Plant vs. Kidney: Evaluating Nephrotoxicity of Botanicals with the Latest Toxicological Tools

Abstract

Botanicals can cause nephrotoxicity via numerous mechanisms, including disrupting renal blood flow, damaging compartments along the nephron, and obstructing urinary flow. While uncommon, there are various reports of botanical-induced nephrotoxicity in the literature, such as from aristolochia (Aristolochia spp.) and rhubarb (Rheum spp.). However, at present, it is a challenge to assess the toxic potential of botanicals because their chemical composition is variable due to factors such as growing conditions and extraction techniques. Therefore, selecting a single representative sample for an in vivo study is difficult. Given the increasing use of botanicals as dietary supplements and herbal medicine, new approach methodologies (NAMs) are needed to evaluate the potential for renal toxicity to ensure public safety. Such approaches include in vitro models that use layers of physiological complexity to emulate the in vivo microenvironment, enhance the functional viability and differentiation of cell cultures, and improve sensitivity to nephrotoxic insults. Furthermore, computational tools such as physiologically based pharmacokinetic (PBPK) modeling can add confidence to these tools by simulating absorption, distribution, metabolism, and excretion. The development and implementation of NAMs for renal toxicity testing will allow specific mechanistic data to be generated, leading to a better understanding of the nephrotoxic potential of botanicals.

1. Introduction

Herbal medicines and botanical supplements are taken around the world, for both disease prevention and treatment, with U.S. sales of these products estimated at $11.2 billion in 2020 (a 17.3% increase from 2019) [1]. Botanicals – which include plant-derived products, algae, and fungi – are composed of hundreds to thousands of chemical constituents [2]. Products derived from botanicals can vary in chemical composition due to multiple factors, including growing conditions, manufacturing and processing methods, and combination with other botanicals. Because of their chemical complexity and variability, botanicals are difficult to characterize and assess for toxicity. Most global regulatory structures do not require toxicity testing for botanicals, typically relying on a history of safe use. Often, evidence of harm post-market, usually in the form of adverse event reporting, is required to induce regulatory action.

Use of botanicals can lead to various unintended toxicological effects, underscoring that having access to safe botanical products is a critical public health need [2]. In extreme cases, permanent organ damage and even death can occur. Multiple reviews highlight known or suspected nephrotoxic botanicals [3–6], although causality is not typically well established. However, most botanical products have not been specifically evaluated for nephrotoxicity. More research is needed to decipher the extent and mechanisms of toxicity from known and potentially nephrotoxic botanicals. Given their chemical complexity and variability, selecting a single representative botanical sample for traditional rodent in vivo testing can be difficult, and testing multiple samples is resource intensive. Thus, predictive techniques suitable for whole complex mixtures are needed to screen for botanical nephrotoxicity that are relatively rapid, effective, and resource (cost, time, animals) appropriate.

This review explores botanical induced renal injuries and introduces new approach methodologies (NAMs) that could be used to identify the nephrotoxic potential of botanicals. Although there is substantial literature on renal toxicity resulting from contaminants in botanical products, such as heavy metals [5,7], this paper will focus solely on the botanicals themselves. This paper is not intended to be a systematic review of botanical-induced nephrotoxicity but rather provides examples of nephropathies caused by botanicals that are documented via clinical studies, animal data, mechanistic information, or a combination thereof. It should also be noted that literature on botanicals is itself varied, and often does not include detailed constituent information. We intend for this review to stimulate additional work to develop and adapt tools to evaluate nephrotoxicity, specifically for complex mixtures like botanicals.

2. Renal vulnerability to botanical-induced toxicity

Kidneys perform functions that are essential in maintaining the body’s internal environment in a healthy balance. This includes the metabolism and excretion of xenobiotics [8]. However, kidneys are innately vulnerable to xenobiotic-induced injury, due in large part to having a high rate of toxicant delivery. Kidneys receive a significant proportion of cardiac output (renal blood flow is 20–25% of cardiac output), which is supplied by an extensive microvascular network [9]. Following delivery, renal tubular epithelial cells are exposed to xenobiotics in both the blood and glomerular filtrate. Tubular cells have a high metabolic rate which powers extensive transport processes that promote intracellular accumulation of toxicants [10]. The nephrotoxic potential of a botanical constituent is dependent on its structure, solubility, charge, dose and duration, metabolism and excretion processes, and the interaction of these factors with intrinsic patient and kidney characteristics.

