Antibacterial and Antibiotic Modifying Potential of Crude Extracts, Fractions, and Compounds from Acacia polyacantha Willd. against MDR Gram-Negative Bacteria

The present study aimed to assess the in vitro antibacterial and antibiotic modifying activities of methanol extracts prepared from the leaf (APL) and bark (APB) of Acacia polyacantha, fractions (APLa-d) and compounds isolated from APL against a panel of multidrug resistant (MDR) Gram-negative bacteria. Leaf extract was subjected to column chromatography for compounds isolation; antibacterial assays were performed on samples alone and with an efflux pump inhibitor (EPI), respectively, and several antibiotics on the tested bacteria. The phytochemical investigation of APL led to the isolation of stigmasterol (1), β-amyrin (2), 3-O-β-D-glucopyranosylstigmasterol (3), 3-O-methyl-D-chiro-inositol (4), epicatechin (5), quercetin-3-O-glucoside (6), 3-O-[β-D-xylopyranosyl-(1→4)-β-D-galactopyranosyl]-oleanolic acid (7), and 3-O-[β-galactopyranosyl-(1→4)-β-D-galactopyranosyl]-oleanolic acid (8). APL and APB had minimal inhibitory concentration (MIC) values ≤ 1024 μg/mL on 73.3% and 46.7% of the tested bacteria, respectively. APLb and APLd were effective against 88.9% of tested bacterial species with compound 8 showing the highest activity inhibiting 88.9% of tested bacteria. The EPI, phenylalanine-arginine-β-naphthylamide (PAßN), strongly improved the activity of APL, APLb, APLd, and compound 8 on all tested bacteria. Synergistic effects were obtained when APL and compounds 7 and 8 were combined with erythromycin (ERY), gentamycin (GEN), ciprofloxacin (CIP), and norfloxacin (NOR). The present study demonstrates the antibacterial potential of Acacia polyacantha and its constituents to combat bacterial infections alone or in combination with EPI.


Introduction
Bacterial drug resistance constitutes a serious concern in the therapy of infectious diseases. Despite the abundance of various classes of antibiotics, the emergence of resistant strains of bacteria is increasing [1]. This phenomenon of resistance has increased the disease burden, and it has become necessary to search for new and cheaper alternatives with fewer side effects [2]. Botanicals (or crude plant extracts) and their secondary metabolites have long been used by humans for medicinal purposes. It is estimated that about 80% of the world's population uses medicinal plants as alternative for their health care [3]. Cameroon's flora is an enormous reservoir of antibacterial botanicals and phytochemicals (or plants secondary metabolites); some Cameroonian medicinal plants previously documented for their antibacterial potential include Treculia obovoidea [4], Vismia laurentii [5], Artocarpus communis [6], Piper nigrum and Vernonia amygdalina [7], Cyperus esculentus [8], and Beilschmiedia obscura [9]. Several botanicals have 2 Evidence-Based Complementary and Alternative Medicine previously been reported for their activity against multidrug resistant (MDR) bacteria and had the ability to potentiate the activity of currently used antibiotics; such plants included Dorstenia psilurus [10], Combretum molle [11], Xanthosoma mafaffa, Moringa oleifera, and Passiflora edulis [12], Rubus fellatae, and Manihot esculenta [13]. It is important to improve our library of botanicals and phytochemicals with promising antibacterial potential, in order to combat MDR phenotypes. In the present study, we selected another Cameroonian medicinal plant, Acacia polyacantha Willd. (Fabaceae). Acacia polyacantha is a deciduous, straight cylindrical, erect tree of about 10-15 m height found in Tropical Africa. It has a geographical distribution, ranging from Gambia to Ethiopia and southwards to Kenya and Zimbabwe [14,15]. The plant is traditionally used to treat livestock diseases and gastrointestinal infections [16]. The plant is also used as a remedy for snakebite and as an infusion to bath children who are restless at night [14]. This is the first report on the antibacterial potential of this plant against MDR bacteria. It was found that this plant had no anthelmintic effect against a levamisole resistant strain of the nematode Caenorhabditis elegans [17]. Previous phytochemical investigations of the leaves of the plant led to the isolation of polyacanthoside A, oleanolic acid, stigmasterol, stigmasterol-3-O--glucopyranosyl, epicatechin quercetin-3-O-glucoside, 3-O-methyl-D-Chiroinositol, and 3-O-[ -D -galactopyranosyl-(1 →4)--Dgalactopyranosyl]-oleanolic acid [15]. The present study was designed to evaluate the antibacterial activity of the leaf and bark extracts of Acacia polyacantha against Gram-negative bacteria expressing MDR phenotypes. The work includes the isolation and identification of the active constituents of the leaf as well as the ability of this plant and its components to potentiate the activity of commonly used antibiotics.

