Esterase-Sensitive Prodrugs of a Potent Bisubstrate Inhibitor of Nicotinamide N-Methyltransferase (NNMT) Display Cellular Activity

A recently discovered bisubstrate inhibitor of Nicotinamide N-methyltransferase (NNMT) was found to be highly potent in biochemical assays with a single digit nanomolar IC50 value but lacking in cellular activity. We, here, report a prodrug strategy designed to translate the observed potent biochemical inhibitory activity of this inhibitor into strong cellular activity. This prodrug strategy relies on the temporary protection of the amine and carboxylic acid moieties of the highly polar amino acid side chain present in the bisubstrate inhibitor. The modification of the carboxylic acid into a range of esters in the absence or presence of a trimethyl-lock (TML) amine protecting group yielded a range of candidate prodrugs. Based on the stability in an aqueous buffer, and the confirmed esterase-dependent conversion to the parent compound, the isopropyl ester was selected as the preferred acid prodrug. The isopropyl ester and isopropyl ester-TML prodrugs exhibit improved cell permeability, which also translates to significantly enhanced cellular activity as established using assays designed to measure the enzymatic activity of NNMT in live cells.


Introduction
Nicotinamide N-methyltransferase (NNMT) is a small molecule methyltransferase enzyme responsible for the conversion of nicotinamide (NA, vitamin b3) to 1-methylnicotinamide (MNA). NNMT utilizes the cofactor S-adenosyl-L-methionine (SAM) as a methyl donor, which is converted to S-adenosyl-L-homocysteine (SAH) upon methylation of nicotinamide [1]. Under normal physiological conditions, NNMT is mainly expressed in the liver and in adipose tissue [2]. One of the primary roles of NNMT is the detoxification of xenobiotics. This function is achieved through NNMT's broad substrate recognition that allows for the methylation of different metabolites, including pyridines, quinolines, and other related heterocyclic aromatics [3].
The overexpression of NNMT has been described in a wide variety of tissues and diseases, generally with detrimental effects, although there are reports suggesting a ben- The carboxylic acid can be masked as an ester and the amine can be masked as an amide using the esterase-sensitive trimethyl-lock (TML). (B) The mechanism of the trimethyl-lock cleavage. Deacetylation by esterases results in spontaneous lactonization, releasing the free amine.

Experimental Procedures
All reagents employed were of American Chemical Society grade or finer and were used without further purification unless otherwise stated. For compound characterization, 1 H NMR spectra were recorded at 400 and 500 MHz with chemical shifts reported in parts per million downfield relative to CHCl3 (7.26) or CH3OH (δ 3.31). 1 H NMR data are reported in the following order: multiplicity (s, singlet; d, doublet; t, triplet; q, quartet; and m, multiplet), coupling constant (J) in hertz (Hz), and the number of protons. Where appropriate, the multiplicity is preceded by br, indicating that the signal was broad. 13 C NMR spectra were recorded at 101 or 126 MHz with chemical shifts reported relative to CHCl3 (77. 16) or CH3OH (δ 49.00). High-resolution mass spectrometry (HRMS) analysis was performed using a Q-TOF instrument. Purity was confirmed to be ≥95% by LCMS performed on a Shimadzu LC-20AD system (Shimadzu, 's-Hertogenbosch, The Netherlands) with a Shimadzu Shim-Pack GISS-HP C18 column (3.0 × 150 mm, 3 µm particle size, Shimadzu, 's-Hertogenbosch, The Netherlands) at 30 °C and equipped with a UV detector monitoring at 214 and 254 nm. The following solvent system, at a flow rate of 0.5 mL/min, was used: solvent A, 0.1% formic acid in water; solvent B, acetonitrile. Gradient elution was as follows: 95:5 (A/B) for 2 min, 95:5 to 0:100 (A/B) over 13 min, 0:100 (A/B) for 2 min, then reversion back to 95:5 (A/B) over 1 min, 95:5 (A/B) for 2 min. This system was connected to a Shimadzu 8040 triple quadrupole mass spectrometer (ESI ionization, Shimadzu, 's-Hertogenbosch, The Netherlands). The compounds were purified via preparative HPLC performed on a BESTA-Technik system with a Dr. Maisch Reprosil Gold 120 C18 column (25 × 250 mm, 10 µm particle size, Dr.Maisch, Ammerbuch-Entringen, Germany) and equipped with an ECOM Flash UV detector monitoring at 214 nm. The following solvent system, at a flow rate of 12 mL/min, was used: solvent A: 0.1% TFA in water/acetonitrile 95/5; solvent B: 0.1% TFA in water/acetonitrile 5/95. Gradient elution was as follows: 95: 5

