Acylphloroglucinols with acetylcholinesterase inhibitory effects from the fruits of Eucalyptus robusta

Abstract

Eleven new acylphloroglucinols, including six new formylated phloroglucinol-monoterpene meroterpenoids, eucalyprobusals A–F (1–6), one monomeric acylphloroglucinol, eucalyprobusone B (7), and four dimeric acylphloroglucinols, eucalyprobusones C–F (8–11) were purified from the fruits of Eucalyptus robusta. The establishment of the structures of 1–11 was achieved by a combination of NMR and HRESIMS data analyses, electron circular dichroism (ECD), and single-crystal X-ray diffraction. Compounds 6, 8, and an inseparable mixture of 10 and 11 were found to be potent AChE inhibitors with IC₅₀ values of 3.22 ± 0.36, 3.82 ± 0.22, and 2.55 ± 0.28 μM, respectively. Possible interaction sites of 6, 8, 10, and 11 with AChE were investigated by means of molecular docking studies, and the results revealed that AChE residues Asn87, Ser125, Thr83, Tyr133, Tyr124, Tyr337, and Tyr341 played crucial roles in the observed activity of the aforementioned compounds.

1. Introduction

Plants of Eucalyptus genus (Myrtaceae) are a prolific resource of structurally intriguing phloroglucinol derivatives, especially formylated phloroglucinol meroterpenoids (FPMs) [1–4]. These Eucalyptus secondary metabolites not only possess multifarious bioactive properties, including protein tyrosine phosphatase 1B inhibitory [2], immunosuppressive [3], antimicrobial [4,5], antiviral [6], anticancer [7–9], AChE inhibitory [10], and anti-leishmanial [11] effects, but also have attracted significant attention from the synthetic organic chemistry community [12–18]. Eucalyptus robusta, a tall arbor indigenous to Australia, is widely cultivated in south China. Its leaves have been traditionally used as a Chinese folk medicine to treat dysentery, malaria, and bacterial diseases [19], whereas its fruits are usually used for the main treatment of malaria.

Alzheimer’s disease (AD), a neurodegenerative disorder associated with memory and other cognitive functions, has been commonly known as one of the most burdensome threats to increasingly elderly people [20,21]. Currently, the causative factors of AD are not fully understood, pathophysiological brain hallmarks mainly include low levels of acetylcholine (ACh), amyloid-β (Aβ) deposits, and neurofibrillary tangles. Despite decades of studies for the basic biology of AD and significant pharmaceutical efforts to develop viable therapies, no effective therapy is available to totally cure AD or to significantly inhibit the progression of AD symptoms. Pharmacologically, three marketed acetylcholinesterase inhibitors (AChEIs) that are approved by U.S. FDA, named donepezil, rivas-tigmine, and galantamine [22], are only relevant medicines for the treatment of ameliorating the symptoms of AD patients. All these AChEIs acting on central nervous system (CNS) cholinergic pathways are now approved for mild to severe dementia, although they are widely used for patients in earlier predementia stages associated with significant progressive memory impairment. Therefore, it would be of great significance to hunt for potent AChEIs from medicinal plant resources. The PE (petroleum ether)–EtOAc (ethyl acetate) extract of E. robusta fruits displayed an AChE inhibitory rate of 68% at the concentration of 500 μg/mL, which prompted further phytochemical investigation with the aim at clarifying its bioactive constituents. As a result, six new FPMs, eucalyprobusals A–F (1–6), one monomeric acylphloroglucinol, eucalyprobusone B (7), and four acylphloroglucinol dimers, eucalyprobusones C–F (8–11) were isolated and structurally characterized (Fig. 1). AChE inhibitory assays of 1–11 were performed, and the possible action sites of 6, 8, 10, and 11 with AChE were also accomplished via molecular docking methods.

Fig. 1.

Structures of 1–11 isolated from E. robusta.

2. Experimental

2.1. General experimental procedures

Optical rotation and UV spectra were measured on a AUTOPOL VI automatic and a SHIMADZU UV-2700 UV–VIS instruments, respectively. CD data were recorded on an Applied Photophysics spectro-polarimeter. A Bruker FT-IR Tensor-27 infrared spectrophotometer was utilized for measuring the IR spectra (KBr disks). NMR spectra were collected on Bruker Ascend 500, 600, and 800 MHz instruments with various solvent (including CDCl₃, methanol-d₄, acetone-d₆, and pyridine-d₅) signals as referenced internal standards. An Agilent 1290 UPLC/6540 Q-TOF system was used for HRESIMS data. Crystallographic data of 1 and a mixture of 10 and 11 were obtained using a Bruker D8 QUEST diffractometer (λ = 1.54178 Å) with Cu Kα radiation. Silica gel, Sephadex LH-20, and MCI were applied as the packing materials for CC (column chromatography). Chiral analysis was performed on an Agilent 1100 instrument with a CHIRALPAK IC column (4.6 × 250 mm, 5 μm). A Hanbon Newstyle preparative HPLC instrument equipped with a SunFire Prep C₁₈ column (10 × 250 mm, 5 μm) was used to purify compounds.

2.2. Plant material

The E. robusta fruits authenticated by Dr. Rong Li (Kunming Institute of Botany, CAS) were collected from Kunming, Yunnan province, People’s Republic of China. A voucher specimen (HY0032) is deposited in the State Key Laboratory of Phytochemistry and Plant Resources in West China, Kunming Institute of Botany, Chinese Academy of Sciences.

2.3. Extraction and isolation

The dried E. robusta fruits (5.0 kg) were powdered and extracted with PE–EtOAc (1:1 v/v, 15 L × 3, each 24 h) to afford an inquinate residue. This crude extract (230.0 g) was subjected to silica gel CC eluting with PE–EtOAc (100:1→ 1:1, v/v) to afford six fractions (Fr. A–Fr. F) as monitored based on TLC by spraying with 10% FeCl₃-EtOH. Fr. D (20.5 g) was separated on an RP-18 column (MeOH–H₂O, 60:40→100:1, v/v, 1‰ FA in H₂O) and was further purified with a Sephadex LH-20 column (MeOH) and semipreparative HPLC (MeCN–H₂O, 80:20 v/v, 1‰ FA in H₂O) to yield 7 (22.2 mg), 8 (10.2 mg), 9 (6.8 mg), and a mixture of 10 and 11 (31.1 mg). Fr. E (10.5 g) was fractionated by an RP-18 column (MeOH–H₂O, 50:50→100:1, v/v, 1‰ FA in H₂O) and further purified via semipreparative HPLC (MeOH–H₂O, 98:2 v/v, 1‰ FA in H₂O) to give 4 (2.4 mg), 5 (3.3 mg), and 6 (54.3 mg). Likewise, Fr. F (16.0 g) was separated on an RP-18 column (MeOH–H₂O, 50:50→100:1, v/v, 1‰ FA in H₂O) and followed by semipreparative HPLC (MeCN–H₂O, 90:10 v/v, 1‰ FA in H₂O) to afford 1 (12.1 mg), 2 (8.8 mg), and 3 (1.4 mg).

