Plant Products as Antimicrobial Agents

Simple phenols and phenolic acids.

Some of the simplest bioactive phytochemicals consist of a single substituted phenolic ring. Cinnamic and caffeic acids are common representatives of a wide group of phenylpropane-derived compounds which are in the highest oxidation state (Fig. 1).

The common herbs tarragon and thyme both contain caffeic acid, which is effective against viruses (245), bacteria (31, 224), and fungi (58).

Catechol and pyrogallol both are hydroxylated phenols, shown to be toxic to microorganisms. Catechol has two −OH groups, and pyrogallol has three. The site(s) and number of hydroxyl groups on the phenol group are thought to be related to their relative toxicity to microorganisms, with evidence that increased hydroxylation results in increased toxicity (76). In addition, some authors have found that more highly oxidized phenols are more inhibitory (192, 231). The mechanisms thought to be responsible for phenolic toxicity to microorganisms include enzyme inhibition by the oxidized compounds, possibly through reaction with sulfhydryl groups or through more nonspecific interactions with the proteins (137).

Phenolic compounds possessing a C₃ side chain at a lower level of oxidation and containing no oxygen are classified as essential oils and often cited as antimicrobial as well. Eugenol is a well-characterized representative found in clove oil (Fig. 1). Eugenol is considered bacteriostatic against both fungi (58) and bacteria (224).

Quinones.

Quinones are aromatic rings with two ketone substitutions (Fig. 1). They are ubiquitous in nature and are characteristically highly reactive. These compounds, being colored, are responsible for the browning reaction in cut or injured fruits and vegetables and are an intermediate in the melanin synthesis pathway in human skin (194). Their presence in henna gives that material its dyeing properties (69). The switch between diphenol (or hydroquinone) and diketone (or quinone) occurs easily through oxidation and reduction reactions. The individual redox potential of the particular quinone-hydroquinone pair is very important in many biological systems; witness the role of ubiquinone (coenzyme Q) in mammalian electron transport systems. Vitamin K is a complex naphthoquinone. Its antihemorrhagic activity may be related to its ease of oxidation in body tissues (85). Hydroxylated amino acids may be made into quinones in the presence of suitable enzymes, such as a polyphenoloxidase (233). The reaction for the conversion of tyrosine to quinone is shown in Fig. 2.

In addition to providing a source of stable free radicals, quinones are known to complex irreversibly with nucleophilic amino acids in proteins (210), often leading to inactivation of the protein and loss of function. For that reason, the potential range of quinone antimicrobial effects is great. Probable targets in the microbial cell are surface-exposed adhesins, cell wall polypeptides, and membrane-bound enzymes. Quinones may also render substrates unavailable to the microorganism. As with all plant-derived antimicrobials, the possible toxic effects of quinones must be thoroughly examined.

Kazmi et al. (112) described an anthraquinone from Cassia italica, a Pakistani tree, which was bacteriostatic for Bacillus anthracis, Corynebacterium pseudodiphthericum, and Pseudomonas aeruginosa and bactericidal for Pseudomonas pseudomalliae. Hypericin, an anthraquinone from St. John’s wort (Hypericum perforatum), has received much attention in the popular press lately as an antidepressant, and Duke reported in 1985 that it had general antimicrobial properties (58).

Flavones, flavonoids, and flavonols.

Flavones are phenolic structures containing one carbonyl group (as opposed to the two carbonyls in quinones) (Fig. 1). The addition of a 3-hydroxyl group yields a flavonol (69). Flavonoids are also hydroxylated phenolic substances but occur as a C₆-C₃ unit linked to an aromatic ring. Since they are known to be synthesized by plants in response to microbial infection (56), it should not be surprising that they have been found in vitro to be effective antimicrobial substances against a wide array of microorganisms. Their activity is probably due to their ability to complex with extracellular and soluble proteins and to complex with bacterial cell walls, as described above for quinones. More lipophilic flavonoids may also disrupt microbial membranes (229).

Catechins, the most reduced form of the C₃ unit in flavonoid compounds, deserve special mention. These flavonoids have been extensively researched due to their occurrence in oolong green teas. It was noticed some time ago that teas exerted antimicrobial activity (227) and that they contain a mixture of catechin compounds. These compounds inhibited in vitro Vibrio cholerae O1 (25), Streptococcus mutans (23, 185, 186, 228), Shigella (235), and other bacteria and microorganisms (186, 224). The catechins inactivated cholera toxin in Vibrio (25) and inhibited isolated bacterial glucosyltransferases in S. mutans (151), possibly due to complexing activities described for quinones above. This latter activity was borne out in in vivo tests of conventional rats. When the rats were fed a diet containing 0.1% tea catechins, fissure caries (caused by S. mutans) was reduced by 40% (166).

