Antibacterial constituents of three Cameroonian medicinal plants: Garcinia nobilis, Oricia suaveolens and Balsamocitrus camerunensis

Background Multidrug resistance is a worrying cause of treatment failure in bacterial infections. The search of bioactive constituents from medicinal plants against multidrug resistant (MDR) bacteria has significantly evolved in the two last decades. In the present study, twenty-two compounds (three terpenoids, eleven phenolics and eight alkaloids) isolated from three Cameroonian medicinal plants, namely Garcinia nobilis, Oricia suaveolens and Balsamocitrus camerunensis, as well as the crude extracts were tested for their antibacterial activities against Mycobacterium tuberculosis and Gram-negative bacteria amongst which were MDR active efflux pumps expressing phenotypes. Methods The microplate alamar blue assay (MABA) and the broth microdilution methods were used to determine the minimal inhibitory concentrations (MIC) and minimal bactericidal concentrations (MBC) of the studied samples. Results The results of the MIC determinations indicate that, the best crude extract was that from G. nobilis (GNB), its inhibitory effects being noted against 12 of the 14 tested bacteria. The extract of GNB also exhibited better anti-tuberculosis (MIC of 128 μg/ml M. tuberculosis against ATCC 27294 strain) and antibacterial (MIC of 64 μg/ml against Escherichia coli ATCC10536) activities compared to the extracts of O. suaveolens and B. camerunensis. Interestingly, 4-prenyl-2-(3,7-dimethyl-2,6-octadienyl)-1,3,5,8-tetrahydroxyxanthone (2), isolated from the most active extract GNB, also showed the best activity amongst compounds, inhibiting the growth of all the fourteen tested microorganisms. The lowest MIC value obtained with compound 2 was 8 μg/ml against M. tuberculosis ATCC 27294 and M. tuberculosis clinical MTCS2 strains. Other compounds showed selective activities with 11 of the 14 tested bacteria being sensitive to the xanthone, morusignin I (5) and the alkaloid, kokusaginine (13). Conclusions The results of the present investigation provide evidence that the crude extract from G. nobilis, O. suaveolens and B. camerunensis as well as some of their compounds, and mostly compound 2 (isolated from G. nobilis,) could be considered as interesting natural antibacterial products.


Background
The continuous emergence of multidrug-resistant (MDR) bacteria drastically reduced the efficacy of our antibiotic armory and, consequently, increases the frequency of therapeutic failure. Drug resistance is a consequence of the worldwide use of antibiotics, and the acute challenge for health care is to find measures to efficiently combat resistant organisms [1]. Several natural compounds have successfully been tested for their abilities to prevent the growth of MDR bacteria [2]. In our continuous search of antibacterial drugs from natural source, we targeted three Cameroonian medicinal plants, Garcinia nobilis Engl. (Clusiaceae), Oricia suaveolens Engl. [Commonly known in West Africa as Abe iolo or Kru-bete parihi (Ivory Coast) or Mende jagbole (Sierra Leone)] and Balsamocitrus camerunensis Letouzey, R. (Rutaceae). Plants of the genus Garcinia, widely distributed in the tropical Africa, Asia, New Caledonia, and Polynesia, have yielded many biologically active and structurally intriguing natural products [3]. Garcinia species are known to contain a wide variety of oxygenated and prenylated xanthones, as well as polyisoprenylated benzophenones such as the guttiferones [4]. Previous studies of the chemistry of the genus Oricia, including O. suaveolens revealed the presence of alkaloids and triterpenes [5,6]. B. camerunensis is a new species recently found in Batouri (Cameroon) and Boukoko (Central African Republic) [7]. Plants of genus Balsamocitrus (including the decoction of the bark of B. camerunensis) are used in traditional African medicine to treat malaria, hypertension, infertility, and influenza [8,9]. Previous phytochemical investigations of this genus revealed the presence of coumarins, quinoline alkaloids, free aliphatic acids and steroids, some of these compounds exhibiting potent antibacterial, fungicidal, and algicidal properties [8].

Methods
General experimental procedure 1 H and 13 C NMR spectra were recorded in chloroform on Digital NMR BRUCKER AVANCE 400 and 500 MHz. EIMS were obtained on Joel the MS route JMS.600H and HREIMS were performed on a thermo finnigan Mat 95 XP. Thin layer chromatography (TLC) and pre-coated TLC were performed on silica gel GF 254 (Merck). Column chromatography (CC) was performed on silica gel (Merck) type 100(70-230 Mesh ASTM) eluted either with gradient system (Hex-Ethyl Acetate-MeOH; Hex-CH 2 Cl 2 -MeOH and Hex-CH 2 Cl 2 -Acetone). All the solvents used were distilled commercial grade. The isolated compounds were crystallized from the same solvent and their purity was checked by TLC. Pre-coated plates of silica gel GF 254 (Merck) were used for this purpose; the spots were detected with UV lamp at 254 and 366 nm and by spraying with 50% H 2 SO 4 or ceric sulfate following by heating.

