Structure–activity relationship expansion and microsomal stability assessment of the 2-morpholinobenzoic acid scaffold as antiproliferative phosphatidylcholine-specific phospholipase C inhibitors

Abstract

Dysregulation of choline phospholipid metabolism and overexpression of phosphatidylcholine-specific phospholipase C (PC-PLC) is implicated in various cancers. Current known enzyme inhibitors include compounds based on a 2-morpholino-5-N-benzylamino benzoic acid, or hydroxamic acid, scaffold. In this work, 81 compounds were made by modifying this core structure to explore the pharmacophore. Specifically, these novel compounds result from changes to the central ring substitution pattern, alkyl heterocycle and methylation of the N-benzyl bridge. The anti-proliferative activity of the synthesised compounds was assessed against cancer cell lines MDA-MB-231 and HCT116. PC-PLC_BC enzyme inhibition was also assessed, and the development of a pharmacokinetic profile was initiated using a microsomal stability assay. The findings confirmed the optimal pharmacophore as a 2-morpholino-5-N-benzylamino benzoic acid, or acid derivative, scaffold, and that this family of molecules demonstrate a high degree of stability following treatment with rat microsomes. Additionally, benzylic N-methylated compounds were the most biologically active compounds, encouraging further investigation into this region of the pharmacophore.

Novel analogues of PC-PLC inhibitors were studied, with the 1-acyl-2-morpholino-5-N-benzylamine central scaffold found to be the optimal pharmacophore. This family demonstrates a high degree of stability following treatment with rat microsomes.

Introduction

Disruption of choline phospholipid metabolism has been identified as a potential biomarker for diagnosing cancer and measuring malignant progression in cancer patients.¹ The specific biomarkers measured are phosphocholine and total choline concentration, both of which are metabolites of phospholipase enzymes, including phosphatidylcholine-specific phospholipase C (PC-PLC). PC-PLC is responsible for metabolising phosphatidylcholine, an abundant mammalian phospholipid, into the secondary messengers phosphocholine and 1,2-diacylglycerol (DAG), both of which have been implicated in neoplasia and metastatic progression.^1–5 The majority of oncogenic activity associated with PC-PLC is attributed to DAG, which has downstream stimulatory effects on biological pathways involved in cellular proliferation such as PKC, NF-κB and mitogen-activated protein kinases (MAPK).^2,3,6–10

In addition to being implicated in activating various oncogenic cell signalling pathways, PC-PLC has also shown to be upregulated in multiple cancer cell lines, demonstrating a six-fold increase in MDA-MB-231 (breast cancer) cells,¹¹ five-fold in OVCAR3 (ovarian cancer) cells,⁴ and two-fold in A431 (epidermoid cancer) cells.⁵ Furthermore, it was demonstrated that in all three cell lines, intervention via PC-PLC inhibition was able to significantly reduce proliferation, and in the case of MDA-MB-231 cells, significantly reduce cell motility and reverse other metastatic traits.^4,11 Moreover, evidence of physical interaction between PC-PLC and HER2 receptors has been reported and PC-PLC inhibition has been shown to induce HER2 internalisation and degradation.^12,13

Due to this evidence establishing PC-PLC as a promising target for novel cancer therapies, our research group has conducted previous investigations into finding novel drug-like PC-PLC inhibitors using a combination of computational molecular modelling and traditional medicinal chemistry techniques.^14–16 These investigations have identified compounds containing a 2-morpholinobenzoic acid scaffold as promising lead compounds that demonstrate good efficacy in both PC-PLC_BC inhibitory and antiproliferative assays. To date, a range of 2-morpholinobenzoic acid derivatives have been synthesised bearing 1-acyl-2-morpholino-5-N-benzyl substituents on a central aromatic ring (Fig. 1A). It was found that compounds 1a–d, bearing a 1-carboxylic acid group, 2-morpholino group, and halogen substituents on the 5-N-benzyl ring were strong inhibitors of PC-PLC_BC (Fig. 1B),¹⁴ and analogous compounds 2a–d bearing a 1-hydroxamic acid moiety were potent anti-proliferative agents in MDA-MB-231 and HCT116 cancer cell lines (Fig. 1C).¹⁵ Furthermore, it was determined that both classes of compounds 1 and 2, were more potent inhibitors of PC-PLC_BC than the literature standard at the time, D609 (tricyclodecan-9-yl-xanthogenate).^14,15

Fig. 1. Summary of previous PC-PLC_BC inhibitors derived from 2-morpholinobenzoic acid.^14–16 (A) General structure showing areas of modification; (B) structures of 2-morpholinobenzoic acid analogues exhibiting the greatest inhibition of PC-PLC_BC enzyme activity; (C) structures of 2-morpholinohydroxamic acid analogues with the most potent anti-proliferative activity against HCT116 and MDA-MB-231 cells.

In this work, the 2-morpholinobenzoic acid family was targeted for further exploration through modification of the substituent pattern around the central aromatic ring. Specifically, the 1,2,5-relationship of the acyl group, morpholine ring and N-benzylamine group, respectively, found in previous compounds from this class would be altered to a 1,2,4-relationship to form a new series of 2-morpholino-4-N-benzylamine compounds 3 (Fig. 2A). A second series of 1-acyl-2-tetrahydropyranyl-5-N-benzylamine compounds 4 (Fig. 2B) was also targeted, with these compounds possessing a tetrahydropyranyl (THP) ring at the 2-position rather than a morpholine ring, in order to investigate the effect of replacing the tertiary nitrogen atom.

Fig. 2. Proposed modifications to the original 2-morpholino-5-N-benzylamine structure in this research. (A) General structure of 2-morpholinio-4-N-benzylamine 3 derivatives. (B) General structure of 2-tetrahydropyranyl-5-N-benzylamine 4 derivatives.

In silico molecular docking simulations indicated that with the planned modifications, co-ordination to the catalytic Zn246 ion would be maintained, along with π-stacking interactions between Phe-66 and the central aromatic ring (see ESI† 2). The validity of these predictions was to be assessed through in vitro determination of PC-PLC_BC inhibition following the synthesis of the new series 3 and 4. Biological screening would also include evaluation of anti-proliferative activity and subsequent microsomal stability assessment of select compounds. The latter investigation was included as it was thought prudent that an initial pharmacokinetic profile should start to be outlined, due to this class of compounds now becoming an established family of drug-like anti-cancer molecules.

Results and discussion

Synthesis of series 3 and 4 compounds

Syntheses of the two series of targeted 2-morpholino-4-N-benzylamine derivatives 3 (R = OMe and R = OH) were achieved through adaptation of methods for structurally related compounds (Scheme 1).¹⁵ Beginning from 2-chloro-4-nitrobenzoic acid 5, an Ullmann-coupling was employed to install the morpholine ring affording intermediate 6,¹⁷ which was then subjected to a Fischer esterification in methanol to afford methyl benzoate 7. Subsequent reduction of the nitro-group on 7 gave aniline 8, which was then used for a sequence of two-step reductive amination reactions with benzaldehydes 9a–m, installing the 4-N-benzylamine moiety and giving compounds 10a–m. Following this, methyl benzoates 10a–m were then saponified to afford the desired benzoic acid derivatives 11a–m.

Scheme 1. Synthesis of 2-morpholino-4-N-benzyl compounds containing ester 10 and carboxylic acid 11 motifs. Reagents and conditions: (i) morpholine (3 eq.), K₂CO₃ (2 eq.), l-proline (0.2 eq.), CuI (0.1 eq.), DMSO, 130 °C, 24 h, 74% (ii) H₂SO₄ (cat.), MeOH, reflux, 24 h, 80%; (iii) 10% Pd/C (10% w/w), H₂ (3 atm), EtOAc, r.t., 24 h, 88%; (iv) 9a–m (1.3 eq.), 4 Å molecular sieves (1 g mmol⁻¹), CH₂Cl₂, r.t., 24 h; (v) NaBH₄ (3 eq.), MeOH/THF (2 : 3), r.t., 24 h. 44–84% (over two steps); (vi) 1 M aq. NaOH (5 eq.) THF/MeOH, 60 °C, 24 h. 67–95%.

An alternate pathway was used to synthesise the targeted hydroxamic acid variants of the 2-morpholino-4-N-benzylamine derivatives 3, where R = NHOH, 12a–m. This alternative was implemented as the coupling of a THP-protected hydroxylamine directly to carboxylic acids 11 proved to be difficult, likely due to a decrease in electrophilicity at the reactive centre as a result of electron donating ortho and para-amino groups. To overcome this issue, an amide coupling using coupling reagent (1-cyano-2-ethoxy-2-oxoethylidenaminooxy)dimethylamino-morpholino-carbenium hexafluorophosphate (COMU) was performed on benzoic acid 6, bearing an electron-withdrawing 4-nitro group, affording O-tetrahydropyranyl-protected N-hydroxybenzamide 13, which was then subjected to catalytic hydrogenation to afford aniline 14 (Scheme 2). Aniline 14 was then utilised for the reductive amination with benzaldehydes 9a–m, giving N-benzylamines 15a–m and subsequent deprotection in mild acidic conditions to afford the desired hydroxamic acids 12a–m.

