Antibacterial and Antifungal Alkaloids from Asian Angiosperms: Distribution, Mechanisms of Action, Structure-Activity, and Clinical Potentials

The emergence of multidrug-resistant bacteria and fungi requires the development of antibiotics and antifungal agents. This review identified natural products isolated from Asian angiosperms with antibacterial and/or antifungal activities and analyzed their distribution, molecular weights, solubility, and modes of action. All data in this review were compiled from Google Scholar, PubMed, Science Direct, Web of Science, ChemSpider, PubChem, and a library search from 1979 to 2022. One hundred and forty-one antibacterial and/or antifungal alkaloids were identified during this period, mainly from basal angiosperms. The most active alkaloids are mainly planar, amphiphilic, with a molecular mass between 200 and 400 g/mol, and a polar surface area of about 50 Å2, and target DNA and/or topoisomerase as well as the cytoplasmic membrane. 8-Acetylnorchelerythrine, cryptolepine, 8-hydroxydihydrochelerythrine, 6-methoxydihydrosanguinarine, 2′-nortiliacorinine, pendulamine A and B, rhetsisine, sampangine, tiliacorine, tryptanthrin, tylophorinine, vallesamine, and viroallosecurinine yielded MIC ≤ 1 µg/mL and are candidates for the development of lead molecules.


Introduction
The resistance of bacteria and fungi to antimicrobial agents necessitates the continuous development of antibiotics and antifungal agents with original chemical frameworks that may come from flowering plants. The flowering plants also termed angiosperms comprise 11 major taxa or clades grouped into three groups: (i) basal angiosperms including Protomagnoliids, Magnoliids, Monocots, Eudicots; (ii) core angiosperms including Core Eudicots, Rosids, Fabids, Malvids; and (iii) upper angiosperms including Asterids, Lamiids, and Campanulids [1]. Within each clade, plants yield both non-specific and specific secondary metabolites such as alkaloids to control the growth of phytopathogenic bacteria and fungi. These antimicrobial principles fall into two main groups: phytoanticipins and phytoalexin. Phytoanticipins are antimicrobial compounds in plants that are present before phytopathogenic microorganism challenge or inactive immediate precursors stored in healthy tissues that are converted into antimicrobial metabolites known as phytoalexins [2]. Phytoanticipins and phytoalexins containing primary, secondary, tertiary, or quaternary amines are called alkaloids, which belong to various chemical classes, principally amides, indoles, piperidines, quinolines, isoquinolines, pyrrolidines, imidazoles, diterpenes, sesquiterpeness, and steroidal alkaloids [3].
Phytopathogenic Gram-negative bacteria are more resistant to alkaloids than Grampositive bacteria due, at least in part, to an outer hydrophilic and negatively charged layer of lipopolysaccharides [4]. In Gram-negative bacteria, porins allow for the entry of water and nutrients through the outer layer and hydrophilic and amphiphilic xenobiotics with a molecular mass below 600 g/mol xenobiotics without nutritional or physiological benefit or toxins are actively cleared from the cytoplasm by efflux pumps [5]. The assessment of the antibacterial and antifungal strength of isolated secondary metabolites in vitro is qualitatively appreciated by the measurement of the diameter of an inhibition zone and quantitively based on the minimum inhibiting concentration (MIC), and several thresholds of activity have been proposed [6][7][8][9]. Angiosperms produce alkaloids that inhibit the growth of both Gram-positive, Gram-negative bacteria, yeasts, and filamentous fungi and these principles are of potential therapeutic value. Among the factors that influence the antibacterial or antibacterial strength, and the targets of alkaloids are the molecular mass and water solubility [10].
Over the last 80 years, enormous research efforts have been devoted to the aim of identifying antibiotic or antifungal lead molecules in flowering plants globally, but to date, none of these have been developed as drugs. The present work attempts to provide a comprehensive review of the main findings regarding the antibacterial and antifungal alkaloids from Asian angiosperms. All data in this review were compiled from Google Scholar, PubMed, Science Direct, Web of Science, ChemSpider, PubChem, and a library search from 1979 to 2022 and were analyzed to address the following points: (i) distribution; (ii) strongest principles identified; (iii) spectrum of activity; (iv) influence of molecular mass, water solubility, and polar surface; (v) the mode of action; (vi) structure-activity; and (vii) efflux pump inhibition.

