1 Introduction

Traditional Chinese medicines with the functions of promoting blood circulation (“Huoxue” in Chinese) and/or removing blood stasis (“Huayu” in Chinese) are claimed to be useful in antiplatelet therapies and the treatment of thrombotic diseases [1]. For example, antiplatelet compounds have been found in a Huoxue herb Selaginella moellendorffii Hieron. (Selaginellaceae) [2, 3].

The genus Piper (Piperaceae) is a medicinally important group of plants consisting of approximately 2000 species worldwide. There are approximately 60 species distributed in the tropical areas of the People’s Republic of China, of which approximately 30 species have been used as traditional Chinese medicines [4]. Some Piper species are used for promoting blood circulation, while Piper mullesua Buch.-Ham. ex D. Don and P. yunnanense Tseng are used for removing blood stasis [5]. As a folk medicine in China with the Chinese name of Duan-Jv (短蒟), the whole plants of P. mullesua are also used to treat bleeding, bone fractures, injuries from falls, rheumatoid arthritis, rheumatic arthralgia, acroanesthesia, asthma, colds, stomach aches, abdominal pain, toothaches, swelling and pain of furuncles, dysmenorrhea, menoxenia, empyrosis, and snake and insect bites [5, 6].

Alcoholic extracts of P. mullesua showed the activity against rabbit platelet aggregation induced by 7.2 nM of the platelet-activating factor (PAF) with an IC50 value of 64.43 μg/mL [7]. Amides including retrofractamide A, chingchengenamide A [6], N-isobutyl-16-phenylhexadeca-2E,4E-dienamide, and N-isobutyldeca-2E,4E-dienamide [8], lignans including (-)-nectandrin A, nectandrin B, galgravin [6], asarinin, fargesin, and sesamin with antifeedant activity [8, 9], a phenylpropanoid myristicin with insecticidal activity [9, 10], and several arylalkenyl carboxylic acid esters [10, 11] have been isolated from the plants. However, the active constituents of P. mullesua responsible for the antiplatelet aggregation remain unclear. In continuing efforts to search for antiplatelet compounds from Piper plants [12, 13], we herein present the results of the analysis of compounds from the aerial parts of P. mullesua and the bioactivity of these compounds.

2 Results and Discussion

2.1 Structure Elucidation

Nine undescribed compounds (19, Fig. 1) and 28 known ones (1037) were isolated from the methanolic extracts of P. mullesua by silica gel, D101 resin and Sephadex LH-20 column chromatography and semipreparative HPLC.

Fig. 1
figure 1

Structures of undescribed compounds (19) from Piper mullesua

Pipermullesine A (1) had the molecular formula C15H15NO4 based on 13C NMR (Table 1) and HREIMS data. Its IR spectrum showed absorption peaks for a tertiary amide at 1643 cm−1 and a phenyl ring at 1595, 1513, and 1461 cm−1. The 1H NMR data (Table 1) indicated a 1,2,4-trisubstituted phenyl ring [δH 7.12 (1H, dd, J = 8.3, 1.8 Hz), 7.01 (1H, d, J = 1.8 Hz), and 6.86 (1H, d, J = 8.3 Hz)], an E double bond [δH 7.67 (1H, d, J = 15.3, 1.8 Hz) and 6.49 (1H, d, J = 15.3 Hz)], a 1,4-oxazine ring [δH 6.61 (1H, dd, J = 5.1, 1.9 Hz), 6.10 (1H, dd J = 5.1, 1.9 Hz), 5.84 (1H, d, J = 5.1 Hz), and 5.70 (1H, d, J = 5.1 Hz)] [14], and two methoxy groups [δH 3.92 (3H, s) and 3.91 (3H, s)]. The above NMR characteristic signals implied that compound 1 might be a cinnamamide derivative.

Table 1 1H (500 MHz) and 13C NMR (125 MHz) NMR Data of 1 in CDCl3

According to the 1H–1H COSY and HMBC correlations of compound 1 (Fig. 2), (E)-3,4-dimethoxycinnamoyl and 1,4-oxazine groups were confirmed. Although the correlations from H-1″ and H-4″ to C-1 were not observed in the HMBC spectrum, the structure of 1 was finally determined as (E)-3-(3,4-dimethoxyphenyl)-1-(4H-1,4-oxazin-4-yl) prop-2-en-1-one by a single-crystal X-ray diffraction analysis (Fig. 3).

