Evidence for Cardiac Glycosides in Foliage of Colorado Potato Beetle-Resistant Solanum okadae

Abstract

Leptinotarsa decemlineata, the Colorado potato beetle (CPB), is a herbivore that primarily feeds on Solanum foliage and is a global pest of the potato agricultural industry. Potato breeding through cross-hybridization with CPB-resistant wild relatives is used for genetic improvement. The wild species Solanum okadae was demonstrated to deter CPB feeding in choice and no choice feeding assays. Liquid chromatography–mass spectrometry (LC-MS) was used for comparative metabolite profiling between S. okadae and CPB-susceptible domesticated potato variety, Solanum tuberosum cv. Shepody. Major foliar metabolites detected were steroidal glycoalkaloids (SGAs) with tomatine and dehydrotomatine produced in S. okadae and solanine and chaconine in S. tuberosum cv. Shepody. Cardiac glycosides were also detected in the foliar metabolite profile of S. okadae but not S. tuberosum cv. Shepody. This class of plant compounds have known insecticidal activity through inhibition of animal Na⁺/K⁺ ATPase. Thin-layer chromatography (TLC) separation of foliar extracts also provided evidence for cardiac glycosides in S. okadae. Cardiac glycosides are known inhibitors of Na⁺/K⁺ ATPase, and foliar extracts from S. okadae (OKA15), but not S. tuberosum cv. Shepody, were able to inhibit the Na⁺/K⁺ ATPase of CPB. These findings suggest a novel mechanism of plant resistance against CPB involving production of cardiac glycosides in S. okadae.

Introduction

Commercial potato varieties are susceptible to Colorado Potato Beetle (CPB) defoliation.¹ CPB are global pests, which show adaptive resistance to insecticides.² Plant resistance to herbivores includes production of secondary metabolites that can mediate deterrence activity, toxicity, or represent precursors to physical defense systems.³ It is thought that secondary metabolites evolve in plants as a response to herbivore-related stresses.⁴ This evolution in plants results in resistance mechanisms evolving in herbivores.⁴ It is through this cycle of resistance that diversity in plant secondary metabolites and insect herbivores is maintained.⁴ The major potato defense metabolites include SGAs, which are secondary metabolites widely produced in Solanum foliage.¹ This metabolite class can disrupt cellular membranes and inhibit acetylcholinesterases in insect pests.⁵ However, CPB have demonstrated the ability to feed on foliar tissue containing high levels of solanine and chaconine, indicating these pests have evolved to tolerate and feed on potato foliage with high levels of SGAs.^5,6

Strategies to reduce yield losses of potato due to CPB defoliation include genetic improvement through interspecies crosses with wild potato species carrying beetle resistance.¹Solanum okadae is a diploid wild tuber-bearing potato species native to Bolivia and Argentina.^7,8 This wild species has the capacity to intercross with Solanum tuberosum.⁹ Previous studies have shown that adult and larval CPB have deterred feeding on S. okadae in the field, resulting in reduced defoliation of S. okadae.¹⁰ Furthermore, CPB larvae feeding on S. okadae have an increased mortality rate compared with CPB larvae feeding on commercial S. tuberosum.¹⁰

The current study provides a comparative analysis of the foliar metabolites produced in CPB-resistant S. okadae and CPB-susceptible S. tuberosum. Differences in glycoalkaloid profiles of the two species were found. Additionally, analysis of foliar metabolites of S. okadae revealed the presence of cardiac glycosides, which are associated with insect resistance in other plant species. This is a rare finding of cardiac glycosides in Solanaceae and the genus Solanum.

Methods

Plant Material for Metabolite Analysis

Wild S. okadae plants were propagated from true botanical seeds sourced from the NRSP-6, the United States Potato Genebank (accession number PI 458367). Seeds were sown and germinated in potting mix in a 1 cm wide plastic trough. Germinated seedlings were transplanted in the greenhouse to potting mix in 6-in. pots at ambient temperature. The domesticated species, S. tuberosum cv. Shepody, was propagated from tuber seed sourced from the Benton-Ridge Substation of the Fredericton Research and Development Centre (formerly Potato Research Centre), Agriculture and Agri-Food Canada (AAFC), Benton, New Brunswick, Canada. A weekly fertilizer application of 20–20–20 was applied to both species left to grow in the greenhouse in 6-in. pots with potting mix. At 12 weeks after planting, sampling of foliage was done on fully grown plants. Apical leaflets from five leaves for each plant in 15 mL conical tubes were flash-frozen in liquid nitrogen. Prior to extraction, the conical tubes containing the samples were stored at −80 °C. Sampling was completed from three S. okadae plants grown from true botanical seed and nine replicates for S. tuberosum cv. Shepody grown from tubers.

True botanical seed from the same accession of S. okadae was sown on MS agar (4.4 g of Murashige and Skoog basal salts with vitamins, 30 g of sucrose, and 7 g/L plant tissue culture agar) and germinated. Germinated seedlings were transferred to potting mix in 6-in. pots in the greenhouse. A single vigorous and fertile seedling, named OKA15, was selected, propagated in vitro, and used for analysis of CPB feeding and analysis of cardiac glycosides. S. okadae (OKA15) plants grown in the greenhouse were also transferred to a short-day growth chamber with 8 h light and 16 h dark to induce tubers after 1 month of growth in the greenhouse. Tubers were harvested after 2 months in the short-day growth chamber.