Distribution
Regarding the distribution of antibacterial and antifungal alkaloids from Asian angiosperms (Figure 1), the following could be observed:

•
All clades yielded antibacterial and/or antifungal alkaloids, except for the Rosids. • Most antibacterial and/or antifungal alkaloids can be found in basal angiosperms, which are isoquinolines. • Some clades yielded a specific class of antibacterial and/or antifungal alkaloids such as the Amaryllidaceae alkaloids in the Monocots, phenanthrene alkaloid in the Magnoliids, Securinega alkaloids in Fabids, carbazoles in the Malvids, and monoterpene indole alkaloids in the Lamiids. • Core angiosperms and upper angiosperms use various classes of alkaloids and phytoalexins. • Most antibacterial and/or antifungal alkaloids have been isolated from medicinal plants ( Table 1, Table S1). • Tripathi et al. (1994) observed changes in the antibacterial alkaloid concentrations in plants over time [11].
• Core angiosperms and upper angiosperms use various classes of alkaloids and phytoalexins. • Most antibacterial and/or antifungal alkaloids have been isolated from medicinal plants ( Table 1, Table S1). • Tripathi et al. (1994) observed changes in the antibacterial alkaloid concentrations in plants over time [11].    • Most antibacterial and/or antifungal alkaloids have been isolated fro plants (Table 1, Table S1). • Tripathi et al. (1994) observed changes in the antibacterial alkaloid conc plants over time [11].  Table 1. The chemical structures of antibacterial and/or antifungal alkaloids and thei origin.

The Strongest Antibacterial and/or Antifungal Alkaloids Identified and Spectrum of Activity
Rios and Recio defined a crude extract with MIC greater than 1000 µg/mL as inactive and suggested interesting antibacterial activity for MICs of 100 µg/mL or lower [6]. Previously, Fabry et al. (1998) defined crude active extracts as having MIC values below 8000 µg/mL [7], while more recently, Kuete (2010) defined crude extracts with MIC values less than 100 µg/mL as active and MICs above 625 µg/mL as weakly active [9]. Here, a compound was very strongly antibacterial or antifungal for a MIC value below or equal to 1 µg/mL; strongly antibacterial (or antifungal for a MIC value above 1 and below or equal to 50 µg/mL; moderately antibacterial or antifungal for a MIC from 50 and below 100 µg/mL; weakly antibacterial (or antifungal) for a MIC from 100 and below 500 µg/mL; very weakly antibacterial or antifungal for a MIC ranging from 500 to below 2500 µg/mL; and inactive for a MIC value above 2500 µg/mL. Following this classification, 8-acetylnorchelerythrine, cryptolepine, 8-hydroxydihydrochelerythrine, 6-methoxydihydrosanguinarine, 2 -nortiliacorinine, pendulamine A and B, rhetsisine, sampangine, tiliacorine, tryptanthrin, tylophorinine, vallesamine, and viroallosecurinine showed very strong activities (Table S2).
Looking at the spectrum of activity, most alkaloids showed activity against Grampositive bacteria, followed by Gram-negative bacteria, yeasts, filamentous fungi, and mycobacteria (Table S2).

Influence of Molecular Mass
The molecular mass of natural products dictates their ability to fit in the catalytic pockets of enzymes and to cross biological membranes. Here, low molecular mass molecules were defined with a molecular mass below 200 g/mol; medium molecular mass molecules were defined with a molecular mass from 200 to 400 g/mol; and high molecular mass molecules were defined for a molecular mass above 400 g/mol. Following this classification, we noted that principles with very strong activity against bacteria and fungi had a molecular mass mainly from 200 to 400 g/mol whereas very strong repressors of mycobacteria mainly had a mass above 400 g/mol (Table S2).

Influence of Solubility and Polar Surface
Because solubility is a fundamental criterion in which to consider the efficacy of alkaloids, there is need to use mathematical values. Log P is equal to the ratio of concentrations of a compound between octanol and water. Hydrophilic compounds (hydrophilic) have low or negative values (about −3) (compounds are mainly found in the water phase). Mid-hydrophilic compounds have a Log P close to 0 (the compound is equally partitioned between the octanol and water layers). Non-hydrophilic (hydrophobic, liposoluble) compounds have a high Log P (up to about 7) (note that lipophilic alkaloids tend to remain in and destabilize the cytoplasmic membrane of bacteria and fungi). However, Log P is only relevant for non-ionizable principles and for an ionized substance, a Log D is preferable (in terms of ADME), but the pH must be fixed. In some databases, one can find a Log P of 3 for the ionic alkaloid berberine, suggesting a lipophilic substance, which is not sensible. Since compounds destined for pharmaceutical development will mainly be exposed to physiological pH and often a weak base or a weak acid, we defined, at pH 7.4, lipophilic compounds for a negative Log D value of 4, amphiphilic (mid-polar) compounds for a Log D up to about 4.5, and lipophilic for a Log D above 4.5. Note that the Log D values given here are predicted values. Following this classification, we noted that principles with very strong activity against bacteria and fungi were mainly amphiphilic, whereas very strong antimycobacterial agents were mainly lipophilic (Table S2). Most antibiotics with very strong antibacterial and/or antifungal effects have a polar surface area around 50 Å (Table S2).

