Three neo-Clerodane Diterpenoids from Tinospora cordifolia Stems and Their Arginase Inhibitory Activities

Abstract

Three neo-clerodane diterpenoids, including two new tinocordifoliols A (1) and B (2) and one known tinopanoid R (3), were isolated from the ethyl acetate-soluble fraction of the 70% ethanol extract of Tinospora cordifolia stems. The structures were elucidated by various spectroscopic methods, including one dimensional (1D) and 2D-NMR, high resolution-electrospray ionization (HR-ESI)-MS, and electronic circular dichroism (ECD) data. The T. cordifolia extract and all isolated compounds 1–3 possessed arginase I inhibitory activities. Among them, 3 exhibited moderate competitive inhibition of human arginase I (IC₅₀ = 61.9 µM). Furthermore, docking studies revealed that the presence of a β-substituted furan in 3 may play a key role in the arginase I inhibitory activities.

Introduction

Arginases I and II are Mn(II)-dependent hydroxylases that convert L-arginine into L-ornithine and urea in the urea cycle. Arginase I is mainly localized in the liver cytosol, and plays a key role in maintaining the nitrogen balance and modulating nitric oxide (NO) homeostasis.¹⁾ In contrast, arginase II is an intramitochondrial enzyme expressed in tissues such as the kidney and prostate, although the physiological reasons for the different localizations of arginases I and II remain unknown.^2,3) Overexpression of arginase I is associated with ailments such as liver disease, cardiovascular disease, and cancers, which are among the leading causes of death.^4,5) Thus, the inhibition of human arginases has been proposed as a potential therapy for various illnesses, such as cardiovascular, anti-inflammatory, autoimmune, oncological, and infectious diseases, characterized by abnormally high arginase activity or abnormally low NO synthase activity.^6,7) Due to the relevance of arginases in these diseases, inhibitors such as L-nor-N-hydroxyarginine (nor-NOHA), 2(S)-amino-6-boronohexanoic acid (ABH), and CB-1158 (numidargistat) have been developed. However, their side effects have constrained their clinical usage.^6–9) Several plant-derived compounds, such as piceatannol, obacunone, and salvianolic acid, have also been discovered as arginase inhibitors.¹⁰⁾ In our previous study, we reported that alismoxide, a guaiane sesquiterpenoid, is a selective arginase I inhibitor in a competitive manner with respect to L-arginine.¹¹⁾ However, these natural inhibitors have not been successfully forwarded to clinical settings, due to their poor pharmacokinetics profiles. Thus, further searches for potent arginase inhibitors are still necessary.

Tinospora cordifolia is a herbaceous vine of the Menispermaceae family.¹²⁾ In Vietnam, it is locally known as ‘Day than thong’ and is widely used in traditional medicine to treat diseases such as malaria, pharyngitis, rheumatism, and diabetes.^13,14) Previous phytochemical investigations revealed that this plant contains clerodane diterpenoids, alkaloids, and lignans with anti-inflammatory, anti-cancer, and anti-tumor properties.^15–20) In our ongoing search for arginase inhibitors from natural resources, the ethyl acetate-soluble portion of the 70% ethanol extract from T. cordifolia stems showed potent inhibitory activity against human arginase I in a colorimetric assay. Further phytochemical investigation of the 70% EtOH extract of T. cordifolia led to the isolation of three neo-clerodane diterpenoids, including two new tinocordifoliols A (1) and B (2) and one known tinopanoid R (3). Herein, we report the isolations and characterizations of the chemical constituents of T. cordifolia, as well as the structural elucidations of the isolated compounds 1–3 and their inhibitory activities against human arginase I.

Results and Discussion

Phytochemical Study

The 70% EtOH extract of T. cordifolia stems was partitioned into n-hexane-, EtOAc-, and Me₂CO-soluble fractions. Chromatographic purification of the EtOAc-soluble fraction led to the isolation of two new compounds, named tinocordifoliol A (1) and tinocordifoliol B (2), together with one known neo-clerodane diterpenoid (Fig. 1). The structure of the known compound was identified as tinopanoid R (3) by comparisons of its NMR spectroscopic data with those reported in the literature.²¹⁾

