HPLC grade hexane was purchased from Sigma Aldrich (St. Louis MO, USA). All chemicals used in preparing various reagents for phytochemical screening were analytical grade and high purity. Two bacteria (gram-positive: Staphylococcus aureus ATCC 25923; gram-negative: Salmonella typhimurium ATCC 7222569) and one fungus ( Candida albicans ATCC 10231) were provided by the Microbiology section of the Department of Public Health, Pharmacology and Toxicology of the University of Nairobi.

Fresh roots of F. hildebrandtii were collected from Kilome constituency, Kitaingo sub location in Makueni County in Kenya (GPS coordinates: 1.7726° S, 37.4498° E). Identification and authentication of the plants were done at the herbarium, Department of Land Resource Management, University of Nairobi. Identification and authentication involved the determination of the botanical origin of the plant, determination of the scientific binomial name, determination of the vernacular name, site of collection, habitat and season of collection, altitude and the parts collected ^{11, 12} . The keys available in the book titled ‘Kenya trees, shrubs, and lianas’ ¹¹ was used to arrive at the said genera and the specific epithet.

The collected roots were gently washed in tap water to remove dirt. They were then shade dried in the laboratory at room temperature for ~2-3 weeks. After completely drying, plant material was pulverized by use of a mechanical mill and subsequently sieved using a 0.45µm pore Whatman ^® membrane filter to obtain a powder of suitable consistency. The powdered material was weighed and divided into two portions. One portion (~500g) was mixed with ~2000 ml of sterile distilled water, allowed to macerate for 72 hours and centrifuged at 3000 rpm for 10 minutes. The supernatant was collected, filtered through a 0.45 µm membrane filter, and the filtrate lyophilized, and appropriately stored in an amber coloured bottle (henceforth known as AQRFH). To the second portion, 1250 ml of hexane was added and the mixture allowed to macerate for 72 hours, centrifuged at 3000 rpm for 10 minutes, and the supernatant filtered using a 0.45 µm membrane filter. The filtrate was then reduced by use of a rotary evaporator at 40℃ to obtain a dried product (henceforth known as HEXRFH).

AQRFH and HEXRFH were analysed for the presence of various phytochemical metabolites, including alkaloids, cardiac glycosides, flavonoids, phenolics, saponins, phytosterols, tannins and triterpenes using standard procedures ¹³ . See Table 1 for the exact tests carried out.

Adult healthy male and female albino mice ( Mus muculus; BALB/c strain; n=66: 42 females, 24 males; age: 8 to 12 weeks; body weight: 20–30 g) were used for this study. The number of animals used were based on the Organization for Economic Corporation and Development (OECD 423) test guidelines ¹⁴ ( Extended data: Table S1 ¹⁵ ). Female mice were nulliparous and not pregnant and all animals were sourced from the Animal House of our institution (University of Nairobi). They were housed in polypropylene cages in the laboratory for at least one week to acclimatize to standard conditions (temperature: 25±3°C; relative humidity: 56–60%; 12 hours of light and 12 hours of darkness). Fluorescent room lights were switched on at 0600h–1800h (light cycle) and off at 1800h–0600h (dark cycle). They were fed on standard mice pellets from a commercial feed supplier (Unga Group Plc, Kenya) and water was provided ad libitum.

All efforts were made to ameliorate any suffering of the animals by adopting the OECD 423 test guidelines ¹⁴ as well as the recommendations of the Biosafety, Animal Use and Ethics Committee (BAUEC) of the University of Nairobi (BAUEC/2018/163) ( Extended data: Figure S1 ¹⁵ ).

All treatments on the experimental animals were carried out in rat cages during the daytime/light cycle (0600h to 1800h) at the animal holding unit of the Department of Public Health, Pharmacology, and Toxicology, University of Nairobi.

Acute toxicity study. The acute toxic class (OECD 423) test guidelines on acute oral toxicity were used ¹⁶ . In total 18, 8–12-week-old female albino mice of 18–20 g were labelled from 1–6 in such a way that there were three animals bearing each number ¹⁷ . The number assigned to each animal was then written on a piece of paper (n=18), folded and placed on a receptacle which was then shaken ¹⁷ . A paper was withdrawn at random and the animal that bore the number was assigned into the respective cages labelled 1 (distilled water control group), 2 (extra virgin oil control group), 3 (300mg/kg AQRFH extract group), 4 (2000mg/kg AQFRH group), 5 (300mg/kg HEXFRH group) and 6 (2000mg/kg HEXFRH group) ¹⁷ . Each animal was individually weighed at day 0, 7, and 14 and the observation of the weights noted ( Extended data: Table S2 ¹⁵ ). The mice were then fasted for 4 hours before dosing.

