Isolation and characterization of related impurities in andrographolide sodium bisulphite injection

Abstract

Andrographolide sodium bisulphite (ASB) injection was widely used in China for the treatment of infectious diseases. The chemical analysis of ASB injection resulted in the isolation of four new related impurities (2, 3, 5, 6), together with two known compounds (1, 4). The structures of the new compounds were elucidated by 1D and 2D NMR, high-resolution mass spectrometry, Mo₂(OAc)₄-induced circular dichroism and ECD calculation. Among them, 1 and 2 were two unprecedented photocyclization derivatives of isocopalane diterpene. The generation of the primary impurity 1 was proved to be an intramolucular 6-exo carbonyl radical cyclization of ASB through a novel sulfonyl group transfer. This finding furnished a facile photocyclization methodology to afford 1 in good yield with excellent regioselectivity. The possible mechanism for the formation of the related impurities was also discussed.

---

Introduction

Andrographolide sodium bisulphite (ASB), a water-soluble sulfonate of andrographolide isolated as the major bioactive diterpenoid lactone of Andrographis paniculata Nees (Acanthaceae),¹ has been used clinically as the injection formulation (Lianbizhi®) in China for the treatment of upper respiratory tract infections, bacillary dysentery, pneumonia and acute tonsillitis.^2,3 Recent research suggested that ASB possessed a wide range of pharmacological activities such as a gastroprotective effect against indomethacin-induced gastric ulceration and competitive thiol groups for urease inhibition.^4,5

Reports of adverse drug reactions including acute renal failure caused by ASB injection have increased in the past decade, leading the issue of quality control upon production and storage of ASB as well as its injection formulation to emerge.^6–9 The potential nephrotoxicity of two kinds of ASB injection with different purities on rats has been reported. ASB injection with higher amount of unknown related substance or high dosage of the drug could increase the risk of renal damage.^10,11 However, the mechanism of renal toxicity of ASB remained unclear.

As the units of α,β-unsaturated γ-butyrolactone and two olefin bonds at C-8 (C-17) and C-13 (C-14) were susceptible to oxidization, ring cleavage and isomerization,^9,14 the active pharmaceutical ingredient ASB may undergo degradation, increasing the risk of clinical adverse events. Thus thorough knowledge of the ASB's related impurities profile were required.¹² A reversed-phase HPLC assay was establish for the determination of ASB under stress conditions.^13,14 RP-HPLC method and liquid chromatography-tandem mass spectrometry (LC-MS/MS) has been successfully used for the determination of ASB in biological samples.¹⁵ However, to the best of our knowledge, there was no report available in the literature regarding the related impurities in ASB injection.

In the analysis of ASB injection, six related impurities were observed (Fig. 1). In view of the fact that the impurity levels were above the acceptance limits of 0.1%,¹⁶ it was necessary to study the structures of related impurities. In this paper, four new related impurities (2, 3, 5, 6) and two known compounds (1, 4) including two unprecedented sulfonate derivative of isocopalane diterpene (1, 2) were purified from ASB injection and fully characterized by 1D and 2D NMR, high-resolution mass spectrometry, Mo₂(OAc)₄-induced circular dichroism and ECD calculation (Fig. 2). The possible pathways for the formation of impurities were also discussed (Scheme 1). In our continuing investigation on the origin of the primary impurity 1, previously synthesised from andrographolide (Scheme 2),¹⁷ it was found that an aqueous solution of ASB under UV irradiation (λ > 300 nm) underwent a novel sulfonate-mediated carbonyl radical cyclization to afford 1 in 67% yield with excellent regioselectivity, providing a convenient and promising eco-friendly photocyclization methodology. A plausible radical cyclization mechanism involving in the generation of 1 was also proposed (Scheme 3).

Fig. 1 HPLC-UV chromatogram of ASB injection.

Fig. 2 Chemical structures of ASB and the related impurities.

Scheme 1 Proposed formation pathways of related impurities isolated from ASB injection.

Scheme 2 Previous ring-closing reaction of ASB to afford 1.

Scheme 3 Plausible mechanism of photochemical cyclization of ASB.