Wild common milkweed plants growing outdoors were harvested for cardiac glycoside analysis. Foliage was sampled and frozen at −20 °C.

CPB Feeding Choice Assay

Freshly laid CPB egg masses were obtained from laboratory colonies at the Fredericton Research and Development Centre. Laboratory colonies were reared on S. tuberosum cv. Kennebec. The egg masses were left to hatch under 24 h light and remained under these conditions until they reached the second instar (L2). The L2 larval stage for CPB is associated with feeding preference behavior in previous studies¹⁰ and was used for feeding choice assays in the current study.

The amount of CPB feeding on leaves from S. okadae (OKA15) was compared to S. tuberosum cv. Shepody, a susceptible species, using a feeding choice assay. The assay was conducted using quantification of CPB feeding on leaf disks 1.2 cm in diameter. The disks were made using a metal cork-borer. Leaflets from the middle section of the plant were used to produce the disks. Disks were arranged following the template in Figure 1a and each disk overlapped another. The leaf area for disks was recorded prior to CPB feeding. A single second instar larva was left to feed in each dish for 24 h under constant light. After the 24 h feeding period, the area of leaf consumed was quantified using LeafByte version 1.3.0 (https://zoegp.science/leafbyte)¹¹ and the percent of the original pre-feeding area for leaf disks was determined. There was a total of eight replicate Petri dishes for each of the treatments (“Shepody”, “Choice Test”, and “Okadae”). The percent leaf area consumed for two S. tuberosum cv. Shepody and the two S. okadae (OKA15) leaf disks in the “Choice Test” dishes was compared using a two-sample t-test (Welch), where the null hypothesis was no differences between S. tuberosum cv. Shepody and S. okadae (OKA15) leaf disks. The two-sample t-test was also used to compare the no choice “Shepody” dish leaf disks and the S. tuberosum cv. Shepody leaf disks in the “Choice Test” dish. The leaf disks in the no choice “Okadae” dish and S. okadae (OKA15) leaf disks in the “Choice Test” dish were also compared. Significant differences at p ≤ 0.01 and 0.05 were noted.

Figure 1.

Leaf disk placement for the CPB feeding choice assay in Petri dishes. “Shepody” is a no choice dish with four disks of S. tuberosum cv. Shepody. “Choice Test” is a dish with two pairs of disks: leaf disks on the bottom left and top right are S. tuberosum cv. Shepody, and leaf disks on the bottom right and top left are S. okadae (OKA15). “Okadae” is a no choice dish with four disks of S. okadae (OKA15). (a) Template for three different placements of S. okadae (OKA15) (O) and S. tuberosum cv. Shepody (T) leaflet disks. (b) Moist filter paper prior to placing the leaf disks in Petri dishes. (c) Petri dishes before 24 h feeding period. (d) Petri dishes after 24 h of CPB feeding. The arrow on the disk template indicates the direction the CPB head faced when placed in the Petri dish.

CPB was weighed on an analytical balance before being placed in the Petri dishes and after 24 h of feeding. The increase in mass of CPB from the “Choice Test” dish was compared to the no choice “Okadae” dish and the no choice “Shepody” dish. Comparison was also made between the CPB mass increase from the no choice “Shepody” dish with the no choice “Okadae” dish. A two-sample t-test (Welch) was done, where the null hypothesis was no differences with mass of CPB from choice and no choice dishes at p ≤ 0.01 and 0.05.

Extraction of Foliar Samples for LC-MS

The frozen foliar samples from each of three S. okadae plants grown from true botanical seed and nine S. tuberosum cv. Shepody plants grown from tubers were ground into a fine powder using a mortar and pestle while immersed in liquid nitrogen. 100 mg of frozen ground tissue was placed in 1.5 mL polyethylene screw-cap tubes and kept frozen in a liquid nitrogen holding station (SPEX Sampleprep, Metuchen, NJ) until all samples were ground. The ground powder was extracted with 400 μL of extraction solution (92% methanol, 0.1% formic acid LC-MS grade v/v) (Sigma-Aldrich, Oakville, ON, Canada). The samples were briefly vortexed and placed on ice until all samples were prepared. Samples were then placed in a sonicating water bath for 15 min and filtered through a 0.2 μm syringe filter into an LC-MS autosampler vial. Sample dilutions were optimized to ensure peak intensities were in the linear range and to avoid detector saturation. S. okadae and S. tuberosum cv. Shepody samples were diluted 4-fold with extraction buffer. The samples were allowed to equilibrate at 23 °C in the dark for 1 h prior to analysis and were maintained under these conditions for the duration of the LC-MS analysis.

