Ethical approval for plant use in this study was waived by the Research Institute for Health Sciences, Walailak University, which also approved the research protocol.

Tamarindus indica L. fruits were collected in the southern part of Thailand during the peak harvest season from December, 2020 to January, 2021. The seeds were removed and thoroughly washed with tap water, dried in a hot air oven overnight and ground with an electric blender. Then, 50 g of the dried seed powder was extracted in 200 ml of 95% ethanol for seven days at room temperature. The extract solution was filtered and then concentrated under reduced pressure at 40°C using a rotary evaporator, Rotavapor ^® R-100 (Buchi Labortechnik AG, Flawil, Switzerland). The sample was air-dried at room temperature to eliminate the remainder of the solvent. ²⁵ Crude extract of Tamarindus indica L. seeds (TS) was dissolved in 100% dimethyl sulfoxide (DMSO) and stored at 4°C for further experiments.

An in vitro free radical scavenging activity was performed by 2,2-diphenyl-2-picrylhydrazyl (DPPH) (Sigma-Aldrich) scavenging assay. ²⁶ Ten millimolar DPPH solution was diluted in absolute methanol and the working reagent of DPPH was adjusted to an optical density of 0.07 ± 0.02 at 517 nm. Ten microliters of different concentrations (0.125–0.4 mg/ml) of the TS extracts and ascorbic acid (used as standard) were thoroughly mixed with 90 μl of freshly prepared DPPH reagent in a 96-well plate. The reaction mixture was incubated in the dark for 30 mins at room temperature. The absorbance was measured at 517 nm with Eon ^TM Microplate Spectrophotometer (BioTek Instruments, Inc. Vermont, USA). The measurement was repeated with three individual experiments. The decrease of absorbance determined scavenging activity. The percentage of DPPH radical scavenging activity was calculated from [(A ₅₁₇ of control – A ₅₁₇ of sample)/A ₅₁₇ of control] × 100. IC ₅₀ value was determined from the percentage inhibition graph, obtained using the y = mx + c formula from the slope of the graph.

The 2,2′-Azino-bis (3-ethylbenzothiazoline-6-sulfonic acid (ABTS) assay is based on decolorization occurring when blue/green ABTS ^•+ is reduced by antioxidants to ABTS ^• (2,2′-Azino-bis (3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt (Sigma-Aldrich). ABTS was dissolved to a concentration of 7 mM in water. ²⁷ The radicals of ABTS ^•+ were generated by reaction of 7 mM ABTS ^• and 2.45 mM potassium persulfate at a 1:1 ratio. The mixture was incubated in the dark for 16 to 18 h at room temperature. Then, the ABTS ^•+ solution was further diluted in absolute methanol until an absorbance value of 734 nm was reached at 0.07 ± 0.02 optical density. The different concentrations (0.125–0.4 mg/ml) of TS extract and ascorbic acid standard were mixed with 90 μl of ABTS ^•+ working solution in a 96-well plate. The reactions were incubated for 45 mins at room temperature and the absorbance was measured at 734 nm. The experiment was performed in triplicate. The percentage of ABTS ^•+ scavenging activity was calculated as [(A ₇₃₄ of control – A ₇₃₄ of sample)/A ₇₃₄ of control] × 100. IC ₅₀ value was determined from the percentage inhibition graph obtained using the y = mx + c formula from the slope of the graph.

The rat cardiomyoblast cell line (H9c2) was obtained from ATCC ^® CRL-1446 (RRID:CVCL_0286). The cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM, Invitrogen) with 10% fetal bovine serum and 1% penicillin/streptomycin (Thermo Fisher Scientific). Cells were cultured for 5 to 7 days and maintained in 5% CO ₂/95%-humidified air at 37°C.

After cells reached 80% confluence, cells were trypsinized with 0.05% trypsin and plated with a density of 5 × 10 ³ cells/well in 96-well plates for 48 h. Cell culture media was replaced with serum free medium before being subjected to hypoxia for 4 h. Hypoxic condition was induced by overlayed mineral oil on the culture media to mimic hypoxic conditions. ^{28 , 29} In some experiments, cells were pretreated with TS extract for a 48-h period before and during hypoxic induction.