Ester Stability Assay
The prodrugs were evaluated for their stability in both PBS buffer at pH 7.4 and Tris buffer at pH 8.4. The compounds were dissolved in DMSO at a concentration of 40 mM and diluted with the respective buffer to a final concentration of 1 mM. Compounds were tested directly (t 0 ) and subsequently every 2 h over a time period of 16 h by HPLC. Compounds were eluted from a Dr. Maisch ReproSil-Pur C18 column (4.6 × 250 mm, 10 µm particle size, Dr. Maisch, Ammerbuch-Entringen, Germany) with the following solvent system at a flow rate of 0.5 mL/min: solvent A, 0.1% trifluoroacetic acid in water/acetonitrile (95:5); solvent B, 0.1% trifluoroacetic acid in water/acetonitrile (5:95). Gradient elution was as follows: 100:0 (A/B) to 0:100 (A/B) over 8 min, 0:100 (A/B) for 1 min, then reversion back to 100:0 (A/B) over 1 min, 100:0 (A/B) for 2 min. The formation of the parent compound was evaluated and normalized by measuring the peak area at 214 nm and comparing it to the initial timepoint.

Esterase-Mediated Hydrolysis
The conversion of the prodrugs to the parent compound by esterases was evaluated using pig liver esterase (PLE, 18 U/mg, Sigma-Aldrich, St. Louis, MO, USA) in PBS at pH 7.4. Compounds were dissolved in DMSO at 40 mM, diluted to a final concentration of 2 mM with PBS, and added to an equal volume of a 10 mg/mL solution of PLE in PBS (pH 7.4), resulting in final concentrations of 2.5% DMSO, 1 mM compound, and 5 mg/mL PLE. At different time-points 50 µL aliquots were taken, added to 100 µL acetonitrile to precipitate the proteins, and centrifuged for 5 min at 10,000 rpm. The supernatant was subsequently analyzed by HPLC as described in Section 2.2 above.

Cell Culture
HSC-2 human oral cancer cell line, T24 human bladder cancer cell line, and A549 human lung cancer line were cultured in DMEM/F12 medium and supplemented with 10% fetal bovine serum and 50 µg/mL gentamicin at 37 • C in a humidified 5% CO 2 incubator. For each compound tested, powder was dissolved in DMSO at 100 mM concentration. This stock solution was then diluted in culture medium to final concentration values ranging between 1 µM and 100 µM. For each sample, DMSO was kept constant at 0.1% final concentration. The day before starting treatment, cells were seeded in 96-well plates, at a density of 2000 cells/well. Cells were allowed to attach overnight and then incubated with compounds at different concentrations, or with DMSO only, for 24, 48, and 72 h. All experiments were performed in triplicate.