3. Botanical-induced nephropathies

Case reports of botanical-induced nephropathies are uncommon, and the risk of damage to the kidneys is presumed low for most marketed herbal dietary supplements. As noted, however, most botanicals are not evaluated for their nephrotoxic potential. Despite that, some botanicals have been shown to impact kidney structures and functions. Damage can be provoked in multiple ways (Figure 1), such as by disrupting renal blood flow (pre-renal disease), inflicting injury to various compartments along the nephron (intrinsic renal disease), and obstructing urinary flow within the urogenital system (post-renal disease) [11]. Table 1 presents specific cases in the reported literature relevant to these injury types that have been induced by botanicals and Table 2 provides appropriate screening models to identify these toxicities. It should be noted that the botanicals listed are examples, however, evidence types vary, and botanicals cited are not often chemically characterized or authenticated. When possible, we noted doses and exposure durations in Table 1. The lack of high-quality botanical nephrotoxicity data emphasizes the need for additional tools to screen for toxicity.

Figure 1:

Botanicals can provoke kidney injury in multiple ways, including by 1) disrupting renal blood flow (pre-renal disease), 2) damaging various compartments along the nephron (intrinsic renal disease), and 3) obstructing urinary flow within the urogenital system (post-renal disease).

Table 1:

Types of nephrotoxicity and botanicals that have been reported to induce nephrotoxicity.

