A simple method for separating sanguinarine and chelerythrine based on preparative acid-base titration

Introduction

The benzophenanthridine alkaloids sanguinarine (SA) and chelerythrine (CHE), produced by plants of the families Papaveraceae, Rutaceae, and Fumariaceae, are of considerable interest to experimental and clinical medicine due to their pronounced antimicrobial [1-3], anti-inflammatory [4, 5], anticancer [6-12] and other pharmacological properties. Although they possess structural similarity and similar spectra of pharmacological effects, they still differ both in the strength of these effects and, in some cases, in their mechanisms. For example, SA exhibits higher cytotoxic and antiproliferative activity against a number of cancer cell lines [10-12], and higher anticholinesterase activity [13]. SA also demonstrates a stronger anti-inflammatory effect, while CHE has a stronger analgesic effect [5]. Therefore, each of these alkaloids is of interest individually. However, in plants they are present together, and their separation is a difficult task.

Column liquid chromatography is a conventional method for the preparative separation of SA and CHE [14, 15]. This approach is common for the separation of alkaloids, but it is associated with the consumption of significant amounts of solvents, which are often toxic. Very small quantities of individual alkaloids can be obtained by preparative thin-layer chromatography [16]. Countercurrent chromatography is a promising method [17-19], but it requires the use of expensive equipment.

It is known that in solution, benzophenanthridine alkaloids can exist in two forms: cationic (iminium) and pseudo-base (alkanolamine). The ratio between them depends on the pH of the medium [20]. A similar equilibrium between the two forms was discovered in biological media [21].

The aim of this study was to find conditions for the preparative separation of SA and CHE based on the difference in their acid-base behavior, and to develop a method for obtaining individual alkaloids without the use of chromatographic techniques.

Materials and methods

Raw material. A mixture of SA and CHE in the form of bisulfates, used as a raw material, was obtained from Macleaya microcarpa leaves using the method previously developed and described [22].

Apparatus. The separation of alkaloids was performed in a laboratory-made prototype of a preparative titration apparatus, consisting of a thermostated glass reaction vial of 50 ml volume, mounted on a magnetic stirrer, a peristaltic pump with a step volume of 20 ml for titrant delivery, and a pH meter. The turbidity of the reaction liquid was measured as light scattering index (LSI), using a QRD1114 reflective object sensor (QT Optoelectronics), mounted on the outer surface of the reaction vial. Since it was not possible to calibrate this sensor against turbidity standards, the LSI was expressed as the ratio of the photocurrent to light source current. All the modules of the apparatus were controlled by a Raspberry Pi 3B computer (Raspberry Pi Ltd.), using an application developed in Lazarus IDE v. 2.2.6 for Linux.

Analysis of the raw material and the fractionated samples was performed using an Agilent 1260 Infinity liquid chromatograph (Agilent Technologies) with a diode-array detector.

To study the equilibrium between the two forms of the alkaloids in buffer solutions, a Lambda 25 (Perkin Elmer) spectrophotometer was used.

Chemicals. Analytical reference standards of SA chloride and CHE chloride, as well as the solvents and reagents of analytical grade used in the study, were purchased from Sigma-Aldrich (USA), Merck, Fluka and Stanchem (Germany).

Methodology. The following technological procedures were developed and used in this study:

Fractionation of the raw material: 2.5 g of the alkaloid mixture in the form of bisulfates was suspended in 25 ml of a mixture of acetonitrile (ACN) and water (60:40) and titrated at 45°C under vigorous stirring with titrant 1 (0.5 M K₂HPO₄ + 0.5 M Na₂HPO₄) at a rate of 30 ml/h until the solid phase was completely dissolved and a pH of 5.5-5.7 was reached; the titrant feed rate was then reduced to 1 ml/h. The titrant feed was stopped immediately before the relative CHE content in the precipitate began to increase. The precipitate of the SA fraction was separated on a glass filter with a pore size of 40 μm, washed with 50% ACN and dried at 40°C. The filtrate was returned to the reaction vial and titrated with the same titrant at 35°C and a rate of 1.5 ml/h until the relative SA content in the liquid phase ceased to decrease. The precipitate of the intermediate fraction, containing both alkaloids in comparable amounts, was separated on a glass filter (40 μm), washed with 50% ACN and dried at 40°C. The filtrate was titrated with titrant 2 (1 M K₂HPO₄ + 1 M NaOH) at 45°C and a rate of 6 ml/h until pH 8.4-8.6 was reached. The precipitate of the CHE fraction was separated on a filter (40 μm), washed with 50% ACN and dried at 40°C.