Fig. 2
figure 2

Key 2D NMR correlations of compounds 19

Fig. 3
figure 3

X-ray crystallographic structures of 1, 2a, and 3a

The molecular formula of pipermullesine B (2), C16H22N2O3, was determined by 13C NMR data (Table 2) and an HREIMS ion at m/z 290.1620 [M]+ (calcd for C16H22N2O3, 290.1630) and required 7 indices of hydrogen deficiency. The 1H NMR data (Table 2) indicated a tetrasubstituted phenyl ring [δH 7.15 (1H, s) and 6.52 (1H, s)], one methoxy group [δH 3.85 (3H, s)], and one acetyl group [δH 1.91 (3H, s)]. The 13C NMR data (Table 2) exhibited 15 signals. However, according to its HREIMS data, compound 2 should have 16 carbon atoms. The disappeared signal for C-9 (δC 33.0) was detected by the HMBC correlation (Fig. 2) from H2-11 to C-9.

Table 2 1H and 13C NMR Data of 2 and 2a in CD3OD

1H–1H COSY correlations (Fig. 2) exhibited two partial structures comprising C-2 to C-3 and C-10 to C-12. On the basis of the HMBC correlations from H2-3 to C-1 and C-4a, H2-4 to C-5 and C-8a, H-5 to C-7 and C-8a, H-8 to C-1, C-4a, and C-6, and 7-OMe to C-7, a 6-hydroxy-7-methoxy-3,4-dihydroisoquinoline fragment with a substituent group at C-1 was confirmed. The group at C-1 was deduced as 4-acetamidobutyl by the HMBC correlations from H2-10 to C-1, H2-11 to C-9 and C-1′, and H3-2′ to C-1′. Thus, the structure of 2 was determined as 1-(4-acetamidobutyl)-6-hydroxy-7-methoxy-3,4-dihydroisoquinoline and given the common name pipermullesine B.

The crystals for pipermullesine B trifluoroacetate (2a) were obtained from methanol. The NMR data of 2a (Table 2) and the result of its single-crystal X-ray diffraction analysis (Fig. 3) further supported the structure elucidation of 2.

Pipermullesine C (3) yielded a molecular formula of C22H30N4O4 with 10 degrees of unsaturation, as deduced by 13C NMR (Table 3) and the HREIMS data. A comparison of the NMR data (Tables 2, 3) of 3 with those of 2 indicated that there were signals for one additional 4-acetamidobutyl group [δC 173.2 (C), 40.1 (CH2), 29.7 (CH2), 28.9 (CH2), 25.2 (CH2), and 22.6 (CH3)] and one more imine (δC 168.8) in 3.

Table 3 1H and 13C NMR Data of 3 and 3a in CD3OD

On the basis of 2D NMR correlations (Fig. 2), a 1-(4-acetamidobutyl)-6-hydroxy-3,4-dihydroisoquinoline moiety was determined. One more ring is needed to meet the unsaturation, and the ring was deduced as an oxazole ring attached to C-7 and C-8 by comparison of the NMR data with those of benzoxazoles in the literature [15, 16]. The additional 4-acetamidobutyl group was located at C-1′′ of the oxazole ring by the HMBC correlation from H-3′′ to C-1′′. Thus, the structure of 3 (pipermullesine C) was determined.

Fortunately, the crystals for pipermullesine C trifluoroacetate (3a) were also obtained from methanol. The NMR data of 3a (Table 3) and the result of its single-crystal X-ray diffraction analysis (Fig. 3) confirmed the chemical structure of 3.

The molecular formula of pipermullamide A (4), C18H28N2O3, was determined by 13C NMR data (Table 4) and an HREIMS ion at 320.2102 [M]+ (calcd for C18H28N2O3, 320.2100), indicating 6 degrees of unsaturation. The 1H NMR data (Table 2) indicated one monosubstituted phenyl ring [δH 7.31 (2H, d, J = 7.4 Hz), 7.25 (2H, dd, J = 7.4, 7.4 Hz), and 7.17 (1H, dd, J = 7.4, 7.4 Hz)], three N-methyl groups [δH 2.85 (9H, s)], and two methyl groups [δH 0.98 (3H, d, J = 5.9 Hz) and 0.95 (3H, d, J = 6.3 Hz)]. By comparing the NMR data of 4 with those of phenylalanine and leucine trimethylbetaine [17, 18], compound 4 might comprise the two fragments, which was confirmed through its 1H–1H COSY and HMBC correlations (Fig. 2). The amino of phenylalanine was acylated by the carboxyl group of leucine trimethylbetaine according to the HMBC correlation from H-8′ to C-1.

