DNA topoisomerase Ⅰ (Topo Ⅰ), a pivotal enzyme orchestrating DNA replication and transcription, is a validated drug target for cancer therapy [
1,
2]. Topo Ⅰ inhibitors exert their therapeutic effects by stabilizing the DNA-Topo Ⅰ cleavable complex, thereby preventing DNA religation and subsequently inducing DNA damage and tumor cell apoptosis [
3–
5]. Currently, the clinical use of Topo Ⅰ inhibitors is largely restricted to camptothecin derivatives such as irinotecan and topotecan [
6,
7]. Although these agents demonstrate significant efficacy against various malignancies, their clinical utility is substantially limited by dose-limiting toxicities and acquired drug resistance, which severely compromise long-term therapeutic outcomes [
8–
11]. Therefore, there is an urgent need for Topo Ⅰ inhibitors with novel chemical scaffolds.
β-Carboline alkaloids are nitrogen-containing heterocycles biosynthesized from tryptophan with a unique tricyclic pyrido[3,4-b]indole skeleton [
12]. This structure is associated with a range of pharmacological activities, with its notable antitumor potential attracting significant attention [
12–
15]. Among natural sources, plants of the
Peganum genus have been identified as exceptionally rich repositories of structurally diverse
β-carboline derivatives [
16]. Critically, many of these compounds, such as harmine and its analogs, have been identified as Topo Ⅰ inhibitors, making them a key focus for phytochemists and medicinal chemists [
17–
20].
The seeds of
Peganum harmala have been traditionally employed in Chinese folk medicine for the treatment of various ailments, including cancer, rheumatism, cough, and asthma [
21–
24]. In a previous study, we reported the isolation of new
β-carboline alkaloids [
24,
25]. As a part of our ongoing research, three novel long conjugated
β-carboline alkaloids, pecanthines A–C (
1−
3), were isolated from the seeds of
P. harmala (
Fig. 1). Their structures were elucidated by spectroscopic analyses and X-ray diffraction. Compounds
1 and
2 represent the first heterodimeric alkaloids merging canthin-6-one and melatonin-type scaffolds. More strikingly, the structure of peganumium B reported from the seeds of
P. harmala was revised to be compound
3 (
Fig. 1) [
26]. Herein, we report the isolation, structural elucidation, hypothetical biogenetic pathway, and biological evaluation of
1–
3. Meanwhile, the first gram-scale synthesis of
3 was achieved, further confirming the correctness of the structure.
Pecanthine A (1) was isolated as pink needle-shaped crystals. The molecular formula of 1 was established to be C28H24N4O4 from its high-resolution electrospray ionization mass spectrometry (HR-ESI-MS) data (m/z 481.1872 [M + H]+, calcd. for C28H25N4O4 481.1870). The 1H and 13C nuclear magnetic resonance (NMR) spectra data showed the presence of two AMX spin systems [δH 8.32 (1H, d, J = 8.7 Hz), δH 8.16 (1H, d, J = 2.4 Hz), δH 7.56 (1H, d, J = 8.7 Hz), δH 7.24 (1H, dd, J = 8.7, 2.4 Hz), δH 6.97 (1H, d, J = 2.3 Hz), δH 6.73 (1H, dd, J = 8.7, 2.3 Hz)], three aromatic protons [δH 8.83 (1H, d, J = 5.0 Hz), δH 8.20 (1H, d, J = 5.0 Hz), δH 8.15 (1H, s)], two methoxys [δH 3.94 and 3.80 (each 3H, s); δC 55.9 and 55.2], a methyl [δH 1.71 (3H, s); δC 22.6], and an indole NH signal [δH 11.14 (1H, s)]. The above data indicated that 1 was a dimeric β-carboline alkaloid. The 1H and 13C NMR data of 1 were definitively assigned by integrating 2D NMR experiments, as documented in Table S1 (Supporting information).