To establish the optimal volumes of titrant 1 for separating the SA and intermediate fractions, a pilot experiment was conducted with each batch of the initial alkaloid mixture at a constant temperature of 45°C, without stopping the titrant feed or separating the fractions. During the titration, 40 µl samples of the suspension were periodically collected for analysis. The samples were filtered through 20 µl pipette tips with polypropylene filters by purging the tips with air and collecting the drained liquid. The tips were washed with 100 µl of 25% ACN, and the filtrate was diluted to a volume of 10 ml with a mobile phase for HPLC analysis consisting of ACN and 0.05% trifluoroacetic acid solution (32:68). The precipitate on the filter was dissolved in 20 μl of acetic acid, the filter was washed twice with 100 μl of 25% ACN, and the liquid was also diluted to 10 ml volume with the mobile phase. The alkaloid concentrations in the samples prepared in this way were determined by the HPLC method described in [22]. The purity of the SA fraction was calculated as the ratio of the SA content to the total alkaloid content in the precipitate, and the SA recovery (at the titration stage) was calculated as the ratio of its content in the precipitate to the total SA content in the sample. For the CHE fraction, the purity was calculated as the ratio of the CHE content to the total alkaloid content in the supernatant, and the CHE recovery was calculated as the ratio of its content in the supernatant to the total CHE content in the sample.

To confirm the scalability of the proposed method, an experiment was conducted with a fivefold increase in the raw material load. The titrant volumes and feed rates were proportionally increased, and the titration was performed in a 500 ml vial.

The reverse conversion of alkaloid pseudo-bases to bisulfates for all fractions was performed in the following ways:

Method 1. A portion of 2.0 g of the pseudo-base was suspended in 40 ml of 0.2 M H₂SO₄ (approximately 35% excess), boiled under reflux with constant stirring until the solid phase was completely dissolved (usually 20-30 min), then slowly cooled to 4°C. The precipitate of the target product was separated on a filter (100 μm), washed with 96% ethanol and dried at 40°C.

Method 2. A portion of 2.0 g of the pseudo-base was mixed with 2.7 ml of 40% acetic acid, and the mixture was heated, stirring occasionally, for 1 h at 60°C for SA or 80°C for other fractions. The still-warm solution was diluted with 40 ml of water, kept for 3-4 hours at 4°C, and centrifuged at 1500 g. The supernatant was heated almost to boiling, 0.9 ml of 9 M H₂SO₄ was added, and then slowly cooled to 4°C. The precipitated target product was filtered (100 μm), washed with 96% ethanol, and dried at 40°C.

Repeated fractionation was performed to process the intermediate fraction and, if necessary, to increase the target alkaloid content in the products. For this purpose, the fractions obtained were used either in the form of bisulfates and as acetic acid solutions of the pseudo-bases. In the latter case, the pseudo-base solution in acetic acid, obtained according to Method 2 (see above), was diluted with 7 ml of water and centrifuged. 15 ml of ACN was added to the supernatant, and the mixture was titrated with titrant 2 at a rate of 1.5 ml/h and a temperature of 45°C.

Repeated fractionation yielded only two fractions: the target fraction (SA or CHE) and an additional fraction, the composition of which corresponded to the intermediate fraction from the primary fractionation. When fractionating the intermediate fraction, the target was the CHE fraction, and when processing the SA fraction, the additional fraction with low alkaloid content was discarded.

Recovery of ACN from liquid waste was accomplished by double distillation after separation of all solid fractions, resulting in an azeotropic mixture with about 87% ACN content, which was used to prepare 60% ACN in subsequent loads.

Descriptive statistical analysis was performed using the statistical tools of Microsoft Excel. Quantitative data were presented as mean ± standard deviation.

Results and discussion

For five successive loads of 2.5 g of raw material with an SA content of 39.2% by weight (53.8% of the total alkaloids) and CHE content of 33.7% by weight, the following products were obtained using the above-described technological procedure: SA fraction – 899 ± 36 mg with an SA content of 90.6 ± 1.1% by weight or 95.4 ± 1.2% of the total alkaloids and an SA recovery of 83.1 ± 4.7% (for the entire technological process); CHE fraction – 463 ± 19 mg with a CHE content of 93.4 ± 0.2% by weight or 98.1 ± 0.2% of the total alkaloids and a CHE recovery of 51.3 ± 4.2%. An intermediate fraction of 344 ± 26 mg, containing 19.1 ± 2.8% SA and 76.1 ± 2.8% CHE by weight, was also obtained.

