Financial support of this work was provided by the School of Biomolecular and Physical Sciences, Griffith University, Australia.

Introduction

Plants have long been recognised as a valuable source of medicines for treating a variety of different diseases and complaints. Most, if not all civilisations, have used plants as medicines. The use of plant natural therapeutics in Asia is wide spread, being used in the treatment of numerous disorders including eczema, malaria and respiratory disorders [₁]. Africa also has a long history of medicinal plant use. For example, Phytolacca dodecandra is used as a moluscicide in the control of schistosomiasis [₂]. The antitumour agent's vinblastine and vincristine (derived from Catharanthus roseus) are currently used in the treatment of a variety of tumours [₃; ₄]. Europe and the Americas also have a history of medicinal plant use. Studies demonstrate the myriad of medicinal plant uses by indigenous North and Central Americans [₅] and South Americans [₆]. Approximately 1500 medicinal plant species are currently in use in Europe [₁].

As a result of its isolation, Australia has a variety of unique and distinctive flora not found elsewhere in the world. Australian Aborigines used a variety of plant medicines to help maintain their health prior to European settlement [₇; ₈]. More than 150 plants from nearly 60 widely varied botanical families were used by Australian Aborigines as antiseptic agents [₈]. Although there is enormous potential for the development of medicinal agents from Australian plants, much of our knowledge of the antimicrobial nature of Australian plants is anecdotal. Recent studies [₉; ₁₀; ₁₁] have demonstrated the antibacterial and antifungal activity of extracts from a wide variety of Australian plants.

To be medically useful as antimicrobial agents, plant preparations should be non-toxic or of low toxicity to human cells [₁₂]. Limited information is available on the toxicity of antibacterial preparations from Australian plants. Recent studies have reported on the low toxicity of Backhousia citriodora essential oils [₁₃] and on the toxicity of extracts from a variety of Australian plants [₁₀] towards human cell lines. A study from this laboratory has indicated the ability of several Australian plants to act as antimicrobial agents [₁₁]. To further assess their potential it is necessary to assess their toxicity. The Artemia franciscana nauplii (brine shrimp larvae) lethality bioassay was used in the current study. This assay has been used to examine the toxicity of a wide variety of compounds [₁₄]. It is an efficient, inexpensive and relatively rapid way to detect toxic compounds, requiring only low amounts of sample (<20 mg). This test correlates well with cytotoxic activity of some human tumours and therefore has the potential to detect new antitumour agents [₁₅].

Materials and Methods

Plant Material

Results

Thirty nine samples from twenty five Australian native plant species were extracted in methanol and dried as previously described [₁₁]. The weight of the dried extracted material is recorded in table 1. The weight of dried extractable material varied across samples, ranging from 83 mg (L. juniperium flowers) extracted per 1 g starting plant material up to 567 g (C. citrinus flowers) from the original 1 g of ground dried plant material. All extracts were resuspended in 15 ml of 20 % methanol, resulting in the crude test extract concentrations reported in table 1.

Figure 1

Table 1: Botanical names of plant species extracted, weight of dried extractable material and the concentrations of each extract.

All extracts were diluted to 2000 µg/ml in artificial seawater for toxicity testing, resulting in a 1000 µg/ml concentration in the Artemia franciscana lethality bioassay. The results of A. franciscana bioassay screening of the Australian plant methanolic extracts are shown in table 2. Previous reports [₁₈; ₁₉] express LC₅₀ values of toxins following 24 h of exposure. Of the 39 extracts tested, 7 (18 %) showed greater than 50 % mortality at 24 h. Of these, A. complanta flowers, A. littoralis leaves, E. baileyana leaves and E. major flowers have been previously shown to have significant antibacterial activity [₁₁]. Of the positive controls, only Mevinphos displayed significant mortality at 24 h. Potassium dichromate treatment resulted in only approximately 10 % mortality at 24 h. Therefore, mortality at later times (48 and 72 h) are also reported here.

An extra four extracts showed greater than 50 % mortality at 48 h compared to 24 h (B. acerifolius leaves, E. major leaves, G. juncifolia flowers and G. robusta leaves). This equates to approximately 28% of the tested extracts showing toxicity at 48 h. Of these, only E. major flowers have been previously shown to have significant antibacterial activity [₁₁]. At 72 h, 8 further extracts (compared to 48 h) induced greater than 50 % mortality. Of these, B. citriodora, C. citrinus leaves, C. salignus leaves and flowers, L. bracteata flowers, and S. leuhmannii leaves have been previously shown to have significant antibacterial activity [₁₁].

