Discovery, Optimization, and Cellular Activities of 2-(Aroylamino)cinnamamide Derivatives Against Colon Cancer

Curcumin and trans-cinnamaldehyde are acrolein-based Michael acceptor compounds that are commonly found in domestic condiments, and known to cause cancer cell death via redox mechanisms. Based on the structural features of these compounds we designed and synthesized several 2-cinnamamido-N-substituted-cinnamamide (bis-cinnamamide) compounds. One of the derivatives, (Z)-2-[(E)-cinnamamido]-3-phenyl-N-propylacrylamide 1512 showed a moderate antiproliferative potency (HCT-116 cell line inhibition of 32.0 µM), good selectivity profile (no inhibition of normal cell lines), and proven cellular activities leading to apoptosis. SAR studies led to more than 10-fold increase in activity. Our most promising compound, [(Z)-3-(1H-indol-3-yl)-N-propyl-2-[(E)-3-(thien-2-yl)propenamido)propenamide] 4112 killed colon cancer cells at IC50 = 0.89 µM (Caco-2), 2.85 µM (HCT-116) and 1.65 µM (HT-29), while exhibiting much weaker potency on C-166 and BHK normal cell lines (IC50 = 71 µM and 77.6 µM, respectively). Cellular studies towards identifying the compounds mechanism of cytotoxic activities revealed that apoptotic induction occurs in part as a result of oxidative stress. Importantly, the compounds showed inhibition of cancer stem cells that are critical for maintaining the potential for self-renewal and stemness. The results presented here show discovery of covalently-acting Michael addition compounds that potently kill cancer cells by a defined mechanism, with minimal effect on normal noncancerous cell.


INTRODUCTION
Propenamide is a Michael acceptor (MA) moiety that is frequently employed in designing drugs and drug candidates [1][2][3]. For instance, afatinib (Figure 1) a targetspecific propenamide that selectively inhibits mutated HER2 kinase [4] has been recently approved by the FDA for the treatment of non-small cell lung cancer [5]. Chidamide is a histone acetylase (HDAC) inhibitor approved in China for the treatment of advanced peripheral T-cell lymphoma. Chidamide incorporates 3-(3-pyridyl)acryloyl group which can be considered bioisosteric to cinnamyl moiety [6]. Other examples, among many, include ibrutinib (BTK inhibitor) and neratinib (HER-2 inhibitor) developed for B-cell cancers and solid tumors, respectively [7,8]. The efficacies of these drugs are, at least in part, depend on the MA functionalities. There are anticancer Mechanistically, the MAs cause cancer cell apoptosis by increasing the oxidative stress inside these cells. The cellular prooxidant induction by MAs is attributed to covalently ligation of SH groups of certain targets involved in regeneration of reduced glutathione (GSH). This result in accumulation of reactive oxygen species (ROS) which cause cancer cell cycle exits and apoptosis [9]. Acrolein (propenal, Figure 1) is the simplest enone MA characterized by high reactivity and non-selectivity to cellular nucleophiles and, therefore, is considered as a toxic environmental pollutant with well-known carcinogenic activities [10]. Interestingly, acrolein has also been reported to inhibit cancer cell growth. This is mediated by inhibition of the β-subunit of NFκB, an anti-apoptotic protein, via covalent binding with the nucleophilic Cys-61 and Arg-307 [11]. Replacement of the aldehyde group of acrolein with amide group (acrylamide moiety) is known to attenuate the reactivity of the MA moiety leading to less toxic derivatives [12]. In another direction, simple modification of acrolein by adding a phenyl group to the β-carbon provides a moiety that is found in natural products such as trans-Cinnamaldehyde (tCA) and curcumin (Figure 1). These domestic condiments are highly safe natural condiments with confirmed antiproliferative activities against several cancer cell lines [9,13,14]. Driven by the lead pharmacophores presence in natural food, attractive anticancer activities, and well-known safety profiles [9,[15][16][17][18][19][20], we envisaged tCA and curcumin as building blocks to develop MA anticancer compounds with superior pharmacologic properties ( Figure 2). We investigated structure activity relationship of these novel compounds on the antiproliferative potency in colorectal cancer (CRC) cells (CaCo-2, HCT-116 and HT-29). Our continued focus on CRC [21,22] was influenced by the prevalence of this malignant disease [23,24], its high morbidity and mortality rate [25], as well as the formidable problems of resistance to current available chemotherapeutic agents [26].
The rationale for designing of these compounds depended on replacement of the metabolically labile aldehyde group of tCA by a cinnamamide that extends to attach another cinnamamide to form the bis-cinnamamide scaffold (Figure-2). We investigated the structure-activity relationship (SAR) in three phases ( Figure 2): the first generation was testing the designed bis-cinnamamide scaffold as a close analogue to tCA, thus R 1 and R 2 were fixed as unsubstituted phenyl attached to two propenamide moieties and the variation was limited to substituents on R 3 that ranged from aliphatic, aromatic, heterocyclic groups. In addition, a compound in this series was an ester (X = O). In the second generation, X-R 3 was fixed as NH-propyl and the distance between the two aromatic groups (R 1 to R 2 ) was varied. In a third generation, the R 1 and R 2 were varied simultaneously by substituted phenyl and heteraryl groups. This stepwise SAR studies helped to identify compounds with high activity against three colon cancer cell lines. Figure 1. The SAR studies are illustrated. The entries are detailed in Table 1 below. Note that numbering of the final compounds follows the original project coding system and therefore, they are not sequential.

