ and CPZ, mice were administered 25 mg/kg intravenously then sacrificed after 15 min. Brain tissue and blood were collected from each mouse and immediately acidified to prevent prodrug degradation. The concentration of each compound in the serum (Figure 6A) and brain (Figure 6B) were quantified by LC–MS/MS. TMZ had a brain:serum ratio of 0.08 ± 0.01 ng/g:ng/mL (Figure 6C), commensurate with other TMZ biodistribution studies in murine systems.^58,72,73 Compound 10 had a near-equivalent ratio of 0.06 ± 0.01 ng/g:ng/mL (Figure S13) indicating that swapping a methyl group at N3 for a propargyl group had a negligible effect on BBB penetrance as suggested by clogBB and CNS MPO scores (Figure 2B). However, CPZ demonstrated a >10-fold increase in brain distribution relative to TMZ with a ratio of 1.2 ± 0.2 ng/g:ng/mL (Figure 6C).

Figure 6.

In vivo biodistribution and hematological toxicity studies with CPZ. (A,B) Serum and brain concentrations of TMZ and CPZ 15 min after administration of 25 mg/kg compound (IV) to mice. Number of mice per cohort ≥3. (C) Brain:serum ratios using data from (A,B). (D–F) White blood cell, lymphocyte, and neutrophil counts in mice 7 days after administration of a single 125 mg/kg IV dose of TMZ and CPZ. Statistical significance was determined by using a two-sample Student’s t-test (two-tailed test, assuming equal variance). **P < 0.01.

Arguably the most important question when designing an alkylating agent able to evade MGMT-mediated resistance is the resultant effect on hematological toxicity. MGMT operates as a systemic protectant against alkylating xenobiotics.⁷⁴ Consequently, its inhibition leads to drastically enhanced sensitivity to the toxic effects of alkylating chemotherapy.^41,42,75 Therefore, an MGMT-independent alkylating agent like CPZ could render off-target cells vulnerable to irreparable DNA alkylation and result in a similar degree of toxicity. We hypothesized that the increased BBB permeability of CPZ would divert enough of the drug to the brain such that it could be tolerated at a therapeutically-useful dose. The hematological effects of a single dose of TMZ or CPZ were compared head-to-head in vivo. Mice were treated with 125 mg/kg TMZ or CPZ intravenously; this dose was selected because 125 mg/kg TMZ is known to induce nonlethal toxicity in mice.^36,58,76 After 7 days, the mice were sacrificed and complete blood counts were obtained. Expectedly, 125 mg/kg of TMZ led to depletions in white blood cells, lymphocytes, and neutrophils relative to vehicle-treated mice, representing drug-induced myelosuppression (Figure 6D–F). CPZ-treated mice exhibited total white blood cell, lymphocyte, and neutrophil counts that were similar to TMZ. Notably, CPZ did not give rise to other hematological symptoms such as thrombocytopenia (Figure S14). This could suggest that more BBB penetrant imidazotetrazines mitigate drug-induced hematological toxicity; however, it is also recognized that the exact mechanism of cell death for CPZ, including the biological processing and alkylation pattern of the propargyl moiety, is not yet clear. Additionally, the maximum tolerated dose (MTD) of CPZ was determined to be 66 mg/kg when dosed intraperitoneally, once-per-day over five days. This matches the reported MTD of TMZ at the same schedule (~66 mg/kg).⁷⁷

DISCUSSION

TMZ has been a mainstay as the standard-of-care therapy for GBM patients since its approval in 2005 and remains a frontline therapy for other brain cancers such as oligodendrogliomas and diffuse astrocytic gliomas.⁷⁸ In patients whose tumors do not express MGMT and have functioning MMR, TMZ extends GBM patient survival by nearly one year compared to patients receiving RT only.⁸ Unfortunately, GBMs that express MGMT and/or have reduced MMR capacity are not sensitive to TMZ, and this describes a majority of the GBM patient population both newly diagnosed and recurrent. Therefore, a variant of TMZ that is able to extend the survival benefits to GBM patients with MGMT-expressing and MMR-deficient tumors would have a transformative clinical impact. The 30+ year absence of an approved imidazotetrazine that appends a group apart from methyl underscores the immensity of the challenges faced when modifying imidazotetrazines: synthetic accessibility, rate of hydrolysis, BBB permeability, toxicity profile, etc. and suggests that simply substituting the methyl group at N3 with another functional group is unlikely to yield MGMT-independent anticancer activity that properly balances all other factors.

