Preparative-cum-quantitative mass-directed analysis of swainsonine and its in situ activity against Sf-21 cell line


Correspondence: Gurvinder Kaur, Department of Biotechnology, Indian Institute of Technology Guwahati, Guwahati 781 039, Assam, India. Tel.: +91 361 2582207; fax: +91 361 2582249; e-mail:


Swainsonine is a polyhydroxy indolizidine alkaloid with various research and potential therapeutic applications. In this work, swainsonine was partially purified (2.5-folds) with acetone–methanol solvent system from Metarhizium anisopliae fermentation broth. The partially purified broth was further subjected to mass-directed preparative-cum-quantitative analysis. Swainsonine was eluted as MS1 fraction [M + H]+ 174.36 ± 0.21 at 4.91 ± 0.04 min with calculated yield of 7.85 ± 1.59 μg mL−1 corresponding to 3.74 × 105 counts. In situ antiproliferative activity of standard and purified swainsonine fractions was tested against Spodoptera frugiperda, Sf-21 cell line with IC50 values of 2.96 μM and 3.28 μM, respectively, at 36 h. This analytical procedure for purification and quantitative analysis of swainsonine may ensure its suitability for routine laboratory studies and research.


The entomopathogenic fungus Metarhizium anisopliae is a well studied and applied species for microbial control of insect pests (Schrank & Vainstein, 2010). Metarhizium spp. produce a variety of secondary metabolites in several chemical classes, including cytochalasins C and D, myroridins, destruxins A, B and E, viridoxin, helvonic acid, 12-hydroxyovalicin, hydroxyfungerin, 7-desmethyl analogues of fusarin C and (8Z)-fusarin C, serinocyclins A and B and aurovertins. These metabolites are toxic to a broad range of animals and microorganisms, including insects, fungi, bacteria and viruses. One of the metabolites is swainsonine, a trihydroxy indolizidine alkaloid with selective inhibition property against Golgi α-mannosidase II that blocks the abnormal formation of complex β-1, 6 branched N-linked glycans, leading to reduced metastasis and tumour growth (Thompson et al., 2012). Swainsonine is also described for its apoptotic activities against cancerous A549 and C6 glioma cells, both in situ and ex situ (Li et al., 2012).

Swainsonine production and purification still remains a challenge from both biological as well as expensive synthetic routes. Earlier swainsonine was reportedly being extracted and purified with various polar solvent systems, physiological conditions and ion-exchange resins (Gardner & Cook, 2010). However, the use of polar solvents such as ethanol or water also imposes another problem of co-extraction of broth components such as glucose, amino acids and other such hydrophilic metabolites. Several methods have been reported for the analysis and quantification of swainsonine including capillary gas chromatography, thin-layer chromatography, HPLC, LC-MS/MS and HPLC with evaporative light-scattering detector (Yang et al., 2012). Conventional HPLC methods have rarely been applied to swainsonine analysis on account of reasons such as (1) the lack of suitable chromophore groups rendering its UV-visible detection almost impossible and (2) the extreme hydrophilicity of swainsonine ensues the co-elution of small polar molecules such as sugar or amino acids (Molyneux et al., 2002). Moreover, preparative LC-MS methods are advantageous over destructive low pH conditions used in ion-exchange chromatography (Donaldson et al., 1990). Hence, the development of sensitive, selective and nondestructive method for the analysis of swainsonine has always been the prerequisite for its application studies. Mass spectrometry is considered to be a specific and universal detection method. Liquid chromatography–mass-spectrometry-directed (LC-MSD) purification provides a high-throughput rapid purification-cum-quantification of diverse library compounds under selected ion monitoring (SIM) mode (John et al., 2010; Chen et al., 2012).

Increasing sensitivity towards secondary metabolites from fungal biological control agents has prompted the toxicological risk assessment of metabolites produced by M. anisopliae (Skrobek & Butt, 2005). In situ assays are important and useful tools in toxicity assessment of various classes of environmental contaminants including fungal metabolites because they significantly reduce evaluation time and also provide information about the mode of action of the toxicant (Fornelli et al., 2004).

In the present study, we propose a novel mass-directed purification-cum-quantification method for the partially purified swainsonine and showed its entomopathogenic activity against Spodoptera frugiperda (Sf-21) cell line.