Plant material
The stem bark of the G. nobilis was collected in Okola-Yaounde (

Extraction
The air-dried and powdered samples from each plant were macerated in either 7.0 L methanol (MeOH) for the stem bark of G. nobilis (2 kg) or in 10 L MeOH/ dichloromethane (CH 2 Cl 2 ) mixture for the wood (4 kg) and roots (3 kg) of O. suaveolens, and the stem bark (3.8 kg) of B. camerunensis. The extraction was done at room temperature for two days. The evaporation under reduced pressure yielded the crude extracts from the stem bark of G. nobilis (GNB; 100 g), wood (OSW; 173 g) and roots (OSR; 145 g) from O. suaveolens, and from the stem bark B. camerunensis (BCB; 128 g).

Antibacterial assays Microbial strains and culture media
The studied microorganisms included reference and clinical strains ( Tables 1 and 2) of Escherichia coli, Enterobacter aerogenes, Klebsiella pneumoniae, Providencia stuartii, Pseudomonas aeruginosa, a drug-susceptible strain of M. tuberculosis H37Rv obtained from the American Type Culture Collection, and two clinical strains of M. tuberculosis MTCS1, MTCS2. M. tuberculosis was plated on Löwenstein-Jensen medium and allowed to grow for 3-4weeks at 37°C. Middlebrook 7H9 broth supplemented with 0.2% glycerol and 10% Oleic Acid-Albumin-Dextrose-Catalase (OADC) (Sigma) was used to determine the MIC and MBC values of the test samples on M. tuberculosis. We previously reported the features [28] of all the tested Gramnegative bacteria. Nutrient agar was used for the activation of bacteria other than M. tuberculosis strains [28]. The clinical strains used in this work are our laboratory collection previously obtained from Yaoundé General Hospital (Cameroon), and from the Mediterranean University (Marseille, France).

Microplate Alamar blue assay (MABA) against M. Tuberculosis
The activity of all samples against M. tuberculosis strains was tested using the MABA [32]. Briefly, each of the above M. tuberculosis strains was cultured at 37°C in Middlebrook 7H9 broth supplemented with 0.2% glycerol and 10% OADC (oleic acid-albumin-dextrosecatalase; Sigma) until logarithmic growth was reached. The homogenous culture was obtained using sterile glass beads and vortex. About 6×10 6 CFU/ml inoculum (100 μl) of M. tuberculosis was then added to the two fold serially diluted samples (100 μl). The final concentration of DMSO in all assays was 2.5% or less and this dilution also served as solvent control. The samples were assayed in triplicate. All tests were carried out in sterile flat-bottomed 96-well microplates. Each microplate was incubated for 5 days at 37°C in a sealed plastic CO 2 -permeable bag. After 5 days of incubation, 32 μl of a mixture of freshly prepared Alamar Blue solution and 20% sterile Tween-80 (Sigma) 1:1 v/v were added to one growth-control well. The microplates were incubated again at 37°C for 24 h. If a color shift from blue to pink was observed in the growth-control sample, 32 μl of alamar blue solution was added to each of the remaining wells, and the microplate was further incubated for 24 h. A well-defined pink color was interpreted as positive bacterial growth, whereas a blue color indicated an absence of growth. The MIC corresponded to the greatest dilution of sample extract in which the color shift from blue to pink was not observed. Samples with recorded MIC values following MABA were assayed for their mycobactericidal effect [32]. Briefly, 5 μl of the undeveloped mycobacterial suspensions were transferred from the former to a new microplate containing 195 μl of fresh culture medium per well. Three wells were inoculated with 100 μl of fresh inoculum as for MABA and three more wells were incubated with 200 μl of culture medium only, as negative controls. The microplates were incubated and developed with alamar blue as for MABA. The MBC corresponded to the minimum sample concentration that did not cause a color shift in cultures that were re-incubated in fresh medium.

