The University of KwaZulu-Natal Ethics Committee provided ethical approval for this study and consent processes (BREC approval number: BREC/00001973/2020).

This is a Type C extract consisting of botanical drugs and their extracts derived from lesser-studied species and the drugs derived from them, which are not included in a national or regional pharmacopeia and are not used commercially at an international level. The traditional medicine was obtained from THPs, Mr S Nkabinde and Mr M Nkabinde. The medicinal plant parts, namely, Sclerocarya birrea stem/leaf, Gnidia sericocephala roots, Senna italica roots, and Pentanisia prunelloides roots were collected in Tugela Ferry in Msinga, KwaZulu-Natal by the healers. The plants were verified and classified by Ms Magda Nel, an ethnobotanist at the H.G.W.J Schweickerdt Plant Herbarium at the University of Pretoria where the voucher specimens were deposited as shown in Table 1 (Setlhare et al., 2024). In the laboratory, PN was filter-sterilized and freeze-dried using a Benchtop Freeze Dryer (VirTis, Sp Scientific, Warminster, PA, United States) to form a powder and stored at −20°C until needed for other experiments. A water: ethanol mixture extract of G. sericocephala was manufactured by Afriplex Pharmaceuticals (https://afriplex.co.za) in Paarl, South Africa, through spray-drying technology. For extract preparation, PN (10 mg/mL) was weighed and dissolved in phosphate-buffered saline (PBS) (Thermo Fisher Scientific, Carlsbad, CA, United States) and filter sterilized. Gnidia sericocephala was weighed (10 mg) and first dissolved in 2% dimethylsulfoxide (DMSO) (Sigma-Aldrich, St Louis, MO, Untied States) then 10 mL of PBS to obtain a final concentration of 10 mg/mL.

A patent for PN has been filed in South Africa with reference number 2023/03587 and a global patent with reference number PCT/IB 2024/052,529 has also been filed for PN. Setlhare et al. (2024) and Tembeni et al. (2022) have described the UPLC-HRMS and HPLC analysis of PN and G. sericocephala phytochemicals, respectively.

Zidovudine (AZT), efavirenz, raltegravir, maraviroc and amprenavir were obtained from the National Institute of Health (NIH). These compounds were dissolved in DMSO (not more than 1%) and diluted in PBS. Each drug was reconstituted to prepare a stock of 20 mg/mL Phytohemagglutinin (PHA) purchased from Sigma-Aldrich (St Louis, MO, United States).

The HEK293 cells are a derivative of the human embryonic kidney cells (HEK293T), established by the expression of a temperature-sensitive SV40 T-antigen mutant (Bae et al., 2020; DuBridge et al., 1987). The expression of the T-antigen permits plasmids that carry the SV40 origin of replication to replicate when transfected into the cell (DuBridge et al., 1987). TZM-bl cell line, also known as the JC.53bl-13 cell line, is a genetically modified HeLa cervical cancer cell line engineered to express CD4, CCR5, and CXCR4 (Platt et al., 1998) to allow for HIV-1 entry. The cells harbor integrated reporter genes for firefly luciferase and β-galactosidase under the control of the HIV-1 long terminal repeat (LTR), allowing quantification of infection through luminescence or colorimetric assays (Xing et al., 2016; Wei et al., 2002). Both cell lines were obtained from the National Institute of Health (NIH AIDS Research and Reference Reagent Programme) and maintained in Dulbecco’s Modified Eagle Medium (DMEM) (Gibco, Grand Island, NY, United States) supplemented with 10% heat-inactivated fetal bovine serum (FBS) (Gibco, Grand Island, NY, United States), 100 mL penicillin-streptomycin (Thermo Fisher Scientific, Carlsbad, CA, United States), and 25 mM Hepes buffer (Thermo Fisher Scientific, Carlsbad, CA, United States) at 37°C in a 5% carbon dioxide (CO₂) atmosphere as previously described (Jiang et al., 2010).

