Cancer remains one of the leading causes of death worldwide, and despite significant advances in cancer research, discovering new and effective therapeutic agents continues to be a pressing challenge [1,2]. In recent years, cancer therapies have increasingly focused on tumor-specific targets, with growing emphasis on personalized treatment strategies. Among emerging modulators, adenosine—accumulating at high levels in the tumor microenvironment—can influence cancer progression via four G protein-coupled adenosine receptors: A1, A2A, A2B, and A3 [3]. The A1AR is highly expressed in glioblastoma cells and associated microglia, and its deletion has been shown to enhance tumor growth [4]. In contrast, A2AAR suppresses anti-tumor immunity by inhibiting T and natural killer cell activity, making it an attractive target for immunotherapy, making it a target for immunotherapy. A2BAR contributes to tumor progression by promoting angiogenesis, although it may also inhibit tumor growth signals through suppression of MAPK pathway activity. Increasing evidence shows that A3AR is overexpressed in various tumor types compared to adjacent normal tissues, with its expression levels correlating with disease severity [[5], [6], [7]]. The high receptor expression in tumors was attributed to overexpression of NF-κB, a known transcription factor for A3AR. The high receptor expression in tumors was attributed to overexpression of NF-κB, a known transcription factor for A3AR. Notably, A3AR agonists exert cell-type-specific effects: in normal cells, they stimulate growth factor production via NF-κB activation, while in tumor cells, they induce apoptosis and suppress tumor growth by disrupting NF-κB and Wnt signaling pathways. Given its complex role, the involvement of A3AR in cancer has been widely investigated through the use of selective agonist and antagonist ligands, frequently leading to contrasting results [8]. The efficacy of A3AR antagonists as anticancer agents has been supported based on the concept of hypoxia, which is characteristic of solid tumor, where adenosine levels are elevated, and HIF-1α is stabilized [9]. In this regard, it has been reported that A3AR stimulation induced HIF-1α accumulation in various cancer cell lines [10,11]. Additionally, A3AR stimulation has been linked to increased microvessel density in glioblastoma cells as well as an enhanced invasion of glioblastoma cells through A3AR and MMP-9 stimulation, a mechanism previously observed in macrophages. Conversely, A3AR agonists also have shown potential as cancer therapies. Previous studies suggested that tumor metastasis was less frequent in striated muscle, leading to the discovery that muscle cells secrete natural A3AR agonists [12,13]. Additional studies has demonstrated that the activation of A3AR suppresses telomerase activity and exerts cytostatic effects in tumor cells [[14], [15], [16], [17]]. Activation of A3AR reduced the migration of prostate cancer cells both in vitro and in vivo, inhibited cell proliferation, and induced G1 cell cycle arrest and apoptosis [[18], [19], [20]]. In the tumor microenvironment (TME), the enzyme ecto-5′-nucleotidase (CD73) catalyzes the conversion of adenosine-5′-monophosphate (AMP) into adenosine and phosphate, leading to elevated adenosine levels. This release of adenosine can activate any of the AR subtypes to modulate cAMP levels [21]. In this regard, overexpression of A3AR in various tumor cells indicates it could be an effective target for inhibiting cancer cell proliferation.

As shown in Fig. 1., both nucleoside and non-nucleoside derivatives are recognized for their ability to modulate A3AR. The A3AR selective agonists IB-MECA and Cl-IB-MECA, which are currently undergoing clinical trials for the treatment of inflammation and cancer, respectively, are prominent examples [14]. Additionally, the agonists CP-532903 and CP-608039 have been explored for their potential in clinical applications related to anti-ischemic cardioprotection [22]. Significant research has been conducted on the 4′-truncation of the A3AR nucleosides, particularly the 4′-thionucleosides, which led to the identification of LJ1251, which were shown to be antagonists using a functional assay of guanine nucleotide binding [23,24]. Furthermore, the triazoloquinazoline-based MRS 1220 [25,26] and quinazoline-based VUF5574 [27] are well-known for A3AR antagonists.