Results and discussion
The chemical structures of the compounds isolated from G. nobilis, O. suaveolens and B. camerunensis (Figure 1) were established by spectroscopic methods. The compounds were isolated from the stem bark of G. nobilis (1)(2)(3)(4)(5), the woods of O. suaveolens (6-11), the roots (12)(13)(14)(15)(16)(17) and the stem bark of B. camerunensis (7,(18)(19)(20)(21)(22). The twenty two isolated compounds were identified as caroxanthone (1), 4-prenyl-2-(3,7-dimethyl-2,6-octadienyl)-1,3,5,8-tetrahydroxyxanthone (2), smeathxanthone A (3), 8-hydroxycudraxanthone G (4), morusignin I (5), stigma-5en-3-ol (6), lupeol (7), evoxanthine (8), norevoxanthine (9), 1-hydoxy-2,3-dimethoxy-10-methylacridone(10), 1,3-dime thoxy-10-methylacridone (11), skimmianine (12), koku saginine (13), montrifoline (14), 1-hydroxy-4-methoxy-10-methylacridone (15), syringaresinol (16), limonin (17), 1-hydroxy-3,6-dimethoxy-8-methylxanthone (18), xantho xyletin (19), imperatorin (20), scoparone (21) and umbe lliferone (22) [11][12][13][14][15][16][17][18][19][20][21][22][23][24][25]. Amongst the twenty-two compounds were three terpenoids (6, 7 and 17), eleven phenolic compounds (1-5, 16, 18-22), and eight alkaloids (10)(11)(12)(13)(14)(15)(16)(17). The isolated terpenoids were steroid (6), triterpenoid (7), and limonoid (17) whilst the alkaloids included five acridones (8)(9)(10)(11)15) and three furanoquinolines (12)(13)(14). The phenolics obtained herein were six xanthones (1)(2)(3)(4)(5)18), one lignan (16) and four coumarins (19)(20)(21)(22). The isolation and identification of compounds 1-5 from G. nobilis was previously reported [10]. The occurrence of alkaloids and terpernoids from O. suaveolens was reported [5,6], and their isolation in the present study is in consistence with previous reports. The occurrence of coumarins, quinoline alkaloids, and free aliphatic acids was also reported in B. camerunensis [9]. However, in this study, only coumarins were isolated. The crude extracts as well as the isolated compounds [excluding compounds 6, 7, 12 (known to have low or no antimicrobial activity), and 14 (isolated in very low quantities)] were tested for the antibacterial activities against Gram-negative bacteria and M. tuberculosis and the results are summarized in Tables 1 and 2. The results of the MIC determinations (Table 1) showed that the crude extract from G. nobilis (GNB) was the most active amongst the studied extracts, its inhibitory effects being noted on 12 of the 14 tested bacteria. GNB also exhibited the best activity against M. tuberculosis ATCC 27294 (MIC of 128 μg/ml) and E. coli ATCC10536 (MIC of 64 μg/ml) than OSR, OSW and BCB. The inhibitory effects of the extracts OSR and OSW from O. suaveolens were noted on 9/14 and 7/14 studied bacteria respectively, meanwhile that of the extract BCB of B. camerunensis was observed on 4/14 pathogens tested. Interestingly, compound 2 isolated from the most active extract GNB, also exhibited the best activity, preventing the growth of all the fourteen tested microorganisms. The lowest MIC obtained with compound 2 was 8 μg/ml against M. tuberculosis ATCC 27294 and the clinical MTCS2 strains. It is noteworthy that compound 2 was more active than chloramphenicol on two Gram-negative MDR bacteria, namely E. coli AG102 and E. aerogenes CM64 ( Phytochemicals are routinely classified as antimicrobials on the basis of susceptibility tests that produce MIC in the range of 100 to 1000 μg/ml [33]. The activity is considered to be significant if MIC values are below 100 μg/ml for crude extract and moderate when the MICs vary from 100 to 625 μg/ml [34,35]. Also, the activity of compounds is considered to be significant when the MIC is below 10 μg/ml, and moderate when such values vary between 10 and 100 μg/ml [34,35]. On the basis of such criteria, the activity of the studied crude extracts can mostly be considered as moderate, though a significant effect was observed with GNB on E. coli ATCC strains. Compound 2 was significantly active against M. tuberculosis ATCC 27294 and MTCS2 strains. However its activities can also be considered as moderate against the majority of the bacteria tested. All the tested compounds were active on at least one of the studied microorganisms, and their presence can explain the activity of the crude extracts. Nonetheless, it can be observed that the activity of GNB was not detected on both P. stuartii ATCC NAE16 and P. aeruginosa PA124, while the extract yielded at least one active compound on these bacteria (Table 1). This can be explained by the fact that the activity of the crude extract does not only depend on the presence of the active compounds, but is also influenced by the quantity and/or possible interaction with other constituents of the plant. This observation can also be applied when