Background: The plant kingdom constitutes an enormous reservoir of bioactive molecules, generally used by plants to prevent or to protect themselves from pathogens’ attacks. To date, several primary or secondary plant metabolites have been already proven to exert antibiotic activities; nonetheless, researchers are still continuing to lavish great efforts to identify and characterize new natural molecules one by one. Aiming at the replacement of synthetic chemi- cal products, the bioactivity of plant extracts should be assessed case by case, and active substances should be tested as individuals to obtain accurate information on the real usefulness of plant metabolites. In this work major glycoalka- loids obtained from Solanum nigrum, glucosinolates from Armoracia rusticana, and cannabinoids from Cannabis sativa were identified. The antimicrobial activity of crude extracts and pure components against Gram + (Bacillus cereus (A1I), Bacillus thuringiensis (B7I2), and Bacillus amyloliquefaciens (A5TI)) and Gram− bacteria (Pseudomonas orientalis (A14-1II), and Stenotrophomonas maltophilia (B9TIII)), employed as model organisms, was tested. Result: Major glycoalkaloids, glucosinolates, and cannabinoids were identified in crude plants’ extracts using high- resolution LC–ESI-FTICR/MS. From antimicrobial assays useful information towards a few of biological activities of crude extracts and individual components were obtained. Solanum nigrum extracts revealed inhibition activity on all bacteria tested as well as the main active glycoalkaloids, solamargine and solasonine, which were found to be active even when tested individually. At assayed concentrations, A. rusticana extract was active towards a few of the microorganisms tested, confirming that the activity of glucosi- nolates can be referred only partially to the mother molecules, while biological efficiency of such kind of compounds is mainly due to their enzymatic breaking off, where myrosinase converts them into isothiocyanates and/or thiocy- anates. Hemp-type C. sativa extract showed antimicrobial activity only against Gram+ bacteria, but the main indi- vidual components tested showed always a limited bioactivity. Conclusion: Promising results were obtained, but tests performed in vitro are only the first step of a wider investiga- tion as required for an extensive application. Further research efforts are necessary to demonstrate the efficiency of natural substances in different target environments. Keywords: Glycoalkaloids, Glucosinolates, Cannabinoids, Solanum nigrum, Armoracia rusticana, Cannabis sativa, LC– ESI-FTICR/MS, Antimicrobial activity, Gram+ bacteria, Gram− bacteria *Correspondence: email@example.com Department of European Cultures (DICEM), University of Basilicata, 85100 Potenza, Italy Full list of author information is available at the end of the article © The Author(s) 2018. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creat iveco mmons .org/licen ses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creat iveco mmons .org/ publi cdoma in/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Lelario et al. Chem. Biol. Technol. Agric. (2018) 5:13 Page 2 of 12 to synergistic action with other chemical compounds Background co-synthesized in the plant cells . Natural products were developed and used to relieve This work deals with the identification of more sickness by the dawn of human history: before the representative glycoalkaloids and glucosinolates in “Synthetic Era”, indeed, 80% of medicine, drugs, and crude extracts obtained from Solanum nigrum and pesticides were obtained from roots, barks, and leaves Armoracia rusticana, respectively, and of several can- (fluid extracts) of plants, contributing to the diseases’ nabinoids extracted from Cannabis sativa. Moreover, restraint [1, 2]. Despite this success, natural products’ the antimicrobial activity of crude extracts and some research has endured a global decline due to the pro- pure components against a limited number of Gram+ duction difficulties and small quantities obtained . and Gram− bacteria, employed as model organisms, The necessity of using drugs and pesticides in high was investigated as preliminary assessment of their quantity (linked also to the population growth) has bioactivity. required large amounts of products, which only labo- ratory synthesis could realize. In this process, however, Methods some negative aspects attributable to a large use of Chemicals synthetic substances were not considered, such as the Solamargine and solasonine standards were pur- organic balance alteration, pollution of various envi- chased from Glycomix (UK); sinigrin monohydrate was ronmental systems, resistance induction, and genetic obtained from Sigma-Aldrich (Steinheim, Germany); changes in living beings in a very short period. The Δ9-tetrahydrocannabinol (THC), cannabidiol (CBD), current trend is the return, when possible, to the pre- cannabinol (CBN), and cannabidiolic acid (CBDa) were ceding system of diseases’ treatment, re-emphasizing purchased from HPC Standards GmbH (Cunnersdorf, the use of metabolic constituents produced by several Germany). Methanol, acetonitrile, and formic acid plant, bacterial, and fungal species . Due to the sci- were obtained from Carlo Erba (Milan, Italy). Ultrapure entific awareness surrounding the use of natural sub- water was produced using a Milli-Q RG system from stances instead of synthetic ones, in recent years many Millipore (Bedford, MA, USA). researchers have undertaken studies on the occurrence of secondary metabolites in plants that are widely used in almost all geographical areas and on their possible bioactivity. Standard preparation −1 Secondary metabolites are organic molecules, not Glycoalkaloids: stock solutions (1 mg L ) of pure involved in the normal growth and development of standards in ultrapure water acidified with 1% acetic an organism, whose functions are largely unknown, acid were prepared as reported elsewhere [10, 11] and although they seem involved in the organism defence kept in the darkness at + 4 °C. Glucosinolates: stock −1 [5, 6]. solutions (1 mg L ) of pure standards in methanol/ The recognition of the biological properties of thou- water (70/30, v/v) were kept