Results and discussion

Structure elucidation

Compound 2 was obtained as colorless needle crystal. The molecular formula was deduced as C₂₀H₃₁O₉SNa by HRESIMS data at m/z 429.1582 [M–H₂O–Na]⁻, (calcd 429.1583). The ¹H NMR spectrum of 2 showed two methyl groups [δ_H 1.06 (3H, s, H-18), 1.13 (3H, s, H-20)]. The ¹³C NMR spectrum displayed 20 carbon resonances, which were categorized by HSQC experiments into two methyls, eight methylenes, five methines and five quaternary carbons. The NMR data of 2 (Tables 1 and 2) were similar to those of 1,¹⁷ clearly indicating the presence of the same isocopalane diterpene skeleton in 2. The difference was a sec/tert diols unit at [δ_C 81.6 (C-12), 73.8 (C-13)] in 2 instead of double bonds at [δ_C 137.4 (C-12), 129.1 (C-13)] in 1. This was further demonstrated by HMBC correlations from H-12 (δ_H 4.77) to C-9 (δ_C 48.9), C-13 (δ_C 73.8), C-14 (δ_C 54.7) and C-16 (δ_C 176.8); and H-15b (δ_H 4.53) to C-13 (δ_C 73.8) and C-16 (δ_C 176.8) (Fig. 3). On the other hand, the correlations of H-15a (δ_H 4.27)/C-8 (δ_C 39.2); H-9 (1.65)/C-17 (δ_C 56.9); and H-17 (δ_H 3.76 and 3.95)/C-14 (δ_C 54.7) showed that the ring formation occurred at C-8 to C-14 and the sulfonate group was linked at C-17. All of the proton and carbon signals were assigned unambiguously from the ¹H–¹H COSY, HSQC and HMBC spectra (Tables 1 and 2). In the ROESY spectrum (Fig. 3), the cross-peaks of 18-CH₃ (δ_H 1.06)/H-3 (δ_H 3.18), H-5 (δ_H 0.93); and H-9 (δ_H 1.65)/H-5 (δ_H 0.93), H-14 (δ_H 2.53) indicated that H-3, H-5, H-9, H-14 and 18-CH₃ were β-orientation. The α-orientation of H-17 was also confirmed by the ROESY correlation of 20-CH₃ (δ_H 1.13)/H-17a (δ_H 3.76), H-17b (δ_H 3.95).

Table 1 ¹H NMR data of compounds 1–6^a (δ in ppm, J in Hz)

No.	1^b	2^c	3^b	4^b	5^b	6^b
a Assignments were made by HSQC, HMBC, ^1H–^1HCOSY and ROESY experiments.b Measured in CD3OD at 300 MHz.c Measured in DMSO-d6 at 300 MHz.
1a	1.03, m	0.97, dd (4.6, 13.8)	1.12–1.21, m	1.46–1.54, m	1.02–1.10, m	1.22–1.35, m
1b	1.70, m	1.56, d (10.1)	1.93–2.01, m	1.87, dt (12.8, 3.2)	1.92, dt (12.9)	1.79–1.86, m
2a	1.71, m	1.52–1.71, m	1.67–1.78, m	1.70–1.86, m	1.70–1.81, m	1.73–1.82, m
2b	1.81, m		1.90–2.02, m
3	3.36, dd (2.5, 6.9)	3.18, m	3.37, dd (5.1, 11.0)	3.41, t (8.5)	3.30, m	3.40, t (8.3)
5	1.07, d (8.2)	0.93, t (4.4)	1.32–1.35, m	1.31, m	1.10–1.17, m	1.30, m
6a	1.66, d (8.4)	1.42–1.48, m	1.85–1.91, m	1.32, m	1.13–1.17, m	1.32, m
6b	1.95, ddd (1.7, 7.9, 16.0)			1.79–1.87, m	1.37–1.53, m	1.78–1.82, m
7a	1.23, dt (2.4, 8.1)	1,21–1.32, m	1.83–1.92, m	1.94–2.05, m	1.95–2.05, m	2.01, m
7b	2.94, d (8.3)	1.88, d (13.3)	2.36–2.41, m	2.38, dt (11.5, 3.7)		2.35, d (11.8)
9	1.41, dd (3.1, 7.1)	1.65, m	1.50, d (11.7)	2.29, d (9.6)		1.81, m
11a	2.13, m	2.05, m	2.16, m	1.73–1.81, m	2.75, dd (11.7, 14.6)	1.99, d (13.7)
11b	2.38, ddd (2.6, 4.9, 12.1)	2.75, dd (10.7, 16.8)	2.37, m	2.47, ddd (2.5, 8.2, 14.6)	3.05, d (14.3)	2.25, t (13.6)
12	6.84, dd (2.0, 4.0)	4.77, d (5.4)	3.97, dd (1.7, 12.0)	3.80, dd (4.5, 8.3)	4.08, dd (2.8, 11.6)	4.26, d (12.2)
14	2.99, m	2.53	7.69, s	7.64, s	7.73, d (1.9)
15a	4.41, t (5.7)	4.27, dd (9.0, 10.0)	4.91, s	4.86, d (1.5)	4.88, d (1.4)
15b	4.91, t (5.7)	4.53, t (10.4)
17a	2.81, d (9.4)	3.76, d (15.3)	4.93, s	4.42, s	1.67, s	4.89, s
17b	3.26, d (9.4)	3.95, d (15.7)		4.79
18	1.21, s	1.06, s	1.16, s	1.20, s	1.18, s	1.19, s
19a	3.40 (1H, d, 6.7)	3.31, d (11.1)	9.95, s	3.33, m	3.35, d (11.3)	3.32, d, m
19b	4.13 (1H, d, 6.7)	3.80, d (11.0)		4.11, d (11.0)	4.13,d (11.1)	4.11, d (11.1)
20	1.06 (3H, s)	1.13, s	0.65, s	0.67, s	0.96, s	0.71, s