UPLC–qTOF-MS

Metabolite analysis was carried out using Acquity ultraperformance liquid chromatography–Xevo quadrupole time-of-flight mass spectrometry (UPLC–qTOF-MS) (Waters, Milford, MA) as described in Tai et al.¹² Using a 5 μL loop, 0.75 μL injections were made for all samples in the study. The same volume of test mixture (rutin hydrate, caffeic acid, benzoic acid, p-coumaric acid, quercetin, l-phenylalanine, resveratrol, ferulic acid, l-tryptophan, sinapic acid, naringenin, trans-cinnamic acid, and isorhamnetin) (Sigma-Aldrich, Oakville, ON, Canada) was injected. All components of the test mixture were present at ∼35 μg/mL in 60:40 acetonitrile/Water. All chromatographic separations were carried out on a 1 mm × 100 mm BEH C18 reversed-phase column. The mobile phase was composed of LC-MS grade water with 0.1% formic acid (phase A) and LC-MS grade acetonitrile with 0.1% formic acid (phase B). The linear gradient consisted of six segments as follows: initial segment 95% A, 5% B; 13:33 min 25% A, 75% B; 13:53 min 5% A, 95% B; 18:00 min 5% A, 95% B; 18:01 min 95% A, 5% B; and 20:00 min 95% A, 5% B. The flow rate was 45.0 μL/min for all segments. The autosampler bed was maintained at 23 °C and the column at 35 °C. Samples were injected in a randomized fashion. Each sample was injected in triplicate except for the test mixture which was injected after every six samples to evaluate the stability of retention time and mass accuracy over the duration of the experiment. Mass spectrometry data were collected over the duration of the LC-ramp from 0 to 800 s. Mass-to-charge ratios (m/z) between 100 and 1500 were detected by electrospray ionization in positive ionization mode using parameters recommended by the manufacturer. A lock mass solution of dilute leucine enkephalin (LE) in acetonitrile/water (50:50) was introduced via the lock-spray probe at 25 μL/min as directed by the MS manufacturer.

Data Processing of Mass Spectra

Mass spectrometry data were processed using Waters Breeze software. The UPLC-MS data were detected and noise-reduced in the UPLC and MS domains. True analytical peaks were further processed by the software. The chromatographic retention time from the chromatogram and m/z of the positive molecular ion was used to identify a feature from each peak. The most common observed molecular ion adduct was [M + H]⁺. Retention time stability was sufficient for UPLC, eliminating the need for retention time correction. Peak intensity was quantified by integrating peaks with a mean retention time in the window of 100–800 s. Column void and washout were excluded from the selected retention time window through visual evaluation of the chromatograms. Compound identities were assigned to features through matching m/z against theoretical masses of compounds in the MetLin (https://metlin.scripps.edu) and MoNA (https://mona.fiehnlab.ucdavis.edu) databases. Features within 6 ppm of the compounds in databases were listed. The Chemical Entities of Biological Interest (ChEBI) (https://www.ebi.ac.uk/chebi),¹³ and The Human Metabolome Database (https://hmdb.ca/metabolites),¹⁴ were also used to identify compound classes and their applications for metabolites.

The metabolite peak intensity data were log₁₀ transformed after adding 1 to all data to remove zeros. The data were grouped by Solanum species. A two-sample t-test (Welch) was used to find features with variation between species at p ≤ 0.01. The null hypothesis tested stated there were no differences between species. The LC-MS features displaying significant species variation were further analyzed. The untransformed peak intensity data for these features were further filtered by excluding features with average peak intensities below a 10-fold difference in S. okadae compared to S. tuberosum.

Steroidal Glycoalkaloid (SGA) Extraction and HPLC Analysis

Approximately 0.25 g of fresh foliar samples of S. okadae (OKA15) were placed in 50 mL centrifuge tubes. The foliar samples were extracted using 10 mL of extraction buffer (92% methanol, 0.1% formic acid LC-MS grade v/v). The samples were homogenized for 3 min using an Omni PREP multi-sample homogenizer (Omni International, Keenesaw, GA). At room temperature, the extracts were centrifuged at 61 rcf for 10 min. The supernatant was then placed in a glass boiling flask and the process was repeated two more times. Using a rotary evaporator at 40–45 °C, 30 mL of the extracts were evaporated to dryness. After 4 mL of ion-pairing reagent (4.0 g of 1-heptanedulgonic acid, sodium salt, 10 mL of glacial acetic acid, H₂O to 1000 mL) was added to the extracts three times, the samples were transferred to 50 mL centrifuge tubes. The samples were centrifuged at 61 rcf for 10 min. Using a pipette, 10 mL of supernatant was inserted into a solid-phase extraction (SPE) unit previously rinsed with 5 mL of methanol and ion-pairing reagent. The samples were rinsed with 5 mL of cartridge wash solution (acetonitrile/water = 20:80). After adding 2 mL of Eluting solution (acetonitrile/water = 50:50) to the SPE unit, the eluate was collected in a 4 mL glass vial. The eluate was filtered using a 0.22 μm syringe filter in preparation for high-pressure liquid chromatography (HPLC) on a Waters Breeze2 system (Waters, Inc.).

Extraction of Cardiac Glycosides

S. okadae (OKA15) and S. tuberosum cv. Shepody leaves from greenhouse-grown plants and milkweed leaves from wild plants were used. Plant tissue was harvested and stored in a −20 °C freezer. The leaves were freeze-dried using a lyophilizer (Labconco, MO). The tissue was then ground (Geno/Grinder, SPEX SamplePrep, NJ) to a powder at 1500 rpm for 12 min in a 50 mL Falcon tube containing two ceramic beads. The plant tissue powder was extracted using a lead acetate extraction method for cardiac glycosides.¹⁵ Briefly, 1 g of plant tissue powder was added to a 50 mL round-bottom flask containing 50% ethanol (20 mL) and 10% lead acetate solution (10 mL) and a small stirring bar. The solution was refluxed for 15 min using a condenser. The solution was filtered using vacuum filtration (Whatman 1, Millipore Sigma, Oakville, Canada). A small quantity (∼5 mL) of acetic acid was added to the filtered solution. The solution was extracted with 15 mL washes of dichloromethane. The extract was filtered over anhydrous sodium sulfate and left to evaporate to dryness in a fumehood. The dried residue was dissolved in 1 mL of dichloromethane/ethanol (1:1) solution for TLC.

