Rapid Bioassay-Guided Isolation of Antibacterial Clerodane Type Diterpenoid from Dodonaea viscosa (L.) Jaeq.

Plant extracts are complex matrices and, although crude extracts are widely in use, purified compounds are pivotal in drug discovery. This study describes the application of automated preparative-HPLC combined with a rapid off-line bacterial bioassay, using reduction of a tetrazolium salt as an indicator of bacterial metabolism. This approach enabled the identification of fractions from Dodonaea viscosa that were active against Staphylococcus aureus and Escherichia coli, which, ultimately, resulted in the identification of a clerodane type diterpenoid, 6β-hydroxy-15,16-epoxy-5β, 8β, 9β, 10α-cleroda-3, 13(16), 14-trien-18-oic acid, showing bacteriostatic activity (minimum inhibitory concentration (MIC) = 64–128 µg/mL) against test bacteria. To the best of our knowledge, this is the first report on antibacterial activity of this metabolite from D. viscosa.


Introduction
The global presence and rise of antibiotic resistant pathogens and decline in antibiotic drug discovery programs by pharmaceutical companies are prompting the scientific community to search for new and also re-examine old sources of bioactive chemicals in order to identify potential drugs [1].Medicinal plants are an area under focus since their secondary metabolites encompass a significant number of drugs used in current therapeutics and there is no doubt in their potential as the source of new medicines [2].
Dodonaea viscosa (L.) Jaeq. is an evergreen shrub in Pakistan and is locally known as Ghawraskay (Pushto).It is traditionally used to treat a wide range of medical conditions from colds to malaria [3,4], and crude extracts have been shown to be active against bacterial, viral, fungal and protozoal pathogens [5][6][7].However, despite numerous phytochemical investigations, only a single compound, hautriwaic acid, with moderate antibacterial activity has been characterized [8].
Isolation and characterization of bioactive molecules from crude plant extracts or fractioned extracts are challenging tasks.This is mainly due to the presence of a large number of metabolites with close physicochemical properties with often complementing bioactivities.However, in order to standardize them, their separation is necessary.Chromatographic methods have revolutionized the separation of such complex mixtures and the combination of chromatography with spectroscopic and spectrophotometric methods such as mass spectroscopy, UV, NMR etc. has further eased the task.In this context, preparative HPLC has shown to be an essential tool for producing high quality natural products and if strict quality control is undertaken, the approach ensures reliable screening results [9,10].
Currently, cell viability assays make use of one of several tetrazolium salts in a microplate format.Metabolically active cells reduce the tetrazolium salts and result in coloured formazan products that can be simply read by a plate reader, thus offering rapid records on the metabolic status of the cell [11].XTT (2,3-bis(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5-carboxanilide) is a tetrazolium salt that does not require the additional solubilisation step as required in the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay and yields a reduced formazan product that is water soluble [11].The XTT assay is exploited in determination of antimicrobial activity of simple plant extracts [12,13].This study describes the combination of automated preparative HPLC in combination with an XTT bioassay for rapid identification of antibacterial fractions and derived active compound from D. viscosa.

Fractionation of Plant Material
Liquid-liquid fractionation of the crude aqueous extract of D. viscosa resulted in 75 g of an n-hexane fraction (H), 70 g of a dichloromethane fraction (D), 92 g of an ethyl acetate fraction (E), 81 g of an n-butanol fraction (B) that spontaneously solidified, and 126 g of thick suspension of an aqueous fraction (A).

Preliminary Antibacterial Screening
Extracts of D. viscosa derived from hexane and ethyl acetate were active against E. coli (the National Collection of Industrial, Marine and Food Bacteria (NCIMB) 8797) and S. aureus (NCIMB 6571), whereas the butanol extract was only active against S. aureus (Table 1).Previous studies on D. viscosa fractions showed bacteriostatic activity against Gram positive bacteria, but not Gram negative such as E. coli [5,6]; our results indicated that the hexane and ethyl acetate fractions inhibited E. coli.These opposite findings may be due to differences in the plant material, extraction protocols and test strains used.As there is growing interest in the generation of libraries of simplified extracts prior to screening [9,10], our approach aimed to select those extracts which exhibited antibacterial activity and use preparative HPLC with the XTT assay for rapid screening of a large number of fractions.

Optimization of XTT (2,3-Bis(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5carboxanilide) Bioassay
Ciprofloxacin (50 µg/mL) and clarithromycin (25 µg/mL) inhibited the test strains of E. coli and S. aureus, respectively, after 24 h incubation in the XTT assay, as shown by the decreased absorbance.The effect of ciprofloxacin on E. coli was evident after 2 h incubation, with absorbance in the treated cells at 0.042 compared to the solvent control at 0.098.However, 24 h incubation was necessary to obtain an evident result on the effect of clarithromycin on S. aureus; therefore 24 h incubation was selected as incubation period.