2.3.1. Eucalyprobusal A (1)

Yellowish crystals (methanol-acetone, 1:1 v/v); ${[\alpha ]}_{\text{D}}^{23}$ + 91.8 (c 0.11, MeOH); UV (MeOH) λ_max (log ε) 207 (4.26), 281 (4.47), 368 (3.57) nm; IR (KBr) ν_max 3440, 2954, 1641, 1180, 781 cm⁻¹; ¹H (500 MHz, CDCl₃) and ¹³C (125 MHz, CDCl₃) NMR spectral data, see Table 1; (+)-HRESIMS m/z 425.1939 [M+Na]⁺ (calcd for C₂₃H₃₀O₆Na, 425.1935).

Table 1.

¹³C and ¹H NMR data for eucalyprobusals A–E (1–5) in CDCl₃

	Eucalyprobusal A (1)a		Eucalyprobusal B (2)a		Eucalyprobusal C (3)b		Eucalyprobusal D (4)a		Eucalyprobusal E (5)a
No.	δC	δH (J in Hz)	δC	δH (J in Hz)	δC	δH (J in Hz)	δC	δH (J in Hz)	δC	δH (J in Hz)
1	76.5		75.9		72.7		82.4		76.3
2	132.6	5.88 d (9.8)	133.4	5.79 d (10.0)	80.2	4.49 d (5.3)	195.5		150.0	6.81 d (10.1)
3	135.9	5.83 dd (9.8, 1.0)	135.3	5.83 d (10.0)	111.6	5.10 d (5.3)	121.9	5.85 s	130.6	6.05 d (10.1)
4	72.8		72.6		155.2		168.7		196.9
5α	27.0	1.54 dd (13.4, 2.4)	35.4	1.51 brd (13.2)	32.4	2.62 dd (18.8, 6.3)	32.9	2.73 ddd (19.4, 5.6)	40.4	2.64 dd (20.1, 7.5)
5β		1.38t (13.4)		1.41 dd (13.2, 2.5)		1.86 brd (18.8)		2.46 brd (3.7)		2.52t (7.5)
6	33.1	2.26 ddd (13.4, 6.2, 3.0)	33.5	2.36 brdd (13.2, 3.2)	40.2	2.23 brt (4.8)	44.0	2.49 m	35.0	2.54 m
7	23.7	1.39 s	27.2	1.55 s	26.4	1.65 s	22.5	1.63 s	25.4	1.67 s
8	37.8	1.76 sept (6.8)	37.6	1.71 sept (6.8)	34.5	2.06 sept. (6.8)	35.6	2.43 sept. (6.8)
9	16.2	0.93 d (6.8)	16.2	0.92 d (6.8)	21.1	0.74 d (6.8)	20.5	1.13 d (6.8)
10	17.3	0.87 d (6.8)	17.3	0.83 d (6.8)	20.9	0.72 d (6.8)	20.7	1.12 d (6.8)
1′	163.5		161.6		165.0		163.9		171.7
2′	104.2		104.0		109.1		104.6		99.3
3′	168.0		168.1		167.3		167.6		160.3
4′	104.2		104.1		105.6		104.8		103.7
5′	171.2		170.1		169.1		169.1		168.1
6′	103.1		103.8		117.4		106.2		103.6
7′a	192.3	9.96 s	192.3	9.99 s	193.7	10.10 s	192.6	10.14 s	21.5	2.41 dd (16.9, 6.0)
7′b										2.81 dd (16.9, 6.0)
8′	191.8	10.14 s	191.7	10.16 s	192.3	10.21 s	192.2	10.15 s	191.4	10.04 s
9′	28.4	3.21 ddd (10.9, 6.2, 3.9)	34.0	2.64 brdd (10.2, 3.2)	35.9	3.52 dt (8.2, 4.9)	31.2	2.66 ddd (9.8, 6.3, 3.5)	206.6
10′a	35.7	2.52 ddd (14.5, 10.9, 3.9)	43.3	1.80 ddd (13.1, 10.2, 3.2)	46.7	1.42 brdd (13.5, 6.8)	43.6	1.61 2H m	52.7	2.98 2H d (6.7)
10′b		1.26 ddd (14.2, 10.9, 3.9)		1.58 ddd (14.2, 10.2, 3.2)		1.27 brdd (13.5, 7.7)
11′	24.5	1.75 m	26.6	1.87 m	25.6	1.47 m	25.6	1.67 m	27.5	1.92 m
12′	20.9	1.00 d (6.5)	21.6	1.01 d (6.5)	22.3	0.93 d (6.3)	22.2	0.97 d (6.4)	22.7	0.98 d (6.7)
13′	24.3	0.96 d (6.5)	23.6	0.96 d (6.5)	22.8	0.85 d (6.3)	24.0	0.85 d (6.4)	22.7	0.98 d (6.7)
OH-3′		13.43 s		13.44 s		13.29 s		13.35 s	OH-1′	15.48 s
OH-5′		13.82 s		13.48 s		13.31 s		13.38 s	OH-5′	14.45 s

^aData were recorded at 500 MHz.

^bData were recorded at 800 MHz.

2.3.2. Eucalyprobusal B (2)

Yellowish amorphous powder; ${[\alpha ]}_{\text{D}}^{24}$−31.4 (c 0.12, MeOH); UV (MeOH) λ_max (log ε) 206 (4.22), 281.5 (4.38), 373 (3.55) nm; ECD (MeOH, Δε) 204 (+3.53), 213 (+1.20), 226 (+3.73), 246 (+0.42), 274 (+10.22), 306 (−6.48) nm; IR (KBr) ν_max 3441, 2952, 1641, 1179, 780 cm⁻¹; ¹H (500 MHz, CDCl₃) and ¹³C (125 MHz, CDCl₃) NMR spectral data, see Table 1; (+)-HRESIMS m/z 425.1942 [M+Na]⁺ (calcd for C₂₃H₃₀O₆Na, 425.1935).

2.3.3. Eucalyprobusal C (3)

Yellowish amorphous powder; ${[\alpha ]}_{\text{D}}^{24}$−254.3 (c 0.12, MeOH); UV (MeOH) λ_max (log ε) 206 (4.28), 273 (4.38), 343 (3.71), 380 (3.53) nm; ECD (MeOH, Δε) 221 + 27.39), 267 (−1.44), 290 (+2.04), 343 (−4.56) nm; IR (KBr) ν_max 3439, 2943, 1632, 1430, 1057 cm⁻¹; ¹H (800 MHz, CDCl₃) and ¹³C (200 MHz, CDCl₃) NMR spectral data, see Table 1; (−)-HRESIMS m/z 401.1978 [M−H]⁻ (calcd for C₂₃H₂₉O₆, 401.1970).

2.3.4. Eucalyprobusal D (4)

Yellowish amorphous powder; ${[\alpha ]}_{\text{D}}^{25}$−307.3 (c 0.13, MeOH); UV (MeOH) λ_max (log ε) 206 (3.15), 236 (3.09), 279 (3.23) nm; ECD (MeOH, Δε) 206 (−5.91), 242 (+9.73), 269 (−20.5), 316 (−0.59), 343 (−3.69) nm; IR (KBr) νmax 3436, 2937, 1721, 1629, 1468, 1024 cm⁻¹; ¹H (500 MHz, CDCl₃) and ¹³C (125 MHz, CDCl₃) NMR spectral data, see Table 1; (+)-HRESIMS m/z 423.1772 [M+Na]⁺ (calcd for C₂₃H₂₈O₆Na, 423.1778).