Flavonoid compounds exhibit inhibitory effects against multiple viruses. Numerous studies have documented the effectiveness of flavonoids such as swertifrancheside (172), glycyrrhizin (from licorice) (242), and chrysin (48) against HIV. More than one study has found that flavone derivatives are inhibitory to respiratory syncytial virus (RSV) (21, 111). Kaul et al. (111) provide a summary of the activities and modes of action of quercetin, naringin, hesperetin, and catechin in in vitro cell culture monolayers. While naringin was not inhibitory to herpes simplex virus type 1 (HSV-1), poliovirus type 1, parainfluenza virus type 3, or RSV, the other three flavonoids were effective in various ways. Hesperetin reduced intracellular replication of all four viruses; catechin inhibited infectivity but not intracellular replication of RSV and HSV-1; and quercetin was universally effective in reducing infectivity. The authors propose that small structural differences in the compounds are critical to their activity and pointed out another advantage of many plant derivatives: their low toxic potential. The average Western daily diet contains approximately 1 g of mixed flavonoids (124); pharmacologically active concentrations are not likely to be harmful to human hosts.

An isoflavone found in a West African legume, alpinumisoflavone, prevents schistosomal infection when applied topically (175). Phloretin, found in certain serovars of apples, may have activity against a variety of microorganisms (101). Galangin (3,5,7-trihydroxyflavone), derived from the perennial herb Helichrysum aureonitens, seems to be a particularly useful compound, since it has shown activity against a wide range of gram-positive bacteria as well as fungi (2) and viruses, in particular HSV-1 and coxsackie B virus type 1 (145).

Delineation of the possible mechanism of action of flavones and flavonoids is hampered by conflicting findings. Flavonoids lacking hydroxyl groups on their β-rings are more active against microorganisms than are those with the −OH groups (38); this finding supports the idea that their microbial target is the membrane. Lipophilic compounds would be more disruptive of this structure. However, several authors have also found the opposite effect; i.e., the more hydroxylation, the greater the antimicrobial activity (189). This latter finding reflects the similar result for simple phenolics (see above). It is safe to say that there is no clear predictability for the degree of hydroxylation and toxicity to microorganisms.

Tannins.

“Tannin” is a general descriptive name for a group of polymeric phenolic substances capable of tanning leather or precipitating gelatin from solution, a property known as astringency. Their molecular weights range from 500 to 3,000 (87), and they are found in almost every plant part: bark, wood, leaves, fruits, and roots (192). They are divided into two groups, hydrolyzable and condensed tannins. Hydrolyzable tannins are based on gallic acid, usually as multiple esters with d-glucose, while the more numerous condensed tannins (often called proanthocyanidins) are derived from flavonoid monomers (Fig. 1). Tannins may be formed by condensations of flavan derivatives which have been transported to woody tissues of plants. Alternatively, tannins may be formed by polymerization of quinone units (76). This group of compounds has received a great deal of attention in recent years, since it was suggested that the consumption of tannin-containing beverages, especially green teas and red wines, can cure or prevent a variety of ills (198).

Many human physiological activities, such as stimulation of phagocytic cells, host-mediated tumor activity, and a wide range of anti-infective actions, have been assigned to tannins (87). One of their molecular actions is to complex with proteins through so-called nonspecific forces such as hydrogen bonding and hydrophobic effects, as well as by covalent bond formation (87, 210). Thus, their mode of antimicrobial action, as described in the section on quinones (see above), may be related to their ability to inactivate microbial adhesins, enzymes, cell envelope transport proteins, etc. They also complex with polysaccharide (247). The antimicrobial significance of this particular activity has not been explored. There is also evidence for direct inactivation of microorganisms: low tannin concentrations modify the morphology of germ tubes of Crinipellis perniciosa (33). Tannins in plants inhibit insect growth (196) and disrupt digestive events in ruminal animals (35).

Scalbert (192) reviewed the antimicrobial properties of tannins in 1991. He listed 33 studies which had documented the inhibitory activities of tannins up to that point. According to these studies, tannins can be toxic to filamentous fungi, yeasts, and bacteria. Condensed tannins have been determined to bind cell walls of ruminal bacteria, preventing growth and protease activity (105). Although this is still speculative, tannins are considered at least partially responsible for the antibiotic activity of methanolic extracts of the bark of Terminalia alata found in Nepal (221). This activity was enhanced by UV light activation (320 to 400 nm at 5 W/m² for 2 h). At least two studies have shown tannins to be inhibitory to viral reverse transcriptases (111, 155).

Coumarins.