Scheme 2. Synthesis of 1,2,4-hydroxamic acids 12a–m from nitrobenzoic acid intermediate 6. Reagents and conditions; (i) NH₂OTHP (1.5 eq.), COMU (1.3 eq.), DIPEA (3 eq.), DMF, r.t., 65 h, 84%; (ii) 10% Pd/C (10% w/w), H₂ (3 atm), MeOH, 0 °C, 45 min, 46%; (iii) 9a–m (5 eq.), EtOH, r.t., 24 h; (iv) NaBH₃CN (3.5 eq.), THF/MeOH (1 : 1), r.t., 24 h, 50–92% (over two steps); (v) 2 M HCl (0.5 eq.), THF/MeOH (1 : 1), r.t., 24 h, 50–90%.

Following this, synthesis of 2-tetrahydropyranyl-5-N-benzylamine 4 derivatives was conducted. This route began from methyl 2-chloro-5-nitrobenzoate 16,¹⁵ where first a Suzuki coupling was employed using 3,6-dihydro-2H-pyran-4-boronic acid pinacol ester to install a dihydropyran ring, affording compound 17 (Scheme 3). A palladium-catalysed hydrogenation was then carried out to reduce both the dihydropyran ring and 5-nitro group of 17, giving tetrahydropyranyl aniline 18. Akin to the previous series, aniline 18 could then be used for reductive aminations with benzaldehydes 9a–m, affording N-benzylamines 19a–m. The methyl ester on 19a–m was then hydrolysed to give benzoic acids 20a–m, which were then successively subjected to amide couplings with O-THP protected hydroxylamine then acid deprotection of the THP group to afford the hydroxamic acids 21a–m.

Scheme 3. Synthesis of 1,2,5-tetrahydropyranyl series 20a–m and 21a–m. reagents and conditions; (i) 3,6-dihydro-2H-pyran-4-boronic acid pinacol ester (1.5 eq.), K₂CO₃ (1.3 eq.), Pd(dppf)Cl₂ (0.1 eq.), 1,4-dioxane/H₂O (5 : 1), 60 °C, 22 h, 86%; (ii) Pd/C (10% w/w), H₂ (3 atm), EtOAc, r.t., 48 h, 96%; (iii) 9a–m (1.2 eq.), EtOH, r.t., 24 h; (iv) NaBH₄ (5 eq.), THF/MeOH (1 : 1), r.t., 24 h, 77–95% (over two steps); (v) 1 M aq. NaOH (4 eq.), MeOH/THF (1 : 1), 80 °C, 2 h, 85–99%; (vi) NH₂OTHP (1.5 eq.), COMU (1.3 eq.), DIPEA (3 eq.), DMF, r.t., 48 h, 78–99%; then 2 M aq. HCl (0.5 eq.), THF/MeOH (1 : 1), r.t., 24 h, 62–95%.

Biological assessment

In vitro PC-PLC_BC inhibitory activity of series 3 and 4

The next step was assessment of in vitro inhibition of PC-PLC_BC of the synthesised compounds, using an Amplex-Red based fluorescence assay.^15,18,19 Of the 2-morpholino-4-N-benzylamine 3 compounds, the most active were those containing a carboxylic acid moiety (11f, 11g), with a small set of ester-functionalised compounds also illustrating comparable PC-PLC_BC inhibitory activity (10g, 10h). The hydroxamic acid bearing compounds 12a–m consistently demonstrated a significant (p-value < 0.05) level of PC-PLC_BC inhibition at 40–60% compared to no treatment, which is a level similar to D609, a known literature standard for PC-PLC inhibition (Table 1).^14,15

Table 1. Amplex-red assay results showing enzymatic activity (%) of PC-PLC_BC following treatment with 2-morpholino-4-N-benzylamine compounds: series 10–12 (10 μM). Data are percentages of enzyme activity (%) relative to vehicle DMSO and are reported as mean ± s.e.m. of three independent experiments performed in triplicate. * p-value < 0.05, ** p-value < 0.01, *** p-value < 0.001 compared to vehicle by one-way ANOVA with a post hoc Dunnett's test.

N-benzyl substitution	Carboxylic acid 11	Ester 10	Hydroxamic acid 12
H (a)	96.6 ± 7.1	101.8 ± 10.7	43.4 ± 4.4**
2-F (b)	82.2 ± 8.8	105.2 ± 3.4	50.9 ± 1.5*
3-F (c)	84.9 ± 1.2	78.9 ± 8.4	49.1 ± 0.7**
4-F (d)	62.4 ± 2.7	80.3 ± 1.9	49.2 ± 4.0**
2-Cl (e)	50.8 ± 4.7*	75.9 ± 4.3	54.1 ± 2.8*
3-Cl (f)	33.1 ± 5.7***	55.8 ± 9.4*	47.5 ± 0.8**
4-Cl (g)	32.4 ± 5.0***	41.3 ± 5.2**	54.4 ± 1.4*
2-Br (h)	95.2 ± 20.1	36.3 ± 6.4***	54.7 ± 1.1*
3-Br (i)	95.8 ± 20.4	97.4 ± 17.8	45.3 ± 2.0**
4-Br (j)	99.1 ± 19.9	103.2 ± 17.8	45.6 ± 2.8**
2-OMe (k)	104.9 ± 15.4	104.4 ± 15.0	46.9 ± 2.5**
3-OMe (l)	105.8 ± 10.2	109.2 ± 15.5	51.4 ± 0.6*
4-OMe (m)	105.9 ± 5.6	103.8 ± 10.3	45.5 ± 5.1**
1ba,b	10.7 ± 1.5***
2ba,c	30.5 ± 7.3***
D609 d	49.8 ± 14.0*

^aPreviously synthesised best-in-class compounds,¹⁵ provided for comparison.

^bStructure shown in Fig. 1; IUPAC name 5-((3-chlorobenzyl)amino)-2-morpholinobenzoic acid.

^cStructure shown in Fig. 1; IUPAC name 5-((3-chlorobenzyl)amino)-N-hydroxy-2-morpholinobenzamide.

^dLiterature standard D609; IUPAC name potassium O-(octahydro-1H-4,7-methanoinden-5-yl) carbonodithioate.

The best-in-class compounds, carboxylic acid 1b and hydroxamic acid 2b, which comprised a 2,5-substitution pattern between morpholine and N-benzyl groups, could be compared to their respective 2,4-substituted analogues 11f and 12f, respectively, due to the conserved 3-Cl substitution at the benzyl ring. This comparison showed altering the ring pattern to a 2,4-relationship resulted in a 3-fold decrease in inhibition in carboxylic acid-bearing compounds, from 10.7 ± 1.5% to 33.1 ± 5.7% enzyme activity relative to vehicle (89.3% to 66.9% inhibition). In hydroxamic acids the same effect occurred albeit to a lesser degree, resulting in a 1.5-fold decrease in enzyme inhibition, from an enzyme activity of 30.5 ± 7.3% to 47.5 ± 0.8%.

2-Tetrahydropyranyl-5-N-benzylamine derivatives, 19, 20 and 21 were poorer inhibitors of PC-PLC compared with the morpholine series 3, with few molecules demonstrating significant inhibition compared to the control (Table 2). In comparison to the respective morpholine analogues 1b and 2b, carboxylic acid 20f containing a THP moiety showed full loss of inhibitory activity. The THP-hydroxamic acid 21f showed a 2.5-fold decrease in inhibition potency, the enzyme activity increasing from 30.5 ± 7.3% to 79.7 ± 0.7%. These findings strongly suggest that the morpholinyl nitrogen is essential for inhibitory activity against PC-PLC_BC.

Table 2. Amplex-red assay results showing enzymatic activity (%) of PC-PLC_BC following treatment with 2-tetrahydropyranyl-5-N-benzylamine compounds: series 19–21 (10 μM). Data are percentages of enzyme activity (%) relative to vehicle DMSO and are reported as mean ± s.e.m. of three independent experiments performed in triplicate. * p-value < 0.05, ** p-value < 0.01, *** p-value < 0.001 compared to vehicle by one-way ANOVA with a post hoc Dunnett's test.

