Synthesis and Biological Activity of 2-Aminothiazoles as Novel Inhibitors of PGE₂ Production in Cells

Abstract

This report presents the synthesis and biological evaluation of a collection of 2-aminothiazoles as a novel class of compounds with the capability to reduce the production of PGE₂ in HCA-7 human adenocarcinoma cells. A total of 36 analogs were synthesized and assayed for PGE₂ reduction, and those with potent cellular activity were counter screened for inhibitory activity against COX-2 in a cell free assay. In general, analogs bearing a 4-phenoxyphenyl substituent in the R² position were highly active in cells while maintaining negligible COX-2 inhibition. Specifically, compound 5l (R¹=Me, R²=4-OPh-Ph, R³=CH(OH)Me) exhibited the most potent cellular PGE₂ reducing activity of the entire series (EC₅₀ = 90 nM) with an IC₅₀ value for COX-2 inhibition of > 5 μM in vitro. Furthermore, the anti-tumor activity of analog 1a was analyzed in xenograft mouse models exhibiting promising anti-cancer activity.

Inflammatory processes are implicated in 25% of all cancers and chronic inflammation has been shown to promote the growth of malignant tissues. Indeed, prostaglandin E₂ (PGE₂) has been identified as a key mediator of pain and inflammation and is overexpressed in various cancers. Definitive evidence suggests that PGE₂ is the dominant prostaglandin involved in the growth of tumors associated with colon, lung, breast, head, and neck cancers.^1-4 Interestingly recent studies support a role for PGE₂ in most if not all of the six hallmarks of cancer named by Hanahan and Weinberg; the mechanisms by which PGE₂ induces these cellular capabilities can vary amongst cancer types.^4-6

Traditionally, prostaglandin synthesis has been modulated primarily through the use of non-steroidal anti-inflammatory drugs (NSAIDs), including non-specific cyclooxygenase (COX) -1 and 2 inhibitors (Aspirin, Tylenol, and Ibuprofen) and specific COX-2 inhibitors (Bextra®, Celebrex®, and Vioxx®). Incidentally, patients taking NSAIDs for chronic inflammatory disorders such as rheumatoid arthritis are associated with a decreased incidence of colon cancer, and high doses of NSAIDs have been shown to cause colorectal tumor regression.^1-4 Furthermore, exogenous treatment with PGE₂ blocks NSAID-induced tumor regression.¹ Therefore, it is believed that the anti-cancer effects of COX inhibitors are ultimately due to a reduction of PGE₂ levels, although there is a downside of NSAID treatment in that the high doses and chronic usage required for anti-cancer activity lead to adverse target-based side effects including gastrointestinal intolerance, heart attacks, and blood clotting.^1-4 Thus, these safety implications have motivated research efforts focused on the reduction of PGE₂ levels through alternate mechanisms.^7-11

Herein we report the biological evaluation of a series of 2-aminothiazole analogs as a novel class of small molecules with PGE₂ reducing character in HCA-7 colon cancer cells. The general structure for analogs 1-5 is shown in Figure 1.

Figure 1.

General structure of aminothiazoles 1-5

2-Aminothiazole analogs depicted in Table 1 (1a-h, 2a-g, 3a-d, 4a-e) as well as structures 5c-i from Table 2 were assembled by employing a microwave assisted Hantzsch reaction (Scheme 1).^{12, 16} Compounds 5j and 5k-l were obtained by reduction of the ketone functionality of 5i and 5f, respectively (Scheme 1). Alternately, compounds 5a and 5b were synthesized in two-steps proceeding through an amidino-thiourea intermediate (6) (Scheme 2).¹³ Purities of all analogs were assessed by HPLC/MS analysis. All compounds showed purity level greater than 95%, as judged by UV absorbance at 254 nm and 214 nm as well as evaporative light scattering (ELS).

Table 1.

