Discovery of a new pyrido[2,3-d]pyridazine-2,8-dione derivative as a potential anti-inflammatory agent through COX-1/COX-2 dual inhibition

Abstract

In this paper, we present the design and synthesis of a novel series of pyrido[2,3-d]pyridazine-2,8-dione derivatives via the annulation of the 2-pyridone pattern. The synthesized derivatives were evaluated for in vivo anti-inflammatory activity using an ear edema model. Compound 7c, which showed a greater inhibition of ear edema (82%), was further tested for its in vitro COX-1/COX-2 inhibitory activity. Compound 7c showed similar inhibitory activities against COX-1 and COX-2 isoenzymes. The structural features that ensure the dual inhibition of COX-1 and COX-2 were elucidated using molecular docking studies. Overall, the ring closing of 2-pyridone pattern I transformed this highly selective COX-2 inhibitor into a dual COX inhibitor (7c), which could serve as a model for determining selectivity for COX-2.

The new pyrido[2,3-d]pyridazine-2,8-dione 7c showed potential anti-inflammatory activity as a COX-1/COX-2 dual inhibitor.

Introduction

Inflammation is the body's response to harmful stimuli and serves as a protective mechanism characterized by pain, heat, swelling, redness, and dysfunction. Upon the initiation of an inflammatory event, the immune system initiates a cascade of physical, chemical, and biological responses.¹ This activation prompts macrophages to generate and release inflammatory mediators, including substances like nitric oxide (NO), prostaglandin H2 (PGH2), tumor necrosis factor (TNF)-α, as well as pro-inflammatory enzymes like inducible NO synthase (iNOS) and cyclooxygenase (COX).² However, excessive levels of inflammatory mediators produced by macrophages can contribute to various skin conditions, such as psoriasis and atopic dermatitis.³ As a result, to modulate these mediators could represent a potential focus for therapeutic intervention in the treatment of inflammatory diseases.⁴

Nonsteroidal anti-inflammatory drugs (NSAIDs) are a family of compounds widely used to alleviate pain, fever, and chronic inflammation. The principal mechanism by which NSAIDs are used to combat inflammation is related to their ability to inhibit the isoforms of the cyclooxygenase enzyme (COX-1 and COX-2).⁵ COX-1 primarily plays a role in synthesizing prostaglandins and maintaining essential functions such as cytoprotective effects in the gastrointestinal tract, vascular stability, and normal renal function. In contrast, COX-2 is associated with the response to pro-inflammatory triggers and the release of prostaglandins, which play a role in the inflammatory process.⁶ Aspirin, indomethacin, ibuprofen, and diclofenac (depicted in Fig. 1a) are non-specific inhibitors that display a greater preference for COX-1, which can lead to significant adverse effects such as gastrointestinal bleeding, ulcers, and kidney issues. Conversely, selective COX-2 inhibitors such as celecoxib, parecoxib, and etoricoxib (Fig. 1b) are used to treat conditions such as osteoarthritis, rheumatoid arthritis, and pain relief. Regrettably, these inhibitors have been associated with undesired side effects, such as elevated systemic blood pressure and the risk of myocardial infarction.⁷ Consequently, there is an urgent need to identify novel COX inhibitors with enhanced safety profiles.

Fig. 1. Representative non-steroidal anti-inflammatory drugs (NSAIDs): a) COX non-selective inhibitors. b) COX-2 selective inhibitors.

Nitrogen-heterocyclic rings are privileged structures that exhibit a strong correlation between physicochemical properties and pharmacological effects. Thus, such structures are commonly found in biologically active compounds.⁷ The pyridone derivatives are of great importance in organic synthesis and drug design, as they exhibit a wide range of pharmaceutical properties, including anti-inflammatory activity.⁸ Furthermore, pyridazine-based compounds have been widely evaluated as anti-inflammatory agents, highlighting the importance of this nitrogen-heterocycle in anti-inflammatory drug discovery.⁹

Recently, we reported the synthesis and anti-inflammatory activity of 3,5,6-trisubstituted 2-pyridone derivatives I, where one of the most active compounds was identified as a selective COX-2 inhibitor.¹⁰ Herein, we envisioned the generation of a new bicyclic fused scaffold by cyclization of I with hydrazines and investigated their anti-inflammatory activity (Fig. 2). To the best of our knowledge, this is the first time that the pyrido[2,3-d]pyridazine-2,8-dione scaffold has been described.

