2.1. Structural Elucidation of the Glycosides
The concentrated ethanolic extract of the sea cucumber
Psolus chitonoides was submitted to hydrophobic chromatography on a Polychrom-1 column (powdered Teflon, Biolar, Latvia). The glycosides were eluted after washing with water as a mobile phase to eliminate salts and inorganic impurities with 50% EtOH. The obtained glycoside fraction was separated by the chromatography on Si gel columns with the stepped gradient of eluents CHCl3/EtOH/H2O (100:75:10), (100:100:17), and (100:125:25) to give the fractions (I–IV). The individual compounds
1–
4 (
Figure 1) were isolated by HPLC of the fractions III and IV on a silica-based column, Supelcosil LC-Si (4.6 × 150 mm) and reversed-phase semipreparative column Supelco Ascentis RP-Amide (10 × 250 mm).
The configurations of the monosaccharide residues in glycosides 1–4 were assigned as D based on their biogenetic analogies with all other known sea cucumber triterpene glycosides.
It was found that chitonoidosides E
1 (
1), F (
2) and H (
4) are characterized by holotoxinogenin as aglycone, which was first found in
Apostichopus japonicus and is broadly distributed in sea cucumber glycosides [
16]. This was deduced from the analyses of their
1H and
13C NMR spectra (
Tables S1–S3, Figures S1–S6, S9–S14 and S25–S31), which coincided with each other as well as with those of the aglycones of chitonoidosides A
1, C, and D isolated earlier from the same species—
P. chitonoides [
9].
The molecular formula of chitonoidoside E
1 (
1) was determined to be C
65H
100O
36S
2Na
2 from the [M
2Na–Na]
− ion peak at
m/z 1543.5315 (calc. 1543.5339) and [M
2Na–2Na]
2− at
m/z 760.2734 (calc. 760.2723) in the (−)HR-ESI-MS (
Figure S8). The
1H and
13C NMR spectra of the carbohydrate chain of chitonoidoside E
1 (
1) (
Table 1,
Figures S1–S7) were coincident with those for chitonoidoside E, isolated recently [
15] and demonstrated six characteristic doublets of anomeric protons at δ
H 4.67–5.11 (
J = 7.1–8.0 Hz) and six signals of anomeric carbons at δ
C 102.3–105.2. The analysis of the
1H,
1H-COSY, 1D TOCSY, HSQC and ROESY spectra of
1 resulted in the assignment of the signals of two xylose residues, one quinovose, one glucose and 3-
O-methylglucose, as well as 3-
O-methylxylose residues. The positions of the sulfate groups were determined based on the deshielding, due to α-shifting effect, of the sulfate group’s doubled signal at δ
C 67.1, which is characteristic of glucopyranose units sulfated by C-6 (C-6 signals of non-sulfated glucopyranose residues are usually observed at ~δ
C 61.2). The signals of C-5 MeGlc4 and C-5 Glc5 were shielded due to the β-effect of the sulfate groups to δ
C 75.3 and 75.5, respectively. These data indicate the presence of a hexasaccharide chain with two sulfate groups attached to C-6 MeGlc4 and to C-6 Glc5, and 3-
O-methylxylose residue as the sixth sugar unit in chitonoidoside E
1 (
1). The sequence of monosaccharides and the positions of the glycosidic bonds were confirmed by the correlations H-1 Xyl1/H-3 (C-3) of the aglycone, H-1 Qui2/H-2 (C-2) Xyl1, H-1 Xyl3/H-4 (C-4) Qui2, H-1 MeGlc4/H-3 (C-3) Xyl3, H-1 Glc5/H-4 (C-4) Xyl1, and H-1 MeXyl6/H-3 (C-3) Glc5 in the ROESY and HMBC spectra of
1, respectively (
Table 1,
Figures S5–S7).
The (
−)ESI-MS/MS of
1 (
Figure S8) demonstrated the fragmentation of [M
2Na–Na]
− ion at
m/z 1543.5 with the ion-peaks observed at
m/z 1423.6 [M
2Na–Na−NaHSO
4]
−, 1266.5 [M
2Na–Na−MeGlcOSO
3Na+H]
−, 1133.5 [M
2Na–Na−MeGlcOSO
3Na−Xyl+H]
−, 987.4 [M
2Na–Na−MeGlcOSO
3Na−Xyl−Qui+H]
−, 841.4 [M
2Na–Na−MeGlcOSO
3Na−Xyl−−Qui−MeXyl+2H]
−, 665.1 [M
2Na–Na–Agl−MeXyl−GlcOSO
3Na+H]
−, 533.1 [M
2Na–Na–Agl−MeXyl−GlcOSO
3Na−Xyl+H]
−, 387.0 [M
2Na–Na–Agl−MeXyl−GlcOSO
3Na−Xyl−Qui+H]
−, 255.0 [M
2Na–Na–Agl−MeXyl−GlcOSO
3Na−Xyl−Qui−Xyl+H]
−. All these ion peaks corroborated the sequence of monosaccharides and the aglycone structure of
1.