Fig. 1. Structures of 1–3 Isolated from T. cordifolia Stems

Compound 1 was obtained as a colorless oil with positive optical rotation [α]_D²⁵ +95.1 (c 0.1, MeOH). The molecular formula was determined to be C₂₁H₂₂O₆ by high resolution-electrospray ionization-time-of-flight (HR-ESI-TOF)-MS (m/z 393.1307 [M + Na]⁺, calcd. for C₂₁H₂₂O₆Na, 393.1314) in conjunction with ¹³C-NMR data. The IR spectrum of 1 displayed absorption bands at 3346 and 1722 cm⁻¹, indicating the presence of the hydroxy group and the ester functional moiety, respectively. The ¹H-NMR spectrum revealed proton signals corresponding to five olefinic protons [δ_H 6.80 (t, J = 4.1 Hz, H-3), 6.65 (br s, H-7), 6.42 (d, J = 1.4 Hz, H-14), 7.43 (t, J = 1.6 Hz, H-15), and 7.47 (d, J = 0.8 Hz, H-16)], one oxygenated methine proton [δ_H 5.09 (dd, J = 11.9, 3.2 Hz, H-12)], one methine proton [δ_H 2.23 (m, H-10)], six methylene protons [δ_H 2.14 (m, H-1α), 1.88 (m, H-1β), 2.45 (m, H-2α), 2.44 (m, H-2β), 2.18 (m, H-11α), and 2.03 (m, H-11β)], one methoxy group [δ_H 3.69 (s, 18-OCH₃)], and two methyl groups [δ_H 1.40 (s, H₃-19) and 1.30 (s, H₃-20)] (Table 1). The ¹³C-NMR and heteronuclear multiple quantum coherence (HMQC) spectra showed signals corresponding to three carbonyl carbons [δ_C 200.3 (C-6), 167.4 (C-17), and 166.7 (C-18)], three quaternary olefinic carbons [δ_C 134.3 (C-4), 149.1 (C-8), and 123.0 (C-13)], seven methine carbons [δ_C 139.0 (C-3), 130.2 (C-7), 49.0 (C-10), 71.3 (C-12), 108.4 (C-14), 143.9 (C-15), and 140.0 (C-16)], two quaternary carbons [δ_C 47.2 (C-5) and 38.1 (C-9)], three methylene carbons [δ_C 17.9 (C-1), 23.1 (C-2), and 45.3 (C-11)], one methoxy carbon [δ_C 52.1 (18-OCH₃)], and two methyl carbons [δ_C 26.1 (C-19) and 26.3 (C-20)] (Table 1). The 1D NMR data of 1 were highly similar to that of tinopanoid R (3), isolated from T. crispa.²¹⁾ The significant difference was the presence of the keto group at C-6 in 1 instead of the hydroxy group in 3, suggesting that 1 was a ketonated analog of 3 with the β-substituted furan ring. The planar structure of 1 was further confirmed by ¹H–¹H correlation spectroscopy (COSY) and heteronuclear multiple bond connectivity (HMBC) experiments (Fig. 2). The ¹H–¹H COSY cross peaks of H₂-1/H₂-2/H-3/H-10 and H₂-11/H-12 and the HMBC correlations from H₂-2 to C-4, H-7 to C-5/C-8/C-9/C-17, H-10 to C-4/C-9, H-12 to C-9, and H₃-20 to C-10/C-11 revealed the presence of a decalin structure fused with a 6-membered lactone ring at C-8 and C-9. The HMBC correlations from H-14 to C-13/C-15 and H-16 to C-13/C-15, as well as from H₂-11 to C-13, implied that 1 had the β-substituted furan ring at C-12, as in the case of 3. Furthermore, the HMBC correlation from H₃-19 to C-4/C-6/C-10 and H₃-20 to C-10/C-9/C-11 confirmed that the two methyl groups were located at C-5 and C-9 in the decalin structure, respectively. The relative configuration of 1 was established by the nuclear Overhauser effect spectroscopy (NOESY) correlations (Fig. 3). The NOESY correlations from H₃-19 to H-10 and H-10 to H-11α/H-12, indicating that these protons were α-oriented. In contrast, the NOESY cross peaks from H₃-20 to H-1β/H-11β suggested that H₃-20, H-1β, and H-11β, and the furan ring adopted the β-orientation. The absolute configuration of 1 was determined by comparing the experimental electronic circular dichroism (ECD) spectrum of 1 with the calculated ECD spectra of (5R,9S,10S,12S)-1a and its enantiomer (1b) (Fig. 4). The ECD spectrum of 1 showed positive Cotton effects at 383 and 305 nm and a negative Cotton effect at 351 nm, due to the presence of the α,β-unsaturated ketone moieties, and matched well with the calculated ECD spectrum obtained from the model of 1a with the 5R,9S,10S,12S configuration. Therefore, 1 was identified as a novel compound and was named tinocordifoliol A.

