Mechanisms of Herbal Nephroprotection in diabetes mellitus

Diabetic nephropathy (DN) is a leading cause of kidney morbidity. Despite the multilayered complexity of the mechanisms involved in the pathogenesis of DN, the conventional treatment is limited to just a few drug classes fraught with the risk of adverse events, including the progression of renal dysfunction. Phytoceuticals offer a promising alternative as they act on the many-sidedness of DN pathophysiology, multitargeting its intricacies. This paper offers a review of the mechanisms underlying the protective action of these phytoagents, including boosting the antioxidant capabilities, suppression of inflammation, averting the proliferative and sclerosing/fibrosing events. The pathogenesis of DN is viewed as a continuum going from the original offense, high glucose, through the noxious products it generates (advanced glycation end-products, products of oxidative and nitrosative stress) and the signaling chains consequently brought into action, to the harmful mediators of inflammation, sclerosis, and proliferation that eventually lead to DN, despite the countervailing attempts of the protective mechanisms. Special attention was given to the various pathways involved, pointing out the ability of the phytoagents to hinder the deleterious ones (especially those leading to, driven by, or associated with TGF-β activation, SREBP, Smad, MAPK, PKC, NF-κB, NLRP3 inflammasome, and caspase), to promote the protective ones (PPAR-α, PPAR-γ, EP4/Gs/AC/cAMP, Nrf2, AMPK, and SIRT1), and to favorably modulate those with potentially dual effect (PI3K/Akt). Many phytomedicines have emerged as potentially useful out of in vitro and in vivo studies, but the scarcity of human trials seriously undermines their usage in the current clinical practice—an issue that stringently needs to be addressed.


Introduction
In most countries, diabetic nephropathy (DN) (also known as diabetic kidney disease) is the main cause of chronic kidney disease (CKD) [1]. DN results from the interplay of several distinct but highly interconnected high glucose-(HG-) induced pathways set into motion by aggressive factors, such as oxidative stress [2] and advanced glycation end-products (AGEs), which trigger signaling chains that generate mediators able to instigate reactive processes, including inflammation, cellular proliferation, and interstitial matrix expansion [3]. Oxidative stress and inflammation enhance each other, resulting in a vicious circle leading to glomerular sclerosis and interstitial fibrosis [4]. Figure 1 illustrates some of the most important mechanisms of DN, although it is by no means exhaustive.
Among the inflammatory mediators involved in DN are nuclear factor kappa-B (NF-κB), monocyte chemotactic protein-(MCP-) 1, and intercellular adhesion molecule-(ICAM-) 1. Attracted and activated by MCP-1 and helped by ICAM-1 (promoted by NF-κB), circulating monocytes invade the kidney [5]. Together with resident mesangial cells proliferation, this leads to the mesangial hypercellularity characteristic for the diabetic glomerulus [6]. Glomerulosclerosis is the hallmark of DN [7] and consists in proteins of extracellular matrix (ECM) (mostly collagen types I, III, and IV and fibronectin [8]) gradually and inexorably encumbering the mesangium, either by lumping together in nodular lesions or by diffusely invading and expanding the interstitial space separating the glomerular loops [9]. The ECM proteins accumulation is the result of excessive production (mostly by mesangial cells) coupled with insufficient proteolysis by mesangial matrix metalloproteinases (MMPs) [10]. However, the pathological deposition of ECM proteins is not limited to the glomerular interstitium but also involves the tubular interstitium and the glomerular basement membrane, explaining its increased thickness [8]. The cellular component of both glomerulus and tubulointerstitium is altered by epithelial-mesenchymal transdifferentiation (EMT) [4]. Growth-promoting pathways involved in DN are those driven by p38 mitogen-activated protein kinase (p38MAPK) [11], mammalian target of rapamycin (mTOR), and phosphatidylinositol 3-kinase (PI3K)/Akt/glycogen synthase kinase-(GSK-) 3β, the last two sometimes acting in concert [12]. It is not clear whether PI3K/Akt activation should be considered protective or deleterious [13]. Endothelial dysfunction [14] and endoplasmic reticulum (ER) stress are other important links in the pathophysiological chain leading to DN [15,16].
Among the pharmaceuticals used for slowing the progression of DN are lipid-lowering agents and renin-angiotensinaldosterone system blockers, especially angiotensin-converting enzyme inhibitors and angiotensin receptor blockers [17], but also mineralocorticoid receptor antagonists [18]. However, the use of these drugs is fraught with the risk of adverse events, including liver and muscle injury, progression of renal dysfunction, and hyperkalemia. Consequently, therapeutic alternatives devoid of such adverse effects are needed. Worldwide herbs have been used for the treatment of diabetes [19], including in the two most practiced systems of traditional medicine: ayurvedic [20] and traditional Chinese medicine [21]. Due to the complexity of their composition, medicinal herbs have, over the modern antidiabetic drugs, the advantage of influencing multiple pathogenic mechanisms [22,23] and being devoid of significant adverse effects [24]. Herbal treatment can complement and even increase the efficiency of the conventional one, the synergistic effect of the combined treatment allowing for dose reduction and delaying the need for insulin therapy [23]. The aim of this article is to review the mechanisms underlying the protective action of these phytomedicines.

