Protein expressions and activities are altered in animal models of T2DM; however, it must be noted that the way T2DM is induced in these animal models, either chemically, by diet or obesity-induced models, or in genetically engineered animals, could lead to different results on p450 mRNA and protein expression [80]. It is also important to note that chemical induction of T2DM in animals will alter and/or induce p450 pathways and that, although there is high conservation of p450 enzymes amongst species, the extent and catalytic activity between species differ substantially [81]. Nevertheless, a number of studies have reported a modulation of various p450 isoforms in rodent animal models including cyp1a1, cyp1a2, cyp2a1, cyp2b1/2, cyp2c6, cyp2c7, cyp3a1/2 and cyp2e1 expression [82,83,84,85]. A summary of alterations reported is presented in Table 1.

The most common alteration noticed in chemically induced rodent T2DM models is the induction of cyp2e1, where it is induced almost eight-fold in streptozocin (STZ)-induced diabetic rats [86]. This has been explained by different mechanisms. First, increased ketone bodies by post translational stabilization of the enzyme by high ketone levels [87,88]. Second, by insulin secretion which does not only decrease cyp2e1 gene transcription but also decreases mRNA half-life [89]. This induction can be suppressed by insulin administration, suggesting that insufficient insulin levels could lead to the induction of cyp2e1 [66,82,88,90]. Lastly, regarding decreased growth hormones levels, as T2DM suppresses growth hormone release, this, in turn, can also contribute to cyp2e1 induction through mRNA stabilization or increased transcription [91]. In support of the role of insulin, others have observed that cyp2e1 and cyp2b expression are decreased in primary cultured rat hepatocytes exposed to insulin [92]. In contrast, cyp2e1 mRNA levels and activity were not significantly modulated in a T2DM model from diet-induced obesity in mice [93].

Induction of cyp1a2 was also observed in STZ diabetic rats. Cyp1a2 induction by T2DM resulted in lower verapamil and lidocaine (which are partially metabolized by cyp1a2) area under the curve (AUC) of drug plasma concentrations in STZ diabetic rats [94,95]. In alloxan-induced diabetic rats, induction of cyp1a2 resulted in lower AUC of oltipraz and higher clearance of antipyrine [96,97]. Theophylline is metabolized through cyp1a2 and cyp2e1 to 1,3-dimethyluric acid. In alloxan diabetic rats, it was also shown that T2DM increases the expression of both p450 isoforms resulting in lower theophylline AUC and thus higher 1,3-dimethyluric acid plasma levels [83].

STZ-induced T2DM animal models have been shown to significantly lower the expression of cyp2c11, resulting in decreased metabolism of diclofenac, glibenclamide and nateglinide [98,99,100]. Studies conducted in a genetically engineered T2DM animal models (db/db mice) demonstrated that the expression of the hepatic glucocorticoid receptor was increased with associated increased cortisone levels [101]. Higher levels of cortisone resulted in increased levels of cyp2b and cyp2c causing a change in the metabolism of substrates such as phenytoin [100,102]. In contrast, in their diet-induced obesity model of T2DM, Maximos et al. have also shown that cyp2c and cyp2b activity and mRNA expression was downregulated in this mouse model [93,103].

Studies in T2DM animal models have shown both increased and decreased levels of cyp3a expression and activity. On one hand, cyp3a activity was shown to be reduced in diet-induced obesity diabetic animals, while other studies in STZ diabetic rats indicated that it is increased [93,104,105,106]. Consequently, discrepancy on the impact of T2DM on the metabolism of various substrates has been reported. Studies have shown that cyp3a1 expression is decreased in male STZ-diabetic rats but increased in female STZ-diabetic rats [94,107]. It was shown that glibenclamide clearance was reduced in Zucker diabetic fatty (ZDF) rats, suggesting a reduction in cyp3a activity [108]. Maximos et al. demonstrated that cyp3a11 and cyp3a25 mRNA levels were decreased 2- to 14-fold in the liver which was associated with a 21-fold decrease in midazolam metabolism [93]. They also showed that cyp3a modulation was tissue-dependent [93].