Table 1. ¹H- (400 MHz) and ¹³C-NMR (100 MHz) Data of 1–3 in CDCl₃ (δ in ppm and J Values in (Hz) in Parentheses)

Position	1		2		3
	δH	δC	δH	δC	δH	δC
1α	2.14, m	17.9	1.98, m	19.3	1.94, m	19.7
1β	1.88, m		1.98, m		1.80, m
2α	2.45, m	23.1	2.28, m	24.5	2.24, m	24.8
2β	2.44, m		2.28, m		2.24, m
3	6.80, t (4.1)	139.0	7.14, t (3.6)	142.6	7.11, t (4.0)	142.1
4		134.3		137.9		137.9
5		47.2		41.9		42.0
6β		200.3	4.50, t (2.0)	71.1	4.48, br s	71.2
7	6.65, br s	130.2	6.84, d (2.0)	143.7	6.71, br s	140.9
8		149.1		135.8		135.9
9		38.1		38.2		37.7
10α	2.23, m	49.0	1.67, m	47.7	1.68, t (5.5)	47.7
11α	2.18, m	45.3	2.70, dd (14.8, 2.8)	40.3	2.43, dd (14.8, 2.8)	42.7
11β	2.03, m		1.69, m		1.93, m
12α	5.09, dd (11.9, 3.2)	71.3	4.90, m	70.3	5.13, dd (11.6, 2.8)	70.2
13		123.0		136.6		124.0
14	6.42, d (1.4)	108.4	7.20, t (1.6)	143.9	6.44, d (1.6)	108.7
15	7.43, t (1.6)	143.9	5.81, t (1.6)	103.3	7.43, t (1.6)	143.8
16	7.47, d (0.8)	140.0		168.8	7.48, d (1.4)	139.9
17		167.4		166.7		168.2
18		166.7		169.9		170.2
19	1.40, s	26.1	1.27, s	24.1	1.28, s	23.2
20	1.30, s	26.3	1.04, s	23.7	1.10, s	25.2
6-OHα			5.34, d (2.3)		5.38, d (1.8)
15-OCH3			3.59, s	57.8
18-OCH3	3.69, s	52.1	3.76, s	52.6	3.79, s	52.6