TLC Analysis of Cardiac Glycosides

Extracts were analyzed using TLC with Millipore Sigma silica gel 60 F₂₅₄ fluorescence indicator plates (Millipore Sigma, Oakville, Canada). The plates were developed using an ethyl acetate/methanol/water (100/13.5/10 mL) mobile phase containing a few drops of acetic acid. The 20 × 20 cm² glass-back plates were cut into quarters using a handheld glass cutter. Plates were developed in a glass TLC chamber (according to instructions from TLC plate manufacturer) and separated compounds were visualized with fluorescence quenching under ultraviolet (UV) light at 254 nm and with fluorescence emission under UV light at 365 nm (Cole Parmer, Montreal, Canada). The cardiac glycoside compound, ouabain (0.4 μM) (Sigma-Aldrich, Oakville, ON, Canada), was spotted on the plate as three applications of 4 μL (4.8 × 10^–6 μmol total).

CPB Microsome Extraction

CPB microsomes were isolated using a commercial kit following the manufacturer’s instructions (product number ab206995, AbCam, Toronto, ON, Canada). Briefly, protease inhibitor cocktail (2 μL/mL for each buffer) was added to the homogenization and storage buffers (∼3 mL homogenization buffer and 0.5 mL storage buffer was required per gram of tissue). The protease inhibitor cocktail was composed of an irreversible (AEBSF) and a reversible (aprotinin) serine protease inhibitor, a metalloprotease inhibitor selective for aminopeptidases (bestatin), an irreversible inhibitor of cysteine proteases (E-64), a reversible inhibitor of serine and cysteine proteases (leupeptin), and a reversible aspartic acid protease inhibitor (pepstatin A). The buffer solutions were kept on ice for the duration of the experiment and all centrifuge steps were performed at 4 °C in a pre-chilled refrigerated centrifuge. Three fresh L4 whole-body CPB larvae (400 mg) were placed in a pre-chilled Dounce homogenizer. Chilled homogenization buffer (200 μL) was added to the sample. The sample was gently homogenized on ice with 10–15 strokes. Additional homogenization buffer (0.6 mL) was added to the homogenizer. The tissue slurry was pipetted up and down to fully suspend the homogenate. The homogenate was transferred to a microcentrifuge tube and vortexed for 30 s and then placed on ice for a 1 min incubation period. The homogenate was centrifuged at 10 000 rcf for 15 min at 4 °C. Using a Pasteur pipette, the floating lipid layer was gently aspirated. The supernatant was transferred to a new pre-chilled microcentrifuge tube and centrifuged at a maximum speed (20 000 rcf) for 20 min at 4 °C. Once centrifugation was complete, the floating lipids were aspirated and the supernatant was discarded. The light beige-pink microsomal pellet (1.568 mg/mL) was maintained. The remaining pellet was gently washed with homogenization buffer (200 μL). The excess buffer was discarded. The microsomal pellet was resuspended in chilled storage buffer (200 μL).

Na⁺/K⁺ ATPase Enzyme Inhibition Assay

The CPB microsome extract was diluted 1:360 in 0.5 M Tris buffer and used for testing for inhibition of Na⁺/K⁺ ATPase activity by compounds in S. okadae (OKA15), milkweed and Shepody foliar extracts and ouabain (Sigma-Aldrich, Oakville, ON, Canada) using an assay kit that quantified release of inorganic phosphate (P_i) from ATP (product number ab270551, AbCam, Toronto, ON, Canada) following manufacturer’s instructions. Control and test samples were run in triplicate. The control reaction, consisting of the substrate mix (100 μL) and 50 μL of diluted CPB microsome extract and 50 μL of 1:1 dichloromethane/ethanol solution, had the maximal amount of Na⁺/K⁺ ATPase activity. The test reactions consisted of substrate mix (100 μL), 50 μL of diluted CPB microsome extract, and 50 μL of ouabain or foliar tissue extracts from either OKA15, milkweed, or Shepody. Three foliar extract concentrations were tested for inhibition of Na⁺/K⁺ ATPase activity: undiluted and 1:5 and 1:125 dilutions. Ouabain was also tested for inhibition of Na⁺/K⁺ ATPase activity in the CPB microsome extract at concentrations of 40 μM and 0.4 μM. The ATP was kept on ice when not in use. After a 30 min incubation period at room temperature, 50 μL of PiColorLock (1800 μL) with accelerator (18 μL) was added to each well. Two minutes after the first addition of the PiColorLock solution, 20 μL of stabilizer was added to each well and mixed by pipetting up and down. After letting sit for 15 min after the first addition of stabilizer, the plate was read using a BioTek Synergy HTX Multi-Mode Reader (Agilent Technologies Canada, Inc., Mississauga, ON, Canada) at 590 nm.

Inhibition of Na⁺/K⁺ ATPase activity with plant extracts and ouabain was compared with the control reaction, which did not contain inhibitors, using a two-sample t-test (Welch) with significance levels of p ≤ 0.01 and p ≤ 0.05. The null hypothesis tested stated there was no difference with the control reaction.