Generic Preparative HPLC-XTT Bioassay
The use of solvents such as methanol and water along with a volume-based fractionation strategy was necessary as the active component(s) may not contain a chromophore.For the extract, a total of 52 primary fractions were collected and each fraction was evaluated against the two test bacteria using the XTT assay.No bioactivity was detected in the fractions from the ethyl acetate extract of D. viscosa, despite the activity against both bacteria reported in disk diffusion assay (Table 1), and they were not investigated further.The butanol extract of D. viscosa yielded three fractions (18, 19 and 20) with moderate antibacterial activity, i.e., 50%-75% reduction in bacterial growth.In contrast, six fractions from the hexane extract inhibited S. aureus, with this antibacterial activity resulting in higher than 75% reduction in bacterial activity (Figure 1).S1.
Figure 2 shows the data from the preparative HPLC and the results of XTT assay, correlating with the active fractions to the corresponding peaks in the chromatogram and indicating the hydrophobic components of the fractions.Even if fractions with antibacterial activity were clearly evident, the majority of fractions enhanced bacterial viability (fractions with no inhibition), most likely due to additional carbon sources.

Focused Gradients-XTT Bioassay
Nine bioactive fractions from D. viscosa were further purified using focused gradients and each of the separated fractions were screened by the rapid XTT assay.The five primary fractions from the

Percentage inhibtion in E. coli growth after 24 h
hexane extract from D. viscosa resulted in 22 fractions, which inhibited growth of S. aureus, and out of which 18 were as active as the positive control clarithromycin.
The combination of preparative HPLC with the XTT assay enabled the rapid isolation of bioactive fractions from plant extracts.In addition, after purification, information on purity and physicochemical properties of bioactive fractions were obtained by LC/MS analysis.Data from the XTT assay (Figure 2) and LC/MS analysis (Figure 3) led to focusing on one compound from D. viscosa.Though this information alone was not sufficient for full characterization of the compound, it provided essential information for monitoring, purifications and obtaining higher amounts of compounds, thus avoiding unnecessary bioassays.
The prime objective of these combination bioassays was to identify fractions with antibacterial activity.At the preliminary stages (fractions and sub-fractions level) quantification was not considered necessary with the reason that synergism and/or potentiation may exist between the respective metabolites that exist in a particular fraction, therefore, during initial screening, fractions were analysed on a qualitative basis.Activity was quantified when a purified bioactive metabolite was obtained.

MIC (Minimum Inhibitory Concentration) and MBC (Minimum Bactericidal Concentration) of Compound 1
The antibacterial activity of the compound 1 was quantified by evaluating the MIC (minimum inhibitory concentration) and MBC (minimum bactericidal concentration) against S. aureus (NCIMB 6571) and E. coli (NCIMB 8797).The results indicated a higher activity against S. aureus compared to E. coli, thereby suggesting it to be more effective against Gram-positive bacteria (Table 3 and Figure 5).The compound was crystalline, white coloured and soluble in 10% (v/v) DMSO in PBS solution and, hence, no color interference was reported with the XTT assay.A few Clerodane furanolactones and diterpenoid derivatives were previously identified in the liverwort Scapania nemorae [15], Solidago spp.[16] and in Aristeguetia spp.[17].However, to the best of our knowledge, this is the first report on its antibacterial activity.Previously, five clerodane diterpenoids isolated from Pulicaria wightiana showed antibacterial activity against Bacillus subtilis, B. sphaericus, S. aureus, Klebsiella aerogenes and Chromobacterium violaceum [18].The results obtained from the bio-guided isolation of furanyl clerodanoic acid in D. viscosa, as well as the antimicrobial activity of hautriwaic acid, a D. viscosa component, reported to be active against B. subtilis, S. aureus and E. coli [8], support the possible use of this plant for the treatment of infectious diseases in the traditional systems of medicines.The application of the XTT bioassay in identification of bioactive metabolite and quantification of its antibacterial activity using standard test bacteria provides an approach that may be utilized in similar kinds of experiments.Since, in case of plant extracts, the commonly utilized assays (agar well diffusion and disk diffusion assays) that only highlight the antimicrobial activity of extracts or their fractions do not demonstrate the actual potential of that particular plant, more detailed standardized assays are essential.Moreover, the potential of isolated compounds can be easily quantified through MIC assays using the XTT bioassay-based approach.The isolated compound belongs to isoprenoids that are known to possess antimicrobial activity.In general, Gram-positive bacteria are more susceptible to these compounds, as confirmed by our results.Putative mechanisms of action may include cell membrane disruption, alteration in the cell membrane fluidity and permeability, increased susceptibility to antibiotics, and disturbances in the respiration chain due to modification of membrane-bound protein localization [19].Moreover, isoprenoids may also be considered for adjuvant therapy with conventional antimicrobials, in order to reduce, delay or impair antibiotic resistance.Studies on antibiotic activity of clerodane type diterpenoid on other microorganisms are still in progress, and further investigations on its possible mechanism(s) of action will be planned and designed.

