Preliminary screening of the antimicrobial activity of nine medicinal plant species from Burkina Faso

Infectious diseases, including skin infections, have always been treated by plants and other natural products since ancient times. This study investigated the antimicrobial activity of nine medicinal plant species from Gonsé Forest and Bangreweogo Park, Burkina Faso on seven microbial strains involved in skin infections. Sixty-eight organic and aqueous extracts from leaves, stem and root barks of the nine plants were tested on five bacterial and two fungal American Type Culture Collection (ATCC) strains using the disk diffusion and microdilution methods. Twenty-eight extracts (41.17%) from eight of the plants differently inhibited bacteria and fungi, the exception being Wissadula amplissima specie. Gram-positive Cocci (GPC) were the most sensitive to the extracts. Eleven extracts had a Minimum Inhibitory Concentration (MIC) lower than 1 mg/ml (MIC < 1 mg/ml). The extracts of Opilia celtidifolia gave the largest zones of inhibition ranging from 21.67 to 24.33 mm at 100 to 300 mg/ml on GPC and Candida albicans. Its MIC varied from 0.03 to 0.06 mg/ml on Streptococcus pyogenes, and from 0.07 to 0.24 mg/ml on Streptococcus agalactiae. Ethanolic extracts (80 and 96%) of Lawsonia inermis had MIC ranging from 0.29 to 9.37 mg/ml on the sensitive strains. The variable inhibitory activity displayed by the plants confirmed in part their traditional use in the treatment of some bacterial and fungal skin infections.


INTRODUCTION
Infectious dermatoses, or skin infections, are a major public health problem in developing countries (Hay et al., 2014;Marks et al., 2019). In tropical regions, the prevalence of skin infections has increased significantly with the occurrence of HIV/AIDS in 1980 (Silva et al., 2012). In Mali it has been reported that skin infections represent 55.10% of consultation causes among children in the dermatology department of the National Centre for Support of Illness Control in Bamako (Fofana et al., 2016). In Senegal, infectious dermatoses accounted for 64.7% of dermatological emergencies in the dermatology department of the Thiès regional hospital center (Dione et al., 2018).

Skin infections are particularly caused by bacteria, such as
Escherichia coli, Pseudomonas aeruginosa Staphylococcus aureus, Streptococcus agalactiae and Streptococcus pyogenes, characterized by high antigenic and antibiotic potential. Moreover, these bacteria are associated with biofilm formation, factor of additional virulence in hostile environments (Fu et al., 2017). S. aureus and P. aeruginosa are frequently involved in the colonization of chronic wounds or severe burns (Tong et al., 2015), spongiotic or acantholytic dermatoses, especially when they are extensive and localized in folds (Morand and Morand, 2017). S. aureus is responsible for a variety of skin and mucosal infections, ranging from mild infection (impetigo and simple cellulitis) (Tong et al., 2015) to complication (acute staphylococcal epidermolysis or Staphylococcal toxic shock syndrome (STSS) and to life threatening (Nhan et al., 2012). P. aeruginosa, is responsible for potentially severe forms of dermatoses, such as Ecthyma gangrenosum, which marks immunosuppression or reveals septicemia, especially in children (Morand and Morand, 2017).
Antibiotic are usually used to treat infections caused by these pathogens microorganisms, but the increase of antibiotic resistance limits their efficacy. The unavailability of antibiotics because of high costs and inaccessibility to some of them is endemic to patients living in rural areas of sub-Saharan countries. In such a context, the search for new non-toxic molecules becomes imperative and natural substances from plants appear as viable alternative. Indeed, it was reported in some studies that some molecules from plants have properties to kill microbial agents (Górniak et al., 2019;Salas et al., 2015) or to inhibit antibiotic resistance mechanisms (Rao et al., 2018). These compounds act by several mechanisms such as the formation of complexes with proteins and polysaccharides possessing polyphenol function (Papuc et al., 2017); the disruption of microbial membranes by lipophilic flavonoids and terpenoids (Górniak et al., 2019), and the inhibition of microbial protein adhesion to host polysaccharide receptors (Lee et al., 2006). Thus, plant molecules because of their specificity play an important role in combating antibiotic resistance by inhibiting efflux pumps (Rao et al., 2018) and biofilm formation (Bhunu et al., 2017;Gupta et al., 2019). For example Fu et al. (2017) noted that Herba patriniae, a medicinal plant from the traditional Chinese pharmacopoeia, inhibited biofilm formation in P. aeruginosa. Elswhere, Escobedo-Martínez et al. (2010) reported that pescaprine, a molecule isolated from Ipomoeapes-capraea had an inhibitory effect on Methicillin-resistant S. aureus (MRSA) efflux pumps, thus, increasing the norfloxacin effect. Some secondary plant metabolites also showed their synergistic role with other antibiotics. This was the case with oleanolic and ursolic acids, two pentacyclic triterpenes that increased the activity of Beta-lactam Youl et al. 523 antibiotics against S. aureus, S. epidermidis and Listeria monocytogenes strains (Kurek et al., 2012). In Burkina Faso, the panorama of bacterial skin infections varies according to the region and is considered frequent (Andonaba et al., 2010;Korsaga et al., 2019). According to statistics from health Ministry skin diseases were the fifth reason for outpatient consultations in primary health Centers, in 2018. In Burkina Faso, Cassia sieberiana, Cymbopogon schoenanthus, Lannea acida, Lawsonia inermis, Opilia celtidifolia, Piliostigma reticulatum, Pupalia lappacea, Sanseveria senegambica, Wissadula amplissima, are frequently used by traditional healers to treat several infections including skin infections. Recipes developed are based on one or several plants (Guigma et al., 2012;Lengani et al., 2010;Zerbo et al., 2011). Decoction, trituration and aqueous maceration, infusion or carbonization is the main methods of preparing herbal concoctions. These herbal concoctions are administered through oral route and external applications. Regarding the widespread use of the nine aforementioned plant species in Burkina Faso, this study evaluated their in vitro antimicrobial activity against seven microbial strains causing skin infections.  (Table 1) were collected from each plant species when possible, then, thoroughly washed with clean water and dried in the shade at room temperature (25°C) until drying (7 to 14 days). The clean, dried plant materials were then pulverized into fine powder by using an electric grinder.