Table 4 1H and 13C NMR Data of 46 in CD3OD

Natural amino acids generally have an l configuration. Compound 10 from the plant is also a derivative of l-phenylalanine. Accordingly, compound 4 (pipermullamide A) was elucidated as l-(N,N,N-trimethyl)leucyl-l-phenylalanine.

The molecular formulae of pipermullamides B to F (5–9) were determined as C18H28N2O3, C17H26N2O3, C20H29N3O3, C20H29N3O3, and C19H27N3O3, respectively, by 13C NMR data (Tables 4, 5) and HRMS analysis. According to 1H–1H COSY and HMBC correlations (Fig. 2), compounds 59 were determined as l-(N,N,N-trimethyl)isoleucyl-l-phenylalanine (pipermullamide B, 5), l-(N,N,N-trimethyl)valyl-l-phenylalanine (pipermullamide C, 6), l-(N,N,N-trimethyl)leucyl-l-tryptophan (pipermullamide D, 7), l-(N,N,N-trimethyl)isoleucyl-l-tryptophan (pipermullamide E, 8), and l-(N,N,N-trimethyl)valyl-l-tryptophan (pipermullamide F, 9), respectively.

Table 5 1H and 13C NMR Data of 79 in CD3OD

The known compounds (+)-phenylalanine betaine (10) [19], (-)-mangochinine (11) [20], xylopinidine (12) [21], (-)-oblongine (13) [22], pellitorine (14) [23], (2E,4E)-N-isobutyl-2,4-dodecadienamide (15) [24], retrofractamide A (16) [23], guineensine (17) [23], brachystamide B (18) [25], retrofractamide C (19) [26], sarmentine (20) [27], 3-(3,4-dimethoxyphenyl)propanoylpyrrole (21) [28], N-trans-feruloyltyramine (22) [29], (-)-machilusin (23) [30], galgravin (24) [31], (-)-nectandrin A (25) [32], methyl 3-(3,4-dimethoxyphenyl)propanoate (26) [33], piperic acid (27) [34], methyl piperate (28) [34], methyl (2E,4E)-7-(1,3-benzodioxol-5-yl)hepta-2,4-dienoate (29) [35], (-)-blumenol B (30) [36], (-)-T-muurolol (31) [37], trans-phytol (32) [38], α-tocopherolquinone (33) [39], γ-tocopherol (34) [40], stigmast-4-ene-3,6-dione (35) [41], (22E)-stigmasta-4,22-diene-3,6-dione (36) [42], and (22E)-stigmasta-4,6,8(14),22-tetraen-3-one (37) [43] were determined by comparing the NMR data of 1037 and the optical rotation values of 1013, 2325, 30, and 31 with those reported in the literature.

2.2 In Vitro Platelet Aggregation Assay

All isolates (137) were evaluated for inhibitory activity against platelet aggregation induced by thrombin (IIa) or PAF. As shown in Tables 6 and 7, compounds 2, 3, 5, 14, 27, 33, and 34 possessed weak inhibitory effects on the aggregation of rabbit platelets induced by thrombin (IIa) (1 U/mL) with inhibition rates from 11.5 to 22.2% at a concentration of 300 µg/mL. Compounds 11, 14, 15, 20, and 25 showed weak inhibitory activity against the rabbit platelet aggregation induced by PAF (0.4 µg/mL) with inhibition rates from 16.8% to 36.4% at a concentration of 300 µg/mL, while (-)-mangochinine (11), pellitorine (14), and (2E,4E)-N-isobutyl-2,4-dodecadienamide (15) have IC50 values of 470.3 µg/mL, 614.9 µg/mL, and 579.7 µg/mL, respectively. The antiplatelet activity of (-)-mangochinine and (2E,4E)-N-isobutyl-2,4-dodecadienamide was reported for the first time. The other tested compounds were inactive.