Peripheral blood mononuclear cells were obtained from participants enrolled in the Females Rising through Education, Support, and Health (FRESH) cohort, established in 2012 in Durban, KwaZulu-Natal, South Africa. The FRESH study is an ongoing prospective cohort that recruits HIV-1 negative adolescent females aged 18–23 years who are at high risk of HIV-1 acquisition. In addition to biomedical monitoring, the study integrates socioeconomic empowerment interventions aimed at addressing structural factors contributing to HIV-1 vulnerability. Longitudinal blood and mucosal samples are collected pre- and post-enrolment to facilitate the detection and characterization of acute HIV-1 infection (Dong et al., 2018). To date, over 2,500 participants have been enrolled, with 95 cases of acute HIV-1 infection identified. Cryopreserved PBMCs obtained at baseline before HIV-1 acquisition, were made available from 6 study participants. Frozen PMBCs were thawed and cultured in complete growth medium composed of RPMI 1640 containing 10% FBS, 100 μg/mL penicillin-streptomycin, 1% L-glutamine, 1% Hepes, and 10 µg (10 μg/mL) recombinant human IL-2, CF (Roche, Germany) and activated using 5 μg/mL phytohemagglutinin (PHA).

To determine the safety profile of the traditional medicines (PN and G. sericocephala) on PBMCs and TZM-bl cell line, a cell viability assay was performed using the CellTiter-Glo^® Luminescent Cell Viability Assay according to the manufacturer’s protocol (Promega, Madison, United States). Briefly, PBMCs were seeded at 1 × 10⁶ cells/mL in 24-well microtiter plates in R10 medium and treated with extracts at defined concentrations of PN (10, 50, 100, 200, 300, and 325 μg/mL), and G. sericocephala (10, 50, 100, and 106 μg/mL) in triplicates and incubated for 24 h at 37⁰C at 5% CO₂. Untreated cells (cells in cell culture medium) served as the negative controls.

Following incubation, 100 µL of treated/control PBMCs were transferred into a white 96-well opaque plate, and 100 µL of CellTiter-Glo^® reagent was added to each well and the plate was placed in the orbital shaker for 2 min to induce cell lysis. The luminescence was then measured using the Victor Nivo Multimode plate reader (PerkinElmer, Massachusetts, United States). Assays were performed in triplicate from three independent experiments and the 50% cytotoxicity concentrations (CC50) values were generated using GraphPad Prism software (Version 10, GraphPad, La Jolla, CA).

TZM-bl cells were seeded overnight (16–18 h) into a 96-well, flat bottom tissue culture plate at a density of (1 × 10⁴) cells per well in 100 μL of supplemented DMEM growth media. The following day, the cell media was removed and replaced with 200 μL of fresh complete DMEM growth media and treated with varying concentrations of PN and G. sericocephala as per the PBMCs above. The treated cells were incubated at 37°C in a 5% CO₂ incubator for 24 h. Following 24-h incubation, the cells were trypsinized and 100 µL was transferred into the white 96-well opaque plate. The number of viable TZM-bl cells was determined using the CellTitre-Glo reagent procedure as mentioned above for PBMCs, and the CC50 concentration was calculated.

HIV-1 env expression plasmids (pSVIII-NL4.3, pSVIII-CM070P.1, pSVIII-YU2, and pSVIII-CM019P.1.2) and the pBR43-ΔEnv-NefSF2GFP backbone were provided by Dr. Katlego Sojane (HIV Pathogenesis Programme, University of KwaZulu-Natal, Durban, South Africa). Each of the five plasmids (pSIII-YU2, pSVIII-NL4.3, pSVIII-CM070P.1, pSVIII-CM019P.1.2, and pBR43-ΔEnv-NefSF2GFP) was propagated by transforming it into the bacterial strain JM109 Escherichia coli (E. coli) competent cells (Promega, United States, Madison) individually, as per the manufacturer’s instructions. Briefly, 2 µL of each of the plasmid DNA was independently added to its respective 50 µL of JM109 E. coli competent cells and immediately placed on ice for 30 min and heat shock the cells at 42°C. Thereafter, 250 µL of super optimal broth with catabolic repression (SOC) media was added into the transformation mixture and placed in the shaking incubator for 60 min at 37°C at a speed of 230 rpm. In an agar plate containing ampicillin, 100 µL of the transformation mixture was spread on the plate, left to dry, and then the plates were placed in the 37°C incubator for 16 h. The following day, colonies were cultured in lysogeny broth (LB) containing Ampicillin antibiotic and incubated for 16 h at 37°C while shaking at 230 rpm.

