Identification of main glycoalkaloids
Figure 1 shows the LC–FTICR/MS separation in positive ion mode of an aqueous extract of black nightshade berries. Analysis of the extracts revealed the presence of two main glycoalkaloids identified by accurate m/z values of protonated species, comparison with authentic standard, and on the basis of IRMPD fragmentation in the ICR cell of precursor ions [M+H]+. In the insets, the mass spectra of two main peaks corresponding to solasonine found at m/z 884.50079 (C45H74NO16, exact m/z 884.50021) and solamargine found at m/z 868.50476 (C45H74NO15, exact m/z 868.50530) are reported. Both compounds were identified with a mass error lower than 1 ppm, which indicates a very good mass accuracy. In the IRMPD MS spectra (data not shown), several common loss from sugar moiety and product ions were observed. Ions generated from fragmentation of B-ring or E-ring of aglycons were diagnostically useful for establishing their membership in the general family of glycoalkaloids . The other intense peak in the TIC (Fig. 1) can be due to a derivative compound of solamargine, the malonyl-solamargine at m/z 954.50525 (C48H75O18N, exact m/z 954.50569).
Quantitative analysis revealed that black nightshade berries extract contains a high amount of solamargine and solasonine (1.35 and 1.52 g kg−1 dry weight, respectively) and a small concentration of other minor known glycoalkaloids, confirming results obtained by Ventrella et al. .
Identification of glucosinolates
The identification of GLSs was based on the study of characteristic fragments of these compounds in IRMPD MS/MS spectra, and on the measure of accurate masses observed using LC/ESI-FTICR/MS, according to Agneta et al. [12, 22]. In Fig. 2, the total ion chromatogram (TIC) acquired in negative ion mode of a horseradish root extract is shown. The qualitative and quantitative analyses of this extract confirmed the occurrence of a high amount of sinigrin (2.04 g kg−1 dry weight), which accounts for more than 90% of the total GLS, and of the other 16 GLSs in trace quantity [12, 22]. In the inset of Fig. 2, the mass spectrum of the peak corresponding to sinigrin, found at m/z 358.02747 (C10H17NO9S2, exact m/z 358.02720, error 0.8 ppm) is shown. By accurate high-resolution mass analysis, the peak eluting at 12.5 min was excluded to be a glucosinolate, but was found to be rustoside, also known as kaempferol 3-lathyroside, which is a member of the class of compounds known as flavonoid-3-O-glycosides normally derived from horseradish.
GLSs exhibited [M−H]¯ as the precursor ion that corresponds to easy deprotonation of the sulphate group. Moreover, the dissociation of [M−H]¯ precursor ion yielded abundant product ions, which gave much information on the structure of the side chain and were of great value for a correct assignment of known and unknown GLSs. Typical fragments of GLS with nominal m/z 97, 195, 241, 259, and 275, which correspond to the fragment ions HSO4¯, C6H11O5S¯, C6H9O8S¯, C6H11O9S¯, and C6H11O8S2¯, respectively, were found in the spectrum examined (data not shown). Other characteristic fragments, such as [M-80-H]¯, [M-162-H]¯, [M-178-H]¯, [M-196-H]¯, and [M-242-H]¯, were very informative for correct molecular identification of GLSs .
Identification of cannabinoids
Using optimized reversed-phase liquid chromatography (RP-HPLC) coupled to electrospray ionization in positive mode (ESI+) and Fourier transform ion cyclotron resonance (FTICR)/MS, together with tandem mass spectrometry (MSn) studies performed using IRMPD and collisional induced dissociation (CID), it was possible to separate and quantify four known cannabinoids (THC, CBD, CBDa, and CBN), useful for the chemotype definition and the classification of C. sativa (Scheme 1). The total ion current (TIC) chromatogram (Fig. 3) revealed the occurrence of three main cannabinoid peaks assigned to CBD, THC, and CBDa; in the insets, the mass spectra of these peaks are shown: CBD, found at m/z 315.23159 (C21H30O2, exact m/z 315.23184, error − 0.8 ppm); THC, found at m/z 315.23148 (C21H30O2, exact m/z 315.23186, error − 1.2 ppm); CBDa, found at m/z 359.22126 (C22H30O4, exact m/z 359.22169 error − 1.2 ppm). CID and IRMPD fragmentation of precursor ions [M+H]+ generates several common species that are diagnostically useful for establishing their identity (data not shown). The wide peak at 7.4 min corresponds to cannabidivarin (CBDV, C19H26O2, m/z [M+H]+ 287.20056), a non-psychoactive cannabinoid homologue of CBD with the side chain shortened by two methylene bridges; it is not useful to determine the chemotype in Cannabis plant destined for human consumption or industrial transformation (Scheme 1).
