Two new secondary metabolites with antibacterial activities from Conyza aegyptiaca (Asteraceae)

Abstract The bio guided fractionation of the dichloromethane/methanol (1:1) crude extract of the air-dried whole plant of C. aegyptiaca led to the isolation of one new flavone derivative designated conyflavone (1) and one new clerodane diterpene type designated conyclerodane (2) along with five known compounds including two flavonoids Gardenin C (3), chrysosplenetin (4) and two steroids glucoside of β-sitosterol (5), the mixture of stigmasterol (6) and β-sitosterol (6′) and ent-2b,18,19trihydroxycleroda-3,13-dien-16,15-olide (7). The structures were established by spectroscopic methods including IR, 1D and 2D NMR in conjunction with mass spectroscopy and by comparison to data of related compounds described in literature. The stereocentres in compound 2 were determined by SC-XRD analysis. Crude extract as well as fractions and pure compounds were evaluated in vitro for their antibacterial activities against four pathogenic and two clinical isolate strains using microdilution methods. Extracts and compounds displayed a moderate antibacterial activity with MIC values ranging from 125 to 500 µg/mL. Graphical Abstract


Introduction
Infectious diseases are the primary cause of mortality in much of the developing countries and, the most important cause of human death worldwide (Yunlei et al. 2020). Invasive infection such as endocarditis, osteomyelitis, bacteremia, and lethal pneumonia bloodstream, urethra, vagina, otitis and sepsis, etc … are caused by Staphylococcus aureus together with Escherichia coli and Shigella flexneri that are bacteria mostly found in health care facilities and in the community (Lowy 1998). The easy propagation of infections due to these bacteria and multidrug resistant strains makes the clinical anti-infective treatments more difficult (Yunlei et al. 2020). Therefore, search of new and safe therapeutic agents and strategies are welcome in order to reduce the incidence of bacterial infections. Low-income people and many ethnic groups in Cameroon and native communities in developing countries use folk medicine for their health problems (Mohammad et al. 2011). Thus medicinal plants such as Conyza are taken as decoction, tea, infusion and so on within this community to faith these diseases. The genus Conyza, belonging to the Asteraceae family, comprises about 50 species, which are found in the tropical and sub-tropical regions of Africa, tropical Asia and Australia . The plants of this genus are used in folk medicine as an anthelmintic, a body-wash for convalescents, a soothing for skin diseases (Burkill 1985) and for their antimicrobial and insecticidal activities (Harraz et al. 2015). Conyza aegyptiaca is also used for the treatment of type 2 diabetes by African populations and also prevents hepatic glucose release and possess anti-hyperglycemic properties (Baï et al. 2019). Previous chemical studies of Conyza genus led to the isolation of secondary metabolites such as; diterpenes (Zdero et al. 1990;Su et al. 2008;Bitchagno et al. 2021), triterpenes (Hammouda et al. 1978;Metwally 1989;Banday et al. 2013), sesquiterpenes (Metwally and Dawidar 1984;Zdero et al. 1990), and flavonoids (El-Karemy et al. 1987), some of which display interesting pharmacological activities including antiviral, antimicrobial (Anani et al. 2000), anti-oxidant, antitumor, antimalarial and anti-inflammatory properties. As part of our continuous search for new antibacterial therapeutic agents in one hand, in view to develop safe, efficient and low cost phyto-medicine able to be used by our population on the other hand. Conyza aegyptiaca was chemicaly and pharmacologicaly investigated. Here in, are reported the results of our investigation.