MTT Cell Viability Assay
Cell proliferation was determined using a colorimetric assay with 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT). The MTT assay measures the conversion of MTT to insoluble formazan by dehydrogenase enzymes of the intact mitochondria of living cells. Cell proliferation was evaluated by measuring the conversion of the tetrazolium salt MTT to formazan crystals upon treatment with compounds or DMSO only for 24, 48, and 72 h. Briefly, cells were incubated for 2 h at 37 • C with 100 µL fresh culture medium containing 5 µL of MTT reagent (5 mg/mL in PBS). The medium was removed and 200 µL isopropanol were added. The amount of formazan crystals formed correlated directly with the number of viable cells. The reaction product was quantified by measuring the absorbance at 540 nm using a plate reader. Experiments were repeated three times. Results were expressed as a percentage of the control (the control equals 100% and corresponds to the absorbance value of each sample at time zero) and presented as mean values ± standard deviation of three independent experiments performed in triplicate. The quantification of 1-methylnicotinamide (MNA), nicotinamide (NA), nicotinic acid (NicA), 1-methyl-2-pyridone-5-carboxamide (Met-2Pyr), and 1-methyl-4-pyridone-5carboxamide (Met-4Pyr) was performed by applying ultra-high pressure liquid chromatography coupled to tandem mass spectrometry (UPLC-MS) according to the methodology previously described with minor modifications [35].

Statistical Analysis
Data were analysed using GraphPad Prism software for Windows (GraphPad Software version 9.2.0, San Diego, CA, USA). Significant differences between groups were determined using the one-way analysis of variance (ANOVA). A p value < 0.05 was considered as statistically significant.

Chemical Synthesis
The prodrug analogues of GYZ-319 were prepared following the routes depicted in Schemes 1 and 2. This synthetic approach was based on the one previously developed during our structure-activity relationship studies with bisubstrate NNMT inhibitors leading to the discovery of GYZ-319 [24], allowing for the convenient modification of different parts of the molecule. The ester building blocks (Scheme 1) were synthesized starting from the Boc-Asp(Bn)-OH 1, which is esterified with the appropriate iodides in the presence of potassium carbonate to produce compounds 2b-e, followed by the deprotection of the benzyl protecting group to obtain compounds 3b-e. Compounds 3a and 3f were obtained from commercial sources. Free carboxylic acids 3a-f were then converted into Weinreb amides 4a-f using BOP-coupling conditions and subsequently reduced to the corresponding aldehyde (5b-f) with DIBAL-H. Compound 4a followed a different route to produce TML-prodrug 8a which contains the free carboxylic acid. In order to do so, the Boc group was selectively deprotected using HCl in dioxane to produce free amine 6a. The free amine was then coupled to the trimethyl-lock acid 7 [36] with BOP and triethylamine to yield Weinreb amide intermediate 8a followed by DIBAL reduction to form aldehyde 9a. The aldehydes were subsequently coupled to intermediate 10 via reductive amination conditions, forming intermediates 11b-f, which were then deprotected to form ester prodrugs 12b-f (Scheme 2). Compounds 11b-f can alternatively be selectively Bocdeprotected using TFA/DCM to form intermediates 13b-f and subsequently coupled to trimethyl-lock acid 7 with BOP and triethylamine. Compound 13a was synthesized through coupling of aldehyde 9a with compound 10. The intermediates were then deprotected under acidic conditions to yield ester-TML dual prodrugs 14a-f. benzyl protecting group to obtain compounds 3b-e. Compounds 3a and 3f were obtained from commercial sources. Free carboxylic acids 3a-f were then converted into Weinreb amides 4a-f using BOP-coupling conditions and subsequently reduced to the corresponding aldehyde (5b-f) with DIBAL-H. Compound 4a followed a different route to produce TML-prodrug 8a which contains the free carboxylic acid. In order to do so, the Boc group was selectively deprotected using HCl in dioxane to produce free amine 6a. The free amine was then coupled to the trimethyl-lock acid 7 [36] with BOP and triethylamine to yield Weinreb amide intermediate 8a followed by DIBAL reduction to form aldehyde 9a.
The aldehydes were subsequently coupled to intermediate 10 via reductive amination conditions, forming intermediates 11b-f, which were then deprotected to form ester prodrugs 12b-f (Scheme 2). Compounds 11b-f can alternatively be selectively Boc-deprotected using TFA/DCM to form intermediates 13b-f and subsequently coupled to trimethyl-lock acid 7 with BOP and triethylamine. Compound 13a was synthesized through coupling of aldehyde 9a with compound 10. The intermediates were then deprotected under acidic conditions to yield ester-TML dual prodrugs 14a-f.

