Tuberculosis (TB) is among the major causes of mortality due to a single infectious bacterium. The burden of TB is higher due to multidrug-resistant Mycobacterium tuberculosis strains, which lead to treatment failures. The present study conducted in silico studies of bioactive compounds isolated from Kigelia africana (Lam.) Benth stem bark.
Methods
Pure compounds were isolated from dichloromethane/methanol stem bark extract of Kigelia africana after repeated column chromatography. The chemical structures of the isolated compounds were established based on 1H NMR, 13C NMR and 2D NMR (COSY, HSQC and HMBC) spectroscopy. In silico analyses were performed to assess the drug-likeness, pharmacokinetics and antibacterial potential of the compounds against target proteins (4QIJ and 5HKF, 5L3J and 8GZY, and 2QIL) from M. tuberculosis, Escherichia coli, and Staphylococcus aureus, respectively.
Results
Four compounds: demethylkigelin, tyrosyl butyrate, stearic acid and stigmasterol were characterized. In molecular docking studies, the compounds showed binding affinities ranging from −4.4 to −9.3 kcal/mol against target proteins 4QIJ, 5HKF, 5L3J, 8GZY and 2QIL. Stigmasterol (L4) had the highest affinity against the highest binding affinity, with a score of −9.3 kcal/mol against S. aureus protein 2QIL. It also showed strong affinities against M. tuberculosis (4QIJ and 5HKF) and E. coli (5L3J and 8GZY) targets. In silico toxicity profiling predicted tyrosyl butyrate and stearic acid to be relatively safe whereas demethylkigelin and stigmasterol showed potential respiratory and cardiotoxic effects that needs further safety evaluation.
Conclusion
Kigelia africana stem bark possesses bioactive compounds that are potential inhibitors of M. tuberculosis with good to better binding affinities and stable interactions. Future studies should validate the in vitro and in vivo bioactivity as well as toxicity of the compounds.
Tuberculosis (TB), whose primary causative agent is Mycobacterium tuberculosis (M. tuberculosis), is among the major causes of mortality due to a single infectious bacterium.1 Initially superseded by COVID-19, TB has now returned to be the major cause of mortality in HIV-infected patients and those related to antimicrobial resistance.2 According to the World Health Organization, the high global burden of TB is due to multidrug (isoniazid/rifampicin)-resistant M. tuberculosis strains, leading to treatment failures when the current regimens are used.2–4 Hepatotoxicity caused by first-line antitubercular drugs used in the treatment of latent as well as active M. tuberculosis infections has also been reported.5 Further, individuals co-infected with TB and HIV require concurrent administration of antiretroviral therapy and antitubercular drugs to improve survival. However, this combination is currently associated with challenges such as drug interactions, cumulative toxicities, and the development of TB-associated immune reconstitution inflammatory syndrome.6 Thus, there is need to find novel or repositioned antimycobacterial agents with unique modes of action and improved safety profiles.7
Natural products have been a source of antimicrobial agents, with several frontline drugs originating or possessing templates from microbial and plant sources.8Kigelia africana (Lam.) Benth. (Synonym: Kigelia pinnata, the so called ‘‘sausage tree’’), is a member of Bignoniaceae family and the sole species in its genus. Its roots, bark, fruits, flowers, leaves and seeds are utilized in African and Asian traditional formularies to manage skin infections, infertility, gynecological complaints, venereal diseases, hypertension, rheumatism, toothache, malaria, cough, asthma and TB.9–11 Bioactivity studies have revealed that extracts from this species has antimicrobial, antioxidant, anti-inflammatory, analgesic, antimalarial, hepatoprotective, antiulcerogenic, antidiabetic and anticancer activities.12,13 These bioactivities are known to be due to the presence of phytochemicals such as phenolics, iridoids and limonoids,13 naphthoquinones, phenyl ethanoglycosides, terpenes, terpenoids, lignans and steroids.14,15
The stem bark of K. africana is utilized for TB treatment,9,10,16 but studies targeting the isolation and evaluation of antitubercular compounds from this species are limited. Only one study by Wadhwani et al17 isolated nine phytochemicals (cycloolivil, cluytyl ferulate, dehydro-α-lapachone, D-sesamin, kigelinone, lapachol, paulownin, tecomaquinone-I, and wodeshiol) from the heartwood of K. africana followed by in silico analyses. The authors reported that tecomaquinone-I was the only compound that exhibited potent antitubercular activity.
In the present study, in silico analyses were performed to investigate the potential antimycobacterial activity of known compounds isolated from K. africana stem bark against key M. tuberculosis, Escherichia coli (E. coli) and Staphylococcus aureus (S. aureus) targets. Through prediction of the binding affinities and interaction profiles, the in silico approach provides results that support the development of K. africana-derived scaffolds as promising candidates for novel and repurposed antitubercular drugs.
Methods
Plant Material
Stem bark of K. africana was collected from Bupeni village, Kaliro District, Eastern Uganda in March 2019. The plant was identified by Mr. Protase Rwaburindore, a botanist at Makerere University Herbarium where a voucher sample of the stem bark (no. IG2019/002) was deposited. The bark was shade-dried for two weeks, and ground using a blender, yielding 1 kg of a brownish-red powder.
Extraction and Isolation of Pure Compounds
The dichloromethane/methanol (1:1) crude extract was obtained by maceration for 24 h. The extraction was repeated twice, and the pooled extract concentrated by rotary evaporation. The crude extract (150 g) was divided into two portions: one half was stored in a refrigerator and the other was subjected to chromatographic fractionation.