Therefore, chitonoidosides E [
15] and E
1 (
1) share the identical oligosaccharide moiety (they belong to the same group of glycosides) and differ from each other through the presence/absence of a carbonyl group at C-18 in the aglycones (the difference of their exact masses by 14
amu in (−) HR-ESI-MS confirmed this). These data indicate that chitonoidoside E
1 (
1) is 3
β-
O-{6-
O-sodium sulfate-3-
O-methyl-
β-D-glucopyranosyl-(1→3)-
β-D-xylopyranosyl-(1→4)-
β-D-quinovopyranosyl-(1→2)-[3-
O-methyl-
β-D-xylopyranosyl-(1→3)-6-
O-sodium sulfate-
β-D-glucopyranosyl-(1→4)]-
β-D-xylopyranosyl}-16-oxo-holosta-9(11),25(26)-diene.
The molecular formula of chitonoidoside F (
2) was determined to be C54H82O28S
2Na
2 from the [M
2Na–Na]
− ion peak at
m/z 1265.4337 (calc. 1265.4337) and [M
2Na–2Na]
2− ion peak at
m/z 621.2244 (calc. 621.2269) in the (−)HR-ESI-MS (
Figure S16). The 1H NMR spectrum of the carbohydrate part of chitonoidoside F (
2) exhibited four characteristic doublets at δ
H 4.67–5.18 (
J = 7.6–8.1 Hz), correlated by the HSQC spectrum with corresponding anomeric carbon signals at δ
C 102.2–104.9. These signals were indicative of a tetrasaccharide chain with
β-configurations of glycosidic bonds (
Table 2,
Figures S9–S15).
An isolated spin system from each monosaccharide residue was analyzed using the 1H,1H-COSY and 1D TOCSY spectra. Further analysis of the HSQC, ROESY and HMBC spectra resulted in the assignment of all monosaccharide NMR signals. Using this algorithm, the monosaccharides composing the carbohydrate moiety of chitonoidoside F (2) were found to be xylose (Xyl1), quinovose (Qui2), glucose (Glc3), and 3-O-methylglucose (MeGlc4). The monosaccharide compositions of the other glycosides, reported herein, were established in the same manner.
The signal of C-6 Glc3 in the
13C NMR spectrum of
2 was deshielded to δ
C 67.2, which is characteristic of sulfation at this position. The signal of C-6 MeGlc4 was observed at δ
C 61.2, indicating the absence of a sulfate group at this position, although the MS data indicated the presence of two sulfate groups in
2. The signal of C-4 MeGlc4 was deshielded to δ
C 76.3 when compared to the same signal of terminal 3-
O-methylglucose residues of the glycosides lacking a sulfate group (δ
C ~70.0) [
12]. Moreover, the signals of C-3 and C-5 MeGlc4 in the spectrum of
2 were shielded to δ
C 85.3 and 76.4, respectively, due to the
β-shifting effect of a sulfate group. Therefore, the 3-
O-methylglucose residue in the sugar part of
2 was sulfated by C-4. This structural feature was only found once in previous research—in the glycosides of
Colochirus quadrangularis [
17]. The observation of 3-
O-methylxylose residue sulfated by C-4 in the glycosides of
P. chitonoides [
15] clearly demonstrates the presence of a specific sulfatase capable of attaching a sulfate group to C-4 of monosaccharides in pyranose form.
The positions of glycosidic linkages, established by the ROESY and HMBC spectra of
2 revealed the uncommon architecture of the sugar chain with disaccharide fragment 4-
O-sodium sulfate-3-
O-methyl-
β-D-glucopyranosyl-(1→3)-6-
O-sodium sulfate-
β-D-glucopyranosyl-(1→4) attached to the first (Xyl1) residue, while quinovose was a terminal unit in the reduced bottom semi-chain (
Table 2). Two tetraosides whose carbohydrate part featured the same architecture were found only in the sea cucumber
Thyonidium kurilensis [
18].