When fractionating an increased load (12.5 g of the raw material from the same batch), 4.75 g of the SA fraction with an SA content of 91.2% by weight or 96.0% of the total alkaloids and an SA recovery of 88.4%, 2.61 g of the CHE fraction with a CHE content of 92.0% by weight or 96.6% of the total alkaloids and a CHE recovery of 57.0%, and 1.79 g of the intermediate fraction containing 21.5% SA and 73.7% CHE by weight were obtained.

The results of pilot experiments on the fractionation of raw material samples with different alkaloid ratios, as well as on the repeated fractionation of the isolated fractions, are presented in Tables 1 and 2.

Repeat analysis of dry samples of the pseudo-bases confirmed their stability for at least 12 months under ambient conditions and at 40°C.

The principle of the proposed method for separating benzophenanthridine alkaloids is based on the difference in the ionization constants of the individual components. The alkaloid cations and the corresponding pseudo-bases in aqueous-ACN solutions have different optical absorption spectra (Fig. 1), which made it possible to determine their ratios in buffer solutions of different pH values by measuring the concentration of the cationic form at 470 nm for SA and 405 nm for CHE. As can be seen in Fig. 2, SA cations are transformed into the pseudo-base form at lower pH values than CHE cations. The differentiating effect of ACN is also clear from Fig. 2. Another factor contributing to the separation is that the CHE pseudo-base is more soluble than the SA pseudo-base in ACN-containing solvents, whereas in water it is less soluble of the two (Table 3). When titration was performed in aqueous buffer without added ACN, the best fractions obtained were SA with a purity of 83.8% and a recovery of 58.9% and CHE with a purity of 88.9% and a recovery of 77.4%. An initial ACN concentration of approximately 60% is optimal for separating the alkaloids.  Further increases are ineffective, and at ACN concentrations above 70%, the solubility of the cationic form of the alkaloids decreases.

Good separation of the alkaloids is ensured by a slow change in pH over time, which is facilitated by the formation of a phosphate buffer during the titration of alkaloid bisulfates with dibasic phosphate, or an acetate-phosphate buffer during the titration of acetate solutions of the pseudo-bases with tribasic phosphate. In the latter case, the use of a more alkaline titrant was necessary due to the higher buffering capacity of the resulting buffer system. The use of mixed potassium-sodium buffers is dictated by the limited solubility of individual phosphates, especially sodium phosphates, since lower titrant concentrations would have led to significant dilution of the reaction mixture, reducing the ACN concentration and impairing alkaloid separation. In the presence of sodium salts, separation of the liquid into two phases was observed during titration (salting-out effect), which did not affect the separation of alkaloids but contributed to the formation of a coarser precipitate of pseudo-bases and facilitated subsequent filtration. Since the passage of reaction products (pseudo-bases) from the liquid to the solid phase affects the acid-base equilibrium, and their crystallization is a relatively slow process, a fairly slow addition of the titrant at a controlled temperature is necessary. Temperatures between 40 and 50°C are optimal for isolating the SA fraction. Higher temperatures, although they speed up the process, significantly impair the alkaloids’ separation, while lower temperatures require excessively long titration times. At the second titration stage, however, lower temperatures (30 to 40°C) combined with a higher titrant feed rate proved optimal for removing residual SA from the solution. In this case, constant slight overtitration of the solution prevents redissolution of the already precipitated SA.

The pH and LSI graphs in Fig. 3 illustrate the processes occurring during the titration of the starting material. In the dissolution phase (I), the alkaloid bisulfates are transformed into more soluble neutral salts. In phase (II), a relatively rapid increase in pH is observed, and towards the end of this phase, the formation of the SA pseudo-base begins. This leads to supersaturation of the solution with the SA pseudo-base, which causes subsequent rapid onset of crystallization (point "a" on the LSI graph) and a decrease in pH at the beginning of the SA fraction precipitation phase (III). In the intermediate fraction precipitation phase (IV), the crystallization of the CHE pseudo-base begins (point "b"). The resulting precipitate still contains significant amounts of SA. Finally, in phase (V), a fairly pure CHE fraction precipitates. The graphs on the right side of Fig. 3 allow one to select the optimal titrant volumes before separating the SA and the intermediate fractions, focusing simultaneously on the purity and recovery of the target alkaloids.