Coumarins are phenolic substances made of fused benzene and α-pyrone rings (161). They are responsible for the characteristic odor of hay. As of 1996, at least 1,300 had been identified (95). Their fame has come mainly from their antithrombotic (223), anti-inflammatory (177), and vasodilatory (152) activities. Warfarin is a particularly well-known coumarin which is used both as an oral anticoagulant and, interestingly, as a rodenticide (113). It may also have antiviral effects (24). Coumarins are known to be highly toxic in rodents (232) and therefore are treated with caution by the medical community. However, recent studies have shown a “pronounced species-dependent metabolism” (244), so that many in vivo animal studies cannot be extrapolated to humans. It appears that toxic coumarin derivatives may be safely excreted in the urine in humans (244).

Several other coumarins have antimicrobial properties. R. D. Thornes, working at the Boston Lying-In Hospital in 1954, sought an agent to treat vaginal candidiasis in his pregnant patients. Coumarin was found in vitro to inhibit Candida albicans. (During subsequent in vivo tests on rabbits, the coumarin-spiked water supply was inadvertently given to all the animals in the research facility and was discovered to be a potent contraceptive agent when breeding programs started to fail [225].) Its estrogenic effects were later described (206).

As a group, coumarins have been found to stimulate macrophages (37), which could have an indirect negative effect on infections. More specifically, coumarin has been used to prevent recurrences of cold sores caused by HSV-1 in humans (24) but was found ineffective against leprosy (225). Hydroxycinnamic acids, related to coumarins, seem to be inhibitory to gram-positive bacteria (66). Also, phytoalexins, which are hydroxylated derivatives of coumarins, are produced in carrots in response to fungal infection and can be presumed to have antifungal activity (95). General antimicrobial activity was documented in woodruff (Galium odoratum) extracts (224). All in all, data about specific antibiotic properties of coumarins are scarce, although many reports give reason to believe that some utility may reside in these phytochemicals (26, 83, 193). Further research is warranted.

In vitro experiments.

(i) Bacteria and fungi.

Initial screening of potential antibacterial and antifungal compounds from plants may be performed with pure substances (2, 23, 117) or crude extracts (73, 203, 182). The methods used for the two types of organisms are similar. The two most commonly used screens to determine antimicrobial susceptibility are the broth dilution assay (18, 89, 219) and the disc or agar well diffusion assay (153); clinical microbiologists are very familiar with these assays. Adaptations such as the agar overlay method (138) may also be used. In some cases, the inoculated plates or tubes are exposed to UV light (221) to screen for the presence of light-sensitizing photochemicals. Other variations of these methods are also used. For instance, to test the effects of extracts on invasive Shigella species, noncytotoxic concentrations of the extracts can be added to Vero cell cultures exposed to a Shigella inoculum (235). The decrease in cytopathic effect in the presence of the plant extract is then measured.

In addition to these assays, antifungal phytochemicals can be analyzed by a spore germination assay. Samples of plant extracts or pure compounds can be added to fungal spores collected from solid cultures, placed on glass slides, and incubated at an appropriate temperature (usually 25°C) for 24 h. Slides are then fixed in lactophenol-cotton blue and observed microscopically for spore germination (179).

Of course, after initial screening of phytochemicals, more detailed studies of their antibiotic effects should be conducted. At this stage, more specific media can be used and MICs can be effectively compared to those of a wider range of currently used antibiotics. The investigation of plant extracts effective against methicillin-resistant S. aureus (190, 229) provides an example of prospecting for new compounds which may be particularly effective against infections that are currently difficult to treat. Sato et al. (190) examined the activity of three extracts from the fruiting bodies of the tree Terminalia chebula RETS against methicillin-sensitive and methicillin-resistant S. aureus as well as 12 other gram-negative and gram-positive bacteria. They found that gallic acid derivatives were more effective against both types of S. aureus than they were against other species.

(ii) Viruses.

Several methods are available to detect either virucidal or inhibitory (antiviral) plant activity. Investigators can look for cytopathic effects or plaque formation (1, 111) or for transformation or proliferative effects on cell lines (48). Viral replication may be assayed by detection of viral products such as DNA, RNA, or polypeptides (127, 171). Vlietinck and Vanden Berghe (238) have noted that the method for assaying antiviral substances used in various laboratories is not standardized, and therefore the results are often not comparable to one another. These authors also point out that researchers must distinguish between merely toxic effects of agents on host cells and true antiviral properties of the plant extracts. Table 5 lists various in vitro antiviral screening assays.

It should be mentioned here that antiviral assays often screen for active substances which prevent adsorption of the microorganism to host cells; this activity is overlooked in screening procedures for antibacterial and antifungal substances. However, the large body of literature accumulating on antiadhesive approaches to bacterial infections (109) suggests that phytochemicals should be assessed for this class of action in addition to their killing and inhibitory activities.

(iii) Protozoa and helminths.