In the
1H–
1H COSY spectrum, the correlations between H-3 and H-4, between H-7 and H-8, between H-5′ and H-6′, and between H
2–11′ (
δH 3.34) and H
2–10′ (
δH 2.98)/NH-12′ (
δH 8.00) indicated the presence of four spin systems (C-3 to C-4, C-7 to C-8, C-5′ to C-6′, and C-10′ to N-12′) in
1 (
Fig. 2). In the heteronuclear multiple bond correlation (HMBC) spectrum, correlations between H-3 and C-1/C-5, between H-4 and C-6/C-13, between H-7 and C-5/C-9/C-11, between H-8 and C-6/C-10, between H-10 and C-6, between H-14 and C-13/C-16, as well as between 9–OCH
3 and C-9 suggested the existence of a 9-methoxycanthin-6-one alkaloid unit
1a (
Fig. 2). The HMBC correlations between H-5′ and C-3′/C-7′/C-9′, between H-6′ and C-4′/C-8′, between H-8′ and C-4′, between H
2–10′ and C-2′/C-4′, between H
2–11′ and C-3′/C-13′, H
3–14′ and C-13′, between 7′–OCH
3 and C-7′, as well as between NH-1′ and C-3′/C-4′ suggested the presence of a melatonin-type unit
1b (
Fig. 2). Furthermore, the HMBC correlations between H-14 and C-2′, and between NH-1′ and C-15 implied that the substructures unit
1a and unit
1b were connected through C-15–C-2′ bonds. Subsequently, crystals of
1 for single-crystal X-ray diffraction were acquired from CH
3OH–CH
2Cl
2 (1:1). Thus, compound
1 was defined as an unprecedented heterodimer alkaloid and further determined by single-crystal X-ray diffraction (Cu K
α) analysis (
Fig. 3).
The molecular formula of 2 was determined to be C27H22N4O4 based on its HR-ESI-MS data (m/z 467.1711 [M + H]+, calcd. for C27H23N4O4, 467.1714). Analysis of the 1H and 13C NMR spectra of 2 was similar to those of 1, except for the absence of a methyl signal (δC 22.6, C-14′). The structure was supported by the 1H–1H correlation spectroscopy (COSY) correlation between NH-12′ and H-13′, as well as the HMBC correlations between H-13′ and C-11′, between H2–11′ and C-13′ (Fig. S1 in Supporting information). Thus, compound 2 was elucidated and named pecanthine B.
Pecanthine C (3) was yielded as yellow powder. The molecular formula of 3 was deduced to be C24H21N3O2 on the basis of its HR-ESI-MS at m/z 384.1709 [M + H]+ (calcd. for C24H22N3O2: 384.1707). The UV spectrum of 3 exhibited absorption maxima at 205, 223, 280, 310 and 432 nm. Analysis of the 1H and 13C NMR spectra of 3 revealed the presence of two AMX spin systems [δH 7.91 (1H, d, J = 8.6 Hz), δH 7.53 (1H, d, J = 8.7 Hz), δH 7.10 (1H, d, J = 2.3 Hz), δH 6.92 (1H, d, J = 2.2 Hz), δH 6.77 (1H, dd, J = 8.7, 2.2 Hz), δH 6.74 (1H, dd, J = 8.6, 2.3 Hz)], two methoxyls [δH 3.84 and 3.83 (each 3H, s); δH 55.1 and 55.3], a methyl [δH 2.82 (3H, s); δH 19.6], and an indole NH signal [δH 11.73 (1H, s)]. Comprehensive analysis of 1H–1H COSY, heteronuclear single quantum coherence (HSQC), and HMBC spectra of 3 resulted in the unambiguous assignment of its 1H and 13C NMR signals as shown in Table S3.