Fig. 2. Key HMBC (Arrows) and ¹H–¹H COSY (Bold Lines) Correlations of 1 and 2

Fig. 3. Key NOESY (Dotted Lines) Correlations of 1 and 2

Fig. 4. Experimental and Calculated ECD Spectra of 1

Compound 2 was obtained as a white amorphous powder with negative optical rotation [α]_D²⁵ −45.4 (c 0.1, MeOH). Its molecular formula was C₂₂H₂₆O₈, according to the HR-ESI-MS data at m/z 441.1520 [M + Na]⁺, in conjunction with the 1D NMR data. The ¹H-NMR spectrum showed signals of three olefinic protons [δ_H 7.14 (t, J = 3.6 Hz, H-3), 6.84 (d, J = 2.0 Hz, H-7), and 7.20 (t, J = 1.6 Hz, H-14)], three oxygenated methine protons [δ_H 4.50 (d, J = 2.0 Hz, H-6), 4.90 (m, H-12), and 5.81 (t, J = 1.6 Hz, H-15)], one methine proton [δ_H 1.67 (m, H-10)], six methylene protons [δ_H 1.98 (m, H-1α), 1.98 (m, H-1β), 2.28 (m, H-2α), 2.28 (m, H-2β), 2.70 (dd, J = 14.8, 2.8 Hz, H-11α), 1.69 (m, H-11β)], two methoxy groups [δ_H 3.59 (s, 15-OCH₃) and 3.76 (s, 18-OCH₃)], two methyl groups [δ_H 1.27 (s, H₃-19) and 1.04 (s, H₃-20)], and one hydroxy group [δ_H 5.34 (d, J = 2.3 Hz, 6-OH)] (Table 1). The ¹³C-NMR and HMQC spectroscopic data of 2 revealed the presence of 22 carbon signals, including three carbonyls [δ_C 168.8 (C-16), 166.7 (C-17), and 169.9 (C-18)], three quaternary olefinic carbons [δ_C 137.9 (C-4), 135.8 (C-8), and 136.6 (C-13)], seven methine carbons [δ_C 142.6 (C-3), 71.1 (C-6), 143.7 (C-7), 47.7 (C-10), 70.3 (C-12), 143.9 (C-14), and 103.3 (C-15)], two quaternary carbons [δ_C 41.9 (C-5), and 38.2 (C-9)], three methylene carbons [δ_C 19.3 (C-1), 24.5 (C-2), and 40.3 (C-11)], two methoxy carbons [δ_C 57.8 (15-OCH₃) and 52.6 (18-OCH₃)], and two methyl carbons [δ_C 24.1 (C-19) and 23.7 (C-20)] (Table 1). The NMR data of 2 were similar to those of 3. The main difference between 2 and 3 was the signals corresponding to the butanolide ring in 2 instead of those of the furan ring in 3 (Table 1). The ¹H–¹H COSY correlations of H₂-1/H₂-2/ H₂-3/H-10 and H₂-11/H-12, as well as the HMBC correlations from H₃-19 to C-4/C-6/C-10, H₃-20 to C-10/C-11, H-7 to C-5/C-9/C-17, H-10 to C-4/C-9, and H-12 to C-9, confirmed the presence of a decalin structure fused with a 6-membered lactone ring at C-8 and C-9 in 2 (Fig. 2). Furthermore, the key HMBC correlations from H-14 to C-12/C-13/C-16, in conjunction with the ¹H–¹H COSY cross peak of H-14/H-15, revealed that the butanolide ring is attached at C-12. In addition, the relative configuration of 2, excepting the C-15 position in the butanolide ring moiety, was assigned on the basis of a NOESY experiment (Fig. 3). The NOESY correlations from H₃-19 to 6-OH/H-10 and H-10 to H-11α/H-12 indicated that H₃-19, 6-OH, H-10, H-11α, and H-12 were α-oriented, while the NOESY cross peaks of H₃-20 to H-6/H-11β suggested that H₃-20 and H-6 were β-oriented. Furthermore, the relative configuration of the C-15 position and the absolute configuration of 2 were determined by ECD spectroscopic analyses (Fig. 5). The comparisons of the experimental ECD spectra of 2 with calculated ECD models of (5R,6S,9S,10S,12S,15R)-2a and (5S,6R,9R,10R,12R,15R)-2b and their enantiomers (2d and 2c) revealed that 2 adopted the absolute configuration of 5R,6S,9S,10S,12S,15R (Fig. 5). Hence, 2 was assigned as shown in Fig. 1 and named tinocordifoliol B.

Fig. 5. Experimental and Calculated ECD Spectra of 2

Arginase I Inhibitory Activities

The arginase I and II inhibitory activities of the three isolated compounds 1–3 were assessed using a colorimetric assay, as reported previously¹¹⁾ (Table 2). The assay revealed that 3 was a moderate arginase I inhibitor with an IC₅₀ value of 69.9 ± 3.2 µM. However, despite the highly structural resemblance, compound 1 exhibited 3.9-fold lower inhibitory activity than that of 3. Compound 2 dramatically decreased the arginase I inhibitory activity with an IC₅₀ value of 1575 ± 3.1 µM, which is 22.5-fold lower than that of 3. The Dixon plot analysis revealed that 3 and 1 inhibited arginase I in competitive and non-competitive manners with respect to L-arginine, respectively (Fig. 6). The mode of inhibition of 2 for arginase I could not be determined, due to its weak inhibitory activity. The structural differences of 3 with 1 and 2 are only the ketone group and the furan ring in 3, compared with the hydroxy group in 1 and the butanolide ring in 2. It is thus strongly speculated that the binding affinity of 3 with the active site in the arginase I was enhanced by the substitution of the ketone group with the hydroxy group, and further augmented by the substitution of the butanolide ring with the furan ring in 3. Similar arginase II inhibitory profiles were observed, with dramatically decreased inhibitory activities of the isolated compounds, where 3 and 1 inhibited the arginase II activity with IC₅₀ values of 272 ± 2.4 and 1813 ± 3.3 µM, respectively, whereas 2 was inactive even at the concentration of 5000 µM (Table 2). These similar profiles suggested that 3 could inhibit arginases I and II in comparable manners.