Preparation of the extracts
The extraction was performed according to the solid-liquid extraction method described by Bruneton Jean (1999) using solvents of increasing polarity (Bruneton, 1999) with slight modifications.
Decoction: 50 g of each plant material powder in 500 ml of distilled Amaranthaceae Voêns-tabdo Leaves water were boiled under reflux for 15 min. The cooled was filtered on a fine-mesh nylon tissue and then centrifuged at 2000 rpm for 10 min. The supernatant obtained was concentrated at 50°C in a ventilated oven. The concentrated extract was frozen at -24°C, and then freeze-dried. The dry extract was weighed and stored in the freezer waiting for biological tests.
Organic extraction: 50 to 100 g of plant powder was macerated for 24 h with 250 to 500 ml of ethanol (80 and 96%) or hexane. The extract was filtered on Whatman filter paper No5 and concentrated under reduced pressure with a rotavapor (BÜCHI RII, Germany). The concentrated extracts were then put in an oven at 40°C to remove the remaining solvent; the dried residues were weighed and also stored in the freezer for later use.

Preparation of the discs with the extracts, antibiotics and controls
The organic and aqueous extracts were each dissolved in dimethyl sulfoxide (DMSO) for analysis ≥ 99.8% mixed with distilled water (1/5) to give an initial concentration of 300 mg/ml. This stock solution was then diluted to give final concentrations of 200 and 100 mg/ml. Erythromycin (15 µg /ml), Ciprofloxacin (5 µg /ml) and Nystatin (100 IU) were used as positive controls. The mixture of DMSO and water (1/5) was used as negative control. Sterile discs (6 mm diameter) were impregnated with plant extracts (300, 200 and 100 mg/ml) and / or positive and negative controls under a microbiological safety cabinet. They were placed in sterile Petri dishes and then dried in the oven at 37°C for 24 h.

Preparation of the bacterial inoculum
A cryo-capsule containing the reference strain previously placed in an oven at 37°C for 10 to 15 min is dissolved in 1 ml of sterile physiological water; 10 μl of the inoculum obtained were cultivated on an appropriate agar according to the growth conditions of microbial strains. The agar media were incubated for 24 h at 37°C for bacteria and 48 h at 37°C for fungi under strain-specific conditions. The morphological, cultural, biochemical and antigenic characteristics of microbial strains were carried out using standard and selective culture media as well as API kit. Colonies (1 to 2 wells) from the preculture were then introduced into a tube containing 5 ml of physiological saline water (0.9% NaCl). The turbidity was adjusted to the McFarland scale (~ 10 8 CFU/ml) using a densitometer. The inoculum was reduced to 10 7 CFU/ml by dilution (1/10).