Chemical Synthesis of 2-cinnamamido-N-substituted-cinnamamide derivatives
The synthetic pathway for Final Compounds (Codes 1501-5612; Entries 1-56; Table 1 and Figure 2), described in scheme 1 utilized Erlenmeyer chemistry for azlactone synthesis [21]. In this scheme, non-commercially available cinnamic acid analogues 2 were prepared by condensation of the corresponding aldehyde with malonic acid [27]. After conversion to the hippuric acid analogues (3), the cyclo-condensation was affected by their reaction with the corresponding aldehydes to afford the 2-arylvinyl-4-benzylidene-5oxazolinone derivatives (azlactones, 4) under Erlenmeyer conditions [21,28]. The key azlactone intermediate 3 was subsequently reacted with a variety of aliphatic and aromatic amines to furnish the final compounds. Generally, the aliphatic and benzylic amines were reacted with the azlactone smoothly at room temperature in ethanol. The less reactive aromatic amines needed microwave heating in the presence of N,N-dimethylformamide (DMF) as a solvent. The ester 15EE was prepared by solvolysis in boiling ethanol in presence of catalytic amount of 4-dimethylaminopyridine (DMAP).

Antiproliferative activities of bis-cinnamamide derivatives
Compounds in this study ( Figure 2, Table 1) were tested for their cytotoxicity activities against colorectal cancer cell line HCT-116, using sulforhodamine-B (SRB) assay [29]. The first twenty compounds (Entries 1-20; Codes 1501 to 15EE), having fixed phenyl groups at R 1 and R 2 positions (Figure 2), showed moderate to weak antiproliferative activities ( Table 1). The compound with the highest activity in this subset was the N-(1-adamantyl) analogue (1505, Entry 4) with IC50 = 14.7 µM, but with poor solubility (ClogP 6.08), while compound 1512 (Entry 2) with moderate activities (IC50 = 32.0 µM ) showed better solubility profile (CLogP 4.48). Interestingly, 1512 exhibited significantly low cancer cell resistance (R-value = 8.2%) [21] and high selectivity in killing cancer cells (HCT-116, Caco-2 and HT-29) versus highly proliferative normal cells (C-166 mouse skin fibroblasts) (Figure 3).  Based on 1512 favorable biological properties (see Supplementary Material, Section S1 for more details on cellular mechanisms of this lead), it was selected as a new lead for the second subset of compounds (Entries 23 to 56; Codes 1812 to 5612) with fixed R 3 as propyl (based on 1512 structure) but diversified at R 1 and R 2 (Figure 2). The new compounds were tested against three CRC cell lines (HCT-116, Caco-2 and HT-29). Some of these compounds showed high potencies comparable to the reference drug doxorubicin (Table   1)

Induction of apoptosis and apoptotic changes in HCT-116 cancer cell lines upon treatment with 4112.
The apoptotic effect of 4112 on HCT-116 cells [30] was investigated by the DAPI staining test after incubation of HCT-116 cells with the compound at 3.00 µM . The result clearly shows that the ratio of cells with fragmented DNA and condensed nuclei are higher than that of untreated cells (control), which significantly increased after 48 h compared to 24 h     Figure 7).