Results and discussion
The bio-guided fractionation monitory by bacterial essays, of the dichloromethane/ methanol crude extract of the air-dried whole plant of C. aegyptiaca by flash chromatography was carried out using n-hexane, EtOAc and methanol as eluent successively. The crude extract exhibited a weak antibacterial potential against S. aureus, Shigella flexneri and E. coli (MIC ¼ 500 mg/mL). Among the tested fractions, only the ethyl acetate fraction showed moderate antibacterial activity (MIC ¼ 500-250 mg/mL) against one reference bacteria strain S. flexneri NR518 two isolates of S. flexneri CPC and S. aureus CPC. For this reason, this fraction was further submitted to repeated column chromatography over silica gel, from which, two new compounds conyflavone (1) and conyclerodane (2) were obtained along with six kwon other including two flavonoids identified as gardenin C (3) (Mamdouh et al. 1991), chrysosplenetin (4) (Ahmed et al. 2016), two phyto-sterol, glucoside of b-sitosterol (5) (Peshin and Kar 2017), the mixture of stigmasterol (6) and b-sitosterol (6 0 ) (Nyigo et al. 2016), and ent-2b,18,19trihydroxycleroda-3,13-dien-16,15-olide (7) (Alembert et al. 2015). They were identified from the analysis of their spectroscopic data which match well to those reported in the literature review.
Compound 1, conyflavone (1) was obtained as a yellow crystal from the n-hexane-EtOAc (9:1) and soluble in methanol. It melts at 210-212 C and gave a positive reaction both to FeCl 3 and Shinoda test suggesting its phenolic and flavonoid nature. Its molecular formula, C 20 H 20 O 9 was deduced from its HRESIMS spectrum, which showed in positive mode the protonated ion peak [M þ H] þ at m/z 405.1160 (calc 405.1186) corresponding to eleven double bond equivalents. Its IR spectrum exhibited vibration bands of hydroxyl group at t 3341 cm À1 , methyl groups t 2997 and 2908 cm À1 and conjugated carbonyl t 1643 cm À1 . The UV spectrum showed absorption bands at k max 280 and 350 characteristic of flavanone chromophore type. This was confirming both by the 1 H NMR which shows one proton singlet at 6.66 ppm correlated in HSQC with carbon at 103.2 ppm, characteristic of proton H-3 and carbon C-3 respectively (Agrawal et al. 1989). Beside H-3 proton, the 1 H NMR also shows two protons singlet at 7.26 ppm (2H, s) due to two protons of ring B having the same environment, and four (04) singlets among which three integrated for three protons each respectively at d H 3.93 ppm (3H, s), 4.00 ppm (3H, s) and 4.12 ppm (3H, s), and the other one integrated for six protons at 3.98 (6H, s), corresponding to five methoxyl groups. The broad band 1 H decouple 13 C NMR spectrum (CD 3 OD, 150 MHz, Supplementary material, Table S1) of compound 1 displayed seventeen (17) carbon signals instead of twenty (20) as figured in the molecular formula suggesting the presence of three isochrone carbons. These signals were sorted by DEPT technique as five (05) methoxy groups respectively at d C 60.8, 61.4, 61.8 ppm and two with the same value at d C 56.1 ppm, three (03) sp 2 methines at d C 103.2 and 103.9 ppm. Thus, there are twelve (12) quaternary carbons including ten (10) oxygenated among which one carbonyl at d C 183.2 ppm and nine other quaternary carbons respectively at d C 164.9, 153.1 149.4, 133.0, 136.3, 145.2, 148.2, 140.1, and 148.2 ppm, two other quaternary carbons appearing at d C 106.7 and 121.1 ppm. It remained for us to establish the positions of the five methoxyl groups as well as the two phenolic hydroxyl groups in our skeleton. The presence of two barring isochronic protons on the first hand, and a singlet of six protons due to methoxyl group, allow us to assign the two aromatic B protons at position 2 0 and 6 0 , and the two isochronic methoxyl group at position 3 0 and 5 0 . The absence in the 1 H NMR spectrum of the signal due to the chelated hydroxyl proton, suggested that one of methoxyl groups is located at position C-5. Since the two hydroxyl group were located at position 4 0 and 7 respectively base on biogenetic pathways and the chemical shift of carbon C-4 0 and C-7 respectively at 140.1 and 149.4 ppm respectively (Harborne and Williams 2000), the two remaing methyl groups were then located at position 6 and 8 of A-ring. Thus, on the basis of the above spectroscopic studies, compound 1, to which trivial name conyflavone was given, was assigned as 6,4 0 -dihydroxy-5,7,8,3 0 ,5 0 -pentam ethoxyflavone.
Compound 2 is soluble in MeOH and was obtained as colourless crystals from the n-hexane-EtOAc (1:1) and melted at 140-141 C. It's molecular formula C 22 H 32 O 6 corresponding to seven double bond equivalents, was determined from its HRESIMS which showed in positive mode a sodiated-molecular ion peak [M þ Na] þ at m/z ¼ 415.2080 (calc 415.2052). The IR spectrum of this compound displayed vibration bands of hydroxyl group at 3269 cm À1 , alkyl group at 2965 cm À1 , carbonyl groups one for a,b-unsaturated-c-lactone at 1751 cm À1 and an ester 1722 cm À1 , and double bound at 1655 cm À1 . The 13 C NMR spectrum displayed 22 carbon signals in the molecular formula among which the acetate moiety carbon signals at 171.6 and 20.5 ppm, indicating that this compound is a diterpenoid. The DEPT 135 presented 16 carbon signals, eight with negative mode including 2 sp 2 methines corresponding to two olefinic carbons at d C 146.1, 129.3 ppm and 6 sp 3 hybridized carbons among which three primary methyl carbons respectively at d C 20.  (Merritt and Ley 1992;Rongtao et al. 2017) were oxidised to oxymethylenes at d H 3.83/3.89 (2H, d)/65.2, and d H 4.67(2H, s)/64.7 respectively; the position of these methylene groups at C-4 and C-5 was based on the observation of HMBC correlations between H-18 and C-3 (d C 129.3) and C-4 (d C 142.2) and between H-19 and C-6 (d C 31.2) and C-10 (d C 45.6). Two olefinic protons at d H 7.25 (1H, t)/146.1 and at d H 5.70 (1H, d, 5.0 Hz)/129.3 respectively were assigned to H-14 and H-3 based on the cross peaks observed on the HMBC spectrum, on one hand, between the proton at d H 7.25 with the carbonyl of the a,b-unsaturatedc-lactone at d C 175.6 (C-16), the oxygenated methylene at d C 70.5 (C-15), the quaternary carbon at d C 134.0 (C-13) and, the second one at 5.70 (1H, d, 5.0 Hz)/129.3, with carbons at d C 27.3 (C-1), d C 64.7 (C-18) and 68.1 (C-2), on the other hand, further HMBC correlations were observed between proton at 4.67(2H, s) and carbon at d C 171.6, 142.2, 129.3 and 43.0 indicating that the acetate moiety was linked to C-18.