In the
1H–
1H COSY spectrum, the correlations between H
2–3 (
δH 4.92) and H
2–4 (
δH 3.26), between H-7 and H-8, and between H-5′ and H-6′ indicated the presence of three spin systems (C-3 to C-4, C-7 to C-8, and C-5′ to C-6′) in
3 (
Fig. 2). In the HMBC spectrum, correlations between H
2–3 and C-1/C-5, between H
2–4 and C-6/C-13, between H-7 and C-5/C-9/C-11, between H-8 and C-6/C-10, between H-10 and C-6, between H-14 and C-1/C-13, between 9–OCH
3 and C-9, as well as between 12-NH and C-5/C-6 suggested the existence of a harmaline unit
3a (
Fig. 2). The HMBC correlations between H-5′ and C-3′/C-7′/C-9′, between H-6′ and C-8′, between H-8′ and C-4′, between H
3–11′ and C-3′/C-10′, as well as between 7′–OCH
3 and C-7′ suggested the presence of a 3-ethylidene-indole unit
3b (
Fig. 2). Furthermore, the HMBC correlations between H
2–3 and C-2′, between H-14 and C-3′, and between H
3–11′ and C-14 indicated that the substructures
3a and
3b were connected through N-2–C-2′ and C-14–C-10′ bonds. Therefore, the planar structure of
3 was established as a unique long conjugated 6/5/6/6/5/6 fused six-ring system.
Peganumium B, recently reported from the seeds of
P. harmala as a
β-carboline dimer with one novel long conjugated structure, has attracted our attention [
26]. Interestingly, compound
3 and peganumium B share identical molecular weights and NMR data, prompting us to re-evaluate peganumium B’s structure. In the original literature, the authors deduced peganumium B as a
β-carboline dimer based on NMR analysis. They further identified unit
a as harmaline through HMBC spectroscopy, which is consistent with our analysis of
3 in this study. Subsequently, the linkage between units
a and
b was inferred from chemical shifts and HMBC correlations. However, we have found that the HMBC spectrum of peganumium B exhibits strong cross-peaks for H
2–3/C-2′, H-14/C-11′, and H
3–11′/C-14 and C-3′ (Fig. S2 in Supporting information), which were not reported in the original study [
26]. These correlations perfectly match those of
3, where each pair of atoms is separated by three bonds, whereas in peganumium B, they would be separated by four or five bonds (
Fig. 4). Thus, a comprehensive re-evaluation of the spectroscopic data led to the revision of peganumium B as structure
3.
Compounds
1–
3 represent two novel classes of dimeric
β-carboline alkaloid analogues. The plausible biogenetic pathway of
1–
3 could be proposed based on the findings of key biogenetic precursors in the plant (
Scheme 1). The biosynthesis initiates with a Mannich reaction between tryptamine and 2-ketoglutaric acid, followed by successive oxidation, methoxylation, and ring condensation to afford the pivotal precursor
ⅰ [
27,
28]. Dimerization of
ⅰ with intermediate
ⅱa/
ⅱb through nucleophilic addition-elimination accounts for the formation of
1 and
2. As for compound
3, its source can be traced back to its precursor, harmaline. First, harmaline (
4) undergoes a Michael addition with crotonyl-CoA, followed by intramolecular condensation to afford the key intermediate
ⅲ [
29,
30]. Subsequent oxidation of
ⅲ yields intermediate
ⅳ, which then engages in a Claisen reaction with anthranilic acid to generate the dehydrated intermediate
ⅴ [
31]. Finally,
3 is formed through Schiff reaction and methoxylation.
