Exploring the Chemistry of Alkaloids from Malaysian Mitragyna speciosa (Kratom) and the Role of Oxindoles on Human Opioid Receptors

Abstract

Ten indole and oxindole alkaloids (1–10) were isolated from the freshly collected leaves of Malaysian Mitragyna speciosa (Kratom). The chemical structures of these compounds were established on the basis of extensive 1D and 2D NMR and HRMS data analysis. The spectroscopic data of mitragynine oxindole B (4) are reported herein for the first time. The spatial configuration of mitragynine oxindole B (4) was confirmed by single-crystal X-ray diffraction. Simultaneous quantification of the isolated alkaloids in the M. speciosa leaf specimens collected from different locations in the northern region of Peninsular Malaysia was also performed using UPLC-MS/MS. The oxindole alkaloids (1–4) and the indole alkaloid (10) were assessed for binding affinity at opioid receptors. Corynoxine (1) showed high binding affinity to μ-opioid receptors with a K_i value of 16.4 nM. Further, corynoxine (1) was 1.8-fold more potent than morphine in rats subjected to a nociceptive hot plate assay. These findings have important implications for evaluating the combined effects of the minor oxindole alkaloids in the overall therapeutic activity of M. speciosa.

Graphical Abstract

Mitragyna speciosa Korth. (Rubiaceae), known also as kratom (English), krathum (Thai), or ketum and biak biak (Malay), is a psychoactive plant indigenous to Southeast Asia, particularly in southern Thailand and the northern states of Peninsular Malaysia. Currently, the plant is widely cultivated in Southeast Asian countries such as Indonesia, Malaysia, and Thailand.¹ Indole and oxindole alkaloids are the primary components of M. speciosa. The corynanthe-type indole mitragynine (6) exists as the major alkaloid, constituting up to 66% of the total alkaloid content.² Additional significant alkaloids are speciogynine (7) and paynantheine (8), which combined account for approximately 16% of the total alkaloid fraction.²

The concentrations of individual alkaloids in commercialized M. speciosa products appear to be largely defined by their origin and environmental factors, although most often their origin is unknown. Nonetheless, there is a widely distinct alkaloid profile between M. speciosa grown in the United States (U.S.) and that grown in Southeast Asia.³ In fact, M. speciosa samples cultivated in the south-central U.S. have a concentration of approximately 0.04% mitragynine (6) (dry sample), several fold lower than concentrations present in the Southeast Asian samples.^2,3 This variance could be attributed to several factors, such as different chemotypes, climates, soil types, and environmental pressures. Recently, an alkaloid profile and simultaneous quantification of 10 different M. speciosa alkaloids were published, providing the first step toward standardization and quality control of marketed preparations and for analysis of wild and/or cultured plant material.⁴

Mitragynine (6) and its epimer speciociliatine (9) (2–4% of the total alkaloid content) together with a minor alkaloid, 7-hydroxymitragynine (11) (<2% of total alkaloid content), are currently known as the primary opioid-like components in M. speciosa.^5,6 These alkaloids are partial agonists at the human μ-opioid receptor (MOR) with K_i values of 161.0, 54.5, and 7.2 nM, respectively.⁶ Efforts to reveal the roles of other minor M. speciosa alkaloids are ongoing.^6,7

This report describes 10 alkaloids (1–10) (Figure 1) isolated from the leaves of Malaysian M. speciosa and the alkaloid profile of leaf samples collected from different locations in the northern states of Peninsular Malaysia. Additionally, this study reports in vitro μ-opioid receptor binding affinity for the following low-abundance Malaysian M. speciosa oxindole alkaloids: corynoxine (1), corynoxine B (2), isospeciofoline (3), mitragynine oxindole B (4), and the oxygenated indole alkaloid speciociliatine N(4)-oxide (10). Corynoxine (1) was also compared to morphine in a rodent (rat) hot plate assay of antinociception.

Figure 1.

Alkaloids (1−10) isolated from the Malaysian samples of Mitragyna speciosa.

RESULTS AND DISCUSSION

Chromatographic purification of the enriched alkaloidal fraction from the methanol extract of the freshly collected Malaysian M. speciosa leaves yielded four known oxindole alkaloids, corynoxine (1),⁸ corynoxine B (2),⁸ isospeciofoline (3),⁹ and mitragynine oxindole B (4).¹⁰ Also obtained were the known indole alkaloids corynantheidine (5),⁴ mitragynine (6),⁴ speciogynine (7),⁴ paynantheine (8),⁴ speciociliatine (9),⁴ and speciociliatine N(4)-oxide (10). The chemical structures of mitragynine oxindole B (4) and speciociliatine N(4)-oxide (10) were confirmed by spectroscopic analysis, with that of mitragynine oxindole B (4) also examined by X-ray crystallography. The structures of the previously described indole alkaloids (5–9) were determined by comparison of their spectroscopic data (MS and ¹H and ¹³C NMR) with previously reported values.