Material and Method
A PubMed (https://www.ncbi.nlm.nih.gov/pubmed) search for ("diabetic nephropathy"[Title/Abstract] OR "diabetic kidney disease"[Title/Abstract]) AND (plant[Title/Abstract] OR herbal[Title/Abstract]) and a ScienceDirect (https:// www.sciencedirect.com/) search for ("diabetic nephropathy" OR "diabetic kidney disease") AND (plant OR herbal) provided the articles employed in this review. The authors endeavored to include (almost) all the relevant papers, preferring those attempting to define the molecular/cellular background of the renoprotective action and those investigating clearly defined active compounds or herbal products (articles on herbal formulas with unclear components were left out). The results of this search are summarized in Table 1.
We considered that a study proved that a herbal product is actually nephroprotective if it demonstrated lower levels of glomerular injury markers (in most cases, albuminuria and/or proteinuria) or improved kidney histology or function (stated as such or reflected by lower serum levels of urea and/or creatinine) in the subjects who took the herbal product. We considered the herbal products for which neither of these effects was demonstrated as only potentially nephroprotective. Practically all the in vivo studies (there was only one exception), done on either human or animal subjects (including those that also had an in vitro component), proved the ability of the investigated phytoceuticals to protect the kidney, while none of the exclusively in vitro studies did so. Consequently, for each mechanism or pathway, the data were separated into in vitro, in vivo, and clinical, pointing out that the in vitro studies suggest potentially protective mechanisms, while the in vivo ones demonstrate actually protective mechanisms.
In most of the in vivo studies, nephroprotection consisted in the ability of the phytoceuticals to lessen the degree of glomerular and/or tubulointerstitial injury at the completion of the study, which may be equated with the ability to delay the progression of DN. No study could demonstrate the complete prevention of DN; therefore, none of the plant products was able to prevent the onset of DN.
In order to assess the strength of the evidence in the area of the clinical trials regarding the efficiency of herbals in DN, a secondary search was performed: on PubMed for ("diabetic nephropathy"[Title/Abstract] OR "diabetic kidney disease"[Title/Abstract]) AND (plant[Title/Abstract] OR herbal[Title/Abstract]) AND (trial[Title/Abstract]) and on ScienceDirect for ("diabetic nephropathy" OR "diabetic kidney disease") AND (plant OR herbal) AND (trial). The Jadad scale was employed for evaluating the quality of the human trials [25]. The trials with Jadad score of 1, 2, 3, 4, and 5 (designated as J1, J2, J3, J4, and J5, respectively) were considered of low, low-to-moderate, moderate, moderate-to-high, and high quality, respectively.

Herbal Nephroprotection
3.1. Herbal Products Decreasing Serum Glucose Level and Peripheral Resistance. As HG is the point of depart in the pathophysiological chain leading to DN, optimal glycemic control prevents CKD or at least delays its onset and slows its pace [26]. Insulin resistance may be involved in the genesis of DN [27].