Two doses (300 and 2000 mg/kg of body weight) of either AQRFH or HEXRFH were prepared by accurately weighing the extracts into 250 ml volumetric flasks, mixing well in distilled water or extra virgin oil respectively, and making up to the mark with either distilled water or extra virgin oil. Dosing of mice followed the OECD 423 test guidelines (i.e. 1ml per 100g body weight of mice) ¹⁴ . Therefore, using the weights of mice as a guide, suitable amounts (in ml, and corresponding to the appropriate doses) of either AQFRH or HEXFRH suspended in distilled water and extra virgin oil, respectively, were drawn into 2ml syringes attached to a gastric gavage and administered to mice. Two groups (Groups 1 and 2) served as controls, and mice in these groups received by gastric gavage a daily dose of distilled water and extra virgin oil. Group 3 mice received a 300 mg/kg dose of AQRFH, Group 4 mice received a 2000 mg/kg dose of AQRFH, Group 5 mice received a 300 mg/kg dose of HEXRFH, and Group 6 mice received a 2000 mg/kg dose of HEXRFH.

Food and water consumption of the mice per treatment was evaluated at 0, 7 and 14 days after treatment ( Extended data: Table S3 ¹⁵ ). At the end of the 14 days, the mice were sacrificed and organs such as the liver, kidney, lungs and spleen were weighed and used to calculate the relative mean organ weight as below:

Relative organ weight = absolute organ weight weight of mice at sacrificing × 100

The data on relative mean organ weight of mice in the acute toxicity protocol is summarized in Extended data: Table S4> ¹⁵ .

Sub-acute toxicity studies. This study was carried out using 24 male and 24 female albino mice (20–30g). They were grouped by randomized complete block design into 6 treatment groups and 2 control groups (n=6 mice/group). Control group mice received an oral daily dose of 0.6ml distilled water (aqueous extract control) and 0.3ml of extra virgin oil (hexane extract control) for 28 days. Treatment group mice received graded doses (250, 500 and 1000mg/kg body weight) of AQRFH and HEXRFH which were administered once daily for 28 days. The body weights of the mice were noted at day 0, 7, 14,21, and 28 and the doses adjusted accordingly ( Extended data: Table S5 ¹⁵ ).

Clinical observations. Changes in fur and skin colour, mucous and eye membranes, respiratory, circulatory autonomic and central nervous systems were noted ¹⁶ . Other observations to be noted included tremors, convulsions, confusion, salivation, diarrhoea, coma and death. Observations were done at intervals of 30 minutes, 4 hours, 24 hours, 48 hours, 7 days, 14 days, 21 days and 28 days ¹⁶ .

Body weight. Individual body weights of mice in the sub-acute toxicity protocol were recorded on day 0, 7, 14 and 28 ( Extended data: Table S5 ¹⁵ ). Body weight was also recorded prior to the first dosing on day 0 and terminally (after fasting) prior to necropsy ( Extended data: Table S5 ¹⁵ ). The food and water consumption in the sub-acute toxicity protocol are summarized in Extended data: Table S6 ¹⁵ .

Necropsy and organ weight. All animals were fasted overnight prior to necropsy. The animals were euthanized in 0.5% v/v halothane (Piramal Enterprises, India; Batch number K37L17B) in a bell jar and vital organs (stomach, kidney, lung, spleen, heart and liver) were excised from both male and female mice, washed gently and observed macroscopically for lesions or any abnormal signs. Organ weights (absolute and relative) were determined for the excised organs by placing them in absorbent paper for a few minutes and weighed. The relative organ weight of each mice was then determined as below:

Relative organ weight = absolute organ weight weight of mice at sacrificing × 100

The relative mean organ weight of mice in the sub-acute toxicity protocol are summarized in Extended data: Table S7 ¹⁵ .