N-benzyl substitution	Carboxylic acid 20	Ester 19	Hydroxamic acid 21
H (a)	54.0 ± 8.7*	77.2 ± 6.2	80.7 ± 5.4
2-F (b)	43.4 ± 7.7**	62.3 ± 7.6	94.0 ± 4.7
3-F (c)	36.2 ± 5.8***	44.5 ± 6.3**	88.2 ± 1.5
4-F (d)	101.6 ± 19.5	52.0 ± 14.7*	93.7 ± 2.1
2-Cl (e)	106.6 ± 17.3	100.9 ± 20.0	82.8 ± 1.1
3-Cl (f)	108.8 ± 15.5	104.2 ± 19.0	79.7 ± 0.7
4-Cl (g)	112.2 ± 14.1	111.9 ± 13.5	92.5 ± 2.2
2-Br (h)	116.3 ± 2.8	111.7 ± 12.0	75.5 ± 1.1
3-Br (i)	112.9 ± 1.5	109.1 ± 9.4	75.2 ± 1.4
4-Br (j)	105.2 ± 3.5	110.4 ± 4.6	71.4 ± 2.6
2-OMe (k)	107.6 ± 0.9	94.7 ± 5.0	80.7 ± 5.3
3-OMe (l)	106.1 ± 4.0	85.5 ± 7.2	80.6 ± 2.9
4-OMe (m)	103.9 ± 2.9	75.4 ± 11.0	80.4 ± 3.5
1ba,b	10.7 ± 1.5***
2ba,c	30.5 ± 7.3***
D609 d	49.8 ± 14.0*

^aPreviously synthesised best-in-class compounds,¹⁵ provided for comparison.

^bStructure shown in Fig. 1; IUPAC name 5-((3-chlorobenzyl)amino)-2-morpholinobenzoic acid.

^cStructure shown in Fig. 1; IUPAC name 5-((3-chlorobenzyl)amino)-N-hydroxy-2-morpholinobenzamide.

^dLiterature standard D609; IUPAC name potassium O-(octahydro-1H-4,7-methanoinden-5-yl) carbonodithioate.

Neither the 2-morpholino-4-N-benzylamine 3 compounds 10, 11, 16 nor the 2-tetrahydropyranyl-5-N-benzylamine 4 compounds 19, 20, 21 were able to achieve a level of inhibition comparable to the previously published 2–5-carboxylic acids,^14,15 indicating the original arrangement (Fig. 1, as present in 1 and 2) remains the optimal known substitution pattern for maximising PC-PLC_BC inhibitory activity.

Antiproliferative activity of series 3 and 4

The anti-proliferative activity of the compound library was assessed using a ³H-thymidine incorporation assay, in which radioactive ³H-thymidine is incorporated into the DNA of dividing cells and therefore a higher percentage of incorporation indicates increased cell growth. The desired outcome is a decreased value, marking inhibition of proliferation. This was tested in triple negative breast cancer cell line, MDA-MB-231 and colorectal cancer cell line HCT116, to align with previous work.^14,15

Considering the 2-morpholino-4-N-benzylamine 3 compounds, acids 11a–m showed minimal activity, with no significant difference in cell proliferation compared to the vehicle. Similarly, literature standard D609 also showed no notable anti-proliferative activity at the 10 μM concentration used for this study. Esters 10a–m exhibited moderate activity, the most potent of these being 10h with 2-bromo substitution on the benzyl ring, which reduced cell proliferation to 51.3 ± 13.3% in MDA-MB-231 cells and 29.5 ± 8.6% in HCT116 cells (Tables 3 and 4). The greatest degree of inhibition was observed in the hydroxamic acid series 12 which were capable of limiting cell proliferation to 21.8 ± 5.5% in MDA-MB-231 cells and 18.1 ± 5.6% in HCT116 cells (4-bromo analogue 12j). It was also seen that anti-proliferative activity did vary according to the substitution on the benzyl ring; 12a (R = H) was approximately four times less effective than 12j. Preference was seen for bromo-substituted compounds which resulted in cell viability ranging from 21.8–30%. The improved potency observed in series 12 was proposed to not be attributed to histone deacetylase (HDAC) binding with the presence of the hydroxamic acid moiety, as all of the synthesised compounds lacked structural homology to known HDAC inhibitors.²⁰ The molecular fingerprints of 12a–m were compared to those of FDA-approved small molecule inhibitors vorinostat, belinostat and panobinostat. Low Tanimoto scores were returned for all analogues, with a maximum of 0.241 between 12a and vorinostat (Table S5†). A score of 0 indicates no structural similarity whilst a score of 1 represents structurally identical molecules.

Table 3. ³H-Thymidine incorporation assay results showing cell proliferation (%) in MDA-MB-231 cells after treatment with 2-morpholino-4-N-benzylamine compounds (series 10–12) at 10 μM. Data are percentages of cell activity (%) relative to vehicle DMSO and are reported as mean ± s.e.m. of three independent experiments performed in triplicate. * p-value < 0.05, ** p-value < 0.01, *** p-value < 0.001 compared to vehicle by one-way ANOVA with a post-hoc Dunnett's test.

N-benzyl substitution	Carboxylic acid 11	Ester 10	Hydroxamic acid 12
H (a)	96.8 ± 4.3	95.2 ± 1.4	85.7 ± 5.3
2-F (b)	96.4 ± 5.1	92.8 ± 3.5	79.6 ± 4.8
3-F (c)	92.4 ± 1.3	85.9 ± 2.5	76.2 ± 4.7*
4-F (d)	94.8 ± 2.0	81.9 ± 2.5	81.6 ± 5.9
2-Cl (e)	88.0 ± 10.4	70.4 ± 3.7*	41.8 ± 2.1***
3-Cl (f)	87.2 ± 4.1	68.9 ± 8.0 **	39.6 ± 2.2***
4-Cl (g)	88.4 ± 2.2	70.3 ± 5.1*	45.5 ± 2.6***
2-Br (h)	90.0 ± 4.7	51.3 ± 13.3***	30.0 ± 0.8***
3-Br (i)	104.3 ± 3.4	62.9 ± 6.4***	25.3 ± 2.8***
4-Br (j)	99.7 ± 5.4	68.9 ± 7.7**	21.8 ± 5.5***
2-OMe (k)	92.6 ± 5.1	73.6 ± 5.3*	70.7 ± 3.8**
3-OMe (l)	101.4 ± 6.9	80.6 ± 0.9	65.7 ± 2.8***
4-OMe (m)	109.9 ± 3.6	87.5 ± 4.8	76.9 ± 1.4*
2aa,b	11.7 ± 1.9***
2ca,c	7.4 ± 1.0***
D609 d	99.0 ± 2.1

^aPreviously synthesised best-in-class compounds,¹⁵ provided for comparison.

^bStructure shown in Fig. 1; IUPAC name 5-((2-chlorobenzyl)amino)-N-hydroxy-2-morpholinobenzamide.

^cStructure shown in Fig. 1; IUPAC name 5-((2-bromobenzyl)amino)-N-hydroxy-2-morpholinobenzamide.

^dLiterature standard D609; IUPAC name potassium O-(octahydro-1H-4,7-methanoinden-5-yl) carbonodithioate.

Table 4. ³H-Thymidine incorporation assay results showing cell proliferation (%) in HCT-116 cells after treatment with 2-morpholino-4-N-benzylamine compounds (series 10–12) at 10 μM. Data are percentages of cell activity (%) relative to vehicle DMSO and are reported as mean ± s.e.m. of three independent experiments performed in triplicate. * p-value < 0.05, ** p-value < 0.01, *** p-value < 0.001 compared to vehicle by one-way ANOVA with a post-hoc Dunnett's test.

N-benzyl substitution	Carboxylic acid 11	Ester 10	Hydroxamic acid 12
H (a)	92.4 ± 5.1	85.4 ± 1.6	89.7 ± 0.7
2-F (b)	95.2 ± 7.9	67.4 ± 3.8***	87.6 ± 1.8
3-F (c)	93.3 ± 5.5	71.0 ± 4.4**	84.3 ± 2.7
4-F (d)	91.6 ± 7.7	65.0 ± 4.7***	91.4 ± 2.3
2-Cl (e)	87.6 ± 8.0	43.2 ± 4.5***	68.4 ± 2.9***
3-Cl (f)	88.6 ± 8.0	44.3 ± 7.9***	48.6 ± 5.4***
4-Cl (g)	88.2 ± 7.3	43.8 ± 3.4***	57.4 ± 5.0***
2-Br (h)	94.4 ± 1.2	29.5 ± 8.6***	54.5 ± 6.1***
3-Br (i)	101.1 ± 3.3	34.7 ± 2.6***	32.3 ± 5.9***
4-Br (j)	101.1 ± 1.3	52.5 ± 3.1***	18.1 ± 5.6***
2-OMe (k)	102.0 ± 1.5	55.5 ± 3.1***	83.2 ± 2.7
3-OMe (l)	102.6 ± 1.0	48.5 ± 0.6***	77.4 ± 4.6**
4-OMe (m)	105.3 ± 1.3	70.4 ± 7.6**	91.5 ± 2.7
2aa,b	2.0 ± 0.4***
2da,c	1.9 ± 0.2***
D609 d	98.4 ± 3.2

^aPreviously synthesised best-in-class compounds,¹⁵ provided for comparison.