Mechanisms of Action and Structure-Activity Relationships
Most antibacterial and/or antifungal alkaloids from Asian angiosperms either bear a quinoline or an indole framework and the main mechanisms of action involve the targeting of the DNA, topoisomerases, and cytoplasmic membrane.
The mode of antibacterial and antifungal action of quinoline alkaloids mainly evokes interaction with DNA. Dictamnine binds to DNA under UV light [21]. Camptothecin stabilizes the topoisomerase I-DNA complex [22]. Liriodenine blocks topoisomerase II [23]. Berberine was active against Actinobacillus pleuropneumoniae and Streptococcus agalactiae (CVCC 1886) via DNA synthesis inhibition and the blockage of synthesis [24,25]. Anonaine induces DNA damage [26] as well as magnoflorine [27]. Aporphine alkaloids are planar and intercalate DNA and inhibit topoisomerase [28] as well as Amaryllidaceae alkaloids [29]. The quaternary ammonium ion of protoberberine alkaloids and their heterocyclic planar framework account for topoisomerase I inhibition [30].
During bacterial division, topoisomerase IV catalyzes the relaxation of the DNA chain and 6,6 -dihydroxythiobinupharidine blocks this enzyme in S. aureus [31]. Phenanthroindolizidine alkaloids interact with DNA [32]. In Gram-positive bacteria, topoisomerase IV is a target for bactericidal quinolone antibiotics [33]. Asian angiosperms produce a vast array of quinoline, isoquinoline, piperidine, and quinolizidine alkaloids, representing a fascinating reservoir of topoisomerase IV inhibitors.
In summary, heterocyclic alkaloids with low to medium molecular mass, close to planar or planar, with the presence of a few hydroxy or ketone groups, almost always target DNA and/or RNA in bacteria and fungi. From this perspective, quinazolone alkaloids in the family Hydrangeaceae (order Saxifragales; clade Core Eudicots) could be examined for their antibacterial and or antifungal properties.

Alkaloids Targeting the Cytoplasmic Membrane
Phenanthroindolizidine alkaloids disturb the cytoplasmic membrane integrity [34]. The long alkyl chain of piperine and piperlongumine could penetrate the membrane of bacteria and fungi. In Candida albicans, piperine affects the membrane integrity, leading to oxidative stress followed by cell cycle arrest and apoptosis [35]. Marques et al. (2010) presented evidence that amides were more active against Cladosporium cladosporoides when the non-substituted aromatic ring, single double bonds, and substitution of nitrogen with alkyl groups were present [36]. In fungi, berberine targets the mitochondrial membrane [24]. Liriodenine in Paracoccidioides brasiliensis evoked cytoplasmic alterations and damage to the cell wall [37] while berberine evoked cytoplasmic insults in Streptococcus agalactiae (CVCC 1886) [34] and targeted the mitochondrial membrane of fungi [25].

Miscellaneous Targets
Aristolochic acids inhibited the H + -ATPase-mediated proton pump in E. coli [12]. Securinine induces mitotic block in cancer cells by binding to tubulin and inhibits microtubule assembly [38], therefore, microtubules could be involved in the antibacterial and/or antifungal properties of cytotoxic monoterpenoid indole alkaloids.

