Antimicrobial activity of extracts from Crotalaria bernieri Baill. (Fabaceae)

This work was designed to study the antimicrobial activity of Crotalaria bernieri Baill. (Fabaceae). Extracts from leaf, root, pod and seed using hexane, ethyl acetate and methanol were tested in vitro for their activity against 17 bacteria, 5 fungi (3 yeasts and 2 molds) using disc diffusion and micro dilution methods. At the concentration of 1 mg/disc, all the extracts exhibited antimicrobial activity depending on the plant part and the extraction method used. The most sensitive germs were Salmonella enteridis , Streptococcus pyogenes and Candida guilliermondii with inhibition zone diameter (IZD) of 11 mm, 15 mm and 13 mm respectively. Most of extracts showed, broad activity spectrum varying from one extract to another. Minimum inhibitory concentration (MIC), minimum bactericidal concentration (MBC) and minimum fungicidal concentration (MFC) of all extracts were recorded. Ten extracts displayed an excellent effect (MIC < 100 µg/ml), 8 a moderate effect (MIC from 100 to 500 µg/ml), 5 a weak effect (MIC from 500 to 1000 µg/ml) and the others were ineffective (MIC > 1000 µg/ml). Leaf methanol extracts were the most efficient and Gram positive bacteria the most sensitive. All extracts had bactericidal (MBC/MIC ≤ 4) or fungicidal action (MFC/MIC ≤ 4) in certain microorganisms and bacteriostatic (MBC/MIC > 4) or fungistatic action (MFC/MIC > 4) in others. Antimicrobial activity might be due to tannins, polyphenols, steroids, triterpenes and flavonoids that were present in most of the plant organs, but alkaloids in leaf and pod and saponosides in root might also be involved. C. bernieri with the effectiveness of all its parts, the variety of its secondary metabolites, the great number of sensitive pathogen microorganisms and its ubiquity make this plant species an interesting source of antimicrobial agents.


INTRODUCTION
Antimicrobial resistance is one of the world's most serious public health problems. There is an urgent need *Corresponding author. E-mail: victor_jeannoda@yahoo.fr.
Author ( to find new disposable and affordable remedies to face this problem (Zongo et al., 2011). Many studies led to systematic screening of plant extracts as a source of antibacterial compounds (Dalmarco et al., 2010;Stefanovic and Comic, 2011). Several Crotalaria species have been reported to display antimicrobial properties. For example, Crotalaria madurensis is active against Bacillus subtilis, Staphylococcus aureus, Escherichia coli and Candida albicans (Bhakshu et al., 2008), Crotalaria capensis against Salmonella typhimurium (Dzoyem et al., 2014), Crotalaria burhia against B. subtilis and S. aureus (Sandeep et al., 2010;Mansoor et al., 2011), Crotalaria juncea against S. aureus (Chouhan and Sushil, 2010), Crotalaria pallida against E. coli and Pseudomonas sp (Pelegrini et al., 2009), and Cladophora trichotoma against Alternaria solani (Ravikumar and Rajkumar, 2013). The purpose of this study was to assess the antimicrobial activity of C. bernieri by testing plant part's extracts obtained in different methods on pathogen bacteria and molds. C. bernieri is one of the 53 Crotalaria species growing in Madagascar, an annual herb which is found in open vegetation, grassy places and roadsides in most regions of Madagascar (Peltier, 1959). It flowers on July to October and December to March (Polhill, 1982;Dupuy et al., 2002).

Plants
C. bernieri ( Figure 1) were harvested in Ibity, District of Antsirabe, Region of Vakinankaratra, 200 km from Antananarivo region. The plant was collected in April and July, 2013 and was identified by Polhill R.M. Voucher specimens (Herizo R. 010) of C. bernieri were deposited in the herbarium of Plant Biology and Ecology Department of the Faculty of Sciences of the University of Antananarivo.

Microorganisms strains
The microorganisms used in this study consisted of 17 strains of bacteria (10 Gram (-) and 7 Gram (+)), 3 yeasts and 2 molds (Table  1). These strains were obtained from the collections of Laboratoire de Chimie des Substances Naturelles et Sciences des aliments (LCSNSA) of La Réunion University. They were maintained on agar slant at 4°C and cultured on a fresh appropriate agar plate during 24 h prior to antimicrobial tests.

Preparation of extracts
The dried leaves, seeds, seed pods, and roots of the plant were grounded into powder. The resulting powder (100 g) was extracted successively with 4x500 mL of hexane, ethyl acetate and methanol for 24 h under stirring at room temperature. After filtration using a Whatman filter paper, each combined extract was evaporated under reduced pressure to dryness. The dry residues, dissolved in hexane, ethyl acetate and sterile distilled water, constituted hexane, ethyl acetate and methanol extracts respectively (Table 2).