Compound 4 (mitragynine oxindole B) was isolated as colorless needle crystals and determined by HRESIMS to have a molecular formula of C₂₃H₃₀N₂O₅ (m/z 415.2227 [M + H]⁺, calcd for C₂₃H₃₁N₂O₅, 415.2155). This is consistent with 10 degrees of unsaturation. The ¹H and ¹³C NMR data (Table 1) resembled those of corynanthe-type oxindole alkaloids.^8,9,11 The ¹H NMR spectrum (Table 1) indicated the presence of an AMX trisubstituted aromatic system at δ_H 7.06 (dd, J = 8.0, 8.0 Hz, H-11), 6.48 (d, J = 8.0 Hz, H-10), and 6.42 (d, J = 8.0 Hz, H-12). The spectrum also showed a methoxy substituent attached to an aromatic ring at δ_H 3.78 (s, OCH₃-9), an ethyl group at δ_H 1.73 and 1.10 (m, H₂-19) and 0.80 (t, J = 7.38 Hz, CH₃-18), and an indolic NH proton at δ_H 7.79 (s, N(1)-H). Additionally, signals at δ_H 7.24 (s, H-17), 3.57 (s, OCH₃-17), and 3.56 (s, OCH₃-22) suggested the presence of a β-methoxy acrylate moiety. The ¹³C NMR spectrum (Table 1) confirmed 4 to have 23 carbon atoms, including a carbonyl (δ_C 180.5), an ester carbonyl (δ_C 168.2), three aromatic carbons (δ_C 128.0, 104.3, and 101.6), three aromatic quaternary carbons (δ_C 154.6, 141.1, and 117.5), three methines (δ_C 69.9, 39.5, and 39.2), five methylenes (δ_C 53.9, 53.4, 30.0, 24.2, and 18.1), and one quaternary carbon (δ_C 55.3). Two-dimensional (2D) NMR experiments, namely, homonuclear ¹H–¹H COSY and heteronuclear ¹H–¹³C HSQC and HMBC, helped assign the ¹H and ¹³C NMR data (Figure 2). The location of the β-methoxy acrylate moiety was confirmed by the long-range HMBC correlations observed for H-15 with C-16 and C-17 (Figure 2A). The methoxy group position was determined by long-range ¹H–¹³C HMBC correlation of the methoxy protons with C-9 in the aromatic ring. From these observations it was concluded that the planar structure is likely to be a tetracyclic oxindole, similar to corynoxine A/B but with an extra methoxy group in the aromatic ring at the C-9 position.

Table 1.

¹H and ¹³C NMR Data of Mitragynine Oxindole B (4) and Speciociliatine N(4)-Oxide (10)

	compound 4a		compound 10b
position	1H δH, mult. (J in Hz)	13C δC, mult.	1H δH, mult. (J in Hz)	13C δC, mult.
2		180.5, C		127.2, C
3	2.49, br d (10.8)	69.9, CH	5.09 br d (12.7)	65.95, CH
4
5	3.16, br t (7.7)	53.4, CH2	3.69–3.64, m	65.93, CH2
	2.42, m		3.57, mc
6	2.23, mc	30.0, CH2	3.57, mc	19.71, CH2
			3.18, m
7		55.3, C		105.0, C
8		117.5, C		116.7, C
9		154.6, C		154.3, C
10	6.48, d (8.0)	104.3, CH	6.46, d (8.1)	98.8, CH
11	7.06, dd (8.0, 8.0)	F128.0, CH	6.96, dd (8.1)	121.7, CH
12	6.42, d (8.0)	101.6, CH	6.90, d (8.1)	104.3, CH
13		141.1, C		138.1, C
14	2.23, mc	24.2, CH2	α, 2.66, ddd (14.5, 12.8, 6.5)	28.6, CH2
	0.98, br d (12.6)		β, 2.17, dt (14.4, 2.1)
15	2.59, dt (13.0; 3.0)	39.2, CH	3.56, mc	29.9, CH
16		110.6, C		110.6, C
17	7.24, s	159.4, CH	7.67, s	161.1, CH
18	0.80, t (7.4)	12.4, CH3	0.94, t (7.8)	10.8, CH3
19	1.73, m	18.1, CH2	1.28, m	22.9, CH2
	1.1, m
20	1.41, br d (10.8)	39.5, CH	2.79, ddd (12.4, 7.3, 5.0)	33.6, CH
21	3.20, br d (9.2)	53.9, CH2	β, 4.10 br t (12.0)	69.0, CH2
	2.04, br d (9.2)		α, 3.18, m
22		168.2, C		169.8, C
NH	7.79, s
OMe-9	3.78, s	60.6, CH3	3.88, s	54.1, CH3
OMe-17	3.57, s	54.2, CH3	3.93, s	60.9, CH3
OMe-22	3.56, s	50.2, CH3	3.75, s	50.7, CH3

^aCDCl₃.

^bCD₃OD.

^cOverlapped.

Figure 2.

(A) COSY and HMBC correlations and (B) selected NOESY correlations of mitragynine oxindole B (4).

The relative configuration of 4 was deduced from a NOESY experiment (Figure 2B), which indicated correlations of H-20 (δ_H 1.41) and H-15 (δ_H 2.59), clearly showing that these protons are in the same plane, while H-3 (δ_H 2.49) correlated with the H-21α (δ_H 3.20) and H-5α (δ_H 3.16) protons. The H-21α proton showed correlation with H-20, suggesting that the protons of the three aliphatic methines at position C-3, C-15, and C-20 are on the same face (3S, 15S, and 20S). This postulated structure was further validated by a single-crystal X-ray diffraction analysis (Figure 3), which also indicated an R configuration at the spiro-carbon, C-7. Prior to this isolation from M. speciosa leaves, compound 4 had previously been synthesized by oxidative conversion of mitragynine and assigned the name “mitragynine oxindole B”.¹⁰ We identified only one other report of the natural occurrence of a mitragynine oxindole, in a Thai M. speciosa specimen, which used TLC to identify mitragynine oxindoles A and B.¹¹ However, in neither of these previous studies had spectroscopic data been used to confirm the structure. The current report is the first to fully characterize mitragynine oxindole B (4), isolated from an authentic Malaysian M. speciosa specimen.

Figure 3.

Projection of mitragynine oxindole B (4), with the displacement ellipsoids drawn at the 50% probability level.