In Vivo Studies-Actually Nephroprotective Mechanisms.
Many of the investigated herbal products are able to decrease glucose level (Allium sativum [28], Artemisia sieberi [29], Bacopa monnieri [30], Hypericum perforatum [31], Punica granatum [32], and Terminalia chebula [33] to name only a few-see Table 2 for a complete list) and some of them are also able to decrease insulin resistance (Cladophora glomerata [34], Panax notoginseng [35], Huangqi decoction [36] etc.-see Table 2) which corroborates or explains their glucose-lowering effect. This is evident especially on the animal models conceived to mimic as good as possible the actual human disease, such as high-fat diet/low-dose streptozotocin type 2 diabetic Wistar albino rats in which some of the most important features of human type 2 diabetes are recognizable: hyperglycemia, insulin resistance, heightened oxidative stress, and structural and functional kidney deterioration. In this animal model, obesity-related insulin resistance, and not lack of insulin, is the cause of altered glucose metabolism. Insulin resistance (that may be attenuated by ellagic acid) is the result of the inflammatory response triggered by adipose tissue infiltration by immune cells spurred by proinflammatory cytokines such as interleukin-(IL-) 1β, IL-6, and tumor necrosis factor-(TNF-) α produced by NF-κB-activated adipocytes [37].

Herbal Products Decreasing Oxidative Stress and AGEs
Production. Oxidative stress is the consequence of antioxidant protective mechanisms being overwhelmed by the reactive oxygen species (ROS) generation. It leads to reactive nitrogen species (such as peroxynitrite) production, lipid peroxidation, inflammatory pathways activation [via NF-κB, protein kinase C (PKC), etc.], apoptosis, and mesangium (both cells and matrix) expansion [4,38]. Acting by means of RAGE (receptor for AGEs), AGEs induce inflammation, oxidative stress, apoptosis, and exuberant ECM protein synthesis [39], one of the mediators being ROS production [8].

In Vitro Studies-Potentially Nephroprotective
Mechanisms. Paeoniflorin and oxypaeoniflora (from Paeonia suffruticosa), improved the protection against AGE-induced inflammatory and oxidative damage by boosting glutathione peroxidase and catalase activities, hampering both the cellular (macrophage migration), and humoral (transcription factors, cytokines) components of inflammation [40]. Curcumin and demethoxycurcumin (from Curcuma longa) prevented AGE-induced apoptosis of mesangial cells [41].
The PI3K/Akt pathway increases insulin sensitivity, averts apoptosis, and is important for recovery after kidney injury, but it may also deleteriously promote cell proliferation and fibrogenesis in DN. A similar action may have mTOR pathway, and these two pathways may act in concert. Inactivating phosphorylation of GSK-3β may be one of the mechanisms mediating the proliferative and fibrogenetic effects of PI3K/Akt activation [12]. Berberine (from Berberis vulgaris) induces PI3K/Akt signaling pathway, which results in the activation of Nrf2 (and its target genes, including heme Table  1: Phytoceuticals effective in preventing diabetic nephropathy in animal models of DN. If not otherwise specified, the indicated dose was the daily dose and the route of administration was oral (generally by gastric gavage). If the strain of animals is not specified, that means that it was not specified in the cited article.   Tyr (14) phosphorylation; ↑ CrCl, stabilization of caveolin-1 and β-catenin [130] 5 Journal of Diabetes Research     HG-(25 mM) cultured mouse podocytes ↓ NOX4, superoxide production, podocyte apoptosis, albuminuria [43] (   10 Journal of Diabetes Research        Cornus officinalis [47], curcumin [108], Fragaria × ananassa [89], myricetin [157] Transforming/tumor growth factor β (TGF-β)

In Vivo Studies-Actually Nephroprotective Mechanisms.
Many studies have pointed out the antioxidant properties of the investigated herbal products (catechin [46], Cornus officinalis [47], dihydroquercetin [48], etc.-see Table 2), translated in their ability to diminish the burden of ROS (quercetin [49], etc.), reactive nitrogen species (tocotrienol [50], etc.), and lipid peroxidation products (breviscapine [51,52], Flammulina velutipes [53], Hibiscus sabdariffa [54], and many more-see Table 2). Beside decreasing malondialdehyde, Artemisia campestris also lowered the level of nitric oxide (NO) and advanced oxidation protein products [55]. Most of these plant-derived products are also able to increase the antioxidant capacity (Anogeissus acuminata [56], Pleurotus eryngii [57], Punica granatum [32], etc.-see Table 2 for a complete list). Suppressing the oxidative stress prevents inflammation and cell death. Therefore, the antioxidant effect maintains cells alive (as silybin does with the HG-damaged podocytes [43]) and decreases inflammation (garlic reduced not only the burden of oxidative stress but also the level of TNF-α [28]). The antioxidant armamentarium of some phytoagents includes the activation of Nrf2, in conjunction with ameliorating mitochondrial dysfunction, dampening the inflammatory response, and bolstering the function of antioxidant enzymes (curcumin [58]) or with suppressing TGF/Smad signaling and the glomerular accumulation of fibronectin and collagen 4 (Hydrangea paniculata [59]). Puerarin (from Pueraria lobata) prevents podocyte foot process effacement and boosts the expression of podocyte slit diaphragm proteins such as nephrin and podocin, beside attenuating oxidative and nitrosative stress, and one of their consequences, the activation of MMP-9 [60].