Phytochemical screening
The reactions for the detection of chemical groups were those developed by Fong et al. (1977) and Marini-Bettolo et al. (1981).

Antimicrobial activity test
The in vitro antimicrobial activity of the extracts was determined using disc diffusion method described by Pyun and Shin (2006) and Ngameni et al. (2009). Two mL of inoculum corresponding to 0.5 MacFarland (10 8 CFU/ml) was uniformly spread on the surface of Columbia Agar medium (for Streptococcus); Mueller-Hinton Agar (MHA) for the other bacteria and Potato Dextrose Agar (PDA) for yeasts. Sterilized filter paper discs 6 mm diameter (BioMérieux,REF 54991) were impregnated with 10 μL of each extract to the concentration of 100 mg/mL (1 mg/disc). The soaked discs were then placed on the surface of the agar and incubated at 37°C during 24 h for bacteria, or at 25°C for yeasts. The inhibition zone diameter (IZD) was measured and the results were interpreted by  Ponce et al. (2003) and Celikel and Kavas, (2008) stating that bacteria are not sensitive for IZD less than 8 mm, sensitive for IZD of 9 to 14 mm, very sensitive for IZD of 15 to 19 mm and extremely sensitive for IZD larger than 20 mm. Antifungal activity was evaluated by a method described by Favel et al. (1994). One ml of each extract was added to 19 ml of medium culture of PDA and maintained at 45°C. The mixture is then poured into Petri dishes and dried for 15 min at 37°C. 10 µl of each tested microorganism corresponding to 0.5 MacFarland were spread on the medium surface. IZD were measured after incubation at 25°C for 72 h. Negative controls were prepared by using the same solvents employed to dissolve the plant extract samples while the reference antibiotics were used as positive controls. All the experiments were performed in triplicate. The results were expressed as mean values ± standard deviations (mm ± SD).

MIC, MBC and MFC determination
For extracts showing antibacterial activity in the disc diffusion method (IZD ≥ 8 mm), MIC (minimum inhibitory concentration), MBC (minimum bactericidal concentration) MFC (minimum fungicidal concentration) were determined by microdilution method ). The concentration of each extract was adjusted to 25 mg/ml. This was serially diluted two-fold to obtain concentration ranges of 0.024 to 25 mg/ml. Each concentration was added in a well (96-well microplate) containing 95 μl of Mueller-Hinton broth (MHB) and 5 μl of inoculum (standardized at 0.5 MacFarland). A positive control containing bacterial culture without the extract and a negative control containing only the medium, were also analyzed. The plates were covered with sterilized aluminum foil, and then incubated at 25°C (yeasts and molds) or at 37°C (bacteria) for 24 h. The MIC of each extract was detected following addition 40 µl of 0.2 mg/ml p-iodonitrotetrazolium chloride and incubation at 25°C (yeasts and molds) or at 37°C (bacteria) for 30 min . Viable bacteria reduced the yellow dye to a pink color. MIC was defined as the lowest sample concentration that prevented this change and exhibited complete inhibition of bacterial growth. For the determination of MBC and MFC, 5 μl from each well that showed no change in color was transferred on MHA or PDA plate and incubated at 25°C (yeasts and molds) or at 37°C (bacteria) for 24 h. The lowest concentration at which no growth occurred on the agar plates corresponded to the MBC or MFC.
The ratios MBC/MIC and MFC/MIC were calculated for each extract, to determine the nature of the effect. The extract is bactericidal or fungicidal for MBC/MIC or MFC/MIC ≤ 4 and bacteriostatic or fungistatic when these ratios are >4 (Djeussi et al., 2013;Bouharb et al., 2014;Chamandi et al., 2015).

Statistical analyses
Results were expressed as mean values ± standard deviations of three separate determinations. One-way analysis of variance (ANOVA) which was followed by Newman Keuls comparison test with Staticf® software was used for statistical analysis. Statistical estimates were made at confidence interval of 95%.

Extraction yields
The extractive yield of different parts of C. bernieri with different solvents varied from 4.2 (PME) to 24.1% (RME) ( Table 3).

Qualitative phytochemical analysis
The major secondary metabolites identified in the different organ extracts are presented in Table 4. Tannins, polyphenols, steroids, triterpenes and unsaturated sterols occurred in all the C. bernieri organs. Flavonoids were found in all organs except root. Alkaloids were present only in leaf and pod while saponins only in root. Iridoïds, leucoanthocyanins, and quinones were not detected in all parts of C. bernieri.

