Chemical Constituents from the Leaves of Eruca sativa (Rocket Salad) Exhibiting Dual Cardioprotective and Antiplatelet Activities

Abstract

Eruca sativa (rocket salad; arugula) is a well-known edible plant commonly used in salads and as a topping for pizza and pasta. In this study, one new cyclic thiocarbamate (1) and four new kaempferol glycosides (2–5), along with 15 known compounds (6–20), were isolated from leaves of E. sativa using chromatographic methods. The structures of all compounds were elucidated based on their HR-ESI-MS, ¹H and ¹³C NMR, and 2D NMR data. All isolates were evaluated for their cardioprotective and antiplatelet activities. Among them, compound 2 exhibited the most potent antioxidant activity in a DPPH assay. Further investigation revealed that compound 2 significantly reduced cellular ROS levels in H9c2 cardiomyocytes under hypoxia/reoxygenation conditions and effectively decreased apoptotic cell death in H9c2 cells, as determined by Annexin V-FITC/PI staining. Additionally, compound 2 demonstrated dose-dependent inhibition of CRP-induced platelet aggregation in human washed platelets, suggesting its potential as a dual-functional candidate for cardiovascular protection.

1. Introduction

Eruca sativa Mill., commonly known as arugula or rocket salad, is an edible vegetable from the Brassicaceae family. Brassicaceae family is globally recognized as the second-largest family in terms of both production and consumption through all other vegetables. In Greek, E. sativa is referred to as “Euzomon”, a name signifying “good broth”, highlighting its delightful and flavorful characteristics. It has been widely spread since Roman ages in the Mediterranean region and is currently distributed globally. The leaves and shoots of this vegetable are commonly incorporated into salads for their spicy and pungent flavor, cooked with meat and processed in pesto, enhancing the taste of otherwise plain salads. In Italy, it is used in pizza toppings, pasta and added as seasoning to meat and fish. , In Brazil, arugula pizza has high acceptability as a complete meal without any supplement to students from public schools. Historically, it has been used in traditional practices for a variety of medicinal applications, including as an astringent, diuretic, digestive, emollient, tonic, depurative, laxative, rubefacient, and stimulant. ,,

Oxidative stress plays a pivotal role in the pathophysiology of cardiovascular diseases. , In cardiomyocytes, ischemia followed by reperfusion leads to excessive production of reactive oxygen species (ROS), primarily through mitochondrial electron transport chain dysfunction. This ROS overload triggers mitochondrial membrane depolarization, cytochrome c release, and caspase-dependent apoptosis, culminating in myocardial injury and impaired cardiac function. , Parallel to this, platelet-derived ROS serve as secondary messengers that amplify platelet activation and aggregation, especially under pathological stimuli such as collagen exposure. Notably, stimulation of glycoprotein VI (GPVI) on platelets by agonists like CRP-XL leads to the activation of NADPH oxidase (NOX), generating ROS that reinforce integrin αIIbβ3 activation and thrombus formation. , While platelet aggregation serves to prevent blood loss (hemostasis) in response to vascular injury, its excessive occurrence in atherosclerotic coronary arteries or other diseased blood vessels could cause the formation of thrombosis that can block brain arteries, resulting in brain damage and stroke symptoms. − Thus, strategies that suppress ROS in both cardiomyocytes and platelets may offer comprehensive protection against thrombotic and ischemic cardiovascular events.

Previous phytochemical investigations have reported various bioactive chemical constituents of glucosinolates, flavonoids and unsaturated fatty acids. , The crude extracts or purified constituents of E. sativa has been reported to possess antiobesity, antibacterial, antihyperuricemia, anti-inflammatory and antioxidant activities. , In addition, Fuentes et al. have shown that E. sativa extract exhibited antiplatelet and antithrombotic activities. Furthermore, E. sativa extract showed antithrombotic activity in murine models. However, the cardioprotective and antiplatelet activities of chemical constituents isolated from E. sativa have not been described.

In the present study, we aimed to isolate and structurally characterize bioactive constituents from the methanolic extract of E. sativa leaves (Figure ) and evaluate their cardioprotective and antiplatelet activities. We further investigated their biological activities in two pathologically relevant models: (1) ROS production and apoptotic cell death in cardiomyocytes under hypoxia/reoxygenation (H/R) conditions and (2) ROS-mediated platelet aggregation following CRP-XL stimulation.

1.

Chemical structures of compounds isolated from leaves of E. sativa Mill.

2. Materials and Methods

2.1. General Experimental Procedure

Optical rotations were measured with a JASCO P-2000 polarimeter (JASCO, Easton, MD, USA). UV spectra were measured using a Chirascan CD spectrometer (Applied Photophysics, Leatherhead, UK). Fourier transform infrared (FT-IR) spectra were measured on a JASCO FT/IR-4700 (JASCO). The NMR spectra were measured on JEOL 500 (JEOL Ltd., Tokyo, Japan) and Bruker AVANCE 800 NMR spectrometers (Bruker, Billerica, MA, USA). High-resolution mass spectrometric data were obtained using an Agilent Q-TOF 6530 mass spectrometer (Agilent, Santa Clara, CA, USA). HPLC was subjected on a Gilson 321 pump and Gilson 172 diode array detector (Gilson, Madison, WI, USA), and various HPLC columns. The MPLC was carried out on Biotage Isolera one (Biotage, Uppsala, Sweden). Biotage Sfär C18 D - Duo 100 Å 30 μm (Biotage) and Silica gel (CHROMATOREX SILICA GEL, grade GS60-40/75, Fuji Silysia Chemical Ltd., Japan) were used for column chromatographic separations; Sephadex LH-20 (GE Healthcare, Uppsala, Sweden) were conducted, and TLC on silica gel 60 F₂₅₄ plates (Merck, Darmstadt, Germany) was performed. Solvents for extraction and isolation were purchased from SK Chemical (Seoul, Republic of Korea). The solvent for NMR was purchased from Cambridge Isotope Laboratories, Inc. (Andover, MA, USA). Hydrochloric acid, sulfuric acid, formic acid and pyridine were purchased from Sigma-Aldrich (Merck). O-Tolyl isothiocyanate and l-cysteine methyl ester hydrochloride were obtained from Tokyo Chemical Industry (Tokyo, Japan).