---

Table 2 ¹³C NMR data of compounds 1–6^a (δ in ppm)

No.	1^b	2^c	3^b	4^b	5^b	6^b
a Assignments were made by HSQC, HMBC, ^1H–^1HCOSY and ROESY experiments.b Measured in CD3OD at 75 MHz.c Measured in DMSO-d6 at 75 MHz.d Missing signals.
1	38.9	37.3	37.6	37.6	37.4	37.8
2	28.1	26.8	29.1	29.0	28.9	29.1
3	80.9	78.1	77.3	81.1	81.0	80.9
4	43.5	42.0	54.5	43.7	43.7	43.7
5	57.1	55.3	57.1	56.5	53.3	56.0
6	20.8	18.3	25.4	25.3	20.1	25.4
7	37.0	42.6	38.9	39.2	35.2	39.3
8	38.2	39.2	148.0	149.6	130.6	149.0
9	56.7	48.9	53.1	55.6	137.4	53.7
10	38.3	36.2	40.3	40.3	39.0	40.0
11	24.6	19.7	27.5	28.4	30.5	26.9
12	137.4	81.6	55.9	56.7	57.6	62.2
13	129.1	73.8	131.8	133.2	131.6	130.6
14	54.3	54.7	151.1	150.6	151.5	145.1
15	70.7	65.8	72.4	72.3	72.3	174.7
16	172.5	176.8	—^d	176.0	176.4	174.5
17	50.4	56.9	108.9	107.7	21.0	108.1
18	23.3	22.6	21.1	23.4	23.2	23.4
19	65.0	62.8	208.7	65.1	65.0	65.1
20	15.6	15.1	15.4	15.5	22.6	15.8

---

Fig. 3 ¹H–¹H COSY, key HMBC and ROESY correlations of 2.

The absolute configuration of the 12,13-diol moiety was determined by the in situ dimolybdenum CD method, developed by Snatzke and Frelek.^18–20 According to Snatzke's empirical rule, band IV (310 nm) showing the same sign with the O–C–C–O dihedral angle in the favored conformation allowed the assignment of the absolute configuration. After the inherent CD contribution was subtracted, the metal complexes of 2 in DMSO gave a significant induced CD spectrum (ICD) with the negative Cotton effect at 310 nm (Fig. 4), permitting the assignment of 12S,13S for 2. Therefore, compound 2 was established as sodium 14-deoxy-12(S),13(S)-dihydroxy-8,14-cycloandrographolide-17-sulfonate.

Fig. 4 CD spectra of in situ formed Mo-complexes of sec/tert diols 2 in DMSO.