Clinical pathology. At the end of dosing, mice were fasted overnight prior to scheduled necropsy. On day 29, mice were anaesthetized using halothane inhalation in a bell jar, blood samples were collected by use of cardiac puncture into vacutainers containing anticoagulant (EDTA for haematological parameters) and without anticoagulant (plain tubes for biochemical parameters). About 0.5 ml of the blood in plain tubes was centrifuged at 3000 rpm for 10 minutes to obtain serum, which was then stored at −20 °C awaiting further use in biochemical assay of aspartate amino transferase (AST), alanine amino transferase (ALT), total protein (TP), urea and creatinine using specific kits (Catalyst AST Aspartate amino transferase-98-11067-01, Catalyst ALT/GPT Alanine amino transferase-98-11069-01, Catalyst TP Total Protein-98-11085-01, Catalyst BUN Urea-98-11070-01, and Catalyst CREA Creatinine-98-11074-01: Magnum Veterinaria, Laagri Arimaja, Vae 16, Estonia). In addition, ~1.3ml of the blood collected in the anticoagulant containing vacutainers was used to evaluate levels of haematocrit (HCT), total white blood cell (WBC) count, total red blood cell (RBC) count, haemoglobin (Hb), mean corpuscular volume (MCV), mean corpuscular haemoglobin (MCH), and mean corpuscular haemoglobin concentration (MCHC) using IDEXX kits (One Idexx drive, Westbrook, Maine, 04092, United States).

Histopathological studies. Histopathological studies were carried out on organ samples of liver and kidney of the mice. These organs were surgically removed from the mice and fixed in 5% buffered formalin. Tissue samples were then dehydrated in graded doses of ethanol (70–99.9%; Fischer Scientific Ltd.; Batch number 0556512, washed in toluene (Fischer Scientific: Lot no: 193377, and enclosed in paraffin (Honeywell Fluka ^TM: 18634H). Thin tissue sections of 5µm were obtained on a rotatory microtome and stained with haematoxylin-eosin dye and analysed microscopically for any pathological alterations. The photomicrographs obtained were labelled for identification.

Antimicrobial activity. Staph. aureus (gram positive bacteria), S. typhimurium (gram negative bacteria) and C. albicans (fungus) were cultured on blood agar (Oxoid Ltd.; Lot number 1623360) and incubated overnight at 37°C in an incubator (Memmert, Germany). The cultures were suspended in 5 ml sterile physiological buffed saline and turbidity adjusted to 0.5 McFarland to give a concentration of approximately 1.5 × 10 ⁸ colony forming units per ml (CFU/ml). In total, 800 mg of AQRFH and HEXRFH were dissolved in 3ml of sterile distilled water and virgin oil, respectively. Serial dilutions ranging from 3.125–400 mg/ml of the extracts were then prepared using sterile peptone water (Oxoid Ltd; Lot number 1721983). Subsequently, using a sterile 1ml pipette, the adjusted individual micro-organisms ( Staph. aureus, S. typhimurium and C. albicans were inoculated into every tube of the diluted plant extracts to give a final concentration of 5×10 ⁶ CFU/ml. The tubes were then incubated at 37°C for 24 hours. Serial dilutions of the plant extracts devoid of micro-organisms were used as negative control. Positive control tubes had the respective microorganisms without serial dilutions of the plant extracts. Peptone water was used to blank the spectrophotometer. Flucloxacillin (10mg/ml; Dawa Pharmaceuticals Ltd-Kenya; Batch number MB18085 and fluconazole (10mg/ml); Universal Corporation Ltd-Kenya; Batch number 5805073) were used as positive reference compounds of antimicrobial activity.

Baseline optical density at a wavelength of 450 ηm readings for each of the tubes were taken immediately after all tubes had been prepared using a spectrophotometer (Spectronic 21D, Milton Roy, USA). Optical density readings were also taken after 24 hours. A comparison between baseline optical density readings and optical density readings after 24 hours of inoculation of the tubes was made. This was done to determine whether there was growth or not. The minimum inhibitory concentration (MIC) was defined as the lowest concentration of the extracts that inhibited any visible growth of bacteria as determined by the absence of change in the optical density readings between baseline values and those obtained after incubation of the tubes. For the determination of the minimum bactericidal/fungicidal concentration (MBC/MFC), 100µl of broth was taken from all the MIC tubes after 24 hours of incubation at 37°C and sub-cultured aseptically to Plate Count Agar (Oxoid Ltd.; Lot number 1626187). The plates (WOM-CHINA; Lot number RT20171103) were incubated at 37°C for 24 hours and checked for bacterial growth. The lowest concentration of the plant extracts that showed no bacterial growth was considered the MBC/MFC, defined as the lowest concentration of the extract which is responsible for killing greater than 99.9% of the initial bacteria inoculums.