Speciociliatine N(4)-oxide (10) was obtained as a white, amorphous solid, and the molecular formula was confirmed by the analysis of the positive-ion HRESIMS (m/z 415.2591, calcd for C₂₃H₃₁N₂O₅, 415.2233). Compound 10 has the same molecular formula as 4 and similar NMR spectra (Table 1). Both compounds exhibit a trisubstituted aromatic AMX system at δ_H 6.96 (dd, J = 8.1, 8.1 Hz, H-11), 6.90 (d, J = 8.0 Hz, H-10), and 6.46 (d, J = 8.0 Hz, H-12), a methoxy substituent attached to an aromatic ring at δ_H 3.88 (s, OCH₃-9), an ethyl group at δ_H 1.23 (m, H₂-19) and 0.94 (t, J = 7.8 Hz, CH₃-18), and signals for the β-methoxy acrylate moiety at δ_H 7.67 (s, H-17), 3.93 (s, OCH₃-17), and 3.75 (s, OCH₃-22). However, compound 10 lacks a lactam carbonyl and a quaternary spiro-carbon, which suggests it has an indole alkaloid skeleton.^12,13 The differences in the NMR spectra between 10 (Table 1) and more typical mitragynine-type alkaloids are the aliphatic proton resonances in rings C and D were shifted to lower field, e.g., H-3 at δ_H 5.09 (br d, J = 12.7 Hz), H-21β at δ_H 4.10 (br d, J = 12.0 Hz), H-21α at δ_H 3.18 (m), H-4β at δ_H 3.67 (m), and H-4α at δ_H 3.57 (m). These shifts are consistent with the presence of an N-oxide moiety.¹² Homonuclear ¹H–¹H COSY and heteronuclear ¹H–¹³C HSQC and HMBC experiments indicated a structure similar to a mitragynine (6)- or speciociliatine (9)-type indole alkaloid with an N-oxide functionality at N(4).^12,14 A 2D NOESY experiment failed to establish the configuration of the methine protons at H-3, H-15, and H-20, although 1D selective NOE studies confirmed the proton configurations. Irradiation of H-20 at δ_H 2.79 correlated with the signals at δ_H2.66 (ddd, H-14α) and 3.18 (occurring as a double of doublets (dd) only in the 1D NOE experiment, H-21α). Moreover, the δ_H 3.56 (H-15) signal showed a broad triplet (br t) but only in the 1D NOE experiment, suggesting that H-20 and H-15 are on the same face of the molecule. Irradiation of the signal assigned for H-3 at δ_H 5.09 (d) showed a clear correlation of the signals at δ_H 2.17 (d, H-14β) and 4.10 (br t, H-21β), being on the opposite face of the molecule in comparison with H-15 (Figure 4). Additional irradiation of the isolated signals confirmed the structure as speciociliatine N(4)-oxide (10). The same low shift for H-3 has also been reported for the β configuration of isopaynantheine N(4)-oxide, a major secondary metabolite found in Mitragyna diversifolia.¹²

Figure 4.

Selective NOE correlations observed for speciociliatine N(4)-oxide (10).

Owing to the high polarity of speciociliatine N(4)-oxide (10), the R_f values on normal silica gel TLC were much lower than those of speciociliatine (9) in various solvent systems, giving indirect evidence for the N-oxide assignment. Speciociliatine and mitragynine N-oxides were first described as oxidative products formed in a harsh acid medium.¹⁰ However, Takayama et al. also derived mitragynine N-oxide from mitragynine using m-chloroperbenzoic acid.¹⁴ During the preparation of this article, speciociliatine N(4)-oxide (10) was identified by Flores-Bocanegra et al. in a commercial M. speciosa product.¹⁵

The reported alkaloidal concentrations in M. speciosa leaf samples collected in Malaysia and Thailand are significantly different from each other,¹⁶ and specimens grown in the United States are even more distinctive.^3,16 The precise reasons for these distinctions are unclear, but potentially include the influence of growing conditions such as seasonal variance and photoperiod lengths, age of the plant at harvest, and sample handling (e.g., whether the leaves are fresh or dried). Postharvest environmental factors, such as storage, transportation, humidity, and contaminating microorganisms, may also significantly affect the alkaloid content of the leaf material. This suggests the opportunity to use authentic M. speciosa leaves from known origins to develop chemical fingerprints that can define biomass.

The specimens used in the current study were collected from different locations in Kedah and Penang, Malaysia (Table S1 and Figure S1, Supporting Information). Interestingly, while mitragynine (6) was the most abundant alkaloid in T1, −2, and −5–16, two samples (T3, T4) did not contain quantifiable levels of mitragynine (6), similar to plants reported that were grown in the south-central U.S.³ Trees T1–T5 were of similar age and located in the same Penang plantation, and the leaves were collected on the same day. Interestingly, mitragynine (6) was below the lower limit of quantification in T3 and T4, and speciogynine (7), and paynantheine (8) concentrations were also low, but speciociliatine (9) was in a higher concentration than in any of the other samples (Table 2). This variance in T3 and T4 suggests that M. speciosa exhibits chemotypes, which could also explain the alkaloidal profile found in M. speciosa growing in the U.S.³ and the recent report from Oberlies’ group that outlined clear differences in the alkaloidal composition on marketed M. speciosa specimens in the U.S.¹⁵ This difference in alkaloidal profiles demonstrates that the historical literature should not be taken as universal. It is worth noting that plants with multiple chemotypes are not unusual. Indeed Cannabis sativa is a prime example; certain chemotypes produce high levels of Δ⁹-trans-tetrahydrocannabinol (THC), while hemp varietals produce little to no measurable amounts of THC.¹⁷

Table 2.