The dual-substituted compound CPZ displays anticancer activity irrespective of MGMT expression and MMR status. Though MGMT promoter methylation is predictive of tumoral response to TMZ, the biomarker is not widely actionable because of extensive intertumoral heterogeneity and the lack of other therapeutic options. As a result, all GBM patients are treated with TMZ despite its ineffectiveness in most of these patients. The importance of MMR for patients receiving TMZ extends beyond futile cycling of O⁶-methylguanine. TMZ chemotherapy in MGMT-negative GBM induces a strong selective pressure to mutate or downregulate the MMR pathway, usually through MSH6, MSH2, MLH1, or PMS2. Recent studies have shown that in addition to conferring resistance to TMZ, loss of MMR function unleashes the mutagenic potential of TMZ as unrepaired O⁶-methylguanine adducts lead to widespread G:C → A:T transition mutations, more malignant hypermethylated tumors,^14–17 and a positive result in the Ames test. Contrastingly, CPZ leads to direct lethality in cancer cells and appears to avoid a lesion-tolerant, hypermutated phenotype as suggested by negative results in two Ames tests, including in a strain more sensitive to mutation. However, it should be noted that these two strains are sensitive to mutation via direct base substitution, and use of other strains to interrogate other possible forms of mutagenicity (e.g. frame shift, cross-linking, oxidative damage, etc.)⁷⁹ will be critical to determine the full mutagenic profile of CPZ. Killing GBM cells through a less mutagenic lesion could mitigate or delay tumor recurrence and/or progression by preventing the acquisition of additional driver mutations.

In addition to alkylating DNA in biochemical experiments, CPZ is also capable of cross-linking DNA, and a cross-linking imidazotetrazine immediately invites comparison to first-generation imidazotetrazine, MTZ.⁸⁰ Despite showing curative anticancer activity in murine models,⁸¹ MTZ met a swift end in the clinic due to extreme myelosuppressive effects that were attributed to DNA cross-linking.^82,83 Not to be understated, the single dose MTD of MTZ in mice is 37.5 mg/kg, whereas CPZ can readily be dosed 66 mg/kg once-per-day over 5 days. Careful consideration of the active cross-linking species reveals key differences between CPZ and MTZ. Both the anticancer activity and toxicity of MTZ has been attributed to interstrand DNA cross-links as only ~20 per cell are required for lethality⁸⁴ (vs >6500 O⁶-methylguanine lesions).⁸⁵ Indeed, a comparison of the cytotoxicity of N3-ethyl imidazotetrazine (2, U87 IC₅₀ = 470 μM) and N3-chloroethyl imidazotetrazine (MTZ, U87 IC₅₀ = 30 μM), which differ only by a single chlorine atom, suggests that MTZ derives nearly all of its anticancer efficacy in culture from its ability to cross-link DNA. In contrast to MTZ, the DNA interstrand cross-links imparted by CPZ are hypothesized to arise from a side reaction of the propargyl diazonium ion (S_N2′ vs S_N2). Therefore, the cytotoxicity of CPZ appears not to be primarily driven by the interstrand cross-links. The contribution of CPZ cross-links to cell death can be estimated by the difference in activity between compounds 10 and 19. The slight difference in cytotoxicity (IC₅₀ = 29 vs 37 μM in U87 and 77 vs 110 μM in T98G for 10 and 19, respectively) observed may be due to the ability of 10 to cross-link where 19 cannot (S_N2′ pathway is blocked). The retention of cytotoxicity by 19 serves as further evidence that the anticancer activity of N3-propargyl imidazotetrazines is likely driven by propargylic monoadducts, although at this time the exact pattern of alkylation is not known.

The observations reported herein challenge a major assumption in the imidazotetrazine literature that N3 derivatives of TMZ operate analogous to TMZ, through formulation of O⁶-dG alkyl lesions; to date, the effect of propargyl and other diazonium ions on DNA has not been explored in depth.^40,86,87 Our cell-free DNA alkylation experiments suggest that DNA alkylation is occurring, but cell-based LC–MS/MS quantitation indicates that O⁶-Prop-dG adducts are not formed at a detectable level, and thus such lesions are not likely to be the driver of cell death, in stark contrast with TMZ and the analogous methyl lesion. Instead, propargyl diazonium ions delivered by CPZ induce rapid cytotoxicity across cell lines. Future studies are required to identify the exact DNA alkylation pattern and to trace the contribution of the major adduct(s) to DNA and cancer cell death.

