Cytotoxic and Apoptotic Activity of Aglaforbesin Derivative Isolated from Aglaia loheri Merr. on HCT116 Human Colorectal Cancer Cells

Background: The genus Aglaia (Meliaceae) is an established source of many anticancer compounds. The study evaluated the leaf extracts of Aglaia loheri, a tree native to the Philippines, as potential source of anticancer compounds. Methods: Using bioassay-guided fractionation, A. loheri leaf extract was subjected to various chromatographic techniques and step-wise application of MTT assay on human colorectal carcinoma cells, HCT116, to determine the cytotoxic fractions. The most cytotoxic HPLC isolate was structurally identified using 1D and 2D NMR and its apoptotic effect was assessed by JC-1 staining, caspase 3/7 assay and TUNEL assay. Results: After stepwise chromatography fractionation, an HPLC isolate, structurally identified as aglaforbesin derivative (AFD), demonstrated potent cytotoxicity against HCT116. AFD exhibited strong toxicity (IC50 = 1.13 ±0.07 µg/mL) and high selectivity on HCT116 than normal human kidney cells (HK-2). AFD-induced toxicity to HCT116 is possibly through the stimulation of the apoptotic signaling pathway via caspase 3/7 activation and DNA fragmentation independent of mitochondrial membrane depolarization. Conclusion: AFD exhibited selective cytotoxicity and apoptotic activity to HCT116 and could be further developed as anticancer drug lead.


Cytotoxic and Apoptotic Activity of Aglaforbesin Derivative Isolated from Aglaia loheri Merr. on HCT116 Human Colorectal Cancer Cells
Aetas, an indigenous tribe in the Philippines, who use the plant for medicinal and nutritional purposes (Dapat et al., 2013). Many Aglaia species have been used in traditional medicine to treat fever, cough, asthma, inflammations, wounds and tumors (Janaki et al., 1999;Proksch et al., 2005;Ebada et al., 2011).
Isolation of compounds from different Aglaia species show highly cytotoxic to non-cytotoxic activity against an array of human cancer cells. Several novel compounds cytotoxic against different cancer cell lines have been isolated from Aglaia species e.g., silvestrol from Aglaia foveolata (King et al., 1982), rocagloic acid from Aglaia elliptifolia (Wang et al., 2001) and aglaroxin A 1-O-acetate from Aglaia edulis (Kim et al., 2006). Previous studies on A. loheri leaves showed high to moderate cytotoxicity of crude extract against HCT116 and A549 cells (Canoy et al., 2011). Isolated compounds -pinasterol, trilinolein, and phytyl fatty acid were found to be cytotoxic against HCT116 (Ragasa et al., 2012) and a phenolic ester (Maldi 531.2[M + H]+) exhibiting cytotoxicity against human leukemic cells CCRF-CEM and its multidrug-resistant subline CEM/ADR5000 (Dapat et al., 2013).
This work is focused on screening for the most active extracts and fractions from A. loheri leaves that possess strong cytotoxic activity against human colorectal cancer cell line, HCT116. The study provides initial information on the mode of cell death induced by the most cytotoxic HPLC isolate obtained as well as its chemical identity. Altogether, the results of the study strengthen our understanding on the cytotoxic properties of A. loheri extracts and the potential of its cytotoxic isolates to be used as drug lead for future development as therapeutic agent.

Plant collection and crude extraction
Leaves of A. loheri were collected from Mt. Lamao, Bataan, Philippines. Identification was verified at the Jose Vera Santos Herbarium of the Institute of Biology, University of the Philippines, Diliman, Quezon City, where voucher specimen (accession number 21418) was deposited. Leaves were washed, air-dried, ground to powder and were macerated in distilled methanol at room temperature for at least 48 hours and filtrate was concentrated using a rotary evaporator at 40°C.