2.3.5. Eucalyprobusal E (5)

Yellowish amorphous powder; ${[\alpha ]}_{\text{D}}^{25}$−53.7 (c 0.09, MeOH); UV (MeOH) λ_max (log ε) 206 (4.00), 276 (4.09), 335 (3.39), 366 (3.20) nm; ECD (MeOH, Δε) 215 (+2.07), 239 (−0.61), 263 (+1.27), 292 (−0.58), 354 (+0.03) nm; 1H (500 MHz, methanol-d₄) and 13C (125 MHz, methanol-d₄) NMR spectral data, see Table 1; (+)-HRESIMS m/z 397.1051 [M+K]⁺ (calcd for C₂₀H₂₂O₆K, 397.1048).

2.3.6. (±)-Eucalyprobusal F (6)

Yellowish gum; ${[\alpha ]}_{\text{D}}^{22}$+ 86.8 (c 0.10, MeOH) for (+)-6; ${[\alpha ]}_{\text{D}}^{22}$−86.2 (c 0.10, MeOH) for (−)-6; UV (MeOH) λ_max (log ε) 202 (4.38), 213 (4.37), 291 (4.27), 388 (3.76) nm; ECD (MeOH, Δε) 204 (+14.46), 227 (+1.06), 246 (+21.48), 264 (+1.75), 274 (+3.08), 322 (−0.21) nm for (+)-6; ECD (MeOH, Δε) 204 (−13.89), 227 (−1.03), 246 (−20.39), 264 (−1.68), 274 (−2.97), 322 (+0.20) for (−)-6; ¹H (methanol-d₄, 500 MHz) NMR δ 0.88 (3H, d, J = 6.6 Hz, H₃-13′), 0.94 (3H, d, J = 6.6 Hz, H₃-12′), 1.17 × 2 (6H, d, J = 7.0 Hz, H₃-8/H₃-9), 1.49 (1H, m, H-11′), 1.80 (1H, ddd, J = 14.5, 8.0, 6.6 Hz, H-10′b), 2.19 (1H, ddd, J = 14.5, 8.4, 5.2 Hz, H-10′a), 2.22 (3H, s, H₃-10), 2.78 (1H, sept., J = 7.0 Hz, H-7), 4.65 (1H, dd, J = 9.6, 6.6 Hz, H-9′), 6.85 (1H, dd, J = 7.8, 1.5 Hz, H-4), 6.91 (1H, d, J = 7.8 Hz, H-5), 7.45 (1H, d, J = 1.5 Hz, H-2), 10.05 (2H, s, H-7′/H-8′); ¹³C (methanol-d₄,125 MHz) NMR δ 19.5 (C-10), 22.9 (C-12′), 23.8 (C-13′), 24.6 (C-8), 24.7 (C-9), 27.4 (C-11′), 35.0 (C-9′), 35.2 (C-7), 42.9 (C-10′), 106.3 × 2 (C-2′/C-4′), 111.3 (C-6′), 124.6 (C-4), 127.7 (C-2), 131.0 (C-5), 134.7 (C-6), 142.9 (C-1), 146.8 (C-3), 169.1 (C-3′), 169.9 × 2 (C-1′/C-5′), 193.1 × 2 (C-7′/C-8′); (−)-HRESIMS m/z 383.1872 [M−H]⁻ (calcd for C₂₃H₂₇O₅, 383.1864).

2.3.7. Eucalyprobusone B (7)

Yellowish gum; UV (MeOH) λ_max (log ε) 205 (3.87), 272 (4.22), 321 (3.59) nm; ¹H NMR (CDCl₃, 500 MHz) δ 1.16 × 2 (6H, d, J = 6.8 Hz, H 3–10/H₃-11), 3.68 (1H, sept., J = 6.8 Hz, H-9), 3.95 (3H, s, OMe-3), 5.91 (1H, s, H-4), 10.20 (1H, s, H-8), 12.99 (1H, s, OH-5), 15.50 (1H, s, OH-1); ¹³C NMR (CDCl₃, 125 MHz) δ 19.1 × 2 (C-10/C-11), 39.6 (C-9), 56.2 (OMe-3), 90.8 (C-4), 103.4 (C-2), 105.3 (C-6), 168.1 (C-3), 169.9 (C-5), 171.8 (C-1), 192.7 (C-8), 210.5 (C-7); (+)-HRESIMS m/z 261.0735 [M+Na]⁺ (calcd for C₁₂H₁₄O₅Na, 261.0733).

2.3.8. ( ± )-Eucalyprobusone C (8)

Yellowish gum; ${[\alpha ]}_{\text{D}}^{22}$+ 72.5 (c 0.10, MeOH) for (+)-8; ${[\alpha ]}_{\text{D}}^{22}$−72.4 (c 0.10, MeOH) for (−)-8; UV (MeOH) λ_max (log ε) 207 (4.48), 302 (4.43) nm; ECD (MeOH, Δε) 213 (−3.46), 234 (+13.65), 264 (−0.15), 308 (+4.11) nm for (+)-8; ECD (MeOH, Δε) 213 (+3.40), 235 (−13.39), 264 (+0.15), 308 (−4.04) for (−)-8; ¹H (600 MHz, acetone-d₆) and 13C (150 MHz, acetone-d₆) NMR spectral data, see Table 2; (+)-HRESIMS m/z 503.2645 [M+H]⁺ (calcd for C₂₈H₃₉O₈, 503.2639).

Table 2.

¹H and ¹³C NMR Data for Eucalyprobusones C (8) and D (9).

	Eucalyprobusone C (8)a		Eucalyprobusone D (9)b
No.	δC	δH (J in Hz)	δC	δH (J in Hz)
1	165.8		163.6
2	104.6		104.0
3	162.5		162.0
4	93.6	6.10 s	92.8	6.07 s
5	164.4		162.8
6	110.3		106.0
7	211.3		211.1
8	39.9	3.82 sept. (6.7)	39.3	3.79 sept. (6.8)
9	19.6	1.13 d (6.7)	19.2	1.17 d (6.8)
10	19.7	1.12 d (6.7)	19.2	1.17 d (6.8)
1′	165.8		169.9
2′	104.6		105.6
3′	162.5		165.3
4′	93.6	6.10 s	104.8
5′	164.4		168.4
6′	110.3		103.5
7′	211.6		15.1	3.74 2H s
8′	27.8	a 1.79 m, b 1.35 m	193.2	10.15 s
9′	46.7	3.69 sext. (6.7)	207.2
10′	16.8	1.11 d (6.7)	52.2	3.00 2H d (6.7)
11′	16.9	1.11 d (6.7)	25.2	2.44 brsept. (6.7)
12′			22.7	0.99 d (6.7)
13′			22.7	0.99 d (6.7)
1″	28.2	5.01 t (8.1)
2″	40.8	2.08 2H m
3″	27.4	1.44 brsept. (6.5)
4″	22.8	0.88 d (6.5)
5″	22.8	0.88 d (6.5)
OH-1		9.50 s		16.76 s
OH-1′		9.50 s		17.23 s
OH-3′				10.31 s
OH-5				8.93 s
OH-5′				14.48 s
OMe-3	56.2	3.90 s	55.8	3.86 s
OMe-3′	56.2	3.89 s

^aData were recorded at 600 MHz in acetone-d₆.

^bData were recorded at 500 MHz in CDCl₃.