Cell viability was determined by 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2-H-tetrazolium bromide (MTT) assay. The medium was replaced with 0.5 mg/ml MTT (Invitrogen, Thermo Fisher Scientific) in phosphate buffer saline (PBS) and incubated in the dark for 4 h at 37 °C. The medium was then removed and 100 μl dimethyl sulfoxide (DMSO) was added into each well. The absorbance values were measured at 570 nm using the Eon ^TM Microplate Spectrophotometer (BioTek Instruments, Inc., VT, USA). The relative percentage of cell viability was compared with the control group.

Cells were seeded in a black-wall 96-well plate at a density of 5 × 10 ³ cells/well and maintained in completed DMEM overnight in 5% CO ₂/95%-humidified air at 37°C. The medium was discarded and replaced with condition media containing 25 μM 2′,7′-dichlorofluorescin diacetate (DCFH-DA, Sigma-Aldrich). In the hypoxia group, mineral oil was overlayed on condition media and incubated in 5% CO ₂/95%-humidified air at 37°C for 4 h. In the presence of ROS, the DCHF-DA was rapidly oxidized to highly fluorescent 2′,7′-dichlorofluorescine (DCF). After removing the leftover dye, PBS was added and the intracellular production of ROS was measured by fluorescence intensity with excitation and emission wavelengths of 498 and 522 nm, respectively, with Synergy Mx Microplate Reader (BioTek Instruments Inc., VT, USA).

To measure hypoxic condition, HIF-1α was determined. GENEzol ^TM Reagent (Geneaid, Taiwan) was added directly to 7 × 10 ⁴ cultured cells in a 12-well plate. The cells were lysed by pipetting several times and the sample mixture was incubated for 5 mins at room temperature (RT). Total RNA from the cells was extracted with a GF-1 extraction kit (catalog #GF-TR-100, Vivantis Technologies, Malaysia). The sample was transferred into an RNA Binding Column assembled in a collection tube and centrifuged at 10,000 × g for 1 min then the flow-through was discarded. Wash buffer provided by the kit was then added and centrifuged at 10,000 × g for 1 min. DNase I Digestion Mix provided by the kit was subsequently added into the RNA Binding Column and incubated at RT for 15 mins. Inhibitor Removal Buffer was then added, centrifuged and the flow-through was discarded. The membrane was washed twice with Wash Buffer before the column was placed into a new microcentrifuge tube. Forty microliters of RNase-free water were added directly into the membrane and incubated for 1 min before being centrifuged at 10,000 × g for 1 min. The RNA concentration was determined by NanoDrop One ^C (Thermo Fisher Scientific, MA, USA). Complementary DNA synthesis was performed using 0.1 μg of RNA with Viva cDNA Synthesis Kit (catalog #cDSK01-050, Vivantis Technologies, Selangor, Malaysia) following the manufacturer’s protocol.

Quantitative PCR (qPCR) reaction mix was prepared using iTaq™ Universal SYBR ^® Green Supermix kit (catalog #1725121, Bio-Rad Laboratories, CA, USA). Ten microliters of iTaq™ Universal SYBR Green Supermix (2X) was mixed with 100 ng of cDNA and 1 μl of 200 mM F + R primers. The total reaction mix volume was adjusted to 20 μl with diethyl pyrocarbonate (DEPC) water. The software StepOnePlus™ Real-Time PCR system (Applied Biosystems, Waltham, MA, USA) was set with a thermal cycle as follows: holding stage 95 °C for 30 s; cycling stage at 95 °C for 15 s; 60 °C for 60 s for 40 cycles; and then melting curve stage at 95 °C for 15 s, 60 °C for 60 s, and 95 °C for 15 s with a temperature increase of 0.3 °C. SYBR Green primers were HIF-1α (Forward primer: 5′ ACACAGAAATGGCCCAGTGAG; Reverse primer: 5′ CACCTTCCACGTTGCTGACTT), ³⁰ IL-1β (Forward primer: 5′ CACCTCTCAAGCAGAGCACAG; Reverse primer: GGGTTCCATGGTGAAGTCAAC) and IL-6 (Forward primer: 5′ TCCTACCCCAACTTCCAATGCTC; Reverse primer: TTGGATGGTCTTGGTCCTTAGCC). ³¹ Measurements were performed in triplicates and relative gene expression levels were normalized to endogenous reference gene, GAPDH (Forward primer: 5′ TTCCTACCCCCAATGTATCCG; Reverse primer: 5′ CATGAGGTCCACCACCCTGTT). ³² Relative fold change in expression was calculated using the 2 ^-∆∆Ct method.