in the darkness at + 4 °C −1 sands of these molecules has increased interest in this [12–15]. Cannabinoids: stock solutions (1 mg L ) of field for new drugs, antibiotics, insecticides, fungi- pure compounds in ethanol were prepared and kept in cides, and herbicides research and brought about a re- the darkness at − 20 °C. Solutions prepared as above evaluation of bacteria, fungi, and plant role, especially were used for analytical purposes. in the ecological context. Terpenes (gums, resins, For antimicrobial assays, solutions of standard carotenoids, etc.), phenols (lignin, flavones, anthocya- compounds were prepared when necessary before nins, tannins, etc.) as well as alkaloids are just some of each test using sterilized ultrapure water, which was the substances currently used in different application also employed to dilute them up to the required fields, giving rise to the world’s growing attention due concentration. to their widespread use and to the concurrent preser- All glass apparatus were heat sterilized by autoclaving vation of both human health and environment. for 60 min at 121 °C before use. Aseptic handling mate- Each family, genus, and species of several plants pro- rials and laboratory facilities were used throughout the duce a characteristic mixture of substances that can study to maintain sterility. be present in active form or in prodrug state and are used for taxonomic classification . Normally, they are activated when wounding or infection in the veg- Plant samples etal body occurs . These compounds can be active Black nightshade (S. nigrum) unripe berries (glycoalka- as single components or strengthen their activity due loids), horseradish (A. rusticana) roots (glucosinolates), Lelario et al. Chem. Biol. Technol. Agric. (2018) 5:13 Page 3 of 12 and hemp (C. sativa) flowers (cannabinoids) were the extract; 0.001–0.3 mM, for cannabis extracts. MIC assays vegetal materials used to obtain the extracts. were performed according to the European Committee Green unripe berries of black nightshade (60 berries for Antimicrobial Susceptibility Testing (EUCAST) and harvested from 15 different plants in a homogeneous Mann and Markham methods [20, 21]. experimental cultivation) and horseradish root (15 plants harvested from a homogeneous experimental cultivation) Analytical procedure were supplied, respectively, from a greenhouse located All analytical experiments were performed using a Sur- in Metaponto village (Italy) and in a field established at veyor LC system coupled to an ESI-FTICR mass spec- the Institute of Plant Genetics-National Research Coun- trometer (Thermo Fisher Scientific, Bremen, Germany), cil close to Policoro village (Italy). The voucher specimens equipped with a 20 W C O -laser IRMPD (Synrad, of both plants were deposited at the Herbarium Lucanum Mukilteo, WA, USA), emission wavelength 10.6 μm. (HLUC in Index Herbariorum) with the ID Code 2320 Glucosinolates and glycoalkaloids LC separations were and the ID Code 9197 for S. nigrum and A. rusticana, performed at ambient temperature by using the same respectively. chromatographic conditions reported elsewhere [11, 14, Two different flower samples from experimental fields 22, 23]. For separation and identification of cannabinoids, located in southern Italy and derived from the registered a new optimized LC–ESI-FTICR/MS method was used. C. sativa accession “Eletta campana” were supplied with Mass spectrometric conditions were optimized by the courtesy of Eletta campana S.r.l. company. direct infusion of standard solutions. The instrument was The “Eletta campana” cannabis accession has been bred tuned to facilitate the ionization process and to achieve and grown during the past century both in insular and the highest sensitivity. The ESI-FTICR mass spectra peninsular areas of Italy, mainly for industrial production obtained were used to characterize the ionization behav- purposes [16, 17]. iour of the compounds. Sample #1: flowers—obtained as representative popu - Data acquisition and analyses were accomplished using lation—of the “Eletta campana” chemical phenotype, 30 the Xcalibur software package (version 2.0 SR1 Thermo flowers from 30 plants. Electron), and total ion current (TIC) acquisition; data Sample #2: field selection of flowers used in varietal were collected in full MS scan mode and processed post- improvement schemes to select plants with a higher con- acquisition to identify the compounds of interest. In centration of THC and CBD compared to the average addition to accurate mass determination and retention composition of the “Eletta campana” accession, 30 flow - times, extensive structural information was obtained by ers from 30 selected plants. MS/MS fragmentation performance of the compounds investigated (data not shown). The chromatographic raw Extraction, purification, and preliminary tests data were imported, elaborated, and plotted by SigmaPlot Glycoalkaloids and glucosinolates extractions were made 10.0 (Systat Software, Inc., London, UK). in five replicates following previously published methods Analytical determination was performed to know, [10–15]. in detail, the composition of crude extracts and subse- Cannabinoids were extracted in five replicates for each quently to permit bioactivity testing of pure standards, sample of hemp flowers crushed using liquid nitrogen, using about the same concentrations present in the sieved, and lyophilized. Ultrasound-assisted extraction extracts. (USAE) was carried out using absolute ethanol as sol- vent according to published methods [18, 19]. Samples Antimicrobial activity assays were centrifuged by using the Hettich Zentrifuge, MIK- The different extracts were tested against five bacterial RO220R (Germany) for 12 min at 2400g and filtered strains of the culture collection stored in the Department (PTFE filters 0.20 µm) to clarify the liquid phase. of Sciences, University of Basilicata, Potenza, Italy. Three The replicates were brought together to constitute a Gram+ bacteria [Bacillus cereus (A1I), Bacillus thur- representative sample of the total material collected. Sub- ingiensis (B7I2), and Bacillus amyloliquefaciens (A5TI)], sequently, analytical determinations and antimicrobial and two Gram− bacteria [Pseudomonas orientalis (A14- assays were conducted. 