2.3.9. Eucalyprobusone D (9)

Yellowish gum; UV (MeOH) λ_max (log ε) 234 (4.42), 301 (3.44) nm; ¹H (600 MHz, CDCl₃) and ¹³C (150 MHz, CDCl₃) NMR spectral data, see Table 2; (−)-HRESIMS m/z 499.1345 [M+K]⁺ (calcd for C₂₄H₂₈O₉K, 499.1365).

2.3.10. (±)-Eucalyprobusones E (10) and F (11)

Colorless crystals (methanol-acetone, 1:1 v/v); [α] −0.67 (c 0.15, MeOH); UV (MeOH) λ_max (log ε) 207 (4.43), 302 (4.38) nm; ¹H (600 MHz, pyridine-d₅) and ¹³C (150 MHz, pyridine-d₅) NMR spectral data, see Table 3; (+)-HRESIMS m/z 525.2467 [M+Na]⁺ (calcd for C₂₈H₃₈O₈Na, 525.2459).

Table 3.

¹³C (150 MHz) and ¹H (600 MHz) NMR data for eucalyprobusones E (10) and F (11) in pyridine-d_5.

	Eucalyprobusone E (10)		Eucalyprobusone F (11)
No.	δC	δH (J in Hz)	δC	δH (J in Hz)
1	166.6		165.6
2	104.6		105.1
3	161.4		161.5
4	92.6	6.27 s	92.6	6.28 s
5	164.4		164.5
6	111.1		111.1
7	210.0		210.2
8	39.5	3.75 sept. (6.7)	39.5	3.75 sept. (6.7)
9	19.4	1.13 d (6.7)	19.4	1.13 d (6.7)
10	19.5	1.10 d (6.7)	19.5	1.10 d (6.7)
1′	166.6		165.6
2′	104.6		105.1
3′	161.4		161.5
4′	92.6	6.28 s	92.6	6.28 s
5′	164.4		164.5
6′	111.1		111.1
7′	210.2		210.0
8′	46.2	3.65 m	46.2	3.65 m
9′	16.6	1.10 d (6.7)	16.7	1.14 d (6.7)
10′a	27.3	1.80 m	27.4	1.82 m
10′b		1.32 m		1.34 m
11′	12.0	0.80 t (7.4)	12.1	0.84 t (7.4)
1″	27.9	6.03 t (8.3)	27.9	6.03 t (8.3)
2″	41.8	2.47 2H t (7.5)	41.8	2.47 2H t (7.5)
3″	26.9	1.92, brsept. (6.6)	26.9	1.92 brsept. (6.6)
4″	22.8	0.90 d (6.6)	23.0	1.09 d (6.6)
5″	23.0	1.08 d (6.6)	23.0	1.09 d (6.5)
OMe-3/3′	55.3	3.64 s	55.3	3.65 s

2.3.11. Crystallographic data for eucalyprobusal A (1)

C 23H₃₀O₆, M = 402.47, a = 10.3730(6) Å, b = 13.0331(8) Å, c = 31.5028(18) Å, α = 90°, β = 90°, γ = 90°, V = 4258.9(4) ų, T = 100.(2) K, wavelength 1.54178 Å, orthorhombic crystal system, space group P2₁2₁2₁, Z = 8, absorption coefficient 0.735 mm⁻¹, μ(Cu Kα) = 0.735 mm⁻¹, F(000) = 1728, crystal size 0.260 × 0.200 × 0.140 mm³, θ range for data collection 3.67 – 72.32°, index ranges −12 ≤ h ≤ 12, −13 ≤ k ≤ 16, −38 ≤ l ≤ 38, 35,845 reflections collected, 8336 independent reflections (R_int = 0.0249), completeness to θ (72.32°) 99.4%, data/restraints/parameters 8336/0/553, largest diff. peak and hole 0.205 and −0.154 e.Å⁻³. The final R₁ values were 0.0247 [I > 2σ(I)]. The final wR(F²) values were 0.0647 [I > 2σ(I)]. The final R₁ values were 0.0250 (all data). The final wR(F²) values were 0.0649 (all data). The goodness of fit on F² was 1.042. Flack parameter = –0.01(2). Crystallographic data for 1 is deposited at CCDC (Cambridge Crystallographic Data Center) with a number of CCDC 2003650.

2.3.12. Crystallographic data for (±)-eucalyprobusones E (10) and F (11)

C 28H₃₈O₈, M = 502.58, a = 11.2889(2) Å, b = 11.5007(2) Å, c = 11.7348(2) Å, α = 81.6690(10)°, β = 78.5010(10)°, γ = 63.8270(10)°, V = 1337.06(4) ų, T = 100.(2) K, wavelength 1.54178 Å, triclinic crystal system, space group P-1, Z = 2, absorption coefficient 0.744 mm⁻¹, μ(Cu Kα) = 0.744 mm⁻¹, F(000) = 540, crystal size 0.630 × 0.480 × 0.270 mm³, θ range for data collection 4.29–72.38°, index ranges −13 ≤ h ≤ 13, −14 ≤ k ≤ 14, −14 ≤ l ≤ 14, 42,727 reflections measured, 5244 independent reflections (R_int = 0.0421) completeness to θ (72.38°) 99.3%, data/restraints/parameters 5244/255/425, largest diff. peak and hole 1.093 and –0.430 e.Å⁻³. The final R₁ values were 0.0707 [I > 2σ(I)]. The final wR(F²) values were 0.1933 [I > 2σ(I)]. The final R₁ values were 0.0712 (all data). The final wR(F²) values were 0.1936 (all data). The goodness of fit on F² was 1.111. Crystallographic data for 10 and 11 is deposited at CCDC (Cambridge Crystallographic Data Center) with a number of CCDC 2003659.

2.4. ECD computational methods

The ECD calculations of 2–6 and 8 were carried out using Gaussian 16 [23]. Conformational analysis of 2–6 and 8 was carried out by CONFLEX 8B software (CONFLEX Corporation, Tokyo, Japan) using MMFF94s molecular force field with a search limit of 1.0 kcal/mol to yield six, six, three, four, two, and 10 conformers, respectively. These initial structures were optimized via the Density Functional Theory (DFT) at the B3LYP/6–31+G(d) level in gas phase. The optimized conformations were used for ECD calculations by the Time Dependent DFT (TDDFT) at the B3LYP/6–311++G (2d, p) level.

2.5. AChE inhibitory assay

AChE inhibitory effects of all the isolated phloroglucinols were carried out on the basis of the spectrophotometric method in 96-well microplates with slightly modification [24]. Each well was filled with human acetylcholinesterase (0.02 U/mL, Sigma-Aldrich Corp., USA), phosphate buffer (pH = 8.0), and tested phloroglucinols (100, 50.0, 30.0, 10.0, 3.0, 1.0, and 0.2 μM) in DMSO and then incubated for 20 min at 37 °C. These reactions were initiated by the addition of 40 μL of solution containing Ellman’s reagent (DTNB, 0.625 mM of 5,5′-di-thiobis-2-nitrobenzoic acid) and acetylthiocholine iodide (0.625 mM) for AChE inhibitory assays, respectively. The results of acetylthiocholine hydrolysis were monitored at 405 nm for 1.0 h (30 s interval readings). DMSO and galanthamine were selected as the negative and positive controls, respectively. The percentage inhibition was calculated as follows:

$$\mathit{\text{inhibition}}(\%)=\frac{\text{E}-\text{S}}{\text{E}}\times 100$$

(E and S are the average absorption values for the enzyme activities treated without and with tested compounds, respectively).