Results of haematological and biochemical tests were expressed as mean ± of standard deviation of three independent observations and analysed using ONE-WAY ANOVA on Statistical Package for the Social Sciences (SPSS, version 21.0). Tukey’s test was used as post hoc. p< 0.05 was considered significant.

AQRFH and HEXRFH yielded about 9% and 2.74% of extract, respectively. Preliminary phytochemical screening of AQRFH revealed the presence of alkaloids, flavonoids, phenolics, saponins, steroids, tannins and terpenoids while the screening of HEXRFH revealed the presence of alkaloids, cardiac glycosides, phenolics, saponins, steroids, tannins and terpenoids ( Table 1).

Acute toxicity studies were carried out for 14 days as per the OECD 423 guidelines ¹⁴ . No mortality of mice was observed within the first 4 hours of continuous observation, nor after 24 hours. There was also no lethal effect observed after the administration of the extracts for the experimental period of 14 days. Morphological characteristics such as the fur, and skin colour appeared normal. There was no salivation, diarrhoea, lethargy, unusual behaviour or altered respiration. Mean body weight ( Extended data: Table S2 ¹⁵ ), food and water intake ( Extended data: Table S3 ¹⁵ ), and relative mean organ weight ( Extended data: Table S4 ¹⁵ ) of treatment group animals were not significantly different from the control group animals. Figure 2 illustrates the changes in the mean body weight of mice receiving different treatments over the 14-day study period for acute oral toxicity. Figure 3 illustrates the feed and water consumption of the experimental mice over the 14-day study period.

On gross macroscopic examination, there was no necrosis, inflammation or change in size of the lungs, liver, spleen, kidney or heart in mice ( Figure 4).

Mean body weight of mice that received AQFRH over 28 days. There was no significant difference in the mean weight of control mice relative to the mice treated with 250 or 500mg/kg doses of AQFRH. However, there was a significant difference in the mean weight of control mice relative to the mice (female) treated with a 1000mg/kg dose of AQFRH after 28 days ( Extended data: Table S5 ¹⁵ ).

Feed and water consumption in mice treated with graded doses of AQFRH and HEXFRH. Feeding mice on graded doses of the root extracts of F. hildebrandtii over a 28-day period had no significant effect on the feed and water consumption of mice in treatment groups relative to the control ( Extended data: Table S6 ¹⁵ ).

Relative mean weight of mice treated with graded doses of AQFRH and HEXFRH. Doses of 250mg/kg and 500mg/kg of AQFRH did not produce any significant effect on the mean weight of mice relative to the controls ( Extended data: Table S7 ¹⁵ ).

None of the doses of HEXFRH had any significant effect on the mean weight of the treatment mice relative to the controls ( Extended data: Table S8 ¹⁵ ).

Relative-organ weight ratios of mice treated with AQFRH and HEXFRH. There was no significant change in the mean relative weight of the stomach, left and right kidneys, lung, spleen, heart and liver in mice treated with a 28-day oral dose of 250mg/kg and 500mg/kg of AQRFH, respectively ( Extended data: Table S9 ¹⁵ ). A 1000mg/kg dose of AQRFH significantly lowered the mean relative weight of the spleen in both male and female mice ( Extended data: Table S9 ¹⁵ ). This dose also significantly lowered the mean relative weight of the liver in male mice, and the stomach and lung in female mice relative to the control group ( Extended data: Table S9 ¹⁵ ).

There was no significant change in the mean relative organ weight of the stomach, left and right kidneys, lung, spleen, heart and liver in mice treated with a 28-day oral dose of 250mg/kg and 500mg/kg of HEXRFH ( Extended data: Table S10 ¹⁵ ). A 1000mg/kg dose of HEXRFH significantly increased the mean relative organ weight of the liver in male mice relative to the control. This dose also produced a significant decrease in the mean relative organ weight of the lungs, spleen and heart in male and female mice, and the stomach of female mice relative to the control group ( Extended data: Table S10 ¹⁵ ).