2.2. Plant Materials

The leaves of Eruca sativa Mill. (Brassicaceae) were purchased in April 2022 from the local market at Ansan-si, Gyeonggi Province, Republic of Korea. The plant material was identified by Professor Young-Won Chin, College of Pharmacy, Seoul National University, Republic of Korea. A voucher specimen (CYWSNU-KP0026) was deposited at College of Pharmacy, Seoul National University, Republic of Korea.

2.3. Extraction and Isolation

The fresh leaves of E. sativa (7.1 kg) were extracted by 30 L of methanol at room temperature for three times (24 h per time), giving the crude methanolic extract (295 g). The crude extract was suspended in H₂O and partitioned with hexane, dichloromethane, ethyl acetate and n-butanol sequentially to give the hexane-soluble extract (56 g), dichloromethane–soluble extract (7.9 g), ethyl acetate-soluble extract (12.6 g), n-butanol-soluble extract (33 g) after removal of the solvents in vacuo. The dichloromethane-soluble extract (7.9 g) was chromatographed with a silica column chromatography using gradients of increasing polarity with hexane–ethyl acetate (20:1 → 2:1) and CHCl₃–MeOH (10:1 → 1:1) as solvents and then fractionated into 14 subfractions (ESD1-ESD14). Fr. ESD13 (30 mg) was subjected on semipreparative HPLC (YMC-Pack Pro-C18(2)­100A-LC, 250 × 10 mm column), eluted with mixture of MeCN–H₂O (15:75 → 20:80), 3.0 mL/min for 30 min to yield 1 (10 mg, t _R 29.5 min) and 16 (0.5 mg, t _R 28.2 min). The n-butanol-soluble extract (33 g) was subjected to Diaion HP-20 eluting with MeOH–H₂O as solvent system (25:75, 50:50, 75:25, 100:0) and then eluting with acetone to obtain five fractions (ESB1, ESB2, ESB3, ESB4, ESB5). Fr. ESB3 (3.0 g) was chromatographed on reverse-phase MPLC column chromatography eluting with a gradient system of MeOH–H₂O (5:95 to 50:50, v/v) to obtain 15 subfractions (ESB3A to ESB3O). Fr. ESB3K (150 mg) was applied on semipreparative HPLC (YMC-Pack Pro-C18(2)­100A-LC, 250 × 10 mm column), eluted with MeCN–H₂O (18:82), 3.0 mL/min to yield 2 (45.0 mg, t _R 42.5 min), 3 (10.0 mg, t _R 47.3 min), 4 (6.0 mg, t _R 49.5 min) and 5 (1.0 mg, t _R 52.1 min). Fr. ESB3G (20 mg) was subjected on semipreparative HPLC (YMC-Pack Pro-C18(2)­100A-LC, 250 × 10 mm column), eluted with mixture of MeCN–H₂O (15:85 → 20:80), 3.0 mL/min for 50 min to yield 15 (1.4 mg, t _R 29.3 min), 17 (1.0 mg, t _R 33.2 min) and 18 (0.6 mg, t _R 38.6 min). Fr. ESB3E (20 mg) was subjected on semipreparative HPLC (YMC-Pack Pro-C18(2)­100A-LC, 250 × 10 mm column), eluted with mixture of MeCN–H₂O (20:80 → 60:40), 3.0 mL/min for 50 min to yield 19 (1.5 mg, t _R 29.5 min) and 20 (1.5 mg, t _R 32.2 min). Fr. ESB3H (100 mg) was subjected on semipreparative HPLC (YMC-Pack Pro-C18(2)­100A-LC, 250 × 10 mm column), eluted with MeCN–H₂O (12:88), 3.0 mL/min to yield 7 (7.0 mg, t _R 31.2 min), 6 (7.0 mg, t _R 33.7 min) and 8 (8.5 mg, t _R 35.8 min). Fr. ESB3M (35 mg) was subjected on semipreparative HPLC (YMC-Pack Pro-C18(2)­100A-LC, 250 × 10 mm column), eluted with mixture of MeCN–H₂O (10:90 → 20:80), 1.0 mL/min for 60 min to yield 12 (3.0 mg, t _R 31.3 min) and 13 (2.5 mg, t _R 42.5 min). Ethyl acetate-soluble extract (12.6 g) was chromatographed on reverse-phase MPLC column chromatography eluting with a gradient system of MeOH–H₂O (5:95 → 80:20, v/v) to obtain 14 subfractions (ESE1 to ESE14). Fr. ESE3 (100 mg) was subjected on semipreparative HPLC (RStech OptimaPak C18, 250 × 10 mm column), eluted with MeCN–H₂O (20:80), 3.0 mL/min to yield 10 (10.0 mg, t _R 32.5 min), 14 (2.5 mg, t _R 38.7 min), 9 (10.0 mg, t _R 53.4 min) and 11 (3.0 mg, t _R 57.8 min).