^bStructure shown in Fig. 1; IUPAC name 5-((2-chlorobenzyl)amino)-N-hydroxy-2-morpholinobenzamide.

^cStructure shown in Fig. 1; IUPAC name 5-((3-bromobenzyl)amino)-N-hydroxy-2-morpholinobenzamide.

^dLiterature standard D609; IUPAC name potassium O-(octahydro-1H-4,7-methanoinden-5-yl) carbonodithioate.

Despite select compounds displaying good activity, none of the 2-morpholino-4-N-benzylamines 3 compounds matched the activity of the previous 1,2,5-substituted hydroxamic acids 2a and 2c. This indicates that moving the N-benzyl ring from the 5- to 4-position has a negative effect on activity and the optimal pharmacophore is the 1,2,5 arrangement.

Compounds containing a THP moiety, 19, 20, and 21, were also assessed for anti-proliferative activity and in both cell lines the carboxylic acids 20a–m were entirely inactive in inhibiting cell growth (Tables 5 and 6). Slight growth inhibition could be observed in hydroxamic acids 21, the most potent compound again being 21j with 4-bromo substitution, the same substitution pattern that was best in the morpholine-containing compounds. Of the THP-containing compounds the ester subseries 19 performed very well. In the MDA-MB-231 cell line, 19h (2-Br) and 19b (2-F) were the most active, with respective activities of 16.4 ± 2.3 and 17.6 ± 2.4%. In HCT116 cells, 19b achieved strong growth inhibition, resulting in only 9.5 ± 0.1% ³H-thymidine incorporation, 19h was next best in inducing cell proliferation, with 20.3 ± 1.0%. Similarly to series 12 compounds which also contained a hydroxamic acid group, 19a–m were structurally compared to known HDAC inhibitors but again all demonstrated low Tanimoto coefficients, the maximum being 0.221 between 19m and vorinostat (Table S6†).

Table 5. ³H-Thymidine incorporation assay results showing cell proliferation (%) in MDA-MB-231 cells after treatment with 2-tetrahydropyranyl-5-N-benzylamine 4 compounds (series 19–21) at 10 μM. Data are percentages of cell activity (%) relative to vehicle DMSO and are reported as mean ± s.e.m. of three independent experiments performed in triplicate. * p-value < 0.05, ** p-value < 0.01, *** p-value < 0.001 compared to vehicle by one-way ANOVA with a post hoc Dunnett's test.

N-benzyl substitution	Carboxylic acid 20	Ester 19	Hydroxamic acid 21
H (a)	107.0 ± 4.4	29.3 ± 4.6***	90.0 ± 2.6
2-F (b)	112.8 ± 7.2	17.6 ± 2.4***	88.0 ± 2.9
3-F (c)	104.5 ± 8.6	41.3 ± 1.1***	84.4 ± 2.4
4-F (d)	107.3 ± 9.1	36.0 ± 2.6***	86.6 ± 3.2
2-Cl (e)	110.6 ± 8.6	25.9 ± 5.0***	86.8 ± 1.5
3-Cl (f)	113.7 ± 5.0	48.3 ± 5.8***	80.1 ± 1.3
4-Cl (g)	114.6 ± 4.5	83.8 ± 10.2	72.2 ± 1.3**
2-Br (h)	107.9 ±1.9	16.4 ± 2.3***	74.0 ± 1.4**
3-Br (i)	107.4 ± 2.7	58.8 ± 12.4***	77.6 ± 2.7*
4-Br (j)	102.0 ± 3.4	79.6 ± 5.4	61.6 ± 2.2***
2-OMe (k)	100.5 ± 4.6	22.0 ± 2.8***	80.7 ± 1.7
3-OMe (l)	104.3 ± 2.2	83.9 ± 4.2	89.2 ± 3.3
4-OMe (m)	102.3 ± 3.1	71.2 ± 5.6**	82.0 ± 1.5
2aa,b	11.7 ± 1.9***
2ca,c	7.4 ± 1.0***
D609 d	99.0 ± 2.1

^aPreviously synthesised best-in-class compounds,¹⁵ provided for comparison.

^bStructure shown in Fig. 1; IUPAC name 5-((2-chlorobenzyl)amino)-N-hydroxy-2-morpholinobenzamide.

^cStructure shown in Fig. 1; IUPAC name 5-((2-bromobenzyl)amino)-N-hydroxy-2-morpholinobenzamide.

^dLiterature standard D609; IUPAC name potassium O-(octahydro-1H-4,7-methanoinden-5-yl) carbonodithioate.

Table 6. ³H-Thymidine incorporation assay results showing cell proliferation (%) in HCT-116 cells after treatment with 2-tetrahydropyranyl-5-N-benzylamine 4 compounds (series 19–21) at 10 μM. Data are percentages of cell activity (%) relative to vehicle DMSO and are reported as mean ± s.e.m of three independent experiments performed in triplicate. * p-value < 0.05, ** p-value < 0.01, *** p-value < 0.001 compared to vehicle by one-way ANOVA with a post-hoc Dunnett's test.

N-benzyl substitution	Carboxylic acid 20	Ester 19	Hydroxamic acid 21
H (a)	99.8 ± 8.0	19.8 ± 4.8***	80.2 ± 2.3*
2-F (b)	103.6 ± 3.6	9.5 ± 0.1***	98.6 ± 2.8
3-F (c)	103.0 ± 2.9	29.2 ± 1.5***	94.4 ± 3.6
4-F (d)	103.8 ± 3.1	25.3 ± 4.1***	91.6 ± 3.4
2-Cl (e)	100.8 ± 4.1	25.9 ± 1.4***	86.7 ± 4.6
3-Cl (f)	103.1 ± 3.3	37.2 ± 1.2***	83.0 ± 3.5
4-Cl (g)	101.3 ± 3.3	56.7 ± 2.3***	69.9 ± 4.3***
2-Br (h)	99.5 ± 2.8	20.3 ± 1.0***	78.3 ± 5.3*
3-Br (i)	104.3 ± 0.1	40.1 ± 4.2***	78.8 ± 6.2*
4-Br (j)	103.8 ± 0.4	61.2 ± 1.7***	60.1 ± 7.1***
2-OMe (k)	106.0 ± 0.1	31.2 ± 2.6***	93.3 ± 3.9
3-OMe (l)	101.1 ± 0.3	62.5 ± 3.1***	94.4 ± 2.1
4-OMe (m)	93.6 ± 0.7	54.5 ± 3.5***	88.6 ± 4.1
2aa,b	2.0 ± 0.4***
2da,c	1.9 ± 0.2***
D609 d	98.4 ± 3.2

^aPreviously synthesised best-in-class compounds,¹⁵ provided for comparison.

^bStructure shown in Fig. 1; IUPAC name 5-((2-chlorobenzyl)amino)-N-hydroxy-2-morpholinobenzamide.

^cStructure shown in Fig. 1; IUPAC name 5-((3-bromobenzyl)amino)-N-hydroxy-2-morpholinobenzamide.

^dLiterature standard D609; IUPAC name potassium O-(octahydro-1H-4,7-methanoinden-5-yl) carbonodithioate.

This result of the ester-containing compounds performing the best is contradictory to previous known trends and in combination with the fact that these compounds performed poorly to moderately in the PC-PLC_BC inhibition assay (Table 2) suggests 19a–m may be acting on an alternate biological pathway to PC-PLC, thereby requiring further investigation.

Synthesis of N-methylated compounds

Overall, the biological assessment of series 3 and 4 compounds revealed that neither modification made to the original 1-acyl-2-morpholino-5-N-benzyl pattern (series 1 and 2) offered improvements in potency. Having completed exhaustive synthesis around the remainder of the central aromatic ring, it was decided to approach the N-benzylamine linker, which is an area of the pharmacophore that has been previously untouched.^14–16,19 To provide a few model examples of a potential new generation, three N-methylated compounds were synthesised with a 1,2,5-arrangement and 2-Cl group on the terminal aryl ring, as these moieties had proven important for potent PC-PLC_BC inhibitory and antiproliferative activity.¹⁵

Synthesis of N-methylated compounds was initiated by treating 22 with potassium carbonate and iodomethane yielding the desired N-methylated methyl benzoate 23, which underwent subsequent hydrolysis, affording N-methyl benzoic acid 24 (Scheme 4). Amide coupling of 24 with O-tetrahydropyranyl protected hydroxylamine using COMU afforded an O-THP protected N-hydroxybenzamide intermediate, which was deprotected using mild acidic conditions to afford the hydroxamic acid derivative 25, finalizing the N-methylated series.