Fig. 2. Ring closing strategy of I to achieve a novel pyrido[2,3-d]pyridazine-2,8-dione framework.

Results and discussion

Herein, we report the synthesis of a series of novel pyrido[2,3-d]pyridazine-2,8-dione derivatives and evaluate their anti-inflammatory properties in silico, in vitro, and in vivo.

Synthesis

Diversified pyrido[2,3-d]pyridazine-2,8-dione derivatives were prepared via the synthetic steps shown in Scheme 1. A series of polyfunctionalized 2-pyridone substrates 2a–g and 3a–g were prepared from the reaction of β-enamino diketones 1a–g with active methylene reagents (malononitrile and ethyl cyanoacetate).¹⁰ Subsequently, cyclocondensation of substrates 2 and 3 with hydrazine monohydrate, carried out in reflux EtOH/MeCN (1 : 1 v/v) for 6–16 hours,¹¹ resulted in the desired 3,5-disubstituted pyrido[2,3-d]pyridazine-2,8-diones 4a–g and 5a–g, respectively, in good to excellent yields (63–99%). The nature of the R substituent on the benzene ring influences both reaction time and yield. For the electron-donating group (R = OMe), a longer reaction time was required (16 h), and products 4g and 5g were obtained in the lowest yields (63% and 69%, respectively). Furthermore, we developed a one-pot synthesis of pyrido[2,3-d]pyridazine-2,8-diones 4 and 5 starting from β-enamino diketones 1a–g. The reaction between 1a–g and active methylene reagents (malononitrile and ethyl cyanoacetate) was performed under reflux EtOH, for 8–24 h, and in sequence, hydrazine monohydrate was added and the reaction was refluxed for the time required for full conversion to the desired products (6–16 h). Overall, 3,5-disubstituted pyrido[2,3-d]pyridazine-2,8-diones 4a–g and 5a–g were obtained in better yields (61–88%) using the one-pot method. To initiate a preliminary structure–activity relationship study, 3-carboxyethyl pyrido[2,3-d]pyridazine-2,8-diones 5a–g were hydrolyzed by reaction with NaOH (4 M), in MeOH at room temperature for 6 h, providing new derivatives 6a–g in good to excellent yields (78–90%).

Scheme 1. Synthesis of pyrido[2,3-d]pyridazine-2,8-diones derivatives 4a–g, 5a–g, and 6a–g. Reagents and conditions: (i) EtOH, reflux, 6–24 h; (ii) EtOH, 20 eq. HCl 37%, rt, 15–120 min, (55–86%); (iii) EtOH/MeCN (1 : 1), reflux, 6–16 h, (63–99%); (iv) malononitrile or ethyl cyanoacetate, EtOH, reflux, 6–24 h, hydrazine monohydrate, reflux, 6–16 h, (61–88%); (v) NaOH 4 M, MeOH, rt, 6 h, (78–90%); ^ayield from one-pot synthesis.

Additionally, we tested the reaction of 2-pyridones 2a–g with phenylhydrazine under the same conditions as those employed for hydrazine monohydrates; however, the desired product was not obtained. Then, the reaction was carried out using Lewis's acid BF₃·OEt₂ (2.0 equiv.), or even acetic acid (4.0 equiv.), under EtOH/MeCN (1 : 1 v/v) reflux, for 24 h. Again, the desired product was not formed. On the other hand, when we used a stronger protic acid (p-toluenesulfonic acid – 1.0 equiv.) the 3,5-disubstituted pyrido[2,3-d]pyridazine-2,8-diones 7a–g were obtained in moderate to good yields (50–71%) without the need for purification step (Scheme 2). This methodology was also applied to 2-pyridones 3a–g, however, the pyrido[2,3-d]pyridazine-2,8-dione derivatives were not obtained. All the synthesized compounds were characterized using spectroscopic and spectrometric data (see ESI† for more details).

Scheme 2. Synthesis of pyrido[2,3-d]pyridazine-2,8-dione derivatives 7a–g. Reagents and conditions: (i) TsOH, EtOH/MeCN (1 : 1), reflux, 24 h, (50–71%); yield of isolated product.