2.3.1. 1,3-Thiazepane-2-one (1)

White amorphous powder; [α]_D –13.6 (c 0.05, MeOH); UV (MeOH) λ_max (log ε) 203 (1.01) nm; FT-IR (ATR) ν_max 3265, 2925, 1633, 1450, 1431, 1383, 1358, 1319, 1240, 1163, 1105, 1039, 937 cm^–1; ¹H NMR (500 MHz, CD₃OD) data, see Table ; ¹³C NMR (125 MHz, CD₃OD) data, see Table ; HR-ESI-MS m/z 132.0481 [M + H]⁺ (calcd for C₅H₁₀NOS, 132.0478).

1. ¹³C and ¹H NMR Spectroscopic Data (125 and 500 MHz) Recorded in Methanol-d ₄ for Compound 1 .

1
No.	δH	δC
1		175.5
2	2.93, m	30.2
3	2.08, m	28.6
4	1.73, m	28.1
5	3.36, m	41.8

2.3.2. Kaempferol 3-O-β-d-Glucopyranoside-4′-O-[β-d-glucopyranosyl-(1→4)-(6-O-E-sinapoyl)]-β-d-glucopyranoside (2)

Yellow amorphous powder; [α]_D –113.4 (c 0.05, MeOH); UV (MeOH) λ_max (log ε) 203 (1.33), 266 (0.71), 328 (0.94) nm; FT-IR (ATR) ν_max 3379, 2924, 1722, 1651, 1600, 1512, 1460, 1368, 1247, 1182, 1070, 1024 cm^–1; ¹H NMR (800 MHz, CD₃OD) data, see Table ; ¹³C NMR (200 MHz, CD₃OD) data, see Table ; HR-ESI-MS m/z 977.2556 [M – H]⁻ (calcd for C₄₄H₄₉O₂₅, 977.2563).

2. ¹³C and ¹H NMR Spectroscopic Data (200 and 800 MHz) Recorded in Methanol-d ₄ for Compounds 2 and 3 .

		2		3
Position	No.	δH	δC	δH	δC
Aglycone	2		157.7		158.6
	3		135.2		135.9
	4		178.8		179.5
	5		162.4		163.2
	6	6.18, s	99.4	6.20, s	100.2
	7		165.3		166.4
	8	6.22, s	94.1	6.25, s	94.9
	9		157.5		158.2
	10		105.1		105.9
	1′		124.9		125.8
	2′	8.01, d (8.8)	131.4	8.03, d (9.0)	132.1
	3′	7.13, d (8.8)	116.5	7.15, d (9.0)	117.3
	4′		159.9		160.6
	5′	7.13, d (8.8)	116.5	7.15, d (9.0)	117.3
	6′	8.01, d (8.8)	131.4	8.03, d (9.0)	132.1
3-O-Glc	1″	5.22, d (7.7)	103.4	5.21, d (7.7)	104.2
	2″	3.41, m	75.0	3.41, m	75.7
	3″	3.40, m	77.4	3.40, m	77.9
	4″	3.37, m	70.6	3.37, m	71.3
	5″	3.20, m	77.8	3.19, m	78.5
	6″	3.53, dd (11.9, 5.5)	61.9	3.53, dd (11.9, 5.5)	62.6
		3.68, m		3.68, m
4′-O-Glc	1‴	5.11, d (7.7)	100.4	5.10, d (7.7)	101.2
	2‴	3.64, m	73.8	3.61, m	74.6
	3‴	3.73, m	75.7	3.70, m	76.4
	4‴	3.74, m	80.4	3.71, m	81.1
	5‴	3.97, m	73.5	3.96, m	74.3
	6‴	4.55, dd (12.0, 6.1)	63.3	4.52, dd (12.0, 6.4)	64.0
		4.69, dd (12.0, 2.4)		4.68, dd (12.0, 2.4)
4‴-O-Glc	1⁗	4.42, d (7.7)	104.1	4.41, d (7.7)	104.8
	2⁗	3.28, m	74.2	3.27, dd (9.3, 7.7)	73.2
	3⁗	3.39, m	77.2	3.38, m	78.4
	4⁗	3.32, m	70.8	3.31, m	71.5
	5⁗	3.38, m	77.6	3.37, m	78.1
	6⁗	3.91, m	61.9	3.91, m	62.6
		3.69 m		3.69, m
6‴-O-Sin/Fer	CO		168.8		168.9
	α	6.37, d (15.9)	114.9	6.36, d (15.9)	115.2
	β	7.51, d (15.9)	146.9	7.55, d (15.9)	147.3
	1		125.7		127.5
	2	6.79, s	106.1	7.11, d (2.0)	111.7
	3		148.6		149.3
	4		138.9		150.8
	5		148.6	6.67, d (8.2)	116.5
	6	6.79, s	106.1	7.01, dd (8.2, 2.0)	124.2
	3-OMe	3.78, s	56.8	3.81, s	56.5
	5-OMe	3.78, s	56.8

2.3.3. Kaempferol 3-O-β-d-Glucopyranoside-4′-O-[β-d-glucopyranosyl-(1→4)-(6-O-E-feruloyl)]-β-d-glucopyranoside (3)

Yellow amorphous powder; [α]_D –119.4 (c 0.05, MeOH); UV (MeOH) λ_max (log ε) 203 (1.02), 267 (0.50), 326 (0.60); FT-IR (ATR) ν_max 3356, 2922, 2853, 1737, 1650, 1594, 1506, 1371, 1249, 1176, 1072, 1046 cm^–1; ¹H NMR (800 MHz, CD₃OD) data, see Table ; ¹³C NMR (200 MHz, CD₃OD) data, see Table ; HR-ESI-MS m/z 947.2466 [M – H]⁻ (calcd for C₄₃H₄₇O₂₄, 947.2457).

