*4.1. OAT1 and OAT3*

β α κ The basolateral membrane of renal proximal tubule cells highly expresses OAT1 and OAT3, mediating the excretion of their substrates including ACE inhibitors, angiotensin II receptor blockers, diuretics, β-Lactam antibiotics, antiviral agents and endogenous compounds [3]. Several studies demonstrated that diabetes also impaired the function and expression of renal OATs. In STZ-induced diabetic rats, it was found that diabetes significantly decreased the membrane expression of renal OAT3, leading to lower uptake of [3H] estrone sulfate in renal cortical slice. These decreases were in line with the activation of PKCα and NF-κB pathways, increased nuclear factor erythroid 2-related factor 2 (Nrf2) and oxidative stress [115,116]. Alteration in OATs by diabetes is dependent on sex. In lean and obese Zucker spontaneously hypertensive fatty (ZSF1) rats, it was found that the levels of renal OAT1 and OAT3 mRNA were higher in lean females than in lean males. Obesity remarkably reduced the expression of renal OAT1 and OAT3 in female ZSF1 rats but not in male ZSF1 rats [117]. Clinical trials showed that expressions of renal OAT1 and OAT3 in patients with diabetic kidney

disease were less than 50% of the normal levels, which were in line with decreased urinary excretion of some organic acid metabolites and increased plasma levels of organic acid metabolites [118].

Diuretics show their diuretic effect via affecting proximal tubular epithelial cells. Some diuretics (such as torasemid and furosemide) are also substrates of OATs [3]. Several studies have showed that OAT1 or OAT3 deficiency damage the natriuretic effects of furosemide and bendroflumethiazide [119,120], inferring that the decreased expression of OATs by diabetes may impair the diuretic effect of diuretics. In line with the deduction, alloxan-induced diabetes significantly increased the plasma exposure of furosemide and decreased recovery from urine following intravenous dose to rats [105]. Importantly, diuretic efficiency, natriuretic efficiency, kaluretic efficiency and chloruretic efficiency of furosemide in diabetic rats showed a trend to decrease. Similarly, the diuretic efficiency, natriuretic efficiency, kaluretic efficiency and chloruretic efficiency of torasemide were significantly impaired both in alloxan-induced diabetic rats and STZ-induced diabetes, although its plasma exposure was unaltered [50]. These findings also may explain the clinical finding that diabetic patients need higher furosemide doses [121]. Moreover, sodium-glucose cotransporter (SGLT2) inhibitor empagliflozin is a substrate of OAT3. A report showed that OAT3 deficiency damaged the glucosuric effect of empagliflozin [122], indicating that the decreased expression of renal OAT3 by diabetes may attenuate the glucosuric effect of empagliflozin, which may also explain the clinic finding that compared with normal renal function and normal-to-mildly reduced renal function, diabetic patients with mild-to-moderately reduced renal function showed the lowest lowering glucose effect of luseogliflozin [123].

#### *4.2. OCTs*

Renal OCT1/2, mainly expressed at the basolateral membrane of tubule cells, transport a variety of organic cations, such as metformin, cisplatin, cephalexin and acyclovir. In STZ/HFD-induced diabetic rats, it was found that OCT2 protein expression was only 50% of control rats [124]. Similarly, in STZ-induced diabetic rats, the expressions of renal OCT1 and OCT2 were decreased by 50% and 70%, respectively [125]. The reduced mRNA and protein expressions of OCT1, OCT2 and OCT3 under diabetic status were also associated with a reduction in the clearance of N<sup>1</sup> -methylnicotinamide [126,127]. The decreased mRNA and protein expressions of renal OCT2/3 were reported to be negatively correlated with the accumulation of renal and plasma advanced glycation end-products (AGEs) [127]. Insulin or AGE inhibitor aminoguanidine could reverse the decreased OCT2/3 by diabetes [125,127], inferring that the accumulation of AGEs may be involved in impaired expression of renal OCTs by diabetes [127]. Expressions of OCT1 and OCT2 are also be dependent on sex. In lean female ZSF1 rats, the mRNA expression level of OCT1 was significantly higher than that in their male counterparts. Obesity significantly reduced renal mRNA expressions of OCT1 and OCT2 in female ZSF1 rats, but not in male rats [117]. The decreased expression of renal OCTs by diabetes may affect pharmacokinetics of their substrates. For example, significantly higher AUC of metformin and lower renal clearance were observed in alloxan-induced diabetic rats than in normal rats following an intravenous dose [128]. A clinic trial [129] showed that renal clearance of metformin decreased along with diabetes progression. A report demonstrated the rank of the renal clearances of metformin in healthy subjects (525 ± 125 mL/min) > newly diagnosed maturity onset diabetic patients (322 ± 166 mL/min) > maturity onset diabetic patients (224 ± 38 mL/min). Interestingly, pregnancy seems to increase the functions of renal MATEs and OCTs [130]. Renal clearances of *N*<sup>1</sup> -methylnicotinamide (endogenous probe for the renal OCTs and MATEs) in both mid (504 ± 293 mL/min) and late pregnancy (557 ± 305 mL/min) were reported to be higher than in postpartum (240 ± 106 mL/min). The renal secretion of *N*1 -methylnicotinamide was 3.5-fold higher in mid pregnancy and 4.5-fold higher in late pregnancy compared with postpartum [130]. In line, gestational diabetes mellitus pregnant women showed higher renal clearance and renal secretion [131], and lower plasma concentrations of metformin [131,132] compared with non-pregnant controls. Another example is cisplatin. Cisplatin is a substrate of OCTs. Its main toxicity is nephrotoxicity. Several studies have demonstrated that both insulin-dependent