Evaluation of haematological parameters. A 250 mg/kg dose of AQRFH significantly lowered the mean RBC levels in male mice, while significantly raising the mean platelet levels in both male and female mice relative to the control group ( Table 2). This dose also non-significantly increased the mean levels of Hb, HCT, MCV, MCH, and the MCHC in both male and female mice as well as mean levels of WBC in male mice relative to the control ( Table 2). This dose also non-significantly lowered the mean RBC levels in female mice relative to the controls. A 500mg/kg dose of AQRFH significantly lowered the mean RBC levels in male mice while significantly raising the mean Hb levels in female mice and the mean platelet levels in male mice relative to the control ( Table 2). This dose also non-significantly increased the mean levels of MCHC, MCH and WBC in both male and female mice, and the mean levels of HCT, platelets and MCV in female mice, as well as the mean Hb levels in male mice relative to the controls. A 1000mg/kg dose of AQRFH resulted in the death of all female mice tested (n=6). Thus, no haematological parameter was evaluated for these animals. On the other hand, this dose of extract significantly lowered the mean levels of all haematological parameters evaluated in male mice relative to the controls ( Table 2).

A 250mg/kg dose of HEXREF significantly lowered the mean RBC and platelet levels in male and female mice relative to the controls ( Table 3). This dose also non-significantly lowered the mean haematocrit, WBC and MCHC levels in both male and female mice, as well as the mean Hb, and MCV levels in female mice relative to the control group. This dose also non-significantly raised the mean Hb and MCV levels in male mice, as well as the MCH levels in both male and female mice relative to the control group. A 500mg/kg dose of HEXRFH significantly lowered the mean RBC levels in both male and female mice, and the mean platelet levels in female mice relative to the control group ( Table 3). This dose non-significantly raised the mean Hb, MCHC and WBC levels in male and female mice, as well as the mean levels of Hb, MCH and WBC in male mice relative to the control group. This dose also non-significantly lowered the mean Hb, MCHC and WBC levels in both male and female mice, as well as the MCH levels in male mice relative to the control group. A 1000mg/kg dose of HEXRFH resulted in the death of all female mice tested (n=6). Thus, no haematological parameter was evaluated for these animals. On the other hand, this dose of extract significantly lowered the mean levels of all haematological parameters evaluated in male relative to the controls ( Table 3).

Evaluation of biochemical parameters. A 250 mg/kg dose of AQRFH did not produce any significant change in the mean levels of any of the biochemical parameters evaluated ( Table 4). This dose produced a non-significant decrease in the mean levels of urea, creatinine and ALT in both male and female mice and AST and TP in male mice relative to the control group. This dose also produced a non-significant increase in the mean levels of AST and TP in female mice relative to the control group. A 500 mg/kg dose of AQRFH significantly raised the mean levels of ALT in female mice relative to the controls ( Table 4). This dose produced a non-significant decrease in the mean levels of AST in male and female mice, and in mean levels of ALT, urea, creatinine, and TP in male mice only relative to the controls. This dose also produced a non-significant increase in the mean levels of urea, creatinine and TP in female mice only relative to the controls. A 1000mg/kg dose of the aqueous extract of F. hildebrandtii resulted in the death of all female mice tested (n=6) ( Table 4). Thus, no biochemical parameter was evaluated for these animals. On the other hand, this dose of extract significantly raised the mean levels of all biochemical parameters evaluated in male mice relative to the controls.

A 250 mg/kg dose of HEXRFH produced a significant increase in the mean levels of ALT in female mice ( Table 5). This dose produced a non-significant decrease in the mean levels of urea in male and female mice, creatinine in female mice and AST in male mice relative to the controls. This dose also produced a non-significant increase in the mean levels of TP in male and female mice, creatinine and ALT in males, and AST in female mice relative to the controls. A 500 mg/kg dose of HEXRFH significantly raised the mean levels of ALT in both male and female mice relative to the controls ( Table 5). This dose produced a non-significant decrease in the mean levels of urea and AST in male mice relative to the controls, and in mean levels of ALT, urea, creatinine, and TP in male mice only relative to the controls. This dose also produced a non-significant increase in the mean levels of urea, creatinine and TP in female mice only relative to the controls. A 1000mg/kg dose of HEXRFH resulted in the death of all female mice tested (n=6) ( Table 5). Thus, no biochemical parameters were evaluated for these animals. On the other hand, this dose of extract significantly raised the mean levels of all biochemical parameters evaluated in male mice relative to the controls.