The plasmid DNA was purified using the GeneJET plasmid miniprep kit (Thermo Fisher Scientific, Carlsbad, CA, United States) as per the manufacturer’s instructions. After purification, the plasmid DNA quality was assessed by using the NanoDrop™ 2000 software version 1.6.198 (Thermo Fisher Scientific, Carlsbad, CA, United States). The purified PCR product was restricted using BamHI-HF and Kpn1 to confirm the inserts (Thermo Fisher Scientific, Carlsbad, CA, United State). The positive colonies were confirmed by using 1% agarose gel electrophoresis.

The HEK293T cell line was used for pseudovirus production as previously described (Ozaki et al., 2012; Li et al., 2005). Briefly, viral stocks for HIV-1 strains NL4.3 (subtype B, using the CXCR4 coreceptor), CM070P.1 (subtype C, using the CXCR4 coreceptor), YU2 (subtype B, using the CCR5 coreceptor) and CM0190P.1.2 (subtype C, using the CCR5 coreceptor) were produced by co-transfecting exponentially dividing 293T cells (0.3 × 10⁶ cells/mL growth medium in a T-75 culture flask) with 0.8 µg envelope-expressing plasmids (pNL4.3 or, pCM070P.1 or, pYU2 or, and pCM0190P.1.2) and 2.4 µg of an env-deficient HIV-1 backbone vector (pBR43-ΔEnv-NefSF2GFP) using polyethyleneimine (PEI) transfection reagent (Sigma-Aldrich, St Louis, MO, United States). The cell culture supernatants containing pseudotype viruses were harvested 48 h post transfection by filtration (0.45 µm) and then stored in 1 mL aliquots at −80°C until further use. The 50% tissue culture infective doses (TCID50) titer of each pseudovirus aliquot was determined using a TZM-bl luciferase-based assay. For TCID50 measurement, fivefold serial dilutions of pseudovirus were made in septuplicate wells in 96-well culture plates in a total volume of 100 µL growth medium for a total of six dilution steps. Trypsinized TZM-bl cells (10,000 cells in 100 µL of growth medium containing 25 μg/mL DEAE-dextran) were added to each well, and the plates were incubated at 37°C in a humidified 5% CO2 for 48 h. The viral titers were determined by measuring the luciferase activity using the Bright-Glo luciferase reagent, according to the manufacturers’ instructions (Promega, Madison, WI, United States). The plate was read with the Victor Nivo Multimode plate reader (PerkinElmer, Massachusetts, United States). The wells that produced relative luminescence units (RLU) > 3X background were scored as positive. Results were analyzed using GraphPad Prism10. The TCID50 was determined by the virus dilution that was able to produce 20,000–50,000 RLUs to ensure a standardized virus dose in the assay (Li et al., 2005).

The anti-HIV activity of PN and G. sericocephala extracts was assessed using the TZM-bl luciferase-based assay as previously described (Li et al., 2005). Briefly, 50 µL TCID50 of each virus was incubated with serially diluted (eight dilutions, threefold stepwise) concentrations of PN (starting from 32.5 μg/mL), G. sericocephala (106.4 μg/mL) or zidovudine (AZT, 262.5 μg/mL) in triplicate for 1 h at 37°C, humidified 5% CO₂ atmosphere, in a total volume of 150 µL DMEM in 96-well plates. Following incubation, freshly trypsinized cells were added per well and incubated for 48 h under the same conditions. Each experiment included a positive control consisting of untreated TZM-bl cells infected a corresponding virus, referred to as virus-only control (cells + virus) while the negative control was untreated and uninfected TZM-bl cells, referred to as cell-only control (uninfected). Infectivity of the virus was quantified using Bright-Glo™ luciferase assay (Promega, Madison, WI, United States) according to the manufacturer’s protocol. Luminescence was measured with a Victor Nivo Multimode Plate Reader (PerkinElmer, Waltham, MA, United States). The half-maximal inhibitory concentration (IC₅₀) was defined as the concentration of compound (PN, G. sericocephala, or AZT) required to reduce chemiluminescence units by 50% compared to virus control wells. The HIV-1 inhibition percentage was expressed using this calculation: %  inhibition of HIV ‐ 1 = 1 − ( Average           treatment − Average           cell control Average             virus control             control − Average           cell control × 100