Statistical Analysis

All statistical analyses were performed using the R statistical analysis software package (version 4.0.0).¹⁶

Results

CPB Deterred from Feeding on S. okadae but Not S. tuberosum

S. okadae (OKA15) and S. tuberosum cv. Shepody plants were tested for differences in CPB feeding using leaf disk choice and no choice feeding assays (Figure 1). The “Choice Test” dish had two different pairs of leaf disks, one pair was S. tuberosum cv. Shepody (“Exp Shep”) and the other pair was S. okadae (OKA15) (“Exp OKA”). A significant increase in CPB leaf disk consumption of “Exp Shep” over “Exp OKA” was found in the “Choice Test” dish (Figure 2). These results show an avoidance of S. okadae (OKA15) leaves by CPB and a preference for S. tuberosum cv. Shepody. CPB leaf disk consumption in the no choice “Shepody” dish (“Control Shep”) was compared to “Exp Shep”, and no significant differences were found. There were also no significant differences between CPB leaf disk consumption in the no choice “Okadae” dish (“Control OKA”) and “Exp OKA” (Figure 2). These results show that CPB avoided feeding on S. okadae (OKA15) even if there was no other choice in food.

Figure 2.

Boxplots of the average percent area consumed of leaf disks in Petri dishes. “Control Shep” (green) is the average for four leaf disks in the no choice Shepody dish with only S. tuberosum cv. Shepody. “Control OKA” (orange) is the average for four leaf disks in the no choice Okadae dish with only S. okadae (OKA15). The Choice Test dish has one pair of leaf disks of S. tuberosum cv. Shepody (“Exp Shep”, blue) and another pair of leaf disks of S. okadae (OKA15) (“Exp OKA”, pink). Significant differences between “Exp Shep” and “Exp OKA” leaf disks in the Choice Test dish found using a t-test are indicated with ** (p-value ≤ 0.01). Outliers are represented by black circles.

The CPB feeding in either the no choice “Shepody” dish or “Choice Test” dish increased in mass compared to CPB in the no choice “Okadae” dish (Figure 3). CPB were also observed to have reduced movement and lack of frass in the no choice “Okadae” dishes. The lack of CPB mass gain further supports a deterrence by S. okadae of CPB feeding. CPB mass in the no choice “Shepody” dish and the “Choice Test” dish were not significantly different.

Figure 3.

Boxplots of the increase in mass of CPB feeding on leaf disks. Average mass increase of CPB is on the y-axis for insects feeding in the no choice S. tuberosum cv. Shepody dish (“Shepody” green), the choice test dish with both S. tuberosum cv. Shepody and S. okadae (OKA15) (“Choice Test”, purple), and the no choice dish with only S. okadae (OKA15) (“Okadae”, orange) as indicated on the x-axis. Significant differences found in the t-tests comparing CPB in the no choice “Shepody” and “Okadae” dishes, and in “Choice Test” and no choice “Okadae” dishes, are indicated with * (p-value < 0.05). Outliers are represented by black circles.

Comparative Analysis of Foliar Metabolites between S. okadae and S. tuberosum

Production of defense metabolites in foliage is a mechanism of resistance for plants against insect herbivores, such as CPB, and likely contributes to deterrence of feeding in S. okadae. A comparative analysis of foliar metabolites between S. okadae and S. tuberosum was done to identify metabolites associated with different CPB feeding preference between these two species. Three S. okadae plants were used for analysis that were grown from seeds from the same accession as S. okadae (OKA15). Peak intensities of the metabolite features from the untargeted LC-MS analysis of foliar extracts from S. okadae and S. tuberosum cv. Shepody were compared. The peak intensities were log₁₀ transformed and features with significant differences between the two species are listed in Table S1. There were 111 features showing peak intensity differences, and 48 of the features were assigned to compounds based on searching the MetLin database. Out of the 111 features, 11 were increased in S. tuberosum cv. Shepody over S. okadae, the rest were increased in S. okadae. Metabolites with the greatest difference in peak intensities were steroidal glycoalkaloids (SGAs) for both species.

SGAs

The retention time for the SGAs in the LC-MS analysis was between 7.00 and 9.00 min (Table S1). The major peaks for S. tuberosum and S. okadae found within this range were examined. Features increased in S. tuberosum cv. Shepody included 7.60/868.5004 and 7.96/868.5099, which had m/z’s matching solanine in the MetLin database. The other metabolites in the MetLin search were not known to be found in potato (Table S1). Features 7.60/868.5004 and 7.96/868.5099 were also assigned to solanidenol chacotriose and 7.46/884.5049 was assigned to solanidenol solatriose, which are SGAs reported in other published studies.^12,17 Also increased in S. tuberosum cv. Shepody was the feature 8.05/852.514, which matched m/z with chaconine. A derivative of chaconine that has a single hexose sugar moiety, γ-chaconine (7.97/560.3957), was also increased in S. tuberosum cv. Shepody as was the feature 7.47/414.335, which was assigned to solasodine. These SGA features had high peak intensities in S. tuberosum cv. Shepody and were not detectable or at background noise levels in S. okadae.

The wild species, S. okadae, produced glycoalkaloids in large quantities that were not detected in S. tuberosum cv. Shepody. The features 7.75/1034.5552 and 7.77/578.4045 had notably high average peak intensities in S. okadae and neither were detected in S. tuberosum. These features had m/z’s that matched with α- and δ-tomatine, respectively. β-Tomatine has an m/z matching that of feature 7.80/902.5148 and was produced in lower levels. The detection of SGA tomatine, the tomato SGA, in the wild species aligns with previous reports.¹⁰ Other features with average peak intensities belonging solely to S. okadae include 7.66/1032.5438, 7.62/738.4479, and 7.63/576.3906. These features match the m/z’s of dehydrotomatine and its disaccharide and monosaccharide derivatives, γ and δ, respectively. The feature 7.78/416.3512 could be assigned to either SGAs tomatidine or soladulcidine.