Over the past decade, natural products have emerged as promising candidates for the development of novel anticancer drugs [28]. Among these natural products, resveratrol and curcumin have been shown to possess notable biological activities [[29], [30], [31], [32], [33]]. Resveratrol (1), a polyphenol, is present in many plant species such as berries, grapes, and peanuts. Plants synthesize resveratrol as a defense mechanism in response to mechanical damage, fungal infection, and UV radiation [34]. As a nutraceutical, resveratrol plays a role in numerous beneficial actions in the body, including prevention against cardiovascular diseases, acting as an anti-aging [35], antioxidant [36], anti-inflammatory agents [37], and exhibiting chemopreventive [29], and anticancer effects [32,38]. Although resveratrol is known to directly alter cell functions such as proliferation and apoptosis in human cancer cells [30], the detailed mechanisms underlying its biological activity are poorly understood. Since it was reported that resveratrol might perturb membranes and alter protein function [39], many specific biological targets for resveratrol have been identified, including NQO2 (alone and in interaction with AKT1), GSTP1, estrogen receptor beta, carbonyl reductase 1, and integrin αVβ [40]. Nevertheless, it is still difficult to specify the exact targets responsible for the observed effects in cells and model organisms [40]. Curcumin is another nutraceutical phytochemical extensively studied over the past half-century. Curcumin (2) is a diarylheptanoid found in turmeric (Curcuma longa L.) that exerts different biological effects through diverse molecular targets, such as transcription factors, enzymes, protein kinases, growth factors, inflammatory cytokines, and receptors [41]. However, both curcumin and resveratrol are known as pan-assay interference compounds (PAINS) and invalid metabolic panaceas (IMPS). They can generate false signals in various assays due to drug-mimicking interactions [[42], [43], [44], [45]]. This may explain why resveratrol and curcuminoids have failed to show proof of concept in more than 300 clinical trials for several diseases [46]. In addition, curcumin derivatives have poor pharmacokinetic/pharmacodynamic (PK/PD) properties and toxic effects under certain testing conditions, making them unsuitable for drug development [47]. Therefore, there is a need for the development of novel compounds with improved biological activity, physicochemical properties, and safety profiles.

The activities of invalid metabolic pancreas (IMPS) are associated with Michael acceptors, such as the chemical structure of curcumin, which displays a broad range of bioactivities. However, in the field of drug discovery, their presumed indiscriminate reactivity often leads them to be overlooked. Although a well-established approach involves targeted covalent modification using Michael acceptors, it requires a sophisticated design for the target protein [48].

Previously, we designed and synthesized 9-cinnamyl-9H-purines to avoid PAINS-related IMPS activities and to have the modified nucleoside's properties [49]. (Fig. 2) In this study, 9-cinnamyl-9H-purines and their modified analogues were designed and synthesized to circumvent PAINS-related IMPS activities while retaining the properties of modified nucleosides. This approach integrates the cinnamyl group found in resveratrol and curcumin with purine, which are essential components of DNA and RNA. The nucleobase part of nucleosides and nucleotides plays a crucial role in various cellular processes, including DNA and RNA synthesis, cell signaling, enzyme regulation, and metabolism [50]. Therefore, these useful purine rings were introduced to minimize complications associated with the PAINS rule. By applying this nucleobase moiety to resveratrol and curcumin structures, it was anticipated that the issues of low efficacy in several disease models and toxicity under certain conditions could be addressed [46,47].

Among the compounds tested, 5e significantly inhibited nitric oxide production in LPS-induced macrophages (IC50 = 6.4 μM). It also reduced the levels of pro-inflammatory cytokines (IL-6, TNF-α, IL-1β) and decreased the protein level of iNOS and COX-2. Mechanistically, 5e disrupted the TLR4-MyD88 protein interaction, leading to suppression of the nuclear factor kappa B (NF-κB). In this regards, NF-κB is a protein complex that critically involved in regulating the expression of genes involved in inflammation, cell growth, and survival. Its overactivation is implicated in various of cancers, contributing to tumor growth, chemotherapy resistance, and metastasis [[51], [52], [53]]. NF-κB activity is modulated by several upstream signaling pathways, including the tumor necrosis factor-alpha (TNF-α) pathway and the IκB kinase (IKK) pathway. Certain anticancer agents, such as proteasome inhibitors and kinase inhibitors, can block these upstream signaling pathways and prevent NF-κB activation [54], thereby reducing tumor growth and promoting cancer cell death [55].

In addition to NF-κB, the signal transducer and activator of transcription (STAT) proteins play an essential role in cancer development and progression. STAT1, a member of the STAT family, governs essential cellular processes such as proliferation, apoptosis, and differentiation [56]. While traditionally viewed as a tumor suppressor orchestrating the transcription of cell cycle inhibitors and pro-apoptotic proteins, emerging evidence suggests that STAT1 may also exhibit tumor-promoting activities. It contributes to tumor growth through diverse mechanisms including suppressing tumor immune surveillance, enhancing invasiveness and metastasis, and enabling the tumor to develop resistance to radiation and chemotherapy treatments. Additionally, upregulation of STAT1 associated with elevated levels of CD74 delineates a subtype of triple-negative breast cancer (TNBC) characterized by enhanced invasive and metastatic capacity [57,58].

Therefore, we investigated the anticancer potential of 9-cinnamyl-9H-purines analogues. We conducted structure-activity relationship studies using 5e, which exhibited the most potent anticancer activity against various cancer cell lines. In our efforts to elucidate the mechanism of action, we employed the AI-based DeepZema® target identification system and verified the results by human adenosine A1, A2A, A2B and A3 receptor competition binding experiments followed by in silico studies.