The quantification of secondary metabolites (THC, CBD, CBN, CBDa) was performed in parallel through low-resolution mass spectra, selected reaction monitoring (SRM), and high-resolution total ion current (TIC).
As described in Table 1, the chemovar analysed as sample #1 does not exceed the THC limit (0.2%) recommended by the European Union regulations [26, 27], confirming previous findings .
The ([THC] + [CBN])/[CBD] ratio (phenotypic index) of samples was used to assess the chemical phenotype (chemotype) of the specific accession .
The high content of cannabidiol (CBD) suggests that the “Eletta campana” accession can be defined as an industrial hemp having a ratio [CBD]/[THC] > 10 (CBD-prevalent chemotype) . In our case, the CBN concentration is not significant for the chemical definition of cannabis quality.
Data reported in Table 1 for the sample #2 indicate that the field selection of plant flowers was able to discriminate a group of plants with a higher content of analysed cannabinoids.
Antimicrobial activity assays
The antimicrobial activities and MICs were evaluated against selected bacterial strains giving different results depending on the type of plant under observation.
Statistics of the antimicrobial activity data (diameters of ZoI) confirmed that the diameter ranges chosen (< 10 mm; 11–15 mm, > 16 mm) were able to well discriminate significant differences among the antimicrobial activities (Tables 2, 3, 4).
Solanum nigrum and C. sativa extracts demonstrated a certain antimicrobial activity, while A. rusticana did not reveal any activity against bacteria in this research.
The Gram− bacteria, P. orientalis and S. maltophilia, were sensitive only to the S. nigrum extract, showing a middle inhibition diameter of 13.5 and 15 mm, respectively; moreover, this extract proved a middle antimicrobial activity against all Gram+ bacteria (inhibition zone ranging from 13.5 to 15.2 mm) (Table 2).
Both flower samples of C. sativa showed a similar effect on Gram+ bacteria with a high antimicrobial activity; these extracts were more effective against B. thuringiensis and B. cereus with 37.5 and 37.0 mm diameter of inhibition zone, respectively, while B. amyloliquefaciens was slightly less sensitive (Table 2).
The active extracts of S. nigrum and C. sativa were subjected to determine MIC by the agar well diffusion method against the respective susceptible bacterial species (Table 2). The results obtained indicated that Gram+ and Gram− bacterial species tested were sensitive to different extracts in a similar way with an MIC of 5–10 mg L−1. The more effective extracts were the two samples of C. sativa with the higher antimicrobial ability and a low inhibitory concentration (Table 2).
The antimicrobial activity of standard pure components of the plants was investigated to understand whether the activity observed in our experiments was due to the synergistic action of more than one constituent in the extracts .
In the case of S. nigrum, solamargine, solasonine, and the solamargine/solasonine mixture (1:1 v/v) were tested. All bacteria were sensitive to both components with a middle antimicrobial activity ranging from 12 to 15 mm (Table 3).
Among Gram+ bacteria, B. cereus was the most sensitive (with an MIC of 5 mg L−1) compared to the other two species, B. thuringiensis and B. amyloliquefaciens, which were inhibited at higher concentrations (ranging from 20 to 40 mg L−1). Gram− bacteria, instead, showed the same behaviour in the presence of standard pure compounds (Table 3).
The C. sativa components were able to inhibit only the Gram + bacteria tested; THC showed a low antimicrobial activity, while CBD and the CBD/THC mixture (1:1 v/v) proved a middle activity, underlining a stronger effect when the mixture was used (Table 4); nevertheless, the bacterial species appeared not very sensitive to the standard pure components, requiring an inhibitory concentration of > 60 mg L−1.