Table 6 The inhibitory effect of all compounds on aggregation of rabbit platelet induced by Thrombin (IIa) (1 U/mL)a
Table 7 The inhibitory effect of all compounds on rabbit platelet aggregation induced by PAF (0.4 µg/mL)a

More than fifty antiplatelet compounds, mainly including alkaloids and amides, lignans and neolignans, and phenylpropanoids, have been isolated from the Piper genus [44]. Pellitorine is a very interesting compound with various biological activities, such as an inhibitory effect on platelet aggregation induced by arachidonic acid (IC50 = 53.0 μg/mL) [45], antituberculosis activity (MIC = 25 µg/mL) [46], antifungal activity against Cryptococcus neoformans (IC50 = 7.7 μg/mL) [47], and α-glucosidase-I enzyme inhibitory activity (IC50 = 34.39 μg/mL) [48]. Although its in vitro activity against platelet aggregation is weak, the amide shows strong in vivo anticoagulant activities at a dose of 4.5 μg/mouse or 9.0 μg/mouse [49]. It is worthwhile to conduct further in vivo antithrombotic studies of pellitorine along with (-)-mangochinine and (2E,4E)-N-isobutyl-2,4-dodecadienamide.

3 Experimental Section

3.1 General Experimental Procedures

The instruments and materials for isolation and identification of compounds from the herb were presented in Supplementary Material.

3.2 Plant Material

The aerial parts of Piper mullesua Buch.-Ham. ex D. Don (Piperaceae) were collected from Mengyuan Village (E101°22′01″, N21°45′22″), Guanlei Town, Mengla County, Xishuangbanna of Yunnan Province, People’s Republic of China, in July 2014, and identified by one of the authors (C.-L.L.). A voucher specimen (No. 201401) was deposited at the Key Laboratory of Economic Plants and Biotechnology, Kunming Institute of Botany, Chinese Academy of Sciences.

3.3 Extraction and Isolation

The air-dried, powdered P. mullesua plant (1.8 kg) was exhaustively extracted with MeOH (4 × 10 L) at room temperature. The MeOH extracts (92.5 g) were suspended in H2O and further partitioned with petroleum ether and CHCl3. The petroleum ether-soluble part (31.2 g) and CHCl3-soluble part (4.8 g) were combined (36.0 g, part B) according to the testing results of thin-layer chromatography. The water phase was partitioned by D101 resin column chromatography to obtain the water-eluted part (discarded) and 95% EtOH-eluted part (7.2 g, part A).

Part A was subjected to column chromatography (silica gel; CHCl3/MeOH, 10:1 → 0:1, v/v) to yield three fractions (A1–A3). Fraction A1 was separated on an RP-18 silica gel column eluted with MeOH/H2O (10% → 100%). The 40% MeOH-eluted portion was purified by Sephadex LH-20 column chromatography (MeOH) and semi-preparative HPLC (Aligent Zorbax SB-C18 column, 10 × 250 mm, 2 mL/min) to obtain 3 (6.5 mg, MeOH/H2O, 80:20, tR = 12.150 min) and 11 (6.9 mg, MeCN/H2O, 15:85, tR = 5.190 min).

Fraction A2 was separated on an RP-18 silica gel column eluted with MeOH/H2O (10% → 100%) to yield two main subfractions. The 5% MeOH-eluted portion was purified by Sephadex LH-20 column chromatography (MeOH) to yield two main subfractions (A2-1-1 and A2-1-2). Subfraction A2-1-1 was performed on preparative TLC (CHCl3/MeOH, 3:1) to obtain 2 (4.3 mg). Subfraction A2-1-2 was recrystallized from MeOH to obtain 10 (23.5 mg). The 30% MeOH-eluted portion was purified by Sephadex LH-20 column chromatography (MeOH) and semipreparative HPLC [Welch Ultimate AQ-C18 column, 5.0 μm, ϕ 4.6 × 300 mm, MeCN/H2O (containing 0.05% TFA), 20:80, 1 mL/min] to obtain 13 (2.0 mg, tR = 5.796 min).