Histopathological studies on kidney and liver sections of mice receiving AQFRH and HEXFRH over 28 days. In Figure 5, kidney sections A, B, C were characterized by regular glomeruli and normal renal tubules with complete smooth epithelia and clear lumen (X400). Kidney section D (×400) was characterized by degeneration of renal tubules and loss of epithelium. Kidney section E (×400) was characterized by mild tubular degeneration. Kidney section F (×400) was characterized by multifocal tubular degeneration with loss of epithelium. Kidney section G (×400) was characterized by degeneration of tubules and haemorrhage.

In Figure 6, liver sections A, B, C (×400) were characterized by normal parenchymal architecture with normal hepatic cells that were evenly distributed and separated by sinusoids. Liver sections D and E (×400) were characterized by diffuse cloudy swelling of the hepatocytes indicative of early necrosis with pyknotic nucleus. Liver section F (×400) was characterized by multifocal hepatocyte necrosis with hepatocytes undergoing cloudy swelling with nuclear pyknosis. Liver section G (×400) was characterized by hepatocyte necrosis and pyknosis, cloudy swelling, necrotic hepatocyte and regenerating hepatocyte around the necrotic hepatocyte.

Minimum inhibitory concentration of root extracts of F. hildebrandtii against Staph. aureus, Salmonella typhimurium and C. albicans. C. albicans was not sensitive to any of the concentrations of AQRFH and HEXRFH. Staph. aureus was sensitive to a concentration of 100 mg/ml, 200mg/ml and 400 mg/ml of both AQRFH and HEXRFH. Salmonella typhimurium was sensitive to concentrations of 200 mg/ml and 400mg/ml of AQRFH and a concentration of 100mg/ml, 200mg/ml and 400mg/ml of HEXRFH ( Table 6).

Minimum bactericidal/fungicidal concentration (MBC/MFC) of the root extracts of F. hildebrandtii against S. aureus, S. typhimurium and C. albicans. C. albicans was not sensitive to any of the concentrations of either AQRFH or HEXRFH. Staph. aureus was sensitive to a concentration of 200mg/ml and 400 mg/ml of both AQRFH or HEXRFH. Salmonella typhimurium was sensitive to a concentration of 400mg/ml only of the AQRFH and a concentration of 200mg/ml and 400mg/ml of HEXRFH ( Table 7). Based on the findings of MBC and MIC of the extracts against the tested pathogens, the MBC/MIC ratio of the extracts was found to be 400/200=2; for AQFRH against Salmonella typhi), 200/100=2; for HEXFRH against Salmonella typhi), 200/100=2; for AQFRH against Staph. aureus), 200/100=2; for HEXFRH against Staph. aureus)

Figshare: Underlying data on the study titled 'Acute and sub-acute toxicity study of the root extracts of Fagaropsis hildebrandtii in mice and evaluation of their antimicrobial effects', https://doi.org/10.6084/m9.figshare.9104894.v1 ³⁸ ; https://doi.org/10.6084/m9.figshare.9101231.v2 ³⁹ ; https://doi.org/10.6084/m9.figshare.9104480.v3 ⁴⁰

This project contains the following underlying data:

Figshare: Photomicrographs of kidney and liver sections of mice treated with distilled water, extra virgin oil, and graded doses of aqueous and hexane root extracts of Fagaropsis hildebrandtii, https://doi.org/10.6084/m9.figshare.9199406.v2 ¹⁵

Figshare: Extended data on the study titled 'Acute and sub-acute toxicity study of the root extracts of Fagaropsis hildebrandtii in mice and evaluation of their antimicrobial effects', https://doi.org/10.6084/m9.figshare.9083348.v1 ⁴¹ .

This project contains the following extended data:

Data are available under the terms of the Creative Commons Zero “No rights reserved” data waiver (CC0 1.0 Public domain dedication).