Human Studies-Actually Nephroprotective Mechanisms.
A study done on human subjects has proved the ability of silymarin to boost the antiproteinuric activity of reninangiotensin system (RAS) inhibitors [65].

Herbal Products with Anti-Inflammatory Activity.
Inflammation is a key pathophysiologic component in the genesis of DN [66][67][68]. The involvement of microinflammation in the pathogenesis of DN justifies the interest for antiinflammatory herbal products in preventing DN-this concept may be correlated with the thousands years old employment of "heat-clearing" herbs for the treatment of DN in Traditional Chinese Medicine [69]. Herbal products may decrease inflammation by multiple ways, such as lowering the level of proinflammatory cytokines (TNF-α, IL-1β, IL-6, IL-12, etc.), decreasing the factors promoting inflammatory cells infiltration (chemokines such as MCP-1 and adhesion molecules including ICAM-1 and vascular cell adhesion molecule-1), and modulating the inflammatory pathways and/or the activity of transcription factors.
Crocin, the carotenoid from Crocus sativus responsible for its saffron color, has also been proven a podocytefriendly substance. It is able to foster the integrity of glomerular filtration barrier, demonstrated by higher levels of slit diaphragms markers: nephrin, podocin, and CD2associated protein. Moreover, it diminishes oxidative stress and the proinflammatory response of the podocytes (by NF-κB inactivation) [73].
Other studies demonstrated, aside from NF-κB inactivation and lower levels of proinflammatory cytokines, a decrease in the expression of chemokines (mainly MCP-1) and of the factors spurring sclerosis/fibrosis (especially TGF-β1), thereby warding off macrophage infiltration (curcumin [79]) and suppressing the production of collagen 4 and fibronectin, and hence ECM expansion and glomerulosclerosis (ellagic acid [37], Shen-Yan-Fang-Shuai Formula [80]). Added to these effects, the inhibition cyclooxygenase-2 and inducible NO synthase boosted the anti-inflammatory potency of green tea polyphenols [81], of Hypericum perforatum [31], and of curcumin analogue B06 (in the latter case probably by means of blocking the JNK (c-Jun N-terminal kinase)/NF-κB signaling [66]). Suppressing the production of vasoactive factors (vascular endothelial growth factor, endothelin-1) resulted in an all-encompassing protective spectrum for North American ginseng, which has been shown to decrease inflammation, fibrosis, and mesangial expansion, to improve the oxidative and metabolic status [82], and to suppress AGEs generation [83]. Other members of the Panax genus have also been shown to prevent diabetic glomerular lesions (P. notoginseng [35]). The aptitude to block both inflammation and fibrosis is also manifested by plant extracts that conjointly inhibit NF-κB and canonical TGF-β1 (i.e. Smad) pathways (Prunella vulgaris [84]), berberine [85]). Arctigenin from Fructus arctii also can improve the viability and function of the podocytes conjointly with attenuating the NF-κB-mediated inflammatory effects due to protein phosphatase 2 A-mediated decrease in p65 NF-κB activating phosphorylation [86].
Several herbal products hampered the production of ROS conjointly to inhibiting NF-κB dependent inflammatory pathways (puerarin [76]) and the production/activation of various proinflammatory cytokines (Paederia foetida [78]), adhesion molecules, and fibrosis promoting cytokines (isorhamnetin, present in onions, sea buckthorn, and various other medicinal plants [87]). Secoisolariciresinol diglucoside (from Linum usitatissimum) is able not only to diminish the inflammatory and oxidative aggression on the kidney cells, but also to prevent their death, as reflected by higher levels of antiapoptotic markers survivin and B-cell lymphoma-2 (Bcl-2) [88].
Hypericum perforatum seems to exert an all-embracing nephroprotective action, the anti-inflammatory and antioxidant effects being complemented by antifibrosing and antiapoptotic activity (inhibition of caspase-3 and cytochrome c), and enhanced expression of PPAR (peroxisome proliferatoractivated receptor)-γ [31]. Strawberry (Fragaria × ananassa) extracts also combine the PPAR-γ pathway activating effect with an anti-inflammatory one, associated with suppression of several fatty acid synthesis genes and of the sterol regulatory element-binding protein (SREBP) transcription factor [89], one of the inducers of TGF-β1 that may be activated by a HG-milieu [90]. The inactivation of both SREBP and NF-κB pathways, coupled with antioxidant activity and decreased production of AGEs, was proven for the fruit of Cornus officinalis, having morroniside, loganin, and 7-O-galloyl-D-sedoheptulose as the main active compounds [47].
The NLRP3 (nucleotide binding and oligomerization domain-like receptor family pyrin domain-containing 3) inflammasome triggers inflammatory events by means of activating caspase-1, which leads to both proinflammatory cytokines (such as IL-1β and IL-18) activation and to pyroptosis, a type of inflammatory cell death [91]. The dihydroflavone dihydroquercetin blocked the activation of NLRP3 inflammasome, besides reducing cell proliferation, ROS generation, and the expression of renal fibrosis-associated proteins [48].
The activation of the GSK-3β pathway is known to ameliorate diabetes-induced kidney injury. GSK-3β inactivation by PI3K/Akt-mediated phosphorylation increases protein synthesis associated with diabetic glomerular hypertrophy and sclerosis [12,92]. Akt (protein kinase B) is involved in metabolism (particularly glucose metabolism), growth, proliferation, and survival/apoptosis. PI3K is a key regulator of the multi-step process controlling Akt activation [93]. Emodin, the main active component of rhubarb (Rheum officinale), exerted its anti-inflammatory, antiapoptotic [decrease in B-cell lymphoma 2-associated X protein (Bax) and caspase-3 expression], and antioxidative activities by triggering PI3K/Akt/GSK-3β signaling pathway [94]. However, it is presently not clear whether PI3K/Akt induction should be 20 Journal of Diabetes Research considered protective or deleterious for the diabetic kidney, as among the phytoagents useful for preventing DN some turn this pathway on, while others turn it off [13].