Fraction A3 was separated on an RP-18 silica gel column eluted with MeOH/H2O (10% → 100%) to yield two main subfractions. The 15% MeOH-eluted portion was purified by Sephadex LH-20 column chromatography (MeOH) and semipreparative HPLC (Aligent Zorbax SB-C18 column, 10 × 250 mm, MeOH/H2O, 20:80, 2 mL/min) to obtain 7 (4.5 mg, tR = 14.369 min), 6 (8.0 mg, tR = 15.398 min), 9 (24.0 mg, tR = 17.255 min), 8 (38.4 mg, tR = 22.038 min), 5 (27.3 mg, tR = 23.055 min), and 4 (8.5 mg, tR = 38.045 min). The 35% MeOH-eluted portion was purified by Sephadex LH-20 column chromatography (MeOH) and semipreparative HPLC [Welch Ultimate AQ-C18 column, 5.0 μm, ϕ 4.6 × 300 mm, MeCN/H2O (containing 0.05% TFA), 20:80, 1 mL/min] to obtain 12 (3.6 mg, tR = 9.300 min).

Part B was subjected to column chromatography (silica gel; petroleum ether/acetone, 20:1 → 0:1, v/v) to yield five fractions (B1–B5). Fraction B1 was separated on an RP-18 silica gel column eluted with MeOH/H2O (10% → 100%) to yield four main subfractions. The 70% MeOH-eluted portion was purified by Sephadex LH-20 column chromatography (MeOH) and silica gel column chromatography (petroleum ether/acetone, 40:1, v/v) to obtain 28 (3.3 mg) and 29 (4.9 mg). The 85% MeOH-eluted portion was purified by Sephadex LH-20 column chromatography (MeOH) and preparative TLC (petroleum ether/EtOAc, 20:1, v/v) to obtain 32 (13.5 mg). The 90% MeOH-eluted portion was purified by Sephadex LH-20 column chromatography (MeOH) to obtain 31 (2.3 mg) recrystallized from MeOH. The 95% MeOH-eluted portion was purified by Sephadex LH-20 column chromatography (MeOH) and semipreparative HPLC (Aligent Zorbax SB-C18 column, 10 × 250 mm, MeOH/H2O, 100:0, 2 mL/min) to obtain 34 (6.0 mg, tR = 19.574 min) and 37 (1.5 mg, tR = 26.868 min).

Fraction B2 was separated on an RP-18 silica gel column eluted with MeOH/H2O (10% → 100%) to yield two main subfractions. The 80% MeOH-eluted portion was purified by Sephadex LH-20 column chromatography (MeOH) and semipreparative HPLC (Aligent Zorbax SB-C18 column, 10 × 250 mm, 2 mL/min) to obtain 14 (86.7 mg, MeOH/H2O, 80:20, tR = 16.778 min), 26 (6.4 mg, MeOH/H2O, 75:25, tR = 8.558 min), 21 (16.0 mg, MeOH/H2O, 80:20, tR = 9.164 min), and 23 (18.9 mg, MeOH/H2O, 80:20, tR = 19.593 min). The 90% MeOH-eluted portion was purified by Sephadex LH-20 column chromatography (MeOH) and semipreparative HPLC (Aligent Zorbax SB-C18 column, 10 × 250 mm, MeCN/H2O, 99:1, 2 mL/min) to obtain 36 (3.4 mg, tR = 37.378 min), 35 (4.0 mg, tR = 42.351 min), and 33 (5.0 mg, tR = 45.179 min).

Fraction B3 was separated on an RP-18 silica gel column eluted with MeOH/H2O (10% → 100%) to yield two main subfractions. The 60% MeOH-eluted portion was purified by Sephadex LH-20 column chromatography (MeOH) to obtain 17 (32.6 mg) and 24 (105.7 mg) recrystallized from MeOH. The 70% MeOH-eluted portion was purified by Sephadex LH-20 column chromatography (MeOH) and semipreparative HPLC (Aligent Zorbax SB-C18 column, 10 × 250 mm, MeOH/H2O, 85:15, 2 mL/min) to obtain 15 (10.3 mg, tR = 19.983 min).