The (
−)ESI-MS/MS of
2 (
Figure S16) demonstrated the fragmentation of [M
2Na–Na]
− ion at
m/z 1265.4 resulting in the ion-peaks appearance at
m/z 1146.5 [M
2Na–Na−NaSO
4]
−, 987.4 [M
2Na–Na−MeGlcOSO
3Na+H]
−, 841.4, [M
2Na–Na−MeGlcOSO
3Na−Qui)+H]
−, corroborating the notion that sulfated 3-
O-methylglucose and quinovose are terminal monosaccharides. All these data indicate that chitonoidoside F (
2) is 3
β-
O-{
β-D-quinovopyranosyl-(1→2)-(4-
O-sodium sulfate-3-
O-methyl-
β-D-glucopyranosyl-(1→3)-6-
O-sodium sulfate-
β-D-glucopyranosyl-(1→4))-
β-D-xylopyranosyl}-16-oxo-holosta-9(11),25(26)-diene.
The molecular formula of chitonoidoside G (
3) was determined to be C
66H
104O
36S
2Na
2 from the [M
2Na–Na]
− ion peak at
m/z 1559.5646 (calc. 1559.5652) and [M
2Na–2Na]
2− ion peak at
m/z 768.2895 (calc. 768.2880) in the (
−)HR-ESI-MS (
Figure S24).
Based on the absence of the signals of 18(20)-lactone at δ
C ~178 (C-18) and ~83 (C-20) in the
13C NMR spectrum of
3, the aglycone with 18(20)-ether bond instead of the lactone was supposed to be present. The NMR spectra of the aglycone part of chitonoidosides G (
3) and A [
15], where this aglycone was first found, were almost coincident with each other (
Table 3,
Figures S17–S22).
The
1H and
13C NMR spectra of the carbohydrate chain of chitonoidoside G (
3) (
Table 4,
Figures S17–S23) demonstrated six signals of anomeric protons at δ
H 4.66–5.18 (d,
J = 6.9–7.9 Hz), corresponding to the signals of the anomeric carbons at δ
C 102.2–104.8. These signals indicated the presence of a hexasaccharide moiety with
β-glycosidic bonds. The monosaccharide composition of
3 was determined as two xyloses (Xyl1 and Xyl3), quinovose (Qui2), glucose (Glc5), and two 3-
O-methylglucoses (MeGlc4 and MeGlc6). The doubled signal at δ
C 67.1 indicated the presence of two sulfate groups. Using the 1H,1H-COSY and 1D TOCSY spectra their positions were deduced as C-6 MeGlc4 and C-6 Glc5. The comparison of the
13C NMR spectra of the carbohydrate parts of the chitonoidosides G (
3) and E
1 (
1) demonstrated the closeness of the signals of the monosaccharide units from first to fifth. The signals of the terminal sixth monosaccharide residue in
3 were assigned as 3-
O-methylglucose instead of 3-
O-methylxylose in
1. The sequence of monosaccharides and the positions of the glycosidic bonds were confirmed by the correlations in the ROESY and HMBC spectra of
3 (
Table 4,
Figures S17–S23). Therefore, chitonoidoside G (
3) is the first compound in the combinatorial library of the glycosides from
P. chitonoides to feature non-sulfated 3-
O-methylglucose as terminal residue in the upper semi-chain.
The (
−)ESI-MS/MS of
3 (
Figure S24) demonstrated the fragmentation of [M
2Na–Na]
− ion at
m/z 1559.6 leading to the ion-peaks appearance at
m/z 1439.6 [M
2Na–Na−NaHSO
4]
−, 1383.6 [M
2Na–Na−MeGlc+H]
−, 1281.6 [M
2Na−Na−MeGlcOSO
3Na+H]
−, 1149.5 [M
2Na–Na−MeGlcOSO
3Na−Xyl+H]
−, 1003.5 [M
2Na–Na−MeGlcOSO
3Na−Xyl−Qui+H]
−, 533.1 [M
2Na–Na–Agl−MeGlc−MeGlcOSO
3Na−Xyl+H]
−, 387.0 [M
2Na–Na–Agl−MeGlc−MeGlcOSO
3Na−Xyl−Qui+H]
−, 255.0 [M
2Na–Na–Agl−MeGlc−MeGlcOSO
3Na−Xyl−Qui−Xyl+H]
−, confirming the sequence of monosaccharides and the aglycone structure of
3.