1II) and Stenotrophomonas maltophilia (B9TIII)] were Aiming at the determination of the minimal inhibitory employed as screening microorganisms for this study. concentration (MIC), preliminary tests were performed All strains were maintained as freeze-dried stocks in using initial concentration ranges of main representative reconstituted (11% w/v) skim milk, containing 0.1% (w/v) compounds under investigation: 0.005–1 mM, for black ascorbic acid, and routinely cultivated in optimal growth nightshade unripe berries; 1–5 mM, for horseradish conditions. Lelario et al. Chem. Biol. Technol. Agric. (2018) 5:13 Page 4 of 12 Antimicrobial activities of all tested extracts were concentration of the extract inhibiting the growth of bac- determined by agar well diffusion assay . terial strains. The MIC values were obtained in triplicate For each strain, a subculture in a specific broth (PCb) tests. was obtained from the active stock culture by 1% (v/v) Results were compared by analysis of variance inoculum and incubated overnight at 30 °C. A volume of (ANOVA) and Bonferroni post hoc test using GraphPad 200 μL for each subculture was used to inoculate the agar Prism 6 software, version for Windows. 9 −1 media (to achieve a final concentration of 10 CFU L ) and distributed into Petri plates. Each extract (60 μL) Results was poured into wells (5 mm ∅) bored in the agar plates, Analytical outcomes and then the plates were incubated at 30 °C. The organic Identification of main glycoalkaloids solvent was used as negative control, while the antibi- Figure 1 shows the LC–FTICR/MS separation in positive otic was used as positive control. The experiments were ion mode of an aqueous extract of black nightshade ber- performed in triplicate and the antimicrobial activity of ries. Analysis of the extracts revealed the presence of two each extract was expressed as mean diameter (mm) of main glycoalkaloids identified by accurate m/z values of the zone of inhibition (ZoI) produced by the respective protonated species, comparison with authentic standard, extract after 24 h of incubation. A value of ZoI < 10 mm and on the basis of IRMPD fragmentation in the ICR cell was stated to indicate a low antimicrobial activity; of precursor ions [M+H] . In the insets, the mass spec- 11 < ZoI < 15 mm, a middle antimicrobial activity; and tra of two main peaks corresponding to solasonine found ZoI > 16 mm, a high antimicrobial activity. at m/z 884.50079 (C H NO , exact m/z 884.50021) and 45 74 16 Extracts producing an inhibition zone were screened solamargine found at m/z 868.50476 (C H NO , exact 45 74 15 to determine the minimum inhibitory concentrations m/z 868.50530) are reported. Both compounds were and evaluate the antimicrobial effectiveness of each identified with a mass error lower than 1 ppm, which extract against different bacterial strains by means of indicates a very good mass accuracy. In the IRMPD MS the agar well diffusion method . The medium inocu - spectra (data not shown), several common loss from lated with the strain subculture was distributed into Petri sugar moiety and product ions were observed. Ions gen- plates, and different concentrations of extracts, ranging erated from fragmentation of B-ring or E-ring of aglycons −1 from 1 to 100 mg L , were poured into wells bored in were diagnostically useful for establishing their member- the agar plates and the plates were incubated for 24 h. ship in the general family of glycoalkaloids . The other After incubation, the MIC was determined as the lowest intense peak in the TIC (Fig. 1) can be due to a derivative Fig. 1 LC/ESI-FTICR TIC chromatogram acquired in positive mode of a black nightshade berries’ extract. Mass spectra of three main peaks, corresponding to solasonine (found at m/z 884.50079), solamargine (found at m/z 868.50476), and malonyl-solamargine (found at m/z 954.50525) along with their molecular structures are reported in the insets Lelario et al. Chem. Biol. Technol. Agric. (2018) 5:13 Page 5 of 12 [C H NO S ] 10 17 9 2 OH 120 358.02719 - - [C [M-H] H O ] 26 27 15 HO O 579.13611 120 O 80 O O [M-H] HO OH 60 HO O OH OH OH OH 3e+7 200400 600800 1000 20 m/z 200400 600800 1000 m/z 0246 81012141618202224 Time (min) Fig. 2 LC/ESI-FTICR TIC chromatogram acquired in negative ion mode of root crude extract of horseradish. Mass spectra of the main peaks corresponding to sinigrin (found at m/z 358.02747) and rustoside (found at m/z 579.13611), and their molecular structures are reported in the insets compound of solamargine, the malonyl-solamargine at class of compounds known as flavonoid-3-O -glycosides m/z 954.50525 (C H O N, exact m/z 954.50569). normally derived from horseradish. 48 75 18 Quantitative analysis revealed that black nightshade GLSs exhibited [M−H]¯ as the precursor ion that berries extract contains a high amount of solamargine corresponds to easy deprotonation of the sulphate −1 and solasonine (1.35 and 1.52 g kg dry weight, respec- group. Moreover, the dissociation of [M−H]¯ precursor tively) and a small concentration of other minor known ion yielded abundant product ions, which gave much glycoalkaloids, confirming results obtained by Ventrella information on the structure of the side chain and were et al. . of great value for a correct assignment of known and unknown GLSs. Typical fragments of GLS with nomi- Identification of glucosinolates nal m/z 97, 195, 241, 259, and 275, which correspond The identification of GLSs was based on the study of to the fragment ions HSO4¯, C H O S¯, C H O S¯, 6 11 5 6 9 8 characteristic fragments of these compounds in IRMPD C H O S¯, and C H O S ¯, respectively, were found in 6 11 9 6 11 8 2 MS/MS spectra, and on the measure of accurate the spectrum examined (data not shown). Other char- masses observed using LC/ESI-FTICR/MS, according acteristic fragments, such as [M-80-H]¯, [M-162-H]¯, to Agneta et al. [12, 22]. In Fig. 2, the total ion chro- [M-178-H]¯, [M-196-H]¯, and [M-242-H]¯, were very matogram (TIC) acquired in negative ion mode of a informative for correct molecular identification of horseradish root extract is shown. The qualitative and GLSs . quantitative analyses of this extract confirmed the −1 occurrence