Briefly, 200 g of silica gel (70-230 mesh size) was used for column chromatography, using hexane as the mobile phase, and increasing polarity using ethyl acetate. Different fractions collected were spotted on TLC plates, and those with similar TLC profiles were combined. Fraction I eluted with 2% ethyl acetate in hexane yielded an amorphous solid (compound 1, 35 mg). Fraction II (eluted with 5% ethyl acetate in hexane) was further purified over Sephadex LH-20 (dichloromethane/methanol, 1:1) to give a colorless, oily liquid (compound 2, 25 mg). Fraction III yielded a white solid (compound 3, 29 mg), following chromatographic separation with 15% ethyl acetate in hexane. Fraction IV eluted with 10% ethyl acetate in hexane yielded a white amorphous solid (compound 4, 47 mg).
Spectroscopic Analysis of Compounds
The isolated compounds were subjected to NMR spectroscopy on a Bruker AV-500 spectrometer, that is, 1D (1H and 13C) and 2D (1H-1H COSY, HSQC and HMBC) analysis of the samples dissolved in deuterated chloroform. The spectra in FID format were processed using MestReNova (version 8.1.1). The residual chloroform peaks were used as references.18
Protein Preparation
The Autodock 4 (1.5.7) was selected and utilized to prepare the Protein Data Bank (PDB) file for the proteins of M. tuberculosis (PDB ID: 4QIJ and 5HKF),19E. coli (PDB ID: 5L3J and 8GZY) and S. aureus (PDB ID: 2QIL).20,21 The structural integrity was modified by removing the water molecules, ions, the Kollman charges were added and the other cofactors, and then followed by addition of the polar hydrogen atoms to the system. The protein structures were subsequently saved in PDBQT format. To realign all the structural loops and determine the binding active sites, BIOVIA Discovery Studio v.2025 Visualizer software was used.22
Ligand Preparation
Information on the antibacterial potential of K. africana constituents against E. coli and S. aureus was obtained from previous studies,23,24 and the phytochemicals selected for this study were those characterized (1-4). The structures of the isolated compounds and the FDA approved antitubercular drugs (Isoniazid and Ethionamide) were downloaded from PubChem.25 The initial SDF format structure of the ligands was converted into PDB format and their structures were subsequently saved in PDBQT format using Autodock 4.
Drug Likeness Properties and ADME Analysis of the Phytochemicals
The SMILES format of the compounds was pasted onto ADMETLAB3.0 (https://admetlab3.scbdd.com/documentation) and SWISSADME (https://www.swissadme.ch/index.php) to generate their pharmacokinetic properties/ADMET and drug-likeness parameters.26,27 The Bioavailability Radar tool was used to determine the oral bioavailability of the compounds.28,29
Molecular Docking
Molecular docking was done using AutoDock Vina (1.1.2) with automation using Shell script, a configuration file containing the grid parameters and center values obtained from AutoDock 4 were passed as parameters.30,31 The binding energies (kcal/mol) were obtained together with the output pdbqt file.32 Vina_Split software was used to separate the output file into individual poses for the ligands. BIOVIA Discovery Studio was used to analyse the interactions and generate the 2D and 3D structures of the best poses with minimal energies determined by AutoDock.
Toxicity Prediction
Toxicity of the compounds was predicted using the ProTox-II web server (https://tox.charite.de/protox3/), an in silico platform for computational toxicology. The chemical structures, in SMILES format, were submitted to the server to predict a wide range of toxicological endpoints. Predictions were based on an integrated methodology that includes molecular similarity, pharmacophore-based analyses, and machine learning models trained on publicly available toxicological data. The endpoints evaluated included hepatotoxicity, respiratory toxicity, cardiotoxicity (hERG inhibition), mutagenicity (Ames test), and organ-specific toxicities, each reported with an associated probability score. The acute oral toxicity (LD50) of each compound was predicted.33
Statistical Analysis
Molecular docking results were analyzed descriptively based on the binding affinities (kcal/mol) and the visualized interactions of ligands with target proteins. Binding affinities of the test ligands were compared to those of reference antitubercular drugs (isoniazid and ethionamide). Drug-likeness and ADMET properties were interpreted according to standard rules (Lipinski, Ghose, Veber, Egan, Muegge). The predicted LD50 values of the isolated compounds were classified according to the Globally Harmonized System (GHS).33 Due to the computational nature of the study, no further inferential statistical analyses were performed.
Results
Isolated Compounds from K. africana Stem Bark
Four compounds were isolated from the stem bark of K. africana (Figure 1). They were conclusively identified as demethylkigelin (1), tyrosyl butyrate (2), stearic acid (3), and stigmasterol (4). Compound 1 was isolated as a white amorphous solid. The 13C NMR spectrum of the compound (Table 1; Figure S1) showed eleven signals including peaks due to an aromatic ring system (δC 102.3, 105.4, 133.1, 135.6, 155.0 and 155.6), an aliphatic cyclic system (δC 20.8, 34.6, 76.0) with oxygen as a heteroatom, and a diortho-substituted aromatic methoxyl peak as collaborated with HSQC data at 61.0 ppm (7-OCH3; Figure S2).
Structure of Compounds Isolated from Dichloromethane/Methanol Extract of K. africana Stem Bark: Demethylkigelin (1), Tyrosyl Butyrate (2), Stearic Acid (3) and Stigmasterol (4).
NMR Spectral Data (1H NMR, CDCl3 500 MHz and 13C NMR, 125 MHz) of Compound 1.