2.3.4. Kaempferol 3-O-(6-O-E-Coumaroyl)-β-d-glucopyranoside-4′-O-β-d-glucopyranoside (4)

Yellow amorphous powder; [α]_D –20.2 (c 0.05, MeOH); UV (MeOH) λ_max (log ε) 201 (1.23), 268 (0.82), 278 (7.49), 314 (1.06); FT-IR (ATR) ν_max 3385, 2921, 2852, 1733, 1632, 1604, 1587, 1362, 1242, 1184, 1074, 1019 cm^–1; ¹H NMR (800 MHz, CD₃OD) data, see Table ; ¹³C NMR (200 MHz, CD₃OD) data, see Table ; HR-ESI-MS m/z 755.1837 [M – H]⁻ (calcd for C₃₆H₃₅O₁₈, 755.1823).

3. ¹³C and ¹H NMR Spectroscopic Data (200 and 800 MHz) Recorded in Methanol-d ₄ for Compounds 4 and 5 .

Position	No.	δH	δC	δH	δC
Aglycone	2		158.5		158.7
	3		135.8		135.7
	4		179.5		179.5
	5		163.1		163.0
	6	6.12, s	100.2	6.12, s	100.1
	7		166.2		166.1
	8	6.29, s	95.0	6.29, s	94.8
	9		158.5		158.5
	10		105.7		105.7
	1′		125.6		125.6
	2′	8.02, d (8.6)	132.0	8.02, d (8.7)	131.9
	3′	7.08, d (8.6)	117.1	7.11, d (8.7)	117.1
	4′		161.3		161.0
	5′	7.08, d (8.6)	117.1	7.11, d (8.7)	117.1
	6′	8.02, d (8.6)	132.0	8.02, d (8.7)	131.9
3-O-Glc	1″	5.20, d (6.7)	104.2	5.24, d (6.7)	104.1
	2″	3.45, m	78.3	3.45, m	78.1
	3″	3.43, m	75.7	3.43, m	75.7
	4″	3.28, m	71.9	3.28, m	71.9
	5″	3.47, m	75.9	3.47, m	75.9
	6″	4.25, dd (11.8, 2.0)	64.6	4.25, m	64.5
		4.16, dd (11.8, 7.2)		4.16, m
6″-O-Fer/p-Coum	CO		168.9		168.8
	α	6.02, d (15.9)	114.7	6.09, d (15.9)	115.0
	β	7.36, d (15.9)	146.7	7.39, d (15.9)	146.9
	1		127.1		127.6
	2	7.28, d (8.1)	131.3	7.05, d (2.0)	111.7
	3	6.77, d (8.1)	116.9		149.3
	4		161.3		150.6
	5	6.77, d (8.1)	116.9	6.81, d (8.1)	116.5
	6	7.28, d (8.1)	131.3	6.93, dd (8.1, 2.0)	124.2
	6-OMe			3.90, s	56.5
4′-O-Glc	1‴	4.95, d (6.7)	101.8	4.95, d (6.7)	101.7
	2‴	3.46, m	74.9	3.46, m	74.9
	3‴	3.48, m	78.0	3.48, m	77.9
	4‴	3.44, m	78.3	3.44, m	78.2
	5‴	3.41, m	71.3	3.41, m	71.3
	6‴	3.88, dd (12.2, 2.3)	62.6	3.90, m	62.5
		3.69, dd (12.2, 5.5)		3.72, dd (12.3, 5.5)

2.3.5. Kaempferol 3-O-(6-O-E-Feruloyl)-β-d-glucopyranoside-4′-O-β-d-glucopyranoside (5)

Yellow amorphous powder; [α]_D –46.6 (c 0.05, MeOH); UV (MeOH) λ_max (log ε) 204 (1.32), 267 (0.82), 327 (1.04); FT-IR (ATR) ν_max 3363, 2922, 2853, 1736, 1649, 1630, 1589, 1361, 1243, 1183, 1069, 1020 cm^–1; ¹H NMR (800 MHz, CD₃OD) data, see Table ; ¹³C NMR (200 MHz, CD₃OD) data, see Table ; HR-ESI-MS m/z 785.1915 [M – H]⁻ (calcd for C₃₇H₃₇O₁₉, 785.1929).

2.4. Sugar Analysis

Sugar analysis of compounds 2–5 was performed using the modified method previously reported by our research group. , After acid hydrolysis and derivatization to their thiocarbamoyl-thiazolidine derivatives, the retention times were compared with those of standard d-glucose and l-glucose samples (Supporting Information, Figures S47–S49).

2.5. DPPH Radical Scavenging Assay

The antioxidant activities of the samples were evaluated using the DPPH antioxidant assay kit (D678-10, Dojindo Laboratories, Kumamoto, Japan), following the manufacturer’s instructions. Briefly, the antioxidant capacities of samples were assessed by measuring their capacity to scavenge the stable free radical, 2,2-diphenyl-1-picrylhydrazyl (DPPH). Antioxidant activity was quantified by calculating the IC₅₀ values from the dose–response curves of each sample. These IC₅₀ values were then compared to that of the standard compound Trolox to determine the Trolox equivalent antioxidant capacity (TEAC). TEAC values were expressed as micromoles of Trolox equivalents per gram of sample (μmol TE/g), based on the following calculation:

$$\mathrm{TEAC}(\mu \mathrm{mol\; TE}/\mathrm{g})={\mathrm{IC}}_{50}\text{\_}\mathrm{Trolox}/{\mathrm{IC}}_{50}\text{\_}\mathrm{sample}$$

2.6. Cell Subculture, Treatment and H/R

H9c2 rat cardio-myoblast cell (CRL-1446, ATCC, Manassas, VA, USA) was maintained in Dulbecco’s modified Eagle’s medium (DMEM, Hyclone, Logan, UT, USA) supplemented with 10% FBS and 1% A/A in a humidified 5% CO₂ atmosphere. Hypoxia was induced by incubating H9c2 cells in a hypoxia chamber (MIC-101; IBscience, Daejeon, Republic of Korea). The cells were maintained at 37 °C with a humidified atmosphere of 5% CO₂, 2% O₂, and 93% N₂ in d-glucose and serum free DMEM. Reoxygenation was subjected to maintain cells with DMEM in 37 °C with a humidified atmosphere of 5% CO₂ followed by hypoxic stress.