The CC50 values represent the concentration of the TM extract or drug that causes 50% toxicity on each cell line. The IC₅₀ values represent the concentration of each treatment that decreases the viral titer by 50%. These values were used to calculate the selectivity index (SI) of each treatment using the following formula: Selectivity index SI = C C 50 I C 50

For confirmation of the anti-HIV-1 activity of crude extracts, the inhibitory effect of PN and G. sericocephala extracts on HIV-1 subtype B and C replication in primary cells was assessed by quantifying p24 antigen levels. Briefly, cryopreserved peripheral blood mononuclear cells (PBMCs) were thawed and rested for 2 h at 37°C in a humidified 5% CO₂ atmosphere in complete growth medium. Rested cells were adjusted to 2 × 10⁶ cells/mL and stimulated with 5 μg/mL phytohemagglutinin (PHA) for 48 h. Activated PBMCs (1 × 10⁶ cells in 100 µL R10) were then infected with HIV-1 subtype B (NL4.3 and YU2) or subtype C (CM070P.1 and CM019P.1.2) and treated with PN, G. sericocephala, or AZT at their CC10 concentrations (corresponding to 90% cell viability), in a final volume of 500 µL. Viral inoculation was performed by spinoculation at 12,000 × g for 30 min, followed by incubation at 37°C for 72 h. Every 3 days, 200 µL of culture supernatant was harvested by centrifugation (1800 rpm, 5 min), replaced with fresh pre-warmed R10 medium containing 10 μg/mL IL-2, and cultures were maintained at 37°C. Collected supernatants were aliquoted and stored at −80°C for quantification of HIV-1 p24 levels using the QuickTiter™ Lentiviral Quantification Kit (Cell Biolabs Inc., San Diego, CA, United States), according to the manufacturer’s instructions.

The cytotoxicity of individual drugs (AZT, raltegravir, maraviroc, and amprenavir) was evaluated using the CellTiter-Glo^® Luminescent Cell Viability Assay. Furthermore, antiretroviral drugs and traditional medicine extracts were combined at a 1:1 ratio to examine potential drug interactions. The IC₅₀ concentrations of each drug and traditional medicine were used individually or in combination to treat TZM-bl cells infected with HIV-1 strains (NL4.3, CM070P.1, YU2, and CM019P.1.2) (Table 2). TZM-bl cells (1 × 10⁵ cells/well) containing 25 μg/mL DEAE-dextran were seeded in 96-well plates and incubated at 37°C with 5% CO₂ for 48 h. Following incubation, the culture medium was replaced with Bright-Glo™ luciferase assay reagent (Promega, Madison, WI, United States), and the plates were incubated for 2 min. Cell lysates were transferred to black 96-well plates and the luminescence was measured using a Victor Nivo plate reader. The IC₅₀ was defined as the drug concentration reducing viral RLUs by 50% compared to virus control wells. after subtracting background RLUs. Drug combination effects were calculated as follows: %  inhibition of HIV ‐ 1 = 1 − ( Average   treatment   − Average   cell   control Average   virus control   − Average   cell   control × 100

The fractional inhibitory concentration index (FICI) for the combination treatments was calculated using the following formula: FICI = ( MICAB ) MICB + ( MICBA ) MICA

Where MICAB is the minimum inhibitory concentration (MIC) of drug A tested in combination, MICA is the MIC of drug A tested alone, MICBA is the MIC of drug B tested in combination and MICB is the MIC of drug B tested alone. Values of FICI ≤ 0.5, FICI > 0.5 but ≤ 4.0, and FICI > 4.0 determine synergistic, additive, and antagonistic interactions, respectively.

All data were analyzed with GraphPad Prism Version 9.5.1 software (GraphPad Software Inc., San Diego, CA, United States) and presented as mean ± SEM (standard error of the mean). Data are representative of three independent experiments done in triplicates. The corresponding dose-response curves were fitted by non-linear regression analysis using a sigmoidal model. Unless otherwise noted, three independent replicates were assessed for each sample and statistical significance was evaluated using a one-way ANOVA test (Kruskal–Wallis), and p-values ≤0.05 were considered statistically significant.