Cells were seeded in a six-well plate at a density of 1 × 10 ⁵ cells/well. At the end of each experiment, the culture plate was placed on ice and then the media was removed. Cells were washed with ice-cold PBS and proteins were collected with 2X Laemmli Sample Buffer (catalog #161-0737, Bio-Rad Laboratories, CA, USA) containing 2-mercaptoethanol and homogenized with insulin syringes. The lysates in 2X Laemmli buffer were denatured at 95°C for 5 mins in a heating box. The cellular protein was separated on 12% SDS-polyacrylamide gel at 90 V for 10 mins followed by 100 V for ~90 mins and then was transferred to polyvinylidene fluoride (PVDF) membranes (100 V; 100 mins). The membranes were blocked with 5% nonfat dry milk in Tris-buffered saline (pH 7.4) containing 0.1% Tween-20 (TBST) for 1 h, followed by incubation with primary antibodies, at the dilution of 1:1,000, cleaved caspase-3 rabbit polyclonal antibody (catalog #9661, RRID:AB_2341188, Cell Signaling Technology, MA, USA), cleaved caspase-8 rabbit monoclonal antibody (catalog #9496, RRID:AB_561381, Cell Signaling Technology, MA, USA), Bax rabbit polyclonal antibody (catalog #sc-493, RRID:AB_2227995, Santa Cruz Biotechnology, Inc., TX, USA) and Bcl-2 rabbit polyclonal antibody (catalog #sc-492, RRID:AB_2064290 , Santa Cruz Biotechnology, Inc., TX, USA) in 1% nonfat dry milk in TBST at 4°C overnight. Membranes were washed and incubated with 1:5,000 dilution goat anti-rabbit IgG secondary antibody, horseradish peroxidase (HRP)-conjugated (catalog #AP132P, Merck & Co., NJ, USA) at room temperature for 1 h. Signals were visualized by Affinity ^® Western blot ECL kit (Affinity Biosciences, OH, USA). The digital images of chemiluminescent Western blots were obtained from ImageQuant ^TM LAS 500 (GE Healthcare, Life Sciences, NJ, USA). Results were expressed relative to β-actin (Santa Cruz Biotechnology) expression bands on the same samples. The protein quantifications were performed using ImageJ (Version 1.53) (RRID:SCR_003070).

All data are expressed in mean ± SEM. Statistical comparisons between each time point and a single control group were performed with one-way ANOVA followed by Tukey’s multiple comparisons test. The comparisons between two groups of studies were performed with unpaired Student’s t tests. Statistical analyses were performed using GraphPad Prism (Version 9.0). All data analyses performed using this software have been included in the manuscript. ³³ A p-value of <0.05 was considered statistically significant. R programming language is an open-access alternative that can perform functions equivalent to GraphPad Prism software.

To optimize timing of hypoxia-induced cell injury, H9c2 cells were subjected to hypoxic conditions for 30 min, 1 h, 4 h or 6 h. Cell viability was determined by MTT assay. Under hypoxic condition, cell viability was significantly decreased to 54 ± 0.6%, 50 ± 0.7%, 49 ± 1.2% and 54 ± 2.7% at 30 min, 1 h, 4 h or 6 h, respectively, compared with the control ( Figure 1: see also Underlying data Fig 1 ³³ ).

Antioxidant activity of TS extract was performed by DPPH and ABTS assays. The crude extract of TS showed potent antioxidant activities obtained from both assays. The DPPH scavenging activity of TS extract displayed 79 ± 5.8% at 0.4 mg/ml and 24 ± 6.1% at 0.1 mg/ml. Similarly, the DPPH radical scavenging activity of ascorbic acid was 89 ± 1.8% at 0.4 mg/ml and 24 ± 3.6% at 0.1 mg/ml ( Figure 2A: see also Underlying data Fig 2A ³³ ). Moreover, the scavenging activity of ABTS ⁺ radicals of TS extract was observed as a potential antioxidant property; TS extract displayed 100 ± 0.1% scavenging activity at 0.4 mg/ml and 63 ± 2.3% at 0.1 mg/ml, while ascorbic acid served as the standard antioxidant, showing maximum scavenging activity of 100 ± 0.1% at 0.4 mg/ml and 54 ± 2.0% at 0.1 mg/ml ( Figure 2B: see also Underlying data Fig 2B ³³ ). The IC ₅₀ values of DPPH free radical scavenging and ABTS radical scavenging were 236 μg/ml and 82 μg/ml, respectively, for TS extract.