2.5.1. Molecular modeling

Discovery Studio was used to carry out molecular docking studies using recently published methods [25].

3. Results and discussion

3.1. Structural elucidation

Dried and powdered fruits of E. robusta were extracted three times by PE–EtOAc at room temperature. The obtained extract was separated using silica gel chromatrography to give six fractions (Fr. A–Fr. F). Fractions D, E, and F were repeatedly chromatographed on silica gel, Sephadex LH-20, and RP-18 columns as well as using semipreparative HPLC to yield 11 new acylphloroglucinols (1–11). The structures of 1–11 were elucidated employing a combination of NMR and HRMS data analyses; the absolute configurations of 1–6 and 8 were established based on X-ray diffraction or ECD calculations. 1–6 are phloroglucinol–monoterpene conjugates, whereas 7 and 8–11 are mono- and dimeric-acylphloroglucinols (Fig. 1), respectively. Among them, 10 and 11 were found to be an inseparable mixture of two pairs of enantiomers.

Eucalyprobusal A (1), a yellowish crystal, had a molecular formula of C₂₃H₃₀O₆ as determined by the observed sodium adduct ion at m/z 425.1939 [M+Na]⁺ (calcd for C₂₃H₃₀O₆Na, 425.1935) in the HRESIMS spectrum. The IR spectrum showed absorptions at 3440 and 1641 cm⁻¹ which indicated the existence of hydroxy and carbonyl functionalities, respectively. The ¹H NMR spectral data (Table 1) disclosed resonances for four secondary methyls (δ_H 0.87, d, J = 6.8 Hz, H 3–10; 0.93, d, J = 6.8 Hz, H₃-9; 0.96, d, J = 6.5 Hz, H₃-13′; 1.00, d, J = 6.5 Hz, H₃-12′), a tertiary methyl (δ_H 1.39, s, H₃-7), two olefinic protons (δ_H 5.83, dd, J = 9.8, 1.0 Hz, H-3; 5.88, d, J = 9.8 Hz, H-2), two aldehyde protons (δ_H 9.96, s, H-7′; 10.14, s, H-8′), and two hydroxy protons (δ_H 13.43, s, OH-3′; 13.82, s, OH-5′). Besides the characteristic signals for a diformylated phloroglucinol (DFPG) scaffold (δ_C 103.1, C-6′; 104.2 × 2, C-2′/C-4′; 163.5, C-1′; 168.0, C-3′; 171.2, C-5′; 191.8, C-8′; 192.3, C-7′), ¹³C NMR spectral data (Table 1) showed 23 carbon resonances ascribed to five methyls (δ_C 16.2, C-9; 17.3, C-10; 20.9, C-12′; 23.7, C-7; 24.3, C-13′), two methylenes (δ_C 27.0, C-5; 35.7, C-10′), four methines (δ_C 24.5, C-11′; 28.4, C-9′; 33.1, C-6; 37.8, C-8), an endocyclic double bond (δ_C 132.6, C-2; 135.9, C-3), and two oxygen-bearing quaternary carbons (δ_C 72.8, C-4; 76.5, C-1). The aforementioned NMR signals of 1 closely resembled those of eucalyptin D [6] recently obtained from E. globulus fruits, the difference being the configuration of the C-4 hydroxy group. Combined with three spin systems (Fig. 2) as furnished by the ¹H−¹H COSY experiment, HMBC correlations from H₃-7 (δ_H 1.39) to C-6 (δ_C 33.1)/C-1 (δ_C 76.5)/C-2 (δ_C 132.6), from H₃-10 (δ_H 0.87)/H₃-9 (δ_H 0.93)/H₂-5 (δ_H 1.54, 1.38)/H-2 (δ_H 5.88) to C-4 (δ_C 72.8), from OH-3′ (δ_H 13.43) to C-2′ (δ_C 104.2)/C-4′ (δ_C 104.2)/C-3′ (δ_C 168.0), from OH-5′ (δ_H 13.82) to C-6′ (δ_C 103.1)/C-4′ (δ_C 104.2)/C-5′ (δ_C 171.2), from H-7′ (δ_H 9.96) to C-2′ (δ_C 104.2), from H-8′ (δ_H 10.14) to C-4′ (δ_C 104.2), and from H-9′ (δ_H 3.21) to C-6′ (δ_C 103.1)/C-1′ (δ_C 163.5) indicated that 1 was a phloroglucinol-monoterpene conjugate. Although the observed ROESY correlations of both H-6 (δ_H 2.26) and H-9′ (δ_H 3.21) with H₃-7 (δ_H 1.39) (Fig. S1, Supporting Information) revealed that these protons occupied the same side of the molecule and were stochastically assigned as β-oriented, no ROESY evidence was used to establish the configuration of the C-4 hydroxy group. Fortunately, needlelike crystals of 1 were obtained from a mixed solution of acetone and methanol (1:1, v/v). Single-crystal X-ray diffraction analysis with Cu Kα radiation (Fig. 3) of 1 not only resolved the configuration of C-4 hydroxyl, but also unequivocally established its absolute configuration (1R,4R,6R,9′S).

Fig. 2.

Selected ¹H–¹H COSY (blue bold line) and HMBC (red arrow) correlations of 1–11.

Fig. 3.

ORTEP drawing of 1.

Eucalyprobusal B (2) was assigned to have the same molecular formula (C₂₃H₃₀O₆) according to its HRESIMS ion at m/z 425.1942 [M +Na]⁺ (calcd for C₂₀H₃₀O₆Na, 425.1935). The 1D NMR data (Table 1) of 2 were highly similar to those of 1, and they shared the same planar architecture (Fig. 2) after detailed analysis of their ¹H–¹H COSY, HMBC, and HSQC data. Examination of the NMR data revealed that C-9’, C-10’, and C-11’ were significantly deshielded by Δδ_C + 5.6, +7.6, and +2.1, respectively, indicating that 2 should be a C-9′ epimer of 1. This assumption was supported by the ROESY correlations (Fig. S1, Supporting Information) of both H-10′a (δ_H 1.80) and H-6 (δ_H 2.36) with H₃-7 (δ_H 1.55). The absolute configuration (1R,4R,6R,9′R) of 2 was substantiated by a comparison of its calculated and experimental ECD spectra (Fig. 4).

Fig. 4.

Calculated and experimental ECD spectra for 2.