Scheme 4. Synthesis of N-Me analogue series. Reagents and conditions; (i) 22 (1 eq.), MeI (4 eq.), K₂CO₃ (2 eq.), MeCN, reflux, 18 h, 80%; (ii) 23 (1 eq.), 1 M aq. NaOH (5 eq.), MeOH/THF (1 : 1, 14 mL), 80 °C, 1.5 h, 96%; (iii) 24 (1 eq.), NH₂OTHP (1.5 eq.), COMU (1.3 eq.), DIPEA (3 eq.), DMF, r.t., 48 h; then 2 M aq. HCl (0.5 eq.), MeOH/THF (1 : 1), r.t., 24 h, 52% (over two steps).

In silico PC-PLC_BC inhibitory activity of synthesised compounds

The new N-methylated series was not part of the original set of proposed compound library and as such was only synthesised following the completion of the in vitro PC-PLC_BC inhibition assay (section 2.2.1). Consequently, enzyme data was not able to be collected for these compounds through the same in vitro approach. To predict how these compounds may perform as PC-PLC_BC inhibitors, they were docked into the binding site of the PC-PLC_BC crystal structure (PDB ID 1AH7).²¹ This approach has been used in previous classes of 2-morpholinobenzoic acid derivatives to examine and direct the synthesis of novel compounds through simulated protein-ligand interactions and binding poses.^14,16,22

All of the synthesised compounds in this study were screened using four genetic algorithms (ESI† part 2, Tables S1–S7). Chemscore was selected for further evaluation due to its incorporation of metal–ligand terms, deemed most suitable for metalloproteases such as PC-PLC_BC which contains a trimetallic catalytic site of Zn246, Zn247 and Zn248.^23,24 Zn246, also denoted Zn1, is responsible for catalytic activity whilst the other two Zn²⁺ are essential structural ions.²⁵

The biological mechanism of PC-PLC_BC can be used to identify potential key protein–ligand interactions and involves hydrolysis of the phosphate bond of phosphatidylcholine (PC). The substrate is stabilised by placement of the phosphocholine head group into a hydrophobic binding pocket lined by Glu4, Phe66 and Tyr56 residues.^25,26 Two mechanisms of hydrolysis are largely considered which both recruit active site waters and are initiated by Asp55 as a general base. Firstly, a direct water activation is proposed in which an activated nucleophilic water molecule W4 (ID: HOH258) attacks the phosphorous centre of PC. The second mechanism proposes the nucleophilic attack is initiated by a bridged hydroxide ion between zinc atoms Zn1 and Zn3, deemed W1 (HOH249).^25,26 The latter option is more recently accepted and is in line with other trinuclear zinc enzymes.²⁷

To identify favourable or disfavoured interactions, the highest and lowest performing compounds in the Amplex red assay were first evaluated. The most potent in vitro inhibitor synthesised in this work was 11g, with a 1,2,4-substitution pattern of the central ring comprising 1-carboxylic acid, 2-morpholinyl and 4-(4-Cl N-benzyl) groups. This analogue reduced enzyme activity to an average of 32.4%; comparatively, the previous best-in-class 2-morpholinobenzoic acid (1b) achieved 10.7% activity. Both 11g and 1b were analysed in silico (Fig. 3 and 4) to outline which interactions may be correlated with strong target inhibition.

Fig. 3. Binding pose and protein interactions of previous best-in-class compound 1b, as calculated using the Chemscore algorithm of GOLD suite software. Dashed lines indicate bonding interactions; hydrogen bonding (green), π–π stacking (dark pink), π–alkyl interaction (light pink), alkyl interaction (light green), water hydrogen bonding (blue). Zinc ions are depicted with ball-and-stick display, waters are depicted with line drawing display/free waters denoted as X.

Fig. 4. Binding pose and protein interactions of most potent in vitro inhibitor 11g, as calculated using the Chemscore algorithm of GOLD suite software. Dashed lines indicate bonding interactions; hydrogen bonding (green), π–π stacking (dark pink), π–alkyl interaction (light pink), alkyl interaction (light green), water hydrogen bonding (blue). Zinc ions are depicted with ball-and-stick display, waters are depicted with line drawing display/free waters denoted as X.

Potent PC-PLC_BC inhibitors in this study demonstrated disruption of active site water molecules, including unfavourable clashes in the calculated binding poses and the rotation of multiple water atoms. Given the good solvent accessibility of the binding site, along with favourable lipophilic and metal interactions between protein and ligand, these water clashes may be overcome by the displacement of waters from the active site. Certain water molecules are critical to the enzyme mechanism and their displacement could result in loss of activity. In compound 1b, unfavourable interactions were made with multiple active site waters including HOH249, 250, 341, and 452, which have been shown to be displaced upon binding of a literature inhibitor, as well as with proposed nucleophile HOH258 (Fig. 3).²⁴ Favoured water interactions, however, could promote substrate binding and stability through extra hydrogen bonding networks.

Importantly for catalytic activity, attractive charge interactions were made between the carboxylic acid moiety of 1b and all three of the binding site Zn²⁺ ions. Other key interactions include those between the ligand and Glu146 and Asp55 which block their ability to act as a general acid/base. Choline binding pocket residues provided further stability with lipophilic interactions; Glu4 through van der Waals forces, Tyr56 through hydrogen bonding to the morpholine oxygen, and Phe66 through π-stacking to the central aromatic ring, plus π–alkyl bonds to the morpholine group. The ChemPLP algorithm performed similarly, with an RMSD of 0.46 Å between algorithms supporting this binding pose.

The most potent inhibitor of the newly synthesised compounds, carboxylic acid 11g, made similar interactions to 1b, including clashes with active site waters, favourable Zn²⁺ and Trp1 attractive charges, and Phe66 π–π and π–alkyl interactions (Fig. 4). The N-benzyl substitution between 1b and 11g was changed from a 4-Cl to 3-Cl pattern and therefore this combined with the other structural changes resulted in an altered position of the benzyl group in the binding site, permitting Leu135 alkyl interactions and a carbon-hydrogen bond between the aromatic ring and Tyr79. However, with the different spatial arrangement, hydrogen bonding with Tyr56 was no longer possible via the morpholinyl oxygen, which may be a reason for the reduced activity of 11g compared to 1b.

Upon modelling 20h, the compound causing the highest enzyme activity in vitro and therefore having no inhibitory effect (>100%), some interactions were retained although at different distances due to conformational changes whilst Leu135, Asn134, Thr133 and Trp1 interactions were absent (ESI† part 2, Table S14). Notable three-dimensional differences between this compound and the inhibitors 1b and 11g were rotation of the carbonyl group that was observed in other low performing analogues as well as a flipped chair conformation of the alkyl ring (ESI† part 2, Fig. S4). Unfavourable water interactions were maintained but only one of the water molecules, HOH254, made favourable hydrogen bonding to stabilise the ligand. Furthermore, the modelling solutions did not induce rotation of water atoms in the docking runs.

The examples of each 2-morpholinobenzoic acid or ester class with 2-Cl substitution of the benzylic ring were next modelled along with methylated N-benzyl analogues. The traditional 1,2,5-morpholinyl benzoic acid scaffold was assessed containing the ester (22), acid (1a), and hydroxamic acid (2a) functionalities to provide reference for their corresponding N-methylated equivalents 23, 24, and 25. Despite a large decrease in hydrogen bonding arising from the loss of the aniline NH functionality (Table 7), the N-Me analogues were calculated to have similar free energy of binding (ΔG), suggesting alternative favourable interactions to largely offset this penalty. Overall, minimal change was seen between esters 22 and 23 and a slight decrease of binding affinity resulted from methylation of hydroxamic acid 2a, however, a notable improvement was observed in the free binding energy of carboxylic acid 1a from −49.84 to 55.63 kJ mol⁻¹ in 24a. In all N-methylated equivalents a lower rotatable bond score was predicted, indicating increased ligand stability and more drug-like character.

Table 7. Chemscore results of N-methylated compounds. Scores are the average of the top 3 docking solutions.