SC-XRD analysis of compound 2
A single crystal of (2) (C 22 H 32 O 6 ) was examined on a Rigaku Supernova diffractometer using Cu Ka (k ¼ 1.54184 Å) radiation. The crystal was kept at 100.0(1) K during data collection. Using Olex2 (Dolomanov et al. 2009) the structure was solved with the ShelXT (George 2015) structure solution program using Intrinsic Phasing and refined with the ShelXL refinement package using Least Squares minimization. Hydrogen atoms were refined isotropically. The absolute configuration (S at C2, R at C5, S at C6, R at C7 and R at C9) could be validated with the following parameters: Hooft y: À0.01(4), Parson's q: 0.00(4), Flack x: 0.02(5). CCDC 2159930 contains the supplementary crystallographic data for this paper. These data can be obtained free of charge from the Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/conts/ retrieving.html

Antibacterial potential of extracts, fractions and compounds
For antibacterial activity of crude extract, fractions and compounds obtained from fractionation were determined and the Minimal Inhibitory Concentrations obtained were recorded in the table (Supplementary material, Table S3). Crude extract was active (MIC ¼ 500 mg/mL) on two sensitives bacteria S. flexneri NR 518 and E. coli ATCC 25922 and one resistant Staphylococcus aureus (ATCC 43300). All tested fractions displayed moderated antibacterial activity against on at list 2 of the 6 pathogenic strains used (MICs ranging from 250 to 500 mg/mL). The n-hexanic and the ethyl acetate fractions only inhibited the growth of S. flexneri species while the methanol fraction acts on S. flexneri and Staphylococcus aureus. The most sensitive bacteria strain was Shigella flexneri. S. aureus ATCC and MRSA strains were the most resistant among all tested bacteria (ATCC and clinical isolates). While comparing the activity of fractions to the activity of crude extract, we may notice that fractionation did not improve the antibacterial potential of fractions compare to the crude extract against S. aureus ATCC 43300, S. flexneri NR 518, and E. coli ATCC 25922. Indeed, it is well documented that fractionation might either positively or negatively affect the bioactivity of a given extract (Nwodo et al. 2010), this owing to a polarity-based grouping that gathered bioactive compounds.
Concerning compounds, they exhibited promising antibacterial activity with MICs ranging from 0.33 to 3.05 mmol/mL. Among the seven compounds analyzed, only five displayed antibacterial potential against at least one of the six tested bacteria. According to the wide range of action and the MICs values obtained, the antibacterial potency could be rated in the following order: 6 > 4>2 > 3>1 > 7>5 Compound 4 improved the antibacterial potential by two time against S. Flexneri CPC, and four time against S. Flexneri NR 518. Compound 6 improved the activity by 4 folds against E. coli ATCC 25922. This compound also inhibited the growth of the methicillin resistant S. aureus (SA MR 33591) compared to all fractions and the crude extract that were inactive against this strain. Globally compound 4 and 6 were the most actives with a broad spectrum of action.

General experimental procedures
Melting points were measured on a standard heating block melting point apparatus. IR spectra were recorded on a Bruker Fourier transform/infrared (ATR) spectrophotometer. HR-ESI-MS spectra were measured with a FTHRMS-Orbitrap (Thermo-Finnigan) mass spectrometer. 1D ( 1 H, APT) and 2D ( 1 H-1 H-COSY, HSQC, HMBC) NMR spectra were recorded in deuterated solvents on a Bruker ARX 600 ( 1 H 600 MHz and 13 C150 MHz). NMR spectrometer equipped with a 5 mm cryoprobe. All chemical shifts (d) were measured in parts per million (ppm) using a residual solvent signal as secondary reference relatively to tetramethylsilane (TMS) as internal standard, while coupling constants (J) are given in Hz. Solvents were distilled prior to use. Analytical grade solvents were used for LC-MS. Column chromatography (CC) was performed using Merck MN silica gel 60 M (0.04-0.063 nm) and thin layer chromatography (TLC) was performed on aluminum silica gel 60 F 254 (Merck) precoated plates (0.2 mm layer thickness). Spots were visualized on TLC either by UV lamp (254 and 365 nm) or by heating after spraying with 20% H 2 SO 4 (v/v) solution. Different mixtures of n-hexane, EtOAc and MeOH were used as eluting solvents.