Eucalyprobusal C (3) was determined to have the same molecular formula (C₂₀H₃₀O₆) as those of 1 and 2 based on its HRESIMS ion at m/z 401.1978 [M−H]⁻ (calcd for C₂₀H₂₉O₆, 401.1970). Inspection of the ¹H and ¹³C NMR data of 3 (Table 1) suggested that it was also a phloroglucinol–monoterpene adduct, of which the latter unit is similar to that of euglobal G6 [26]. The ¹H–¹HCOSY spectrum (Fig. 1) revealed the presence of three structural fragments, H-2–H-3, H₃-9–H-8–H₃-10, and H₂-5–H-6–H-9′–H₂-10′–H-11′–H₃-12′ (H₃-13′), for the monoterpene scaffold. In the HMBC spectrum, the observed correlations from H 3–7 (δ_H 1.65) to C-6 (δ_C 40.2)/C-1 (δ_C 72.7)/C-2 (δ_C 80.2), from H₂-5 (δ_H 2.62, 1.86) to C-3 (δ_C 111.6)/C-4 (δ_C 155.2), and from H₃-10 (δ_H 0.72)/Me-9 (δ_H 0.74) to C-4 validated the existence of a γ-terpinene derivative with a C-2 hydroxy group in 3. Compared with the remarkably different ¹³C NMR data for 1 and 2, the downfield chemical shifts of C-9’ (δ_C 35.9) and C-10’ (δ_C 46.7) indicated an α-oriented configuration for the H-9’ in 3. The ROESY correlations (Fig. S1, Supporting Information) of H-10′a (δ_H 1.42)/H-6 (δ_H 2.23)/H-2 (δ_H 4.49) with H₃-7 (δ_H 1.65) proved that these protons were all β-oriented. The experimental ECD spectrum with two positive Cotton effects at 221 (+27.39) and 290 (+2.04) nm as well as two negative Cotton effects at 267 (−1.44) and 343 (−4.56) nm (Fig. 5) of 3 defined its absolute configuration (1S,2R,6R,9′R).

Fig. 5.

Calculated and experimental ECD spectra for 3.

Eucalyprobusal D (4) was shown to possess a molecular formula of C 20H₃₀O₆ due to its observed HRESIMS ion at m/z 423.1772 [M+Na]⁺ (calcd for C₂₃H₂₈O₆Na, 423.1778). The NMR data (Table 1) of 4 highly resembled those of 3, with the exception for the presence of a ketone carbonyl group (δ_C 195.5, C-2) in 4, instead of an oxygenated methine (δ_C 80.2, C-2; δ_H 4.49, H-2) in 3. The placement of this ketone carbonyl carbon at C-2 was proved by the evidently observed HMBC correlations from H₃-7 (δ_H 1.63) to C-6 (δ_C 44.0)/C-1 (δ_C 82.4)/C-2 (δ_C 195.5). Similarly, Me-7 in 4 was stochastically assigned a β-orientation, and the observed ROESY correlations (Fig. S1, Supporting Information) of H₃-7 with both H-10′ (δ_H 1.61) and H-6 (δ_H 2.49) revealed the β-orientations for the C-9′ isopentyl group and H-6 (Fig. 1). The absolute configuration (1S,6R,9′R) of 4 was established by a comparison of its experimental and calculated ECD spectra (Fig. 6).

Fig. 6.

Calculated and experimental ECD spectra for 4.

Eucalyprobusal E (5) was proved to share a molecular formula of C 20H₂₂O₆ owing to its HRESIMS ion at m/z 397.1051 [M+K]⁺ (calcd for C₂₀H₂₂O₆K, 397.1048). Its NMR data (Table 1) were highly similar to those of euglobal IIc [27], except for the presence of a ketone carbon (δ_C 196.9, C-4) and the disappearance of signals for the C-4 isopropyl functionality. Together with two fragments of H-2–H-3 and H₂-5–H-6 revealed by the ¹H–¹H COSY spectrum, HMBC correlations from H₃-7 (δ_H 1.67) to C-6 (δ_C 35.0)/C-1 (δ_C 76.3)/C-2 (δ_C 150.0) and from both H-2 (δ_H 6.81) and H₂-5 (δ_H 2.64, 2.52) to C-4 (δ_C 196.9) indicated the existence of an α-phellandrene derivative with the loss of a C-4 isopropyl group (Fig. 1). In the ROESY spectrum, the key correlations of H 3–7 (δ_H 1.67) with H-6 (δ_H 2.54) suggested that they shared -β-configurations. The experimental ECD curve with three positive Cotton effects at 215 (+2.07), 263 (+1.27), and 354 (+0.03) nm as well as two negative Cotton effects at 239 (−0.61) and 292 (−0.58) nm (Fig. 7) defined the absolute configuration (1S,6S) of 5.

Fig. 7.

Calculated and experimental ECD spectra for 5.

Eucalyprobusal F (6) had a molecular formula of C₂₃H₂₈O₅ as deduced from its HRESIMS ion at m/z 383.1872 [M−H]⁻ (calcd for C 20H₂₇O₅, 383.1864). The ¹H NMR spectrum displayed resonances for four secondary methyls (δ_H 0.88, d, J = 6.6 Hz, H₃-13′; 0.94, d, J = 6.6 Hz, H₃-12′; 1.17 × 2, both d, J = 7.0 Hz, H₃-8/H₃-9), a tertiary methyl (δ_H 2.22, s, H₃-10), three aromatic protons (δ_H 6.85, dd, J = 7.8, 1.5 Hz, H-4; 6.91, d, J = 7.8 Hz, H-5; 7.45, d, J = 1.5 Hz, H-2), and two aldehyde protons (δ_H 10.05 × 2, s, H-7′/H-8′). Apart from the readily discernable signals attributable for a diformylated phloroglucinol unit (δ_C 106.3 × 2, C-2′/C-4′; 111.3, C-6′; 169.1, C-3′; 169.9 × 2, C-1′/C-5′; 193.1 × 2, C-7′/C-8′), the ¹³C NMR data indicated the occurrence of five methyls (δ_C 19.5, C-10; 22.9, C-12′; 23.8, C-13′; 24.6, C-8; 24.7, C-9], one methylene (δ_C 42.9, C-10′), three methines (δ_C 27.4, C-11′; 35.0, C-9′; 35.2, C-7), and a trisubstituted benzene ring (δ_C 124.6, C-4; 127.7, C-2; 131.0, C-5; 134.7, C-6; 142.9, C-1; 146.8, C-3). Along with two spin systems in the ¹H–¹H COSY spectrum (Fig. 1), the observed HMBC correlations from H-4 (δ_H 6.85) to C-2 (δ_C 127.7)/C-6 (δ_C 134.7), from H-5 (δ_H 6.91) to C-1 (δ_C 142.9)/C-3 (δ_C 146.8), from H₃-8/H₃-9 (δ_H 1.17, both) to C-3 (δ_C 146.8), and from H₃-10 (δ_H 2.22) to C-5 (δ_C 131.0)/C-6 (δ_C 134.7)/C-1 (δ_C 142.9) allowed the establishment of the monoterpene moiety as p-cymene.⁷ The linkage of monoterpene and phloroglucinol units via a C-1–C-9′ bond was determined by the key HMBC correlations (Fig. 1) from H-9′ (δ_H 4.65) to C-2 (δ_C 127.7)/C-6 (δ_C 134.7)/C-1 (δ_C 142.9)/C-6′ (δ_C 111.3)/C-1′ (δ_C 169.9)/C-5′ (δ_C 169.9). Meroterpenoid 6 was determined to be a racemic mixture by HPLC analysis using a CHIRALPAK IC column (Fig. S2, Supporting Information). Chiral separation followed by ECD calculations determined the absolute configurations (9′S) and (9′R) for (+)-6 and (−)-6, respectively (Fig. 8).

Fig. 8.

Calculated and experimental ECD spectra of (±)-6.