2.7. Cellular ROS Measurement

Following H/R treatment, detached cells were stained with 5 μM of 5-(and-6)-chloro­methyl-2′,7′-dichloro­dihydro­fluore­scein diacetate (CM-H₂DCFDA; Invitrogen, Carlsbad, CA, USA) for 15 min at 37 °C in the dark. Fluorescence was analyzed using a FACS Lyric flow cytometer (BD Biosciences, San Jose, CA, USA) with excitation and emission wavelengths of 492 and 525 nm, respectively. Relative changes in fluorescence intensity were analyzed using FlowJo software 10.9 (FlowJo LLC, Ashland, OR, USA).

2.8. Apoptotic Cell Death Assay

Following H/R treatment, detached cells were costained with Annexin V conjugated with fluorescein isothiocyanate (Annexin V-FITC) and propidium iodide (PI; 25 μg/mL) for 20 min at 37 °C in the dark. The percentage of apoptotic cell was determined using a FACS Lyric flow cytometer with excitation and emission wavelengths of 485/535 and 493/636 nm for Annexin V-FITC and PI, respectively.

2.9. Human Platelet Preparation

All procedures involving human blood samples from healthy volunteers were approved by the Seoul National University Institutional Review Board (protocol no. 2206/001-006) and conducted in accordance with relevant guidelines and regulations. Written informed consent was obtained from all participants prior to sample collection. Blood collection and platelet preparation were performed following the method previously published by our group.

2.10. Platelet Aggregation Assay

Washed platelets in Tyrode’s-HEPES buffer were incubated with 1 mM CaCl₂ for 2 min before stimulating with CRP-XL (CambCol Laboratories, Cambridge, United Kingdom). Platelet aggregation was assessed in siliconized cuvettes at 37 °C using a 4-channel aggregometer (Chrono-Log Model 700; Havertown, PA, USA) with continuous stirring (1,000 rpm). Aggregation profiles were recorded via Aggrolink software (Model# 700; Chrono-Log, Havertown) and quantified as percentage light transmission change relative to buffer controls (set as 100% transmission).

2.11. Statistics

Statistical analysis was performed using GraphPad Prism software (version 10.4.2; Boston, MA, USA). Statistical significance was assessed using the Student’s t test for two-group comparisons and one-way ANOVA followed by Bonferroni’s post hoc test for multiple comparisons. p values less than 0.05 were considered statistically significant.

3. Results and Discussion

3.1. Structural Elucidation of Isolated Compounds

Compound 1 was obtained as a white amorphous solid. The molecular formula of 1 (C₅H₉NOS) was deduced from the protonated ion peak at m/z 132.0481 [M + H]⁺ (calcd for C₅H₁₀NOS, 132.0478) from the HR-ESI-MS data. A characteristic isotopic peak at [M + 2 + H] ⁺ peak (4.03%) also indicated the presence of one sulfur atom (Figure S3). The ¹³C NMR and DEPT-135 spectra indicated one carbonyl carbon [δ_C 175.5 (C-1)] and four methylene carbons [δ_C 30.2 (C-2), 28.6 (C-3), 28.1 (C-4) and 41.8 (C-5)]. The ¹H NMR spectrum of 1 showed four signals including δ_H 2.93 (2H, m, H-2), 2.08 (2H, m, H-3), 1.73 (2H, m, H-4) and 3.36 (2H, m, H-5) (Table ). The ¹H and ¹³C NMR spectra of 1 were similar to those of 1,3-thiazepane-2-thione (sativin), a seven-membered cyclic structure isolated from E. sativa sprout, except for the pronounced upfield-shifted chemical shift of C-1 signal from δ_C 201.4 to δ_C 175.5. In the ¹H–¹H COSY spectrum, the correlations between H-2/H-3, H-3/H-4 and H-4/H-5 showed that the methylene units are connected to each other, indicating a 1,4-disubstituted n-butane moiety. Furthermore, the HMBC correlations from H-2 to C-1 and from H-5 to C-1 (Figure ) indicated a seven-membered cyclic thiocarbamate. Compound 1 was therefore structurally determined as 1,3-thiazepane-2-one.

2.

Key HMBC (solid arrow) and ¹H–¹H COSY (bold line) correlations of compounds 1–5.