The intestine also expresses a variety of p450 enzymes, mainly cyp3a, that contributes to first-pass metabolism. It was found that cyp3a expression was significantly decreased in the intestine of diabetic rats and mice [106,109]. However, insulin was able to partially reverse the decreased in intestinal cyp3a expression in rats [110].

Most studies have indicated that cyp4a expression and activity are induced in T2DM rats. According to Kroetz et al., cyp4a mRNA levels were induced about eight-fold in STZ-induced diabetics rats, which translated into an associated increase in cyp4a protein expression and activity [111]. The molecular mechanism proposed of the induction of cyp4a in diabetic rats is related to the activation of peroxisome proliferator activated receptor-alpha (PPAR-α) increasing cyp4a transcription while normal regulation of cyp3a is due to the activation of the pregnane X receptor (PXR), which is known to alter the expression profiles of cyp3a [111].

It has been shown that inflammation-related conditions suppress CYP3A expression and activity in human hepatocytes [57,121]. CYP3A4 expression and activity were significantly lower in hepatic microsomes obtained from diabetic livers. One study reported, using human liver microsomes and midazolam as a probe drug, that CYP3A activity and mRNA and protein expression levels in samples from T2DM patients were decreased 1.6- to 3.3-fold compared to samples from non-diabetic patients [20]. In contrast, CYP2E1 mRNA and protein expression levels as well as activity (using chlorzoxazone as probe drug) were increased significantly [20]. One reason for diabetic induction of CYP2E1 could be hyperketonemia as demonstrated in animal models [122]. CYP2B6 mRNA level has been shown to be reduced under inflammatory conditions in human hepatocytes [57]. Glyburide, a second-generation sulfonylurea, is mainly metabolized by CYP2C9 to cis-3-hydroxycyclohexyl glyburide and trans-4-hydroxycyclohexyl glyburide [123,124]. According to Jain et al. the metabolism of glyburide is decreased in microsomes prepared from the placenta of pregnant women with T2DM compared to that of non-diabetic pregnant women suggesting a decrease in CYP2C9 activity [114].

Together, these P450 enzymes metabolize multiple drugs, and alteration in their enzyme activity or expression could modulate the biotransformation of several drugs in T2DM patients, thus making them either toxic or ineffective (prodrugs).

Marques et al. studied 17 hypertensive patients, 9 of whom were also diabetic, to measure the pharmacokinetics of the anti-hypertensive agent nisoldipine, a known CYP3A4 substrate [48]. Additionally, lidocaine disposition was assessed as a marker of CYP3A4 activity. These researchers found that nisoldipine clearance was stereospecific in non-diabetic controls, with the (−) enantiomer exhibiting approximately six-fold higher clearance compared to the (+) enantiomer. However, when non-diabetic and T2DM patients were compared, clearances of the (−) and (+) nisoldipine were reduced by 52% and 38% in T2DM patients, respectively [48]. Consequently, AUCs of (−) and (+) nisoldipine were increased by 143% and 94% in T2DM patients compared to non-diabetic controls [48]. The suggested decrease in CYP3A4 activity in T2DM patients was somewhat confirmed by an increase (~21%) in plasma concentration of lidocaine in T2DM patients 15 min after intravenous infusion of the drug [48]. The fact that lidocaine—a relatively high liver extraction ratio drug—was administered intravenously could explain the difference in the magnitude of changes observed between lidocaine and nifedipine although a decrease in the clearance was observed for both drugs. Moises et al. compared the clearance of lidocaine in pregnant woman vs. non-pregnant woman. They also noticed that lidocaine clearance was lower in pregnant women with gestational diabetes, suggesting that gestational diabetes reduced the activity of CYP3A4 (and possibly CYP1A2) which is responsible for the metabolism of this drug although the combinatorial effect of pregnancy and gestational diabetes (i.e., diabetes with a different origin than T2DM), cannot be excluded [115].