Elevation of oxidative stress indicators within HCT-116 cells upon treatment with 4112.
ROS release, as an indicator for increased oxidative stress, was monitored over time (

Effect of Pre-treatment with Antioxidants on Cellular Viability
To investigate the potential effects of compound 4112 on ROS levels in HCT-116 colon cancer cells, we employed a viability assay in which cells were pre-incubated with the antioxidant N-acetyl cysteine (NAC) or the vitamin E derivative Trolox (6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid) prior to treatment with 4112 ( Figure 1). It worthy to mention that the antioxidant concentrations used in this test had no effect on cell viability.
As shown in Figure 9, pre-treatment with NAC for 1 h had a dramatic protective effect on

Inhibition of cancer stem cell proliferation in HT-29 cancer cell lines upon treatment with 3712 and 4112
We also evaluated 4112, and another potent compound 3712 for their ability to inhibit cancer stem cell proliferation using a primary (1°) colonosphere formation assay. In this assay, HT-29 colon cancer cells were grown in low adhesion plates to form spheres as previously described. [31] Wells treated with our compounds were compared to vehicle

Lethal dose toxicity test for 4112
Since compounds with electrophilic properties, e.g. MA are usually involved in variety of toxic effects on mammalians, we decided to investigate the general acute toxicity (modified, adopted March 23, 2006). As observed in our cellular assay, the compound safety was confirmed to be in the highest category, as the LD50 for all compounds were between 2000-5000 mg/kg (Category 5) [32].

DISCUSSION
Toward the goal of developing novel chemotherapeutic agents, we designed and synthesized several 2-cinnamamido-N-substituted-cinnamamide (bis-cinnamamide) compounds. Some of these compounds showed high potency against three CRC cancer cell lines, and in a few cases, were comparable to doxorubicin. The straight forward Erlenmeyer chemistry enabled SAR studies in three stages. First set of compounds (Entries 1-20, Table 1)  properties. The primary amide 1508 and the ester 15EE also were among the weakly active compounds.
The promising results of 1512 suggested that this compound is a viable lead for further optimization. In a limited set of only 2 compounds, the scaffold optimal size was investigated that have the distance between the two aromatic rings were changed by contracting (1612) and expanding this distance (1712) respectively. Both compounds showed almost no activity at concentrations as high as 100 µM against HCT-116 cells, suggesting that keeping the distance between the two aromatic rings in compound 1512 at 6 atoms is important for activity.
The third-generation compounds (Fixed R 3 as n-propyl and varied at R 1 and R 2 ) also showed a consistent SAR when tested against the CRC cell lines HCT-116, Caco-2 and HT-29. For instance, the variation in R 2 position between electron-deficient and electron-withdrawing aromatic moieties demonstrated a strong correlation between electron density over the R 2 ring and activity. Rings with higher electron density (4dimethylaminophenyl, 4-hydroxy-3-methoxyphenyl and 3-indolyl) as exemplified by compounds 2312, 2712, 3312, 3712-4112 exerted potent antiproliferative activities against cancer cells in the low micromolar range (0.89 to 6.9µM). Some of these compounds exhibited low activity against the non-cancerous cell lines BHK (Table 2). On the other hand, electron-deficient R 2 aryl groups showed almost no antiproliferative activity against the cancer cell lines, exemplified by compounds 4212 to 4612 (Entries 46-51, Table 1) where R 2 = 3-pyridyl; and 4712 to 5112 (Entries 52-56, Table 1) where R 2 = 4-nitrophenyl.
Notably, unlike R 2 , variations at the R 1 position affected anticancer activities less significantly. In this regard, all compounds bearing 3-indolyl group at R 2 were highly active regardless of the structure of the groups at R 1 . Meanwhile, all compounds with electron withdrawing groups at R 2 (such as 3-pyridyl and 4-nitrophenyl) were much less active, and this was also irrelevant to the structure of R 1 .
The both of which are critical for maintaining the potential for self-renewal and stemness. [33] Consistently, CSCs show low intracellular ROS levels, suggesting that maintaining a reduced intracellular environment is associated with an undifferentiated state [34]. We, therefore, tested the effect of 3712 and 4112 on CSCs in a cancer colonosphere formation assay, which is designed to screen for molecules that selectively target colorectal CSCs was also confirmed by increased levels of cleaved caspase-3 and its release through the nuclear membrane (Figure 7). In agreement with the above mentioned Annexin-V staining assay result, 4112 induces apoptosis via activation of caspases 3.
In conclusion, this work presents a model of molecular discovery of promising MA compounds to be further optimized as anticancer agents especially for colorectal cancer.
These compounds showed selective toxicity against cancer cell lines and CSC over noncancerous cells, low animal toxicity, and clear cell death mechanism. The two most potent compounds, 3712 and 4112 will be subjected to more SAR investigations, as well as antitumor activities in future studies.