Validation for SGAs detected in the LC-MS analysis was done using other plant extracts from the same clones using another HPLC instrument with standards for solanine, chaconine, dehydrocommersonine, tomatine, and dehydrotomatine. The analysis shows the presence of solanine and chaconine in S. tuberosum cv. Shepody but not S. okadae, and the presence of tomatine and dehydrotomatine in S. okadae but not S. tuberosum cv. Shepody (Figure 4), which was similar to results from untargeted LC-MS. The SGA dehydrocommersonine, which was produced in other wild Solanum,^12,18 was shown to be absent from S. okadae (Figure 4).

Figure 4.

HPLC chromatograms of SGAs. (a) Standards: 1. Solamarine, 2. Solamargine, 3. Dehydrocommersonine, 4. Solanine and 5. Chaconine. (b) Standards: 6. Dehydrotomatine, 7. Tomatine. SGAs from foliar extracts of (c) S. tuberosum cv. Shepody. (d). S. okadae (OKA15).

Other Notable Metabolites

Features 7.60/1227.6069, 7.41/743.4609, 7.24/901.4805, 7.24/1063.5354, 7.41/1229.6195, and 7.41/1067.5678 had an m/z matching that of compounds in the steroidal saponins group (Table S1). Fistuloside B (7.24/901.4805), hypoglaucin G (7.24/1063.5354), and melongoside P (7.41/1067.5678) were among the compounds assigned. Feature 7.41/743.4609 had an m/z matching that of alliosterol 1-rhamnoside 16-galactoside, a steroidal glycoside. Furthermore, several flavonoid glycosides were identified. Features 5.65/611.1611, 5.66/757.2223, and 5.65/465.1003 were assigned to flavonoids rutin, cyanidin 3-O-rutinoside 5-O-β-d-glucoside, and quercetin 3-O-glucoside, respectively. Features 4.36/163.0387 and 4.02/163.0393 had an m/z matching that of hydroxycoumarin. Other assigned coumarins include 1,3,8-naphthalenertriol, 2-propenal, and 10-hydroxy-2,8-decadiene-4,6-diynoic acid, which were assigned to feature 5.24/177.0541. Features 7.76/193.0498, 7.76/223.0593, and 6.15/147.0444 were assigned compounds belonging to the coumarin class. Other notable metabolites include retinoids, which were assigned to features 8.42/287.2371 and 8.31/301.2147. These features had m/z matching those of 13-cis-retinol and retinoic acid, respectively. Furthermore, features 7.84/275.2364, 7.41/401.3396, and 9.20/449.3252 were assigned compounds pertaining to the steroid class. Additionally, 7.78/295.1018 was assigned to tuliposide B, a lipopolysaccharide.

Cardiac Glycosides

Cardenolide glycosides, a subclass of cardiac glycosides (Figure 5), were assigned to features 8.41/535.2898 and 8.41/553.2995. Corchoroside A, helveticoside, and corotoxigenin-3-O-α-l-rhamnopyranoside were assigned to feature 8.41/535.2898. Bipindogulomethyloside, desglucocheirotoxol, antioside, lokundjoside, and panogenin-3-O-α-l-rhamnopyranoside were assigned to feature 8.41/553.2995. The presence of cardiac glycosides was a rare finding for a Solanum species. The discovery of cardiac glycosides in the CPB-resistant S. okadae was of high interest as these compounds have known insecticidal activity and indicate another mode of insect resistance for the genus Solanum.

Figure 5.

General structure of (a) cardiac glycosides, (b) cardenolides, and (c) bufadienolides.

TLC Analysis of Cardiac Glycosides

Cardiac glycosides absorb short-wave UV light, which can be visualized as fluorescence quenching on TLC plates embedded with a fluorescent indicator.^15,19 TLC of these three spots had approximately the same R_f value (0.66 and 0.70) and similar size and shape for all three extracts. However, emission under long-wave (365 nm) UV light of the spot for S. okadae (OKA15) and milkweed was a blue-white fluorescence, whereas emission was a red fluorescence for S. tuberosum cv. Shepody (Figure 6b), indicating a different class of compound. The fluorescence quenching of the S. okadae (OKA15) X_B1 spot was stronger and more distinct than for milkweed (X_C1) and S. tuberosum cv. Shepody (X_D1). Additional fluorescence emitting spots were visible for milkweed and S. tuberosum cv. Shepody under 365 nm (Figure 6 and Table 1).

Figure 6.

(a) Fluorescence quenching under 254 nm UV light of TLC separated spots with ouabain (A) and foliar extracts of S. okadae (OKA15) (B), milkweed (C), and S. tuberosum cv. Shepody (D). The R_f values of the indicated quenching spots are as follows: x_A1 = 0.87, x_B1 = 0.70, x_B2 = 0.74, x_C1 = x_D1 = 0.66. (b) Fluorescence emission under 365 nm UV light for ouabain (A) and foliar extracts of S. okadae (OKA15) (B), milkweed (C), and S. tuberosum cv. Shepody (D). y indicates the solvent front.