Compound	ΔG (kJ mol−1)	Fitness score	Hbond scorea	Rotatable bond score
N–H acid (1a)	−49.84	42.60	−18.96	1.12
N-Me acid (24)	−55.63	40.83	−6.03	1.09
N–H ester (22)	−43.85	41.86	−17.89	1.26
N-Me ester (23)	−42.93	33.45	0.00	1.21
N–H hydroxamic acid (2a)	−62.05	51.57	−17.56	1.60
N-Me hydroxamic acid (25)	−60.17	47.59	−9.00	1.55

^aWeighted Hbond term.

Visualisation of docking solutions revealed that introduction of the methyl group most often allowed more alkyl interactions to Phe70, His128 and His142 (Fig. 5). Both of these histidine residues directly ligand Zn2. Specifically in N-Me carboxylic acid 24, His142 also π-stacked to the central aromatic ring which may assist in stabilising the binding conformation overall. This increase in favourable protein–ligand interactions suggest that methylation of the aniline group may increase enzyme inhibition of PC-PLC_BC in carboxylic acid-containing compounds.

Fig. 5. Binding pose and protein interactions of N-methylated compound 24, as calculated using the Chemscore algorithm of GOLD suite software. Dashed lines indicate bonding interactions; hydrogen bonding (green), π–π stacking (dark pink), π–alkyl interaction (light pink), alkyl interaction (light green), water hydrogen bonding (blue). Zinc ions are depicted with ball-and-stick display, waters are depicted with line drawing display/free waters denoted as X.

Antiproliferative activity of N-methylated compounds

Anti-proliferative data was obtained following the same methods as previous series for compounds 23, 24 and 25 (Table 8). The hydroxamic acid-containing compound 25 was found to induce the highest degree of anti-proliferative activity, following the established trend for the 2-morpholinobenzoic acid family.¹⁵ In MDA-MB-231 cells, treatment with N-methylated 25 reduced the cell proliferation to 4.3 ± 4.6% relative to vehicle. For reference, the previous generation best-in-class compound 2c in this cell line resulted in 7.4 ± 1.0%. To directly compare between NH and N-Me analogues, 2a in MDA-MB-231 gave 11.7 ± 1.9%, and 2.0 ± 0.4% in HCT116. With regard to IC₅₀ values, 25 demonstrated low-micromolar concentrations of 1.1 ± 0.1 μM and 3.4 ± 0.3 μM in MDA-MB-231 cells and HCT116 cells, respectively. These results suggest that alteration of the N-benzyl bridge can afford compounds with similar biological activity to the best-in-class compounds such as 2a and 2d, and that further investigations into novel compounds with alternative benzylic linkers regions should be conducted.

Table 8. Anti-proliferative activity of N-methylated compounds 23, 24, 25 at 10 μM in MDA-MB-231 and HCT116 cell lines. Cell proliferative activities after treatment with best-in-class compounds 2a, 2d are provided for comparison.¹⁵ Data is presented as cell growth (%) relative to vehicle (DMSO) and is reported as the mean ± s.e.m of three independent experiments performed in triplicate.

R group	3H-thymidine uptake at 10 μM (%)		IC50 (μM)
	MDA-MB-231	HCT116	MDA-MB-231	HCT116
OMe (23)	83.8 ± 3.8	54.5 ± 7.8	—	—
OH (24)	95.5 ± 2.1	87.8 ± 5.7	—	—
NHOH (25)	4.3 ± 4.6	4.8 ± 2.3	1.1 ± 0.1	3.4 ± 0.3
2a (2-Cl HA)	11.7 ± 1.9	2.0 ± 0.4	2.9 ± 1.2	4.9 ± 1.4
2d (3-Br HA)	13.5 ± 1.6	1.9 ± 0.2	3.3 ± 1.0	5.8 ± 1.2

In total, 39 2-morpholino-4-N-benzylamine 3 compounds 10a–m, 11a–m, 12a–m and 39 2-terahydrohydropyranyl-5-N-benzyl 4 compounds 19a–m, 20a–m, 21a–m, combined with three N-Me analogues 23–25 afforded 81 novel drug-like compounds to be assessed for biological activity and microsomal stability.

Microsomal stability analysis

To assess drug metabolism, a stability assay was carried out using freshly prepared liver microsomes over a 1 h time period, with concentration measurements taken at 15-minute intervals. Compounds chosen to represent the different core scaffolds included the morpholino-containing 1a and 11e, and THP containing 20e, which all contained a carboxylic acid group on the central aromatic ring and a 2-chloro substituent on the N-benzyl ring. 7-Ethoxycoumarin (7-ETC) was used as a positive control and demonstrated a significant reduction in concentration at all time points of the 1 h time course (Fig. 6A). The negative control was conducted using 7-ETC with no NADPH co-factor present in the reaction and this demonstrated no significant metabolism after 1 h. The previous best-in-class PC-PLC inhibitor 1a demonstrated only a small degree of degradation (17%) after the 1 h time-period (Fig. 6B), and a high degree of stability was seen for 11e and 20e, both of which did not show significant metabolism after 1 h (Fig. 6C and D). Conversely, N-methylated 24 exhibited a high level of instability, with a notable drop even at the first 15-minute time point and had degraded by 94% by the end of the time-course (Fig. 6C). This result, in combination with the high degree of biological activity demonstrated by N-methylated compounds, suggest that while further investigation into alternative N-benzyl linker moieties is warranted, emphasis should be placed on improving the metabolic stability of these compounds with alternative linkers.

Fig. 6. Results from microsomal stability assays, displayed as concentration of tested compounds over time relative to initial concentration. (A) 7-Ethoxycoumarin (positive control) and negative control; (B) 1a; (C) 24; (D) 11e; (E) 20e; (F) comparative area under each curve. Values are reported as mean ± s.e.m. of three individual experiments performed in triplicate. For (A)–(E) * p-value < 0.05, ** p-value < 0.01, *** p-value < 0.001 by two-way ANOVA with post-hoc Dunnett's test comparing each time point to baseline (0 minutes) for each compound. For (F) * p-value < 0.05, ** p-value < 0.01, *** p-value < 0.001 by one-way ANOVA with post hoc Dunnett's test comparing the area under the curve of each compound with the negative control.

Conclusions

Overall, the core structure of the 2-morpholinobenzoic acid scaffold underwent two major forms of modification to gain new SAR information regarding PC-PLC_BC inhibition and anti-proliferative activity of this class. The first being a change in the relative position of the N-benzyl group from the 1̲-acyl-2̲-morpholino-5̲-N-benzyl substitution pattern seen in previous inhibitors (1 and 2), to the 1̲-acyl-2̲-morpholino-4̲-N-benzyl arrangement in the newly synthesised compounds (10–12). The second major modification involved substitution of the morpholine ring for a tetrahydropyranyl ring at the 2-position (19–21). Within each class, further modifications were made to the benzyl ring substituents and the acyl group to produce carboxylic acids, methyl esters, and hydroxamic acid moieties which afforded a total of 78 novel analogues: 10a–m, 11a–m, 12a–m, 19a–m, 20a–m, 21a–m. Finally, a small subset of compounds in which the benzylic aniline was methylated (23–25) was synthesised to contribute another 3 analogues to the 2-morpholinobenzoic acid library and explore potential future leads. Biological evaluation revealed that derivatives with the 1-acyl-2-morpholino-4-N-benzyl substitution pattern had diminished activity in both PC-PLC inhibition and anti-proliferative screening assays, when compared to previously published compounds containing the 1-acyl-2-morpholino-5-N-benzyl pattern. It was also determined that replacing the morpholine ring at the 2-position with a tetrahydropyranyl ring led to diminished PC-PLC inhibitory activity. A similar trend was visible in the anti-proliferative assays, with the majority of compounds synthesised in this study demonstrating diminished activity compared to previously published compounds. The exception to this being 19a–m, which unexpectedly showed significant cell growth inhibition, and N-methylated 25, which was the most potent anti-proliferative compound synthesised in this study. In silico molecular modelling found that many binding interactions were shared between the strongest PC-PLC_BC inhibitors in this study and previous potent derivatives, albeit with loss of a key interaction with Tyr56, which may be the cause of reduced PC-PLC inhibition in this study. N-Me compounds were also modelled and shown to form similar interactions to known strong PC-PLC_BC inhibitors. Additional microsomal stability assessment indicated that all series of compounds were highly stable, with the exception of the N-methylated compounds, which showed a half-life of <15 min. Despite lacking stability, the benzylic N-methylated compounds were found to be the most biologically active compounds in the study, which encourages further investigation into this region of the pharmacophore, with the primary goal of improving the stability of these potent anti-cancer molecules.