Alkaloid Content in Freshly Collected M. speciosa Leaf Specimens from Malaysia

			mg compound/1000 mg dried leaves
state	place	code	1a	2a	3a	4a	5a	6a	7a	8a	9a	10a	11b	12b	13b	14b
Penang	Kubang Semang	T1	0.37	0.12		0.10	0.08	10.89	1.12	1.17	3.51		0.05	0.07
		T2	0.27	0.10	0.03	0.18	0.24	12.12	1.05	1.07	6.38	0.02	0.06	0.17
		T3	0.46	0.15		0.11	0.08				13.49	0.06		0.09
		T4	0.28	0.09		0.08	0.12			0.01	14.76	0.06		0.10
		T5				0.16	0.09	18.85	1.23	0.93	2.04		0.15	0.10
Penang	Bukit Mertajam (Pak Teh Plantation)	T6	0.03			0.14	0.03	9.38	0.99	1.01	1.65		0.08	0.07
	Bukit Mertajam (Mat Plantation)	T7	0.13	0.05		0.19	0.05	13.00	1.53	1.47	1.81		0.11	0.07
	Bukit Mertajam (Zalem Plantation)	T8	0.10	0.05		0.05	0.03	12.22	1.74	1.63	0.43		0.06	0.04
	Bukit Mertajam (Permatang Pauh Red Vein)	T9				0.05	0.22	18.45	1.95	1.90	1.64		0.04	0.05
	(Permatang Pauh Super Red Vein)	T10	0.25	0.12		0.07	0.03	9.73	1.25	1.40	1.12		0.03	0.05
Kedah	Kulim (Halim Plantation)	T11	0.15	0.06		0.09	0.03	17.29	1.99	1.78	1.00		0.06	0.05
	Baling	T12				0.18	0.06	15.68	1.33	1.09	1.25		0.11	0.05
		T13				0.16	0.13	11.02	0.97	0.88	1.94		0.12	0.03
		T14				0.14	0.42	12.42	1.23	1.09	1.39		0.09	0.12
		T15	0.07			0.21	0.16	16.52	1.55	1.18	0.97		0.15	0.04
		T16	0.16	0.06		0.17	0.23	13.28	1.12	1.12	2.22		0.11	0.07

^aIsolated alkaloids from M. speciosa: 1, corynoxine; 2, corynoxine B; 3, isospeciofoline; 4, mitragynine oxindole B; 5, corynantheidine; 6, mitragynine; 7, speciogynine; 8, paynantheine; 9, speciociliatine; 10, speciociliatine N(4)-oxide.

^bOther alkaloid standards: 11, 7-hydroxymitragynine; 12, isocorynantheidine; 13, ajmalicine; 14, mitraphylline. Non value = below lower limit of quantification (LLOQ) (<0.1%)

Mitragynine oxindole B (4) was present in all the specimens examined in the current study, but in highly variable concentrations (0.05–0.21 mg/g; Table 2). Similarly, the other oxindoles examined, corynoxine (1) (0–0.42 mg/g) and its epimer, corynoxine B (2) (0 to 0.17 mg/g), were inconsistent in their representation with previously investigated specimens. In Uncaria species, the oxindole content is known to be affected by environment and seasonal factors such as total precipitation and plant water content.^18,19 It is therefore possible that oxindole alkaloid biosynthesis from their indole precursors requires certain levels of water and oxygen. A short discussion about the biosynthesis of the kratom alkaloids (Figure S2) is provided in the Supporting Information.

7-Hydroxymitragynine (11) was present in small amounts in all samples (0.03–0.15 mg/g) except for T3 and T4, which supports recent reports suggesting 7-hydroxymitragynine could be a postharvest artifact of mitragynine.⁵ Two pentacyclic alkaloids, mitraphylline (13) and ajmalicine (14) (Figure S2, Supporting Information), have previously been reported in M. speciosa,² but were below the lower limit of quantification in all M. speciosa samples in the present study (Table 2).

Oxindoles (1–4) and the N-oxide (10) were screened at concentrations of 0.1 and 10 μM for their capacity to displace tritiated opioid agonists from cultured cells expressing recombinant human δ-, κ-, and μ-opioid receptors (Table 3). In addition, each compound’s binding displacement constant (K_i) for μ-opioid receptors was accessed over a more complete concentration range (Table 4). At 10 μM, all compounds except isospeciofoline (3) displaced >50% [³H]DADLE at δ-opioid receptors (53.5–87.1%), and all except isospeciofoline (3) and mitragynine oxindole B (4) displaced >50% [³H]U69,593 (51.8–80.9%) from κ-opioid receptors (Table 3). At human μ-opioid receptors (hMOR), corynoxine (1) exhibited the highest affinity with a K_i value of 16.4 ± 1.4 (mean ± SEM) nM. Interestingly, corynoxine B (2), which differs only in its configuration at C-7, demonstrated about 6 times less affinity (K_i = 109.8 ± 8.1 nM; Table 5). The O-methyl group at C-9 in mitragynine oxindole B (4) also greatly reduced human μ-opioid receptor affinity (1715.9 ± 912.0 nM), and was 16-fold lower than corynoxine B (2) (Table 4). In a recent study to evaluate the μ-opioid affinity for some selected M. speciosa alkaloids, corynoxine B (2) was found to have an affinity approximately 15-fold lower than that found in the present study.⁷ Since corynoxine (1) had the highest affinity for human μ-opioid receptors (Table 4), its in vivo activity was compared with that of morphine in a rat hot plate antinociception assay. Both compounds were administered via the intravenous (i.v.) route in order to minimize differences in bioavailability and time to maximum effect. Morphine dose-dependently increased antinociception, with an ED₅₀ value (95% confidence interval) of 12.1 (10.2–14.4) mg/kg, while the ED₅₀ value of corynoxine (1) was 6.75 (5.90–7.62) mg/kg (Figure 5, Table 5). Both morphine and corynoxine (1) antinociception was fully reversed by 0.1 mg/kg (i.v.) naltrexone (opioid receptor antagonist). These data demonstrate that corynoxine (1) is as antinociceptive as morphine in the hot plate assay, despite having slightly less affinity in the in vitro studies (Table 4). It is unclear whether this reflects a difference in the affinity of corynoxine for rat vs human opioid receptors or a difference in the efficacy of morphine and corynoxine (1) to produce antinociception in rats.