General Procedure.
Optical rotation was measured with a Horiba SEPA-300 polarimeter (Horiba, Kyoto, Japan). NMR spectra were recorded on Bruker DMX Avance 600 instruments equipped with an autotune probe and using the automation mode aided by the Bruker program. HREI-SMS spectra were determined on a micrOTOF-Q 98 spectrometer (Bruker-Daltonics, Bilerica, MA). For column chromatography, silica gel 60 particles size 0.04-0.063 mm (Merck, Darmstadt, Germany) and Sephadex LH-20 purchased at Sigma-Aldrich (St Louis, MO) were used. The plates were visualized using UV (254 and 366 nm) and revealed by spraying with vanillin-sulphuric acid (1% ethanolic solution of vanillin + 10% ethanolic sulphuric acid).

Plant Material and Extraction.
The bark and leaf of Acacia polyacantha Willd. (Fabaceae) were collected on February 2016 in Kaéle, in the Far North Region of Cameroon. The plant was then identified at the National Herbarium of Cameroon by Mr. Nana Victor and a voucher specimen was deposited under the registration number 58985/SRF/CAM. Air-dried and powdered leaf and bark of A. polyacantha (2 kg) were extracted twice at room temperature with 4 L of methanol (MeOH) for 48 hours. The solvent was evaporated under reduced pressure to yield 229 g and 108 g of crude leaf (APL) and bark (APB) extracts, respectively. These extracts were then kept at 4 ∘ C until further use.

Microbial Strains and Culture Media.
In this study, 15 Gram-negative bacterial strains belonging to five species were used. They included reference (from American Type Culture Collection, ATCC) and clinical (Laboratory collection) strains of Escherichia coli (ATCC8739, ATCC10536, AG102, and AG100Atet), Enterobacter aerogenes (ATCC13048, CM64, EA27 and EA289), Klebsiella pneumoniae (ATCC11296, KP55 and KP63), Providencia stuartii (ATCC29916 and NEA16), and Pseudomonas aeruginosa (PA01 and PA124). Bacterial features or resistance profiles previously reported [19] are shown as supporting information (Table S2). Bacterial cultures were maintained on agar plates at 4 ∘ C and subcultured on a fresh appropriate agar plates 24 h prior to any antimicrobial assay. The activation of bacteria prior to any assay was done in Mueller Hinton Agar (Sigma) meanwhile antibacterial assays were carried out using Mueller Hinton broth (MHB; Sigma) [20].

Determination of Minimum Inhibitory Concentration (MIC) and Minimum Bactericidal Concentration (MBC). The
MICs and MBCs of extract, fractions and isolated compounds against the tested bacteria were determined by microplate dilution method using the rapid INT colorimetric assay according to previously described methods [21] with some modifications [19,22,23]. In general, the concentrations ranges were 8-1024 g/mL for crude extracts, 4-512 g/mL for fractions, and 2-256 g/mL for chloramphenicol.
The role of efflux pumps in the susceptibility of Gramnegative bacteria to the most active samples (APL, APLb, APLd, compound and CHL) was evaluated by testing the studied samples in the presence of an EPI, PA N (at 30 g/mL) using the rapid INT colorimetric assay as earlier described [7,19]. A preliminary study showed that the concentration of 30 g/mL did not affect the growth of selected bacteria [23]. Nine selected bacterial strains including E. coli ATCC8739 and AG102, E. aerogenes ATCC13048 and CM64, K. pneumoniae KP55 and KP63, P. aeruginosa PA01 and PA124 and P. stuartii ATCC29916 were used. Increase of activity was determined as the ratio of MIC in the absence of EPI versus MIC in the presence of EPI.
To evaluate the potentiating or antibiotic resistance modulating effect of samples, a preliminary assay was performed against a problematic bacterium, P. aeruginosa PA124 (see supporting information S3); the selected samples were tested at various subinhibitory concentrations in combination with antibiotics. MIC/2 and MIC/4 were selected as the best subinhibitory concentrations [6,24] and were further used for the best samples (APL, compounds and ) in combination with antibiotics against the seven other bacteria. Briefly, the MIC was determined as described above. The 96-wells microplate rows receiving antibiotic dilutions without extracts were used for the determination of the MICs of the antibiotics. The concentrations ranges of antibiotics were generally 2-256 g/mL. The MIC was determined as described using INT colorimetric method as earlier described [3,19]. The modulation factor was defined as the ratio of the MIC for the antibiotic alone and that of the antibiotics in the presence of the extract (RHL). Modulation factor ≥ 2 was set as the cutoff for biological significance of antibiotic resistance modulating effects [19,20].