Anti-inflammatory activity

The topical anti-inflammatory activity of the newly synthesized pyrido[2,3-d]pyridazine-2,8-dione derivatives was evaluated using an experimental model of ear edema induced by croton oil (CO), which contains 12-o-tetradecanoylphorbol-13-acetate (TPA) and other phorbol esters with irritating properties. This method is straightforward, easily reproducible, and provides rapid and sensitive results, requiring only a minimal amount of each compound for assay purposes.¹² Topical administration of the compounds was at doses of 1.25 mg per ear and 0.625 mg per ear. The effectiveness of the compounds was expressed as percent edema inhibition (EI%) after 6 hours and compared to the reference drug indomethacin.

In our earlier work,¹⁰ 2-pyridone derivative I with R = Cl and NO₂ exhibited significant activity; therefore, we evaluated a related series of pyrido[2,3-d]pyridazine-2,8-diones 4a, 4c, 5a, 5c, 6a, 6c, 7a, and 7c. The results are presented in Table 1. For compounds 4a, 4c, 5a, 5c, 6a, and 6c obtained from hydrazine monohydrate (R² = H), the presence of the carboxylic acid group in R¹ produced better results than the nitrile and ester groups. Compound 6a showed an EI% of 56.2% at a dose of 1.25 mg per ear, whereas compound 6c showed an EI% of 75.2%. Furthermore, 6a and 6c were more potent than parent compound I (entries 5 and 6, Table 1). N-Phenyl-substituted compounds 7a and 7c (R² = Ph) demonstrated superior anti-inflammatory properties against the unsubstituted analogs 4a and 4c (R² = H). Compound 7c was the most effective, with an EI% of 81.8% at a dose of 1.25 mg per ear. Interestingly, the inhibition of ear edema by bicyclic structures 7a and 7c was almost identical to that by congener compound I.

Effects of pyrido[2,3-d]pyridazine-2,8-dione derivatives on croton oil-induced ear edema.

Entry	Compound	R	R1	R2	Dose (mg per ear)	EI%	EI% compound Ib
1	4a	NO2	CN	H	1.25	—a	93.0
					0.625	—a	—a
2	4c	Cl	CN	H	1.25	50.4	73.9
					0.625	49.0	—a
3	5a	NO2	CO2Et	H	1.25	nd	87.9
					0.625	64.2	66.3
4	5c	Cl	CO2Et	H	1.25	—a	—a
					0.625	—a	—a
5	6a	NO2	CO2H	H	1.25	56.2	—a
					0.625	63.2	—a
6	6c	Cl	CO2H	H	1.25	75.2	—a
					0.625	65.9	—a
7	7a	NO2	CN	Ph	1.25	78.5	93.0
					0.625	—a	—a
8	7c	Cl	CN	Ph	1.25	81.8	73.9
					0.625	62.8	—a
9	Indomethacin				1.00	98.2

^aDid not present significant results; nd: not determined because of its low solubility in the vehicle.

^bAccording to the available data in ref. 10.

Cytotoxicity assay

Utilizing in vitro assays with mammalian cell cultures can serve as a means to assess the toxicity and impact of target compounds as skin irritants.¹³ Additionally, this approach is cost-effective and helps reduce the excessive use of laboratory animals. Thus, to assess the effects of 7c on cell viability, experiments were performed on mammalian fibroblast L929 cells (ATCC® CCL-1, Manassas, USA) and macrophage J774A.1 cells (TIB-67; American Type Culture Collection, Manassas, VA, USA) by using the MTT method. Compound 7c showed CC₅₀ values for fibroblast (L929) and macrophage (J774A1) cells of 248.87 ± 22.56 μM and 364.98 ± 31.54 μM, respectively. Moreover, 7c showed a significant reduction in macrophage cytotoxicity compared with congener I (<100 μM).