Experimental procedures

General synthetic details

All reactions were carried out using freshly dried AR grade solvents under an atmosphere of nitrogen unless otherwise specified. All chemical reagents were used as purchased from Sigma Aldrich or AK scientific. Thin-layer chromatography (TLC) was carried out on Merck silica gel F354 aluminium plates pre-coated with silica and flash chromatography was conducted with Silica Gel 60 (40–63 μM, 230–430 mesh ASTM) using solvent systems specified in the experimental procedures.

All novel compounds were characterized using ¹H NMR, ¹³C NMR, HSQC and HBMC NMR experiments in combination with high-resolution mass spectrometry (HRMS). All NMR spectra were recorded using a Bruker Avance DRX 400 MHz spectrophotometer at ambient temperature. Chemical shifts are reported relative to the solvent peaks of DMSO-d₆ (δ 2.52 for ¹H and δ 39.52 for ¹³C) or CDCl₃ (δ 7.26 for ¹H and δ 77.16 for ¹³C). ¹H NMR data is reported in ascending order using the following sequence: position (δ), relative integral, multiplicity (s, singlet; d, doublet; t, triplet; q, quartet; dd, doublet of doublets; m, multiplet; br s, broad singlet; br d, broad doublet), coupling constant (J, Hz), and proton assignment. ¹³C NMR data was reported in ascending order using the following sequence: position (δ), multiplicity (d, doublet), coupling constant (J, Hz), and the carbon assignment. NMR assignments were made using a combination of ¹H NMR, ¹³C NMR, HSQC and HMBC experiments. HRMS was carried out using electrospray ionisation (ESI) on a MicroTOF-Q mass spectrometer. Melting points were measured using a Reicher–Kofler block, with the capacity to measure melting points up to 230 °C; melting points reported in this study are uncorrected. IR spectra were recorded using a Perkin–Elmer Spectrum 1000 series Fourier Transform Infra-Red ATR spectrometer, the absorption maxima are expressed as wave numbers (cm⁻¹).

All compounds are numbered according to their systematic IUPAC names.

General synthetic procedures

The synthesis of previous generation inhibitors 1a–b and 2a–d used for comparison in this article has been previously reported.¹⁵

Synthesis of methyl 4-N-benzylamino-2-morpholinobenzoates 10a–m

To a stirred solution of methyl 4-amino-2-morpholinobenzoate 8 (1.25 mmol) and 4 Å molecular sieves (1 g mmol⁻¹ of 8) in CH₂Cl₂ (5 mL) was added the corresponding benzaldehyde 9a–m (1.3 equiv.) and the resulting mixture was then stirred for 24 h at room temperature. The mixture was filtered through a plug of Celite, and filtrate evaporated in vacuo to give the imine intermediate as a yellow residue. The imine was redissolved in THF/MeOH (1 : 1, 6 mL), NaBH₄ (3 equiv.) was added portion-wise at 0 °C and the suspension was allowed to warm to room temperature and left to stir overnight. Following this, the mixture was quenched with ice-cold water and extracted with ethyl acetate (3 × 20 mL). The combined organic layers were washed with H₂O (10 mL), brine (10 mL), dried (MgSO₄) and the solvent removed in vacuo to give a crude product, which was purified by flash chromatography to afford amine 10a–m.

Synthesis of 4-N-benzylamino-2-morpholinobenzoic acids 11a–m

To a stirred solution of 10a–m (0.75 mmol) in THF/MeOH (1 : 1, 5 mL) was added 1 M aq. NaOH (5 equiv.), the mixture was stirred at 60 °C for 24 h. The resulting solution was redissolved in H₂O (10 mL) and separated using a biphasic mixture of ethyl acetate/1 M NaOH (1 : 1, 30 mL). The aqueous layer was collected, and organic layer washed further with 1 M aq. NaOH (2 × 10 mL). The combined aqueous solutions were acidified to pH ∼3 using 2 M aq. HCl, resulting in the formation of a white precipitate, which was collected via vacuum filtration. The precipitate was washed with ice-cold water, then collected and dried in vacuo affording the desired carboxylic acid 11a–m, no further purification was required.

Synthesis of 4-benzylamino-2-morpholino-N-hydroxy-tetrahydropyran amides 15a–m

To a solution of 4-amino-2-morpholino-N-((tetrahydro-2H-pyran-2-yl)oxy)benzamide 14 (0.35 mmol) in EtOH (3 mL) was added the corresponding benzaldehyde 9a–m (5 equiv.) and the subsequent mixture was allowed to stir overnight at room temperature. The mixture was then evaporated in vacuo and the residue was triturated with petroleum ether (2 × 15 mL) to afford the imine intermediate, which was dissolved in THF (2.5 mL) and added dropwise to NaBH₃CN (3.5 equiv.) in MeOH (2.5 mL) at 0 °C. This mixture was allowed to warm to room temperature and stirred overnight, before quenching with H₂O (10 mL) at 0 °C. This subsequent suspension was extracted with ethyl acetate (3 × 15 mL), the combined organic layers were washed with H₂O (10 mL), brine (15 mL), dried (Na₂SO₄) and evaporated in vacuo to give a crude residue, which was purified by flash chromatography, affording the desired amide intermediate 15a–m.

Synthesis of 4-benzylamino-2-morpholino-N-hydroxybenzamides 12a–m

Following this, the amide intermediate 15a–m was dissolved in THF/MeOH (1 : 1, 5 mL) and treated with 2 M aq. HCl (0.5 equiv.) and left to stir overnight at room temperature. The mixture was then evaporated in vacuo, redissolved in 1 M aq. NaOH (10 mL), washed with diethyl ether (2 × 10 mL) and the aqueous layer collected. The aqueous layer was then adjusted to pH 4 with 2 M aq. HCl and extracted with chloroform (4 × 15 mL) and the combined organic layers were washed with brine (10 mL), dried (Na₂SO₄) and concentrated in vacuo. The resulting solid was triturated with 1 : 1, petroleum ether/diethyl ether (2 × 5 mL) to give the desired hydroxamic acid 12a–m, no further purification was required.

Synthesis of methyl 5-benzylamino-2-tetrahydropyranylbenzoates 19a–m

To a stirred solution of methyl 5-amino-2-(tetrahydro-2H-pyran-4′-yl)benzoate 18 (0.85 mmol) in anhydrous EtOH (5 mL) was added the corresponding benzaldehyde 9a–m (1.2 equiv.). The mixture was then stirred for 3 h at room temperature under a nitrogen atmosphere. Following this, the mixture was evaporated to dryness in vacuo, affording the imine intermediate as a white solid. NaBH₄ (5 equiv.) was added portion-wise to the imine in anhydrous MeOH/THF (1 : 1, 6 mL) at 0 °C under an atmosphere of nitrogen. The suspension was then warmed to room temperature and stirred overnight under an atmosphere of nitrogen_. Following this, the mixture was quenched with ice-cold water (10 mL) and extracted with ethyl acetate (3 × 20 mL), the combined organic layers were washed with H₂O (10 mL), brine (10 mL) and dried (Na₂SO₄). The solvent was removed in vacuo to give a crude product, which was purified by flash chromatography to afford amine 19a–m.

Synthesis of 5-benzylamino-2-tetrahydropyranylbenzoic acids 20a–m

1 M NaOH (4 equiv.) was added to a solution of ester 19a–m (0.45 mmol) in THF/MeOH (1 : 1, 6 mL) and mixture stirred at 80 °C for 2.5 h. The mixture was then concentrated in vacuo, dissolved in 1 M NaOH and washed with diethyl ether (2 × 10 mL). The aqueous layer was collected then acidified to ∼pH 4 with 2 M HCl forming a white precipitate, which was extracted using CHCl₃ (4 × 15 mL). The combined organic layers were dried (MgSO₄) and evaporated in vacuo to afford carboxylic acid 20a–m.

Synthesis of 5-benzylamino-2-tetrahydropyranyl-N-hydroxybenzamides 21a–m

To a mixture of corresponding acid 20a–m (0.5 mmol), O-(tetrahydro-2H-pyran-2-yl) hydroxylamine (1.5 equiv.) and DIPEA (3 equiv.) in dry DMF (3 mL) was added a solution of COMU (1.3 equiv.) in dry DMF (3 mL) dropwise over 10 min. The mixture was allowed to stir for 2 days at room temperature under nitrogen, then diluted with ethyl acetate (15 mL) and washed with sat. aq. NaHCO₃ (10 mL), followed by brine (10 mL), dried (Na₂SO₄). The solvent was removed in vacuo to give a crude product, which was purified by flash chromatography to afford the amide intermediate. This intermediate was then dissolved in THF/MeOH (1 : 1, 3 mL), 2 M HCl was added (0.5 equiv.) and the mixture was allowed to stir overnight at room temperature. The reaction mixture was then evaporated in vacuo, redissolved in 1 M NaOH (15 mL) and washed with diethyl ether (2 × 10 mL). The aqueous extract was then acidified to pH 4 with 2 M HCl and extracted with ethyl acetate (3 × 15 mL). The combined organic extracts were washed with brine (10 mL), dried (Na₂SO₄) and evaporated in vacuo, triturated with 1 : 1, petroleum ether/diethyl (2 × 5 mL), affording the desired pure hydroxamic acid 21a–m.