Type	Compartment	Phenotype/Pathology	Botanical	Evidence type(s)	Details(s) (e.g., dose, preparation, usage)
Pre-renal disease	Hemodynamic disturbance	Intravascular volume depletion	Lemongrass (Cymbopogon citratus)	Human data [14]	Lower creatinine clearance rates and GFR in 105 subjects consuming lemongrass tea made from 2, 4 or 8 g of powdered lemongrass leaves for 30 days.
		Vasoconstriction/vasodilation	Licorice (Glycyrrhiza glabra)	Case report [16,60] Mechanistic data [61]	Case report of 39 year old female taking 50–100 g herbal products which contained licorice every day for 8 weeks. Licorice induced hypokalemia. Other sources support mechanisms of vasoconstriction. Glycyrrhizic acid in licorice induces lower potassium levels due to increasing levels of cortisol.
			Mexican prickly poppy (Argemone mexicana)	Human data [62] Animal data [63]	In humans, epidemic dropsy associated with argemone oil contaminated mustard oil was characterized by edema in the lower limbs and featured vascular dilation leading to renal hypoperfusion and acute tubular necrosis. In rats given 0.1% argemone oil, glomerular and vascular congestion was observed; tubules displayed swelling of epithelium and occlusion of lumen.
Intrinsic renal disease	Glomerulus	Glomerular disease	Yellow oleander (Cerbera thevetia)	Case report [22] Animal data [64]	4 Case reports of yellow oleander poisoning with jaundice and renal failure. Study in rats reported proliferation of glomerular endothelium, hypercellularity of the glomerulus, necrosis of convoluted tubular epithelium, disappearance of nuclei and pyknosis. However, these were at doses that killed 16/20 and 18/20 of rats within 10 days.
	Tubules	Acute tubular necrosis (ATN)	Aristolochia (Aristolochia spp.)	Epidemiological survey, animal data, case-control, cohort study [3]	Well-documented for nephrotoxicity.
			Impila (Callilepis laureola)	Case reports [65,66] retrospective study [3,14]	Three case reports (dosage not provided) and a retrospective study in 50 children. Nephrotoxicity due to the constituent atractyloside.
			Himalayan yew (Taxus wallichiana)	Case reports [23]	Two case reports describing ingesting “extractions” from 120 g and 150 g of Himalayan yew.
			Felty germander (Teucrium polium)	Case report [3] Animal data [24]	Case report of a 46-year-old diabetic woman hospitalized with AKI who drank a tea made with a “pinch” of felty germander for an unknown duration. Kidney toxicity is also documented in several animal studies.
			Wild yam (Dioscorea quartiniana)	Animal data [67] Case studies [68–70]	Renal tubulo-interstitial fibrosis in rats following 28-day treatment with 0.79 g/kg Dioscorea villosa extract. Multiple cases of acute renal failure (likely acute interstitial nephritis).
			Trypterygium (Tripterygium wilfordii)	Clinical studies, mechanistic data [14,71]	2018 systematic review of clinical studies showed 5.8% evidence of nephrotoxicity. Many clinical studies do not report concentrations of triptolide. Trypterygium is used to treat kidney disease (immunosuppressive & antiproteinuric; active constituent (triptolide). Mechanism of action has been well-studied -- can disturb Ca2+ channels, lead to oxidative stress and kidney damage, particularly via alteration of permeability in the proximal tubule.
		Proximal tubule (PT) dysfunction and/or Fanconi syndrome	Aristolochia (Aristolochia spp.)	Case report [31]	60-year-old consumed Aristolochia for 5 months. Laboratory investigation showed hypokalemia (1.8 mEq/L), hyperchloremic metabolic acidosis (Cl−, 111 mEq/L, and HCO3−, 14.0 mEq/L), hypophosphatemia (0.9 mg/dL) with hyperphosphaturia, hypouricemia (1.3 mg/dL) with hyperuricosuria, and glycosuria, consistent with Fanconi’s syndrome.
			Rhubarb (Rheum spp.)	Animal Data [32]	Anthraquinones in rhubarb are responsible for observed renal proximal tubule cell swelling and degeneration in rats treated with 0.165, 0.33, or 0.66 mg/kg rhubarb extract.
		Distal tubule (DT) dysfunction	Limited clinical reports of botanicals inducing this injury type
	Interstitium	Acute tubulointerstitial nephritis (ATIN)	Aristolochia (Aristolochia spp.)	Animal Data [34] Case report [72]	Salt-depleted rats received daily subcutaneous injections of 1 mg/kg aristolochic acid (AA) or 10 mg/kg AA for 34 days. High-dose AA induced glucosuria, proteinuria, and elevated serum creatinine levels and reduced leucine aminopeptidase enzymuria. Tubular necrosis associated with lymphocytic infiltrates (day 10) and tubular atrophy surrounded by interstitial fibrosis (day 35) were the histologic findings for the high-dose AA-treated rats. In both AA groups, urothelial dysplasia was also observed, as well as fibrohistiocytic sarcoma at the injection site. Case report on a 10-year-old boy who presented with severe anemia, Fanconi’s syndrome, and progressive renal failure. Renal biopsy revealed typical findings of aristolochic acid nephropathy (AAN). Aristolochic acids I and II were identified from a Chinese herb mixture ingested by the boy. AAN was diagnosed after other etiologies had been excluded.
			Himalayan yew (Taxus wallichiana)	Case reports [23]	Case reports describing eight patients on constituent, sciadopitysin. Renal biopsy was performed and showed acute tubulointerstitial nephritis with acute tubular necrosis in one patient.
			Impila (Callilepis laureola)	Case reports [65,66] retrospective study [3,14]	Three case reports (dosage not provided) and a retrospective study in 50 children. Nephrotoxicity due to the constituent, atractyloside.
		Chronic tubulointerstitial nephritis (CTIN)	Aristolochia (Aristolochia spp.)	See other Aristolochia entries
			Lemongrass (Cymbopogon citratus)	Clinical trial [36]	Decrease in glomerular filtration rate in men and women taking 2, 4, or 8 g lemongrass for 30 days.
Post-renal disease	Obstructive nephropathy (This can also affect the nephron)	Nephrolithiasis (kidney stones)	Ephedra, ma huang (Ephedra sinica)	Case reports [38,73]	40–3,000 mg/day over several months
			Star fruit (Averrhoa carambola)	Case reports [74]	Two case reports with patients drinking 1.6 L and 3 L, respectively, of star fruit juice corresponding to 13.1 g and 9.2 g, respectively of oxalate. The lethal dose of soluble oxalate for humans varies from 2 to 30 g.
			Rhubarb (Rheum spp.)	Case report [39]	Secondary oxalate nephropathy - Kidney biopsy showed multiple birefringent crystallinous casts were found within the tubular lumen and serum oxalate levels were elevated following 500 mg of rhubarb (fresh weight) per day in the last 4 weeks (patient had pre-existing diabetes).
	Carcinoma	Renal cell carcinoma	Aristolochia (Aristolochia spp.)	Well documented in humans, animal data, and mechanistic data [75,76]

Note: Although many other botanicals have been reported to cause nephrotoxicity, this table provides an overview of a few prototypical examples of each type that are well-documented via clinical studies, animal data, mechanistic information, or a combination thereof.