Biological activity of 2-aminothiazoles analogs 1-4. ^abcdef

Cmpd	R1	R2	R3	R4	PGE2 (%)c	COX-2 (%)d	COX-2 IC50μMe	PGE2 EC50 /μMf
1a	4-Cl-Ph	4-OPh-Ph	H	H	75.5 ± 7	32.1 ± 5	> 10	0.28 ± 0.11
1b	4-Cl-Ph	3-OCH2O-4-Ph	H	H	94.2 ± 3	68.8 ± 8	0.84 ± 0.20	-
1c	4-Cl-Ph	3-OMe-Ph	H	H	78.3 ± 3	55.1 ± 3	1.39 ± 0.34	-
1d	4-Cl-Ph	Ph	H	H	51.0 ± 8	-	-	-
1e	4-Cl-Ph	1-naphthyl	H	H	51.4 ± 8	-	-	-
1f	4-Cl-Ph	4-CN-Ph	H	H	23.1 ± 10	-	-	-
1g	4-Cl-Ph	4-OH-Ph	H	Me	85.8 ± 8	84.2 ± 13	1.08 ± 0.37	-
1h	4-Cl-Ph	4-OMe-Ph	Me	H	89.3 ± 2	12.0 ± 1	> 5	0.29 ± 0.10
2a	Ph	4-OPh-Ph	H	H	79.6 ± 10	31.0 ± 1	> 5	0.33 ± 0.15
2b	4-F-Ph	4-OPh-Ph	H	H	87.9 ± 1	43.4 ± 8	> 5	0.12 ± 0.04
2c	2-OMe-Ph	4-OPh-Ph	H	H	82.0 ± 3	38.9 ± 5	> 5	0.18 ± 0.02
2d	3-OMe-Ph	4-OPh-Ph	H	H	88.9 ± 1	40.3 ± 4	> 5	0.29 ± 0.13
2e	4-OMe-Ph	4-OPh-Ph	H	H	80.4 ± 2	40.8 ± 5	> 5	0.24 ± 0.05
2f	2-napthyl	4-OPh-Ph	H	H	53.5 ± 1	-	-	-
2g	4-Cy-Ph	4-OPh-Ph	H	H	41.6 ± 16	-	-	-
3a	4-Me-Ph	4-OMe-Ph	H	H	37.3 ± 6	-	-	-
3b	3,4-diMe-Ph	4-OMe-Ph	H	H	73.9 ± 2	69.3 ± 5	1.20 ± 0.19	-
3c	4-OMe-Ph	4-OMe-Ph	H	H	73.6 ± 3	66.8 ± 3	1.37 ± 0.08	-
3d	4-OMe-Ph	4-Me-Ph	H	H	91.5 ± 1	73.1 ± 6	0.81 ± 0.16	-
4a	Ph	3-OCH2O-4-Ph	H	H	98.2 ± 1	114.2 ± 1	0.41 ± 0.11	-
4b	2-Cl-Ph	3-OCH2O-4-Ph	H	H	85.9 ± 3	84.8 ± 3	0.91 ± 0.10	-
4c	2-OMe-Ph	3-OCH2O-4-Ph	H	H	89.0 ± 4	91.3 ± 4	0.93 ± 0.02	-
4d	3-OMe-Ph	3-OCH2O-4-Ph	H	H	94.4 ± 2	86.0 ± 8	0.92 ± 0.38	-
4e	4-OMe-Ph	3-OCH2O-4-Ph	H	H	96.2 ± 1	94.4 ± 2	0.91 ± 0.06	-

^aCelecoxib was used as a positive control (PGE₂, 5.70 ± 3.28% and EC₅₀ = 1.15 ± 0.06 nM; COX-2, 85.6 ± 4.4% inhibition and IC₅₀ = 0.03 ± 0.01)¹².

^b- not determined.

^c% of inhibition of PGE₂ levels in HCA-7 cells at 1μM concentration ± SD (n = 3).

^d% of inhibition of COX-2 levels in vitro at 5 μM concentration ± SD (n = 3).

^ein vitro IC₅₀ for COX-2 inhibition ± SD (n = 3).

^fEC₅₀ for PGE₂ level reduction in HCA-7 cells ± SD (n = 3).

Table 2.

Biological activity of 2-aminothiazoles analogs 5a-l. ^abcdefgh

Cmpd	R1	R2	R3	PGE2 (%)d	COX-2 (%)e	COX-2 IC50 /μMf	PGE2 EC50/μMg
5a	Ph	3-OCH2O-4-Ph	C(O)Ph	NR	-	-	-
5b	Ph	4-CN-Ph	C(O)Ph	NR	-	-	-
5c	Ph	4-OMe-Ph	C(O)Ph	NR	-	-	-
5d	Ph	4-OPh-Ph	C(O)Ph	42.0 ± 6	-	-	-
5e	Me	Ph	C(O)Me	29.9 ± 1	-	-	-
5f	Me	4-OPh-Ph	C(O)Me	87.9 ± 1	28.9 ± 15	> 5	0.39 ± 0.07
5g	Me	3-OCH2O-4-Ph	C(O)Me	25.3 ± 3	-	-	-
5h	Me	4-CN-Ph	C(O)Me	22.4 ± 4	-	-	-
5i	Me	4-OMe-Ph	C(O)Me	58.5 ± 4	-	-	-
5j	Me	4-OMe-Ph	Et	94.4 ± 1	49.3 ± 4	> 5	0.32 ± 0.14
5k	Me	4-OPh-Ph	Et	87.1 ± 3	41.3 ± 5	> 5	0.37 ± 0.07
5l	Me	4-OPh-Ph	CH(OH)Me	80.4 ± 3	-16.9 ± 10 h	> 5	0.09 ± 0.04

^aCelecoxib was used as a positive control (PGE₂, 5.70 ± 3.28% and EC₅₀ = 1.15 ± 0.06 nM; COX-2, 85.6 ± 4.4% inhibition and IC₅₀ = 0.03 ± 0.01)¹².