of a high amount of sinigrin (2.04 g kg Identification of cannabinoids dry weight), which accounts for more than 90% of the Using optimized reversed-phase liquid chromatogra- total GLS, and of the other 16 GLSs in trace quantity phy (RP-HPLC) coupled to electrospray ionization in [12, 22]. In the inset of Fig. 2, the mass spectrum of the positive mode (ESI ) and Fourier transform ion cyclo- peak corresponding to sinigrin, found at m/z 358.02747 tron resonance (FTICR)/MS, together with tandem (C H NO S , exact m/z 358.02720, error 0.8 ppm) mass spectrometry (MS ) studies performed using 10 17 9 2 is shown. By accurate high-resolution mass analysis, IRMPD and collisional induced dissociation (CID), it the peak eluting at 12.5 min was excluded to be a glu- was possible to separate and quantify four known can- cosinolate, but was found to be rustoside, also known nabinoids (THC, CBD, CBDa, and CBN), useful for as kaempferol 3-lathyroside, which is a member of the the chemotype definition and the classification of C. Intensity (cps) % Relative Abundance % Relative Abundance Lelario et al. Chem. Biol. Technol. Agric. (2018) 5:13 Page 6 of 12 Abbreviation Name Molecular structure CBD Cannabidiol CBDaCannabidiolic acid CBDVCannabidivarin CBN Cannabinol THCΔ9-tetrahydrocannabinol Scheme 1 Common names and molecular structures of cannabinoids detected sativa (Scheme 1). The total ion current (TIC) chro - destined for human consumption or industrial trans- matogram (Fig. 3) revealed the occurrence of three formation (Scheme 1). main cannabinoid peaks assigned to CBD, THC, and The quantification of secondary metabolites (THC, CBDa; in the insets, the mass spectra of these peaks CBD, CBN, CBDa) was performed in parallel through are shown: CBD, found at m/z 315.23159 (C H O , low-resolution mass spectra, selected reaction monitor- 21 30 2 exact m/z 315.23184, error − 0.8 ppm); THC, found at ing (SRM), and high-resolution total ion current (TIC). m/z 315.23148 (C H O , exact m/z 315.23186, error As described in Table 1, the chemovar analysed as 21 30 2 − 1.2 ppm); CBDa, found at m/z 359.22126 (C H O , sample #1 does not exceed the THC limit (0.2%) recom- 22 30 4 exact m/z 359.22169 error − 1.2 ppm). CID and IRMPD mended by the European Union regulations [26, 27], con- fragmentation of precursor ions [M+H] generates firming previous findings . several common species that are diagnostically use- The ([THC] + [CBN])/[CBD] ratio (phenotypic index) ful for establishing their identity (data not shown). The of samples was used to assess the chemical phenotype wide peak at 7.4 min corresponds to cannabidivarin (chemotype) of the specific accession . (CBDV, C H O , m/z [M+H] 287.20056), a non- The high content of cannabidiol (CBD) suggests 19 26 2 psychoactive cannabinoid homologue of CBD with the that the “Eletta campana” accession can be defined as side chain shortened by two methylene bridges; it is not an industrial hemp having a ratio [CBD]/[THC] > 10 useful to determine the chemotype in Cannabis plant (CBD-prevalent chemotype) . In our case, the CBN Lelario et al. Chem. Biol. Technol. Agric. (2018) 5:13 Page 7 of 12 [C H O ] 21 31 2 120 315.23159 [M+H] CBD 135.97108 [C H O ] 22 31 4 [M-H] 359.22126 120 435.06299 [M+H] 20 CBDa 80 100 200 300 400 500 600 m/ z 60 135.97108 [C H O ] 21 31 2 [M-H] 315.23148 435.06299 + [M+H] 100 200 300 400 500 600 THC 60 135.97108 m/ z [M-H] 435.06299 2e+6 100 200 300 400 500 600 m/ z TIC 0246 81012141618202224 Time (min) Fig. 3 Total ion current ( TIC) obtained using LC–ESI-FTICR MS in positive ionization of an ethanol extract obtained from representative flowers of C. sativa accession “Eletta campana”. The insets show the mass spectra corresponding to peaks of CBD (found at m/z 315.23159), THC (found at m/z 315.23148), and CBDa (found at m/z 359.22126). CBN (occurring at 12.3 min) is not of significant concentration (see Table 1). The wide peak at 7.4 min corresponds to cannabidivarin (CBDV, C H O , found at m/z 287.20056) a homologue of CBD with the side chain shortened by two 19 26 2 methylene bridges Table 1 Composition of the two samples of C. sativa analysed: Δ9-tetrahydrocannabinol (THC), cannabidiol (CBD), and (CBN + THC)/CBD ratio identifying the chemotype Samples THC CBD CBN (CBN + THC)/ CBD −1 −1 −1 1 0.021% (2.1 ± 0.5 g kg dw) 0.781% (78.1 ± 1.3 g kg dw) 0.008% (0.81 ± 0.12 g kg dw) 0.037 −1 −1 −1 2 0.050% (5.0 ± 0.8 g kg dw) 0.967% (96.7 ± 1.7 g kg dw) 0.010% (0.98 ± 0.28 g kg dw) 0.061 concentration is not significant for the chemical defini - Antimicrobial activity assays tion of cannabis quality. The antimicrobial activities and MICs were evaluated Data reported in Table 1 for the sample #2 indicate that against selected bacterial strains giving different results the field selection of plant flowers was able to discrimi - depending on the type of plant under observation. nate a group of plants with a higher content of analysed Statistics of the antimicrobial activity data (diame- cannabinoids. ters of ZoI) confirmed that the diameter ranges chosen Relative Abundance % Intensity (cps) 128.93228 [C H O S] 6 11 9 Relative Abundance % 128.93228 [C 32S] H O 6 11 9 Relative Abundance % 128.93228 [C 32S] H O 6 11 9 Lelario et al. Chem. Biol. Technol. Agric. (2018) 5:13 Page 8 of 12 Table 2 Antimicrobial activity and MIC of S. nigrum, A. rusticana, and C. sativa extracts tested a b Bacteria S. nigrum A. rusticana C. sativa Sample #1 C. sativa Sample #2 c −1 c −1 c −1 c −1 ZoI (mm) MIC (mg L ) ZoI (mm) MIC (mg L ) ZoI (mm) MIC (mg L ) ZoI (mm) MIC (mg L ) Gram+ A H C H C H Bacillus cereus (A1I) 15.2 ± 1.0 5 ± 1 NI – 37.0 ± 1.0 5 ± 1 36.5 ± 1.0 5 ± 1 B K C K C H Bacillus thuringiensis (B7I2) 13.5 ± 0.8 10 ± 1 NI – 37.5 ± 0.9 10 ± 1 37.5 ± 0.7 5 ± 1 A K D H D H Bacillus amyloliquefaciens (A5TI) 15.0 ± 0.9 10 ± 1 NI – 34.5 ± 0.9 5 ± 1 34.0 ± 0.9 5 ± 1 Gram− B K Pseudomonas orientalis (A14-1II) 13.5 ± 1.0 10 ± 1 NI – NI NI A K Stenotrophomonas maltophilia 15.0 ± 1.0 10 ± 1 NI – NI NI (B9TIII) Assays were performed in triplicate and results are the mean of three values ± standard deviation NI no inhibition zone was observed Different letters in superscript to numerical data indicate significant differences (p value < 0.05) Sample #1 is composed of flowers representative of the “Eletta campana” cannabis accession Sample #2 was obtained from a field mass selection of “Eletta campana” flowers having a higher THC and CBD content ZoI: diameter of inhibition zone obtained after 24 h of incubation in agar well diffusion assays. ZoI < 10 mm: low antimicrobial activity; 11 < ZoI < 15 mm: middle antimicrobial activity; ZoI > 16 mm: high antimicrobial activity Table 3 Antimicrobial activity and MIC of pure compounds as components of S. nigrum Bacteria Solamargine Solasonine Solamargine/solasonine (1:1 v/v) a −1 a −1 a −1 ZoI (mm) MIC (mg L ) ZoI (mm) MIC (mg L ) ZoI (mm) MIC (mg L ) Gram+ A H A H A H Bacillus cereus (A1I) 15.0 ± 1.0 5 ± 1 14.0 ± 0.9 5 ± 1 15.0 ± 1.0 5 ± 1 B L B M B M Bacillus thuringiensis (B7I2) 12.0 ± 0.9 20 ± 1 12.0 ± 0.8 40 ± 1 13.0 ± 1.0 40 ± 1 B L B M A L Bacillus amyloliquefaciens (A5TI) 12.0 ± 0.9 20 ± 1 12.0 ± 1.0 40 ± 1 14.5 ± 1.0 20 ± 1 Gram− B M B M B M Pseudomonas orientalis (A14-1II) 13.0 ± 0.8 40 ± 1 13.0 ± 0.7 40 ± 1 13.5 ± 1.0 40 ± 1 A K A K A H Stenotrophomonas maltophilia (B9TIII) 15.0 ± 0.9 10 ± 1 14.0 ± 0.8 10 ± 1 15.0 ± 1.1 10 ± 1 Assays were performed in triplicate and results are the mean of three values ± standard deviation Different letters in superscript to numerical data indicate significant differences (p value < 0.05) ZoI: diameter of inhibition zone obtained after 24 h of incubation in agar well diffusion assays. ZoI < 10 mm: low antimicrobial activity; 11 < ZoI < 15 mm: middle antimicrobial activity; ZoI > 16 mm: high antimicrobial activity (< 10 mm; 11–15 mm, > 16 mm) were able to well dis- Both flower samples of C. sativa showed a similar criminate significant differences among the antimicrobial effect on Gram+ bacteria with a high antimicrobial activities (Tables 2, 3, 4). activity; these extracts were more effective against Solanum nigrum and C. sativa extracts demonstrated a B. thuringiensis and B. cereus with 37.5 and 37.0 mm certain antimicrobial activity, while A. rusticana did not diameter of inhibition zone, respectively, while B. reveal any activity against bacteria in this research. amyloliquefaciens was slightly less sensitive (Table 2). The Gram− bacteria, P. orientalis and S. maltophilia, The active extracts of S. nigrum and C. sativa were were sensitive only to the S. nigrum extract, showing a subjected to determine MIC by the agar well diffusion middle inhibition diameter of 13.5 and 15 mm, respec- method against the respective susceptible bacterial tively; moreover, this extract proved a middle antimi- species (Table 2). The results obtained indicated that crobial activity against all Gram+ bacteria (inhibition Gram+ and Gram− bacterial species tested were sensi- zone ranging from 13.5 to 15.2 mm) (Table 2). tive to different extracts in a similar way with an MIC −1 of 5–10 mg L . The more effective extracts were the Lelario et al. Chem. Biol. Technol. Agric. (2018) 5:13 Page 9 of 12 Table 4 Antimicrobial activity and MIC of standard pure compounds as components of C. sativa Bacteria THC CBD CBD/THC a −1 a −1 a −1 ZoI (mm) MIC (mg L ) ZoI (mm) MIC (mg L ) ZoI (mm) MIC (mg L ) Gram+ A B C Bacillus cereus (A1I) 8.5 ± 1.2 > 60 12.0 ± 1.0 > 60 14.5 ± 1.4 > 60 A B C Bacillus thuringiensis (B7I2) 9.5 ± 1.3 > 60 13.0 ± 1.1 > 60 14.5 ± 1.3 > 60 A B C Bacillus amyloliquefaciens (A5TI) 8.0 ± 0.8 > 60 11.0 ± 0.9 > 60 15.0 ± 1.3 > 60 Gram− Pseudomonas orientalis (A14-1II) NI – NI – NI – Stenotrophomonas maltophilia (B9TIII) NI – NI – NI – Assays were performed in triplicate and results are the mean of three values ± standard deviation NI no inhibition zone was observed, THC Δ9-tetrahydrocannabinol, CBD cannabidiol Different letters in superscript to numerical data indicate significant differences (p value < 0.05) ZoI: diameter of inhibition zone obtained after 24 h of incubation in agar well diffusion assays. ZoI < 10 mm: low antimicrobial activity; 11 < ZoI < 15 mm: middle antimicrobial activity; ZoI > 16 mm: high antimicrobial activity two samples of C. sativa with the higher antimicrobial it is distributed widely in nature and is commonly found ability and a low inhibitory concentration (Table 2). in the soil as a saprophytic organism. As a soil bacte- The antimicrobial activity of standard pure com - rium, B. cereus can spread easily to many types of foods ponents of the plants was investigated to understand such as vegetables, eggs, meat, and dairy products, and whether the activity observed in our experiments was is known to cause 2–5% of food-borne intoxications due due to the synergistic action of more than one constitu- to its secretion of emetic toxins and enterotoxins. Food ent in the extracts . poisoning occurs when food is left without refrigera- In the case of S. nigrum, solamargine, solasonine, tion for several hours before it is served. The remaining and the solamargine/solasonine mixture (1:1 v/v) were spores of contaminated food from heat treatment grow tested. All bacteria were sensitive to both components well after cooling and are the source of food poisoning. In with a middle antimicrobial activity ranging from 12 to addition, B. cereus is an opportunistic human pathogen 15 mm (Table 3). and is occasionally associated with infections, causing Among Gram+ bacteria, B. cereus was the most sen- periodontal diseases and other more serious infections −1 sitive (with an MIC of 5 mg L ) compared to the other . The availability of natural substances active towards two species, B. thuringiensis and B. amyloliquefaciens, this microorganism, but well tolerated by the human which were inhibited at higher concentrations (ranging body, could be useful to increase the food storage time. −1 from 20 to 40 mg L ). Gram− bacteria, instead, showed B. thuringiensis is a Gram+, rod shaped, and aerobic the same behaviour in the presence of standard pure spore-forming soil bacterium producing crystalline pro- compounds (Table 3). teins (endotoxins) that have insecticidal properties; on The C. sativa components were able to inhibit only the the other hand, this bacterial species synthesizes several Gram + bacteria tested; THC showed a low antimicrobial enzymes and toxins that give them a wide adaptation to activity, while CBD and the