Agar disk-diffusion method
The antimicrobial activity testing was carried out using the diskdiffusion method of Kirby Bauer (SFM, 2011). Petri dishes containing Agar were inoculated with each inoculum previously prepared (~ 10 7 CFU/ml) over their entire surface. Then, the dried discs containing the extracts (300, 200 and 100 mg/ml) as well as the control discs (negative and positive) were delicately placed on the agar surface of each petri dish. Petri dishes were incubated 24 h under growth conditions required for each microorganism. Zones of inhibition were measured (Figure 1). The test was validated if the negative control did not inhibit bacterial growth, and if the positive control had an inhibition diameter within the range given by the Clinical and Laboratory Standards Institute (CLSI).

Microdilution method
The minimal inhibitory concentration (MIC) and the minimal bactericidal concentration (MBC) of extracts (inhibition diameter ≥ 10 mm) were determined by the microdilution method (CLSI, 2012). In a 96-well microplate, successive dilutions of reason 2, were made from wells to wells (wells 1 to 6), leading to final concentrations varying between 150 and 0.0183 mg / ml. The wells 7 and 8 were used for positive and negative controls. The bacterial inoculum in a broth (Mueller Hinton, Heart-Brain or Sabouraud) at 10 6 CFU/ml was added to each well; the microplates were incubated at 37°C for 18 h. The MIC was the lowest concentration of extract that completely inhibits the growth of the bacteria in wells, estimated visually compared to controls. Plates containing MH, Sabouraud or chocolate + polyvitex agar, inoculated with 5 μl of the inoculum from wells which did not show visible growth when determining the MIC were sub-cultivated; CMB was the lowest concentration of the extract with no colony growth.The MBC/MIC ratio allows to specify the action mode of a substance. If the MBC/MIC ratio is ≤ 2, the substance is bactericidal. On the other hand, if it is ≥2, the substance is bacteriostatic (Fauchere, 2002).

Statistical analysis
Data were analyzed with Excel 2010. Results are expressed as mean (3 tests) plus standard deviation. The Student t-test was applied for means comparison with the significance level set at 5%.