Compound 2 was isolated as a yellow amorphous solid and its molecular formula was inferred as C₄₄H₅₀O₂₅ based on the deprotonated ion peak at m/z 977.2556 [M – H]⁻ (calcd for C₄₄H₄₉O₂₅, 977.2563) from the HR-ESI-MS data. The ¹H and ¹³C NMR spectra for 2 were shown in Table . The ¹H NMR signals of the aglycone at [δ_H 8.01 (2H, d, J = 8.8 Hz, H-2′ and 6′), 7.13 (2H, d, J = 8.8 Hz, H-3′ and 5′), 6.22 (1H, s, H-8) and 6.18 (1H, s, H-6)] showed the typical pattern of kaempferol derivatives. Subsequently, the signals of aromatic protons at δ_H 6.79 (2H, s, Sin-2), olefinic protons at δ_H 6.37 (1H, d, J = 15.9 Hz, Sin-α) and 7.51 (1H, d, J = 15.9 Hz, Sin-β), as well as signals of two methoxy groups at δ_H 3.87 (6H, s, OCH₃) indicated the presence of a trans-sinapoyl moiety. Furthermore, the ¹H and ¹³C NMR spectra for 2 showed signals of three sugar moieties, combined with the HR-ESI-MS data, suggesting that compound 2 is a sinapoyl acylated kaempferol-O-triglycoside. The large coupling constant for the anomeric protons of sugar moieties at δ_H 5.22 (1H, d, J = 7.7 Hz, H-1″), 5.11 (1H, d, J = 7.7 Hz, H-1‴) and 4.41 (1H, d, J = 7.7 Hz, H-1⁗) indicated the β-configuration. All sugar moieties were determined as d-glucose by comparison of the retention time of their thiocarbamoyl-thiazolidine derivative of the acid hydrolysate with that of the standard sample of d-glucose and l-glucose using a HPLC-UV-based method. , The sites of glycosidation for the aglycone were provided by the HMBC correlation (Figure ) from H-1‴ to δ_C 159.9 (C-4′) and from H-1″ to δ_C 135.2 (C-3). Furthermore, the HMBC correlation from H-1⁗ to δ_C 80.4 (C-4‴) indicated that the interglycosidic linkage is glucosyl-(1⁗ →4‴)-glucose, which was well supported by the apparently downfield-shifted chemical shift of C-4‴ when compared with δ_C 70.6 (C-4″) and δ_C 70.8 (C-4⁗). The position of the sinapoyl moiety was determined at C-6‴–OH via an ester linkage based on the HMBC correlations from the glucosyl C-6‴ protons (δ_H 4.55 and δ_H 4.69) to the sinapoyl carbonyl carbon at δ_C 168.8. Compound 2 was therefore elucidated as kaempferol-3-O-β-d-glucopyranoside-4′-O-[β-d-glucopyranosyl-(1→4)-(6-O-E-sinapoyl)]-β-d-glucopyranoside.

Compound 3 was obtained as a yellow amorphous solid and its molecular formula C₄₃H₄₈O₂₄ was deduced from the deprotonated ion peak at m/z 947.2466 [M – H]⁻ (calcd for C₄₃H₄₇O₂₄, 947.2457) from the HR-ESI-MS data. The ¹H NMR spectrum of 3 (Table ) was similar to those of 2 except for the signals at δ_H 6.36 (1H, d, J = 15.9 Hz, Fer-α), 7.55 (1H, d, J = 15.9 Hz, Fer-β), 7.11 (1H, d, J = 2.0 Hz, Fer-2), 6.67 (1H, d, J = 8.2 Hz, Fer-5), 7.01 (1H, d, J = 8.2, 2.0 Hz, Fer-6) and 3.81 (3H, s, OCH₃), which indicated the presence of a trans-feruloyl moiety. The β-orientation of sugar moieties was determined by the large coupling constant for the anomeric protons at δ_H 5.21 (1H, d, J = 7.7 Hz, H-1″), 5.10 (1H, d, J = 7.7 Hz, H-1‴) and 4.41 (1H, d, J = 7.7 Hz, H-1⁗). All sugar units in 3 were confirmed as d-glucose based on the aforementioned method for 2. Furthermore, the trans-feruloyl moiety was determined at C-6‴–OH via an ester linkage based on the HMBC correlations from the glucosyl C-6‴ protons (δ_H 4.52 and δ_H 4.68) to the trans-feruloyl carbonyl carbon at δ_C 168.9. By detailed interpretation of 2D NMR spectra as shown in Figure , compound 3 was elucidated as kaempferol-3-O-β-d-glucopyranoside-4′-O-[β-d-glucopyranosyl-(1→4)-(6-O-E-feruloyl)]-β-d-glucopyranoside.

Compound 4, a yellow amorphous solid, possessed the molecular formula of C₃₆H₃₆O₁₈ based on its HR-ESI-MS deprotonated ion peak at m/z 755.1837 [M – H]⁻ (calcd for C₃₆H₃₅O₁₈, 755.1823). The ¹H and ¹³C NMR spectra for 4 were shown in Table . Compound 4 was determined to be an acylated kaempferol-O-diglycoside by comparing its 1D NMR spectra and HR-ESI-MS data with those of Compound 2. Subsequently, the ¹H NMR signals of aromatic protons at δ_H 7.28 (2H, J = 8.6 Hz, Coum-2/6) and 6.77 (2H, J = 8.6 Hz, Coum-3/5), olefinic protons at δ_H 6.02 (1H, d, J = 15.9 Hz, Coum-α) and 7.36 (1H, d, J = 15.9 Hz, Coum-β) indicated the presence of a trans-p-coumaroyl moiety. The large coupling constant for the anomeric protons of sugar moieties at δ_H 4.95 (1H, d, J = 6.7 Hz, H-1‴) and 5.20 (1H, d, J = 6.7 Hz, H-1″) indicated the β-configuration and both sugars were confirmed as d-glucose using the same method mentioned for 2. The position of glycosidation for the aglycone was deduced from the HMBC correlation from H-1‴ to δ_C 161.3, (C-4′) and from proton H-1″ to δ_C 135.8, (C-3). The coumaroyl ester was determined to be connected to C_6″–OH based on the HMBC correlations from glucosyl C-6″ protons (δ_H 4.25 and δ_H 4.16) to the coumaroyl carbonyl carbon (δ_C 168.9) (Figure ). Compound 4 was therefore elucidated as kaempferol 3-O-(6-O-E-coumaroyl)-β-d-glucopyranoside-4′-O-β-d-glucopyranoside. The prediction of 4 was previously reported by Aguinagalde et al. using paper chromatograms and alkaline hydrolysis. However, to the best of our knowledge, this is the first report of the isolation and spectroscopic characterization of 4.