Chemical Synthesis
Solvents and reagents were purchased from Sigma-Aldrich (USA), VWR (USA) or Purity of compounds were first assessed qualitatively using Thin Layer Chromatography (TLC), 1 H NMR and quantitatively using LC/MS (UV detection). The compounds were screened only if purity was confirmed to be above 95%. The compounds subjected to all biological screenings were used as a single Z isomer as detected by TLC, LC/MS and NMR.
The product 1558 was an off-white solid, (0

Cytotoxicity assay
The sulforhodamine B (SRB) assays were performed according to Skehan et al. 19

Nuclear fragmentation by DAPI staining
Cells were cultured on sterile 22 mm 2 cover slips (Harvard Apparatus, MA, USA) in sterile six well plates at a density of 2×10 5 cells/well. 24 h after seeding, cells were exposed to

Cell morphology
Cells were cultured on sterile 22 mm 2 cover slips (Harvard Apparatus, MA, USA) in sterile six well plates at a density of 2×10 5 cells/well. 24 h after seeding, cells were exposed to IC50 of the tested compounds in fresh medium for 24 h. At the end of the exposure, cells attached to cover slips were washed with PBS and visulized under leica light microscope (Leica, Buffalo Grove, IL, USA).

Primary (1°) Colonosphere Formation Assay
For primary sphere formation, cells were plated in nontreated, low adhesion, 96 wells plate at the concentration of 100 cells/100 μL/well in stem cell media (SCM) that consisted of DMEM:F12:AA (Gibco), supplemented with 1× B27 (Gibco), 20 ng/mL epidermal growth factor, and 10 ng/mL fibroblast growth factor (Sigma). After 4 h of incubation, vehicle (control) or 3712 and 4112 at the desired concentrations were added to each well (at least in triplicates for each sample). On day five, numbers of spheres ranging from 50 to 150 mm in diameter were counted using phase contrast microscope and percent inhibition was calculated compared to control.

Determination of ROS accumulation
To determine the effect of the newly synthesized compounds on the cellular redox status,

Cell viability assay including pre-treatment with N-acetyl cysteine or Trolox
N-Acetylcysteine (Sigma) was dissolved in sterile water (100 mM). Trolox (Sigma) was dissolved in ethanol (1 mM). Fresh solutions were prepared for each experiment. HCT-116 cells were seeded in 96-well plates at a density of 2.5  10 4 cells/mL. After 23 h, cells were pre-treated with NAC or Trolox (2 μL) for 1 h. The remainder of the assay was carried out as described for the previous cell viability section (7.2.2).

Apoptotic cell determination
Apoptosis was determined by staining cells with Annexin V-fluorescein isothiocyanate

A. Animals and Compounds
The experiment was conducted on 12 healthy Swiss albino mice (males and females) The compounds 1512, 3712 and 4112 of purity 95% or more (LC/MS) were prepared for this study as 10% suspension in water containing 0.5% tween 80.

B. Acute oral toxicity test
The animals were distributed randomly into four groups (3 mice at each group). One group did not receive any drug (control group). The second group received an oral dose