To determine the optimal concentration of TS on H9c2 cells that does not cause cellular injury, H9c2 was treated with different concentrations of TS crude extract, varying from 10 μg/ml to 200 μg/ml, for 48 h. The cytotoxicity of TS was measured with MTT assay. The results showed that TS extract at concentrations of 10 μg/ml, 20 μg/ml and 50 μg/ml did not affect cell viability when compared with the untreated control group, 101.3 ± 3.6%, 100.9 ± 8.8% and 83.5 ± 4.3%, respectively. However, TS extract at 100 μg/ml and 200 μg/ml significantly reduced cell viability by 76.8 ± 6.4% (p = 0.0272) and 26.2 ± 2.1% (p <0.0001), respectively, compared with the untreated control group ( Figure 3A: see also Underlying data Fig 3A ³³ ).

To examine the effect of TS extract treatment on cell viability under hypoxia, H9c2 cells were pretreated with TS extract at a concentration of 10 μg/ml for 48 h before being subjected to hypoxia for 4 h. The results showed that hypoxic condition significantly reduced cell viability by MTT assay to 60 ± 4.2%, compared with control (p = 0.0008). Pretreatment with 10 μg/ml of TS extract increased cell viability to 72 ± 2.5%, compared with CTL (p = 0.0151). Therefore, TS extract increased cell viability by 21% compared with hypoxia-induced cell injury ( Figure 3B: see also Underlying data Fig 3B ³³ ).

In order to investigate whether TS extract could reduce intracellular ROS production, H9c2 cells were cultured under hypoxic condition for 4 h and measured with DCFH-DA. It was shown that hypoxic condition significantly increased relative ROS production to 29.7 ± 3.4%, compared with CTL (p <0.0001, Figure 4). Pretreatment of 10 μg/ml of TS extract with an additional treatment during hypoxia significantly reduced relative intracellular ROS production to 18.9 ± 2.0 compared with hypoxic group (p = 0.0136, Figure 4: see also Underlying data Fig 4 ³³ ).

The key transcription factor, HIF-1α, is a response to hypoxic injury. The effects of TS extract-mediated cellular inflammation under hypoxic stimulation on HIF-1α mRNA expression were measured. The effects are shown as fold-changes relative to the control (see also Underlying data Fig 5A–C ³³ ). The results showed that hypoxic condition significantly increased HIF-1α expression by 2.5 ± 0.37 when compared with control (p = 0.0421, Figure 5A), confirming a hypoxic response in H9c2 cells. Treatment with TS extract significantly reduced HIF-1α expression by 3.7-fold, compared with hypoxia group (p = 0.0160, Figure 5A).

To determine whether TS extract mediates inflammation, the mRNA expression of inflammatory cytokines, IL-1β and IL-6, were measured. Under hypoxic condition, the mRNA expressions of IL-1β and IL-6 increased by 3.7 ± 0.89 and 1.6 ± 0.23, respectively, compared with the control ( Figure 5B, C). Most importantly, pretreatment with TS extract significantly attenuated mRNA expression of IL-1β and IL-6 by 5.2- and 3.1-fold, compared with hypoxia, with nearly complete reversal of changes ( Figure 5B, C: see also Underlying data Fig 5A–C and supplementary file S1–S7 ³³ ). Overall, these results indicate that TS extract protected against hypoxia-induced cytokine production.

To assess the effect of TS extract on H9c2 cells apoptosis under hypoxic condition, TS extract at a concentration of 10 μg/ml was pretreated for 48 h prior to hypoxia. The levels of apoptotic regulatory protein cleaved-caspase 3, cleaved-caspase 8 and Bax were assessed by Western blot. The representative immunoblots were normalized to actin. The results showed that hypoxic condition significantly increased the level of cleaved-caspase 3, cleaved-caspase 8 and Bax to 0.34 ± 0.04, 0.40 ± 0.08 and 1.35 ± 0.19, respectively, compared with the control ( Figure 6A– C: see also Underlying data Fig 6A–6C ³³ ). Interestingly, TS extract was markedly decreased in cleaved-caspase 3, cleaved-caspase 8 and Bax levels by 1.4-fold, 4.3-fold and 1.8-fold, respectively, compared with hypoxic condition. However, the level of Bcl-2 protein expression was not different between groups ( Figure 6D– E: see also Underlying data Fig 6D–6E and supplementary file S8 ³³ ). Our results suggest that TS extract plays an important role in inhibiting cellular apoptosis associated with hypoxic-promoted cell death.