Читать книгу Drug Transporters - Группа авторов - Страница 126

In spite of the apparent multi‐specific nature of the OATs, there exist specific structural/molecular determinants within their substrates, which help target them for elimination via a particular OAT. This view is supported by studies of knockouts of Oat1 and Oat3, which have revealed altered handling of a variety of pharmaceuticals and toxins (Table 4.2). For example, while in vitro studies suggested that PAH is handled by the majority of Oats, analysis of the Oat1 knockout animal found that in vivo handling of this “prototypical” Oat substrate was largely dependent on Oat1 [124, 125]. As expected, the handling of a number of drugs is affected in the Oat1 and Oat3 knockout animals or tissues (e.g., kidney, choroid plexus) derived from them. These include diuretics, antibiotics, antivirals, and methotrexate [38,115–117,125–128]. The knockout of Oat1 protects the kidney from injury mediated by mercury conjugates, suggesting that Oat1 is the primary renal transporter of this environmental toxin [119]. Certain metabolites such as urate appear to be handled in concert by Oats, Rst/Urat1, and other Slc and/or Abc transporters [38, 122]. In addition, metabolomics analysis of urine and plasma samples from Oat1‐null and wild‐type mice identified dozens of small molecule metabolites with altered concentration in the mutant mice [120, 123, 129]. However, the concentrations of these metabolites were not altered in Oat3‐null mice [125], providing support for the notion that Oat1 and Oat3 each have its own substrate spectrum.

TABLE 4.2 Characterization of Oat1/Oat3 knockout mice with respect to drugs, toxins, and endogenous metabolites

Compound	Oat1 KO	Oat3 KO	In vivo or ex vivo	Reference
Antivirals	✓	✓	Ex vivo	[115]
Antibiotics		✓	In vivo	[116]
Diuretics	✓	✓	In vivo	[117]
Fluoroquinolone antibiotics		✓	In vivo	[118]
Mercury	✓		In vivo	[119]
Uremic Toxins	✓	✓	In vivo	[120, 121]
Uric acid	✓	✓	In vivo	[122]

Adapted with permission from Ref. 123.

A machine learning‐based chemoinformatic analysis of the altered metabolites in each knockout revealed distinct differences in substrate specificity [39]. For example, compared to Oat1, Oat3 appears to prefer larger, more hydrophobic (endogenous) molecules with a larger number of rings. This raises the possibility that one can exploit this molecular/structural data inherent to the substrates to gain insight into Oat‐mediated transport. This approach, which represents a strategy for designing inhibitors or for designing drugs targeting specific Oats, analyzes the substrates that interact with the Oats and identifies common chemical features/molecular determinants among the substrates to generate hypothetical models called pharmacophore models. Pharmacophore models for Oat1, Oat3, and Oat6 have been built, and these have proven quite helpful in increasing our understanding of transporter–substrate interactions [36, 120,130–133]. For example, initial computational characterization and comparative molecular field analyses of the substrate binding preferences of Oat1 versus Oat6 suggested that the (mostly) pharmaceutical substrate selectivity of these transporters was influenced by mass, net charge, and hydrophobicity [131]. Other approaches have based pharmacophores on endogenous metabolites, which accumulate in the plasma of Oat1 knockout animals [120, 132]. In these studies, metabolites were clustered into distinct groups and consensus pharmacophores were created. Chemical databases were screened virtually for molecules which fit the pharmacophores, and several of these molecules were subsequently shown to interact with Oat1 in in vitro assays [120, 132]. In addition, to analyze the cationic binding potential of Oat3, a pharmacophore model based on cations that bind Oat3 was created. The model was used to screen a commercially available compound database, leading to the identification of molecules that proved to be potent highly selective in vitro inhibitors of Oat3 [36].