Efflux Pumps Inhibitors
P. aeruginosa and Gram-negative bacteria can resist a broad spectrum of natural products because they have extra classes of efflux pumps such as ABC (ATP binding cassette), RND (resistance nodulation cell-division), MF (major facilitator), SMR (small multidrug resistance) and MATE (multidrug and toxic compound extrusion) pumps [39]. ABC efflux pumps located in the cytoplasmic membrane of both Gram-positive and Gram-negative bacteria that use the energy derived from ATP hydrolysis to expel xenobiotics. RND efflux pumps located in the cytoplasmic and outer membrane of the Gram-negative bacteria (specific to Gram-negative bacteria) that expel xenobiotics using the H + gradient (antiporters). MF efflux pumps located in the cytoplasmic membrane of both Gram-positive and Gramnegative bacteria that expel xenobiotics using the H + gradient (antiporters). SMR efflux pumps located in the cytoplasmic membrane of both Gram-positive and Gram-negative bacteria that expel xenobiotics using the H + gradient (antiporters). MATE efflux pumps located in the cytoplasmic membrane and are antiporters (the exit of the xenobiotic coincides with the entry of a Na + ).
Tryptanthrin inhibits efflux P-glycoprotein in Caco-2 cells [16] and as such may inhibit bacteria and or fungal efflux pumps. Apocynaceous monoterpene indole alkaloids are often vasorelaxant [40], and therefore, with some inhibition levels of bacterial and or fungal efflux-pumps. For instance, the reserpine from Rauvolfia serpentina (L.) Benth. ex Kurz decreased the resistance of S. aureus (1199B, NorA hyperproducer) to ciprofloxacin and norfloxacin [10]. Reserpine is a calcium channel antagonist and an inhibitor of efflux pumps in Gram-negative bacteria and mycobateria [41]. From Rauvolfia serpentina (L.) Benth. ex Kurz, ajmaline and yohimbine are neuroactive and efflux pump inhibitors in Gram-negative bacteria [42]. Verapamil is another example of a calcium channel antagonist that inhibits the efflux pump in bacteria [42]. The reason why calcium channel antagonists have the tendency to inhibit the bacterial efflux pump could be because of the correlations between the bacterial efflux pumps and bacterial calcium transport [43]. Tetrandrine, which is a calcium channel antagonist in mammalian cells, inhibited the efflux pumps in S. aureus Rv2459 (jefA), Rv3728, and Rv3065 (mmr) efflux pumps in Mycobacterium species [44]. Therefore, natural products known for being calcium channel inhibitors should be screened as antibiotic potentiators. 4 -O-Methyldopamine inhibits NorA [45], showing that N-caffeoylphenalkylamide with the strongest efflux pump inhibitor activities presented hydroxyl substitution on the aromatic rings of the caffeic acid part and methoxy substitution on the aromatic ring of the dopamine moiety, which led to an increase in activity. Dopamine is a neurotransmitter, and it could be argued that neuroactive principles are a first line candidate for the development of efflux pump inhibitors. L-dopa increased the resistance of C. neoformans toward amphotericin B [46]. In line, erythrinan-type alkaloids and amide alkaloids in the family Piperaceae interact with GABAergic receptors and as such may be able to inhibit bacterial efflux pumps as in pellitorine, which at 16 µg/mL increased the sensitivity of S. aureus (RN4220) to erythromycin at 16 µg/mL via inhibition of the efflux pumps [5]. Canthin-6-one has a chemical structure with some similarity with serotonin and thus might be able to inhibit bacteria and/or fungal efflux pumps. In mammalian cells, tryptanthrin inhibits the expression of P-glycoprotein efflux pumps [46] and one could investigate its effect on the expression of efflux pumps in bacteria and fungi.  [47]. Pellitorine (50 µg/disk) inhibited the growth of Listeria monocytogenes with an inhibition zone diameter of 9 mm and the MIC value of 500 µg/mL [48]. Pellitorine (20 µL of a 2 µg/mL solution/6 mm well) inhibited the growth of Aspergillus flavus, Aspergillus fumigatus, Coniophora puteana, Fibrophoria vaillentii, Fusarium proliferatum, and Rhisopus sp. with the inhibition zone diameters of 27,29,26,28,29, and 27 mm, respectively [49]. Piperlonguminine suppressed B. sphaericus (ATCC 14577), B. subtilis (ATCC 6051), S. aureus (ATCC 9144), E. coli (ATCC 25922), P. syringae (ATCC 13457), and S. typhimurium (ATCC 23564) with the MIC values of 20, 9, 12.5, 150, 75, and 175 µg/mL, respectively [47]. Piperlonguminine inhibited Mycobacterium tuberculosis with a MIC value of 50 µg/mL [50]. 8Z-N-isobutyleicosatrienamide and pellitorine had moderate potencies with S. aureus (MIC: 34 µM) [51]. Pellitorine restrained M. tuberculosis (H 37 Ra) with the MIC value of 25 µg/mL [52]. di-p-Coumaroyl-caffeoylspermidine weakly inhibited the mycelial growth of Pyrenophora avenae and Blumeria graminis [53]. In the clade Clampanulids, Spilanthes paniculata Wall. ex DC yielded N-Isobutyl-2 (E), 6 (Z), 8 (E)-decatrienamide (also known as spilanthol), which was bactericidal for Streptococcus mutans with MIC/MBC values of 125/125 µg/mL and weakly repressed C. albicans (ATCC 10231) [44]. The condensation of ferulic acid and dopamine yielded N-trans-feruloyl-4-methyldopamine 200 µg/disk that developed halos with a broad-spectrum of bacteria [54] and increased the susceptibility of multidrug-resistant S. aureus (overexpressing the multidrug efflux transporter NorA) to norfloxacin at 100 µg/mL [44].