Following the structural revision of peganumium B to
3 and unsuccessful attempts to obtain X-ray-quality crystals, we designed a biomimetic synthesis of
3 according to the biogenic pathway (
Scheme 2). The synthesis commenced with harmaline (
4) from
P. harmala seeds. Using commercially available Meldrum’s acid derivative
6 as an acylketene source, the key intermediate
7 was synthesized in 77% yield [
32]. Subsequent Boc protection of the indole NH followed by iodination at the
α-position of the carbonyl group afforded intermediate
8 in 83% yield over two steps. Next, a Suzuki-Miyaura coupling of
8 with boronic ester
9 was performed using Pd(dppf)Cl
2 and K
2CO
3 in 1,4-dioxane at 100 ℃, affording the desired
10 in 90% yield. Subsequent reduction with iron powder and NH
4Cl cleanly afforded intermediate
11 in 96% yield. The final transformation involved POCl
3-mediated intramolecular cyclization under reflux conditions, simultaneously effecting N-Boc deprotection. Ammonia workup delivered
3 in 87% yield on gram scale. The
1H and
13C NMR spectra of synthetic
3 matched those of the natural isolate.
β-Carboline alkaloids have attracted considerable attention due to their notable neuroprotective and antitumor activities. Our study demonstrated that compound
1 exhibited dose-dependent neuroprotective effects in an L-glutamate-induced HT-22 cell model (Fig. S3 in Supporting information). Meanwhile, compound
3 displayed significant cytotoxicity against PC-9 and H1299 cells, with the half maximal inhibitory concentration (IC
50) values of 0.856 ± 0.13 and 1.312 ± 0.08 µmol/L, respectively (
Fig. 5a). Colony formation assays (
Figs. 5b and c) and EdU proliferation assays (Figs. S4a and b in Supporting information) revealed that compound
3 effectively suppressed tumor cell clonogenicity and DNA synthesis. Furthermore, wound healing and transwell invasion assays confirmed that
3 significantly inhibited the migratory and invasive capabilities of PC-9 cells (Figs. S4c–g in Supporting information).
To elucidate the anti-tumor molecular mechanism of compound 3, we focused on Topo Ⅰ, a classic anti-tumor target. The results of relaxation experiments showed that 3 had significantly better inhibitory activity against Topo Ⅰ than the positive control drug camptothecin (
Fig. 5d). Molecular docking analysis revealed that 3 formed hydrogen bond interactions with Topo Ⅰ and produced
π-
π stacking effects with DNA base pairs (
Fig. 5e). These molecular interactions were the structural basis for its inhibitory activity against Topo Ⅰ. Given that compound 3 has potent Topo Ⅰ inhibition and antiproliferative activity, we further examined its effect on DNA damage. Immunofluorescence detection and Western blot revealed that the expression level of
γ-H2AX, a DNA damage marker, was significantly increased in PC-9 cells treated with 3 (
Figs. 5f–h). Flow cytometry analysis showed that 3 could arrest the PC-9 cell cycle at the S phase (
Figs. 5i and j) and significantly induce cell apoptosis (
Figs. 5k–m).
In summary, three novel long-conjugated
β-carboline alkaloids, pecanthines A–C (
1–
3), were isolated from the seeds of
P. harmala. Compounds
1 and
2 represent the first heterodimeric alkaloids merging canthin-6-one and melatonin-type scaffolds. The structure of peganumium B reported from the seeds of
P. harmala was revised to be compound
3. Meanwhile, a gram-scale synthesis of
3 was achieved in six steps, leveraging a key acyl-ketene imine condensation and Suzuki coupling, further corroborating its revised structure based on NMR analysis. The bioassays revealed that compound
1 exhibited dose-dependent neuroprotective effects in an L-glutamate-induced HT-22 cell model. Compound
3 displayed significant cytotoxicity against PC-9 and H1299 cells, with IC
50 values of 0.856 ± 0.13 and 1.312 ± 0.08 µmol/L, respectively. Significantly, compound
3 inhibits tumor cell proliferation by suppressing Topo Ⅰ activity, inducing DNA damage, arresting the cell cycle, and promoting apoptosis (
Fig. 5n). This finding provides a new structural template for the development of Topo Ⅰ inhibitors.
PDF
Full
98
53
关闭全屏
No. 86 Xueyuan South Road, Haidian District, Beijing
010-62662699(Operation and Technical Support)
qkjq@cast.org.cn