AKI = acute kidney injury, ATN = acute tubular necrosis, ATIN = acute tubulointerstitial nephritis, CTIN = chronic tubulointerstitial nephritis, AA = aristolochic acid, AAN = aristolochic acid nephropathy, GFR = glomerular filtration rate

Table 2:

Botanicals can injure structures and functions of the entire nephron via various mechanisms. New approach methodologies (NAMs) are in vitro and in silico models used in combination with assays and diagnostic endpoints to evaluate the potential for xenobiotic-induced toxicity [58]. The models highlighted here are complementary and compatible with many different assays. These include comprehensive high-throughput molecular screening methods (transcriptomics, metabolomics, and proteomics), high-content imaging, and biomarker analysis, as well as specific phenotypic assays that provide insights into toxicological mechanisms and modes of action.

		Pre-renal disease		Intrinsic renal disease						Post-renal disease
		Hemodynamic disturbance		Glomerulus	Tubules			Interstitium		Obstructive nephropathy	Carcinoma
New Approach Methodologies (NAMs)		Intravascular volume depletion	Vasoconstriction/vasodilation	Glomerular disease	Acute tubular necrosis (ATN)	Proximal tubule (PT) dysfunction and/or Fanconi syndrome	Distal tubule (DT) dysfunction	Acute tubulointerstitial nephritis (ATIN)	Chronic tubulointerstitial nephritis (CTIN)	Nephrolithiasis (kidney stones)	Renal cell carcinoma
				Examples of botanicals linked to each nephropathy
In vitro models for toxicity screening		Lemongrass (Cymbopogon citratus)	Licorice (Glycyrrhiza glabra), Mexican prickly poppy (Argemone mexicana)	Yellow oleander (Cerbera thevetia)	Aristolochia (Aristolochia spp.), Impila (Callilepis laureola), Himalayan yew (Taxus wallichiana), Felty germander (Teucrium polium),	Aristolochia (Aristolochia spp.), Rhubarb (Rheum spp.)	Limited clinical reports of botanicals inducing this injury type	Aristolochia (Aristolochia spp.), Himalayan yew (Taxus wallichiana), Impila (Callilepis laureola),	Aristolochia (Aristolochia spp.), Lemongrass (Cymbopogon citratus)	Ephedra (Ephedra sinica), Star fruit (Averrhoa carambola), Rhubarb (Rheum spp.)	Aristolochia (Aristolochia spp.)
2D cultures	Immortalized cell lines			X [44]	X [44]	X [44]				X [77]	X [78]
	Primary cells			X [44]	X [45]	X [45]	X [79]			X [77]	X [78]
	iPSCs and ESCs			X [46]	X [46]	X [46]
3D cultures (and co-cultures)	Supporting scaffold		X [80]	X [81]	X [45]	X [45]		X [82]	X [82]
	Spheroids			X [83]	X [42]	X [42]
	Organoids			X [55–57]	X [55–57]	X [55–57]	X [55–57]				X [84]
	Flow based microphysiological devices	X [85]	X [86]	X [44,50]	X [44,48]	X [44,48]	X [44,52]	X [87]	X [87]	X [52]
Computational approaches	Physiologically Based Pharmacokinetic (PBPK) modeling	PBPK modeling establishes the relationship between external and internal doses to query exposure scenarios of toxicological concern [54]. Thus, toxicologically driven diseases are dependent on a given exposure scenario (e.g., no effect level [NOEL], point of departure [POD], lowest observable adverse effect level [LOAEL].

Note: The cited references provide examples of NAMs that have been developed or used to model different nephron segments and/or nephropathies. However, these examples are not exhaustive, and many different models have been – and continue to be – developed.