The data in Table 1 demonstrate the feasibility of achieving good separation of alkaloids at despite considerable variability in their ratio in the raw material. However, the optimal volumes of titrant required for the precipitation of the SA and intermediate fractions depend on this ratio. The table also shows that repeated fractionation of the separated SA and CHE fractions significantly improves their purity and, in the case of the intermediate fraction, yields an additional amount of CHE, which compensates for its relatively low recovery in the primary fractionation. The data in Table 2 confirm that acetic acid solutions of pseudo-bases, as well as alkaloid bisulfates, are suitable for repeated fractionation.

In this study, the alkaloid pseudo-bases were converted back to their bisulfate form after separation. However, given their acid-base behavior in biological media, the pseudo-bases themselves may be of interest as prodrugs, since they are sufficiently stable and, according to our observations, probably do not exhibit local irritant properties.

Conclusions

A simple method for separating a mixture of SA and CHE into individual alkaloids was developed, based on their sequential precipitation as pseudo-bases from aqueous-ACN solutions during acid-base titration. It allows the effective separation of the alkaloids without the use of chromatographic or other expensive techniques, while minimizing the use of toxic solvents.

Competing interests

None declared.

Authors’ contributions

IC designed the study, performed the technological part, prototyped the preparative titration installation, developed a software application, interpreted the data, and drafted the first manuscript. AC performed the analytical part of the laboratory work and critically revised the manuscript. The final version of the manuscript was approved by all authors.

Ethics approval

Not needed for this study

Acknowledgements and funding

The research was carried out within Subprogram 080301, funded by the Ministry of Education and Research of the Republic of Moldova.

Provenance and peer review

Not commissioned, externally peer-reviewed.

Authors’ ORCID IDs

Igor Casian – https://orcid.org/0000-0001-6392-3804

Ana Casian – https://orcid.org/0000-0001-8876-3691

References

1.     Frolova A, Kosinets A. Antimicrobial resistance. New antimicrobial agents in the treatment of chronic trophic ulcers. Lietuvos chirurgija [Lithuanian Surgery]. 2015;14(1):31-37. https://doi.org/10.15388/LietChirur.2015.1.5329

2.     Falchi FA, Borlotti G, Ferretti F, Pellegrino G, Raneri M, Schiavoni M, Caselli A, Briani F. Sanguinarine inhibits the 2-ketogluconate pathway of glucose utilization in Pseudomonas aeruginosa. Front Microbiol. 2021;12:744458. doi: 10.3389/fmicb.2021.744458.

3.     Gong Y, Yin S, Sun S, Li M. Chelerythrine reverses the drug resistance of resistant Candida albicans and the biofilm to fluconazole. Future Microbiol. 2022;17(16):1325-1333. doi: 10.2217/fmb-2021-0203.

4.     Wang Q, Dai P, Bao H, Liang P, Wang W, Xing A, Sun J. Anti-inflammatory and neuroprotective effects of sanguinarine following cerebral ischemia in rats. Exp Ther Med. 2017;13(1):263-268. doi: 10.3892/etm.2016.3947.

5.     Danielewski M, Zielinska S, Merwid-Lad A, Szandruk-Bender M, Slupski W, Wlodarczyk M, Sozanski T, Ziolkowski P, Szelag A, Nowak B. Sanguinarine–Chelerythrine from Coptis chinensis offers analgesic and anti-inflammatory effects without gastrotoxicity. Pharmaceutics. 2025;17(3):323. https://doi.org/10.3390/pharmaceutics17030323

6.     Malikova J, Zdarilova A, Hlobilkova A. Effects of sanguinarine and chelerythrine on the cell cycle and apoptosis. Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub. 2006;150(1):5-12. https://doi.org/10.5507/bp.2006.001

7.     Zhou J, Qiu Z, Zhou F, Xin G, Kong N, Lv G, Lin H, Lin Z, Fu Y, Wang Y, Huang X. Chelerythrine induces apoptosis and ferroptosis through Nrf2 in ovarian cancer cells. Cell Mol Biol. 2024;70(3):174-181. https://doi.org/10.14715/cmb/2024.70.3.26

8.     Pallichankandy S, Thayyullathil F, Cheratta AR, Subburayan K, Alakkal A, Sultana M, Drou N, Arshad M, Tariq S, Galadari S. Targeting oxeiptosis-mediated tumor suppression: a novel approach to treat colorectal cancers by sanguinarine. Cell Death Discov. 2023;9:94. https://doi.org/10.1038/s41420-023-01376-3

9.     Pallichankandy S, Rahman A, Thayyullathil F, Galadari S. ROS-dependent activation of autophagy is a critical mechanism for the induction of anti-glioma effect of sanguinarine. Free Radic Biol Med. 2015;89:708-720. doi: 10.1016/j.freeradbiomed.2015.10.404.