CBD/THC mixture (1:1 v/v) natural habitats . The intrinsic resistance and adapt - proved a middle activity, underlining a stronger effect ability of this bacterium makes it an ideal model for the when the mixture was used (Table 4); nevertheless, the tests performed in this research. B. amyloliquefaciens is a bacterial species appeared not very sensitive to the stand- non-pathogenic Gram+ soil bacterium. Similar to other ard pure components, requiring an inhibitory concentra- Bacillus species, it is capable of producing endospores −1 tion of > 60 mg L . allowing it to survive for extended periods of time. The species also shows some antifungal properties, which Discussion are influenced by environmental nitrogen availability. The antimicrobial activity and MIC were evaluated It synthesizes a natural antibiotic protein active against against selected bacterial strains of significant envi - other photogenic Bacillus spp. and is used in agriculture, ronmental and health concern, used as model of target aquaculture, and hydroponics to fight root pathogens organisms. B. cereus is an endospore-forming Gram+ . For this reason, it was important to verify if it was bacterium that can cause food poisoning. Capable of inhibited by secondary metabolites produced by plants. adapting to a wide range of environmental conditions, P. orientalis is a Gram−, rod-shaped bacterium placed Lelario et al. Chem. Biol. Technol. Agric. (2018) 5:13 Page 10 of 12 in the P. fluorescens group. It shows antagonistic activity researchers investigating the same topic agree that against several phytopathogenic bacteria  and as B. these compounds are, in most cases, slightly more amyloliquefaciens can be inhibited by secondary metab- active against Gram+ than Gram− bacteria. How- olites produced by plants. S. maltophilia is an aerobic, ever, Wilkinson et al.  remarked that some excep- non-fermentative, Gram− bacillus possessing flagella in tions may occur, since Aeromonas hydrophila (Gram−) a multitrichous formation, and is found naturally in the appears to be one of the most sensitive species to the rhizosphere. However, it is also the third most common action of essential oils obtained from different natu - nosocomial pathogen with multi-drug resistance that tar- ral essences (thyme, cinnamon, bay, clove, almond, gets immune-compromised patients in hospitals, making etc.). Moreover, the authors have postulated that indi- it important in medical bacteriology . vidual components of natural extracts exhibit different Antimicrobial trials have demonstrated that black degrees of activity due to their chemical composition, nightshade extract is active on assayed microorganisms. which can vary according to the geographical origin Solamargine and solasonine, main components of black and harvesting period. nightshade extract, were very active also when tested It should also be emphasized that biopharmaceu- individually or as binary mixture. Glycoalkaloids, which ticals and biopesticides may have different action are produced by widely cultivated Solanaceae plants, are mechanisms than those of the conventional synthetic confirmed to be bioactive substances useful for different products, even if the corresponding compounds are applications acting as cellular membrane disrupting fac- similar. A low or inadequate dosage could cause fail- tors or inhibitors of acetylcholinesterase activity [6, 11, ure of protection, which could lead to the abandoning 13]. of natural products in favour of conventional meth- Horseradish (A. rusticana) extract was not active ods. Therefore, using bioactive substances efficiently towards any of the tested microorganisms at assayed requires specific knowledge of the agent and the target concentrations as expected for glucosinolates in the disease for optimizing their application time, doses, absence of the enzymatic reaction needed for the pro- and rates. duction of active derivatives . Hemp (C. sativa) extract showed antimicrobial activ- Conclusion ity only against Gram+ bacteria, as Gram− bacteria In conclusion, the bioactivity of plant extracts to replace seem to be more resistant to the secondary metabolites synthetic chemical products should be assessed case contained in the extract [38, 39]. The different compo - by case, and active substances should be tested as indi- sitions of the samples analysed did not influence their viduals to obtain more extended information on the bioactivity. real applicability of plant metabolites against pathogens. The main components tested, either as individual Promising results were obtained, and glycoalkaloids compounds or as CBD–THC mixture, showed a bio- antimicrobial activity was confirmed herein, in line with activity about three times lower compared to the raw previous reports against insects [6, 11, 13]. But tests per- extract; the rationale behind such a behaviour could formed in vitro are only the first step of a deeper research be that antibacterial properties were due to the syner- aimed at extending the use of natural substances to com- gistic effect of many components (such as terpenoids, bat plant or animal diseases. Further research efforts are carboxylic moieties, and simple or complex phenols) necessary to demonstrate plant secondary metabolite present in the extract even if the prenyl moiety of can- efficiency in the target environments, to better under - nabinoids has been highlighted as effective in antimi - stand their biological activities and to develop actions crobial activity [40–43]. strategies of such complex mixtures usage. The results of the antimicrobial activity of natural extracts against both Gram+ and Gram− microorgan- Abbreviations isms of this study are in agreement with the research CBD: cannabidiol; CBDa: cannabidiolic acid; CBN: cannabinol; CID: collisional work of Tajkarimi et al. . Gram+ bacteria tend to induced dissociation; ESI: electrospray ionization; EUCAST: European Commit- tee for Antimicrobial Susceptibility Testing; GLS: glucosinolate; IRMPD: infrared be more sensitive to the antimicrobial properties of multiple photon dissociation; LC–FTICR/MS: liquid chromatography–Fourier natural extracts [45, 46], while Gram− are less suscep- transform ion cyclotron resonance/mass; MIC: minimal