Ethno-botanical surveys on traditional medicinal plants used for treating common ailments, those most traded and threatened were carried out in five districts: Bulilima, Chimanimani, Chipinge, Mangwe and Matobo and the plants identified and authenticated by the National Herbarium and Botanic Gardens, Harare, as previously reported [8] . The plant materials roots, leaves, twigs, tuber, bark or whole, were dried, finely ground and stored.

Selected plant parts were powdered (30 g), macerated in methanol (200 ml) [13] , filtered and the filtrate evaporated off under reduced pressure, freeze dried and stored at −20˚C. Part of the lyophilised extracts were dissolved in DMSO (10 mg/ml), filtered under aseptic conditions and stored for further bioactive use at −20˚C.

Bioactivity testing using the Brine Shrimp (Artemia salina) Lethality Test (BSLT) method was carried out in triplicate with different concentrations of the sterilized plant extracts in brine solution (10 µg/ml - 1000 µg/ml) and the percentage lethality of the nauplii determined [14] . The concentration of the extract that kills 50% of the shrimps, “Lethal Concentration₅₀” (LC₅₀) values were recorded using Nerium oleander as the positive control by GraphPad Prism 5.0 linear regression and Pearson’s two tailed analysis for 95% confidence limits (95% CI) [15] .

The lyophilized methanol extracts were dissolved in DMSO to a concentration of 10 mg/ml. Vero cells (African green monkey kidney) (Highveld Ltd., South Africa) were grown and maintained in essential growth medium supplemented with 5% foetal calf serum and incubated under 5% carbon dioxide at 37˚C. The confluent cells were removed and trypsinated with phosphate buffered saline solution to achieve a cell concentration of 1 - 2 × 10⁴ cells per well for VERO cells.

Herpes simplex virus 2 (HSV-2) (Highveld Ltd., South Africa) was propagated on Vero cells and the recovered virus suspension (1 ml) diluted by infecting Vero confluent monolayers grown on microtitre plates with 0.1 ml of serial tenfold dilutions of the virus suspension quadruplicated and observed for 7 days for cytopathic effect (CPE). The HSV-2 virus titre was obtained from the 50% tissue culture infections dose per ml (TCID_50/ml) [16] .

Cytotoxicity of the different plant extracts that could cause non-specific cytopathic effect (CPE) on confluent Vero cells in growth medium was determined by using 100 ml of serial two fold dilution of the plant extracts onto confluent Vero cells on microtitre plates in quadruplicate. The 50% cytotoxicity concentration (CC₅₀) was defined as the plant extract concentration causing 50% CPE compared to uninfected cells [16] . The maximum nontoxic dilution (MNTD) was taken as the next dilution after the TCID₅₀ of the plant extract and used for the antiviral assays.

Antiviral assays were carried out using the Neutralization test (NT) to determine non-cytotoxic concentration (ID₅₀) that inhibits/protects 50% of the monolayer cells against destruction by the virus compared to uninfected cells using the Spearman-Karber formula [17] and by the End point titration technique (EPTT) [18] . In the EPPT assay, confluent monolayer Vero cells were grown on 96-well microtitre plates with the medium removed from the wells. 0.05 ml (MNTD) of the plant extract and 0.05 ml of growth maintenance medium followed 1 h later by 0.05 ml of tenfold HSV-2 serial diluted virus suspension and the mixture incubated at 37˚C. Controls consisted of Vero cells infected with virus and not treated with plant extract and virus uninfected and untreated cells. All tests were compared with a positive control, acyclor (Sigma). The antiviral activity was determined as the reduction factor (RF) of the viral titre: the ratio of virus titre in the absence and presence of the plant extract. A promising antiviral has a RF ≥ 10³ [13] [18] .

The bioactivity of the plant extracts using the Brine shrimp lethality test (BSLT) indicated that plants in this category were relatively toxic with 15 of the 26 extracts recording LC₅₀ < 500 µg/ml (Table 1). Only 6 extracts had LC₅₀ > 1000 µg/ml. Potentially very toxic plants were T. sericea (LC₅₀ = 66.7 µg/ml) and K. africana (LC₅₀ = 117.4 µg/ml) showing greater toxicity than the positive control N. oleander (LC₅₀ = 141.7 µg/ml). The BSLT bioassay is widely used as a basic screening test for toxicity in prospecting for suspected biological activity in crude and isolated plant extracts from traditional folklore [4] [14] [15] [19] - [22] .

Cytotoxicity effects for the extracts on Vero cells showed all 26 extracts had CC₅₀ < 500 ug/ml, 19 with CC₅₀ < 100 ug/ml and 12 with CC₅₀ < 50 ug/ml (Table 1) indicating potential antitumour and/or antiviral activity. Antiviral screening tests indicated some plant extracts possess potential activities to protect the cells against HSV-2

Plant part investigated: lf = leaf; rt = root; tb = tuber; bk = bark; Environmental status S: Cm = common; Th = threatened; VTh = very threatened; BSLT, Brine shrimp lethality test LC₅₀, lethal concentration that kills 50% of the shrimps with corresponding 95% confidence intervals (95% CI); CC₅₀ (50% CPE), plant extract cytotoxicity concentration that kills 50% tissue cells; NT, neutralisation test: NA = no activity; ID₅₀, non-cytotoxic concentration that inhibits/protects 50% uninfected cells, NA = no activity; TI: therapeutic index = CC₅₀/ID₅₀ EPPT, End point titration test. RF^b antiviral reduction factor: ratio of viral titre of control in absence of extract over viral titre in presence of extract.

induced destruction with EPTT reduction factors (RFs) ≥ 10³ (Table 1). D. cinerea (RF = 10⁴) and K. Africana (RF = 10⁴) were particularly active, however, most plants in the group had RFs ≥ 10³ which activities were also mirrored in the neutralization test (NT) ID₅₀s where acyclovir the positive control had ID₅₀ = 1.5. G. senegalensis, T. sericea and W. salutaris previously noted for their high antimicrobial activities [8] were also in the group of potential antiviral plants with high cytotoxicities (CC₅₀ < 50 µg/ml) and high reduction factors (RF = 10³) for some of their extracts. However, in addition to having a high reduction factor, a plant extract should also have a high therapeutic index (TI)/high sensitivity value in order to prevent cell destruction and be considered for further potential antiviral investigation. From this group (Table 1), Cassia abbreviata had TI values of 7 and 15 for leaf and root extract respectively, Dichrostachys cinerea TI = 7.5 for leaf extract and Hypoxis rooperi had TI = 15 for the tuber. Kigelia fruit showed a high reduction factor 10⁴ but low therapeutic efficacy 1.2 in line with other reports where its anti-tumour and other biological properties are highlighted [21] - [28] . Plant extracts Gymnosporia senegalensis (Celastraceae), Warburgia salutaris (Canellaceae) and Terminalia sericea (Combretaceae) are popular plants in traditional practice. Recent studies are focusing on the toxicity, antiviral and antitumor properties of crude and isolates against different viral species especially HIV-1, HSV-1, HSV-2, anti- Dengue [29] - [38] . In general, all the plants in this study displayed both antiinfective and toxic properties (Table 1) whilst reports from phytochemical studies reflect flavonoids and condensed tannins largely responsible for the observed antiviral and anticancer activities [25] - [28] [38] [39] . Isolated fractions and derivatives can have greater toxicity and/or activity than the crude extracts, the case of artemisinin from Artemisia annua in the treatment of malaria [39] and the isolated phytoconstituents from Kigelia and Hypoxis species [24] - [27] .