Compound 3 was obtained as white powder. Its molecular formula was determined as C₂₀H₂₇O₇SNa by HRESIMS [M–Na]⁻ at m/z 411.1483 (calcd 411.1483). The NMR data of 3 (Tables 1 and 2) were almost identical to those of 4,²¹ an C-12 epimer of ASB, except for the 19-CHO function at (δ_C 208.7/δ_H 9.95) instead of the 19-CH₂OH (δ_C 65.1/δ_H 3.33 and 4.11) in 4. These attachments were corroborated by the chemical shift of C-4 dramatically down from 43.7 ppm to 54.5 ppm (+10.8 ppm) while C-3 up from 81.1 ppm to 77.3 ppm (−3.8 ppm) in 3. These shift results indicated that 3 was the product of 4 after dehydrogenation at C-19, which was further confirmed by its HRESIMS data. In addition, the HMBC correlations of 20-CH₃ (δ_H 0.65)/C-1 (δ_C 37.6), C-5 (δ_C 57.1), C-9 (δ_C 53.1); 18-CH₃ (δ_H 1.16)/C-3 (δ_C 77.3), C-5 (δ_C 57.1); and H-17 (δ_H 4.93)/C-7 (δ_C 38.9), C-9 (δ_C 53.1) could be observed. The HMBC correlation from H-14 (δ_H 7.69) to C-12 (δ_C 55.9) and C-15 (δ_C 72.4) indicated that the sulfonate group was linked at C-12. ROESY cross-peaks of 18-CH₃ (δ_H 1.16)/H-3 (δ_H 3.37) and H-5 (δ_H 1.32-1.35); H-9 (δ_H 1.50)/H-5 (δ_H 1.32-1.35) and H-14 (δ_H 7.69); and 20-CH₃ (δ_H 0.65)/19-CHO (δ_H 9.95) revealed that H-3, H-5, H-9, H-14 and 18-CH₃ were β-orientation, while 19-CHO and 20-CH₃ were α-orientation.

To determine the absolute configuration of C-12, the CD spectrum of 3 was recorded in MeOH, which exhibited Cotton effects at λ_max 219 nm (Δε +3.7). Then the ECD spectrum was calculated using TDDFT method at the B3LYP/6-31G(d) level.²² The calculated ECD of H-12α-orientation matched the experimental result very well (Fig. 5), allowing the assignment of C-12 as S. Thus, the structure of 3 was determined as sodium 19-dehydro-14-deoxy andrographolide-12(S)-sulfonate.

Fig. 5 Calculated and experimental CD spectra of 3, 5 and 6.

Compound 5, was isolated as white amorphous powder. It gave the molecular formula of C₂₀H₂₉O₇SNa, determined by HRESIMS [M–Na]⁻ at m/z 413.1639 (calcd 413.1640). The NMR data of 5 (Tables 1 and 2) were similar to those of 4 expect for the presence of an additional 17-CH₃ (δ_C 21.0/δ_H 1.67) and carbon–carbon double bond unit [δ_C 130.6 (C-8), 137.4 (C-9)] in 5, which implied rearrangement of the terpene skeleton that the exocyclic double bond Δ^8,17 in 4 was replaced by an double bond Δ^8,9 in 5. This was further confirmed by the chemical shift of C-11 (δ_C 30.5) and C-17 (δ_C 21.0) down to lower field coupled with the HMBC interactions from 17-CH₃ (δ_H 1.67) to C-7 (δ_C 35.2), C-12 (δ_C 57.6) and C-20 (δ_C 22.6). The sulfonate group substituted at C-12 was deduced from the HMBC cross-peaks from δ_H 7.73 (H-14) to C-12 (δ_C 57.6) and C-15 (δ_C 72.3). The relative configuration of 5 was consistent with that of 4 by the ROESY experiment. The calculated ECD curve for 12R-enantiomer excellently agreed with the experimental ECD curve of 5 (Fig. 5). Compound 5 was therefore characterized as sodium 14-deoxy-8,9-didehydro andrographolide-12(R)-sulfonate.

Compound 6 was obtained as a white powder. The HRESIMS of 6 exhibited a quasi-molecular ion [M–Na]⁻ at m/z 445.1537 (calcd 445.1537) corresponding to the molecular formula of C₂₀H₂₉O₉SNa. The ¹H and ¹³C NMR data indicated that 6 was a labdane-type bicyclic skeleton with an exocyclic 17-CH₂ group (δ_C 108.1/δ_H 4.89), a 18-Me group (δ_C 23.4/δ_H 1.19), a 19-CH₂OH group (δ_C 65.1/δ_H 3.32 and 4.11) and an angular 20-Me group (δ_C 15.8/δ_H 0.71),²³ structurally similar with those of 4 and 5. However, the absence of the lactone C=O signal and the presence of two COOH signals at δ_C 174.5 and 174.7 implied that the α,β-unsaturated γ-lactone ring had been opened and transformed to COOH groups in positions 15 and 16, respectively. With the aid of ¹H–¹H COSY, HSQC and HMBC spectra, the full assignments of the signals were summarized in Tables 1 and 2. The relative configurations at C-3, C-5, C-9, C-14, C-18, C-19 and C-20 of 6 were identical to those of 4 and 5 by the analysis of the ROESY spectra.