Bioassay-guided fractionation
A. loheri leaf methanolic extract was sequentially advanced through liquid partitioning, gravity column chromatography (GCC), flash chromatography (FC), size exclusion chromatography (SEC) and reverse-phase high performance chromatography (RP HPLC). Fractions with significant yield were constituted at 10 mg/mL DMSO and screened for toxicity against selected cell lines by 3-(4,5-dimethyl2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) assay. The most cytotoxic and high yielding fractions were tested against each cancer cell line to determine the concentrations inhibiting viability by 50% (IC 50 ).
The crude extract was partitioned successively with n-hexane, ethyl acetate, and water. The partitions were concentrated using rotary evaporator at 40°C while the aqueous partition was lyophilized. The most active ethyl acetate extract was subjected to GCC eluted with gradient of n-hexane, ethyl acetate and methanol. The fractions were monitored by thin layer chromatography (TLC) and similar eluents were combined. The high yielding, cytotoxic fraction 26 was subjected to FC in Biotage Isolera™ (Biotage, Sweden) using gradient of n-hexane, ethyl acetate and methanol. Similar eluents were pooled, affording 14 fractions. Fraction FC10 demonstrated strong cytotoxicity and was subjected to SEC using Sephadex LH-20, affording 10 fractions. The cytotoxic and high yielding SEC fraction 7 was further purified by semipreparative HPLC (mobile phase: 70% ACN and 30% H 2 O) using Luna ® 5 µm C18 100 Å, 250 mm length, 10 mm internal diameter (Phenomenex Inc., CA, USA). HPLC isolation afforded the cytotoxic HPLC7 fraction (6 mg; t R =13.5 min).

NMR analysis
Nuclear magnetic resonance (NMR) spectra of HPLC7 was recorded on a JEOL ECZR spectrometer at 600 MHz for 1 H NMR and 150 MHz for 13 C NMR spectra. Chemical shifts values are given in ppm relative to residual DMSO ( 1 H NMR: δ 2.50; 13 C NMR: δ 39.51) solvent. Coupling constants (J) were reported in Hz with the following splitting abbreviations: s = singlet, d = doublet, and m = multiplet. HPLC7

Assessment of cytotoxicity using MTT assay
Cytotoxicity of A. loheri against HCT116 was determined using MTT assay, a colorimetric test to measure the reduction of MTT by mitochondrial dehydrogenase to purple formazan crystals in live cells. The National Cancer Institute (NCI) established that extracts with IC 50 values <30 µg/mL against cancer cell lines are promising for isolation of bioactive compounds (Suffness and Pezzuto, 1990).
Cells were seeded in 96-well plates at 8 x 10 4 cells/well then incubated for 24 hours at 37°C, 5% CO 2 and 95% humidity. Cells were then treated with the different fractions dissolved in DMSO (10 µg/mL). DMSO was used as negative control and doxorubicin as positive control. Three independent experiments were carried out. DOI:10.31557/APJCP.2021.22.1.53 Cytotoxicity of Aglaforbesin Derivative carbonyl cyanide 3-chlorophenylhydrazone (CCCP) for 5 min prior to JC-1 staining. Cells were stained with JC-1 according to the manufacturer's instruction with some modifications. Briefly, after exposure to treatments, JC-1 solution was added to the wells to a final concentration of 2μM followed by 30 min incubation. Cells were washed with PBS and fluorescence was measured using Varioskan LUX Multimode Microplate Reader (Thermo Fisher Scientific). Fluorescence was detected at Ex/Em 535/595nm for red and 485/535nm for green. Reduction in red to green fluorescence ratio is indicative of mitochondrial membrane depolarization. Three independent experiments were done with triplicates per experiment. Three representative micrographs per replicate were taken using a fluorescence microscope.

Detection of apoptosis by Caspase 3/7 assay
Assessment of caspase 3/7 activation is essential in determining the execution of apoptosis in cells treated with a cytotoxic agent. CellEvent Caspase 3/7 ReadyProbes™ Reagent (Invitrogen, CA, USA), a fluorometric assay that allows monitoring of caspase 3/7 activation in treated cells over time, was used to determine if HPLC7 promotes apoptosis by activating caspase 3/7. Cells seeded in a 96-well plate were treated with HPLC7 (IC 50 and IC 80 ) after 24 hours. CellEvent Caspase 3/7 ReadyProbes™ reagent was added to each well to a final concentration of 2 µM. Plates were incubated and progression of apoptosis was checked every 24 hours. After 72 hours, cell nuclei were counterstained with Hoechst 33342 to a final concentration of 1 µg/mL (Invitrogen, CA, USA) and viewed using FITC/Alexa Fluor™ 488 filter settings in a fluorescence microscope at 400X magnification. Three independent experiments were performed in triplicates. Five representative micrographs were taken per replicate. The number of caspase positive cells were counted and the percentage of apoptotic cells was computed using the formula:

Detection of apoptosis by TUNEL assay
DNA breaks in apoptotic cells treated with HPLC7 was determined using terminal deoxynucleotidyl transferase (TdT) mediated-16-deoxyyuridine triphosphate (dUTP) Nick-End Labelling (TUNEL) to visualize DNA fragmentation. DeadEnd™ fluorometric TUNEL system (Promega, WI, USA) was used according to the manufacturer's protocol with some modification. HCT116 cells were seeded in 96-well plates and treated with the cytotoxic HPLC7 (IC 50 , and IC 80 ) for 72 hours. Cells were harvested using trypsin, fixed in 4% formaldehyde, permeabilized with 0.2% Triton X-100 and labeled with TdT reaction mix. Hoechst 33342 was used to counterstain the nuclei. Stained cells were viewed using fluorescence microscope at 400X magnification. Three independent experiments in triplicates were performed. Five representative images per replicate were taken. The percentage of apoptotic cells was computed using the formula: The plates were incubated for 72 hours. After treatment, the media were withdrawn and 20 µL MTT at 5 mg/mL PBS was added to each well. Plates were incubated for four hours, then 150 µL DMSO was added per well to dissolve the formazan crystals. Absorbance was read at 570nm using LeDetect microplate reader (Labexim, EU). Assuming 100% viability in control cells, inhibition was calculated using the formula: IC 50 was determined using plots of percentages inhibition against concentrations with non-linear regression analysis using Graphpad Prism 6.0 software (San Diego, CA).

Cell selectivity of the cytotoxic HPLC isolate
A promising anticancer compound should be specifically toxic against cancer cells but with minimal or no effect on non-cancer cell models. Recent researches are primarily directed towards discovering potent anticancer agents that are non-toxic to normal cells. The selectivity index (SI) value shows the selective toxicity of the sample to the cell lines tested. SI value ≥ 2 indicates promising selectivity against cancer cell lines (de Oliveira et al., 2015).
The cytotoxicity of HPLC7 fraction against HK-2 cells was determined as previously described. The SI of HPLC7 was calculated using the formula:

Apoptotic effect of HPLC7 on HCT116
HPLC7 demonstrated cytotoxicity and good selectivity against HCT116 hence, its apoptotic effect was investigated.

Documentation of cellular morphology
The morphological features of cells treated with HPLC7 were observed through light microscopy. Apoptotic morphologies noted included cell shrinkage, nuclear condensation, cell rounding and detachment from substratum, membrane blebbing and formation of apoptotic bodies. Three independent trials in triplicates were performed. Five micrographs per replicate were captured after the experiment.

Analysis of mitochondrial membrane depolarization by JC-1 assay
Since the intrinsic apoptosis pathway is characterized by decreased mitochondrial membrane potential, the effect of HPLC7 on mitochondrial membrane potential was investigated by JC-1 Assay (Invitrogen, CA, USA). Mitochondrial membrane depolarization, an early apoptotic event (Suzuki-Karasaki et al., 2013), was monitored after 24-hour treatment. HCT116 cells were seeded in a 96 well plate and were treated with DMSO and HPLC7 at IC 50 (1.13 µg/mL) and IC 80 (3.38 µg/mL) for 24 hours. Cells were treated with the positive control,

Statistical analysis
The IC 50 values are presented as means ± standard deviation (SD) of three independent experiments. Statistical differences were determined by one-way ANOVA followed by Tukey's multiple comparison test or by the non-parametric Kruskal-Wallis test with a posthoc Dunn's test. All statistical analyses were performed using Graph Pad Prism 6.0. Differences were considered significant at P<0.05.

Bioassay-guided purification of A. loheri extract
The methanolic crude extract of A. loheri leaves was subjected to a bioassay-guided fractionation and cytotoxic activity of the fractions was evaluated on HCT116 by MTT assay. The cytotoxic HPLC isolate, a white amorphous powder, obtained after fractionation was determined as aglaforbesin derivative (AFD) by NMR analyses.
The 1 H NMR spectrum of HPLC7 isolate indicated the presence of two methyl signals at δ 0.85 (d, J = 6.6 Hz), and 0.90 (d, J = 6.6 Hz), two methoxy signals at δ 3.66 The 13 C NMR spectrum showed 31 carbon signals including two for methoxy carbons and two for amide carbons at δ 58.0, 55.5, 170.5, and 174.6, respectively. HSQC shows correlation between the benzodioxolo methylene protons at δ 5.92 (d, J = 1.8 Hz) and 6.18 (d, J = 1.8 Hz) and oxygenated carbon at δ 105.4. The oxymethine proton at δ 3.93 (s) is also directly bonded to the oxygenated carbon at δ 85.9 based on HSQC. Spectral data were carefully compared with previously published values for compounds isolated from Aglaia species giving a good match with the aglaforbesin derivative (Figure 1) previously isolated from the twigs of Aglaia oligophylla by Dreyer et al. (2001). The presence of piriferine-like substituent attached to C-3, and the aromatic substituents attached to C-4, and C-2 are supported by 1D and 2D NMR data.