These data indicate that chitonoidoside G (3) is 3β-O-{6-O-sodium sulfate-3-O-methyl-β-D-glucopyranosyl-(1→3)-β-D-xylopyranosyl-(1→4)-β-D-quinovopyranosyl-(1→2)-[3-O-methyl-β-D-glucopyranosyl-(1→3)-6-O-sodium sulfate-β-D-glucopyranosyl-(1→4)]-β-D-xylopyranosyl}-16-oxo-18(20)-epoxylanosta-9(11),25(26)-diene.
The molecular formula of chitonoidoside H (
4) was determined to be C
59H
90O
32S
2Na
2 from the [M
2Na–Na]
− ion peak at
m/z 1397.4749 (calc. 1397.4760) and [M
2Na–2Na]
2− ion peak at
m/z 687.2442 (calc. 687.2434) in the (
−)HR-ESI-MS (
Figure S33).
The
1H and
13C NMR spectra of the carbohydrate chain of chitonoidoside H (
4) (
Table 5,
Figures S25–S32) demonstrated five signals of anomeric protons at δ
H 4.66–5.18 (d,
J = 6.9–8.1 Hz), corresponding to the signals of the anomeric carbons at δ
C 102.3–104.8, which indicated the presence of a pentasaccharide oligosaccharide chain with
β-glycosidic bonds between the monosaccharides. The analysis of the 1H,1H-COSY, 1D, and 2D TOCSY and HSQC spectra resulted in the assignment of the signals of two xyloses (Xyl1 and Xyl3), quinovose (Qui2), glucose (Glc4), and 3-
O-methylglucose (MeGlc5). One sulfate group in chitonoidoside H (
4) was attached to C-6 Glc4 (in the upper semi-chain), which is a typical position for chitonoidosides that feature a branch point in the sugar chain at C-4 Xyl1, with the exception of chitonoidoside B [
15], containing one sulfate group. The position of the second sulfate group was determined as C-4 MeGlc5, based on the deshielding of the signal of C-4 MeGlc5 to δ
C 76.2 (α-shifting effect of sulfate group) and the shielding of the signals C-3 and C-5 MeGlc5 to δ
C 85.2 and 76.4 (β-shifting effect of sulfate group), respectively, compared to the signals in the chitonoidoside G (
3). The sequence of the monosaccharides and the positions of the glycosidic bonds deduced by the ROESY and HMBC spectra of
4 displayed the bottom semi-chain, composed of two monosaccharide units, and the upper semi-chain, composed of three monosaccharide units, forming a chain with an uncommon architecture (
Table 5,
Figures S25–S32).
The comparison of the 13C NMR spectra of the carbohydrate parts of 4 and 2 displayed their difference only in the presence of the signals of additional xylose residue (in the bottom semi-chain of 4), as well as the glycosylation effect at C-4 Qui2, whose signal was observed at δC 85.6, instead of δC 76.2 (C-4 Qui2), observed in 2. Therefore, the chitonoidosides F (2) and H (4) can be considered as sequential steps in the biosynthesis of the glycosides in P. chitonoides.
The (
−)ESI-MS/MS of
4 (
Figure S33) demonstrated the fragmentation of [M
2Na–Na]
− ion at
m/z 1397.5 resulted in the fragmentary ion-peaks at
m/z 1277.5 [M
2Na–Na−NaHSO
4]
−, 1119.5 [M
2Na–Na−MeGlcOSO
3Na+H]
−, 1119.5 [M
2Na–Na−Xyl−Qui]
−, 987.4 [M
2Na–Na−MeGlcOSO
3Na−Xyl+H]
−, 841.4 [M
2Na–Na−MeGlcOSO
3Na−Xyl−Qui+H]
−, 519.0 [M
2Na–Na–Agl−MeGlcOSO
3Na−Xyl+H]
−, 399.0 [M
2Na–Na–Agl−MeGlcOSO
3Na−Xyl−NaHSO
4]
−, 373.0 [M
2Na–Na–Agl−Xyl−Qui−MeGlcOSO
3Na+H]
−, 255.0 [M
2Na–Na–Agl−MeGlcOSO
3Na−GlcOSO
3Na–Xyl+H]
−, confirming both the sequence of monosaccharides and the aglycone structure of
4.
These data indicate that chitonoidoside H (4) is 3β-O-{β-D-xylopyranosyl-(1→4)-β-D-quinovopyranosyl-(1→2)-[4-O-sodium sulfate-3-O-methyl-β-D-glucopyranosyl-(1→3)-6-O-sodium sulfate-β-D-glucopyranosyl-(1→4)]-β-D-xylopyranosyl}-16-oxo-holosta-9(11),25(26)-diene.