In vitro viability testing of new drug candidates is an essential step in the process of drug discovery and clinical trials (Hughes et al., 2011). The cytotoxicity of each crude extract was evaluated as a percentage of cell viability. PN and G. sericocephala were screened for their effects on TZM-bl cells and human PBMC viability using the CellTiter-Glo^® (ATP) assay (Table 2). The assay demonstrated dose-dependent cellular viability across a concentration range of 16–2000 μg/mL for PN and 1–250 μg/mL for G. sericocephala in both cell types. PN exhibited dose-dependent cytotoxicity with CC50 values of 325.5 μg/mL in PBMCs and 499 μg/mL in TZM-bl cells (Figures 1A,C). Similarly, G. sericocephala showed dose-dependent cytotoxicity with CC50 values of 106.4 μg/mL in PBMCs and 309.2 μg/mL in TZM-bl cells (Figures 1B,D). These findings suggest that both extracts exhibit favorable safety profiles as assessed by the CellTiter-Glo^® assay.

Despite the availability of ART, evaluating medicinal plant extracts for anti-HIV-1 activity remains critical due to limitations such as high cost, toxicity, and the emergence of drug resistance. Plant-based therapies may serve as complementary interventions, especially in resource-limited settings where ART access is restricted, improving treatment efficacy and clinical outcomes (Saehlee et al., 2024; Palshetkar et al., 2020; Kurapati et al., 2016). In this study, the anti-HIV-1 activity of crude plant extracts was assessed using a TZM-bl luciferase assay, which measures LTR-driven luciferase expression as an indicator of viral replication relative to virus-only controls (Tang et al., 2015; Platt et al., 2009). Pseudotype viruses (NL4.3, CM070P.1, YU2, and CM019P.1.2) were only used for single round of infection and allowed us to assess the viral infection levels by measuring the luciferase value in the cell supernatant. As expected, AZT, a licensed antiretroviral drug, effectively inhibited all tested HIV-1 strains (Figure 2A). PN exhibited dose-dependent inhibition across all four viral strains, with maximal inhibition ranging from 93% to 100%. Specifically, PN inhibited NL4.3 by 96% (IC₅₀ = 7.4 ± 33.1 μg/mL), YU2 by 93% (IC₅₀ = 5.97 ± 40.1 μg/mL), CM070P.1 by 100% (IC₅₀ = 0.5 ± 32.1 μg/mL), and CM019P.1.2 by 96% (IC₅₀ = 7.94 ± 27.8 μg/mL) (Figure 2B). G. sericocephala also demonstrated inhibitory activity against both subtype B and C viruses; however, its effect was generally less potent compared to AZT and PN, with inhibition of CM070P.1 at 99% (IC₅₀ = 0.43 ± 30.9 μg/mL), CM019P.1.2 at 83% (IC₅₀ = 1.93 ± 21.5 μg/mL), NL4.3 at 47% (IC₅₀ = 219.3 ± 15.1 μg/mL), and YU2 at 34% (IC₅₀ = 959.4 ± 10.9 μg/mL) (Figure 2C). Taken together these data show that PN extract inhibited the replication of all four viruses, whereas G. sericocephala inhibited the replication of HIV-1 subtype C viruses and showed reduced inhibition activity for HIV-1 subtype B viruses. The selectivity index (SI = CC50/IC50) was also calculated to determine the therapeutic index in vitro, attaining maximum antiviral activity with minimal cytotoxicity. Thus, PN demonstrated enhanced potency in inhibiting HIV-1 replication for subtype B and C strains, with SI values being 44 (NL4.3), 651 (CM070P.1), 54.5 (YU2), and 41 (CM019P.1.2) (Table 2).