Buffer Stability
All prodrugs prepared were first tested for their inhibitory activity against NNMT using a biochemical enzymatic activity assay. Somewhat surprisingly, several prodrugs showed significant inhibition when tested at fixed concentrations of 5 and 25 µM. In order to evaluate the validity of these results, the prodrugs were subsequently evaluated for their hydrolytic stability in both PBS buffer at pH 7.4 and Tris buffer at pH 8.4. Both these buffers and pHs have been used in the different biochemical and cellular assays described in this report. Compounds were dissolved in DMSO at a concentration of 40 mM and diluted with the respective buffer to a final concentration of 1 mM. Compound stability was then assessed by RP-HPLC directly (t 0 ) and subsequently every 2 h over a time period of 16 h. The results presented in Table 1 show that over time, a significant amount of hydrolysis occurs for most of the prodrugs containing only the ester modification. Only the isopropyl ester was found to be stable under these conditions. Interestingly, the stability of the prodrugs increased significantly in the presence of the trimethyl-lock (TML) moiety at the amine position. Even for the rather labile methyl ester 12b, the TML group in 14b results in a decrease in hydrolysis of the methyl ester. Benzyl ester 8f was found to be the least stable, and due to its poor aqueous solubility, the benzyl ester-TML dual prodrug 14f was not evaluated further. The most stable prodrugs were found to be the isopropyl ester (compound 12e), the trimethyl-lock (compound 14a), and the isopropyl-trimethyl-lock dual prodrug (compound 14e). However, as compound 14a was found not to improve the cellular activity of the parent compound (see Figure S1 in the Supplementary Information), this compound was not evaluated further.

Buffer Stability
All prodrugs prepared were first tested for their inhibitory activity against NNMT using a biochemical enzymatic activity assay. Somewhat surprisingly, several prodrugs showed significant inhibition when tested at fixed concentrations of 5 and 25 µM. In order to evaluate the validity of these results, the prodrugs were subsequently evaluated for their hydrolytic stability in both PBS buffer at pH 7.4 and Tris buffer at pH 8.4. Both these buffers and pHs have been used in the different biochemical and cellular assays described in this report. Compounds were dissolved in DMSO at a concentration of 40 mM and diluted with the respective buffer to a final concentration of 1 mM. Compound stability was then assessed by RP-HPLC directly (t0) and subsequently every 2 h over a time period of 16 h. The results presented in Table 1 show that over time, a significant amount of hydrolysis occurs for most of the prodrugs containing only the ester modification. Only the isopropyl ester was found to be stable under these conditions. Interestingly, the stability of the prodrugs increased significantly in the presence of the trimethyl-lock (TML) moiety at the amine position. Even for the rather labile methyl ester 12b, the TML group in 14b results in a decrease in hydrolysis of the methyl ester. Benzyl ester 8f was found to be the least stable, and due to its poor aqueous solubility, the benzyl ester-TML dual prodrug 14f was not evaluated further. The most stable prodrugs were found to be the isopropyl ester (compound 12e), the trimethyl-lock (compound 14a), and the isopropyl-trimethyllock dual prodrug (compound 14e). However, as compound 14a was found not to improve the cellular activity of the parent compound (see Figure S1 in the Supplementary Information), this compound was not evaluated further.

Esterase-Mediated Hydrolysis
The next step was to establish whether compounds 12e and 14e are converted to the parent compound in the presence of an esterase. Using commercially available pig liver esterase (PLE), both compounds were found to be readily converted to the parent GYZ-319 as measured by RP-HPLC ( Figure 2). Within 4 h, the prodrugs were fully converted to the parent compound, while no hydrolysis was observed in the absence of PLE.   Of note is the sequential conversion of dual prodrug 14e in which the TML is hydrolyzed first followed by the isopropyl ester ( Figure 2B). No trace of compound 14a could be observed in which the ester is cleaved and the TML is still in place. This finding suggests that the TML group hinders the esterase-mediated hydrolysis of the isopropyl ester moiety and only after deacetylation of the TML moiety followed by its spontaneous loss, can the ester moiety be cleaved by the esterase.