Compound 5 was isolated as a yellow amorphous solid with the molecular formula of C₃₇H₃₈O₁₉ based on its deprotonated ion peak at m/z 785.1915 [M – H]⁻ (calcd for C₃₇H₃₇O₁₉, 785.1929) from the HR-ESI-MS data. The 1D NMR spectrum of 5 (Table ) was similar to those of 4 except for the signals corresponding to a trans-feruloyl moiety. The sugar units in 5 were confirmed as d-glucoses using the aforementioned method for 2. The position of the trans-feruloyl moiety was confirmed at C-6″–OH via an ester linkage based on the HMBC correlations from the glucosyl C-6″ protons (δ_H 4.25 and δ_H 4.16) to the trans-feruloyl carbonyl carbon at δ_C 168.8. Based on the detailed interpretation of 2D NMR spectra (Figure ) and comparison with those of 4, compound 5 was confirmed as kaempferol 3-O-(6-O-E-feruloyl)-β-d-glucopyranoside-4′-O-β-d-glucopyranoside.

On the basis of comprehensive analysis of spectroscopic data and comparison with literature, the chemical structures of the 15 known compounds (6–20) (Figure ) were identified as kaempferol 3,4′-di-O-β-d-gluco­pyrano­side (6), quercetin 3,4′-di-O-β-d-gluco­pyrano­side (7), isorhamnetin 3,4′-di-O-β-d-gluco­pyrano­side (8), kaempferol 3-O-β-d-gluco­pyrano­side (9), quercetin 3-O-β-d-gluco­pyrano­side (10), isorhamnetin 3-O-β-d-gluco­pyrano­side (11), rutin (12), isorhamnetin 3-O-rutinoside (13), isorhamnetin 3-O-β-d-glucuronide (14), pomegralignan (15), (−)-epiloliolide (16), (R)-1-(4-ethylphenyl)-1,2-ethanediol (17), (S)-1-(3-ethylphenyl)-1,2-ethanediol (18), 3-methoxy-4-O-β-d-gluco­pyranosyl-methyl benzoate (19), and 1-O-β-sinapoyl glucopyranose (20).

3.2. Antioxidant Activity

In the present study, the biological activities of compounds isolated and characterized from E. sativa were comprehensively investigated. The antioxidant potential of 20 isolated compounds from E. sativa was initially evaluated using DPPH radical scavenging assay, and TEAC values were calculated accordingly. Compound 12 (rutin), a well-known flavonoid, exhibited the highest antioxidant activity, followed by compounds 2, 5 and 3 (Table ). Among the newly identified compounds, compound 2 showed the most potent antioxidant effect. These findings align with previous literature demonstrating flavonoids, such as rutin and kaempferol derivatives, are potent antioxidants capable of scavenging free radicals. This in vitro free radical scavenging activity indicated the strong antioxidative potential of compound 2, likely due to the presence of multiple hydroxyl groups and acyl substitutions that enhance radical stabilization. The strong antioxidant activity of compound 2 suggests its potential application in preventing oxidative stress-related diseases by mitigating oxidative stress in both systemic circulation and localized tissue environments.

4. Antioxidant Activity of Compounds 1–20 .

No.	DPPH IC50 (mg/mL)	TEAC (μmol TE/g)
1	5.56	11.9
2	1.98	33.4
3	4.14	16.0
4	28.7	2.31
5	2.71	24.4
6	>100	<0.1
7	>100	<0.1
8	>100	<0.1
9	>100	<0.1
10	>100	<0.1
11	>100	<0.1
12	0.172	385.0
13	>100	<0.1
14	21.9	3.02
15	>100	<0.1
16	>100	<0.1
17	>100	<0.1
18	>100	<0.1
19	103.0	0.644
20	4.59	14.4
Trolox (positive control)	0.066	1000

3.3. Cardioprotective Activity

Hypoxia/reoxygenation (H/R) injury is a widely used in vitro model that mimics ischemia-reperfusion conditions in cardiomyocytes. During hypoxia/reoxygenation, mitochondrial oxidative phosphorylation becomes compromised, leading to electron leakage and the generation of reactive oxygen species (ROS) upon reoxygenation. This surge in ROS not only causes direct oxidative damage to lipids, proteins, and DNA, but also activates pro-apoptotic signaling cascades that culminate in cardiomyocyte death and contribute to myocardial dysfunction. , To investigate whether compound 2 can mitigate oxidative stress under these pathological conditions, H9c2 rat cardiomyocytes were subjected to 24 h of treatment with compound 2 (1 or 2 mM), followed by H/R stimulation. Intracellular ROS levels were measured using CM-H₂DCFDA, a fluorescent probe sensitive to peroxides. As shown in Figure A, H/R significantly elevated ROS levels compared to normoxic controls. However, pretreatment with compound 2 resulted in a dose-dependent reduction in fluorescence intensity, indicating that compound 2 effectively suppressed H/R-induced ROS generation. In parallel, we examined whether the reduction in ROS was accompanied by a decrease in apoptotic cell death. Annexin V-FITC/PI double staining, followed by flow cytometric analysis, revealed that H/R markedly increased apoptosis in H9c2 cells (Figure B). Pretreatment with compound 2 significantly reduced both early and late apoptotic cell populations, again in a dose-dependent manner. The protective effect was more pronounced at the 2 mM, nearly restoring levels to baseline.

3.