Figure 2

Table 2: Toxicity of Australian plant extracts (1 mg/ml) to . Numbers indicate the mean % mortality of at least triplicate experiments ± standard deviation.

Ten plant extracts induced greater than 50% mortality by 48 h (A. complanta leaves, A. littoralis leaves, Brachychiton acerifolius flowers, Eucalyptus baileyana leaves, Eucalyptus major leaves and flowers, Grevillea juncifolia leaves and flowers, Macadamia integriflora flowers and Melaleuca quinquenervia leaves). These were considered sufficiently toxic to warrant further investigation to determine the dependence of toxicity on the concentration of the extract the A. franciscana is exposed to. Table 3 shows the LC₅₀ values of these extracts towards A. franciscana. E. baileyana leaf extract was the most toxic of the plant extracts tested with 24, 48 and 72 h LC₅₀ values of 216 µg/ml (± 22). Macadamia integriflora, whilst taking longer to induce mortality, proved to be similarly toxic with 48 h and 72 h LC₅₀ values almost identical to those of E. baileyana leaf extract.

Figure 3

Table 3: LC (95% confidence interval) for brine shrimp nauplii exposed to Australian plant extracts and the reference toxins Mevinphos and potassium dichromate.

- indicates that LC₅₀ values were unable to be obtained as no increase in mortality above seawater controls was evident.

Discussion

A previous study from this laboratory [₁₁] has reported on the antibacterial activity of methanolic extracts from some Australian native plants. That study indicated that A. aulacocarpa, B. citriodora, B. celsissima, A. littoralis, as well as members of the Callistemon, Eucalyptus, Leptospermum and Syzygium genuses are particularly promising as antimicrobial agents. The current study reports on the toxicity of these and other Australian plant extracts. A. aulacocarpa leaf extract was previously reported to be a good antimicrobial agent [₁₁], being capable of inhibiting the growth of both Gram-positive and Gram-negative bacteria. The current studies indicate this extract has low toxicity towards A. franciscana. The Artemia franciscana bioassay has been reported to be a good indication of toxicity towards human cells [₁₄]. These results indicate that A. aulacocarpa leaf extract has potential for medicinal use as an antibiotic agent.

Likewise, studies within this laboratory [₁₁] have also shown L. bracteata leaves, L. juniperium leaves and flowers and S. australe leaves to be versatile antibacterial agents towards Gram-positive and Gram-negative bacteria. B. celsissima leaves were also good antibacterial agents, preferentially inhibiting Gram-positive bacteria. All of these extracts showed low toxicity in the Artemia bioassay, indicating their potential as antibiotic agents.

B. citriodora, A. littoralis, C. citrinus, C. salignus, E. baileyana and E. major extracts have all also been reported to have good antibacterial activity [₁₁]. The current study shows that all of these extracts were toxic towards A. franciscana. The B. citriodora toxicity results contrast with the low toxicity previously reported [₁₀]. However, it is worth noting that the current study examined the toxicity of higher extract concentrations (1000 µg/ml) compared to the Setzer et al [₁₀] report (250 µg/ml), which studied the toxicity of B. citriodora extracts towards human cell lines as well as towards brine shrimp. Similarly, it is worth noting that toxicity was only seen in the current studies following 72 h of exposure of A. franciscana to B. citriodora extract. No mention of Artemia exposure time is made in the previous study of B. citriodora toxicity [₁₀]. However, 24 h and 48 h LC₅₀'s have been reported for other plant extracts in many previous studies [₁₄; ₁₅]. It is therefore likely that these authors recorded mortality following a shorter exposure time than reported here. These differences highlight the necessity to further evaluate the toxicity of these extracts against human cell lines. Extracts that show toxicity towards A. franciscana at the concentrations tested, may have low toxicity towards human cell lines and therefore may have potential as antibiotic agents. Even if this is not the case, these extracts still have potential value as antiseptic and cleaning agents.

Toxicity towards A. franciscana could also indicate other potential medicinal uses for these plant extracts. The A. franciscana bioassay has previously been shown to be a good indicator of antitumour activity [₁₅]. Therefore, extracts toxic towards A. franciscana should also be tested for toxicity towards human tumour cell lines. Whilst the extracts examined in this report appear promising as antimicrobial agents and possibly as anti-tumour agents, further studies using human cell lines are needed to determine the suitability of these extracts for these purposes.

Acknowledgements