Position
C (in ppm)
H (mult., J in Hz)
Integration
HMBC
1
170.2
–
3
76.0
4.66 (m)
1H
C-5a (weak)
4a
34.6
2.83 (m)
1H
C-3, C-5, C-5a, C-8a, C-9
4b
2.84 (m)
1H
5
105.4
6.32 (s)
1H
C-4, C-6, C-7, C-8a
6
155.0
–
7
133.1
–
8
155.6
–
8a
102.3
–
9
20.8
1.51 (d, 6.1)
3H
C-3, C-4
7-OMe
61.0
3.98 (s)
3H
C-7
8-OH
–
11.38 (br s)
1H
C-7, C-8, C-8a
The presence of a signal at δC 170.2 in the 13C NMR spectrum, was suggestive of a lactone moiety. The 1H NMR spectrum (Figure S3) exhibited broad singlet at δH 11.38, attributed to the 8-OH group involved in an intermolecular hydrogen bond with the lactonyl carbon oxygen (C-1). Both the 1H NMR and the 1H-1H COSY spectra (Figure S4) revealed the presence of 2-hydroxy-propyl moiety of a typical cyclic ester [δH 1.51, d (J = 6.1 Hz, 3H, H-9), 2.83-2.84, m (2H, H-4a & H-4b), 4.66, m (1H, H-3)] whose attachment at C-5a of the aromatic ring was based on the HMBC correlation (Figure S5) between H-4a, 4b with carbons resonating at δC 102.3 (C-8a), 105.6 (C-5) and 135.6 (C-5a). These partial structures were joined to give the final structure of compound 1.
Compound 2 was isolated as a colorless and odorless oily liquid. It was identified to be an ester based on its 13C NMR spectrum that displayed a carbonyl peak at δC 174.0 (Figure S6, Table 2). The 1H NMR spectrum (Figure S7) showed a characteristic AA′XX′ spin system [δH 7.08 (m, 2H, H-2/6) and 6.76 2H, (H-3/5)] which is attributable to a para-substituted phenyl ring. The presence of an ethenyloxy group was deduced from the 1H NMR spectrum which showed corresponding signals of an A2X2 spin system at δH 2.86 (t, J = 7.1 Hz, H-1′) and 4.23 (t, J = 7.1 Hz, H-2′). This was further supported by the 13C NMR and HSQC spectra (Figure S8) which exhibited carbon peaks at δC 34.4 (C-1′) and 65.1 (C-2′). The 1H NMR spectrum further displayed features of an alkanoate with a terminal methyl group [δH 0.90, (t, J = 6.9 Hz, H-4′′). This was also supported by the HMBC correlation observed between methylene protons at δH 2.27 (H-2′′) and the carbonyl carbon at δC 174.0 (C-1′′). Complete assignment was achieved on the basis of HSQC, HMBC and 1H-1H COSY correlations (Figures S8, S9 and S10).
NMR Spectral Data (1H NMR, CDCl3 500 MHz and 13C NMR, 125 MHz) of Compound 2.
Position
δC (in ppm)
δH (mult., J in Hz)
1
130.2
–
2/6
130.2
7.08 (m)
3/5
115.6
6.76 (m)
4
154.3
–
1′
34.4
2.86 (t, 7.1)
2′
65.1
4.23 (t, 7.1)
1′′
174.0
–
2′′
34.5
2.27 (t, 7.5)
3′′
22.8
1.30 (m)
4′′
14.3
0.90 (t, 6.9)
Compound 3 was isolated as a white amorphous solid. The 1H NMR spectrum of compound 3 (Table 3; Figure S11) displayed terminal methyl protons at δH 0.88 (t, J = 6.9 Hz, H-18). The presence of a carboxylic acid was deduced from the 13C NMR spectrum of compound 3 (Figure S12) which exhibited a signal at δC 180.5 (C-1). The 1H NMR and HSQC spectra of compound 3 (Figure S13) revealed fourteen CH2 groups at δH 1.17-1.40 (m, 28H, CH2-4 to CH2-17) characteristic of an aliphatic chain. The signals at δH 2.34 (t, J = 7.5 Hz) and 1.63 (quintet, J = 7.5 Hz) in the 1H NMR spectrum correspond to the methylene protons at H-2 and H-3, respectively, located adjacent to the carboxyl group. The HMBC correlations (Figure S14) observed between the carbonyl carbon at δC 180.5 (C-1) with both protons resonating at δH 2.34 and 1.63 supported the assignment of H-2 and H-3. Together with the information from the 1H-1H COSY spectrum (Figure S15), the partial structures were joined to give the final structure of compound 3.
NMR Spectral Data (1H NMR, CDCl3 500 MHz and 13C NMR, 125 MHz) of Compound 3.
Position
δC (in ppm)
δH (mult., J in Hz)
1
180.5
–
2
34.0
2.36 (t, 7.5)
3
24.7
1.63 (quintet, 7.5)
4-15
29.2-29.9
1.17-1.40 (m)
16
32.0
1.28 (m)
17
22.8
1.28-1.30 (m)
18
14.3
0.88
Compound 4 was isolated as a white amorphous solid. It was active against iodine vapor which implied presence of unsaturation. Its 13C NMR spectrum (Table 4; Figure S16) showed twenty-nine signals. Four of the signals at δC 141.5, 138.9, 130.0 and 122.0 ppm were for olefinic carbon atoms. The other signals included the six methyl carbons δC 12.2, 12.3, 19.8, 21.6, 22.7 and 23.05, one of which at δC 72.2 ppm was for an oxygenated C-3 of steroids. The 1H NMR spectrum of compound 4 (Figure S17) showed the presence of two methyl singlets at δH 0.71, and 1.01; three methyl doublets that appeared at δH 0.80, 0.82, and 0.91; and a methyl triplet at δH 0.83.
NMR Spectral Data (1H NMR, CDCl3 500 MHz and 13C NMR, 125 MHz) of Compound 4.