Esterase-Mediated Hydrolysis
The next step was to establish whether compounds 12e and 14e are converted to the parent compound in the presence of an esterase. Using commercially available pig liver esterase (PLE), both compounds were found to be readily converted to the parent GYZ-319 as measured by RP-HPLC ( Figure 2). Within 4 h, the prodrugs were fully converted to the parent compound, while no hydrolysis was observed in the absence of PLE.
Of note is the sequential conversion of dual prodrug 14e in which the TML is hydrolyzed first followed by the isopropyl ester ( Figure 2B). No trace of compound 14a could be observed in which the ester is cleaved and the TML is still in place. This finding suggests that the TML group hinders the esterase-mediated hydrolysis of the isopropyl ester moiety and only after deacetylation of the TML moiety followed by its spontaneous loss, can the ester moiety be cleaved by the esterase. Of note is the sequential conversion of dual prodrug 14e in which the TML is hydrolyzed first followed by the isopropyl ester ( Figure 2B). No trace of compound 14a could be observed in which the ester is cleaved and the TML is still in place. This finding suggests that the TML group hinders the esterase-mediated hydrolysis of the isopropyl ester moiety and only after deacetylation of the TML moiety followed by its spontaneous loss, can the ester moiety be cleaved by the esterase.

Cellular Assays
With the suitability of prodrugs 12e and 14e established, the compounds were next tested using an MTT assay to evaluate their effect on cell viability in three different cancer cell lines: HSC-2 (oral cancer), T24 (bladder cancer), and A549 (lung cancer) (Figure 3). The results of these assays revealed that neither the reference NNMT inhibitor 6-methylamino-nicotinamide (6-MANA), nor prodrugs 12e or 14e or the parent GYZ-319, showed any appreciable effect on cell viability when tested at 1 and 10 µM concentrations. At the highest concentration tested of 100 µM, 6-MANA again showed no effect while the isopropyl ester prodrug 12e showed an effect comparable to GYZ-319. By comparison, administration of compound 14e at 100 µM did cause a time dependent reduction in cell