Cytotoxicity of A. loheri extracts
The study revealed the plant to be a source of cytotoxic agents against cancer cells as demonstrated by the abundance of active fractions (Table 1). AFD isolated in the present study exhibited strong cytotoxicity against HCT116 with IC 50 value of 1.13 ± 0.07 µg/mL but low toxicity to the normal HK-2 cells with IC 50 value of 6.81 ± 1.8 µg/mL. The preliminary selectivity testing demonstrated in a high selectivity index of 6.04 against HCT116 (Supplementary Figure 1), increasing its potential for development as anticancer drug lead. AFD also demonstrated strong cytotoxicity but less selectivity against breast cancer (MCF7) and lung cancer (A549) cells ( Supplementary Table1, Supplementary Figure 1). This is the first study to report on the potent cytotoxic activity of AFD against cancer cell lines. Previous investigation on AFD reported no bioactivity when tested against the larvae of Spodoptera littoralis (Dreyer et al., 2001).

AFD-induced apoptosis in HCT116 cells
The effect of AFD on HCT116 cell morphology was evaluated after 72 hours treatment to determine apoptotic features. Figure 2 shows that cells treated with AFD exhibited characteristic features of apoptosis such as rounding and detachment from the substratum of culture plate, membrane blebbing and formation of apoptotic bodies also observed in other studies (Kroemer et al., 2009;Zhang et al., 2018). Hence, biochemical markers of apoptosis were investigated including mitochondrial membrane depolarization, caspase 3/7 activation, and DNA fragmentation. To determine the effect of AFD treatment on mitochondrial membrane potential of HCT116 cells, JC-1 assay was used. Micrographs of AFD-treated cells showed no visible difference in red and green fluorescence from that of DMSO-treated cells ( Figure 3A). The computed red to green fluorescence ratio of cells treated with AFD showed no significant difference from that of DMSO ( Figure 3B). This shows that AFD did not induce mitochondrial membrane depolarization in HCT116 cells after 24 hours.
To further investigate apoptosis as the mode of cell death induced by AFD, caspase 3/7 activation was   visualized by fluorescence microscopy. Representative micrographs and comparison of caspase 3/7 activation are shown in Figure 4A. Treatment of HCT116 with AFD IC 50 for 72 hours showed caspase 3/7 activation in about 10% of cells while treatment with AFD IC 80 showed increased caspase 3/7 activation to 35% ( Figure 4B). However, caspase 3/7 activity of doxorubicin (93%) was significantly higher compared to DMSO and AFD IC 50 treatments.
Although activation of caspases is suggestive of apoptosis, the presence of active caspases is not sufficient to define apoptosis (Kroemer et al., 2009). Hence, TUNEL assay was conducted which subsequently confirmed the apoptosis-inducing activity of AFD evidenced by DNA fragmentation. Activation of caspases, especially caspase 3, is known to induce DNA fragmentation and that leads to cell death (Kitazumi and Tsukahara 2011;Slee et al. 2001). Additionally, caspase 3 is essential to membrane blebbing by activating Rho-activated serine/ threonine kinase ROCK1 that promotes the movement of  DNA fragments to the bleb and the formation of apoptotic bodies (Zhang et al., 2018). Representative micrographs in Figure 5A showing intense green fluorescence of nuclei, were clearly detected in cells treated with AFD and DNase (positive control) but not in cells treated only with DMSO. Figure 5B showed occurrence of apoptotic DNA fragmentation in HCT116 cells treated with AFD IC 80 (54.35%) that exhibit no significant difference from that of DNase (88.23%). However, treatment with AFD IC 50 (14.61%) showed no significant difference from that of DMSO (0.62%).

Discussion
The genus Aglaia has received considerable attention in the past decades after the isolation of rocaglamide and their related compounds which demonstrated potential significance in agro-chemistry and pharmacology (Ebada et al., 2011). In the present study, the bioassay-guided fractionation of A. loheri leaf extracts revealed the potential of the species as source of cytotoxic agents against human colorectal cancer as demonstrated by the abundance of cytotoxic fractions against HCT116. Most importantly, this investigation lead to the isolation of the selectively cytotoxic agent which was structurally elucidated as aglaforbesin derivative.