Table 2. IC₅₀ Values of the Isolated Compounds against Human Arginases I and II

Sample	IC50 (µM)
	Arginase I	Arginase II
1	268 ± 2.4	1813 ± 3.3
2	1575 ± 3.1	>5000
3	69.9 ± 3.2	272 ± 2.4
Nor-NOHAa)	7.0 ± 0.8	10.2 ± 1.0

a) Positive control.

Fig. 6. Dixon Plot Analyses of the Binding Mode of (A) 1 and (B) 3 with Arginase I

To further investigate the possible inhibitory mode of action of 3, we performed the molecular docking of 3 against human arginase I (Fig. 7). The docking study predicted that the 6-member lactone and furan rings of 3 are accommodated in the substrate/product (L-ariginine/L-ornitine) binding site at the active site cavity of arginase I, as in the case of nor-NOHA and ABH in the crystal structures of arginase I complexed with the inhibitors (PDB IDs: 3KV2, 6Q92),^6,8) respectively, and that of arginase II complexed with (S)-(2-boronoethyl)-L-cysteine (BEC) (1PQ3).²²⁾ Interestingly, the docking studies suggested that 3 interacted not only the dinuclear Mn(II) ions (O-Mn distances: 3.4 Å) at the catalytic center,²³⁾ but also the side chains of Asn130 (3.3 Å) and Ser137 (3.1 Å), with the oxygen atom of the furan ring and the carbonyl group at C-17 (Fig. 7A), in a manner very similar to those of nor-NOHA, ABH, and BEC in the aforementioned crystal structures (Figs. 7B, C). We also found that the hydroxy group at C-6 in 3 formed hydrogen bonds with the side chains of Thr136 (2.9 Å) and Asp183 (3.0 Å) (Fig. 7A), suggesting that the inhibitory activity of 3 would be mainly caused by these interactions at the active site cavity. Presumably, the slightly bulkier substitution of the ketone group at C-6 with the hydroxy group in 1 led to somewhat insufficient interaction with the side chains of Thr136 and Asp183 in arginase I. Thus, 1 likely showed the non-competitive manner with slightly lower arginase I inhibitory activity, compared with that of 3, although the docking study of 1 with arginase I did not predict its possible binding site. On the other hand, the significantly bulkier butanolide ring than the furan moiety might hinder the access of 2 to the active site or allosteric binding site, suggesting that 2 thereby showed the significantly lower inhibitory activity for the arginase I, compared with 1 and 3.

Fig. 7. Interactions of 3 with Human Arginases I and II in Model Structures

(A) Close-up view of the substrate binding site of the model structure of human arginase I complexed with 3 (white: arginase I, dark gray: carbon-backbone of 3). (B, C) Comparisons of the substrate binding site of the model structure of human arginase I complexed with 3 with those of crystal structures of arginase I complexed with (B) nor-NOHA(PDB ID: 3KV2, pale cyan: arginase I, blue: carbon-backbone of nor-NOHA) and (C) ABH (PDB ID: 6Q92, pink: arginase I, purple: carbon-backbone of ABH). (D) A comparison of the substrate binding site of the model structure of human arginase I complexed with 3 and that of crystal structure of human arginase II complexed with BEC (PDB ID: 1PQ3, light green: arginase II, green: carbon-backbone of BEC). Dashed lines represent the distances in Å.

However, we could not identify the critical residues responsible for the lower inhibitory activity of 3 against arginase II compared with arginase I (Fig. 7D). Further crystallographic studies are required to clarify the significantly lower inhibitory activities of 1, 2, and 3 against arginase II, compared with arginases I.

Conclusion

In this study, the phytochemical investigation of T. cordifolia stems led to the isolation of two new neo-clerodane diterpenoids, tinocordifoliols A and B, together with one known neo-clerodane diterpenoid, tinopanoid R. The evaluation of the arginase I inhibitory activities revealed that 3 is a natural arginase I inhibitor in a competitive manner with respect to L-arginine. To the best of our knowledge, this is the first report on the arginase I inhibitory activities of a neo-clerodane diterpenoid. Thus, this study provided new insights into a naturally occurring arginase I inhibitor.

Experimental

Chemicals and Reagents

Chemical reagents were purchased from FUJIFILM Wako Pure Chemical Corporation (Osaka, Japan). Nor-NOHA was purchased from Bachem AG (Bubendorf, Switzerland).