^bNR, no observed reduction.

^c- not determined.

^d% of inhibition of PGE₂ levels in HCA-7 cells at 1 μM concentration ± SD (n = 3).

^e% of inhibition of COX-2 levels in vitro at 5 μM concentration ± SD (n = 3).

^fin vitro IC₅₀ for COX-2 inhibition ± SD (n = 3).

^gEC₅₀ for PGE₂ level reduction in HCA-7 cells ± SD (n = 3).

^hto be considered as no inhibition of activity rather than induction of activity.

Scheme 1.

Synthesis of aminothiazoles 1a-h, 2a-g, 3a-d, 4a-e, 5c-l

Scheme 2.

Synthesis of aminothiazoles 5a-b

Initially, all compounds (Figure 2-3) were screened for their ability to reduce PGE₂ production in HCA-7 colon cancer cells at 1 μM concentration; activities are summarized in Tables 1-2 as percentage reduction of PGE₂ levels.¹⁷ As a means for tuning out COX-2 activity, compounds that exhibited reduction of PGE₂ levels higher than 70% were tested for COX-2 inhibition at 5 μM in an in vitro cell free assay,¹⁸ with Celecoxib incorporated as a positive control in both PGE₂ and COX-2 assays.¹⁴ IC₅₀ values for COX-2 inhibition were determined only for compounds that exhibited inhibitory activity against COX-2 greater than 50%. Only compounds that exhibited more than 70% reduction of PGE₂ levels, but did not show more than 50% COX-2 inhibition were tested further for EC₅₀ determination.

Figure 2.

Structure of 2-aminothiazoles analogs 1-4.

Figure 3.

Structure of 2-aminothiazoles analogs 5.

Aminothiazole analogs 1a-h consist of a para-chlorophenyl group at the C-4 position (R¹) of the thiazole ring and varying substituents at other positions. Presence of 4-phenoxyphenyl (1a), 3,4-methylenedioxyphenyl (1b), or 3-methoxyphenyl (1c) at the R² position leads to good activity (reduction in PGE₂ levels = 76-94%). Compounds 1d-e were characterized by lower activity (reduction in PGE₂ levels = 51%) when compared to compounds 1a-c. Substitution with a polar cyano group (1f) also resulted in significant loss of activity (reduction in PGE₂ levels = 23%) possibly due to a decreased ability to permeate the cellular membrane. Surprisingly, analog 1g bearing a methyl group on the amino functionality (R⁴), resulted in good cellular activity despite the polar 4-hydroxyphenyl substituent at R² (reduction in PGE₂ levels = 86%). Similarly, substitution of a methyl group at the C-5 position (R³) of the aminothiazole ring, as in compound 1h, is also well tolerated (reduction in PGE₂ levels 89%). Compounds from this group exhibited relatively strong inhibition of COX-2 in vitro (IC₅₀ = 0.84-1.39 μM), with exceptions represented by 1a and 1h (IC₅₀ > 5 μM). These two compounds were then tested for EC₅₀ determination and show similar EC₅₀ values for cellular PGE₂ reduction (0.28 μM and 0.29 μM, respectively).

Compounds 1a and 2a-g consist of a para-phenoxyphenyl group on the amino functionality (R²). These compounds show a moderate to high reduction of PGE₂ levels (42-89%). In this series, compounds characterized by a para-fluorophenyl (2b) or a methoxyphenyl (2c-e) substituent at C-4 of the 2-aminothiazole core (R¹) exhibited the highest level of PGE₂ reduction (80-89%). Movement of the methoxy group amongst ortho, meta, and para positions on the phenyl ring at R¹ did not significantly influence the observed biological activity (2c-e). However, the activity of compound 2a (reduction of PGE₂ levels 80%) suggested that the presence of a substituent on the phenyl ring of R¹ was not essential. Presence of bulky lipophilic substituents (naphthyl, 2f; 4-cyclohexylphenyl, 2g) at the R¹ position results in lower activity (reduction in PGE₂ levels = 54% and 42%, respectively) when compared to compound 2a. None of these compounds 2a-g showed significant COX-2 inhibitory activity (IC₅₀ values > 5 μM). Notably, analogs 2b and 2c exhibited the lowest EC₅₀ values for PGE₂ reduction of this group (0.12 and 0.18 μM, respectively). Other analogs showed EC₅₀ values similar to analog 1a (0.24-0.33 μM). In general, analogs bearing a para-phenoxyphenyl substituent at R² (1a and 2a-e) showed negligible inhibitory activity on COX-2 while still strongly reducing PGE₂ levels in cells.