Position
δH (mult., J in Hz)
δC (in ppm)
1
1.84 (m)
36.8
2
1.88 (m)
30.2
3
3.52 (m)
72.2
4
2.31 (m)
42.9
5
–
141.5
6
5.35 (m)
122.0
7
1.50 (m)
32.5
8
1.45 (m)
30.5
9
0.94 (m)
50.7
10
–
36.7
11
1.06, 1.56 (m)
24.4
12
1.18 (m)
39.9
13
–
40.1
14
1.00 (m)
57.4
15
1.54 (m)
24.5
16
1.84 (m)
29.5
17
1.12 (m)
56.6
18
0.80 (s)
12.2
19
1.01 (s)
19.8
20
1.36 (m)
40.2
21
0.91 (m)
23.1
22
5.01 (dd, 15.4, 8.8)
138.9
23
5.15 (dd, 15.4, 8.6)
130.0
24
0.95 (m)
51.3
25
1.68 (m)
33.9
26
0.79 (d, 6.4)
21.6
27
0.84 (d, 6.5)
22.7
28
0.99 (m)
25.1
29
1.01(m)
12.3
The compound also showed protons at δH 3.53, 5.15, and 5.35, suggesting the presence of three protons corresponding to that of a tri-substituted and a di-substituted olefinic bond. The proton corresponding to the H-3 of a sterol moiety appeared as a triplet of doublet of doublets at δH 3.52. The above spectral data supported the presence of a sterol skeleton having a hydroxyl group at C-3 position with two double bonds at C-5/C-6 and C-22/C-23 with six methyl groups which was supported by the key COSY (Figure S18) and HMBC (Figure S19) correlations. The 1H and 13C NMR values for all the protons and carbons (Table 4) were assigned on the basis of COSY, HSQC (Figure S20) and HMBC correlations.
Molecular Docking Results
The interactions between ligands of the compounds and specific target protein residues by molecular docking simulation were investigated to validate the antitubercular and antibacterial efficacy of the isolated compounds 1-4 (L1-L4). The binding modes and the orientations of the compounds in the receptor pockets for M. tuberculosis (PID: 4QIJ and 5HKF), E. coli (PID: 5L3J and 8GZY) and S. aureus (PID: 2QIL), respectively were investigated (Table 5). A number of bonding interactions were observed between ligands and the amino acids present at the binding sites of the proteins, such as conventional hydrogen bonds, carbon hydrogen bonds, van der Waals, Pi-Pi stacked, Pi-sigma, Alkyl and Pi-Alkyl bonding (Figures S21-S25). In molecular docking, a negative score signifies a favorable interaction, with more negative values indicating stronger binding. In the present study, affinity values more negative than −7.0 kcal/mol were interpreted as high, between −5.0 and −7.0 kcal/mol as moderate, and above −5.0 kcal/mol as low.
Docking Score Results of M. tuberculosis, E. coli and S. aureus Receptors with Potential Inhibitors.
Ligand
Binding affinity (kcal/mol)
M. tuberculosis
E. coli
S. aureus
4QIJ
5HKF
5L3J
8GZY
2QIL
L1
−6.4
−5.9
−6.7
−6.7
−6.8
L2
−5.3
−4.4
−6.0
−5.2
−5.9
L3
−5.3
−3.6
−4.8
−5.2
−5.5
L4
−8.5
−7.2
−7.8
−8.0
−9.3
Isoniazid
−5.6
−5.0
−5.2
−5.1
−5.4
Ethionamide
−4.8
−4.4
−5.7
−5.5
−5.1
Ligand 4 (L4) demonstrated a strong binding affinity (binding score of −8.5 kcal/mol) in comparison to −5.9, −4.4 and −3.6 kcal/mol of L1 to L3 against 4QIJ, a protein target for M. tuberculosis. Similarly, it had the strongest binding affinity (binding score of −7.2 kcal/mol) against 5HKF, another protein target for M. tuberculosis. The same ligand (L4) demonstrated strong binding affinities (binding scores of −7.8 and −8.0 kcal/mol, respectively) against 5L3J and 8GZY, which are protein targets for E. coli. Interestingly, the same ligand had the best inhibitory activity with the highest binding affinity (binding score of −9.3 kcal/mol) against 2QIL, the protein target for S. aureus.
Pharmacokinetic Studies Results
The pharmacokinetic and physicochemical characteristics of the compounds (lipophilicity, blood-brain barrier (BBB) penetration potential, and interactions with various cytochrome enzymes) were evaluated (Table S1). These characteristics are essential for determining the compounds’ overall drug-likeness, potential adverse effects, and bioavailability, all of which are important factors in drug discovery and development. The bioavailability scores of the four compounds ranged from 0 to 0.56. All the compounds stand out for having zero violations of the Lipinski, Ghose, Veber, Egan and Muegge rules.
Toxicity Prediction Results
In order to predict the potential toxicity of the compounds, a computational toxicology analysis was performed using the ProTox-II web server. This in silico platform integrates various computational methods (such as molecular similarity searches, fragment-based analysis, and machine learning models) to predict a wide range of toxicological endpoints. Each compound was evaluated for acute oral toxicity, hepatotoxicity, respiratory toxicity and cardiotoxicity, with corresponding probabilities provided to indicate the reliability of each prediction. The results showed that compounds 1 and 4 could cause respiratory toxicity, with the former possibly causing cardiotoxicity (Table 6).
In silico Predicted Toxicities of the Compounds Isolated from K. africana.