Table 3.

Percent Displacement of Radioligands by Selected M. speciosa Alkaloids at Opioid Receptor Subtypes in a Screening Profile^a

M. speciosa alkaloid concentration	δ-opioid	κ-opioid	μ-opioid
corynoxine (1)
100 nM	−3.9	3.8	80.1
10 000 nM	73.8	75.3	97.8
corynoxine B (2)
100 nM	−4.4	4.5	45.4
10 000 nM	65.5	51.8	94.4
isospeciofoline (3)
100 nM	0.0	0.1	13.1
10 000 nM	27.1	37.3	83.3
mitragynine oxindole B (4)
100 nM	10.4	5.0	7.3
10 000 nM	87.1	23.0	66.5
speciociliatine N(4)-oxide (10)
100 nM	5.1	6.6	10.5
10 000 nM	53.5	80.9	84.6

^aFor subtype δ, human recombinant Chem-1 (RBL) cells. [³H]DADLE at 0.5 nM was used as a ligand. Naltrexone (10 μM) was used as a nonspecific binder, with 60 min incubation. For subtype κ, human recombinant RBL cells. [³H]U69,593 at 0.5 nM was used as a ligand. Naloxone (10 μM) was used as a nonspecific binder, with 60 min incubation. For subtype μ, human recombinant HEK-293 cells. [³H]DAMGO at 0.5 nM was used as a ligand. Naloxone (10 μM) was used as a nonspecific binder, with 120 min incubation. All assays were performed at room temperature.

^bBold font in body of the table represents >50% displacement.

Table 4.

K_i Values of Alkaloids 1–10 at the Human μ-Opioid Receptor

alkaloid	μ-opioid, Ki ± SEM (nM)	average abundance in malaysian samples (% w/w) (except T3 and T4)
corynoxine (1)	16.4 ± 1.4	0.18
corynoxine B (2)	109.8 ± 8.1	0.06
Isospeciofoline (3)	789.4 ± 293.8	0.03
mitragynine oxindole B (4)	1715.9 ± 12.0	0.13
corynantheidine (5)6	1910 ± 50.0	0.22
mitragynine (6)6	161 ± 10.0	14.11
speciogynine (7)28	728 ± 61.0	1.48
paynantheine (8)28	410 ± 15.2	1.39
speciociliatine (9)6	54.5 ± 4.4	3.4
speciociliatine N(4)-oxide (10)	552.7 ± 22.5	0.01
DAMGO	0.403 ± 0.008
morphine	4.19 (2.03–11.1)a

^a95% confidence limit.

Table 5.

Statistical Analyses of Dose–Effect Functions for the Antinociceptive Effects of Test Compounds (i.v.) as Shown in Figure 5^a

drug	sex	dose (mg/kg, i.v.)	interaction	post hoc test
morphine + 0.1 mg/kg naltrexone	F1,16 = 0.0129; p = 0.915	F4,16 = 14.3; p < 0.001	F4,16 = 0.205; p = 0.932	17.8 mg/kg vs vehicle (t = 6.05, p < 0.001), 17.8 mg/kg vs naltrexone (t = 5.72, p < 0.001)
corynoxine (1) + 0.1 mg/kg naltrexone	F1,20 = 1.73; p = 0.259	F5,20 = 54.3; p < 0.001	F5,20 = 0.882; p = 0.478	17.8 mg/kg vs vehicle (t = 10.2, p < 0.001), 17.8 mg/kg vs naltrexone (t = 10.2, p < 0.001)

^aThe sample sizes are described in each figure legend. Comparisons were made using a two-way repeated-measures ANOVA followed by post hoc Bonferroni t tests with results shown when effects were statistically significant. Significant differences are in bold.

Figure 5.

Effects of i.v. injection of morphine and corynoxine (1) on hot plate latency in catheterized rats. Abscissae: Vehicle, cumulative drug dose in mg/kg (i.v., log scale), and 0.1 mg/kg naltrexone (i.v.). Ordinates: Percentage of maximum possible effect (MPE). The hot plate was maintained at 52 ± 0.1 °C. Each point represents the mean ± SEM (n = 6, three subjects per sex). Following vehicle, cumulative doses were administered, followed by naltrexone (0.1 mg/kg, i.v.) at 5 min intervals. The antinociceptive effects of morphine and corynoxine (1) were antagonized by naltrexone. ***p < 0.001 compared with vehicle. ^###p < 0.001 compared with morphine (17.8 mg/kg) or corynoxine (1, 10 mg/kg). Details for statistical analyses are shown in Table 5.

This study has revealed that the oxindole alkaloids corynoxine (1) and corynoxine B (2) have high binding affinities toward μ-opioid receptors. Since these oxindoles are found in very low concentrations within M. speciosa leaves, they may not exert a significant pharmacological effect in the overall translation to human consumption of M. speciosa. However, these oxindole alkaloids are reported as major components in other medicinal plants of the Uncaria genus, including Uncaria rhynchophylla, U. macrophylla, and U. tomentosa, popularly known as Cat’s Claw or Guo Teng.^20-22

In conclusion, 10 alkaloids were isolated from freshly collected Malaysian M. speciosa leaves including mitragynine oxindole B (4) and speciociliatine N(4)-oxide (10) and fully characterized. Additionally, a UPLC-MS method was described and demonstrated to simultaneously quantify 14 different M. speciosa alkaloids from 16 samples collected in Peninsular Malaysia. A high variability in the composition of the alkaloid fractions taken from different leaf specimens was observed, suggesting the existence of at least two different chemotypes, one with high content of mitragynine (6) and another with high concentrations of speciociliatine (9). This is also the first report to demonstrate that the oxindole alkaloid corynoxine (1) exhibits a high affinity for human μ-opioid receptors (16.4 ± 1.4 nM) and opioidergic efficacy in an antinociception assay comparable to that of morphine. This series of compounds may be valuable leads in the development of new analgesics.