Replacement of the previous para-phenoxyphenyl group at R² (2a-g) with a para-methoxphenyl group (3a-c) resulted in an analogous activity trend. Compounds 2e and 3c bearing a para-methoxyphenyl at R¹ both show similar reduction of PGE₂ levels (80% and 74%, respectively). Additionally, compounds 3a and 3b bearing 4-methylphenyl and 3,4-dimethylphenyl substituents, respectively, showed cellular activities comparable to analogs 2a, 2f, and 2g also containing hydrophobic moieties at C-4 (reduction in PGE₂ levels = 37-74%, 3a-b and 42-80%, 2a, f-g). Interestingly, when the 4-methoxyphenyl group on R² of analog 3c was exchanged with a 4-methylphenyl to give analog 3d, an increase in cellular PGE₂ reducing activity was observed (reduction in PGE₂ levels = 74% and 92%, respectively). Although transitioning from a para-phenoxyphenyl (2a-g) to a para-methoxyphenyl (3a-c) at R² does not significantly affect PGE₂ reducing activity, the COX-2 inhibition of 3b-c (IC₅₀ = 1.20 and 1.37 μM, respectively) is considerably stronger than for 2a-e (IC₅₀ values > 5 μM).

Introduction of a 3,4-methylenedioxyphenyl substituent on the amine nitrogen (R²) resulted in a dramatic reduction of cellular PGE₂ levels (4a-e, 98-86%). Compound 4a, characterized by an un-substituted phenyl ring at the C-4 position of the 2-aminothiazole (R¹) showed the highest activity (reduction in PGE₂ levels = 98%). Unfortunately, the high levels of cellular PGE₂ reducing activity observed with this group of compounds is always associated with relatively strong inhibition of COX-2 activity (4a-e; IC₅₀ = 0.41- 0.93 μM). Thus, the strong cellular activity of 4a-e as well as analogs 3b-c may arise, at least in part, from inhibitory effects on COX-2. However the overall cellular PGE₂ reducing activity of these analogs cannot be fully explained by COX-2 inhibition and rather is likely the result of inhibitory activity on more than one target. Furthermore, although tuning out COX-2 inhibition has proven to be clinically important in terms of safety, PGE₂ reducing therapeutics functioning primarily through a mechanism not involving COX, but maintaining weak COX-2 inhibition, may still be therapeutically useful if PGE₂ biosynthesis can be blocked while COX activity remains at least at a basal level.^1-4

Compounds 5a-i containing a keto moiety (acetyl or benzyl group) at the R³ position (Table 2) were moderately active to inactive (0-59%; except 5f, reduction of PGE₂ levels = 88%). Overall, analogs characterized by the presence of a methyl at C-4 (R¹) and an acetyl group at C-5 (R³) of the aminothiazole (5e-i) exhibited higher PGE₂ reducing character than analogs 5a-d. Of interest, when the acetyl group of 5i was reduced to the corresponding ethyl chain (R³, 5j) a marked reduction of PGE₂ levels was observed (94%). In contrast, reduction of the ketone functionality of 5f to the corresponding ethyl group (5k) did not lead to a significant change in activity (reduction in PGE₂ levels = 88% and 87%, respectively). Interestingly, compound 5l with -CH(OH)Me at R³ was the most potent compound identified in this report (EC₅₀ = 0.09 μM). Compound 5l was slightly more active than its parent analog 5f (EC₅₀ = 0.39 μM) and its fully reduced derivative 5k (EC₅₀ = 0.37 μM). Active compounds 5f and 5j-l were analyzed for COX-2 activity, showing little to no inhibition at 5 μM.

In conclusion, we have synthesized and tested a series of 2-aminothiazole congeners with the ability to reduce the production of PGE₂ in HCA-7 cells. A total of 36 aminothiazoles were evaluated, and active compounds with limited COX-2 inhibition were identified. Compound 1a was also evaluated for its effect on tumor growth in mouse xenograft models under three different cell lines, and was confirmed to have promising anti-cancer activity. Tumor vs. control values (T/C) were 61% and 40% for HCA-7 human colonic adenocarcinoma and SW837 rectum adenocarcinoma cell lines, respectively, under a dosing schedule of 200 mg/kg over 5 days (P<0.05). Additionally, a T/C value of 38% was observed in A549 human lung adenocarcinoma cells after treatment with 100 mg/kg over a period of 10 days (P<0.05).¹⁵ Work is on-going to identify alternate modes of action that may feasibly lead to reduced PGE₂ levels. Future synthetic efforts will be focused on further development of aminothiazoles with diverse substitution patterns that will continue to reveal important structure activity relationship features and aid in the identification of more potent compounds with anticancer activity in vivo.