Human Studies-Actually Nephroprotective Mechanisms.
In one of the very few human trials, berberine lowered the level of vascular cell adhesion molecule-1 and C-reactive protein, as well as the urinary markers of kidney injury. It also favorably tipped the oxidative stress balance (decreasing lipid peroxidation and nucleic acid oxidation and augmenting the total-antioxidant capacity) and improved renal hemodynamics. Hence, berberine may emerge as a nephroprotective agent able to complement standard hypotensive and hypoglycemic treatment [95]. In another trial on human subjects, Dioscorea bulbifera provided a better control than fosinopril of blood pressure, metabolic, biologic, and inflammatory parameters in patients with DN [96].

Preventing Endothelial Dysfunction.
The nephroprotective effect of some herbal products includes the prevention of endothelial dysfunction, an attribute of DN epitomized by the imbalance between endothelium-derived vasodilators and vasoconstrictors, the former preventing thrombosis and proliferation and the latter promoting atheroma formation [14]. Endothelial dysfunction is reflected by biomarkers such as endothelin 1 [97], homocysteine [98], and IL-6 (a proinflammatory cytokine) [99]. Induced by angiotensin II [100], oxidative stress, inflammation, and hypoxia [46], endothelin 1 promotes afferent and efferent arterioloconstriction with declining glomerular filtration rate [101].

In Vivo Studies-Actually Nephroprotective Mechanisms.
A component of green tea, (+)-catechin, has been proven to mitigate endothelial dysfunction, as reflected by diminished endothelin 1 levels, beside exerting antifibrosing and antioxidant activity [46].

Human Studies-Actually Nephroprotective Mechanisms.
A study on stable diabetic CKD patients revealed the ability of Salacia chinensis to forestall endothelial dysfunction, demonstrated by a decrease in the levels of homocysteine and IL-6 ([102]).