Ligand
LD50 (mg/kg)
Hepatotoxicity
Probability
Respiratory toxicity
Probability
Cardiotoxicity
Probability
L1
500
Inactive
0.67
Active
0.75
Active
0.63
L2
5000
Inactive
0.71
Inactive
0.97
Inactive
0.70
L3
900
Inactive
0.52
Inactive
0.85
Inactive
0.99
L4
890
Inactive
0.87
Active
0.82
Inactive
0.85
Discussion
Kigelia africana, a member of the trumpet creeper or catalpa family is a heavily exploited medicinal plant for its medicinal values. In the present study, chromatographic fractionation of its dichloromethane/methanol stem bark extract followed by extensive structural elucidation yielded four compounds (1-4). Compound 1 was identified as demethylkigelin, an isocoumarin derivative with a methoxyl group at C-7 and a hydroxyl group at C-6. The 13C NMR resonance at δC 61.0 and the HSQC correlation confirmed the methoxylation at C-7, while the deshielded phenolic proton at δH 11.38 indicated the presence of a hydrogen-bonded hydroxyl group at C-6, which is adjacent to the lactone carbonyl at C-1. This substitution pattern is consistent with previously reported data for 6-demethylkigelin isolated from roots and bark of K. africana.34,35 This dihydroisocoumarin has been previously isolated from fungi such as Aspergillus terreus36 and Phaeospheriopsis sp. ZYX-Z-811.37 The structural features of demethylkigelin (especially the methoxyl substitution at C-7 and the phenolic hydroxyl group at C-6) are central to its biological activity because it influences the molecule's ability to form intramolecular hydrogen bonds and interactions,36 as well as π-π stacking interactions with aromatic rings in molecular targets or hydrogen bonding.38
Extensive comparison of spectral data of compound 2 with available literature led to its conclusive identification as 4-hydroxyphenethyl butyrate, also known as tyrosyl butyrate.39,40 The presence of an AA′XX′ spin system in the aromatic region of the 1H NMR spectrum, with signals at δH 7.08 and 6.76, is characteristic of a para-substituted phenyl ring. This confirmed the presence of 4-hydroxyphenyl moiety in compound 2. The ethylene bridge between the aromatic ring and the ester group was confirmed by a pair of triplets at δH 2.86 and 4.23, corresponding to the benzylic and oxygenated methylene protons, respectively, which were also supported by HSQC correlations at δC 34.4 ppm and 65.1 ppm.
The 13C NMR signal at δC 174.0 ppm indicated the presence of an ester carbonyl, which was further substantiated by the HMBC correlations between the methylene protons of the butyrate chain and the carbonyl carbon. Together, these spectral data with the terminal methyl group signal at δH 0.90 (t, J = 6.9 Hz) confirmed the attachment of a butyrate group, completing the structure assignment. Compound 2 is hereby isolated for the first time from K. africana. Although mostly synthesized,40 phenethyl esters are known to occur naturally in plants. They possess anti-inflammatory, antioxidant and mild antibacterial activities owing to the phenolic hydroxyl group and the lipophilic butyrate side chain.39
Compound 3 was unequivocally identified as stearic acid (octadecanoic acid) based on its spectral data when compared with published literature.41,42 The 1H NMR spectrum displayed characteristic signals of a long saturated aliphatic chain, such as the terminal methyl triplet at δH 0.88 and the broad multiplet between δH 1.17 and 1.40 (corresponding to methylene protons). The presence of methylene protons adjacent to the carboxyl group was confirmed by resonances at δH 2.34 (H-2) and δH 1.63 (H-3), which exhibited long-range HMBC correlations with the carbonyl carbon at δC 180.5. Stearic acid is a saturated fatty acid that has been isolated from plants such as Albizia amara (stem bark) and Firmiana colorata (leaves).41,42 It is known to elicit antibacterial, anti-inflammatory, and emollient properties among other bioactivities.43
Compound 4 was identified as stigmasterol after detailed spectroscopic analysis and comparison with literature data.44,45 The presence of two double bonds at positions C-5/C-6 and C-22/C-23, and a hydroxyl group at C-3, confirmed the typical structural features of stigmasterol. The 1H and 13C NMR data (ie, the presence of methyl, methine and olefinic protons, and carbon signals) were in complete agreement with previous reports.44,45 Stigmasterol is widely distributed in medicinal plants and is known for a broad range of pharmacological properties such as anti-inflammatory, antitumor, antimutagenic, neuroprotective and immunomodulatory effects.46
In molecular docking studies, the binding interaction achieved for ligand-protein targets provide good insights into their corresponding shapes and electrostatic relationships.47 In other words, docking score indicates the total energy realized during the interaction, with lower negative values signaling stronger interactions (see Table S1). The binding results of compounds 1-4 in the present study were compared with isoniazid and ethionamide as standard antitubercular drugs (known inhibitors),48,49 with the approximate average binding affinities of −5.2 and −5.1 kcal/mol, respectively. It was interesting to note that the compounds showed better binding energies than both drugs. The binding affinity of the complexes of the ligands (L1-L4) ranged from −4.4 to −9.3 kcal/mol, which confirmed their antitubercular potency. Wadhwani et al17 recently isolated nine compounds from K. africana heartwood, of which tecomaquinone-I exhibited the most potent antitubercular activity with a binding energy of −12.7 kcal/mol. Compared to isoniazid and a co-crystallized PDB ligand used as reference compounds (binding energies −10.6 and −5.7 kcal/mol), the nine compounds had binding energies varying from −12.7 to −5.7 kcal/mol. The binding energy reported for tecomaquinone-I is higher than for stigmasterol (compound 4), which was the most potent isolated compound in the present study. However, stigmasterol had a stronger bonding affinity than isoniazid, similar to the results of Wadhwani et al17 In another study, Wang et al50 reported molecular docking results which are comparable to our findings on anti-M. tuberculosis activities of the four isolated compounds in the present study. Therefore, these molecular docking results are indicative of the potential of the identified compounds (especially compounds 1 and 4) as therapeutic candidates for M. tuberculosis treatment. The observed binding interactions and high binding scores suggest that these compounds may play a significant role in the development of novel strategies for M. tuberculosis treatment.