Phytochemistry.
The chemical structures of compounds isolated from the leaf of Acacia polyacantha were determined using NMR ( 1 H and 13 C) data, in comparison with the literature ( Figure 1). Compounds were identified as stigmasterol  [21]. The 1 H NMR, 13 C NMR spectra and major chemical shifts of these compounds are available as supplementary data (S2).

Antibacterial
Activity. Extracts, fractions, and isolated compounds were tested for their antimicrobial activity against the studied Gram-negative bacteria. The results are   (Table 1). APLc as well as compounds , and were not active (Table 2). MIC values ≤ 512 g/mL for fractions and ≤256 g/mL for compounds were obtained against 8/9 (88.9%) tested bacteria for both APLb and APLd and against 7/9 (77.8%) for compound ( Table 2). Analysis of data from Tables 1 and 2 indicated bacteriostatic effects as MBC/MIC ratios were generally above 4 with no MBC value ≤ 1024 g/mL recorded for the crude extract, fractions, and compounds.

Role of Efflux Pumps in the Susceptibility of Gram-Negative Bacteria Botanicals and Phytochemicals.
The most active extracts, fractions, isolated compound and RA (APL, APLb, APLd, Compound , CHL) were tested in the presence of EPI against 9 bacterial strains including reference strains and MDR phenotypes ( Table 3). The results showed that PA N improves the activity (decrease of MIC values) of APL, APLb, APLd and compound on all tested bacteria with the highest MIC values of 256 g/mL for crude extract (APL) and compound , and 64 g/mL for fractions (Table 3). A preliminary study showed that the MIC of PA N was above 256 g/mL on the selected bacteria and that the concentration of 30 g/mL did not affect their growth.

Antibiotic Resistance Modulating Effects of Botanicals and Phytochemicals.
A preliminary study against P. aeruginosa PA124 (see supporting information S3) allowed choosing the appropriate subinhibitory concentrations of MIC/2 and MIC/4 as well as APL, compounds and for further studies. These samples were combined with eight antibiotics to evaluate their possible synergistic effects. The results summarized in Tables 4, 5, and 6 showed that the synergistic effects were noted with all the tested samples and many antibiotics. When tested at their MIC/2, the percentages of bacterial strains on which synergism was observed (PBS) were ≥50% when APL was combined with ERY and CIP (Table 4), when compound was combined with ERY, KAN and GEN (Table 5), and when compound was combined with ERY, and DOX (Table 6). At their MIC/4, the PBS ≥50% was obtained when APL was combined with GEN and CIP (Table 4), when compound was combined with ERY and GEN (Table 5), and when compound was combined with ERY and NOR (Table 6).   , , , and and [15]. However, few compounds were reported in this study, probably due to the fact that we focussed only on the bioactive fractions of the leaf extract; meanwhile, less active fractions were not further investigated.

Antibacterial Activity.
It is important to take into consideration the development of resistance by Gram-negative bacteria when searching for new antimicrobial agents. In the present study, several clinical MDR bacteria expressing MDR phenotypes were used. The MIC values of chloramphenicol were above 10 g/mL on most of the bacterial strains (Table 1), confirming their resistance phenotypes. Established cutoff points for antibacterial activity of botanicals consider that the inhibitory effect is significant when MIC values are below 100 g/mL, moderate when 100 ≤ MIC ≤ 625 g/mL, and weak when MIC > 625 g/mL [18,29]. On this basis, the antibacterial activity of the crude extracts (APL and APB) could mostly be considered as moderate or poor (Table 1). Nevertheless, MIC values below 100 g/mL were obtained with APL on the problematic bacterial strain P. aeruginosa (PA01) and P. stuartii NAE16 (Table 1), as well as with APB against E. aerogenes ATCC13048. This data suggested that these extracts could be useful to fight bacterial infections, especially in traditional medicine where they are utilised. It is worth nothing that the MIC values of APL against P. aeruginosa (PA01) and P. stuartii ATCC29916, or APB against E. aerogenes ATCC13048 (Table 1) were lower than those of chloramphenicol, confirming this hypothesis. This was the rational for carrying out bioguided fractionation in order to isolate more active compounds from the leaf extract. Fractions APLa, APLb, and APLd had MIC values below 100 g/mL against 1, 3, and 7 of the 9 tested bacterial species, respectively (Table 2). This was an indication that fractionation led to more active samples. The activity of phytochemicals was set as significant when MIC<10 g/mL, moderate when 10<MIC<100 g/mL, and low when MIC>100 g/mL [18,29]. However, less active compounds were obtained from the purified fractions, with none of them displaying MIC value below 10 g/mL ( Table 2). This was an indication that constituents of the extract and fractions synergistically inhibited bacterial growth. All tested samples had MBC/MIC ratios above 4 (Tables 1 and 2), showing that they mostly exerted bacteriostatic effects [30]. To the best of our knowledge, the antibacterial activity of A. polycantha as well as that of its most active constituent (compound ) was reported for the first time.
Concerning the structure-activity relationship, it appeared that terpenoids and (with no sugar in their chemical structures) and (with only one sugar) as well as      (Table 2). Flavonoids and as well as saponins and had selective and poor antibacterial. Within saponins, it can be noted that the presence of a second galactopyranosyl substituent (compound ) instead of xylopyranosyl (compound 7) significantly increased the antibacterial activity, with compound displaying MIC values ≤ 64 g/mL against 7/9 tested bacteria versus 0/9 for compound ( Table 2).