Cellular Assays
With the suitability of prodrugs 12e and 14e established, the compounds were next tested using an MTT assay to evaluate their effect on cell viability in three different cancer cell lines: HSC-2 (oral cancer), T24 (bladder cancer), and A549 (lung cancer) (Figure 3). The results of these assays revealed that neither the reference NNMT inhibitor 6-methylaminonicotinamide (6-MANA), nor prodrugs 12e or 14e or the parent GYZ-319, showed any appreciable effect on cell viability when tested at 1 and 10 µM concentrations. At the highest concentration tested of 100 µM, 6-MANA again showed no effect while the isopropyl ester prodrug 12e showed an effect comparable to GYZ-319. By comparison, administration of compound 14e at 100 µM did cause a time dependent reduction in cell viability for all three cell lines tested. These findings show that both the parent compound and the corresponding prodrug inhibitors have little effect on cell viability unless tested at very high concentrations, more than 10,000 times higher than the activity of the parent compound in biochemical assays (IC 50 = 3.7 nM). One explanation for this finding may be that the prodrugs are still not effectively entering the cells. An alternative explanation, however, could be that the compounds do in fact enter the cells but that subsequent NNMT inhibition is simply not toxic to the cells. If this is the case, the impact on cell viability observed when applying the compounds at the highest concentration tested (100 µM) could instead by ascribed to a non-specific toxic effect.
To more directly investigate the cellular activity of GYZ-391 and the corresponding prodrugs, we next turned to a recently developed assay that allows for the detection and quantification of NNMT activity in live cells [7]. Using this assay, the levels of MNA produced by NNMT in an A549 lung carcinoma cell line can be quantified by means of a sensitive LC-MS method. Using this approach, we evaluated the effect of the reference NNMT inhibitor 6-MANA, parent compound GYZ-319, and prodrugs 12e or 14e on cellular MNA production. As illustrated in Figure 4, high concentrations of both 6-MANA and GYZ-319 are required to substantially decrease the levels of MNA in A549 lung cancer cells. In contrast, when the isopropyl ester prodrug 12e was tested, a clear and significant decrease in MNA levels was observed. This effect is even further enhanced by the introduction of the TML moiety as present in prodrug 14e. compound in biochemical assays (IC50 = 3.7 nM). One explanation for this finding may be that the prodrugs are still not effectively entering the cells. An alternative explanation, however, could be that the compounds do in fact enter the cells but that subsequent NNMT inhibition is simply not toxic to the cells. If this is the case, the impact on cell viability observed when applying the compounds at the highest concentration tested (100 µM) could instead by ascribed to a non-specific toxic effect. To more directly investigate the cellular activity of GYZ-391 and the corresponding prodrugs, we next turned to a recently developed assay that allows for the detection and quantification of NNMT activity in live cells [7]. Using this assay, the levels of MNA produced by NNMT in an A549 lung carcinoma cell line can be quantified by means of a sensitive LC-MS method. Using this approach, we evaluated the effect of the reference NNMT inhibitor 6-MANA, parent compound GYZ-319, and prodrugs 12e or 14e on cellular MNA production. As illustrated in Figure 4, high concentrations of both 6-MANA and GYZ-319 are required to substantially decrease the levels of MNA in A549 lung cancer cells. In contrast, when the isopropyl ester prodrug 12e was tested, a clear and significant decrease in MNA levels was observed. This effect is even further enhanced by the introduction of the TML moiety as present in prodrug 14e.  These findings indicate that the employed prodrug strategy is able to convert a potent, but non-permeable, inhibitor into a compound with cellular activity. Notably, our findings also seem to suggest that the capacity for a small molecule to inhibit NNMT in cells does not per se lead to an impact on cell viability. This is in keeping with recent reports showing that the addition of NNMT inhibitors to cancer associated fibroblasts do not kill the cells but rather cause a reversion of cell morphology to one that more closely resembles normal fibroblasts [17].

Conclusions
In this report, we describe a prodrug approach to translate the biochemical activity of the potent bisubstrate NNMT inhibitor GYZ-319 into cellular activity. The prodrug  These findings indicate that the employed prodrug strategy is able to convert a potent, but non-permeable, inhibitor into a compound with cellular activity. Notably, our findings also seem to suggest that the capacity for a small molecule to inhibit NNMT in cells does not per se lead to an impact on cell viability. This is in keeping with recent reports showing that the addition of NNMT inhibitors to cancer associated fibroblasts do not kill the cells but rather cause a reversion of cell morphology to one that more closely resembles normal fibroblasts [17].

Conclusions
In this report, we describe a prodrug approach to translate the biochemical activity of the potent bisubstrate NNMT inhibitor GYZ-319 into cellular activity. The prodrug strategy focused specifically on masking the amino acid functionality of GYZ-319. The carboxylic acid was masked with an ester using a variety of alkyl and benzyl groups and the amine was masked with a trimethyl-lock group, both of which can be released by esterase activity. The different combinations of esterase-cleavable motifs led to the selection of the isopropyl ester 12e and the isopropyl ester/TML dual prodrug 14e as the compounds with the most promising profile in terms of stability and cellular activity. When tested against different cancer cell lines, the prodrugs were found to have little impact on cell viability. However, when evaluated in an assay allowing for the direct quantification of cellular MNA production, a clear dose-dependent effect was observed for the prodrugs. Notably, the prodrugs exhibit a significant enhancement of cellular activity compared to the parent compound. The data presented here point to the potential for using such prodrug strategies for the delivery of polar NNMT inhibitors into cells. On-going efforts are focused on further establishing the potential of such NNMT inhibitor prodrugs in a range of cellular and in vivo assays relevant to both oncology and metabolic disease.