In the pharmacokinetic studies, all the compounds stood out for having zero violations of the Lipinski, Ghose, Veber, Egan and Muegge rules. This indicates that the compounds are compliant with established drug-likeness criteria.51 Furthermore, the drug's permeability and oral bioavailability depend on the values of topological polar surface area (TPSA) and molecular lipophilicity potential (log P). In this context, log P indicates the compound's distribution coefficient between n-octanol and water, and it plays a vital role in the cellular membrane's interactions with other proteins.52,53 All the compounds exhibited log P values above five, except compound 1 which has an affinity for hydrophilic environments.
During drug metabolism, there are essential enzymes under Cytochrome P450 (CYP) family which are involved in the phase 1 metabolism of xenobiotics. It is known that some phytochemicals isolated from different sources have the ability to inhibit the CYP isoforms, leading to drug-drug interactions.54 In the present study, it was predicted that compounds 1 and 2 would not inhibit the CYP1A2 isoform, compound 2 would not inhibit the CYP2C19 isoform while compounds 2 and 3 would not inhibit the CYP2C9 isoform. Similarly, compound 2 was predicted not to inhibit the CYP2D6 and CYP3A4 isoforms while compounds 3 and 4 were predicted not to inhibit the CYP2B6 isoform. Thus, the isolated compounds are unlikely to affect drug-drug interactions that could, otherwise, result in loss of efficacy.
Computational toxicity predictions showed that L1 (demethylkigelin) exhibited a moderate acute oral toxicity with a predicted LD50 of 500 mg/kg (toxicity class 4 according to GHS criteria). This compound was also predicted to be active for both respiratory and cardiotoxicity, although with lower probabilities for the latter, indicating potential adverse effects on these systems. In contrast, L2 (tyrosyl butyrate) was the safest compound, with a high predicted LD50 of 5000 mg/kg, classifying it as non-toxic (class 5). The high probability scores for its predicted inactivity across all the tested endpoints further support its low toxicity profile. On the other hand, L3 (stearic acid) had a relatively low acute oral toxicity (LD50 of 900 mg/kg) and was predicted to be inactive for all the other tested endpoints, with a very high confidence score for its inactivity in cardiotoxicity. Stigmasterol (L4) had a similar LD50 to L3 (stearic acid), but was predicted to be active for respiratory toxicity with a high probability. Overall, the results suggested that while L2 and L3 showed promising low toxicity profiles, L1 and L4 warrant further investigation, especially for their potential respiratory and cardiotoxic effects.
Despite these interesting results, the present study has some limitations. The study was largely in silico, relying on molecular docking and ADMET predictions. These methods provide useful preliminary insights into drug-target interactions, pharmacokinetic and toxicity properties but they do not substitute for experimental validation. Further, we did not extend the analysis to KEGG pathway or STRING network enrichment, as these require a broad and validated target set to generate meaningful mechanistic insights. Future studies integrating experimental activity data with target prediction will allow more reliable pathway and network analyses. The predicted binding affinities, ADMET and toxicity properties may differ under physiological conditions, and therefore, in vitro and in vivo studies are required to confirm the antimycobacterial activity, safety, and pharmacological potential of the compounds identified from K. africana stem bark.
Conclusion
This study found that K. africana stem bark possesses bioactive compounds that are potential inhibitors of M. tuberculosis with good to better binding affinities and stable interactions. We recommend that future studies should perform in vitro and in vivo validation of the bioactivity and toxicity of the characterized compounds.
Supplemental Material
sj-docx-1-npx-10.1177_1934578X251388844 - Supplemental material for In silico Antimycobacterial Evaluation of Compounds Isolated from Kigelia africana Stem Bark
Supplemental material, sj-docx-1-npx-10.1177_1934578X251388844 for In silico Antimycobacterial Evaluation of Compounds Isolated from Kigelia africana Stem Bark by Ivan Gumula, Mary Achiro, Sarah Kiwanuka Nanyonga, Denis Akampurira, Patrick Onen, Ronnie Tumwesigye and Timothy Omara in Natural Product Communications
Footnotes
Acknowledgments
The authors are grateful to the Institute of Chemistry, University of Potsdam, Germany for NMR analysis of the isolated compounds.
ORCID iD
Ivan Gumula
Ethical Approval
Ethical approval is not applicable for this article.
Statement of Informed Consent
There are no human subjects in this article and informed consent is not applicable.
Statement of Human and Animal Rights
Not applicable for this article.
Author Contributions
Conceptualization: Ivan Gumula, Mary Achiro, Patrick Onen; Methodology: Ivan Gumula, Sarah Kiwanuka Nanyonga, Patrick Onen, Ronnie Tumwesigye, Timothy Omara; Investigation: Mary Achiro, Patrick Onen, Ronnie Tumwesigye; Formal analysis: Mary Achiro, Patrick Onen, Ronnie Tumwesigye; Writing—original draft: Ivan Gumula, Timothy Omara; Writing—review and editing: Ivan Gumula, Timothy Omara. All authors read and approved the final manuscript.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by Kyambogo University through the Competitive Research Grant awarded to the first author.
Declaration of Conflicting Interest
The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: The authors declare no financial or personal conflicts of interest. Timothy Omara discloses his role as a member of the Editorial Review Board of Natural Product Communications. He had no involvement in the editorial decision making for this article.
Data Availability
Data supporting the conclusions of this study are available within this article and its .
Supplemental Material
Supplemental material for this article is available online.
References
1.