Fraction B4 was separated on an RP-18 silica gel column eluted with MeOH/H2O (10% → 100%) to yield four main subfractions. The 60% MeOH-eluted portion was purified by Sephadex LH-20 column chromatography (MeOH) to obtain 1 (169.9 mg) recrystallized from MeOH. The 70% MeOH-eluted portion was isolated by column chromatography (Sephadex LH-20, MeOH; silica gel, petroleum ether/acetone, 30:1, v/v) and further purified by semipreparative HPLC (Agilent Zorbax SB-C18 column, 9.4 × 250 mm, 2 mL/min) to yield 20 (8.5 mg, MeOH/H2O, 80:20, tR = 19.115 min), 25 (3.0 mg, MeOH, 75:25, tR = 15.549 min), and 18 (16.0 mg; MeOH/H2O, 80:20, tR = 15.356 min). The 80% MeOH-eluted portion was purified by Sephadex LH-20 column chromatography (MeOH) obtain 19 (2.0 mg) recrystallized from MeOH. The 90% MeOH/H2O-eluted portion was purified by Sephadex LH-20 column chromatography (MeOH) and semipreparative HPLC (Aligent Zorbax SB-C18 column, 10 × 250 mm, MeOH/H2O, 95:5, 2 mL/min) to obtain 16 (14.7 mg, tR = 14.172 min).

Fraction B5 was separated on an RP-18 silica gel column eluted with MeOH/H2O (10% → 100%) to yield four main subfractions. The 10% MeOH-eluted portion was purified by column chromatography (Sephadex LH-20, MeOH; preparative TLC, CHCl3/MeOH, 5:1, v/v) and semipreparative HPLC (Aligent Zorbax SB-C18 column, 10 × 250 mm, MeOH/H2O, 40:60, 2 mL/min) to obtain 30 (18.2 mg, tR = 18.052 min). The 20% MeOH-eluted portion was purified by Sephadex LH-20 column chromatography (MeOH) to obtain 27 (52.3 mg) recrystallized from MeOH. The 30% MeOH-eluted portion was purified by Sephadex LH-20 column chromatography (MeOH) and semipreparative HPLC (Aligent Zorbax SB-C18 column, 10 × 250 mm, MeOH/H2O, 50:50, 2 mL/min) to obtain 22 (18.0 mg, tR = 6.746 min).

3.4 Spectroscopic Data of Compounds

3.4.1 Pipermullesine A (1)

Pale yellow needles (CHCl3); mp 90–93 °C; UV (MeOH) λmax (logε) 334 (4.22), 243 (4.11), 224 (3.94) nm; IR (KBr) νmax 1643, 1595, 1513, 1461, 1439, 1415, 1376, 1345, 1315, 1287, 1269, 1228, 1160, 1141, 1047, 1023, 907, 803 cm−1; 1H NMR and 13C NMR data, see Table 1; ESIMS m/z 296 [M + Na]+, 569 [2 M + Na]+; HREIMS m/z 273.0997 [M]+ (calcd for C15H15NO4, 273.1001).

Crystal data for pipermullesine A (1): C15H15NO4·H2O, M = 291.30, monoclinic, a = 4.9028(8) Å, b = 21.159(3) Å, c = 13.455(2) Å, α = 90.00°, β = 92.120(2)°, γ = 90.00°, V = 1394.9(4) Å3, T = 100(2) K, space group P21/n, Z = 4, μ(MoKα) = 0.105 mm−1, 14679 reflections measured, and 3883 independent reflections (Rint= 0.0331). The final R1 value was 0.0434 (I > 2σ(I)). The final wR(F2) value was 0.1212 (I > 2σ(I)). The final R1 value was 0.0575 (all data). The final wR(F2) value was 0.1322 (all data). The goodness of fit on F2 was 1.030. The crystallographic data for the structure of 1 have been deposited in the Cambridge Crystallographic Data Centre (deposition number CCDC 1529565). Copies of the data can be obtained free of charge from the CCDC via www.ccdc.cam.ac.uk.

3.4.2 Pipermullesine B (2)

Pale yellow powder; UV (MeOH) λmax (logε) 402 (3.62), 310 (3.07), 268 (3.43) nm; IR (KBr) νmax 3424, 1622, 1511, 1466, 1441, 1368, 1354, 1235, 1208, 1180, 1039 cm−1; 1H and 13C NMR data, see Table 2; ESIMS m/z 291 [M + H]+, 313 [M + Na]+; HREIMS m/z 290.1620 [M]+ (calcd for C16H22N2O3, 290.1630).