Eucalyprobusone B (7) possessed a molecular formula of C₁₂H₁₄O₅ as revealed by an HRESIMS ion at m/z 261.0735 [M+Na]⁺ (calcd for C 12H₁₄O₅Na, 261.0733). With the assistance of HSQC spectrum, the ¹H and ¹³C NMR data showed the characteristic resonances for an isopropenyl (δ_H 1.16 × 2, d, J = 6.8 Hz, H₃-10/H₃-11; δ_C 19.1 × 2, H₃-10/H₃-11; δ_H 3.68, sept., J = 6.8 Hz, H-9; δ_C 39.6, C-9), one methoxy group (δ_H 3.95, s, OMe-3; δ_C 56.2, OMe-3), one pentasubstituted aromatic ring (δ_H 5.91, s, H-4; δ_C 90.8, CH-4; δ_C 103.4, C-2; δ_C 105.3, C-6; δ C 168.1, C-3; δ_C 169.9, C-5; δ_C 171.8, C-1), an aldehyde group (δ_H 10.20, s, H-8; δ_C 192.7, CH-8), a ketone carbonyl (δ_C 210.5, C-7), and two hydroxy protons (δ_H 12.99, s, OH-5; δ_H 15.50, s, OH-1). The aforementioned data indicated that 7 was a monomeric formylated phloroglucinol similar to 1,5-dihydroxy-2-(2′-methylpropionyl)-3-methoxy-6-methylbenzene [28], except for the replacement of a C-8 methyl (δ_H 1.97, s; δ_C 7.4) in the former by a formyl group (δ_H 10.20, s; δ C 192.7) in 7. The HMBC correlations (Fig. 1) from H₃-10 (δ_H 1.16)/H₃-11 (δ_H 1.16) to C-7 (δ_C 210.5), from OMe-3 (δ_H 3.95) to C-3 (δ_C 168.1), from H-8 (δ_H 10.20) to C-6 (δ_C 105.3), from OH-5 (δ_H 12.99) to C-4 (δ_C 90.8)/C-5 (δ_C 169.9), and from OH-1 (δ_H 15.50) to C-2 (δ_C 103.4)/C-6 (δ_C 105.3)/C-1 (δ_C 171.8) established the structure of 7.

Eucalyprobusone C (8) was deduced to have a molecular formula of C₂₈ H₃₈O₈ by its HRESIMS ion at m/z 503.2645 [M+H] ⁺ (calcd for C₂₈H₃₉O₈, 503.2639). The 1D NMR spectral data of 8 (Table 2) indicated it was a dimeric resorcinol analogue. Similar to 7, the discernable signals for an isopropenyl (δ_H 1.12, d, J = 6.7 Hz, H₃-10; δ_H 1.13, d, J = 6.7 Hz, H₃-9; δ_H 3.82, sept., J = 6.7 Hz, H-8), an isobutyl (δ_H 1.11 × 2, d, J = 6.7 Hz, H₃-10′/H₃-11′; 1.79 and 1.35, both m, H₂-8′; 3.69, sext., J = 6.7 Hz, H-9′), an isopentyl (δ_H 0.88 × 2, d, J = 6.5 Hz, H₃-4′’/H₃-5′’; 1.44, brsept., J = 6.5 Hz, H-3′’; 2.08, 2H, m, H 2–2′’; 5.01, t, J = 8.1 Hz, H-1′’), two methoxy groups (δ_H 3.89, s, OMe-3′; 3.90, s, OMe-3), two aromatic protons (δ_H 6.10 × 2, s H-4/H-4′), and two hydroxy protons (δ_H 9.50 × 2, s, OH-1/OH-1′) were readily recognized in the ¹H NMR spectrum (Table 2) of 8. Together with three fragments in blue bold lines (Fig. 1) as revealed by the ¹H–¹H COSY spectrum, HMBC correlations from both H₃-10 (δ_H 1.12) and H₃-9 (δ_H 1.13) to C-7 (δ_C 211.3), from H-9′ (δ_H 3.69) to C-7′ (δ_C 211.6), from OMe-3′ (δ_H 3.89) to C-3′ (δ_C 162.5), from OMe-3 (δ_H 3.90) to C-3 (δ_C 162.5), from both H-4 (δ_H 6.10) and H-4′ (δ_H 6.10) to C-2 (δ_C 104.6)/C-2′ (δ_C 104.6)/C-6 (δ_C 110.3)/C-6′ (δ_C 110.3)/C-3 (δ_C 162.5)/C-3′ (δ_C 162.5)/C-5 (δ_C 164.4)/C-5′ (δ_C 164.4), and from H-1′’ (δ_H 5.01) to C-6 (δ_C 110.3)/C-6′ (δ_C 110.3)/C-3 (δ_C 162.5)/C-3′ (δ_C 162.5)/C-5 (δ_C 164.4)/C-5′ (δ_C 164.4)/C-1 (δ_C 165.8)/C-1′ (δ_C 165.8) not only verified the presence of two methoxy resorcinol moieties, but also substantiated that these two units were connected via a C-6–C-1′’–C-6′ bond. An HPLC analysis equipped with a CHIRALPAK IC column (Fig. S2, Supporting Information) indicated that 8 was a racemic mixture, and ECD calculations (Fig. 9) was used to establish the absolute configurations (1′’S) and (1′’R) for (+)-8 and (−)-8, respectively.

Fig. 9.

Calculated and experimental ECD spectra of (±)-8.

Eucalyprobusone D (9) possessed a molecular formula of C₂₄H₂₈O₉ as inferred from an HRESIMS ion at m/z 499.1345 [M+K]⁺ (calcd for C₂₄ H₂₈ O₉K, 499.1365). The ¹H–¹H COSY spectrum (Fig. 1) indicated two coupled systems of H₃-9–H-8–H₃-10 and H₂-10′–H-11′–H₃-12′(H₃-13′). In the HMBC spectrum (Fig. 1), correlations from H₃-9/H₃-10 (both δ_H 1.17) to C-7 (δ_C 211.1), from OMe-3 (δ_H 3.86)/H-4 (δ_H 6.07) to C-3 (δ_C 162.0), from OH-1 (δ_H 16.76) to C-1 (δ_C 163.6)/C-2 (δ_C 104.0)/ C-6 (δ_C 106.0), and from OH-5 (δ_H 8.93) to C-4 (δ_C 92.8)/C-6 (δ_C 106.0) revealed the occurrence of an isobutyryl methoxy resorcinol moiety, whereas the observed correlations from H₂-10′ (δ_H 3.00) to C-9’ (δ_C 207.2), from H-8′ (δ_H 10.15) to C-3′ (δ_C 165.3)/C-5′ (δ_C 168.4), from OH-5′ (δ_H 14.48) to C-6′ (δ_C 103.5)/C-4′ (δ_C 104.8), and from OH-1′ (δ_H 17.23) to C-6′ (δ_C 103.5)/C-2′ (δ_C 105.6) allowed the establishment of a mono-formylated isovaleryl phloroglucinol unit. The key HMBC correlations from H₂-7′ (δ_H 3.74) to C-5 (δ_C 162.8)/C-1 (δ_C 163.6)/C-3′ (δ_C 165.3)/C-1′ (δ_C 169.9) unequivocally revealed that the two mono-phloroglucinol derivatives were connected by a C-7′–C-6 bond.