To assign the absolute configuration of C-12, the ECD spectra of two possible isomers were calculated respectively. Comparison of the experimental and calculated ECD curves (Fig. 5) for 6 allowed the assignment of C-12 as R-enantiomer. Thus the structure of 6 was determined as sodium 14-deoxy andrographolide-15,16-dioic acid-12(R)-sulfonate.

The proposed formation pathways of impurities isolated from ASB injection were depicted in Scheme 1. ASB underwent the epimerization of C-12 configuration to give 4, followed by isomerization of exocyclic double bond to furnish 5. 3 was produced as the regioselective primary alcohol oxidation product of ASB. As expected, photoinduced intromolecular cyclization of ASB led to 1. Then 1 subsequently underwent epoxidation and hydrolysis to afford trans-diols product 2. The formation of 6 was supposed to include the water-assisted hydrolytic cleavage of α,β-unsaturated lactone²³ followed by the oxidation of 4.

Intramolecular cyclization of ASB

A typical HPLC chromatogram of ASB injection was shown in Fig. 1, revealing one major related impurity 1 in significantly larger amounts about 2%. To gain sight into the nature of the impurity 1, forced degradation studies of ASB were carried out under thermal hydrolysis (60 °C for 10 days), acidic hydrolysis (0.1 N HCl for 5 h), basic hydrolysis (0.1 N NaOH for 2 h), oxidation (30% H₂O₂ for 5 h) and photolysis (UV light, 300 mW cm⁻², 10 days). The results showed that only photolysis led to the formation of 1 in small amount. Then an aqueous solution of ASB was irradiated using a 125 W high-pressure mercury lamp (λ > 300 nm) in a quartz test tube at room temperature to afford the major cyclized product 1 within 5 h in 67% yield with excellent regioselectivity. This finding attracted us to investigate the mechanism of this photocyclization reaction. A previous synthetic method to obtain 1 was performed using K₂O₂S₈ at 35–40 °C for 21 hours.¹⁷ However, the overall utility of reaction might be limited by the toxicity of the reagent, longer reaction time and the remaining unclear reaction mechanism.

In this study, the solvent had a key impact on the reaction outcome. Specifically, no cyclized products were observed in nonaqueous solvent (methanol, acetone, acetonitrile, and DMF), whereas good yields were achieved in aqueous solution. No reaction was observed in the absence of UV irradiation. Furthermore, experiments were performed in aqueous solution by using 2,2,6,6-tetramethylpiperidinooxy (TEMPO, 0.5 equivalent) as radical scavenger.^24,25 The cyclization reaction was terminated immediately, which demonstrated that the intramolecular cyclization underwent a radical process. These results indicated that 1 was produced via radical cyclization pathway induced by UV light, wherein the presence of water solution played a critical role in accelerating the sulfonyl group transfer.

Thus, a plausible mechanism of the photoinduced radical cyclization reaction was displayed in Scheme 3. Under UV irradiation, carbon–sulfur bond of ASB was homolytically cleaved into carbon radical A and sulfonyl anion radical (SO₃⁻˙). Once the allyl radical rearrangement occurred, the 6-exo attack resulted in the six-membered-ring radical intermediate B. A strong NOE observed between 20-CH₃ and 17-CH₂ suggested that 17-CH₂ residue was in α-orientation. Subsequently, group transfer would be accomplished when radical B abstracted the sulfonyl group from another substrate ASB to give the 6-exo-selective radical cyclization product C.

Conclusion

Four new related impurities (2, 3, 5, 6) and two known compounds (1, 4) including two unprecedented sulfonate derivative of isocopalane diterpene (1, 2) were isolated from ASB injection and fully characterized by 1D and 2D NMR, high-resolution mass spectrometry, Mo₂(OAc)₄-induced circular dichroism and ECD calculation. The possible pathways for the formation of related impurities were discussed in detail. This study may be instructive for quality control upon production and storage of ASB as well as its injection formulation. Furthermore, the generation of the primary impurity 1 was proved to be a novel sulfonate-mediated intramolecular radical cyclization. This finding provided a facile phytocyclization access to 1 in good yield with excellent regioselectivity, which integrated radical cyclization and group transfer and was expected to be more attractive and practical for the synthesis of polycyclic natural products. Further mechanism and synthetic application are currently under investigation in our laboratory.