General Experimental Procedures

Specific optical rotations were observed on a P2100 polarimeter (JASCO, Tokyo, Japan). ECD spectra were obtained with a J-805 spectropolarimeter (JASCO). Fourier-transform IR spectroscopy (FTIR) spectra were measured on FT/IR-460 Plus spectrometer (JASCO). UV spectra were recorded on NanoPhotometer NP80 (IMPLEN). NMR spectra were recorded on an ECX400P spectrometer (JEOL, Tokyo, Japan). Chemical shift values calibrated to the residual proton and carbon resonances based on the δ residuals of CDCl₃ at 7.26 for ¹H-NMR and 77.2 for ¹³C-NMR. HRMS data were obtained in the negative mode on a Thermo Scientific LTQ Orbitrap XL ETD spectrometer with an ESI interface. Open column chromatography was performed with normal-phase silica gel (silica gel 60N, spherical, neutral, 40–50 µm) (Kanto Chemical, Tokyo, Japan). TLC was conducted on pre-coated silica gel 60 F₂₅₄ plates (Merck, Billerica, MA, U.S.A.), with spots visualized with a UV lamp (254 and 365 nm), and by spraying with a p-anisaldehyde stain solution and heating at 120 °C for 10 min. A MiniAmp thermal cycler (Thermo Fisher, Waltham, MA, U.S.A.) was used to incubate the arginase assays. The absorbance of the urea produced in the arginase inhibition activity assay was measured using ab SH-1200 microplate reader (Corona Electric, Ibaraki, Japan).

Plant Materials

The T. cordiforlia stems were purchased from a traditional market in Tan Binh District, HoChiMinh City, Vietnam, in 2022 and positively identified by Dr. Hoang Xuan Thao, a lecturer at the Faculty of Biology, University of Education, Hue University (Vietnam). A voucher specimen (TMPW 32111) was deposited in the Museum for Materia Medica, Analytical Research Center for Ethnomedicines, Institute of Natural Medicine, University of Toyama, Japan.

Extraction and Isolation

Dried T. cordiforlia stems (1.8 kg) were macerated with 70% EtOH (1 L × 5) for 3 h by sonication (90 min each). The extract was evaporated in vacuo to obtain the residue (15.8 g), which was successively partitioned into n-hexane (31.0 g), EtOAc (3.5 g), and Me₂CO (28.0 g). The EtOAc extract was then subjected to silica gel column chromatography using an n-hexane: EtOAc (2 : 1, 1 : 1, 1 : 3, 1 : 5, and 1 : 7) solvent system, to give five fractions: F₁ (1.2 g), F₂ (1.1 g), F₃ (0.5 g), F₄ (0.3 g), and F₅ (0.8 g). The F₁ fraction was purified on a normal phase silica gel open column eluted with an isocratic n-hexane: EtOAc solvent system (2 : 1), to afford 1 (11 mg). Fraction F₂ was separated on a normal phase silica gel open column eluted with a gradient n-hexane: EtOAc solvent system (2 : 1, 1 : 1, 0.5 : 1), to obtain 2 (4 mg). Fraction F₃ was further purified on a normal phase silica gel open column eluted by an n-hexane: EtOAc solvent system (1 : 1, 0.5 : 1, 0 : 1) with increasing polarity, to afford 3 (0.8 mg).

Tinocordifoliol A

White amorphous powder; Rf = 0.51 (n-hexane : EtOAc = 50 : 50); [α]_D²⁵ +95.1° (c 0.1, MeOH); CD (c 0.01, MeOH): 383 (Δε + 3.20), 351 (Δε − 0.65), 305 (Δε + 0.79); UV (MeOH) λ_max (log ε): 213 (1.52) nm; IR (KBr) ν_max: 3346, 2896, 1722, 1634, 1435, 1260, 1060, 802 cm⁻¹; ¹H-NMR (400 MHz, CDCl₃) and ¹³C-NMR (100 MHz, CDCl₃) data, see Table 1; HR-ESI-MS: m/z 393.1307 [M + Na]⁺ (calc. for C₂₁H₂₂NaO₆, 393.1314).