EXPERIMENTAL SECTION

General Experimental Procedures.

¹H and ¹³C NMR spectra were obtained using Bruker Avance NEO 600 NMR and Bruker Avance II 600 MHz NMR spectrometers operating at 600 MHz in ¹H and 151 MHz in ¹³C (Bruker Co., Billerica, MA, USA). The chemical shift values were reported in parts per million units (ppm) from tetramethylsilane (TMS) using known solvent chemical shifts. Coupling constants were recorded in hertz (Hz). Standard pulse sequences were used for COSY, HMQC, HMBC, TOCSY, NOESY, and DEPT. HRMS and purity (≥98%) were determined using an Agilent 1290 Infinity series ultraperformance liquid chromatograph (UPLC) system equipped with a photodiode array detector and quadrupole-time-of-flight (QTOF) Agilent 6540 mass spectrometer (Agilent Co., Santa Clara, CA, USA). Column chromatography was carried out on silica gel (70–230 mesh, Merck) and Sephadex LH-20 (Mitsubishi Kagaku, Tokyo, Japan). TLC (silica gel 60 F254, Sorbtech, Norcross, GA, USA) was used to monitor obtained fractions from column chromatography. Preparative TLC was carried out on silica gel 60 PF254 + 366 plates (20 × 20 cm, 1 mm thick). Visualization of the TLC plates was achieved with a UV lamp (λ = 254 and 365 nm) and anisaldehyde/acid spray reagent (MeOH–acetic acid–anisaldehyde–sulfuric acid, 85:9:1:5). All solvents and reagents used for plant extraction and compound isolation were of analytical grade (Fisher Scientific, Fair Lawn, NJ, USA). Corynoxine (1), corynoxine B (2), isospeciofoline (3), mitragynine oxindole B (4), corynantheidine, (5), mitragynine (6), paynantheine (7), speciogynine (8), speciociliatine (9), and speciociliatine N(4)-oxide (10) (purity ≥98%) used for UPLC-MS analysis were extracted and purified from the leaves of M. speciosa following the procedures described below. 7-Hydroxymitragynine (11) (≥98%), used for UPLC-MS analysis, was synthesized according to the method described by Maxwell et al.²³ Another three alkaloid standards, (≥98%) isocorynantheidine (12), ajmalicine (13), and mitraphylline (14), were obtained from an in-house alkaloid library. LC-grade MeOH, acetonitrile, acetic acid, ammonium acetate, and water were purchased from Fisher Scientific. Corynoxine (1) was converted to its hydrochloride salt for animal studies.

Plant Material.

Fresh leaves of M. speciosa were collected on Permatang Rawa, Seberang Perai Tengah (5°22′17″ N; 100°27′1″ E), Penang, Malaysia, for the extraction and purification of alkaloids. The plant was authenticated by a botanist, Dr. Rosazlina Binti Rusly, from the School of Biological Sciences, Universiti Sains Malaysia (USM), and a voucher specimen (NEL-(K)-2019(02)) was deposited at the herbarium of the School of Biological Sciences, USM. The additional 16 samples (T1–T16) were collected from several kratom plantations (Kubang Semang, Bukit Mertajam, Kulim, and Baling areas) in the states of Kedah and Penang, Malaysia. The exact location and dates of collection are summarized in Table S2 and Figure S1 (Supporting Information).

Extraction and Isolation.

M. speciosa leaves (500 g) were extracted with hot MeOH (5 L) in a circulating water bath (45 °C) for 12 h followed by maceration at room temperature for another 12 h. The same extraction procedure was repeated three times with fresh MeOH until the solution became colorless. After evaporation in vacuo, the crude extract was partitioned between hexane (3 × 4 L) and 10% CH₃COOH (4 L). The water-soluble portion was then adjusted to pH 9.0 with 25% NH₄OH and extracted with CHCl₃ (3 × 4 L). The concentrated CHCl₃-soluble portion was lyophilized to yield a crystalline alkaloid fraction (6.8 g). The alkaloid fraction was subjected to a passage over silica gel column chromatography (CC) (hexane–EtOAc–MeOH, 100:0:0 to 0:0:100, v/v) to give seven fractions (Fr. 1–Fr. 8). Fr. 2 (2.0 g) was purified by a silica gel CC (hexane–EtOAc, 80:20, v/v) to afford corynantheidine (5) (14 mg) and mitragynine (6) (1.72 g). Fr. 3 (0.35 g) was subjected to flash silica gel CC (hexane–EtOAc, 90:10 to 20:80, v/v) to afford paynantheine (8) (20 mg). Fr. 4 (0.42 g) was further separated by silica gel CC (hexane–EtOAc, 70:30 to 0:100, v/v) to yield three subfractions (SFr. 4a, 4b, and 4c). SFr. 4a was repeatedly purified by silica gel CC (hexane–EtOAc, 40:60, v/v) to yield compound corynoxine (1) (30 mg). SFr. 4b was further purified by a normal-phase preparative TLC (CHCl₃–EtOAc, 85:15 v/v) to afford speciogynine (7) (22 mg). Fr. 5 (0.43 g) was separated by silica gel CC and eluted with CHCl₃–EtOAcMeOH mixtures (10:0 to 0:10) to afford three subfractions (SFr. 5a, 5b, and 5c). SFr. 5a was recrystallized to yield corynoxine B (2) (23 mg), while SFr. 5b was purified using preparative TLC (CHCl₃–MeOH, 95:5, v/v), to obtain isospeciofoline (3) (10 mg). SFr. 6 (0.82 g) was separated by silica gel CC (CHCl₃–EtoAc, 10:1 to 5:5, v/v) to give two subfractions (SFr. 6a and 6b). SFr. 6a was recrystallized using a mixture of CHCl₃–EtOAc (5:5, v/v) to give a white needle, mitragynine oxindole B (4) (20 mg). Speciociliatine (9) (48 mg) was obtained from SFr. 6b through silica gel CC (EtOAc–MeOH, 10:0 to 0:100, v/v). Speciociliatine N(4)-oxide (25 mg) (10) was obtained from the Fr. 7, as a white solid precipitate, which was cleaned by washing with a mixture of hexane–EtOAc (1:1) and filtered.