Experimental section

General experimental procedures

¹H, ¹³C NMR, ¹H–¹H COSY, HSQC, HMBC, ROESY spectra were measured on a Bruker Avance 300 (300 MHz for ¹H and 75 MHz for ¹³C NMR) spectrometer with TMS as internal standard. Chemical shifts were recorded as parts per million (ppm) with the solvent signal used as a standard (DMSO-d₆, δ_H 2.49, δ_C 39.5; CD₃OD, δ_H 4.73, δ_C 49.0). Coupling constants were given in hertz. Low and high resolution ESIMS were performed on Agilent 1100 series LC/MSD-Trap/SL and Agilent 6210 Q-TOF Premier™, respectively. The CD spectra were obtained on a JASCO J-810 spectropolarimeter. Photochemistry was carried out in a quartz test tube using 125 W high-pressure mercury lamp. Column chromatography was performed with Silica gel (200–300 mesh, Qingdao Marine Chemical Factory, P. R China), Sephadex LH-20 (25–100 mm, GE Healthcare, Sweden) and YMC-Pack ODS-AQ (S-50 μm, 12 nm, YMC, Japan). TLC was performed on precoated HSGF₂₅₄ plates (0.25 mm thick, Yinlong, China). Spots were detected on TLC under UV light or by heating after spraying with 5% H₂SO₄ in EtOH (v/v). Analytical HPLC was performed on an Agilent ZORBAX SB-C18 column (250 × 4.6 mm, 5 μm) at 30 °C. Mobile phase consisted of 0.02 mol L⁻¹ KH₂PO₄–500 μl L⁻¹ H₃PO₄ aqueous solution (A) and methanol (B) using isocratic elution in the ratio of 55 : 45 (v/v). The flow rate was 1 mL min⁻¹, the injection volume was 10 μL, and the detection was monitored at 220 nm.

Material

The ASB injection was manufactured by Wuxi Jiminkexin Shanhe pharmaceutical enterprise Co., Ltd China. All related impurities were isolated from ASB injection by column chromatography in our laboratory.

Extraction and isolation

The concentrated ASB injection (200 g) was recrystallized with 95% EtOH–H₂O (v/v, 95 : 5) repeatedly. Then the mother solution was evaporated under reduced pressure to obtain an extract (80 g), which was subjected to silica gel column chromatography (100–200 mesh) eluting with a gradient mixture CHCl₃/MeOH/H₂O (20 : 8 : 0.1 → 8 : 8 : 0.1) to yield six fractions (Fr. 1–6). Fr. 3 (9.3 g) was separated on a silica gel column (200–300 mesh) using a gradient of CHCl₃/MeOH/H₂O (20 : 10 : 0.1 → 10 : 8: 0.1) to give eight subfractions (Fr. 3.1–3.8). Fr. 3.6 was further subjected to CC over macroporous resin MCI chp20p (MeOH/H₂O, 60 : 40 → 90 : 10) to obtain five subfractions (Fr. 3.6.1–3.6.5). Fr. 3.6.2 was applied onto Sephadex LH-20 (MeOH–H₂O, 70 : 30) to afford 2 (10.5 mg), 4 (16.2 mg). Fr. 3.6.4 was then run on a RP-C₁₈ column using a step gradient of MeOH–H₂O (60 : 40 → 90 : 10), followed by Sephadex LH-20 (MeOH–H₂O, 50 : 50) to yield 5 (5.1 mg), 6 (9.7 mg). Fr. 5 (7.5 g) was subjected to CC over silica gel column using CHCl₃/MeOH/H₂O (20 : 10 : 0.1 → 10 : 8: : 0.1) to afford five subfractions (Fr. 5.1–5.5). Fr. 5.3 was further separated by RP-C₁₈ column (MeOH/H₂O, 60 : 40 → 90 : 10) to obtain 3 (6.5 mg) and 1 (52.3 mg).

Sodium 14-deoxy-12,13-didehydro-8,14-cycloandrographolide-17-sulfonate (1). Colorless needle crystal; [α]²⁵_D +6.0 (c 0.1, MeOH); UV (MeOH) λ_max (log ε) 225 (3.71) nm; IR (KBr) ν_max 3408, 2949, 2861, 1726, 1692, 1338, 1207, 1042, 990, 626 cm⁻¹; ¹H NMR (CD₃OD, 300 MHz) see Table 1; ¹³C NMR (CD₃OD, 75 MHz) see Table 2; HRESIMS: m/z 413.1638 [MNa]⁻ (calcd 413.1640).