Eucalyprobusones E (10) and F (11) were isolated as two pairs of enantiomers with the same molecular formula (C₂₈H₃₈O₈) as that of 9 by HRESIMS (m/z 525.2467 [M+Na]⁺, calcd for C₂₈H₃₈O₈Na, 525.2459). A comparison of the 1D NMR data (Table 3) with those of 9 revealed that the isovaleryl in the latter was replaced by a sec-isovaleryl in the former ones. This was confirmed by the HMBC correlations from H 3–9′ (δ_H 1.10, J = 6.7 Hz for 10; 1.14, J = 6.7 Hz for 11) to C-8′ (δ_C 46.2 for both 10 and 11)/C-7′ (δ_C 210.2 for 10, 210.0 for 11), and from Me-11′ (δ_H 0.80, t, J = 6.7 Hz for 10; 0.84, t, J = 6.7 Hz for 11) to C-10′ (δ_C 27.3 for 10, 27.4 for 11)/C-8′ (δ_C 46.2 for both 10 and 11). Interestingly, an optical rotation value of 0.67 (c 0.15, MeOH) obtained for the mixture of 10 and 11 indicated that they should be racemic. Fortunately, a triclinic crystal obtained from a mixed solvent of acetone-MeOH (1:1, v/v) of 10 and 11 was selected for subsequent X-ray diffraction study. The results revealed that C-7′ sec-isobutyl and C-1′’ isobutyl units were both unordered, suggesting the occurrence of two pairs of enantiomers (Fig. 10). Nevertheless, it was not feasible to obtain (+)-10, (–)-10, (+)-11, and (–)-11 by a chiral column after several attempts (Fig. S2, Supporting Information). Taking the relationships between the specific rotation values and absolute configurations of (+)-8 and (–)-8 into consideration, the absolute configurations of (+)-10, (–)-10, (+)-11, and (–)-11 could be provisionally assigned as (8′R,1′’S), (8′R,1′’R), (8′S,1′’S), and (8′S,1′’R), respectively, owing to the fact that the existence of the C-7′ sec-isobutyl could not strikingly affect the holistic absolute configurations that were determined by the specific rotation values [29].

Fig. 10.

ORTEP drawing of (±)-10 and (±)-11.

3.2. AChE inhibitory effects

Given the PE–EtOAc extract of E. robusta fruits was AChE inhibitory (500 μg/mL, 68%), all the isolated acylphloroglucinols were screened for AChE inhibitory effects. At a concentration of 40.0 μM, only acylphloroglucinols 6 and 8–11 showed AChE inhibitory activities with inhibition rates ranging from 93.02 ± 0.71 to 71.97 ± 2.20%. Further studies indicated these compounds were AChE inhibitory with IC₅₀ values ranging from 2.55 ± 0.28 to 36.22 ± 2.29 μM, with the mixture of 10 and 11 being the most effective possessing an IC₅₀ value of 2.55 ± 0.28 μM (Table 4). Taking their structural characteristics and AChE inhibitory data into consideration, the observable structure–activity relationships can be summarized as follows (i) both FPMs featuring with a dihydropyran ring and acylphloroglucinol monomer were inactive; (ii) acylphloroglucinol dimers that be connected via an isopentyl moiety showed stronger AChE inhibitory effects than that of being linked by C-7′; (iii) the mixtures of (+)-6/(−)-6 and (+)-8/(−)-8 showed stronger AChE inhibitory activities than (+)-6, (−)-6, (+)-8, or (−)-8. Compared with structurally diverse acylphloroglucinol-like compounds reported from various plants [30–34], acylphloroglucinols 6, 8, and the mixture of 10 and 11 isolated from E.robusta fruits displayed more potential AChE inhibitory effects. With regard to acylphloroglucinol derivatives obtained from species of Myrtaceae, apart from polymethylated phloroglucinol meroterpenoids (PPMs) isolated from Rhodomyrtus tomentosa [35], the current findings indicated that FPM and acylphloroglucinol heterodimers [10] connected only by an isopentyl unit are more likely to be AChE inhibitors.

Table 4.

AChE inhibitory effects of acylphloroglucinols 1–11.

Compound	IC50 ± SD(μM)	Compound	IC50 ± SD (μM)
1	> 40.0	7	> 40.0
2	> 40.0	8	3.82 ± 0.22
3	> 40.0	(+)-8	4.96 ± 0.68
4	> 40.0	(−)-8	6.02 ± 0.54
5	> 40.0	9	36.22 ± 2.29
6	3.22 ± 0.36	10 + 11	2.55 ± 0.28
(+)-6	4.79 ± 0.57
(−)-6	5.85 ± 0.76	Galantaminea	1.05 ± 0.06

^aPositive drug.

3.3. Molecular docking investigation

Considering acylphloroglucinols 6, 8, and the mixture of 10 and 11 displayed good AChE inhibitory properties, molecular modeling investigations were used to better understand their mechanism of action and the binding modes with AChE (Fig. 11). The results revealed that all these isolates could be buried into the hydrophobic pocket of AChE. More specifically, (i) the acylphloroglucinol unit of 6 appears to form hydrogen bonds with the Tyr337, Tyr341, Thr83, and Ser125 residues, the phenyl ring of the monoterpene moiety was bound to the Trp86 residue via the π–π stacking interactions, and the terminal methyl fragments of the isopentyl moiety formed π-σ stacking interactions with Phe297 and Tyr124 residues; (ii) both the C-5 and C-5′ hydroxy groups of 8 could form hydrogen bonds with only the Tyr124 residue and two phenyl rings showed π–π interactions with Tyr341 and Trp86 residues, respectively, and the terminal methyl fragments of isopentyl, isobutyl, and isopropyl showed π–σ stacking interactions with Tyr337, Trp286/Phe297, and Trp86/Tyr337 residues, respectively; (iii) the phloroglucinol units of both 10 and 11 could form hydrogen bonds with Ser125, Tyr124, Tyr133, Tyr337, and Asn87 residues, the phenyl rings bearing a sec-isovaleryl group displayed π–π interactions with Trp86 and Tyr341 residues; (iv) the phenyl rings bearing a sec-isovaleryl group of 10 also showed π–π interaction with Tyr337 residue; (v) the terminal methyl fragments of 10 exhibited π-σ stacking interactions with Phe295, Phe297, Tyr124, and Trp86 residues, whereas those of 11 displayed π-σ stacking interactions with only Tyr124 and Trp286 residues. Through docking analysis, the racemic acylphloroglucinols 10 and 11 shared more interaction sites with AChE than 6 and 9 did, which were also consistent with the results of their AChE inhibitory assay.

Fig. 11.

The binding modes of 6 (A), 8 (B), 10 (C), and 11 (D) with human AChE (PDB ID: 4M0F). Hydrogen bond interactions are depicted with red dashes, while π–π and π–σ stacking interactions are displayed with yellow and green dashes, respectively.

4. Conclusion

In summary, the systematically phytochemical investigation of E. robusta fruits resulted in the isolation of 11 new acylphloroglucinols, including six new formylated phloroglucinol-monoterpene meroterpenoids (1–6), one monomeric acylphloroglucinol (7), and four dimeric acylphloroglucinols (8–11). Although all attempts to separate the 10 and 11 mixture have failed, X-ray diffraction was crucial for confirming their structures and absolute configurations. Compounds 6, 8, and the mixture of 10 and 11 displayed significant AChE inhibitory effects, and the possible interaction sites of these four compounds with AChE were investigated by molecular docking, which could be recognized as lead compounds for treatment of Alzheimer’s disease.