The compounds described herein (CPZ and others) now provide the appropriate tools to test the hypothesis that an MGMT-evading imidazotetrazine can be safe and efficacious in intracranial in vivo models of GBM. In addition, the GL261/GL261 MGMT+ isogenic cell line pair reported here will be an extremely useful tool for in vivo studies, since a syngeneic model provides the opportunity to assess compound activity in the presence of an intact immune system while recapitulating the chemoresistance exerted by MGMT. While CPZ and other compounds reported herein provide reason for optimism, it is notoriously difficult to out-perform TMZ in intracranial models; almost exclusively, imidazotetrazines that show promise in cell culture fail in in vivo models,⁷² with only one compound showing a meaningful difference compared to TMZ.⁵⁸

In summary, we have identified a novel imidazotetrazine that circumvents the two primary resistance mechanisms to TMZ, expression of MGMT and loss of MMR function. Despite its drawbacks, TMZ remains the backbone of GBM treatment; CPZ is an exciting alternative that may improve the clinical situation of patients that cannot be successfully treated beyond surgery and RT. The assessment of TMZ and CPZ head-to-head in intracranial tumors will be reported in due course.

MATERIALS AND METHODS

Chemistry.

Chemical reagents were purchased from commercial sources and used without further purification. Flash chromatography was performed using silica gel (230–400 mesh). Anhydrous solvents were dried after being passed through columns packed with activated alumina under positive pressure of nitrogen. Unless otherwise noted, all reactions were carried out in oven-dried glassware with magnetic stirring under nitrogen atmosphere. ¹H and ¹³C NMR spectra were recorded on Bruker 500 (500 MHz, ¹H; 125 MHz, ¹³C) or Varian Unity Inova 500 (500 MHz, ¹H) MHz spectrometers. Spectra are referenced to residual chloroform (δ = 7.26 ppm, ¹H; 77.16 ppm, ¹³C) or dimethyl sulfoxide (δ = 2.50 ppm, ¹H; 39.52 ppm, ¹³C). Multiplicities are indicated by s (singlet), d (doublet), t (triplet), q (quartet), m (multiplet), and br (broad). Coupling constants J are reported in Hertz (Hz). High resolution mass spectrometry was performed on a Waters Q-Tof Ultima or Waters Synapt G2-Si instrument with electrospray ionization or electron impact ionization (EI).

Safety Profile of Chemical Reagents and Intermediates.

As noted in our previous publication (Angew. Chem. Int. Ed. 2020, 132(5), 1873–1878), TMZ and related imidazotetrazines can be advantageous as diazomethane surrogates on 0.2 mmol scale with 60 °C reaction temperatures. A recent paper from Sperry et al. (Org. Process Res. Dev. 2021, 25(7), 1690–1700) reports safety precautions of TMZ, noting that on larger scales (>500 g) and elevated temperatures (reported onset temperatures of 125–200 °C) its handling can be hazardous. Differential scanning calorimetry (DSC) and thermogravimetric (TGA) data are two methods that can facilitate a better understanding of shock sensitivity and explosivity of compounds. Data obtained via these methods are very sensitive to slight variations, including but not limited to sample morphology and crystal structures.⁸⁸ Figure S15 shows parameters determined from DSC and TGA, with calculated explosion potentials and shock sensitivities based on the Yoshida correlations.⁸⁹ With this data, 30 and CPZ have the potential to be explosive and shock-sensitive. However, it is prudent to note that a “null” prediction still can have the potential to exhibit explosive and/or shock sensitive behavior, and more conservative estimations (such as the Pfizer-modified correlation)⁸⁸ could predict more of the following compounds to be shock sensitive and/or explosive.

Cell Culture and Reagents.