Tinocordifoliol B

Colorless oil; Rf = 0.68 (n-hexane : EtOAc = 50 : 50); [α]_D²⁵ −45.4° (c 0.1, MeOH); CD (c 0.01, MeOH): 266 (Δε − 4.63), 252 (Δε + 16.9); UV (MeOH) λ_max (log ε): 209 (0.75) nm; IR (KBr) ν_max: 3374, 2988, 1733 cm⁻¹; ¹H-NMR (400 MHz, CDCl₃) and ¹³C-NMR (100 MHz, CDCl₃) data, see Table 1; HR-ESI-MS: m/z 441.1520 [M + Na]⁺ (calc. for C₂₂H₂₆NaO₈, 441.1525).

Arginase Inhibitory Assay

Arginase inhibitory activities were assessed through a colorimetric assay designed to quantify urea production, as reported previously.¹¹⁾ Compounds 1–3 were dissolved in dimethyl sulfoxide (DMSO). The arginases I and II were first activated by an incubation at 55 °C for 10 min. Afterwards, 0.2 ng of activated arginases I or II was added to a final volume of 35 µL reaction buffer (28 mM Tris–HCl, pH 7.5) containing 4.3 mM MnCl₂, 14.3 mM L-arginine, 0.015% bovine serum albumin, and inhibitors 1–3 (25, 50, 100, 200, and 400 µM) or DMSO as a control. After an incubation for 30 min at 37 °C, the reaction was terminated by adding 60 µL of a mixture of H₂SO₄/H₃PO₄/H₂O (1 : 3 : 7), followed by the addition of 5 µL of α-isonitrosopropiophenone (5% in absolute ethanol). After an incubation at 95 °C for 45 min in the dark, the reaction mixture was centrifuged for 10 min to eliminate the precipitates. The supernatants were transferred to a 96-well plate and shaken for 2 min, and then the arginase inhibition activities were measured using a spectrophotometer at 550 nm. The arginase inhibitory activity was calculated as the percentage of inhibition relative to “100% arginase activity.” The inhibitory modes of 2 and 3 were determined by a Dixon plot analysis. Experimental data were gathered in triplicate, covering concentrations ranging from 6.25 to 100 µM for compound 3 and 2.5 to 10 mM for L-arginine.

ECD Calculation

The absolute configurations of 1 and 2 were determined using Time-Dependent Density Functional Theory-Electronic Circular Dichroism (TDDFT-ECD) as described previously, with some modifications.²⁴⁾ Conformation search using molecular mechanics calculations were performed by Winmostar 10. Merck molecular force field (MMFF) calculations were performed by using Avogadro 1.2. These conformers were optimized by the Gaussian 16 program at the B3LYP/6-31G(d) level with the Polarizable Continuum Model (PCM) in MeOH. The conformers within 2 kcal/mol were then selected and used for ECD calculations (Supplementary Figs. S10, S18). The ECD calculations of all conformers were performed by using the TDDFT method at the B3LYP/6-31G(d) level with PCM in MeOH, and the weighted-average spectra were compared with the experimental ECD spectra recorded in MeOH.

Docking Simulation

The docking simulations were performed using the reported crystal structure of arginase I (PDB ID: 6Q9P).⁸⁾ The Avogadro 1.2 software was used to construct the three-dimensional model of the compound 3. Autodock Vina 1.0.2²⁵⁾ was employed to dock the compound into the arginase I structure. In the docking procedure, the side chains of the Arg21, Asn130, Thr136, Ser137, Asn139, His141, Asp181, Asp183, Glu186, Thr246, and Glu277 residues were set as flexible. The docking model of arginase I complexed with nor-NOHA revealed that the simulated nor-NOHA adopted the binding interaction of that in the reported arginase I crystal structure complexed with nor-NOHA (PDB ID: 3KV2),⁶⁾ with the binding energy of −6.6 kcal/mol, suggesting that this docking methodology might be suitable for human arginase I. The binding affinity of compound 3 was calculated to be −7.8 kcal mol⁻¹.

Acknowledgments

This work was supported in part by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology, Japan (JSPS KAKENHI Grants JP22H02777 to H.M., JP23K06179 to T.K., and JP22K15303 to Y.N.) and by a Grant-in-Aid for the Cooperative Research Project from Japan Preventive Medical Laboratory Company, Ltd.

Author Contributions

NNH performed all experiments. SH, KMD, and TK supported the structure elucidation of compounds. HXT collected and identified the plant sample. YN performed the docking simulation. SH, TK, YN, and HO wrote the manuscript draft. HM designed this study and wrote the final manuscript. All authors commented on the manuscript. All authors read and approved the final manuscript.