GBD 2021 Tuberculosis Collaborators. Global, regional, and national age-specific progress towards the 2020 milestones of the WHO End TB strategy: a systematic analysis for the global burden of disease study 2021. Lancet Infect Dis. 2024;24(7):698–725.
GulumbeBHAbdulrahimAAhmadSKLawanKADanlamiMB. WHO report signals tuberculosis resurgence: addressing systemic failures and revamping control strategies. Decod Infect Transm. 2025;3(14):100044.
4.
BartekaGBwayoDMatovuJKB, et al.Treatment outcomes and predictors of success for multidrug resistant tuberculosis MDR TB in Ugandan regional referral hospitals. Sci Rep. 2025;15(1):14144.
5.
MeadowsIMvungiHSalimK, et al.N-acetylcysteine to reduce kidney and liver injury associated with drug-resistant tuberculosis treatment. Pharmaceutics. 2025;17(4):516.
6.
NavasardyanIMiwalianRPetrosyanAYeganyanSVenketaramanV. HIV-TB coinfection: current therapeutic approaches and drug interactions. Viruses. 2024;16(3):321.
7.
SachanRSKMistryVDholariaM, et al.Overcoming Mycobacterium tuberculosis drug resistance: novel medications and repositioning strategies. ACS Omega. 2023;8(36):32244–32257.
8.
ChaachouayNZidaneL. Plant-derived natural products: a source for drug discovery and development. Drugs Drug Candidates. 2024;3(1):184–207.
9.
AliRKFMOdjoubéréJSagboEM. Ethnobotanical study of the false Baobab (Kigelia africana Lam. Benth) in the Issaba district (Pobè Commune). Open J For. 2025;15(2):181–198.
10.
KandeBGrahZMMoyabiAGASoroYKoneWM. An ethnobotanical survey of traditional medicinal plants used against tuberculosis and symptoms associated in Abidjan, Côte d'Ivoire. Asian J Ethnobiol. 2023;6(1):36–45.
11.
ObakiroSBKipropAKowinoI, et al.Ethnobotany, ethnopharmacology, and phytochemistry of traditional medicinal plants used in the management of symptoms of tuberculosis in East Africa: a systematic review. Trop Med Health. 2020;48(68):1-21.
12.
BelloIShehuMWMusaMZaini AsmawiMMahmudR. Kigelia africana (Lam.) Benth. (Sausage tree): phytochemistry and pharmacological review of a quintessential African traditional medicinal plant. J Ethnopharmacol. 2016;189(1):253–276.
13.
PicernoPAutoreGMarzoccoSMeloniMSanogoRAquinoRP. Anti-inflammatory activity of verminoside from Kigelia africana and evaluation of cutaneous irritation in cell cultures and reconstituted human epidermis. J Nat Prod. 2005;68(11):1610–1614.
14.
SidjuiLSMelongRMahiou-LeddetV, et al.Triterpenes and lignans from Kigelia africana. J Appl Pharmaceut Sci. 2015;5(suppl 2):001–006.
15.
NabatanziANkadimengSMLallNKabasaJDMcGawLJ. Ethnobotany, phytochemistry and pharmacological activity of Kigelia africana (Lam.) Benth. (Bignoniaceae). Plants (Basel). 2020;9(6):753.
16.
BunalemaLObakiroSTabutiJRWaakoP. Knowledge on plants used traditionally in the treatment of tuberculosis in Uganda. J Ethnopharmacol. 2014;151(2):999–1004.
17.
WadhwaniBDNuniaVJoshiK, et al.Identification of potent antitubercular secondary metabolites from Kigelia africana: an in-silico investigation. ChemistrySelect. 2023;8(37):e202302269.
18.
GumulaIKyarimpaCNanyongaSK, et al.Antibacterial properties of phytochemicals isolated from leaves of Alstonia boonei and aerial parts of Ipomoea cairica. Nat Prod Commun. 2024;19(9):1–9.
19.
DoniniSFerrarisDMMiggianoRMassarottiARizziM. Structural investigations on orotate phosphoribosyltransferase from Mycobacterium tuberculosis, a key enzyme of the de novo pyrimidine biosynthesis. Sci Rep. 2017;7(1):1180.
20.
RoyAAnandAGargS, et al.Structure-based in silico investigation of agonists for proteins involved in breast cancer. Evid Based Complement Alternat Med. 2022;2022(1):7278731.
21.
ZouWHassanIAkramB, et al.Validating interactions of pathogenic proteins of Staphylococcus aureus and E. coli with phytochemicals of Ziziphus jujube and Acacia nilotica. Microorganisms. 2023;11(10):2450.
22.
IqbalSOmaraTKahwaIKhanUM. Protective effects of Sphaeranthus indicus floral extract against BPS-induced testicular damage in rats occurs through downregulation of RIPK1/3-MLK-driven necroptosis and fas-FasL-mediated apoptosis. Adv Tradit Med. 2025;25(1):579–595.
23.
AlinomugashaE. In vitro Antibacterial Efficacy of Kigelia africana Extraction Pathogenic Escherichia coli Isolated from Poultry in Soroti City. BSc Thesis. Busitema University.
24.
HussainTFatimaIRafayMShabirSAkramMBanoS. Evaluation of antibacterial and antioxidant activity of leaves, fruit and bark of Kigelia africana. Pak J Bot. 2016;48(1):277–283.
25.
KimSChenJChengT, et al.Pubchem 2023 update. Nucleic Acids Res. 2023;51(D1):D1373–D1380.
26.
IbrahimMDetrojaABhimaniA, et al.In silico discovery of potential novel anti-tuberculosis drug candidates from phytoconstituents of Chlorophytum borivilianum and Asparagus racemosus. Heliyon. 2025;11(4):e42859.