All cell lines were grown in a 37 °C, 5% CO2, humidified environment, in media containing 1% penicillin/streptomycin. Stable cell line generation of GL261 MGMT+ was purchased from Creative Biogene. Cell culture conditions are as follows: cell lines U87, T98G, and RKO were grown in EMEM with 10% FBS. D54, GL261, GL261 MGMT+, and U118MG cell lines were grown in DMEM with 10% FBS. HCT116 cells were grown in RMPI with 10% FBS. D341 Med cells were grown in EMEM with 20% FBS. TMZ was purchased from AK Scientific. TMZ analogues were synthesized as described in the Supporting Information. Compounds were dissolved in DMSO (1% final concentration, Fisher Chemical) for cell culture studies.

Bacterial Strains.

S. typhimurium 14028 was obtained from the American Type Culture Collection (ATCC).

Antibodies.

Antibodies used herein: MGMT: CST-2739, anti-rabbit IgG HRP-linked: CST-7074, GADPH: CST-2118.

Procedures for Biological Assays.

Cell Viability Assays.

Cells were harvested, seeded in a 96-well plate and allowed to adhere. After three hours, compound was added to each well in DMSO (1% final concentration). Cells were incubated for seven days (unless otherwise noted) before viability was assessed by the Alamar Blue assay. Raptinal (20 μM) was used as a dead control.

Hydrolytic Stability Studies.

Compound (100 μM) was incubated at 37 °C in PBS at pH 7.4 (unless otherwise noted). At the indicated time point, an aliquot was collected and analyzed by HPLC. Compound peaks were integrated and compared to the integration at t₀ to calculate the percentage of prodrug remaining.

Killing Kinetics Study.

U87 cells were seeded in seven 96 well plates starting at 19,200 cells/well (for 24 h time point) and serially diluted by two-fold for each subsequent 24 h time point. At each time point, cell viability was assessed via Alamar Blue assay.

Quantitation of O⁶-Guanine DNA Adducts in Cells.

GL261, A172, or D54 cells were plated at 1 × 10⁶ c/w in a 6-well plate before they were treated with compound at the indicated concentration (1% final concentration DMSO). After incubation for the indicated time, the cells were harvested and pelleted. Genomic DNA was extracted using the DNeasy Blood & Tissue Kit (Qiagen, ID: 69504). DNA was then precipitated using the following procedure: 1/10 v/v 3M sodium acetate (pH 5.2) and 2.5× v/v ethanol was added to each sample which was then kept at −80 °C for 1 h. The mixture was centrifuged at max at 4 °C for 30 min and decanted to afford a pellet of DNA, which was re-suspended in ddH₂O containing 10 mM Tris base (pH7.5) and 1 mM EDTA. The concentration of DNA in each sample was quantified by measuring absorbance on a NanoDrop 2000 UV–Vis Spectrophotometer (Thermo Fisher). DNA (10 μg) from each sample was added to DNA hydrolysis buffer⁹⁰ and incubated at 37 °C for 6 h. Hydrolyzed samples were then submitted for LC–MS/MS quantitation along with a synthetic standard. Samples were analyzed with a 5500 QTRAP LC/MS/MS system (AB Sciex) with a 1200 series HPLC system (Agilent).

Immunoblotting/Western Blot Procedure.

Cells were lysed using RIPA buffer containing protease inhibitor cocktail (Calbiochem). Protein concentrations were determined using the BCA assay (Pierce). Lysates containing 10 μg of protein were loaded onto 4–20% gradient gels (BioRad), and sodium dodecyl sulfate-polyacrylamide gel electrophoresis was run. For GL261 MGMT+, the lysate was diluted x20 for optimal visualization of MGMT band. Proteins were then transferred onto membrane (PDVF Millipore) for Western Blot analysis. Blots were blocked with BSA solution (2 g in 40 mL TBST) for 1 h followed by primary antibody addition (1:1000) and incubation overnight. Following overnight incubation, blots were washed with TBST, and incubated with HRP-linked secondary antibody for 1 h in TBST. Blots were washed, then imaged with ChemiDoc after incubation with SuperSignal West Pico Solution following manufacturer’s procedures.

Antimicrobial Susceptibility Tests.

Susceptibility testing was performed in biological triplicate, using the micro-dilution broth method as outlined by the Clinical and Laboratory Standards Institute. Bacteria were cultured with cation-adjusted Meuller-Hinton broth (Sigma-Aldrich; catalogue number: 90922) media in round-bottom 96-well plates (Corning; catalogue number: 3788).

Ames Test.