3.1. Pre-renal disease

Acute kidney injury (AKI) – defined as an abrupt decline in kidney function – can be caused by hemodynamic disturbances that reduce renal blood flow and the glomerular filtration rate (GFR) [12]. Renal hypoperfusion decreases the delivery of oxygen and nutrients to nephrons, which can damage tubular epithelial cells and promotes the accumulation of nitrogenous waste products [13]. This can be caused by true intravascular volume depletion provoked by severe incidents of diarrhea, vomiting, sweating, or bleeding [11]. The nephrotoxic effects of several botanicals, such as lemongrass (Cymbopogon citratus), are associated with volume depletion [14]. Volume depletion can also promote injury by increasing toxicant concentration, due to excessive dosing. Furthermore, sluggish flow rates of insoluble compounds can lead to crystal precipitation and disposition in tubular lumens [15].

Changes in blood pressure can also disrupt renal perfusion. Various botanicals can have a vasoconstrictive or vasodilative effect, including licorice (Glycyrrhiza glabra) indirectly via hypokalemia [16] and Mexican prickly poppy (Argemone mexicana) [17], respectively. Vasoactivity can be augmented in a scenario of volume depletion [4]. In order to maintain renal filtration pressure when a decrease in systemic pressure is caused by volume depletion, the afferent arteriole vasodilates and efferent arteriole vasoconstricts [18]. Thus, prolonged vasoconstriction of the afferent arteriole can reduce renal blood flow and cause AKI [11]. Similarly, efferent arteriole vasodilation combined with reduced systemic blood pressure can decrease renal perfusion and provoke AKI [19].

3.2. Intrinsic renal disease

The nephron is the functional unit of the kidney. Some botanicals can injure structures and functions of the nephron, including those associated with the glomerulus, tubules, and interstitium. Xenobiotics can cause glomerular disease by damaging mesangial, endothelial, and epithelial cells, and by initiating an immune mediated injury [20]. Glomerular disease can decrease the GFR, and provoke proteinuria and hematuria [21], as incited by yellow oleander (Cerbera thevetia) [3,22].

Acute tubular necrosis (ATN) is a common cause of AKI. Cases of botanical-induced ATN have been reported following exposure to Himalayan yew (Taxus wallichiana) [23] and felty germander (Teucrium polium) [24]. ATN can be caused by direct toxicity of a xenobiotic to tubular epithelial cells. This can happen via various mechanisms, including DNA damage, inhibited mitochondrial function, disturbed cellular and lysosomal membranes, destabilized ion gradients, and oxidative stress [25]. ATN can disrupt the transport of solutes that are reabsorbed by tubular epithelial cells, leading to inadequate reabsorption of clinically important filtered substances [26]. Histological findings of ATN include tubular cell deterioration characterized by brush border loss, irregular Na⁺/K⁺-ATPase localization, vacuolization, nuclear abnormalities, and cellular necrosis and apoptosis [27]. ATN frequently provokes inflammatory responses within surrounding tissues.

Adult human kidneys filter ~180 liters of fluid every day, of which ~98% is reclaimed before being excreted [28]. Proximal tubule (PT) cells perform the vast majority of this reabsorption and are reliant on mitochondria to generate the large amounts of ATP needed to power solute transport, making them sensitive to disrupted oxidative phosphorylation [29]. Due to these extensive excretory and resorptive transport capabilities, substances can accumulate within PT cells, making them frequent targets of xenobiotic-induced injury [26]. Toxic insults that inhibit PT cell function can disrupt the movement of solutes that are normally reabsorbed by PT cells [26], such as phosphate, glucose, amino acids, and urate [30]. PT dysfunction is associated with a spectrum of clinical presentations dependent on insult severity. Mild insults typically cause isolated transport defects, while more severe insults provoke a partial or global breakdown of transport function. This clinical scenario is known as Fanconi syndrome [26]. Xenobiotic-induced causes of PT dysfunction include impaired endocytosis, receptor/transporter inhibition, lysosome disruption, oxidative stress and depletion of antioxidant defenses, and mitochondrial dysfunction [26]. Botanicals associated with PT dysfunction include aristolochia (Aristolochia spp.) [31] and rhubarb (Rheum spp.) [32].