and insulin-independent diabetes show resistance against cisplatin-induced nephrotoxicity [133–135]. Animal experiments showed that diabetes prevented nephrotoxicity partly via decreasing the renal accumulation of cisplatin [136,137]. However, cisplatin treatment was not beneficial in diabetes due to its compromising its antitumor effect [136].

#### *4.3. Other Transporters*

Diabetes induced an approximately two-fold increase in renal PepT1 protein and slightly induced mRNA of PepT1 and PepT2 [138]. Reports on the expression of ABC transporters are often contradictory. For renal P-gp, it was reported that spontaneous type 1 and type 2 diabetic mice showed lower expressions of renal P-gp protein. The high glucose was considered to be a reason reducing the expression and function of P-gp [139]. However, the expression of P-gp protein was also reported to be unchanged [17,111,138], although level of Mdr1a mRNA was increased [17,138].

For MRP2, increased expression of renal MRP2 protein was found in STZ/HFD-diabetic rats [124] and in STZ-induced diabetic rats [23]. For BCRP, HFD/STZ-induced diabetes significantly increased the expression of BCRP protein in rats, but STZ-induced diabetes significantly decreased mRNA level of renal BCRP in rats [136]. These discrepancies may result from the duration of diabetes, sex and the type of diabetes. For example, in female rats, obesity decreased the expressions of MRP4 and Mdr1b mRNA, but the expression of MRP2 mRNA was unaltered. In male ZSF1 rats, expressions of MRP2, MRP4 and Mdr1b mRNA remained unchanged in obese rats compared to lean rats [117].

### *4.4. CYP450*

The identified renal CYP450s include CYP2C, CYP2J, CYP2E1, CYP4As and CYP4Fs. They mainly mediate local biotransformation of endogenous compounds such as arachidonic acid. For example, arachidonic acid is metabolized by CYP4As and CYP4Fs to 20-hydroxyeicosatetraenoic acid (vasoconstrictor) and by CYP2C and CYP2J to epoxyeicosatrienoic acid (vasodilator), synergistically regulating renal vasoactivity. Several reports have shown that diabetes also affect the expressions of renal CYP450s. In mice, it was found that HFD feeding significantly decreased activities of CYP3A (1′ -hydroxymidazolam), CYP2E1 (hydroxychlorzoxazone), CYP2J (ebastine hydroxylation) and CYP2B6 (hydroxybupropion) and CYP4A (12-hydroxydecanoic acid), but increased the activities of CYP2C (hydroxytolbutamide) and CYP2D (Hydroxybufuralol) [140]. Diabetes affects the expressions and activities of renal CYP450s in an isoform- and species-specific manner. In STZ-induced diabetic rats, formations of hydroxyandrostenedione and 2α-hydroxytestosterone were significantly increased by 250% and 300% compared to control rats, but formations of androstenedione and 6β-hydroxytestosterone were significantly decreased. The formations of 16α-hydroxytestosterone and 7α-hydroxytestosterone were unaltered [141]. Induction of CYP2E1, 4A2 and K-4 as well as increases in omega- and (omega-1) hydroxylation of lauric acid were also observed in kidney of STZ-induced rats [26]. Insulin treatment partly reversed these alterations by diabetes [26,141]. However, a recent report showed that expression of CYP4A protein in diabetic mice induced by HFD/STZ was downregulated by 29.16% of control mice [142]. The roles of altered renal CYP450s by diabetes in disposition of drugs needed further investigation.

#### *4.5. Transporter Interplay in Kidney*

Transporter interplay occurs in the kidney. For example, methotrexate is eliminated mainly via renal active excretion, which is involved in OAT1/3 at basolateral membrane of tubule and BCRP and MRP2 at brush-border membrane of tubule, these transporters in series work to regulate renal active excretion of methotrexate. Both higher plasma concentration of methotrexate and higher toxicity of methotrexate in STZ-diabetic rats [143] may partly be explained by the imbalance of expressions of renal OATs, BCRP and MRP2. Another example is β-lactam antibiotics. OATs mediate the uptake of β-lactam antibiotics into the tubule from the blood and PepT1/2 located in the brush-border membrane of the tubule, mediate reabsorption of antibiotics from the urine, and regulate urinary excretion of β-lactam antibiotics. Downregulation of renal OATs may partly contribute to a decrease in the renal clearance of DA-1131 (a carbapenem antibiotics) [144] and lower renal cortical cephaloridine accumulation in diabetic rats [145]. Creatinine is also substrate of OCT1/2, MATEs and OAT1/3, indicating that the decreased expressions of renal OATs and OCTs at least partly contribute to decreased clearance of creatinine under diabetic status.