Compounds were assessed at the indicated concentrations in the TA100 or WP2 uvrA pKM101 tester strains using the Muta-ChromoPlate 96-well microplate version of the Ames test (Environmental Bio-Detection Products, prod. no. 5051) according to manufacturer’s instructions. In brief, TA100 S. typhimurium bacteria were grown overnight and plated with compound, growth medium, and indicator (no S9 activation) in 96 well plates. Bacteria were incubated for 5 days before the number of wells per 96 well plate containing revertant colonies were identified colorimetrically.

DNA Cross-Linking/Alkylation Assay.

Procedure was adapted from Healy et al.⁶⁶ In brief, pBR322 plasmid DNA was linearized with EcoRI (New England BioLabs) in NEB EcoRI buffer (New England BioLabs) according to the manufacturer’s instructions. The cut plasmid DNA was purified using a PCR cleanup kit (QIAquick PCR Purification Kit, Qiagen) and eluted into DNA buffer (10 mM Tris Cl, pH 8.5). The concentration of linearized DNA was quantified by measuring absorbance on a NanoDrop 2000 UV–Vis Spectrophotometer (Thermo Fisher). For each reaction, compound (in DMSO, 5% v/v) was added to DNA (150 ng) in DNA buffer (total volume 10 μL) and incubated at 37 °C for the indicated time. Upon completion, 30 μL DNA denaturation buffer (60 mg/mL sucrose, 10 mg/mL NaOH, and 0.4 mg/mL bromophenol blue in water) was added to each sample (the same buffer without NaOH was added to non-denatured controls). Samples were vortexed and denatured for 15 min at 4 °C before they were loaded onto a 1% neutral agarose gel (containing ethidium bromide) and run in TBE buffer (containing ethidium bromide) at 124 V for 1 h. Gels were imaged using a Gel Dox XR+ system (Bio-Rad).

Acrolein Trapping Assay.

Acrolein (1 mmol) was incubated with 2′-deoxyguanosine (0.09 mmol) at 37 °C in 10X PBS (10 mL, pH =7) overnight. The reaction mixture was directly injected on LC/MS without prior purification to visualize acrolein adducts. 2′-Deoxyguanosine (0.09 mmol)/TMZ (1 mmol) and 2′-deoxyguanosine(0.09 mmol)/CPZ (1 mmol) mixtures were separately incubated in 10× PBS (10 mL, pH = 7) overnight. These reaction mixtures were directly injected on LC/MS without prior purification to look for the same adduct generated as the acrolein adduct standard as described.

BBB Permeability.

All experimental procedures were reviewed and approved by the University of Illinois Institutional Animal Care and Use Committee. CD-1 IGS mice were administered compound in 5% DMSO, 10% Tween-20, and 85% SPE-βCD in sterile water (30% w/v) at 25 mg/kg via lateral tail vein injection. 15 min post injection, mice were sacrificed and blood was collected by lacerating the right auricle with iris scissors. An 18 gauge angiocatheter was inserted through the left ventricle, and all residual circulatory volume was removed by perfusing 0.9% saline solution via an analog peristaltic pump. Blood samples were immediately centrifuged at 13,000 rcf for 5 min, and the supernatant was collected and acidified with 8.5% aqueous H₃PO₄. Brains were harvested from the cranial vault, acidified with 0.3% aqueous H₃PO₄ and flash frozen. Homogenized brain samples were centrifuged twice at 13,000 rcf for 10 min, and the supernatant and tissue debris were separated. The resultant supernatant was analyzed, along with plasma, by LC–MS/MS to determine compound concentrations.

Assessment of Hematological Toxicity.

Male CD-1 IGS mice (n ≥ 3 mice/group) were administered a single dose of 125 mg/kg compound intravenously. Imidazotetrazines were formulated with DMSO, Tween-20, and 30% (w/v) SBE-βCD in sterile water immediately prior to injection. Seven days post-treatment, mice were humanely sacrificed and whole blood was collected for assessment of total white blood cells, lymphocytes, neutrophils, platelets, and red blood cells.

Assessment of MTD.

Female C57BL/6 mice were administered compound via intraperitoneal injection in 35% PEG400, 25% propylene glycol, 6% Tween-80, and 34% sterile water (10 mL/kg injection volume) at 66 mg/kg once per day over 5 days. Mice were monitored for weight loss and other signs of toxicity; no significant weight loss or signs of toxicity were observed for this dosing schedule.