Preventing ER stress. Endoplasmic reticulum (ER)
stress is an important link in the pathophysiological chain leading to DN, consisting in improperly folded proteins piling up in the ER lumen and consequently unchaining the unfolded protein response (UPR), which may lead to either cell protection or cell death by apoptosis [15,16]. SERCA dysfunction is an important inductor of ER stress [103].
3.6. Herbal Products with Antifibrosing/Antiproliferative Activity. TGF-β1 has a crucial role in the progression of DN [105] leading to EMT [106], ECM expansion, collagen biosynthesis, and renal cell growth [10]. The result is glomerular and whole kidney hypertrophy [107] that, together with thickening of the glomerular basement membrane, are emblematic for DN [108].

In Vivo Studies-Actually Nephroprotective Mechanisms.
Tea catechins, a class of flavonoids, have been shown to decrease interstitial fibrosis [3]. Coccinia indica could prevent the accumulation of ECM by decreasing fibronectin and laminin in association with antioxidant activity [111].
Arguably the most efficient mechanism for preventing fibrosis/sclerosis is downregulation of its key promoter, TGF-β1, and of the related signaling pathways. Activation of TGF-β1 type I receptors leads to Smad2 and Smad3 phosphorylation, followed by binding to Smad4, the resultant oligomeric complex translocating into the nucleus, where it regulates the transcription of key proteins responsible for renal fibrogenesis [112]. Abating TGF-β1 expression is typically associated with lower levels of fibronectin (rhein [113]), collagen (Cordyceps militaris [114]), or both, sometimes accompanied by antiapoptotic effects, as is the case for Psoralea corylifolia seed extract and two of its major components, isopsoralen and psoralen. These phytoagents downregulated proapoptotic proteins [such as cleaved poly (ADP-ribose) polymerase and Bcl-2-associated death promoter (Bad)] and buttressed cell viability, as indicated by higher levels of survival markers p-Bad (ser112) and B-cell lymphoma-(Bcl-) 2 [115].
Modulation of G protein/AC (adenylyl cyclase)/cAMP (cyclic adenosine monophosphate) signaling pathway by means of altering G protein-coupled receptor kinases may be connected with dwindling TGF-β1 and collagen 4 expression in the case of berberine [121]. Berberine also interfered with prostaglandin E 2 /E prostanoid receptor (EP) 1/Gαq/Ca 2+ pathway, its depressing effects being associated with a decline in mesangial cells proliferation [122]. By blocking the SREBP, which is known to upregulate TGF-β1 [90], curcumin was linked to diminished expression of TGF-β1, EMT suppression (diminished vimentin), and podocyte protection (low desmin and high synaptopodin and connexin 43) [108]. Tangshen Formula combines the antifibrosing effect with an autophagy-inducing one, achieved by inhibiting the pathway leading from promyelocytic leukemia zinc finger protein (PLZF) activation to a decline in autophagy and autophagy-induced collagen 3 dissolution. Hence, the effect of Tangshen Formula is more autophagy and less collagen 3 accumulation, and consequently less tubulointerstitial ECM deposition [123].

Human Studies-Actually Nephroprotective Mechanisms.
A multicenter randomized, double-blind, placebo-controlled trial on the efficacy of Tangshen Formula in DN yielded promising results: after a mere 6 months, the improvements in proteinuria and estimated glomerular filtration rate (eGFR) were significant [124], as was the decrease in livertype fatty acid binding protein [125], a biomarker correlated with the severity and prognosis of DN [126]. Longer duration interventions are clearly needed for defining the role of this formula in preventing DN.

Herbal Products Blocking Epithelial-Mesenchymal
Transdifferentiation. Fibrosis and ECM synthesis are bolstered by EMT, in which the epithelial cells trade their epithelial features (reflected by proteins such as E-cadherin) for mesenchymal ones (correlated with the expression of αsmooth muscle actin and vimentin, among others) [4]. TGF-β1 is one of the main inducers of EMT by means of both the canonical pathway (TGF-β/Smad2/3) and the non-canonical one (MAPK/PI3K) [106]. Consequently, all herbal products able to forestall TGF-β1 generation are potentially apt to thwart EMT.