Antioxidant and antiapoptotic effects of compound 2 in H9c2 cardiomyocytes under hypoxia/reoxygenation (H/R) conditions. (A) Intracellular ROS levels were assessed using CM-H₂DCFDA staining followed by flow cytometry. H9c2 cells were pretreated with compound 2 (1 or 2 mM) for 24 h and then subjected to H/R (1 h hypoxia/4 h reoxygenation) stress. Representative histograms (top) and quantification of relative fluorescence intensity (RFI, bottom) are shown. (B) Apoptotic cell death was evaluated by Annexin V-FITC and propidium iodide (PI) double staining. H9c2 cells pretreated with compound 2 were exposed to H/R, and apoptotic populations were quantified by flow cytometry. Representative dot plots (top) and the percentage of apoptotic cells (bottom) are shown. Quantitative data are presented as mean ± standard deviation (n = 4); ***p < 0.001.

These results demonstrate that compound 2 exerts potent antioxidant activity in cardiomyocytes under ischemia-reperfusion-like conditions. The reduction in intracellular ROS likely contributes to the observed suppression of apoptosis, supporting the notion that compound 2 confers cytoprotective effects via attenuation of oxidative stress. Given that oxidative injury is a central mechanism in many cardiovascular pathologies, such as myocardial infarction and heart failure, , these findings highlight the therapeutic potential of compound 2 as a naturally derived antioxidant agent.

3.4. Antiplatelet Activity

Platelet aggregation is a tightly regulated process that plays a central role in thrombus formation. , One of the primary activation pathways is mediated through the glycoprotein VI (GPVI) receptor, which is engaged by collagen and collagen-related peptides such as CRP-XL (cross-linked collagen-related peptide). Upon ligand binding, GPVI triggers downstream signaling cascades including activation of NADPH oxidases (NOX1/2), leading to a burst of intracellular reactive oxygen species (ROS). These ROS function as secondary messengers that enhance platelet activation and aggregation by promoting integrin αIIbβ3 activation, granule secretion, and phosphatidylserine exposure.

To examine whether compound 2 exerts antiplatelet effects by modulating ROS-dependent aggregation, human washed platelets were pretreated with compound 2 and stimulated with CRP-XL. As shown in Figure A, all five tested compounds reduced CRP-XL-induced aggregation to varying degrees, with compound 2 exhibiting the most potent inhibitory effect. This suggests a structure–activity relationship favoring the sinapoyl-acylated kaempferol glycoside. To further validate the concentration-dependent effect of compound 2, aggregation was assessed at three concentrations: 0.5, 1, and 2 mM. As shown in Figure B, compound 2 inhibited aggregation in a dose-dependent manner, with maximal inhibition observed at 2 mM. To assess whether this inhibitor effect involves ROS-dependent mechanisms, N-acetylcysteine (NAC, 1 mM)-a well-known ROS scavenger-was included as a positive control. As shown in Figure C, NAC significantly reduced CRP-XL-induced aggregation. Notably, compound 2 exhibited a comparable level of inhibition to NAC at the same concentration (1 mM), further supporting the hypothesis that the antiplatelet activity of compound 2 is at least partially mediated by suppression of ROS-dependent signaling.

4.

Inhibitory effect of compound 2 on CRP-XL-induced platelet aggregation. Human washed platelets (5 × 10⁸/mL) were pretreated with 1 mM compounds 1–5 (A), with compound 2 at 0.5, 1, or 2 mM (B) or 1 mM N-acetylcysteine (NAC) (C) for 5 min, followed by stimulation with CRP-XL (1 μg/mL). Platelet aggregation was monitored over time using light transmission aggregometry. Each panel shows representative aggregation traces (top) and quantification of maximum aggregation (bottom). Data are presented as mean ± standard deviation (n = 3–4); *p < 0.05 and ***p < 0.001.

These findings align with the hypothesis that compound 2 reduces ROS production during platelet activation, thereby attenuating downstream pro-aggregatory signaling. Previous studies have confirmed that GPVI-mediated NOX activation results in intracellular ROS generation, which enhances platelet aggregation. , The ability of compound 2 to inhibit CRP-XL-induced aggregation suggests interference with ROS-dependent signaling, possibly through NOX inhibition or scavenging of superoxide and hydrogen peroxide. Given its potent antioxidant effects in cardiomyocytes (Figure ), it is plausible that compound 2 exerts a similar ROS-suppressive mechanism in platelets.

Collectively, these results underscore the central role of oxidative stress in the pathogenesis of cardiovascular diseases, where excessive ROS generation contributes to both myocardial injury and thrombus formation. In cardiomyocytes, ROS are rapidly produced upon H/R and act as key mediators of mitochondrial dysfunction and apoptotic signaling. In parallel, platelet activation via GPVI pathways induces a ROS burst that amplifies integrin activation and aggregation, further exacerbating vascular occlusion and ischemic damage.

4. Conclusion

In this study, one new cyclic thiocarbamate (1) and four new kaempferol glycosides (2–5), were isolated and structurally characterized from the leaves of rocket salad (E. sativa), alongside 15 known compounds (6–20). Among them, compound 2 exhibited potent cardioprotective and antiplatelet activities. Specifically, it suppressed ROS production in both cardiomyocytes and platelets, attenuated hypoxia/reoxygenation-induced apoptosis, and inhibited CRP-XL-induced platelet aggregation. However, the absence of in vivo validation and uncertainty regarding its metabolic stability, indicate the further pharmacological and pharmacokinetic evaluations are essential before its therapeutic potential can be fully realized. Nonetheless, these dual protective effects highlight compound 2 as a promising dual-functional agent for cardiovascular health promotion. Importantly, our findings support the traditional and emerging role of E. sativa not merely as a culinary herb, but as a valuable source of functional food ingredients or nutraceuticals.

Data are contained within the article or the Supporting Information.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.5c05585.

• UV, HR-ESI-MS and NMR spectra of compounds 1–5; sugar analysis of compounds 1–5 (PDF)

#.

R.Z. and J.W.P. are first authors and contributed equally.

The authors declare no competing financial interest.