Crude extracts of PN and G. sericocephala demonstrated potent inhibition of HIV-1 subtype B and C replication in the TZM-bl luciferase assay. PN, a traditional medicine used by the healer for several years to manage HIV-1 in patients, is typically administered at a dose of three cups per day. Setlhare et al. (2024) reported that PN at 325.3 μg/mL significantly upregulated activation markers and cytokine secretion in healthy PBMCs. To validate the antiviral activity observed in TZM-bl cells, we assessed PN and G. sericocephala in PBMCs infected with HIV-1 subtype B (NL4.3, YU2) and subtype C (CM070P.1, CM019P.1.2) strains using the HIV-1 p24 antigen ELISA. AZT was included at its 50% inhibitory concentration for each viral strain (Table 2). Over an 11-day period, PN, G. sericocephala, and AZT treatments significantly reduced p24 levels compared to untreated controls, which ranged from 37,000 to 75,000 pg/mL (Figures 3A–D). Partial inhibition was observed for NL4.3, where p24 levels exceeded the 20,000 pg/mL threshold, likely due to suboptimal dosing (Figure 3A). In contrast, p24 levels for CM070P.1, YU2, and CM019P.1.2 remained below 20,000 pg/mL over 7 days, indicating sustained viral suppression (Figures 3B–D). Taken together, these findings highlight the strain-dependent antiviral efficacy of the extracts and AZT in PBMCs.

African Traditional Medicines (ATMs) are commonly used by people living with HIV/AIDS (PLWH), often alongside antiretroviral therapy (ART). However, their clinical efficacy and safety in combination with ARVs remain underexplored. Understanding these interactions is essential for optimizing therapeutic outcomes and minimizing adverse effects, particularly in the context of combination regimens aimed at enhancing antiviral efficacy (Sibanda et al., 2016; Martínez et al., 2014; Park et al., 2014; Lee et al., 2006). To address this gap, we evaluated the cytotoxicity and antiviral efficacy of four antiretroviral drugs including AZT, raltegravir, maraviroc, and amprenavir alone or in combination with traditional medicinal plant extracts (PN and G. sericocephala) on HIV-1 subtype B (strains NL4.3 and YU2) and HIV-1 subtype C (strains CM070P.1 and CM019P.1.2) in TZM-bl cells. Non-cytotoxic concentrations of each anti-HIV-1 drug were used to assess the anti-HIV-1 effects in TZM-bl cells (Supplementary Figure S1). An antiviral assay was conducted to assess the impact of raltegravir, maraviroc, and amprenavir on HIV-1 replication. As shown in Figures 4A–C, the inhibitory effects of raltegravir, maraviroc, and amprenavir against HIV-1 were ∼37.5–100% and the IC50 values were determined (Table 2), representing 50% inhibition of viral replication of all viruses. Raltegravir, maraviroc, and amprenavir exhibited a dose-dependent inhibition of HIV-1 strains NL4.3, CM070P.1, YU2, and CM190P.1.2 (Figures 4A–C). No cytotoxicity was observed for drug and plant extract combinations in TZM-bl cells (Supplementary Figure S2).

PN extract combined with AZT exhibited different percentage inhibition of HIV-1 subtype B strain NL4.3 (64%) or amprenavir (93%), while PN extract combined with raltegravir (72%) or maraviroc (81.3%, p = 0.0361) inhibited subtype B strain YU2 (Figures 5A,C). For HIV-1 subtype C, the combination of PN with maraviroc (90%), raltegravir (98.7%, p = 0.0083), and amprenavir (66.6%) inhibited the replication of HIV-1 strain CM070P.1, and PN with AZT (99%, p = 0.0428) inhibited replication of CM019P.1.2 relative to the administration of PN alone (Figures 5B,D). G. sericocephala combined with AZT (80.3%, p = 0.0105) significantly inhibited HIV-1 NL4.3 replication. Its combination with AZT (93.7%), raltegravir (92%), and amprenavir (93%) inhibited the replication of HIV-1 YU2 (Figures 5E,G). Additionally, G. sericocephala combined with maraviroc (87%, p = 0.0093), raltegravir (86%, p = 0.0168), or amprenavir (90%, p = 0.0006) significantly inhibited subtype C strain CM070P.1, while its combination with AZT (89.3%) inhibited HIV-1 strain CM019P.1.2 relative to G. sericocephala and the drugs alone (Figures 5F,H). FICI analysis indicated no synergy, additivity, or antagonism in crude extract and antiretroviral drug combinations. The FICI values for all combinations ranged from 1 to 3.5, indicating an indifferent effect (Supplementary Table S3). In summary, combination of PN and G. sericocephala extracts with antiretroviral drugs, showed enhanced inhibition of HIV-1 subtype B and C strains compared to individual treatments.