Role of Bacterial Efflux Pumps.
The clinical MDR bacteria tested in this work overexpressed efflux mechanism via the efflux pumps of the resistance nodulation cell division (RND) family, namely, AcrAB-TolC for enterobacteria such as E. coli, E. aerogenes, K. pneumoniae, and P. stuartii and MexAB-OprM for P. aeruginosa [31][32][33][34][35][36]. These efflux pumps expel toxic compounds (including antibiotics) out of the bacterial cytoplasm, preventing them from reaching their intracellular target [37]. Efflux Pump Inhibitors, such as Pa N, could be used to restore the intracellular concentration of antibacterials acting on intracellular target by blocking the bacterial efflux pumps. In the presence of PA N, it was observed that the activity of the crude extract (except against P. aeruginosa PA01), fractions APLb and APLd as well as compound and CHL strongly increased on almost all tested bacteria ( Table 3). The fold increase ranged from 2 to ≥ 4 for APLb, from ≥ 1 to 8 for CHL, from 2 to ≥ 128 for compound , from 0.5 to 64 for APL, from 2 to 128 for APLd. This clearly indicated that compound as well as other active constituents of the APL are substrates of bacterial efflux pumps and that they may have an intracellular target [38]. Consequently, the development of an antibacterial drug combination of compound , as well as extracts or fractions with an EPI, could be an interesting strategy to tackle MDR bacterial infections. In effect, modulation factor ≥ 2 define a biologically significant antibiotic resistance modulating substance [19,20]. Previous study demonstrated that PA N could also restore the activity of several natural compounds on MDR bacteria expressing active efflux pumps, with MIC values decreasing below 10 g/mL in most of the tested bacteria for the coumarin, MAB3, the xanthone, laurentixanthone B, the naphthoquinones: diospyrone and plumbagin and the flavonoids: 4-hydroxylonchocarpin and isobavachalcone [22,23].

Antibiotic Resistance Modulating Effects.
Difficulties in the field of novel antibacterial drug discovery, for combating resistant pathogens, have propelled the search for new alternative medicine to improve or to restore the activity of commonly used antibiotics. Combining antibiotics with botanicals and phytochemicals is an attractive strategy as regards the diversity of secondary metabolites from natural source. If an antibacterial substance improves the activity of at least 70% of the tested antibiotics on more than 70% tested bacterial strains, it might be considered as a potential efflux pump inhibitor [39]. However, this was not the case in the present study, as neither APL nor compounds and were able to exert such degree of synergistic effects with antibiotics (Tables 4-6). However, synergistic effects were observed between APL, compounds and with at least one of the eight tested antibiotics against at least 50% of the MDR bacterial strains (Tables 3-6). This suggests that possible combination of these samples with specific antibiotics could help in antibacterial chemotherapy.

Limitations
Our study has limitations. The toxicity of this plant also needs to be performed to evaluate its safety.

Conclusion
In the present study, the antibacterial activity of the crude extract, fractions, and compounds from the leaf of Acacia polyacantha Willd. (Fabaceae) was investigated. It was found that the leaf extract was more active than the bark extract. -D -galactopyranosyl]-oleanolic acid. Saponin was the major antibacterial constituent of the plant and acted as a substrate of bacterial EPI. Although the crude extract and its constituents were not EPI, they showed synergistic effects with several antibiotics and could be selectively used in bacterial chemotherapy. The overall results demonstrated that Acacia polyacantha is a source of antibacterial drug that should be explored further to develop novel substances to combat both sensitive and MDR phenotypes.