Mitragynine oxindole B (4).

Transparent crystals (EtOAc); mp 167–169 °C; UV (MeOH) λ_max (log ε) 220 (3.9), 245 (3.2), 290 (1.2) nm; IR (film, NaCl) ν_max 3236, 2948, 2791, 1728, 1707, 1615, 1466, 1450, 1251, 1152, 1098, 773 cm⁻¹; ¹H NMR and ¹³C NMR data, see Table 1; positive HRESIMS m/z 415.2227 [M + H]⁺ (calcd for C₂₃H₃₀N₂O₅, 415.2155).

X-ray Crystallography of Mitragynine Oxindole B (4).

Colorless crystals of compound 4 were obtained in ethyl acetate at room temperature. X-ray intensity data were collected at 100 K on a Bruker DUO diffractometer using Cu Kα radiation (λ = 1.541 78 Å), from an ImuS power source, and an APEXII CCD area detector. Raw data frames were read by the program SAINT²⁴ and integrated using 3D profiling algorithms. The resulting data were reduced to produce hkl reflections and their intensities and estimated standard deviations. The data were corrected for Lorentz and polarization effects, and numerical absorption corrections were applied based on indexed and measured faces. The structure was solved and refined in SHELXTL-2014, using a full-matrix least-squares refinement.²⁵ The non-H atoms were refined with anisotropic thermal parameters, and all of the H atoms were calculated in idealized positions and refined riding on their parent atoms. The correct stereochemistry of the molecule was refined evidenced by the following parameters: Flack x = 0.01(9), Parsons z = 0.06(9), Hooft y = 0.06(8). Additionally, the P2 and P3 for refining the true stereochemistry are 1.000, while the P3 (rac-twin) is 0.1 × 10⁻⁶ and P3 (false) is 0.1 × 10⁻³¹. In the final cycle of refinement, 3220 reflections (of which 3137 are observed with I > 2σ(I)) were used to refine 279 parameters, and the resulting R₁, wR₂, and S (goodness of fit) were 4.51%, 12.52%, and 1.031, respectively. The refinement was carried out by minimizing the wR₂ function using F² rather than F values. R₁ is calculated to provide a reference to the conventional R value, but its function is not minimized. Crystallographic data for 4 were deposited at the Cambridge Crystallographic Data Centre (deposition number: CCDC 2021386). Copies of the data can be obtained, free of charge, on application to the Director, CCDC, 12 Union Road, Cambridge CB2 1EZ, UK [fax: + 44(0)-1233-336033 or deposit@ccdc.cam.ac.uk].

Sample Preparation.

The leaves collected from the 16 different M. speciosa trees (T1—16) (Table S1, Supporting Information) were used for the sample preparation. A total of 25 g of powder of air-dried leaves from each sample was extracted with 250 mL of hot MeOH (45 °C) for 12 h, followed by maceration at room temperature for another 12 h. The same procedure was repeated three times to ensure all the alkaloids were extracted as indicated by Mayer’s reagent. The extracts (3 × 250 mL) from each sample were filtered through Whatman filter paper (No. 4, Fisher Scientific). The filtrate from each sample was evaporated in vacuo and freeze-dried to give a lyophilized powder. The powdered extracts were kept at −80 °C prior to UPLC-MS analysis.

Instrumentation and Analytical Conditions for Quantitative Analysis.

A UPLC-MS/MS method reported by Sharma et al.⁴ was implemented for the simultaneous quantification of 14 alkaloids in the collected M. speciosa samples. In brief, a Waters Acquity Class I plus UPLC coupled with a Xevo TQ-S Micro triple quadrupole mass spectrometer (Milford, MA, USA) was used for the analysis. Chromatographic separation of the analyzed alkaloids was achieved on a Waters Acquity BEH C₁₈ column (1.7 μm, 2.1 × 100 mm) along with the mobile phase consisting of 10 mM aqueous ammonium acetate buffer of pH 3.5 (A) and acetonitrile (B) using a gradient method over 22.5 min. The column was equilibrated initially with 80% A, which was held for 1 min, then linearly decreased to 72% until 19.5 min, then further linearly decreased to 10% reaching 21.5 min. The conditions were returned back to the initial conditions by sharply increasing A up to 80% by 21.5 min and maintaining them until 22.5 min to re-equilibrate the column.

A multiple reaction monitoring (MRM) mode was utilized for the detection of analytes using a triple-quadrupole mass spectrometer. All the analyzed alkaloids showed an adequate response with electrospray ionization (ESI) in the positive-ion mode. The monitored parent to corresponding daughter ion mass transitions (m/z) for 14 alkaloids and internal standards along with instrument parameters are shown in Table S2, Supporting Information. Collected M. speciosa samples were analyzed along with freshly prepared calibration and quality control standards. Control of UPLC/MS-MS, acquisition, and analysis of data was performed using MassLynx XS software version 4.2.