Less is known about botanical-induced nephrotoxicity in the loop of Henle, distal tubule, and collecting duct. While there are some limited clinical reports of distal renal tubular acidosis following botanical exposure, additional mechanistic evidence is needed to provide specifics on this type of kidney injury. This again points to the need for tools that could be used to screen for renal toxicity.

Acute tubulointerstitial nephritis (ATIN) is a type of AKI caused by a non-dose-dependent toxin-induced allergic reaction which promotes interstitial lesions and inflammation, and tubular cell damage [33]. Botanicals associated with ATIN include aristolochia [34] and Himalayan yew [23]. Disease within the interstitial compartment can develop from either an acute or chronic insult. Chronic tubulointerstitial nephritis (CTIN) follows persistent unrecognized and/or untreated ATIN [35]. Botanicals linked with CTIN include aristolochia and lemongrass (Cymbopogon citrullus) [35,36].

3.3. Post-renal disease

Post-renal disease occurs when urinary flow is obstructed within the urinary collection system, from the renal tubule to the urethra. This increases the intra-tubular pressure and decreases the GFR, which can lead to AKI, a buildup of nitrogenous waste products, extreme pain, and infection [37]. Xenobiotics promote obstructive nephropathy when poorly soluble compounds precipitate in renal tubules and when metabolic disturbances alter urinary pH and the excretion of calcium and other purines, leading to nephrolithiasis (i.e., kidney stone formation). In addition to obstructing urinary flow, kidney stones can disturb surrounding tissues. Most patients form stones composed of calcium oxalate or calcium phosphate [37]. Botanicals associated with the development of calcium caliculi include ephedra (Ephedra sinica) [38] and rhubarb [39].

Botanical-induced cancers can also affect the urinary tract. Aristolochic acids (Aristolochia spp.) are associated with a well-established syndrome of kidney injury, including AKI, CKD, tubulopathies, and urothelial carcinomas [3]. Upper urinary tract cancers have been reported in over 40% of patients consuming aristolochic acid, which are nephrotoxic constituents occurring in Aristolochia spp. [40].

3.4. Chronic kidney disease

Most nephropathies – including those described in the previous paragraphs – can progress to chronic kidney disease (CKD) [41]. The loss of kidney function associated with CKD can disrupt fluid homeostasis and cause a dangerous accumulation of waste products. In addition, CKD can lead to other significant health problems, such as heart disease. Underlying AKI and CKD increases the likelihood of developing xenobiotic-induced injury. This is caused by excessive dosing, decreased number of functioning nephrons, ischemic tubular cells, and increased oxidative stress.

4. New approach methodologies for botanical-induced nephrotoxicity screening

NAMs include in vitro and in silico systems used in combination with assays and diagnostic endpoints, with translational relevance for a given context of use. In Table 1, we have collated the evidence for botanical-induced nephropathies. In Table 2, we identify fit-for-purpose NAMs that could be used to screen for and investigate the mechanisms underlying these toxicities. These tools have the potential to make botanical products safer for consumers by improving the efficacy of nephrotoxicity screening assessments. In addition to botanicals, NAMs are already being applied to other chemical types, including industrial chemicals and pharmaceuticals.

Historically, tools for nephrotoxicity assessment have consisted mostly of 2D monocultures of undifferentiated human- or animal-derived renal epithelial cells and mammalian animal models. However, these systems often fail to predict nephrotoxicity due to a limited ability to recapitulate fundamental aspects of human kidney structure and function [42,43]. In recent years, significant efforts have focused on exploring the potential of human-derived cells and emerging technologies to enhance the efficacy of nephrotoxicity safety assessments. Immortalized human kidney cells represent a convenient tool for nephrotoxicity assessment. However, when cultured in 2D static conditions, these cells can lack expression and function of key renal transport proteins, biomarkers, and enzymes [44]. In comparison, freshly isolated human primary cells have enhanced genetic and phenotypic preservation [45]. New technologies use layers of physiological complexity to enhance the functional viability and differentiation of cell cultures. By more accurately emulating the in vivo milieu, these approaches increase the physiological relevance of cells in vitro, and improve sensitivity to nephrotoxic insults [46].