In Vitro Studies-Potentially Nephroprotective
Mechanisms. Matrix metalloproteinase-(MMP-) 9 promotes EMT by degrading type IV collagen of the basement membranes (thereby facilitating the migration of the transdifferentiated epithelial cells through the altered basement membrane) and by preventing the dissolution of ECM, which furthers renal interstitial fibrosis. The overactivity of MMP-9 results from insufficient inhibition by TIMP-1. The MMP-9/TIMP-1 imbalance is induced by integrin-linked kinase (ILK), a downstream mediator on a pathway driven by TGF-β1 [127]. Rhein hampers EMT by suppressing ILK, thus correcting the excessively high MMP-9/TIMP-1 [127].
3.8. Herbal Products that Restore Autophagy. By degrading damaged cellular components, autophagy, an adaptive response of cells exposed to various environmental stresses, is essential for maintaining the structural and functional well-being of podocytes [103,131,132]. Consequently, its malfunction in DN leads to podocyte loss and proteinuria [133]. As it acts as a protective mechanism [134], its inactivation contributes to disease progression [135]. Nonetheless, some authors have found autophagy activation in DN-this seems to be an early, short-lived event, followed by autophagy depression in the long run [135].

In Vitro Studies-Potentially Nephroprotective
Mechanisms. Several pathways converge in regulating autophagy and are the points of impact of the various herbal products that enhance cell viability by activating autophagy: AMPK (berberine [136]), mTOR (hispidulin [137]), PI3K/AKT (curcumin [132]), and β-arrestin-1 (tripterygium glycosides [138]). Besides increasing autophagy, celastrol also dampens inflammation, oxidative stress, and apoptosis and hence increases podocyte viability, as reflected by enhanced expression of nephrin [139]. Still, there are herbal products whose favorable action is associated with reduced autophagy, such as astilbin, a flavanonol found in several plants, including Hypericum perforatum. The beneficial effect of astilbin on HG-treated human proximal tubular epithelial cells was associated with diminished apoptosis and autophagy, resulting from Akt induction (mirrored by higher levels of p-Akt) [140]. This may be explained by autophagy activation early in the course of DN.

In Vitro Studies-Potentially Nephroprotective
Mechanisms. Activated by environmental stressors, p38MAPK stimulates cellular growth, differentiation, and apoptosis [151,152]. Activation of p38MAPK signaling is associated with the development of DN, being involved especially in the progression of interstitial fibrosis [11]. Emodin (from Rheum palmatum) interferes with p38MAPK pathway, thereby quelling proliferation and fibrosis, besides switching on the protective action of PPAR-γ signaling system [153].
PPARs, important regulators of lipid and glucose metabolism, are ligand-activated nuclear transcription factors demonstrated to have a protective role in DN. PPAR-γ ameliorates hyperglycemia, insulin resistance, hypertension, and albuminuria, and inactivates diacylglycerol (DAG)/protein kinase C (PKC)/extracellular signal-regulated kinase (ERK) pathway with subsequent decrease in TGF-β1, fibronectin, and type IV collagen. It also dampens inflammation and oxidative stress and blocks atheroma-promoting mechanisms. PPAR-α activation reduces TGF-β1 and type IV collagen expression and prevents fatty acid build-up and lipidinduced toxicity in the diabetic kidneys [147]. Several herbal products are able to block the proliferation/fibrosis/sclerosis-inducing p38MAPK pathway, coupled with the activation of signaling pathways that may have protective effects, such as PPAR-γ (Danhong, an extract of Salvia miltiorrhiza and Carthamus tinctorius [154]) or PI3K/Akt. The upregulation of the latter by (-)-epigallocatechin-3-gallate was accompanied by a favorable effect on the RAS and on the oxidative status [155]. However, Abelmoschus manihot extract decreased the level of p-Akt, while still exerting a renoprotective effect associated with quenching of the p38MAPK pathway and anti-inflammatory, antifibrosing, and antioxidant effects [156]. An experiment done on db/db mice seems to indicate that Huangqi decoction shields the kidney from the consequences of diabetes through the agency of PI3K-Akt signaling. However, the activation of PI3K was associated with lower levels of the active form of Akt (p-Akt), which may explain why the positive effects of PI3K upregulation (higher expression of GLUT4 receptors that mediate intracellular glucose uptake) were not accompanied by the undesirable ones (protein synthesis and cellular proliferation driven by p-Akt) [36].
The flavonol myricetin activates PPAR-α, improves the biological and histopathological markers of kidney injury and fibrosis, reduces the levels of SREBPs, and favorably interferes with lipid metabolism [157]. As mentioned above, the antifibrosis effect of Hypericum perforatum was accompanied by the activation of PPAR-γ signaling pathway [31]. Abelmoschus manihot extract augmented the transcriptional activity of PPAR-α and PPAR-γ, improved the lipid status, and decreased the expression of inflammatory genes, ER stress, and c-Jun NH2-terminal kinase activation [104].
EP4/Gs/AC/cAMP signaling pathway (a Gs proteincoupled EP4 receptor that generates cAMP by activating AC) exerts a kidney-protective effect in diabetics ameliorating renal injury and slowing the progression of experimentally induced CKD [148]. At least some of the renoprotective effects of berberine are due to its ability to activate this pathway [158].