inhibitory concentra- tible to the antibacterial action of natural substances tion; PCb: potato and carrot broth; RP-HPLC: reversed-phase liquid chromatog- raphy; SRM: selected reaction monitoring; THC: Δ9-Tetrahydrocannabinol; TIC: since they possess an outer membrane surrounding total ion current; USAE: ultrasound-assisted extraction; ZoI: zone of inhibition. the cell wall, which restricts diffusion of hydrophobic compounds through their lipopolysaccharide cover- Authors’ contributions All authors of this research paper have directly participated in the plan- ing. In addition, Gram− microorganisms generally ning, execution, or analysis of this study and read and approved the final present higher MICs than the Gram+ ones. Numerous manuscript. FL performed analytical MS determinations and wrote the relative Lelario et al. Chem. Biol. Technol. Agric. (2018) 5:13 Page 11 of 12 paragraphs in “Methods”. MGB carried out microbial assays and wrote the bioinsecticidal activity of Solanaceae alkaloids. Toxins. 2016;8(3):60. https relative paragraphs in “Methods”. SM prepared crude extracts and surrogates ://doi.org/10.3390/toxin s8030 060. and wrote the relative paragraphs in “Methods”. SDeF selected all plant species 7. Croteau R, Kutcahn TM, Lewis NG. Natural products. In: Buchanan B, Gruis- tested for the experiments and provided the registration of plants’ materials. sem W, Jones R, editors. Biochemistry and molecular biology of plants. GS discussed the bioactivity results. LM, SAB, and LS planned this research and Rockville: American Society of Plant Physiologists; 2000. p. 1250–318. discussed a part of the results. LS and SAB revised the whole manuscript. All 8. Osbourn AE. Saponins and plant defense—a soap story. Trends Plant Sci. authors read and approved the final manuscript. 1996;1:4–9. 9. Braga LC, Leite AAM, Xavier KGS, Takahashi JA, Bemquerer MP, Chartone- Author details Souza E, Nascimento MAA. Synergic interaction between pomegranate Department of Sciences, University of Basilicata, 85100 Potenza, Italy. extract and antibiotics against Staphylococcus aureus. Can J Microbiol. Department of European Cultures (DICEM), University of Basilicata, 2005;51(7):541–7. https ://doi.org/10.1139/w05-022. 85100 Potenza, Italy. 10. Cataldi TRI, Lelario F, Bufo SA. Analysis of tomato glycoalkaloids by liquid chromatography coupled with electrospray ionization tandem mass Acknowledgements spectrometry. Rapid Commun Mass Spectrom. 2005;19(21):3103–10. We are grateful to the Company “Eletta campana S.r.l.” for providing the flower 11. Adamski Z, Halamunda J, Marciniak P, Nawrocka M, Ziemnicki K, Lelario samples of their registered C. sativa accession. F, Scrano L, Bufo SA. Eec ff t of various xenobiotics on hatching success of We would like to thank the Herbarium Lucanum for the prompt supply Spodoptera exigua eggs as compared to a natural plant extract. J Toxicol of the voucher codes relating to the deposited plants of S. nigrum and A. Environ Health Part A. 2009;72:1132–4. rusticana. 12. Agneta R, Rivelli AR, Ventrella E, Lelario F, Sarli G, Bufo SA. Investigation of glucosinolate profile and qualitative aspects in sprouts and roots of horseradish (Armoracia rusticana) using LC–ESI–hybrid linear ion trap with Competing interests Fourier transform ion cyclotron resonance mass spectrometry and infra- The authors declare that they have no competing interests. red multiphoton dissociation. J Agric Food Chem. 2012;60(30):7474–82. 13. Ventrella E, Marciniak P, Adamski Z, Rosiński G, Chowański S, Falabella P, Availability of data and materials Bufo SA. Cardioactive properties of Solanaceae plant extracts and pure Additional data may be available on request to the authors; please contact the glycoalkaloids on Zophobas atratus. Insect Sci. 2015;22(2):251–62. corresponding author. 14. Bianco G, Lelario F, Battista FG, Bufo SA, Cataldi TRI. Identification of glucosinolates in Capers by LC–ESI using a hybrid linear ion trap with Consent for publication Fourier-transform ion cyclotron resonance mass spectrometry (LC–ESI- The authors agreed to the publication of the manuscript in this journal. LTQ-FTICR-MS) and infrared multiphoton dissociation. J Mass Spectrom. The authors grant the accessibility of the scientific article in agreement of 2012;47(9):1160–9. https ://doi.org/10.1002/jms.2996. the springer policy. 15. Cataldi TRI, Lelario F, Orlando D, Bufo SA. Collision-induced dissociation of the A+ 2 isotope ion facilitates glucosinolates structure elucidation by electrospray ionization-tandem mass spectrometry with a linear quadru- Ethics approval and consent to participate pole ion trap. Anal Chem. 2010;82(13):5686–96. The authors declare that this study does not involve human subjects, human 16. Barbieri R, Tedeschi P. Eletta Campana e T4, nuove cultivar di canapa per material, and human data. l’ambiente campano. Sementi Elette. 1968;14(6):412–7. This manuscript is an original article and has not published in other jour- 17. De Meijer EPM. Fibre hemp cultivars: a survey of origin, ancestry, nals. The authors agree to complying with the copyright rule. availability and brief agronomic characteristics. J Int Hemp Assoc. 1995;2(2):66–73. 18. Kim HK, Verpoorte R. Sample preparation for plant metabolomics. Phyto- Funding chem Anal. 2010;21(1):4–13. BIOMON Project, granted by SAIPEM S.p.A. 19. Mandal V, Mohan Y, Hemalatha S. Microwave assisted extraction: an innovative and promising extraction tool for medicinal plant research. Pharmacogn Rev. 2007;1(1):7–18. Publisher’s Note 20. European Committee for Antimicrobial Susceptibility Testing (EUCAST ) Springer Nature remains neutral with regard to jurisdictional claims in pub- of the European Society of Clinical Microbiology and Infectious Diseases lished maps and institutional affiliations. (ESCMID). Determination of minimum inhibitory concentrations (MICs) of antibacterial agents by broth dilution. Clin Microbiol Infect. 2003;9(8):1–7. Received: 22 February 2018 Accepted: 24 May 2018 https ://doi.org/10.1046/j.1469-0691.2003.00790 .x. 21. Mann CM, Markham JL. A new method for determining the mini- mum inhibitory concentration of essential oils. J Appl Microbiol. 1998;84:538–44. 