Sodium 14-deoxy-12(S),13(S)-dihydroxy-8,14-cycloandrographolide-17-sulfonate (2). Colorless cluster crystals; [α]²⁵_D +41.8 (c 0.08, MeOH); IR (KBr) ν_max 3452, 2924, 1762, 1640, 1384, 1156, 1028, 978 cm⁻¹; ¹H NMR (DMSO-d₆, 300 MHz) see Table 1; ¹³C NMR (DMSO-d₆, 75 MHz) see Table 2; HRESIMS: m/z 429.1582 [M–2H₂O–Na]⁻ (calcd 429.1583).

Sodium 19-dehydro-14-deoxy andrographolide-12(S)-sulfonate (3). White amorphous powder; [α]²⁵_D −16.3 (c 0.1, MeOH); UV (MeOH) λ_max (log ε) 202 (3.95) nm; CD (MeOH): 219 nm (Δε = +3.7); IR (KBr) ν_max 3433, 2920, 2850, 1744, 1637, 1469, 1204, 1042, 890, 628 cm⁻¹; ¹H NMR (CD₃OD, 300 MHz) see Table 1; ¹³C NMR (CD₃OD, 75 MHz) see Table 2; HRESIMS: m/z 411.1483 [M–Na]⁻ (calcd 411.1483).

Sodium 14-deoxy andrographolide-12(R)-sulfonate (4). White amorphous powder; [α]²⁵_D −34.8 (c 0.1, MeOH); UV (MeOH) λ_max (log ε) 202 (4.01) nm; IR (KBr) ν_max 3434, 2932, 2852, 1741, 1641, 1448, 1194, 1041, 891, 726 cm⁻¹; ¹H NMR (CD₃OD, 300 MHz) see Table 1; ¹³C NMR (CD₃OD, 75 MHz) see Table 2; HRESIMS: m/z 413.1639 [M–Na]⁻ (calcd 413.1640).

Sodium 14-deoxy-8,9-didehydro andrographolide-12(R)-sulfonate (5). White amorphous powder; [α]²⁵_D −18.8 (c 0.1, MeOH); UV (MeOH) λ_max (log ε) 202 (3.99) nm; CD (MeOH): 214 nm (Δε = −7.4); IR (KBr) ν_max 3423, 2938, 1758, 1644, 1444, 1352, 1201, 1034, 893, 625 cm⁻¹; ¹H NMR (CD₃OD, 300 MHz) see Table 1; ¹³C NMR (CD₃OD, 75 MHz) see Table 2; HRESIMS: m/z 413.1639 [M–Na]⁻ (calcd 413.1640).

Sodium 14-deoxy andrographolide-15,16-dioic acid-12(R)-sulfonate (6). White amorphous powder; [α]²⁵_D −4.4 (c 0.1, MeOH); UV (MeOH) λ_max (log ε) 201 (3.90), 224sh (3.75) nm; CD (MeOH): 209 nm (Δε = −6.8), 241 nm (Δε = +4.0); IR (KBr) ν_max 3442, 2944, 2865, 1726, 1638, 1575, 1403, 1207, 1042, 919, 636 cm⁻¹; ¹H NMR (CD₃OD, 300 MHz) see Table 1; ¹³C NMR (CD₃OD, 75 MHz) see Table 2; HRESIMS: m/z 445.1537 [M–Na]⁻ (calcd for C₂₀H₂₉O₉S, 445.1537).

Acknowledgements

The authors gratefully acknowledge the financial support by the National Natural Science Foundation of China (81573309, 81172955), the Major National Science and Technology Projects of the Chinese twelve five-year Plan (2013ZX09301303-003), and “Qinlan Project” Plan of Jiangsu Province 2012.