Radioligand Receptor Binding Assays.

The alkaloids from M. speciosa were screened at Eurofins Cerep (Celle l’Evescault, France) for their in vitro receptor-binding affinities at opioid receptor subtypes. Briefly, each cell membrane homogenate was incubated with a radioligand in the absence or presence of one of the isolated alkaloids in a buffer. Nonspecific binding was determined in the presence of a specific agonist or antagonist at the target receptor. Following incubation, each sample was filtered rapidly under vacuum through glass fiber filters presoaked in a buffer and rinsed several times with an ice-cold buffer using a 48- or 96-sample cell harvester. The experimental conditions that were used for the binding assay are summarized in Table 3.

In Vivo Functional Studies.

Salt and enantiomeric forms of the compounds used in the present study were as follows: (−)-morphine sulfate pentahydrate (National Institute on Drug Abuse, Drug Supply Program, Rockville, MD, USA), (−)-naltrexone hydrochloride (Sigma-Aldrich Co., St. Louis, MO, USA), and corynoxine (1) hydrochloride (see General Experimental Procedures). Doses of the compounds used in the present study are expressed as the weight of the salt or free base form listed above. All the compounds were dissolved in a vehicle consisting of sterile water (HyClone water for injection, 1 L, Fisher Scientific, Waltham, MA, USA) containing 5% Tween 80 (polyoxyethylenesorbitanmonooleate, Fisher Scientific) and 5% propylene glycol (Sigma-Aldrich Co.). All the compounds and the vehicle were administered i.v. Each solution was filtered with a 0.2 μm pore size syringe filter (Millex-LG, 0.20 μm, SLLG025SS, Sigma-Aldrich Co.) prior to injections. The injection volumes were 1.0 mL/kg of subject body weight. The dose ranges and pretreatment times were referenced to our preliminary data and the literature.²⁶

Animals.

Adult, naive female and male (n = 6 per sex) Sprague–Dawley rats (Envigo, Indianapolis, IN, USA; age 10 weeks) were preimplanted with a catheter into a right external jugular vein upon receipt and singly acclimated in ventilated cages for at least 3 days to a temperature- (21.9 ± 1.9 °C) and humidity-controlled (53 ± 14%) vivarium with a 12 h light/dark cycle (lights on at 07:00 h) during which food (2918 Teklad Global 18% protein rodent diets, Envigo) was available in their housing cages at all times and reverse osmosis water was available in the vivarium at all times. The animal protocol (#201710059) was approved by the Institutional Animal Care and Use Committee at the University of Florida and was in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals, which is fully accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International (AAALAC). All experiments were conducted in the light cycle (08:00 to 17:00 h) using drug naïve subjects.

Hot-Plate Testing.

The hard black anodized aluminum plate with built-in digital thermometer used in the present study (hot plate analgesia meter, 1440 analgesia hot plate with RS-232 port and software, Columbus Instruments, Columbus, OH, USA) was 33.1 cm long, 28.8 cm wide, and 9.9 cm high. A clear acrylic cage (26.7 cm long, 26.7 cm wide, 31.0 cm high) surrounded the square plate to confine an animal on the plate (total height: 40.9 cm). The size of the heated surface was 25.4 cm long, 25.4 cm wide, and 1.9 cm high, while the surface temperature range was 30 to 79.9 °C across the plate. The temperature accuracy was ±0.1 °C. The temperature, humidity, and light/dark cycle in an experimental room were equivalent to those in the vivarium. Prior to commencement of each experiment, the body weight of each animal was measured. The sample size of each experiment group per treatment was six (three subjects per sex). Following a baseline measurement for hot plate latency in each subject (see below for more details), vehicle and then cumulative injections of ascending doses of a test compound were administered i.v. at an interval of 5 min to assess the effects of compounds alone. Following are the baseline values for hot plate latency [mean (standard error of mean; SEM) seconds]: morphine [21.0 (3.11)] and corynoxine (1) [16.5 (3.31)]. For the baseline hot plate latency, two-way analysis of variance (ANOVA) indicated no significant difference in sex (F_1,8 = 1.94, p = 0.201), group (F_1,8 = 1.16, p = 0.312), or their interaction (F_1,8 = 1.86, p = 0.210). Because there was no significant difference in baseline values and sex, data from the hot plate assay were combined per treatment (see data analysis below).

Following the baseline measurement for hot plate latency, vehicle, cumulative injections of morphine or corynoxine (1), and naltrexone (0.1 mg/kg) were administered i.v. every 5 min. The measurement of each baseline was repeated five times at a 5 min interval. Immediately following assessment of the effects of vehicle, the cycle of injections of each test compound was repeated until the hot plate latency was maximized at 60 s in all subjects to avoid tissue damage, or any deemed toxicity (i.e., seizure) or death was observed. Finally, naltrexone was injected to see reversibility of the antinociceptive effects of both test compounds.

Data Analysis.

The hot plate latencies were normalized to the percentage of the maximal possible effect (MPE), using the following formula: % MPE = 100(postinjection latency – preinjection baseline latency)/(maximum latency 60 s – preinjection baseline latency). All data per treatment group in Figure 5 were presented as means (±SEM). Using the SigmaPlot software (ver. 14.0; Systat Software Inc., San Jose, CA, USA), the significance of effects on each variable above was assessed by ANOVA, with post hoc Bonferroni t tests. A two-way (repeated-measures) ANOVA was used as appropriate (factors were compound dose and sex). Outcomes were considered significant at p < 0.05. The ED₅₀ values (95% confidence intervals) for hot plate latency were calculated by linear regression²⁷ of the dose–effect functions using the GraphPad Prism software (ver. 8; GraphPad Software, San Diego, CA, USA).