Plant material
The whole plant of Conyza aegyptiaca was harvested in August 2019 at Bandjoun, a locality of the West Region of Cameroon and was identified by Mr. NGANSOP a botanist at the National Herbarium of Cameroun where a voucher specimen was deposited under the reference number 5604SRF/Cam.

Extraction and isolation
The whole plant of Conyza aegyptiaca (1.2 kg) were cut in small parts and dried at room temperature, powdered and extracted by maceration twice with the dichloromethane/methanol (1/1) v/v. The solution resulting was evaporated under reduce pressure, using a rotary evaporator to afford a green extract (96 g). A part of the crude extract obtains (95 g) was subjected to flash chromatography n-hexane, EtOAc and methanol to afford three fractions indexed F 1 (34 g; n-hexane), F 2 (41 g; EtOAc) and F 3 (16 g; methanol). These fractions were evaluated for their antibacterial potential; all the fractions F 1 , F 2 and F 3 give a moderate activity against testing bacteria. Fractions F 1 and F 2 were chromatographed on a silica gel column, using as eluent a mixture of nhexane/EtOAc for F 1 and dichloromethane/methanol for F 2 in increasing polarity. A total of 338 fractions, each of 100 mL for fraction F 1 and 136 each of 200 mL for F 2 were collected and combined based on their TLC profiles to give a total of 13 sub- . From sub-fractions S 8 and S 11 after filtration we obtained compounds 1 (15.06 mg), 2 (8.05 mg) and compound 3 (25.3 mg) from fraction S 4, compound 4 was obtained from S 7 , compound 5 (30.5 mg) from S 12 , also crystallized at room temperature, the mixture of compounds 6 and 6 0 (17.23 mg) from S 2 . Compound 7 (4.41 mg) was obtained from sub-fraction FS 4 . 3.5. Biological assays 3.5.1. Preparation of stock solutions of plant extract, fractions, compounds and ciprofloxacin Stock solution for crude extract was prepared at 10 mg/mL using 10% DMSO, fractions and compounds were prepared at 5 mg/mL in 10% DMSO. As positive control, ciprofloxacin was prepared at 1 mg/mL in sterile distilled water. Each solution was sterilized by filtration through 0.22 mm sterile filter (Acrodisc Syringe Filter).

Test bacteria and culture conditions
The antibacterial activity of extract, fractions and obtained compounds was tested against the methicillin resistant Staphylococcus aureus (SA ATCC 43300, SA MR 33591), and the reference strains of Escherichia coli (EC ATCC 25922) and Shigella flexneri (SF NR 518) all obtained from American Type Culture Collection. Bacteria isolates were provided by the Centre Pasteur of Cameroon (Yaound e, Cameroon), and consisted of clinical isolates of Staphylococcus aureus (SA CPC) and Shigella flexneri (SF CPC). Prior to used, bacteria were cultivated on Mueller Hinton Agar (Sigma Aldrich) and incubated at 37 C for 24 hours.

Determination of the Minimum Inhibitory Concentration (MIC)
The MICs of the crude extract, fractions, compounds and antibiotic were determined using the broth microdilution method as described by the protocol M07-A9 of the CLSI (2012). One in two serial dilutions of the extract, fractions, compounds and antibiotics were prepared. For the crude extract, fractions, and compounds, the concentrations used ranged from 0.25 mmol/mL to 10 mmol/mL, while, for ciprofloxacin the concentrations used ranged from a.156 lg/mL to 10 mg/mL. One hundred microliters of the extract, fractions, compounds and ciprofloxacin was added to wells in microtiter plates. To each well, 100lL of bacteria was added to achieve a final concentration of 1 Â 10 6 CFU/mL. Wells containing 200 lL of broth only were used as sterility control and wells containing 100lL of broth and 100 lL of bacteria were used as the negative control. The plates were incubated at 37 C for 24 hours. After incubation, Minimal inhibitory concentration (MIC) was defined as the lowest concentration with no visible bacterial growth.

Statistical analysis
For the bioactivity assay, each assay was performed in triplicate, and data were expressed as mean ± standard deviation (SD). Analysis of variance (ANOVA) was performed, and the mean separation was done by the least significant difference (LSD) (p < 0.05) using GraphPad Prism 5.3 software for Windows.