27.
DongJWangNNYaoZJ, et al.ADMETlab: a platform for systematic ADMET evaluation based on a comprehensively collected ADMET database. J Cheminform. 2018;10(1):29.
28.
LipinskiCALombardoFDominyBWFeeneyPJ. Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings. Adv Drug Deliv Rev. 1997;23(1-3):3–25.
29.
VeberDFJohnsonSRChengHYSmithBRWardKWKoppleKD. Molecular properties that influence the oral bioavailability of drug candidates. J Med Chem. 2002;45(12):2615–2623.
30.
MorrisGMHueyRLindstromW, et al.Autodock4 and AutoDockTools4: automated docking with selective receptor flexibility. J Comput Chem. 2009;30(16):2785–2791.
31.
TrottOOlsonAJ. Autodock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. J Comput Chem. 2010;31(2):455–461.
32.
PolákLŠkodaPRiedlováKKrivákRNovotnýMHokszaD. Prankweb 4: a modular web server for protein-ligand binding site prediction and downstream analysis. Nucleic Acids Res. 2025;53(W1):W466–W471.
33.
BanerjeePEckertAOSchreyAKPreissnerR. ProTox-II: a webserver for the prediction of toxicity of chemicals. Nucleic Acids Res. 2018;46(W1):W257–WW26.
34.
GovindachariTRPatankarSJViswanathanN. Isolation and structure of two new dihydroisocoumarins from Kigelia pinnata. Phytochemistry. 1971;10(7):1603–1606.
35.
HigginsCABellTDelbederiZ, et al.Growth inhibitory activity of extracted material and isolated compounds from the fruits of Kigelia pinnata. Planta Med. 2010;76(16):1840–1846.
36.
ShimadaAInokuchiTKusanoM, et al.4-Hydroxykigelin And 6-demethylkigelin, root growth promoters, produced by Aspergillus terreus. Z Naturforsch C J Biosci. 2004;59(3-4):218–222.
37.
QuanMMaQYangL, et al.Study on the secondary metabolites of marine fungus Phaeospheriopsis sp. ZYX-Z-811. J Trop Oceanogr. 2024;43(2):121–127.
38.
GiordanettoFTyrchanCUlanderJ. Intramolecular hydrogen bond expectations in medicinal chemistry. ACS Med Chem Lett. 2017;8(2):139–142.
39.
MateosRTrujilloMPereira-CaroGMadronaACertAEsparteroJL. New lipophilic tyrosyl esters. Comparative antioxidant evaluation with hydroxytyrosyl esters. J Agric Food Chem. 2008;56(22):10960–6.
40.
BerniniRCarastroISantoniFClementeM. Synthesis of lipophilic esters of tyrosol, homovanillyl alcohol and hydroxytyrosol. Antioxidants (Basel). 2019;8(6):174.
41.
AbdurrahmanICai-XiabY. Isolation and characterization of fatty acid derivatives from the stem barks of Albizia amara (Fabaceae), Sudanese medicinal plant. Chem Methodol. 2020;4(4):369–377.
42.
HajiraBBGeraldineMBettadaiahBKNareshKS. Isolation and identification of bioactive compounds from leaf and flower extracts of Firmiana colorata. RJAHS. 2024;4(2):23–29.
43.
IvanovaEPNguyenSHGuoY, et al.Bactericidal activity of self-assembled palmitic and stearic fatty acid crystals on highly ordered pyrolytic graphite. Acta Biomater. 2017;59(1):148–157.
44.
ChaturvedulaVSPrakashI. Isolation of Stigmasterol and β-Sitosterol from the dichloromethane extract of Rubus suavissimus. Int Curr Pharmaceut J. 2012;1(9):239–242.
45.
DasSCHossainMSTabassumNRahmanMSBacharSCIslamMS. Phytochemical and preliminary pharmacological evaluation of Diospyros blancoi fruits. Dhaka Univ J Pharm Sci. 2024;23(2):113–122.
46.
BakrimSBenkhairaNBouraisI, et al.Health benefits and pharmacological properties of stigmasterol. Antioxidants (Basel). 2022;11(10):1912.
47.
KannanRAlharbiNSKadaikunnanSRajaramSKAlexanderRA. In silico analysis of phytoconstituents from Allium sativum as potential inhibitors of Inha in Mycobacterium tuberculosis. Braz Arch Biol Technol. 2016;59(1):e16160109.
48.
RamanKRajagopalanPChandraN. Flux balance analysis of mycolic acid pathway: targets for anti-tubercular drugs. PLoS Comput Biol. 2005;1(5):e46.
49.
ChungBKDickTLeeDY. In silico analyses for the discovery of tuberculosis drug targets. J Antimicrob Chemother. 2013;68(12):2701–2709.
50.
WangALvKLiL, et al.Design, synthesis and biological activity of N-(2-phenoxy)ethyl imidazo[1,2-a]pyridine-3-carboxamides as new antitubercular agents. Eur J Med Chem. 2019;178(1):715–725.
51.
MueggeI. Selection criteria for drug-like compounds. Med Res Rev. 2003;23(3):302–321.
52.
PiccirilloEAmaralATD. Virtual screening of bioactive compounds: concepts and aplications. Química Nova. 2018;41(6):662–677.
53.
GuoYFanYLiGWangZShiXShenM. “Cluster bomb” based on redox-responsive carbon dot nanoclusters coated with cell membranes for enhanced tumor theranostics. ACS Appl Mater Interfaces. 2021;13(47):55815–55826.
54.
EstevesFRueffJKranendonkM. The central role of cytochrome P450 in xenobiotic metabolism: a brief review on a fascinating enzyme family. J Xenobiot. 2021;11(3):94–114.
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