Bacterial Strains
Preliminary screening and XTT bioassays were carried using E. coli (NCIMB 8797) and S. aureus (NCIMB 6571) obtained from the National Collection of Industrial, Marine and Food Bacteria (NCIMB; Aberdeen, UK).

Preparative HPLC
Preparative HPLC was performed using a Biotage Parallex Flex (Biotage, Cardiff, UK) where Flex V3 software was used for instrument control and data acquisition.Separations were performed on a Hyperprep HS C18 column (10 mm I.D. (internal diameter) × 150 mm long; 8 µm particle size; Thermo scientific, Cheshire, UK) with a pre column (KR 100-13 C18-10 CP; Hichrom Ltd., Berkshire, UK).Mobile phase included Milli-Q water (A), and methanol (B).The crude samples were separated using a gradient increasing from 10% to 100% B over 50 min at a flow rate of 10 mL/min.The concentration of crude fractions was 0.2 g/mL, and each injection volume was 0.25 mL.Eluent was monitored at 220 and 254 nm and fractionation was volume-based, collecting 9 mL fractions into deep well microtitre plates (24 × 10 mL; Whatman, Kent, UK).The butanol extract of D. viscosa had three fractions (18, 19 and 20) with moderate antibacterial activity (50%-75% reduction in bacterial growth), while six fractions from the hexane extract inhibited only S. aureus (>75% reduction in bacterial activity) (Figure 2 and Table 2).Using the data available on the percentage of methanol eluting the bioactive fractions, shorter (30 min) focused gradients were selected for further purification of all the primary bioactive fractions, as summarized for each fraction in Table 4.

Focused Separation of Bioactive Fractions
Fractions from the generic separations that exhibited antibacterial activity were dried, weighed and re-dissolved in 0.5 mL methanol.These were further purified using polarity focused gradients based on the proportion of methanol in which they were eluted during the generic gradient (Table 2).Column and solvents were as described in Section 3.4.1 with an injection volume of 0.25 mL.Fraction collection was based on volume (9 mL fraction).
A total of nine bioactive fractions from D. viscosa were further purified using focused gradients over 30 min.Each separation resulted in 35 fractions which were screened by the rapid XTT assay (Figure 2).The five primary fractions from the hexane extract from D. viscosa resulted in 22 fractions (Table 5) which inhibited S. aureus, out of 18 as active as to the positive control clarithromycin, indicating the potential of this plant as a source of new drugs or drug scaffolds.Slight or no activity was observed in E. coli (Figure 2).The HPLC system consisted of a Waters Alliance 2695 with a 2996 photodiode array detector (PDA) and a ZQ 2000 mass spectrometer (Elstree, Hertfordshire, UK).A Sunfire C18 column (Length 150 mm × 2.1 mm I.D., particle size 5 µm) which was maintained at 40 °C was used for separation.Mobile phase was Milli-Q water (A) and acetonitrile (B) both containing 0.05% TFA.A gradient increasing from 5% to 100% B over 30 min at a flow rate of 0.3 mL/min was used for separation of samples (10 µL).Eluent was monitored from 210 to 400 nm with a resolution 1.2 nm and by positive ion electrospray (ESI+) in series, scanning from m/z 100 to 1600 with a scan time of 2 s and inter-scan delay of 0.1 s.Ion source parameters: Sprayer voltage, 3.07 KV; cone voltage, 40 V; desolvation temperature, 300 °C; source temperature, 100 °C.Instrument control, data acquisition and processing were done using MassLynx v4.1 (Waters, Elstree, Hertfordshire, UK).The active component in the n-hexane extract of D. viscosa was found in H42 (Figure 1) and sub-fraction 12 (Figure 2), where the relative purity (PDA-210 to 400 nm) achieved was >90% (Figure 3).16), 14-trien-18-oic acid: Crystalline solid (20 mg): UV (λmax, MeOH) = 217.7 nm.HR-EIMS m/z = 332.1879(calcd.for C20H28O4, 332.2013).The 1 H-NMR (400 MHz) and 13 C-NMR (100 MHz) were recorded in MeOD-d4 on a BRUKER Ultrashield 400 (Bruker UK Limited, Coventry, UK) (Table 2).