Amplex-red assay

PC-PLC_BC inhibitory activities for the compounds synthesised in this study were assessed using an Amplex-Red assay kit (Molecular Probe, Inc.), following a protocol as described by the manufacturer. Briefly, 0.01 U PC-PLC_BC was added to a well containing of 10 μM of synthesised inhibitor, 4 U mL⁻¹ alkaline phosphatase, 1 U mL⁻¹ horseradish peroxidase, 0.1 U mL⁻¹ choline oxidase, 0.5 mM phosphatidylcholine, and 0.4 mM Amplex-Red reagent in 1× reaction buffer (50 mM Tris-HCl, pH 7.4, 0.14 M NaCl, 10 mM dimethylglutarate, 2 mM CaCl₂). PC-PLC_BC inhibitory activity was measured through monitoring the amount of Resorufin fluorophore formed following inhibitor treatment, relative to vehicle (0.1% DMSO). Resorufin was excited at a wavelength of 560 nm and emits fluorescence at 590 nm, which was measured using an EnSpire multimode plate reader. All experiments were performed in triplicate, repeated three times, and reported as mean ± (s.e.m.).

Anti-proliferative assay

The triple-negative breast cancer cell line (MDA-MB-231) and colorectal cancer cell line (HCT116) were purchased from the ATCC, grown in alpha-MEM containing 5% foetal calf serum (FCS), penicillin/streptomycin (100 U mL⁻¹ and 100 μg mL⁻¹, respectively), and insulin/transferrin/selenium supplement (Roche). Cell proliferation assays for compounds synthesised in this study were conducted according to literature methods.^15,28 Briefly, 3000 cells per well were seeded into 96 well plates, which were tissue culture-treated for monolayer culture and incubated in the presence of 10 μM of each inhibitor for 3 days. Prior to harvest, 0.04 μCi of ³H-thymidine was added to each well and incubated for 5 hours, after which the cells were transferred using an automated TomTec harvester onto glass fibre filters. Filters were incubated with Betaplate Scint, and incorporation of radiolabelled thymidine was measured using a Trilux/Betaplate counter. All experiments were performed using triplicate wells, repeated three times, and reported as mean ± s.e.m.

Molecular modelling

The crystal structure of PC-PLC_BC was obtained from the Protein Data Bank (PDB) ID: 1AH7,²¹ resolution 1.50 Å and prepared using Genetic Optimisation for Ligand Docking software (GOLD) v2021.2.0.²³ Hydrogen atoms were added and crystallographic solvents removed, with the exception of waters W1–W9 which were extracted for docking: HOH249, 250, 258, 259, 287, 341, 346, 452 and 453. The binding site centroid was defined at the position of Zn²⁺ ion ZN248 (x = 42.482, y = 22.996, z = 8.556) with a radius of 10 Å. Basic amino acids lysine and arginine were defined as protonated, and acidic residues aspartic and glutamic acids as deprotonated to mimic the physiological environment. Ligands were prepared using Signals Chem3D v23.01.10 and energy minimised using the MM2 forcefield. Each ligand was subjected to 50 docking runs at 100% search efficiency using the scoring functions Goldscore,²⁹ Chemscore,³⁰ Astex Statistical Potential (ASP)³¹ and Piecewise Linear Potential (ChemPLP).³² Docking run solutions were visualised using Biovia Discovery Studio Visualizer v21.1.0.20298.

Methods used for determining microsomal stability

Purification of microsomal fractions

Four male and two female adult Sprague Dawley (SD) rats (246–556 g) were obtained from the Integrated Physiology Unit (University of Auckland, Auckland, New Zealand). All procedures involving the use of animals were conducted in accordance with the New Zealand Animal Welfare Act (1999) and approved by the University of Auckland Animal Ethics Committee. All procedures involving the use of rodents, and their care were approved and conducted in accordance with the ARRIVE2 guideline.³³ Rats were euthanised by CO₂ inhalation and cervical dislocation. Livers were dissected immediately after euthanasia, flash-frozen with liquid nitrogen and stored at −80 °C.

Approximately 100 mg of liver tissue from each rat was homogenised on ice using a 1 mL Dounce glass homogenizer in 1 mL phosphate buffered saline (PBS) containing 0.25 M sucrose, complete mini EDTA-free protease inhibitor cocktail (Cat #4693159001, 1 : 10 000; Roche Applied Science), and PhosSTOP phosphatase inhibitor cocktail (Cat #4906845001, 1 : 10 000; Roche Applied Science). The homogenate was centrifuged at 10 000 × g for 10 min at 4 °C. The supernatant was transferred to a new microfuge tube and centrifuged at 30 000 × g for 90 min at 4 °C. The pellet (the microsomal fraction) was resuspended in 200 μL of PBS, containing protease and phosphatase inhibitor, and the microsomal fractions from each rat were then pooled, aliquoted and stored at −80 °C in protein LoBind tubes (022431081, Eppendorf, Hamburg, DE). The protein concentration of the pooled rat microsomes was quantified using a bicinchoninic acid protein assay kit (ThermoFisher Scientific).

Microsomal stability testing assay

To obtain the T = 0 time point for each compound, compounds were added to PBS incubated at 37 °C to give a final concentration of 10 μM, mixed gently by inversion, and 75 μL taken and diluted/quenched with 75 μL of acetonitrile containing the internal standard (20 μM vanillic acid). For the T = 15, 30, 45, and 60 min time points, microsomal preparations were thawed and diluted to 2 mg mL⁻¹ in PBS on ice. Tubes containing 360 μL of microsomes were incubated at 37 °C were generated for each compound. NADPH was added to the microsomal preparations to give a final concentration of 1.6 mM, with the exception of the negative control tube. 40 μL of each compound was then added to their respective tube, giving a final concentration of 10 μM. The positive control was 10 μM of 7-ETC in microsomes with 1.6 mM NADPH, the negative control was 10 μM of 7-ETC in microsomes without NADPH. The tubes containing compound and microsomes were incubated at 37 °C for 60 minutes with 75 μL of each solution removed at corresponding time points for analysis. Microsomal activity in the 75 μL samples removed at each time point was quenched immediately after addition a 75 μL solution of acetonitrile, containing the internal standard (20 μM vanillic acid) and stored on ice until all samples for each compound and time point were collected. Samples were centrifuged at 15 000 rpm for 10 min at 4 °C and 120 μL of supernatant was transferred to HPLC sample tubes and stored at 4 °C until analysis.

HPLC analysis of microsomal stability

The samples were analysed using a Shimadzu LC-2050C 3D HPLC system with a Phenomenex Luna 5 μm C18(2) column (100 Å, 250 × 4.6 mm). The flowrate was 0.8 mL min⁻¹, and the injection volume was 10 μL. The system operated at 30 °C. The solvent profile was as follows: mobile phase A, water with 0.3% formic acid; mobile phase B, acetonitrile. Separation was afforded using the following binary gradient (Fig. S1†): 0–7 min, 15–31% B; 7–14 min, 31–47% B; 14–21 min, 47–63% B; 21–25 min, 63–95% B; 25–30 min, 95% B; 30–35 min, 95–15% B; 35–40 min, 15% B. Peaks were detected at 254 nm for purity analysis and at λ_max for microsomal degradation analysis (Table S1†). Concentrations at each time point were calculated using their relative peak areas compared to an internal standard (vanillic acid). Standard curves for each compound are given in the supporting information (Fig. S103–S106†).

Data availability

The data supporting this article have been included as part of the ESI.†

Author contributions

Shaun W. P. Rees: Writing – original draft, Writing – review & editing, Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Validation, Visualization. Tayla A. Rees: Writing – original draft, Writing – review & editing, Visualization, Validation, Methodology, Investigation, Formal analysis, Data curation. Emily K. Paulin: Writing – original draft, Writing – review & editing, Methodology, Investigation, Visualization. Olivia R. Arnerich: Writing – review & editing. Euphemia Leung: Data curation, Formal analysis, Investigation, Methodology, Resources. Christopher S. Walker: Supervision, Resources. David Barker: Writing – review & editing, Supervision, Resources, Conceptualization. Lisa I. Pilkington: Writing – review & editing, Supervision, Resources, Funding acquisition, Conceptualization.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could appear to influence the work reported in this article.