Monoterpene Indole Alkaloids
Plants in the order Gentianales produce monoterpene indole alkaloids with moderate broad-spectrum antibacterial properties (Figure 1).

Simple Quinolines
Members of the genus Gomphandra Wall. ex Lindl. are produced from the condensation of tryptamine and secologanin yields via strictodamide camptothecin, which is a strong broad-spectrum antifungal pyrrolquinoline alkaloid [102,103].

Bisbenzylisoquinolines
The radical coupling of benzylisoquinolines gives birth to antibacterial and antifungal bisbenzylisoquinoline alkaloids in plants in the clade Eudicots (Figure 1). For instance, the coupling of N-methyl coclaurine yields tetrandine, which is weakly bactericidal for S. aureus (ATCC 25923) and MRSA (ATCC 33591) [107] and is a bacterial efflux pump inhibitor [90,107].

Hasubanans
Within the family Menispermaceae, intramolecular coupling of benzylisoquinolines form hasubanan alkaloids. One such alkaloid is glabradine, isolated from the tubers of Stephania glabra (Roxb.) Miers, which suppressed S. aureus and S. mutans with the MIC value of 50 µg/mL as well as M. gypseum, M. canis, and T. rubrum with the MIC values of 25, 25, and 50 µg/mL, respectively [154].
The imidazole alkaloid allantoin very strongly hindered B. subtilis with the MIC value of 4 µg/mL as well as S. aureus, E. coli, and K. pneumoniae with the MIC value of 8, 8, and 8 µg/mL, respectively [172] (Table 1).

Steroidal Alkaloids
N-formylconessimine and conimine suppressed MSSA and MRSA with MIC values of 32 and 128 µg/mL, respectively, and conimine increased the sensitivity of MSSA to vancomycin [174].

Concluding Remarks
Weinstein and Albersheim (1983) presented evidence that phytoalexins, especially flavonoids from angiosperms, act as nonspecific membrane antimicrobials that alter the structural integrity of the cytoplasmic membrane, thereby causing the membrane to be a less efficient matrix for membrane-dependent processes [175]. They also argue that phytoalexins with non-specific antimicrobial targets targeting the cytoplasm and proteins also makes it difficult for bacteria or fungi to develop resistance. Furthermore, phytoalexins are often toxic to herbivorous predators as well as repellent and therefrom exhibit low therapeutic indices. In the case of alkaloids, none have been known act on specific antibacterial or antifungal targets [176]. It is for this reason that antibiotic or antifungal alkaloids for systemic use in humans working at micromolar plasmatic concentrations have not been found yet. Research efforts, however, need to continue with the determination of the selectivity indices. Alkaloids from Asian angiosperms represent yet another mind-blowing source of original chemical frameworks that can be used for the hemisynthesis of clinical antibiotics, antimycobacterial agents, and antifungal drugs as well as efflux pump inhibitors of clinical value. 8-Acetylnorchelerythrine, cryptolepine, 8-hydroxydihydrochelerythrine, 6-methoxydihydrosanguinarine, 2 -nortiliacorinine, pendulamine A and B, rhetsisine, sampangine, tiliacorine, tryptanthrin, tylophorinine, vallesamine, and viroallosecurinine with a MIC ≤1 µg/mL are first line candidates.

Supplementary Materials:
The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/antibiotics11091146/s1, Table S1: Medicinal Plants of Asia and the Pacific yielding antibacterial and/or antifungal alkaloids; Table S2: Antibacterial and/or antifungal alkaloid in vitro and from Asian Angiosperms with MIC ≤ 5 µg/mL.