Human Studies-Actually Nephroprotective Mechanisms.
A systematic review and meta-analysis that took into account 66 studies involving 4785 participants concluded that Astragalus membranaceus (Huang Qi) increased the effectiveness of conventional therapies in lowering albuminuria, proteinuria, and serum creatinine levels, with no increase in adverse effects. However, methodological flaws resulted in a low quality of the available studies [159].

Strength of the Evidence in Various Areas
The evidence provided by in vitro and in vivo animal studies is abundant and strong. Much fewer are the human studies. We have found 12 references regarding clinical trials on herbal products in DN. Only 5 of them included information about putative action mechanisms. Two of them are high quality (J5), one on silymarin (decrease in albuminuria) [65], the other on Tangshen Formula (decrease in proteinuria, increase in eGFR) [124,125]. The other three are low quality (J1): one on berberine (decrease in albuminuria) [95], one on Dioscorea bulbifera (decrease in proteinuria) [96], and one on Salacia chinensis (increased creatinine clearance, slowed CKD progression) [102]. Five other references describe study protocols on the efficiency of various traditional Chinese herbal medicines in DN [160][161][162][163][164]. The remaining two articles present the results of clinical trials that did not have as an objective the identification of action mechanisms. One is a moderate power randomized placebocontrolled trial demonstrating the ability of a Chinese herbal formula to decrease microalbuminuria [165]; the trial is of low-to-moderate quality, i.e., J2, as it was not doubleblinded (only the participants were blinded) and no information on dropouts/withdrawals was provided. The other one was a moderate-powered double-blinded, randomized, high-quality (J5) clinical trial on saffron (Crocus sativus L.) but failed to show any significant change in kidney function parameters or albuminuria [166].
A recent meta-analysis of randomized placebo-controlled trials of Chinese herbal medicine for DN selected 20 studies including 2719 patients and concluded that plant products may favorably influence renal function and albuminuria beyond the beneficial effect of conventional therapy (mainly RAS inhibitors). However, the evidence was deemed of moderate-to-low quality. The medicinal plants most frequently employed in the various combinations were Astragali radix, Rehmanniae Radix, and Rhei Radix et Rhizoma [167]. The most studied herbal in DN seems to be Astragalus membranaceus, but the available studies are of low quality [159].
We may conclude that most of the existing studies are of inferior quality, but there are a few moderate-or high-quality ones proving that herbals may help prevent kidney injury and maintain kidney function in diabetic patients. Some of 23 Journal of Diabetes Research the most promising phytoceuticals are silymarin, Astragalus membranaceus, and Tangshen Formula. Given the vast experience with silymarin in the treatment of other (mainly liver) disorders, the wide availability of financially affordable silymarin containing preparations, and the lack of side effects, at least silymarin should be probably included in the standard treatment of diabetic patients, pending the completion of ongoing clinical trials on various other herbal combinations.

Concluding Remarks
Many agents of plant origin have shown the ability to prevent DN, the mechanisms involved being, to a great extent, similar to those of the conventional drugs used for this aim: blocking RAAS, antioxidant, anti-inflammatory, and antiproliferative, preventing sclerosis/fibrosis, improving lipid profile, and activating protective pathways. Consequently, herbal medicines may emerge as healthy alternatives to the agents cur-rently employed for slowing the progression of DN. The most important setback is the scarcity of human experiments, which is surprising and regrettable as these agents are generally devoid of side effects. The dearth of human studies is compounded by the low quality of most of the existing ones. Both these issues should be urgently addressed by future research in this area. A graphical abstract of this paper is provided as Figure 2.