Dostalek et al. showed using population pharmacokinetic modelling that T2DM patients had a significantly reduced clearance of atorvastatin lactone when compared to non-diabetic patients [20]. CYP3A4 is responsible for the metabolism of atorvastatin lactone and acid forms into corresponding hydroxylated metabolites with a higher intrinsic metabolic clearance for the lactone form compared to the acid form [125]. CYP3A4 downregulation due to T2DM could result in less atorvastatin lactone clearance thus forcing it to equilibrium with the acid form, which could lead to a higher risk of 3-hydroxy-3-methyl-glutaryl-CoA reductase (HMG-CoA reductase) inhibitor (statin)-induced myotoxicity.

Diabetes is one of leading causes of kidney diseases; consequently, 30–40% of all kidney transplant recipients in the United States are diabetic at the time of transplantation surgery, and an additional 15–20% develop new-onset diabetes after transplant [126]. Cyclosporine is a calcineurin inhibitor used as an immunosuppressant to prevent organ transplant rejection. Cyclosporine is metabolized in both the intestine and the liver by CYP3A4, resulting in the formation of multiple metabolites [127]. Of the most important metabolites of cyclosporine are the 2-hydroxylated metabolites (AM1 and AM9) [59]. Factors that can contribute to its pharmacokinetics (PK) variability include the extent of cyclosporine pre-systemic elimination by CYP3A4 and/or by the efflux transporter ABCB1. Hryniewieck et al. investigated the associations of cardiovascular diseases with cyclosporine metabolites and indices of cyclosporine metabolism [128]. Their results showed lower AM9 metabolite/cyclosporine ratio of their plasma concentrations in patients with diabetes. This finding suggests a decreased hepatic metabolism in T2DM patients.

Tacrolimus, another calcineurin inhibitor used to prevent organ transplant rejection, is metabolized by CYP3A4/5 to active and inactive metabolites [129,130,131]. There is a wide intersubject variability in the tacrolimus dose needed to achieve therapeutic trough levels. Jacobson et al. found that T2DM patients exhibit increased trough levels suggesting a decrease in CYP3A activities [117].

Gravel et al. have shed light on the clinical impact of T2DM on P450s by investigating activities of seven of the major CYPs (CYP1A2, CYP2B6, CYP2C9, CYP2C19, CYP2D6, CYP2E1, and CYP3A4/5) in T2DM compared to non-T2DM patients following the administration of oral probes (caffeine, bupropion, tolbutamide, omeprazole, dextromethorphan, chlorzoxazone, and midazolam) [120]. They concluded that the activity of CYP2C19, CYP2B6, and CYP3A in T2DM patients was decreased by 46%, 45%, and 38%, respectively; these effects were attributed to some cytokines such as IL-6 [120].

Korrapati et al. found that, while a group of T2DM patients did not differ significantly in their theophylline clearance, theophylline metabolism was positively associated with the degree of HbA1c within the diabetic group suggesting that CYP1A2 may be induced by poor glycemic control [118]. Urry et al. and Gravel et al. have demonstrated that an increase in paraxanthine to caffeine ratio is suggestive of an increase in CYP1A2 activity in patients with T2DM and that CYP1A2 activity can be dependent on glycemic levels as well as HbA1c [78,120]. Hong et al. described a clear association between endogenous insulin levels and CYP1A2 activity [132]. The exact mechanism of CYP1A2 modulation by inflammatory factors is not known but studies may suggest the role of aryl hydrocarbon receptor activation in immune responses or under oxidative stress [133].

In regard to other P450 isoforms, CYP2D6 activity was not changed in patients with T2DM using dextromethorphan as a probe substrate [134]. CYP2E1 has been reported to show no change in activity and expression and that it is less sensitive to disease state (e.g., inflammation) or T2DM [23,120,135,136]. In contrast, as mentioned previously, an induction in cyp2e1 has been noticed in STZ-diabetic rats.