In vitro COX inhibition assays

After determining that compound 7c exhibited the most promising anti-inflammatory effects in vivo, we evaluated its ability to inhibit COX-1 and COX-2 using an in vitro inhibition assay (Table 2). To assess its potency and selectivity, we measured the percentage of enzyme inhibition at two different concentrations of 7c: 1.95 μM and 31.25 μM. At lower concentrations, compound 7c showed a significant inhibitory activity on both COX enzymes, with no significant differences in COX-1 and COX-2 inhibition. At a concentration of 31.25 μM, 7c showed enhanced inhibition of both enzymes, with a stronger effect on COX-2. Therefore, 7c was found to be a nonselective COX inhibitor, while parent compound I exhibited significant inhibitory activity toward COX-2.¹⁰

Evaluation of pyrido[2,3-d]pyridazine-2,8-dione 7c as COX inhibitor.

Compound	Concentration (μM)	COX inhibitiona (%)
		COX-1	COX-2
7c	1.95	22.09 ± 1.14	20.19 ± 0.29
	31.25	30.59 ± 0.45	40.29 ± 0.50b
Indomethacinc		59.83 ± 0.24	32.89 ± 0.43
Rofecoxibd		13.66 ± 6.66	59.34 ± 3.39

^aResults are expressed as the mean ± standard deviation of two duplicates.

^b p < 0.05 vs.7c COX-2 minor concentration (Student's t-test).

^cIndomethacin at 10 μM (ref. 14).

^dRofecoxib at 100 μM for COX-1 and 1 μM for COX-2 (ref. 14).

Docking molecular study with COX-1 and COX-2

NSAIDs demonstrate distinct modes of binding to the COX active site; however, they establish contact with the enzyme in two or more binding pockets: proximal binding, central binding, and COX-2 side. The number of interactions depend on the size and orientation of the ligands. The amino acid residues that constitute the proximal binding pocket include Arg120 and Tyr355, where the majority of the inhibitors establish polar contacts, and Val349, Ser353, Tyr355, Val523 (COX-2), Ile523 (COX-1), and Ala527, which form a hydrophobic interaction network. The central binding pocket is define by Leu352, Leu384, Tyr385, Trp387, Phe381, Phe518, Met522, Gly526, and Ser530, where the inhibitors form hydrophobic interactions. The primary distinction between COX-1 and COX-2 is attributed to a substitution in the amino acid residue at position 523, where COX-1 has a larger Ile523 and COX-2 has a smaller Val523. This alteration in residue size leads to the emergence of a novel pocket (COX-2 side) lined by His90, Gln192, Leu352, Ser353, Tyr355, Arg513, Ala516, Phe518, and Val523. The side pocket is exploited as a binding site by COX-2 selective inhibitors.¹⁵

Docking of compound 7c and indomethacin (see Fig. S57, Tables S1 and S2, ESI†) into the crystal structures of COX-1 (PDB: 2OYE) and COX-2 (PDB: 5KIR) was performed to validate the biological data and elucidate the potential binding modes.

Molecular docking results show that compound 7c docked to the active site of COX-1 (Fig. 3C and D) exhibited similar binding behavior to indomethacin (see Fig. S57 and Table S1 – ESI†), with its pyrido-pyridazinedione moiety projecting up into the proximal binding pocket where the carbonyl group located on the pyridazinone moiety was able to establish a hydrogen bond with Tyr355. The 4-chlorophenyl moiety projects to the central binding pocket and interacts with Leu384, Tyr385, Trp387, Gly526, and Leu352, while the nitrile group forms polar interactions with Ser530 and Leu531 (Table 3).

Fig. 3. (A) 3D docking diagram of 7c (orange sticks models) and COX-2 residues (blue sticks models). (B) 2D mode of interaction of the compound 7c into COX-2 analyzed by Discovery Studio Client v20.1.0. (C) 3D docking diagram of 7c (orange sticks models) and COX-1 residues (yellow sticks models). (D) 2D mode of interaction of the compound 7c into COX-1 analyzed by Discovery Studio Client v20.1.0.

Docking results of 7c into COX-1 (PDB: 2OYE)^a.

Pocket	Amino acid	Interaction	Distance (Å)
Proximal binding	Val349	π-Alkyl with pyrido ring
	Ser353	CH-bond with C(O) of pyridazinone ring
	Tyr355	H-bond with C(O) of pyridazinone ring	2.69
	Ile523	π-Sigma with pyridazinone ring
	Ala527	π-Alkyl with pyrido and pyridazinone rings
Central binding	Leu352	π-Alkyl with 4-chlorophenyl and pyridazinone rings
	Leu384	Alkyl with Cl atom
	Tyr385	π with Cl atom
	Trp387	π with Cl atom
	Phe518	π–π stacked with the phenyl ring
	Gly526	Amide-π stacked with 4-Cl-phenyl ring
	Ser530	CH-bond with N of CN
	Leu531	H-bond with N of CN	2.91

^aBinding score −71.19 kcal mol⁻¹.