Screening plant extracts for activity against protozoa and helminths can be more complicated than screening for activity against bacteria, fungi, or viruses. Growing the organisms is often more difficult, and fewer organisms are obtained. Assays are generally specific for the microorganism. For instance, Freiburghaus et al. (73) used two methods to assay compounds for effectiveness against Trypanosoma brucei. First, 66-h tissue cultures of the trypanosomes were exposed to various concentrations of extracts, and then the MICs were determined with an inverted microscope. (In this case, the MIC was defined as the lowest concentration of extract which completely inhibited the growth of the trypanosomes.) Second, these authors used a fluorescence assay of trypanosome viability in microtiter wells.

There is great interest in plant chemicals which may have anti-infective properties for Plasmodium species. One method for testing extracts which may be toxic to plasmodia uses radiolabelled microorganisms in an erythrocyte infection assay with microtiter dishes (72). Parasitized erythrocytes are incubated with and without test substances, and the numbers of Plasmodium organisms after the incubation period are quantitated. Another laboratory directly measures decreases in the growth of Plasmodium under the influence of the extracts (29).

In vivo testing of phytochemicals.

Two prime examples of animal studies from the recent literature are the descriptions of the effects of oolong tea polyphenols on dental caries in rats (186) and cholera in mice (226) (discussed above). Another bacterial infection model, P. aeruginosa lung infection of athymic rats, was used to test subcutaneous administrations of ginseng. Treated rats had decreased lung pathology and bacterial load, although enhanced humoral immunity was thought to be at least partially responsible (207).

Perhaps because of the long-standing interest in alternative pharmaceuticals for use in developing and tropical regions of the world, many of the in vivo studies to date have examined the antiprotozoal effectiveness of plant extracts (55). Two recent studies with rats used experimental Entamoeba infections to test the effectiveness of a crude Indian herbal extract (205) and of an ethanolic extract of Piper longum fruits, probably containing the alkaloid piperine (78). After applying the chloroform extract of the legume Milletia thonningii topically to the skin of mice, Perrett et al. (175) found that it prevented the establishment of subsequent Schistosoma mansoni infection. There are also animal studies of the activity of various extracts of the bark of Pteleopsis suberosa against ulcers (53), of papaya latex against infections with the helminth Heligmosomoides polygyrus (191), and of Santolina chamaecyparissus essential oil against candidal infections (213).

An Ayurvedic formulation, Cauvery 100, was administered orally to rats either before or after diarrhea was induced by castor oil. While this trial did not test antimicrobial activities of the preparation but, rather, its effect on gut motility and ion loss in the sick animals, it showed that Cauvery 100 had both a protective and a curative effect in rats (134). A similar study with similar results was performed on rats with chemically induced ulcers (133). The Ayurvedic Ashwagandha was found to prolong the survival period of BALB/c mice infected with Aspergillus, an effect believed to be mediated by increased macrophage function (54).

Clinical trials in humans.

Of course, plants have been used for centuries to treat infections and other illnesses in humans in aboriginal groups, but controlled clinical studies are scarce. In some cases, traditional healers working together with trained scientists have begun keeping records of the safety and effectiveness of phytochemical treatments, but these are generally uncontrolled and unrandomized studies. Examples of peer-reviewed trials of the use of phytochemicals in humans are briefly reported here.

A cross-sectional epidemiological study (thus, not a randomized trial) of the effectiveness of chewing sticks versus toothbrushes in promoting oral hygiene was conducted in West Africa. The study’s authors found a reduced effectiveness in chewing-stick users compared to toothbrush users and concluded that the antimicrobial chemicals known to be present in the sticks added no oral health benefit (156). Also, regarding oral health, mouthrinses containing various antimicrobials have been evaluated in humans (reviewed in reference 240). Mouthrinses containing phytochemicals were not found to be as effective in decreasing plaque or clinical gingivitis as were Listerine or chlorhexidine.

To date, few randomized clinical trials of plant antimicrobials have been performed. Giron et al. (79) compared an extract of Solanum nigrescens with nystatin, both given as intravaginal suppositories, in women with confirmed C. albicans vaginitis. The extract proved to be as efficacious as nystatin. In 1996 the long-recognized effects of cranberry juice on urinary tract infections were studied in a group of elderly women. No data are provided by the author about clinical outcomes, but treated women had fewer bacteria in their urine than did the untreated group (17).

Four different Ayurvedic formulations were tested against a placebo for their effectiveness against acne vulgaris. One of the formulations, Sunder Vati, produced a significantly greater reduction in lesion count than did the other three and the placebo. No toxic side effects were observed (170). Two proprietary compounds derived from tropical plants, Provir, for the treatment of respiratory viral infections, and Virend, a topical antiherpes agent, were tested in clinical trials in 1994 (114). Since then, the efficacy and safety of Virend have been established in phase II studies (167).