22. Agneta R, Lelario F, De Maria S, Möllers C, Bufo SA, Rivelli AR. Glucosi- References nolate profile and distribution among plant tissues and phenological 1. Newman DJ, Cragg GM, Snader KM. Natural products as sources of new stages of field-grown horseradish. Phytochemistry. 2014;106:178–87. drugs over the period 1981–2002. J Nat Prod. 2003;66(7):1022–37. 23. Lelario F, Bianco G, Bufo SA, Cataldi TRI. Establishing the occurrence of 2. Cragg GM, Kingston DG, Newman DJ. Anticancer agents from natural major and minor glucosinolates in Brassicaceae by LC–ESI hybrid linear products. Boca Raton: CRC Press; 2011. ion-trap and Fourier-transform ion cyclotron resonance mass spectrom- 3. McChesney JD, Venkataraman SK, Henri JT. Plant natural products: back etry. Phytochemistry. 2012;73(1):74–83. to the future or into extinction? Phytochemistry. 2007;68(14):2015–22. 24. Russo D, Bonomo MG, Salzano G, Martelli G, Milella L. Nutraceutical 4. Tripathi P, Dubey NK. Exploitation of natural products as an alternative properties of Citrus clementina juices. Pharmacologyonline. 2012;1:84–93. strategy to control postharvest fungal rotting of fruit and vegetables. 25. Lelario F, Labella C, Napolitano G, Scrano L, Bufo SA. Fragmentation Postharvest Biol Technol. 2004;32(3):235–45. https ://doi.org/10.1016/j. study of major spirosolane-type glycoalkaloids by collision-induced posth arvbi o.2003.11.005. dissociation linear ion trap and infrared multiphoton dissociation Fourier 5. Agostini-Costa TD, Vieira RF, Bizzo HR, Silveira D, Gimenes MA. Secondary transform ion cyclotron resonance mass spectrometry. Rapid Commun Metabolites. In: Sasikumar D, editor. Chromatography and its applications. Mass Spectrom. 2016;30(22):2395–406. London: InTech; 2012. https ://doi.org/10.5772/35705 . 26. Commission Regulation (EEC) No 1164/89 of 28 April 1989, laying 6. Chowński S, Adamski Z, Marciniak P, Rosiński G, Büyükgüzel E, Büyük- down detailed rules concerning aid for fibre flax and hemp. OJ L 121, güzel K, Falabella P, Scrano L, Ventrella E, Lelario F, Bufo SA. A review of 29.4.1989:4–10. Lelario et al. Chem. Biol. Technol. Agric. (2018) 5:13 Page 12 of 12 27. Commission Regulation (EC) No 2860/2000 of 27 December 2000 38. Van Klingeren B, Ten Ham M. Antibacterial activity of amending Regulation (EC) No 2316/1999 laying down detailed rules for Δ9-tetrahydrocannabinol and cannabidiol. Antonie Van Leeuwenhoek. the application of Council Regulation (EC) No 1251/1999 establishing a 1976;42(1–2):9–12. support system for producers of certain arable crops, to include flax and 39. Gershenzon J, Dudareva N. The function of terpene natural products in hemp grown for fibre, specifying the rules on set-aside areas and amend- the natural world. Nat Chem Biol. 2007;3(7):408–14. ing the base areas for Greece and Portugal. 40. Ali EM, Almagboul AZ, Khogali SM, Gergeir UM. Antimicrobial Activity 28. Avico U, Pacifici R, Zuccaro P. Variations of tetrahydrocannabinol content of Cannabis sativa L. Chin Med. 2012;1:61–4. https ://doi.org/10.4236/ in cannabis plants to distinguish the fibre-type from drug-type plants. cm.2012.31010 . Bull Narc. 1985;37(4):61–5. 41. Appendino G, Gibbons S, Giana A, Pagani A, Grassi G, Stavri M, Rahman 29. De Meijer EPM, Van Soest LJM. The CPRO Cannabis germplasm collection. MM. Antibacterial cannabinoids from Cannabis sativa: a structure-activity Euphytica. 1992;62(3):201–11. study. J Nat Prod. 2008;71(8):1427–30. https ://doi.org/10.1021/np800 30. De Meijer EP, Bagatta M, Carboni A, Crucitti P, Moliterni VC, Ranalli P, 2673. Mandolino G. The inheritance of chemical phenotype in Cannabis Sativa 42. Pollastro F, Taglialatela-Scafati O, Allara M, Munoz E, Di Marzo V, De L. Genetics. 2003;163(1):335–46. Petrocellis L, Appendino G. Bioactive prenylogous cannabinoid from 31. Russo EB, Taming THC. Potential cannabis synergy and phyto- fiber hemp (Cannabis Sativa). J Nat Prod. 2011;74(9):2019–22. https ://doi. cannabinoid-terpenoid entourage effects. British J Pharmacol. org/10.1021/np200 500p. 2011;163(7):1344–64. 43. Docimo T, Consonni R, Coraggio I, Mattana M. Early phenylpropanoid bio- 32. Bottone EJ. Bacillus cereus, a volatile human pathogen. Clin Microbiol Rev. synthetic steps in Cannabis sativa: link between genes and metabolites. 2010;23(2):382–98. https ://doi.org/10.1128/CMR.00073 -09. Int J Mol Sci. 2013;14(7):13626–44. https ://doi.org/10.3390/ijms1 40713 33. Keshavareddy G, Kumar ARV. Bacillus thuringiensis. In: Omkar O, editor. 626. Ecofriendly pest management for food security. London: Academic Press; 44. Tajkarimi MM, Ibrahim SA, Cliver DO. Antimicrobial herb and spice 2016. p. 443–73. https ://doi.org/10.1016/B978-0-12-80326 5-7.00014 -2. compounds in food. Food Control. 2010;21(9):1199–218. https ://doi. 34. Priest F, Goodfellow M, Shute L, Berkeley R. Bacillus amyloliquefaciens sp. org/10.1016/j.foodc ont.2010.02.003. nom., nom. rev. Int J Syst Bacteriol. 1987;37:69–71. 45. Compean KL, Ynalvez RA. Antimicrobial activity of plant secondary 35. Dabboussi F, Hamze M, Elomari M, Verhille S, Baida N, Izard D, Leclerc metabolites: a Review. Res J Med Plants. 2014;8:204–13. https ://doi. H, et al. Taxonomic study of bacteria isolated from Lebanese spring org/10.3923/rjmp.2014.204.213. waters: proposal for Pseudomonas cedrella sp. nov. and P. orientalis sp. 46. Sasso S, Scrano L, Bonomo MG, Salzano G, Bufo SA. Secondary nov. Res Microbiol. 1999;150(5):303–16. https ://doi.org/10.1016/S0923 metabolites: applications on cultural heritage. Commun Appl Biol Sci. -2508(99)80056 -4. 2013;78(2):101–8. 36. Ryan RP, Monchy S, Cardinale M, Taghavi S, Crossman L, Avison MB, Berg 47. Wilkinson JM, Hipwell M, Ryan T, Cavanagh HMA. Bioactivity of Back- G, van der Lelie D, Dow JM. The versatility and adaptation of bacteria housia citriodora: antibacterial and antifungal activity. J Agric Food Chem. from the genus Stenotrophomonas. Nat Rev Microbiol. 2009;7:514–25. 2003;51(1):76–81. https ://doi.org/10.1021/jf025 8003. 37. Sotelo T, Lema M, Soengas P, Cartea ME, Velasco P. In vitro activity of glu- cosinolates and their degradation products against brassica-pathogenic bacteria and fungi. Appl Environ Microbiol. 2015;81(1):432–40.
Chemical and Biological Technologies in Agriculture
– Springer Journals
Published: Jun 5, 2018