Antibacterial and antifungal activity of the extracts
In total, 68 extracts among which leaf extracts (32), root extracts (20) and stem bark extracts (16) were tested against five strains of bacteria and two strains of fungi, American Type Culture Collection (ATCC). At 300 mg/ml, 28 extracts (41%) from eight plant species had at least a diameter of inhibition ≥10 mm) on four of the bacteria strains; at 200 mg/ml, 18 (27%) extracts had this diameter, and at 100 mg /ml, only 7 (10%) extracts. The active extracts showed a dose-dependent inhibition on bacteria and fungi. The inhibitory activity was variable regarding the plant parts. Among extracts with an inhibition zone ≥ 10 mm, leaf extracts were the most represented, (14) followed by stem bark extracts (09); while only 5 extracts from root bark were recorded.
According to Habbal et al. (2011), the variable antimicrobial activity displayed by the extracts may depend on the type of extraction solvent, the part of the plant, the concentration and the microorganism strain (Habbal et al., 2011). It was also shown that chlorophyll, which is one of the constituents of fresh leaves possess antimicrobial activity (Maekawa et al., 2007). However, the high proportion of leaves among the collected samples (32 samples out of 68), could justify these results.
Extracts with medium to high polarity (inhibition diameter ≥10 mm) were the most active. There were 18 ethanol extracts, 9 decoctions and one non-polar extract (hexan extract).
The ethanolic and aqueous extracts of the leaves of L. inermis displayed the highest effect on S. pyogenes (20 mm) and S. aureus (19.67 mm), at 100-300 mg/ml. Only, the hexane extract of C. schoenanthus (Le1) ( Table 2) gave a visible inhibition zone (9-10 mm) on S. agalactiae and (10-12 mm) on C. tropicalis at 100-300 mg/ml. This reflected the ability of polar solvents to retain the active antimicrobial ingredients; which in polar extracts are mainly tannins, saponins, flavonoids and alkaloids However, only the aqueous extract of L. inermis inhibited the growth of P. aeruginosa with an inhibition diameter of 10.67 mm at 300 mg/ml (Table 2). Twenty (29%) extracts inhibited the growth of S. aureus with an inhibition diameter ranging from 10 to 19.67 mm, while nine others (18%) diversely inhibited the growth of S. pyogenes. The results on Gram-negative bacilli (GNBs) contrasted with those of other studies with the same plants species ;(Chowdhury et al., 2014;Habbal et al., 2011;Hoekou et al., 2012). Biotic and abiotic factors, harvesting period, parts of the plant used and extraction method may explain these differences (Mukherjee and Houghton 2009). The activity displayed by L. inermis on P. aeruginosa may be attributed to free hydroxyls that have the capability to combine with the carbohydrates and proteins in the bacterial cell wall. They may get attached to enzyme sites rendering them inactive (Al-Rubiay et al., 2008). In addition, to providing a source of stable free radicals, compounds such us quinones are known to complex irreversibly with nucleophilic amino acids in proteins, often leading to inactivation of the protein and loss of function (Habbal et al., 2011).
Most of the extracts fell to inhibited Gram-negative bacilli; this result was not surprising. Indeed, it was reported that Gram-negative bacilli (GNBs) have an intrinsic resistance to biocidal agents because of the nature of their outer membranes (Dzidic et al., 2008). The same reason may explain the high susceptibility of GPCs found in the study. Gram-positive bacteria walls are made exclusively of peptidoglycans, which seems to be more favorable to the penetration of phytomolecules in bacteria (Dzidic et al., 2008). 7 (10%) extracts were able to inhibit the growth of fungi among which the aqueous extracts of the leaves and stem barks of O. celtidifolia (1Le4 and 1SB4) had almost similar antifungal activity as the control (nystatin) on C. albicans (21,67 and 24,33 mm). The antifungal activity of these extracts may be due to the different chemical compounds (phenolics, tannins, phlobatannins, alkaloids, triterpenoid saponins, anthraquinones and flavanoid) present in O. celtidifolia (Mazadu et al., 2018). It has been reported that two flavonoid classes (flavans and flavanols) owing to their special structure formed by hydroxyl groups in position 2, 4, 5 and 7 on the aromatic core possess significant antimicrobial power (Bouterfas et al., 2016). However, the optimal efficacy of a medicinal plant may not be due to one main active ingredient, but to the combined action of different compounds originally in the plant (Essawi et Srour, 2000). Therefore, there is no explanation for the difference in inhibition depending on the fungal species observed.
Fifteen (22%) of the extracts had an inhibition zone less than 10 mm (d< 10 mm) and were considered as inactive to weakly active. They included all leaf extracts of W. amplissima (4) and 11 extracts from the other 8 plants. It is unusual to notice significant differences in data for the same plant species because important variations of activity can be observed with extracts of plants treated by solvents presenting very polar characters. Thus, between an aqueous extract and an alcoholic extract of low titer, important pharmacological differences appear and can particularly be concretized by a lack of activity (Bouharb et al., 2014). These variations might be due to many factors, including the method of extraction, climatic, seasonal and geographical conditions, and harvest time as previously notified. There were no similar studies to compare the non-activity of W. amplissima extracts. However, the period (June and October) of samples collection that corresponds to the middle and end of the winter season could be a limiting factor because the plants are still full of large quantity of water which may dilute the active ingredients.
All the strains were sensitive to reference antibiotics (ciprofloxacin, erythromycin and antifungal (Nystatin). Their activity was clearly superior to that of all the plant extracts (21.5 to 34 mm for bacteria and 23.51 to 23.76 mm for fungi). The difference of activity between pure chemical compounds and plant extracts has been argued and the most plausible explanation was the low concentration of active ingredients and the presence of both synergistic and antagonistic chemical constituents in the same extracts.

Minimal inhibitory concentration and minimal bactericidal concentration of the plant extracts
The MIC and MBC values varied depending on the types of strains. Extracts with the best MIC (< 1 mg/ml) and MBC are presented in Table 4, showing that 11 extracts from 7 of the plants had a MIC < 1 mg/ml, among which 8 had a bactericidal effect (MBC/MIC = 2) and the 3 others a bacteriostatic effect (MBC/MIC  4). The susceptibility of Gram-positive cocci (GPC) to the extracts was confirmed with the MIC. Ethanol extracts (96 and 80%) of O. celtidifolia root barks were the most effective (MIC between 0.03 and 0.24 mg/ml) on Gram-positive cocci, S. pyogenes and S. agalactiae (Table 3). Ethanolic extracts (96 and 80%) of the leaves of L. inermis exercised moderate inhibitory activity (MIC between 0.29 and 0.78 mg/ml) on susceptible strains. Among all the extracts, only, the 96% ethanolic extract of the leaves of L. inermis had a bactericidal effect on S. aureus. The MICs of ethanol leaf extracts of L. inermis for susceptible strains were higher than what reported by previous studies (Al-Rubiay et al., 2008). This discrepancy may be due to many of the factors previously mentioned.

Conclusion
The results obtained during this study confirmed the antimicrobial properties of some of the selected plants and could justify their use in traditional medicine against skins infectious of bacterial and fungal causes. The extracts from Opilia celtidifolia and Lawsonia inermis with the most promising antimicrobial activity will be further investigated for their development into improved traditional phytomedicines.