3.4.3 Pipermullesine C (3)

Pale yellow powder; UV (MeOH) λmax (logε) 400 (3.62), 310 (3.10), 268 (3.42) nm; IR (KBr) νmax: 3376, 1721, 1630, 1607, 1590, 1562, 1511, 1462, 1438, 1384, 1337, 1271, 1237, 1211, 1101, 1076, 1037 cm−1; 1H and 13C NMR data, see Table 3; ESIMS m/z 415 [M + H]+, 437 [M + Na]+; HREIMS m/z 414.2223 [M]+ (calcd for C22H30N4O4, 414.2223).

3.4.4 Pipermullamide A (4)

White solid; \( [\alpha ]_{\text{D}}^{25} \)− 16.2 (c 0.08, MeOH); UV (MeOH) λmax (logε) 204 (3.35) nm; ECD Δε (c 0.08, MeOH) +1.35 (217); IR (KBr) νmax 3429, 1713, 1626, 1460, 1415, 1384, 1299, 1274, 1126, 1079, 1046 cm−1; 1H and 13C NMR data, see Table 4; ESIMS m/z 321 [M + H]+, 343 [M + Na]+; HREIMS m/z 320.2102 [M]+ (calcd for C18H28N2O3, 320.2100).

3.4.5 Pipermullamide B (5)

White solid; \( [\alpha ]_{\text{D}}^{19} \)− 7.6 (c 0.18, MeOH); UV (MeOH) λmax (logε) 203 (4.00) nm; ECD Δε (c 0.012, MeOH) +1.48 (214); IR (KBr) νmax 3442, 1666, 1609, 1494, 1456, 1385, 1312, 1256, 1226, 1091, 1031 cm−1; 1H and 13C NMR data, see Table 4; ESIMS m/z 321 [M + H]+, 343 [M + Na]+; HREIMS m/z 320.2102 [M]+ (calcd for C18H28N2O3, 320.2100).

3.4.6 Pipermullamide C (6)

White solid; \( [\alpha ]_{\text{D}}^{20}{-}33.1 \) (c 0.09, MeOH); UV (MeOH) λmax (logε) 206 (4.07) nm; ECD Δε (c 0.014, MeOH) +2.50 (214); IR (KBr) νmax 3426, 1673, 1608, 1494, 1456, 1383, 1315, 1257, 1225, 1096, 1032 cm−1; 1H and 13C NMR data, see Table 4; ESIMS m/z 329 [M + Na]+; HRESIMS m/z 307.2017 [M + H]+ (calcd for C17H27N2O3, 307.2022).

3.4.7 Pipermullamide D (7)

White solid; \( [\alpha ]_{\text{D}}^{21}{-}13.7 \) (c 0.08, MeOH); UV (MeOH) λmax (logε) 281 (3.47), 222 (4.28), 206 (4.16) nm; ECD Δε (c 0.037, MeOH) +1.40 (233), -1.15 (222), -0.94 (214); IR (KBr) νmax 3426, 1674, 1611, 1488, 1459, 1384, 1259, 1228, 1127, 1101 cm−1; 1H and 13C NMR data, see Table 5; ESIMS m/z 360 [M + H]+, 382 [M + Na]+; HRESIMS m/z 360.2288 [M + H]+ (calcd for C20H30N3O3, 360.2287).

3.4.8 Pipermullamide E (8)

White solid; \( [\alpha ]_{\text{D}}^{21} {-}32.2\) (c 0.06, MeOH); UV (MeOH) λmax (logε) 398 (1.74), 282 (3.09), 220 (3.90), 205 (3.89) nm; ECD Δε (c 0.028, MeOH) +1.21 (233), +1.87 (225), -0.51 (220), -3.34 (200); IR (KBr) νmax 3428, 1681, 1619, 1452, 1422, 1384, 1209, 1139, 1046 cm−1; 1H and 13C NMR data, see Table 5; ESIMS m/z 360 [M + H]+, 382 [M + Na]+; HRESIMS m/z 360.2293 [M + H]+ (calcd for C20H30N3O3, 360.2287).

3.4.9 Pipermullamide F (9)

White solid; \( [\alpha ]_{\text{D}}^{20}{-}12.2 \) (c 0.21, MeOH); UV (MeOH) λmax (logε) 281 (3.53), 221 (4.31), 206 (4.15) nm; ECD Δε (c 0.013, MeOH) +1.30 (228), -1.80 (212), -4.48 (200); IR (KBr) νmax 3415, 3266, 1672, 1603, 1491, 1459, 1384, 1255, 1229, 1098, 961, 747 cm−1; 1H and 13C NMR data, see Table 5; ESIMS m/z 346 [M + H]+, 368 [M + Na]+; HRESIMS m/z 346.2128 [M + H]+ (calcd for C19H28N3O3, 346.2131).