Table 1. R_f Values of TLC Plate Spots for Ouabain and the S. okadae (OKA15), Milkweed, and S. tuberosum cv. Shepody Plant Foliar Extracts Fluorescently Quenching under 254 nm (Short-Wave) UV Light and Emitting Fluorescence under 365 nm (Long-Wave) UV Light^a.

sample	254 nm	365 nm
ouabain	0.87
OKA15	0.70	0.70 (blue-white)
	0.74
milkweed	0.66	0.32 (dark blue)
		0.40 (light blue)
		0.48 (yellow)
		0.66 (blue-white)
Shepody	0.66	0.66 (red)
		0.94 (red)

^aThe colors of the fluorescent bands under 365 nm are in parentheses.

Inhibition of CPB Na⁺/K⁺ ATPase

Cardiac glycosides inhibit the activity of Na⁺/K⁺ ATPase.²⁰ Therefore, additional evidence for the presence of cardiac glycosides in milkweed and S. okadae (OKA15) foliar extracts was obtained using an assay for inhibition of Na⁺/K⁺ ATPase from CPB microsomes. Ouabain at 0.4 and 40 μM along with foliar extracts from S. okadae (OKA15), milkweed, and S. tuberosum cv. Shepody was assayed (Figure 7). Three different concentrations of foliar extracts were used: undiluted and 1:5 and 1:125 diluted (Figure 7). The highest level of Na⁺/K⁺ ATPase activity in the CPB microsomes was in the control reaction with no inhibitors, which produced P_i that had an absorbance at 590 nm of 2.339 in the presence of PiColorLock. The t-test showed that ouabain at 40 μM had inhibition of Na⁺/K⁺ ATPase activity, but not at 0.4 μM. Undiluted and 1:5 dilutions of S. okadae (OKA15) extracts had significant inhibition of CPB Na⁺/K⁺ ATPase, but not the 1:125 dilution (Figure 7). Undiluted milkweed extracts displayed significant inhibition of CPB Na⁺/K⁺ ATPase, but not the 1:5 and 1:125 diluted milkweed extracts (Figure 7). None of the S. tuberosum cv. Shepody extracts inhibited the CPB Na⁺/K⁺ ATPase (Figure 7).

Figure 7.

Inhibition of the Na⁺/K⁺-ATPase channel in CPB microsomes by S. okadae (OKA15), milkweed, and S. tuberosum cv. Shepody foliar tissue extracts. Inhibition in the Na⁺/K⁺ ATPase assay was quantified by measuring the amount of inorganic phosphorus released from ATP, which changes the color of the PiColorLock reagent, which was measured using absorbance at 590 nm. The control (light blue) consisted of equal volumes of 1:1 dichloromethane/ethanol solution and 1:360 diluted CPB microsomal extract had the maximal amount of Na⁺/K⁺-ATPase without inhibition. Ouabain (40 μM and 0.4 μM, dark blue) and foliar extracts from S. okadae (OKA15) (green), milkweed (pink), and S. tuberosum cv. Shepody (yellow) were tested undiluted and diluted at 1:5 and 1:125. The assay was done in triplicate. Samples with significant inhibition of CPB Na^+/K⁺-ATPase compared to the control with no added inhibitors are indicated with asterisks (**p-value < 0.01).

Discussion

CPB showed avoidance of S. okadae (OKA15) in a feeding choice assay with S. tuberosum. These results are consistent with previously published results on CPB resistance in S. okadae.¹⁰ The feeding choice assay done in the current study used a single clone of S. okadae (OKA15), which is from the same accession (PI 458367) as was used in Pelletier et al.¹⁰S. okadae (OKA15) is conserved in vitro for use in germplasm enhancement in potato breeding. These results also demonstrate consistency in CPB resistance in the S. okadae accession that was used in the study. The CPB also avoided feeding on S. okadae (OKA15) when there was no choice, demonstrating the CPB were not just showing a preference for S. tuberosum cv. Shepody, but had a strong avoidance of S. okadae (OKA15).

The foliar metabolite analysis showed variation in the production of SGAs, which are major defense metabolites in Solanaceae species.²¹ High levels of tomatine and dehydrotomatine, which are characteristic of tomato, were found in S. okadae foliage. This contrasts with potato, which produces solanine and chaconine SGAs. CPB feeds on both tomato and potato,²² hence the production of tomatine is likely not underlying CPB resistance observed in S. okadae. However, S. okadae may show differences in defense against other pests and pathogens from S. tuberosum as a result of these differences in foliar SGAs.

Cardiac glycosides are naturally occurring compounds found in plants and animals, where they serve as a defense mechanism.²³ Cardiac glycosides have structures derived from the steroid skeleton.²³ The skeleton typically consists of three 6-membered carbon rings and one 5-membered ring (Figure 5), with a lactone ring at the C-17 position and a sugar moiety at C-3.²³ Additionally, the type of lactone generates two subcategories of cardiac glycosides: an unsaturated five-membered furanone ring represents a cardenolide and an unsaturated six-membered pyranone ring indicates a bufadienolide.²³ Structurally, cardiac glycosides are similar to steroidal saponins and thus have similar properties.²⁴ Cardiac glycosides can be distinguished from other steroidal glycosides by the presence of a hydroxyl group at C-14.²⁴ Additionally, these compounds are described as having a “U-Shape” due to unique stereochemistry at their ring junctions.²⁴