Antimicrobial activity
At 1 mg/disc, a concentration generally used for antimicrobial activity assessment in plants (Sandeep et al., 2010;Govindappa et al., 2011;Linthoingambi and Mutum, 2013;Marimuthu et al., 2014), the large majority of C. bernieri extracts (16 of 22) inhibited the microorganism growth with IZD ranging from 8 to 15 mm (Tables 5 to 7). However, activity depended on the microorganism, the plant parts and extraction method used. The most sensitive germs were S. enteridis (IZD=11 mm), S. pyogenes (IZD=15 mm) and C. guilliermondii (IZD=13 mm) in Gram (-) bacteria, Gram (+) bacteria and yeasts, respectively. Gram (-) strains C. jejuni and E. coli, E. faecalis, Gram (+) L. monocytogenes and the two molds A. fumigatus and A. niger were resistant to all the extracts. REA, with an IZD of 15 mm against S. pyogenes, displayed the highest antibacterial activity.

Tannins and Polyphenols
Gelatin 1% In yeasts, most of leaf extracts were active against the three Candida strains tested, but seed and pod extracts were active only against C. guilliermondii. Antibiotics used as references in this study (amoxicillin 25 µg, chloramphenicol 30 µg, penicillin 6 µg and miconazole 50 µg) were more effective than most of C. bernieri extracts. MIC, MBC, MFC and MBC or MFC/MIC ratio values are presented in Tables 8 to 10. MIC ranged from 0.048 to 25 mg/ml. MIC maximum values registered was 12.5 mg/mL except for RHE on S. pyogenes (MIC=25 mg/ml). Concerning MBC or MFC, maximum values for all extracts were 25 mg/ml except for root extracts on some Gram (+) bacteria and C. guilliermondii (MBC>25 mg/ml). The ratio MBC or MFC/MIC varied from 1 to more than 100.
All methanol extracts were active. This is also the case for ethyl acetate extracts except LEA. As to hexane extracts, PHE and RHE were efficient but not LHE and SHE. Pod extracts had the broadest spectrum of activity with 10 sensitive microorganisms and seed extracts the narrowest ones with 8 sensitive microorganisms.

DISCUSSION
The present study shows that the C. bernieri extracts inhibited the growth of most tested microorganisms, indicating the presence of antimicrobial compounds in all parts of the plant. Phytochemical screening showed the presence of diverse secondary metabolites, reported to have antimicrobial property. At this stage of the work, results did not yet allow to state whether the same or different compounds are involved in the different parts of the plant. However, they suggested that C. bernieri antimicrobial activity might be mainly due to tannins, polyphenols, steroids, triterpenes and flavonoids, which were present in all or most of the plant organs. Alkaloids might also be concerned in leaves and pods and saponosides in root.
C. bernieri extracts showed generally a broad antimicrobial spectrum. They were capable of inhibiting the growth of different Gram (-) and Gram (+) bacterial strains as well as some yeasts. However, each extract Table 5. In vitro Antimicrobial Activity (IZD in mm) of extracts (1 mg/disc) on Gram (-) bacteria.

Extracts/ controls
Cj  displayed a specific activity spectrum that could be due to difference between the chemical nature and concentration of bioactive compounds in extracts. The results obtained with microdilution method were more reliable than those with disc diffusion. That might be due to the fact that bioactive compounds were in direct contact with germs in liquid medium whereas they diffused little or not at all in solid medium. There was no consensus on the acceptable level of inhibition for natural products (Benko and Crovella, 2010). For Dalmarco et al. (2010), for crude extracts and fractions, a MIC lower than 100 µg/mL was considered as an excellent effect, from 100 to 500 µg/ml as moderate, from 500 to 1000 µg/mL as weak, and over 1000 µg/ml as inactive. According to Kouitcheu et al. (2013), when a crude extract was used, the MIC values of 8 mg/mL or below against any microorganism tested was considered as active.
However, if the interpretation of Kouitcheu et al., (2013) was taken into account, only nine extracts had MIC higher than 8 mg/mL on some germs, which means that all the other extracts of C. bernieri used showed    All the extracts had bactericidal action (MBC/MIC ≤ 4) in certain bacteria and bacteriostatic action (MBC/MIC >) 4) in other ones. For example LME was bactericidal against B. cereus and C. perfringens but bacteriostatic against S. aureus and S. pneumoniae. The comparison of A. bernieri extract activities to foreign Crotalaria species was not easy because antimicrobial activity was assessed under different conditions (other microorganism strains and extract doses used).

Conclusion
This study clearly demonstrates the potential of C. bernieri as a source of interesting natural wide spectrum antimicrobial molecules. All its parts were efficient and could be easily found in significant amounts for the plant grows in fields, in the vicinity of homes, on roadsides and can be cultivated. Furthermore, according to our survey of local populations, C. bernieri is consumed by zebus but no cases of poisoning have yet been reported. At present, our works are concerned with the isolation of pure compounds from different extracts of C. bernieri and the elucidation of their structures in order to better evaluate their pharmacological activity. In view of later therapeutic use of C. bernieri, study on various experimental models of animals is also on going to assess the harmful effects it might have.