A promising technology in this space are kidney microphysiological systems (MPS). MPS integrate microfluidics with human-derived cells and tissue engineering to create 3D microenvironments that mimic the dynamics, functions, and pathologies experienced by specific nephron segments. Biologically relevant flow rates promote barrier function, appropriate cell morphology, and the expression and activity of key transporters [47,48]. In addition to providing fluid flow, MPS have enabled renal epithelial cells to be co-cultured alongside microvascular endothelial cells, which has been shown to further enhance cell differentiation. PT cells co-cultured with kidney-specific microvascular endothelial cells were found to have elevated expression of key renal transporters, when compared to PT monocultures or co-cultures of PT cells with umbilical vein endothelial cells [49]. Similarly, a MPS of the glomerular filtration barrier – developed by co-culturing podocytes with glomerular endothelial cells – achieved differential albumin and inulin clearance, and was able to accurately model pathophysiological responses to a toxic agent [50]. Additional layers of physiological complexity are being developed to further recapitulate the in vivo kidney environment. This includes linking MPS representing different organ systems. Interestingly, the mechanisms underlying the nephrotoxic effects of aristolochic acids were first identified by connecting PT and liver MPS, as these compounds must first undergo hepatic bioactivation before causing kidney injury [51]. In addition to PT and glomerular MPS, a distal tubule MPS has been developed to study kidney stone formation [52]. This system could be used to examine botanical-induced nephrolithiasis.

Physiologically Based Pharmacokinetic (PBPK) modeling is a computational approach that uses mathematical equations to predict in vivo concentrations by simulating the processes of compound absorption, distribution, metabolism, and excretion [47]. To be successful, this method requires accurate in vitro approximations of pharmacokinetic parameters, including transporter mediated drug disposition and drug-drug interactions [53]. Data generated by kidney MPS combined with PBPK modeling has accurately reproduced the kinetics of renal reabsorption and excretion observed in vivo for numerous nephrotoxic compounds [54]. This approach could be used to predict nephrotoxicity and optimize dosing strategies for novel botanicals prior to human exposure. For botanicals, PBPK models will typically be based on individual marker constituents. Dose information can be estimated from sources such as pharmacopoeia monographs, chemical markers if supplied by the manufacturer, or from toxicokinetic studies based on individual constituents. These data can help estimate parameters for the mixture.

As discussed, traditional 2D monoculture systems often overlook important toxicological information. Kidney organoids overcome the simplicity of 2D cultures and represent a promising tool with which to screen for botanical efficacy and safety [47]. Aggregations of human pluripotent stem cells directed towards a renal fate self-organize to form 3D arrangements with characteristic features of the entire nephron and related structures [47]. Multicellular kidney organoids can develop a glomerulus, proximal and distal tubule, alongside a collecting duct, interstitium, and vasculature. Encouragingly, organoids have been able to reproduce human phenotypes provoked by known nephrotoxic agents [55,56], including from the botanical tripterygium (Tripterygium wilfordii) [57]. However, at present, there are some limitations for nephrotoxicity assessment. For example, organoids have been more representative of the early developing nephron rather than a mature fully-functional nephron. In addition, organoids have lacked a functional vascular system [47].

While challenges exist, MPS and organoids demonstrate great promise for nephrotoxicity screening by more closely emulating human renal physiology, xenobiotic metabolism and transport, and pathophysiological responses [44,58]. The extended functional viability afforded by these systems could support modeling of nephropathies induced by prolonged botanical exposure. Furthermore, MPS and organoids are becoming increasingly compatible with comprehensive high-throughput non-targeted molecular screening methods, such as transcriptomics, metabolomics, proteomics, and high-content imaging. These techniques enable sensitive and robust automated multidimensional phenotyping to identify the molecular events underlying disease [57,59].

5. Conclusions

Botanical-induced nephropathies can occur via multiple mechanisms, and given the increasing use of botanicals (e.g., dietary supplements), more research is needed to develop and validate tools that can be used for nephrotoxicity screening. Although there are numerous case reports and reviews on botanical-induced nephrotoxicity in the literature, many rely on a very small number of clinical reports that often lack ingredient identification, supporting information, or mechanistic investigation. The development of NAMs will enable more specific mechanistic data and improve safety, thereby limiting the potential for people to develop nephrotoxicity from botanicals.