carefully analysing the activities of the crude extracts from O. suaveolens and B. camerunensis and their constituents ( Table 1). The activity of the crude extracts and compounds studied herein (and mostly compound 2) can be considered interesting when regarding the medical importance of the studied bacteria. In fact, the clinical MDR Gram-negative bacteria tested express active efflux pumps and are involved in many therapeutic failures [28]. To the best of our knowledge, the antibacterial activities of the extracts of G. nobilis, O. suaveolens and B. camerunensis as well as those of most of the studied compounds are being reported for the first time. Nevertheless, some of the isolated compounds were reported for their antibacterial properties. In effect, lupeol (7) is known to have low antibacterial activities, the lowest MIC value obtained against Enterococcus faecalis ATCC 29212 being 63 μg/ ml [36]. The activity of compound 7 was also not detected against the sensitive Mycobacterium smegmatis when it was tested at a concentration up to 312 μg/ml [37] and consequently, this compound was not tested again in this work. Previously we also reported the moderate activity of the alkaloid 13 on some sensitive Gramnegative bacteria as well as it low effect against M. smegmatis [37]. This compound was not more active against MDR bacteria as observed in the present work, confirming its low antibacterial potential. It has also been demonstrated that compounds 3 (active only on two of 21 tested microorganisms) [38] and 16 (inhibition zone diameters for 10 μl discs at 10 4 ppm reported as 0.0; 0.2; 0.4 and 0.5 mm against K. pneumoniae, S. aureus, Pseudomonas syringae and Bacillus subtilis respectively) [39] have low antibacterial activities. Such reports are in accordance with the results obtained in the present study.
Compounds 19 (xanthoxyletin) and 22 (umbelliferone) were also found inactive against M. tuberculosis and poorly active against Gram-negative bacteria, consolidating the previously reported data [41,42]. It is noteworthy that compound 12 was reported not to have any bacterial activity [43] and was not tested in the present work.
When regarding the structure-activity relationship, it appears that coumarins had the lowest activities, none of them been active against M. tuberculosis. The tested tetranortriterpenoid, compound 17 also showed very weak antibacterial activities. These results are in consistence with previous studies, showing the low antibacterial activities of terpenoids and coumarins against bacteria expressing MDR phenotype [2]. Alkaloids also showed low antibacterial activities. However moderate inhibitory effects were noted with 10 (1-hydoxy-2,3-dimethoxy-10methylacridone) and 13 (kokusaginine) respectively against E. coli ATTC and AG100 strains (Table 1). Amongst alkaloids, compound 13, one of the three isolated furanoquinolines, showed the best spectrum of activity (contrary to the five acridone alkaloids) and was found active on both M. tuberculosis and Gram-negative bacteria. Within the acridone alkaloids, and when comparing the effects of compounds 10 and 15 [antibacterial spectra (7/14 of the tested bacteria for 10 and 4/14 for 15); lowest MIC value (64 μg/ml for 10 and 128 μg/ml for 15)], it appears that the presence of methoxyl-(−OCH 3 ) group in C2 (compound 10) increases the bacterial susceptibility (Figure 1, Table 1). A comparison of the activities of compounds 8 and 9 [the presence of -OCH 3 group in C1 (compound 8) instead of -OH group (compound 9)] on one hand, and those of 11 and 15 [the presence of -OCH 3 group in C1 (compound 11) instead of the hydroxyl (−OH) group (compound 15)] on another hand seems to confirm the fact that the natural substitution of -OCH 3 by -OH in the studied alkaloids increases their activities ( Figure 1, Table 1). The best antibacterial activities were recorded with xanthones. Within the studied xanthones and when regarding the activities of compounds 2 and 3, it appears that the presence of additional prenyl-group in C4 significantly increases the antibacterial activity ( Figure 1, Table 1). Between compounds 4 and 5, it can also be deduced that the cyclisation increase the activity (Figure 1, Table 1).

Conclusion
The results of the present investigation are important, in regards to the medical importance of the studied microorganisms. Hence, these data provide evidence that some of the constituents of G. nobilis, O. suaveolens and B. camerunensis and mostly compound 2 (isolated from G. nobilis), as well as some of the crude extracts from the three plants could be potential antimicrobial products to fight MDR bacteria. The combination of the three plants as used locally in the treatment of infections will further be investigated to provide better understanding to the traditional use.