Notes and references

1. Z. M. Meng, Acta Pharmacol. Sin., 1981, 16, 571–575.
2. C. Y. Shi and L. Y. Dong, Chinese Journal of Ethnomedicine and Ethnopharmacy, 2009, 12, 5–6.
3. J. X. Pan, K. P. Zhang and S. X. Gao, China Practical Medical, 2007, 2, 70–73.
4. Y. H. Liu, Z. B. Zhang, Y. F. Zheng, H. M. Chen, X. T. Yu, X. Y. Chen, X. Zhang, J. H. Xie, Z. Q. Su, X. X. Feng, H. F. Zeng and Z. R. Su, Int. Immunopharmacol., 2015, 26, 384–391.
5. Z. Z. Mo, X. F. Wang, X. Zhang, J. Y. Su, H. M. Chen, Y. H. Liu, Z. B. Zhang, J. H. Xie and Z. R. Su, BMC Complementary Altern. Med., 2015, 15, 238.
6. J. W. Zhao, Z. H. Ni, L. O. Cao, S. Mou, W. Fang and Q. Y. Zhang, Chin. J. Integr. Tradit. Western Nephrol., 2005, 6, 529–531.
7. J. Sun and K. Sun, Chin. J. Pharmacoepidemiol., 2007, 16, 77–79.
8. W. B. Yang, X. W. Wang, J. Y. Yu, Z. W. Jin, Z. Wang, J. L. Cao and F. Jiang, Chinese Journal of Pharmacovigilance, 2013, 10, 46–50.
9. Y. C. Jiang, Asia-Pacific Traditional Medicine, 2014, 10, 109–110.
10. (a) Z. H. Hu, C. Q. Wu, Q. J. Wang, X. Q. Wang, Y. W. Luo, B. H. Yang and M. Y. Liao, J. Toxicol., 2010, 24, 433–437; (b) Z. H. Hu, C. Q. Wu, Q. J. Wang, Q. X. Wang, Y. W. Luo, B. H. Yang and M. Y. Liao, Adverse Drug React. J., 2010, 12, 10–16.
11. (a) H. Lu, X. Y. Zhang, Y. Q. Zhou, X. Wen and L. Y. Zhu, Acta Pharmacol. Sin., 2011, 32, 888–894; (b) H. Lu, X. Y. Zhang, Y. Q. Zhou and S. S. Jin, Chin. J. Pharmacol. Toxicol., 2010, 24, 223–227.
12. S. W. Baertschi, K. M. Alsante and R. A. Reed, in Pharmaceutical stress testing: predicting drug degradation, Informa Healthcare, New York, 2nd edn, 2011, ch. 2, pp. 10–49.
13. M. Z. Jin, X. X. Pang, J. Yi, B. H. Guo, P. L. Tang and S. Y. Zhang, Chin. J. New Drugs, 2005, 14, 890–892.
14. X. Q. Hu, X. Y. Ba, F. Hao, Y. Wu and X. G. Zhou, Drugs Clin., 2013, 28, 523–527.
15. (a) S. Q. Zhang, X. H. Chen, M. Yu, X. Sun and Z. G. Li, J. Chromatogr. B., 2012, 880, 163–167; (b) S. Q. Zhang and Y. M. Fan, J. Chromatogr. B., 2012, 907, 173–177.
16. ICH Guideline Q3A (R), Impurities in New Drug Substances, February 7, 2002.
17. X. Xiao, Z. Y. Xu, Q. D. Zeng, X. B. Chen, W. H. Ji, Y. Han, P. Y. Wu, J. M. Ren and B. B. Zeng, Chem.–Eur. J., 2015, 21, 8351–8354.
18. M. Gorecki, E. Jablonska, A. Kruszewska, A. Suszczynska, Z. Urbanczyk-Lipkowska, M. Gerards, J. W. Morzycki, W. J. Szczepek and J. Frelek, J. Org. Chem., 2007, 72, 2906–2916.
19. G. Snatzke, Angew. Chem., Int. Ed. Engl., 1979, 18, 363–366.
20. L. D. Bari, G. Pescitelli, C. Pratelli, D. Pini and P. Salvadori, J. Org. Chem., 2001, 66, 4819–4825.
21. X. J. He, J. K. Li, H. Gao, F. Qiu, K. Hu, X. M. Cui and X. S. Yao, Tetrahedron, 2003, 59, 6603–6607.
22. C. Wang, C. J. Li, J. Ma, J. Z. Yang, X. G. Chen, L. Li and D. M. Zhang, RSC Adv., 2015, 5, 30046.
23. L. X. Chen, F. Qiu, H. Wei, G. X. Qu and X. S. Yao, Helv. Chim. Acta, 2006, 89, 2654–2664.
24. M. E. Van Loo, J. Lugtenburg and J. Cornelisse, Eur. J. Org. Chem., 2000, 2000, 713–721.
25. P. Bruni, P. Carloni, C. Conti, E. Giorgini, L. Greci, M. Iacussi, P. Stipa and G. Tosi, Tetrahedron, 1996, 52, 6795–6802.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra28178d

---