3.5 Preparation of Pipermullesine B Trifluoroacetate (2a)

Compound 2 (1.6 mg, 0.00551 mmol) was performed on semipreparative HPLC [Welch Ultimate AQ-C18 column, 5.0 μm, ϕ 4.6 × 300 mm, MeCN/H2O (containing 0.05% TFA), 20:80, 1.0 mL/min] to obtain 2a (2.0 mg, tR = 6.344 min; 0.00516 mmol, 94% yield): pale yellow needles (MeOH); mp 142–145 °C; 1H and 13C NMR data, see Table 2.

Crystal data for pipermullesine B trifluoroacetate (2a): C16H23N2O3·C2F3O2, M = 404.38, a = 7.6163(3) Å, b = 8.8383(3) Å, c = 14.8434(5) Å, α = 82.5370(10)°, β = 89.4170(10)°, γ = 74.5720(10)°, V = 954.71(6) Å3, T = 100(2) K, space group P-1, Z = 2, μ(CuKα) = 1.046 mm−1, 13115 reflections measured, 3381 independent reflections (Rint= 0.0623). The final R1 value was 0.1027 (I > 2σ(I)). The final wR(F2) value was 0.2945 (I > 2σ(I)). The final R1 value was 0.1048 (all data). The final wR(F2) value was 0.2990 (all data). The goodness of fit on F2 was 1.412. The crystallographic data for the structure of 2a have been deposited in the Cambridge Crystallographic Data Centre (deposition number CCDC 1529558). Copies of the data can be obtained free of charge from the CCDC via www.ccdc.cam.ac.uk.

3.6 Preparation of Pipermullesine C Trifluoroacetate (3a)

Compound 3 (4.0 mg, 0.00965 mmol) was performed on semipreparative HPLC [Welch Ultimate AQ-C18 column, 5.0 μm, ϕ 4.6 × 300 mm, MeCN/H2O (containing 0.05% TFA), 20:80, 1.0 mL/min] to obtain 3a (4.5 mg, tR = 7.460 min; 0.00880 mmol, 91% yield): pale yellow needles (MeOH); mp 157–159 °C; 1H and 13C NMR data, see Table 3.

Crystal data for pipermullesine C trifluoroacetate (3a): C22H31N4O4·C2F3O2, M = 528.53, a = 8.5097(10) Å, b = 12.3631(15) Å, c = 12.8583(15) Å, α = 68.354(2)°, β = 78.084(2)°, γ = 80.814(2)°, V = 1225.1(3) Å3, T = 100(2) K, space group P-1, Z = 2, μ(MoKα) = 0.118 mm−1, 13522 reflections measured, 6727 independent reflections (Rint= 0.0383). The final R1 values were 0.0528 (I > 2σ(I)). The final wR(F2) value was 0.1229 (I > 2σ(I)). The final R1 value was 0.0879 (all data). The final wR(F2) value was 0.1414 (all data). The goodness of fit on F2 was 1.020. The crystallographic data for the structure of 3a have been deposited in the Cambridge Crystallographic Data Centre (deposition number CCDC 1589949). Copies of the data can be obtained free of charge from the CCDC via www.ccdc.cam.ac.uk.

3.7 In vitro Platelet Aggregation Assay

The inhibitory effects of compounds against rabbit platelet aggregation induced by PAF or Thrombin (IIa) were evaluated according to the published methods [50,51,52,53]. The details were presented in Supplementary Material.

4 Conclusion

Thirty-seven compounds were isolated from the folk Chinese medicine Piper mullesua with the “Huayu” function associated with the antiplatelet therapies. The antiplatelet compounds, especially (-)-mangochinine, pellitorine, and (2E,4E)-N-isobutyl-2,4-dodecadienamide, might be scientific evidence to support the traditional use of the plant as folk medicine. In order to make better use of the folk medicine to serve for human health, further research needs to be conducted on bioguided isolation of compounds from the plant, based on both in vitro and in vivo bioassay testing.