Conflict of Interest

The authors declare no conflict of interest.

Supplementary Materials

This article contains supplementary materials.

References

• 1) Caldwell R. B., Toque H. A., Narayanan S. P., Caldwell R. W., Trends Pharmacol. Sci., 36, 395–405 (2015).
• 2) Cederbaum S. D., Yu H., Grody W. W., Kern R. M., Yoo P., Iyer R. K., Mol. Genet. Metab., 81, 38–44 (2004).
• 3) Morris S. M., Bhamidipati D., Lenhart D. K., Gene, 193, 157–161 (1997).
• 4) Caldwell R. W., Rodriguez P. C., Toque A. A., Narayanan S. P., Caldwell R. B., Physiol. Rev., 98, 641–665 (2018).
• 5) Detroja T. S., Samson A. O., Molecules, 27, 5134 (2022).
• 6) Di Costanzo L., Ilies M., Thorn K. J., Christianson D. W., Arch. Biochem. Biophys., 496, 101–108 (2010).
• 7) Collet S., Carreaux F., Boucher J. L., Pethe S., Lepoivre M., Bougot R. D., Danion D., J. Chem. Soc., Perkin Trans. 1, 24, 4197–4525 (2000).
• 8) Grobben Y., Uitdehaag J. C. M., Seegers N. W., Tabak W. W. A., Man J. D., Buijman R. C., Zaman G. J. R., J. Struct. Biol. X, 4, 100014 (2020).
• 9) Abdelkawy K. S., Lack K., Elbarby F., Eur. J. Drug Metab. Pharmacokinet., 42, 355–370 (2017).
• 10) Thernier C. G., Pham T. N., Demougeot C., Mini Rev. Med. Chem., 15, 798–808 (2015).
• 11) Hoang N. N., Kodama T., Nakashima Y., Do M. K., Hnin S. Y. Y., Lee Y. E., Prema, Ikumi N., Morita H., J. Nat. Med., 77, 891–897 (2023).
• 12) Sharma P., Dwivedee B. P., Bisht D., Dash A. K., Kumar D., Heliyon, 5, e02437 (2019).
• 13) Kiem P. V., Minh C. V., Dat N. T., Kinh L. V., Hang D. T., Nam N. H., Cuong N. X., Huong H. T., Lau T. V., Fitoterapia, 81, 485–489 (2010).
• 14) Panchabhai T. S., Kulkarni U. P., Rege N. N., Phytother. Res., 22, 425–441 (2007).
• 15) Sivasubramanian A., Narasimha K. K. G., Rathnasamy R., Campos A. M. F. O., Nat. Prod. Res., 27, 1431–1436 (2013).
• 16) Pate M. B., Mishra S., Phytother. Res., 26, 1342–1347 (2012).
• 17) Hanuman J. B., Mishra A. K., Sabata B., J. Chem. Soc., Perkin Trans. 1, 1181–1185 (1986).
• 18) Phillip S., Tom G., Vasumathi A. V., J. Pharm. Pharmacol., 70, 1113–1125 (2018).
• 19) Palmieri A. P., Scapoli L., Lapichino A., Mercolini L., Mandrone M., Poli F., Gianni A. B., Baserga C., Martinelli M., Int. J. Immunopathol. Pharmacol., 33, 2058738419855567 (2019).
• 20) Singh N., Singh S. M., Srivastava P., Immunopharmacol. Immunotoxicol., 27, 1–14 (2005).
• 21) Zhu Y. L., Deng L., Dai X. Y., Song J. Q., Zhu Y., Liu T., Kong X. Q., Zhang L. Z., Liao H. B., Bioorg. Chem., 140, 106812 (2023).
• 22) Cama E., Colleluori D. M., Emig F. A., Shin H., Kim S. W., Kim N. N., Traish A. M., Ash D. E., Christianson D. W., Biochemistry, 42, 8445–8451 (2003).
• 23) Khangulov S. V., Sossong T. M. Jr., Ash D. E., Dismukes G. C., Biochemistry, 37, 8539–8550 (1998).
• 24) Hoang N. N., Kodama T., Win N. N., Prema, Do M. K., Abe I., Morita H., Chem. Biodivers., 18, e2100401 (2021).
• 25) Trott O., Olson A. J., J. Comput. Chem., 31, 455–461 (2010).