10.  De Stefano I, Raspaglio G, Zannoni GF, Travaglia D, Prisco MG, Mosca M, Ferlini C, Scambia G, Gallo D. Antiproliferative and antiangiogenic effects of the benzophenanthridine alkaloid sanguinarine in melanoma. Biochem Pharmacol. 2009;78(11):1374-1381. doi: 10.1016/j.bcp.2009.07.011.

11.  Almeida IV, Fernandes LM, Biazi BI, Vicentini VEP. Evaluation of the anticancer activities of the plant alkaloids sanguinarine and chelerythrine in human breast adenocarcinoma cells. Anticancer Agents Med Chem. 2017;17(11):1586-1592. doi: 10.2174/1871520617666170213115132.

12.  Senchina DS, Shah NB, Busch MG. Effects of berberine, chelerythrine, and sanguinarine on proliferation in four human immortalized cell lines. J Iowa Acad Sci. 2012;119(1-4):22-27. https://scholarworks.uni.edu/jias/vol119/iss1/6

13.  Amat-ur-Rasool H, Ahmed M, Hasnain S, Ahmed A, Carter WG. In silico design of dual-binding site anti-cholinesterase phytochemical heterodimers as treatment options for Alzheimer’s Disease. Curr Issues Mol Biol. 2021;44(1):152-175. https://doi.org/10.3390/cimb44010012

14.  Pi G, Ren P, Yu J, Shi R, Yuan Z, Wang C. Separation of sanguinarine and chelerythrine in Macleaya cordata (Willd) R. Br. based on methyl acrylate-co-divinylbenzene macroporous adsorbents. J Chromatogr A. 2008;1192(1):17-24. doi: 10.1016/j.chroma.2008.03.039.

15.  Zhong M, Wang YH, Wang L, Long RQ, Chen CL. Preparation and application of magnetic molecularly imprinted polymers for the isolation of chelerythrine from Macleaya cordata. J Sep Sci. 2018;41(16):3318-3327. https://doi.org/10.1002/jssc.201800245

16.  Gołkiewicz W, Błażewicz A, Jóźwiak G. Isolation of milligram quantities of sanguinarine and chelerithrine from the roots of Chelidonium majus L. by zonal micropreparative TLC. J Planar Chromatogr-Mod TLC. 2001;14(2):95-99. https://www.researchgate.net/publication/235959359

17.  Liu Y, Wang W, Che F, Lu Y, Li A, Li H, Liu J, Wei Y. Isolation and purification of alkaloids from the fruits of Macleaya cordata by ionic-liquid-modified high-speed counter-current chromatography. J Sep Sci. 2020;43(12):2459-2466. https://doi.org/10.1002/jssc.201901242

18.  Liu Q, Sun C, Meng F, Zhao W. Preparative separation of chelerythrine and sanguinarine from Macleaya cordata by pH-zone-refining counter-current chromatography. J Liq Chromatogr Relat Technol. 2015;38(20):1789-1793. https://doi.org/10.1080/10826076.2015.1105257

19.  Ali I, Li J, Cui L, Zhao H, He Q, Wang D. Efficient extraction and purification of benzo[c]phenanthridine alkaloids from Macleaya cordata (Willd) R. Br. by combination of ultrahigh pressure extraction and pH-zone-refining counter-current chromatography with anti-breast cancer activity in vitro. Phytochem Anal. 2021;32(3):423-432. https://doi.org/10.1002/pca.2990

20.  Absolínová H, Jančář L, Jančářová I, Vičar J, Kubáň V. Acid-base behaviour of sanguinarine and dihydrosanguinarine. Cent Eur J Chem. 2009;7(4):876-883. doi: 10.2478/s11532-009-0079-y.

21.  Wu Y, Liu ZY, Cao Y, Chen XJ, Zeng JG, Sun ZL. Reductive metabolism of the sanguinarine iminium bond by rat liver preparations. Pharmacol Rep. 2013;65(5):1391-1400. https://doi.org/10.1016/S1734-1140(13)71498-1

Casian IG, Casian AC, Valica VV. Isolation of benzophenanthridine alkaloids from macleaya leaves without using toxic solvents. Khimiia Rastitel'nogo Syr'ia [Chem Plant Raw Mater]. 2019;(3):79-84. doi: 10.14258/jcprm.2019034524.