Materials and methods

Microorganism and Cultivation conditions

Metarhizium anisopliae ARSEF 1724 was procured from ARS Collection of Entomopathogenic Fungal Cultures (ARSEF) USDA, Ithaca, New York, and maintained in Sabouraud dextrose agar slants at 4 °C. Basal oatmeal media (60 mg mL−1) supplemented with glucose (20 mg mL−1) at 28 °C and 180 r.p.m. were used for the production of swainsonine (Singh & Kaur, 2012).

Cell culture and treatments

Spodoptera frugiperda (Sf-21) cell line was procured from National Centre for Cell Sciences (NCCS), Pune, India. These were maintained as suspension (T-25 flask BD biosciences) at nonhumidified, non-CO2 conditions and at 28 °C in TNH-F media with 10% foetal bovine serum. Cultures in the early stationary phase (typical cell density, 1.6 × 106 cells mL−1 with cell viability approximately 80%) were split every 3–4 days with a starting density of 4 × 105 cells mL−1.


The ingredients required for maintenance and swainsonine production media were purchased from Hi-Media, India. Swainsonine standard from M. anisopliae, α-d-mannosidase from jack bean, p-nitrophenyl-α-d-mannopyranoside and l-glutathione reduced were all obtained from Sigma-Aldrich. All chemicals for purification were of HPLC grade (Merck, Germany). All cell culture-related materials were obtained from Sigma-Aldrich.

Swainsonine extraction and purification

The culture broth (85–90 mL) was centrifuged at 16 000 g for 10 min at 4 °C to separate the cells. The supernatant was mixed with three different solvent systems, viz. acetone (A), acetic acid (2%) in chloroform (B) and ethanol (C) at the fixed proportion of 1 : 4. Cold (−20 °C) and acidic (approximately pH 5.5) conditions were employed up to 8 h for efficient extraction (Fellows & Fleet, 1989). The mixture was immediately centrifuged to remove the high molecular weight components. The extract was then concentrated in vacuum rotary evaporator (Ika HB 10) at 45 °C and freeze-dried before final dissolution in 1 mL methanol.

Enzymatic quantification of swainsonine

The swainsonine titres in the culture supernatant, broth extracts and purified samples were determined using α-mannosidase inhibition assay (Sim & Perry, 1995).

Qualitative mass spectrometry–electro spray ionization (MS-ESI+) analysis

Partially purified broth extracts were filtered (0.2 μm) and MS-analysed (Agilent technologies Inc.). The MS-ESI+ ion source parameters were adjusted for source and desolvation gas temperatures at 100 and 250 °C, respectively. The sampling, capillary and extraction cone voltages were set at 35 V, 3 KV and 3 V, respectively, with ion guide at 1 V. The flow injection rate was fixed at 20 μL min−1. Instrument control and data acquisition was performed with Chemstation analytical software.

LC-MSD purification-cum-quantification

The MSD system (1260 Infinite series, Agilent technologies Inc.) consists of binary solvent pump and autosampler with thermostatted oven, reverse-phase C18 HPLC column (21.2 × 250 mm × 7 A°) and diode array detector (DAD). Injection volume of 25 μL was loaded in all the runs in LC column. Swainsonine was eluted using linear gradient programme (100% water from 0 to 5 min and 100% acetonitrile from 5 to 20 min) at a flow rate of 10 mL min−1. The optimized capillary voltage 3 KV, fragmentor voltage 70 KV and collision energy 25 KV values were applied. The mass spectrometer was run in SIM mode for the mass range 173–174.36 ± 0.21. Instrument control and data acquisition was performed with Chemstation analytical software. LC chromatogram of total ion count (TIC) vs. elution time (min) was analysed in commercial MassHunter software (Agilent Technologies Inc.). The area count for characteristic ion was extracted from the composite total ion peak in SIM mode (Fattorusso & Taglialatela-Scafati, 2007). Calibration curve of known concentration range, 1–8 μg mL−1, was constructed using standard swainsonine. The observed [M + H]+ 174.36 ± 0.21 intensities were plotted against the corresponding concentrations.

Analytical LC-MS validation of swainsonine MS1 fraction 174.36 ± 0.21

Swainsonine mass fractions were further analysed by LC-MS (Triple quad 6410, Agilent Technologies Inc.) with analytical C18 column (4.6 × 250 mm × 5 A°) fitted in thermostatted oven and quaternary pump. The run programme and solvent systems were all being the same as above except the flow rate (1 mL min−1).