The orientation and interactions of compound 7c with the COX-2 binding pocket were similar to those of COX-1. In the proximal binding pocket, the pyrido-pyridazinedione core interacts with Val349, Ser353, Tyr 355, Val523, and Ala527, and the carbonyl group of the pyridazinone moiety forms a hydrogen bond with Tyr355. The para-chlorophenyl ring projects to the central pocket where it interacts with Leu352, Leu384, Tyr385, Trp387, and Phe381, whereas the nitrogen atom of the nitrile group interacts with Ser530. The phenyl group ring attached to the pyridazinone ring is in close proximity to the side pocket but forms only weak π-alkyl interactions with Val523 and Ala516 (Fig. 3A and B, Table 4).

Docking results of 7c into COX-2 (PDB: 5KIR)^a.

Pocket	Amino acid	Interaction	Distance (Å)
Proximal binding	Val349	π-Alkyl with pyrido and pyridazinone rings
	Ser353	CH-bond with C(O) of pyridazinone ring
	Tyr355	H-bond with C(O) of pyridazinone ring	2.58
	Val523	π-Alkyl with pyridazinone ring
	Ala527	π-Alkyl with pyrido and pyridazinone rings
Central binding	Leu352	π-Alkyl with 4-chlorophenyl and pyridazinone rings
	Leu384	Alkyl with Cl atom
	Tyr385	π with Cl atom
	Trp387	π with Cl atom
	Phe381	π with Cl atom
	Ser530	CH-bond with N of CN
COX-2 side	Ala516	π-Alkyl with phenyl ring

^aBinding score −72.30 kcal mol⁻¹.

To gain insights into the structural causes of the different binding mechanisms between nonselective inhibitor 7c and selective inhibitor of COX-2 I, molecular superposition was performed. As shown in Fig. 4, compounds 7c and 2-pyridone I are perfectly aligned. However, the 3-cyano-2-pyridone moiety of compound I fills the selectivity pocket of COX-2 and forms hydrogen bonds with the key amino acids Arg513 and His90. In contrast, the unsubstituted phenyl ring in 7c was oriented toward the side pocket and probably did not form electrostatic interactions because of its hydrophobic nature. The loss of this key interaction may explain the nonselective in vitro activity of compound 7c. Interestingly, the geometry of compound 7c allows the further introduction of substituents with hydrogen-bonding potential to improve the affinity for COX-2 and modulate the selectivity profile of the inhibitor.

Fig. 4. Pose overlap of pyridone pattern I (carbon shown in yellow) and pyrido-pyridazinedione 7c (carbon shown in orange) docked in COX-2.

Conclusion

A novel series of molecules possessing a fused pyrido[2,3-d]pyridazine-2,8-dione system was designed, synthesized, and evaluated for potential anti-inflammatory activity. Interestingly, N-phenyl-substituted pyrido[2,3-d]pyridazine-2,8-diones 7a and 7c exhibited higher anti-inflammatory activity than their unsubstituted analogs 4a and 4c and comparable potency to that of their pattern 2-pyridones I. Moreover, the most potent compound 7c was further tested for its in vitro cytotoxicity and COX-1/COX-2 inhibitory activity. The results revealed that 7c had no significant toxic effects and exhibited dual inhibition of COX-1 and COX-2. Importantly, 7c showed a significant reduction in macrophage cytotoxicity compared with 2-pyridone pattern I, a selective COX-2 inhibitor. Furthermore, the docking results revealed that 7c displayed a favorable fit with COX-1/2, with similar orientation and binding interactions. The dual inhibition of COX-1 and COX-2 could provide a balance between reducing inflammation and preserving COX-1-mediated protective functions. Overall, 7c could serve as a promising candidate for further development of new effective anti-inflammatory agents through structural optimization.

Conflicts of interest

The authors declare no conflict of interest.