Both cardenolides and bufadienolides cardiac glycosides are produced by plants and animals.²⁵ The most common cardiac glycosides are digoxin and digitoxin, which are isolated from Digitalis plants.²³ The plant from which cardiac glycosides were first isolated, foxglove (Digitalis purpurea), contains a variety of cardiac glycosides including digoxin, digitoxin, and digitonin.²⁶ Other common cardiac glycosides derived from plants include ouabain, neriine, and oleandrin.²⁶ Foliar extracts were prepared from S. okadae (OKA15), milkweed, and S. tuberosum cv. Shepody using a procedure that enriched for cardiac glycosides. Spots on TLC with fluorescence quenching were found in the plant foliar extracts. The S. okadae (OKA15) foliar extracts had two spots showing fluorescence quenching under 254 nm UV light. One of these spots also had fluorescence emission under 365 nm UV light that was blue-white. Some cardiac glycosides can produce fluorescent emission under long-wave UV.²⁴ Therefore, both fluorescently quenched spots for S. okadae (OKA15) could be cardiac glycosides. Milkweed are plants known to produce high levels of foliar cardiac glycosides,²⁷ and the milkweed extract was found to produce a fluorescently quenched spot under 254 nm UV light that emitted blue-white fluorescence under 365 nm UV light. Milkweed produces several cardiac glycosides;²⁸ however, a single fluorescence quenching spot was detected under 254 nm UV light in the current study. This result with milkweed was likely due to the optimization of conditions for foliar extraction and TLC for S. okadae in the current study.

For centuries, cardiac glycosides have been isolated from plants for use as insecticides or rodenticides.²⁰ Ingestion of cardiac glycosides leads to toxicity in the cardiovascular and autonomic nervous system.²⁰ These steroids can seriously limit cardiac muscle contraction—leading to heart blockage or cardiac arrest—through inhibition of the Na⁺/K⁺-ATPase channel.²⁰ The Na⁺/K⁺ pump is responsible for maintaining ion concentrations in cells and neurotransmitters, therefore making cardiac glycosides toxic to animals.²⁰ Species specificity of Na⁺/K⁺ ATPase for cardiac glycoside inhibition was reported;²⁹ therefore, the assay in the current study used microsomes prepared from CPB in enzyme inhibition studies. S. okadae (OKA15) foliar extracts were found to inhibit CPB Na⁺/K⁺ ATPase enzyme, which provided additional evidence for the presence of cardiac glycosides in OKA15 foliar tissue. Inhibition of Na⁺/K⁺ ATPase by ouabain at 40 μM was also found for CPB Na⁺/K⁺-ATPase. The concentration of ouabain observed to inhibit the Na⁺/K⁺ ATPase from guinea pig myocardium was in the same range.³⁰ These results confirm the presence of the cardiac glycoside-sensitive Na⁺/K⁺ ATPase in CPB microsomes. Inhibition by the milkweed extract was also found for the CPB Na⁺/K⁺-ATPase. However, inhibition was lost with dilution of the milkweed extract, unlike S. okadae (OKA15), which maintained inhibition at a 1:5 dilution. The results suggest that S. okadae cardiac glycosides may have greater specificity for CPB Na⁺/K⁺ ATPase compared to milkweed. The study provides further evidence that S. okadae produces cardiac glycosides as defense compounds, which is uncommon for the genus Solanum.

S. tuberosum cv. Shepody had a weak fluorescently quenched spot under 254 nm UV light. However, cardiac glycosides were not identified in the LC-MS metabolite profiling of S. tuberosum cv. Shepody. Additionally, fluorescence emission at 365 nm UV light was red rather than blue-white, indicating difference from the fluorescently quenched cardiac glycosides in S. okadae (OKA15) and milkweed foliar extracts and ouabain. Furthermore, the Na⁺/K⁺ ATPase from CPB was insensitive to inhibition from the S. tuberosum cv. Shepody extracts, concurring with lack of cardiac glycoside production.

Previously, cardiac glycosides have been used in the treatment of heart failure and cardiac arrhythmias, and rigorously studied as possible anticancer treatments.²³ Since the SARS-CoV-2 pandemic began in 2019, cardiac glycosides have been of interest for their antiviral activity against enveloped and nonenveloped respiratory viruses.³¹ The antiviral and anti-inflammatory effects of cardiac glycosides have the potential to decrease viral replication of SARS-CoV-2 and modulate the inflammatory response in COVID-19 patients.³¹ Continual research on cardiac glycosides remains relevant due to their wide variety of applications.³¹S. okadae is a possible source of cardiac glycosides for medicinal use, but further investigation will be needed. Future isolation and structural analysis will be performed to confirm the presence of cardiac glycosides in S. okadae.

Glossary

Abbreviations

CPB

Colorado potato beetle

OKA15

Solanum okadae

SGAs

steroidal glycoalkaloids

LC–MS

liquid chromatography–mass spectrometry

TLC

thin-layer chromatography

UPLC–qTOF-MS

ultraperformance liquid chromatography–quadrupole time-of-flight mass spectrometry

LE

leucine enkephalin

SPE

solid-phase extraction

ACN

acetonitrile

AEBSF

4-(2-aminoethyl)-benzenesulfonyl fluoride

P_i

inorganic phosphorus

ATP

adenosine triphosphate

HPLC

high-performance liquid chromatography

UV

ultraviolet

SARS-CoV-2

severe acute respiratory syndrome coronavirus 2

COVID-19

coronavirus disease 19

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.jafc.2c04311.

• Untargeted metabolite profiling of foliar compounds in S. okadae and S. tuberosum (Table S1) (PDF)

The project was funded by the Agriculture and Agri-Food Canada Genomics Research and Development Initiative.

The authors declare no competing financial interest.