Evaluation of swainsonine cytotoxicity using MTT assay

The MTT (3-(4, 5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide) assay was used to describe the in situ cytotoxicity of swainsonine against Sf-21 cell line (established from ovaries of S. frugiperda). The standard and MSD-purified swainsonine was tested at concentrations ranging from 2.0 to 10.0 μM for 36 h with appropriate solvent controls. A 100 μL of 1 × 105 cells mL−1 was seeded into 96-well microtitre plates with 100 μL of swainsonine of varying concentrations in complete growth media. A 20 μL of MTT solution (5 mg mL−1 in PBS) was added after the desired time points into each well, and the cells were incubated for 4 h. The microtitre plate was centrifuged (250 g for 5 min), and the media were removed gently before finally adding 100 μL of DMSO. The plate was then rotated on an orbital shaker for 10 min to dissolve the precipitate completely. The absorbance was detected at 570 nm with reference wavelength of 660 nm using microplate reader (Tecan, Infinite 200).

Statistical analysis

One-way anova followed by Dunnett's test of pairwise multiple comparisons was used for MTT data analysis. Cell viability plots were fitted in spss, version 11 (Science Inc., Chicago, IL), with an exponential decay linear combination equation to generate graphs. Each data point represents mean of multiple wells, as mentioned in legends to Fig. 5.


Microscopic analysis and imaging were performed at 40× magnification using inverted microscope (Nikon TS 100-F, Japan).

Results and discussion

Swainsonine extraction and purification

Solvent Extraction, enzymatic quantification and MS analysis

A prior subtractive purification strategy was applied to precipitate endogenous proteins, sugars and nucleotides, etc. in fermentation broth using organic solvents under chilled conditions. A maximum 2.5-fold purification was achieved using solvent A (acetone), with 7.16 μg mL−1 of swainsonine in extract supernatant. However, only 1.88- and 1.16-fold purification of swainsonine was observed with solvents B (5.19 μg mL−1) and C (3.35 μg mL−1), respectively (Supporting Information, Table S1). The purified fractions were further subjected to mass spectrometry analysis and compared with mass spectrum of swainsonine standard for characteristic MS1 counts (Fig. 1a). The average swainsonine MS1 [M + H]+ 174.36 ± 0.21 ion counts were most abundant in solvent A purified broth (Fig. 1b). The mass spectrum for solvent B purified broth fraction was much noisy with number of unknown higher mass fragments (Fig. 1c), whereas solvent C purified fractions exhibited a comparatively clear mass spectrum, but with insignificant average MS1 ion counts for swainsonine (Fig. 1d). There are certain extra peaks, which may be hypothesized to be of glucose units [M + Na]+, 180 + 23 = 203, or other such sodium or formate adducts of higher sugars in fermentation broth.

Figure 1.

Qualitative ESI-MS+ of (a) swainsonine standard, (b) acetone-extracted, (c) 2% acetic acid-chloroform-extracted and (d) ethanol-extracted samples.

LC-MSD purification

The selective mass-directed purification can be used directly with excellent recovery and purity without involving the tedious packing of ion resins or prepacked cartridges (Kagan et al., 2009; Gardner & Cook, 2010). Here, the approach is based on unique fragmentation patterns requiring either a precursor MS1 ion scan or a neutral loss scan observed under MS/MS2 to generate the species-specific quantification methodology as reported by Chen et al., 2012;. Solvent A (acetone)-extracted broth was selected for LC-MSD purification. Swainsonine MS1 fraction [M + H]+ 174.36 ± 0.21 was eluted at an average (te), 4.91 ± 0.04 min, as the major peak. Swainsonine MS1 fractions were collected between te = 4.41 and te = 4.78 min in vial 1 and te = 4.80 and te = 5.15 min in vial 2. Native mass fractions of (m/z 173) were eluted between te = 5.66 and te = 6.14 min in vial 4, but at negligible signal counts (Fig. 2). The purified fractions were further analysed and validated for swainsonine purification with analytical LC-MS system. Swainsonine was eluted as a single peak at te = 6.38 ± 0.04 min (Fig. 3a) with characteristic m/z = 174.10 under MS1 scan showing 3.74 × 105 counts (Fig. 3b).

Figure 2.

DAD chromatogram of acetone-extracted swainsonine fractions in LC-MSD purification.

Figure 3.

LC-MS validation of MSD-purified fractions (a) TIC chromatogram (b) ESI+ scan.