Preliminary Antibacterial Screening
Crude extracts (16.7 mg/mL) were re-suspended in methanol and sterilized using Millex, 33 mm filter (0.22 µm pore size; Millipore, Watford, UK).Aliquots (6 µL) were applied to sterile disks (6 mm diameter punched from Whatman No. 1 filters).Negative control disks had same volume of methanol on each disk.Ciprofloxacin (10 µg/disk) and clarithromycin (15 µg/disk) served as the positive controls for E. coli NCIMB 8797 and S. aureus NCIMB 6571, respectively.The bacteria were kept on nutrient agar and fresh standardized inocula, turbidity equivalent to 0.5 McFarland (≈1.5 × 10 8 CFU/mL) [21], were used for each experiment.Disks containing the crude extracts, negative control disks (methanol) and positive controls were aseptically placed on the seeded plates.Prior to overnight incubation at 37 °C, seeded plates with disks were kept in inverted position in refrigerator for 2 h to allow adequate diffusion of extracts.All the tests were performed in triplicate.

XTT Bioassay
XTT solution (1 mg/mL) in PBS was prepared and sterilized using Millex, 33 mm filter (0.22 µm pore size; Millipore, Watford, UK) and stored at −80 °C in cryotubes (Cole-Parmer, London, UK) until use.Menadione (173 μg/mL) in acetone was prepared fresh before to every assay.XTT aliquots were thawed at room temperature and, then, menadione was added to XTT (1:12).From each fraction collected during prep-HPLC, an aliquot (300 µL) was placed in glass vials (2 mL; Alltech, Carnforth, UK) and dried overnight in a laminar flow cabinet and re-suspended in 10% v/v DMSO/PBS (400 µL) [22].Samples (100 µL) were transferred into the wells of 96-well flat bottom microtiter plates (Sterilin, Aberbargoed, UK).E. coli NCIMB 8797 and S. aureus NCIMB 6571 inocula were prepared as mentioned earlier.Inoculum (100 µL) of each test bacteria was added to each of the test well, into negative control (DMSO/PBS) and positive control (ciprofloxacin for E. coli, and clarithromycin for S. aureus) wells.XTT/menadione mixture (18 µL) was added to each test well and plates were gently shaken.Absorbance at 490 nm was recorded using a BioTek microplate reader (Fisher, Loughborough, UK) at the start of the test and after overnight incubation at 37 °C.All tests were performed in duplicate.Percent increase in absorbance (X) was calculated using formula 1, where AtN and At0 represented mean absorbance values after 24 h incubation and at the start of test, respectively.Similarly, percentage increase in negative control (N) was also calculated using Equation (1).Percent reduction (R) in cell metabolism was calculated using Equation (2), where XN and TN represented percent increase in absorbance of test material and negative control, respectively, at time (N) which was standardized at 24 h.
Fractions were considered to have an antibacterial effect when they gave a mean (n = 2) reduction in the XTT colorimetric bioassay of ≥50% (threshold level for antibacterial activity), and <50% value was termed to have slight or no antibacterial activity [13,23].

Determination of Minimum Inhibitory Concentration (MIC) of Purified Compound
MIC is defined as the lowest antimicrobial concentration that completely inhibits visible bacterial growth, as detected visually or via an automated or semiautomated method.One hundred µL of purified compound solution (256 µg/mL) in DMSO/PBS was transferred aseptically to the microtiter plates and, sequentially, half diluted in sterile solution of DMSO/PBS such that final dilution of 2 µg/mL was achieved.Then 100 µL of standardized inoculum (≈1.5 × 10 8 CFU/mL) was added to each of the test wells and into negative control (DMSO/PBS) and positive control (ciprofloxacin for E. coli and clarithromycin for S. aureus) containing wells.Then, to each test well, 18 µL of XTT/menadione reagent were added and plates were gently shaken.Absorbance at 490 nm was recorded using microplate reader at the start of the test and, then, after 24 h of incubation at 37 °C.The tests were run in duplicate.The wells having no visible bacterial growth, detected visually and having no change in absorbance before and after 24 h of application of test defined the MIC of the compound.

Determination of Minimum Bactericidal Concentration (MBC) of Purified Compound
The MBC was defined as the concentration of the antimicrobial agent that results in the 99.9% reduction in CFU/mL, compared with the organism concentration in the original inocula.The plates used for the evaluation of MIC were utilized such that, after 24 h incubation, the wells that were visibly clear and showed no metabolic activity of bacteria, ascertained from the absorbance and % reduction in bacterial population readings (evaluated using formulas given in Section 3.5.2.), were selected and 50 µL aliquots from each well were spread separately on the entire surfaces of sterile

Table 1 .
Antibacterial activity (average zones of inhibitions in mm ± SD) of crude solvent fractions determined by disk diffusion assay.

Table 3 .
MIC (minimum inhibitory concentration) and MBC (minimum bactericidal concentration) of compound 1 isolated from D. viscosa against S. aureus and E. coli.
10% (v/v) DMSO in PBS served as negative control.

Table 5 .
Summary of bioactive sub-fractions from focused preparative HPLC runs of fractions from D. viscosa hexane fraction.