Mass-directed LC-ESI-MS+ quantification

LC-MSD purification was simultaneously accompanied with mass-directed quantification using the MS1 ion counts [M + H]+ 174.36 ± 0.21 in SIM mode. The standard curve with regression equation;

display math

where Y is MS1 ion counts (intensity) for swainsonine, and X is swainsonine concentration (μg mL−1) with correlation coefficient R of 0.99 (inset of Fig. 4). Swainsonine concentration was expressed as an average of three replicate runs ± SD, 7.85 ± 1.59 μg mL−1 (Fig. 4). Thus, the mass-directed technique was found to be more rapid with high accuracy, selectivity and sensitivity for both qualitative and quantitative analyses.

Figure 4.

Intensity ion counts vs. retention time (min) curve for mass-directed quantification in SIM mode. Calibration curve of swainsonine for precursor ion [M + H]+ 174.36 ± 0.21 as an inset figure.

Method validation for mass-directed LC-ESI-MS+ quantification under SIM mode


The selected ion count [M + H]+ 174.36 ± 0.21 followed linear relationship with the corresponding standard concentration (1–8 μg mL−1) with correlation coefficient (R) of 0.99. Thereafter, regression values altered significantly from linearity (inset of Fig. 4). The limit of detection (LOD) was defined as the lowest concentration of swainsonine that gave an average signal-to-noise ratio > 3 over three replicate injections based on MS1 ion counts. The LOD for swainsonine was 0.25 μg mL−1.

Selectivity and precision

Selected mass fractions [M + H]+ 174.36 ± 0.21 were only observed in SIM mode. The relative standard deviation (RSD) values for MS1 count intensity and elution time were 0.79% and 4.04%, respectively, for triplicate runs in LC-MSD purification and quantification method (Fig. 4).


Swainsonine concentration derived from MSD quantification was in agreement with that determined from enzymatic assay method (Table. S1). The RSD value between the two methods was found to be 6.5%.

In situ cytotoxicity of purified and standard swainsonine against Sf-21 cell line

The MTT assay results showed that swainsonine treatment caused inhibition of Sf-21 growth with as low as 2 μM of standard and purified swainsonine for 36 h. Further increase in swainsonine concentration (10 μM) inhibited the growth by nearly 40% (Fig. 5). Thus, the alkaloid treatment induced antiproliferative effects on Sf-21 cells in a dose- and time-dependent manner for 36 h. Furthermore, cells viability data curve for both standard and purified swainsonine was fitted to the exponential decay linear combination equation at R = 0.99 for the calculation of IC50:

display math

where in case of purified swainsonine, Y is cell viability (%), and X is swainsonine concentration (μM) with coefficients Y0 = 54.14, a = 44.33, b = 1.07 and c = −1.51. The observed IC50 values for purified and standard swainsonine were nearly 3.28 μM and 2.96 μM, respectively, at 36 h (Fig 5).

Figure 5.

MTT viability assay of Sf-21 cells measured after 36 h with different doses of standard and purified swainsonine. One-way anova with Dunnett's pairwise multiple comparison with control indicated that purified swainsonine affected Sf-21 cell viability in a dose dependent fashion (P < 0.001) with = 405.64, signifying the test.

Phase contrast (40×) images showed characteristic membrane blebbing and released apoptotic bodies early after 12 h (Fig. 6a), which became more distinct 24 h onwards (Fig. 6b–c). The observed blebbing in cellular membrane might affect the capacity of the cell to transport and exchange substances with its environment. Generalized vacuolization and the toxic effect on the mitochondria are the main symptoms of insect intoxication by beauvericin (Valencia et al., 2011). To date, there are no reports on the use of swainsonine for insect cell viability assays, and hence, the present study suggested its potential applications for pest control.

Figure 6.

Phase contrast microscopic (40×) image of standard (2.96 μM) and purified swainsonine (3.28 μM)-treated Sf-21 cells at (a) 12 h, (b) 24 h and (c) 36 h.


A nondestructive purification-cum-quantification method has been developed for swainsonine. Satisfactory validation results were obtained in terms of precision and accuracy using the LC-MSD. We also pointed out an in situ target-specific biological activity of swainsonine against S. frugiperda. The biological evaluation of swainsonine produced by M. anisopliae provides better criteria to design more effective formulations for pest management.


We thank Indian Institute of Technology Guwahati for providing the experimental facilities and Council of Scientific and Industrial research (CSIR), PUSA, New Delhi, India, for providing the research fellowship. We are also thankful to NCCS, Pune, India, for providing the insect cell line. The authors have no conflict of interest.


The part of this work has been accepted in poster presentation category at 5th congress of European microbiologists, FEMS 2013, Leipzig, Germany.