Glutathione transferases GST have been discovered in many ti
Glutathione transferases (GST) have been discovered in many tissues including the cornea of humans and several other mammals (Awasthi et al., 1980, Bilgihan et al., 2003; Gondhowiardjo and van Haeringen, 1993; Saneto et al., 1982, Sastry et al., 1995, Singh et al., 1985, Watkins et al., 1991). To date, three mammalian GST superfamilies have been identified, consisting of cytosolic proteins, mitochondrial proteins and membrane-associated proteins in eicosanoid and glutathione metabolism (MAPEGs). The cytosolic GSTs form the most highly differentiated superfamily. Cytosolic Alpha (α), Zeta (ζ), Theta (θ), Mu (μ), Pi (π), Sigma (σ) and Omega (ω) bifonazole sale have been identified in humans. Those seven classes of soluble proteins comprise 17 different enzymes (Hayes et al., 2005). Based on the GST nomenclature proposed by Mannervik et al., GST classes are abbreviated in Roman capital letters and the specific enzymes are differentiated using Arabic numbers (Mannervik et al., 2005). GSTs exist as dimeric proteins. Generally, these enzymes form homodimeric structures, which are indicated using the same Arabic number for each subunit (Mannervik et al., 2005). For example, the homodimeric Pi glutathione transferase, which is composed of two 1 subunits, is abbreviated as GSTP1‐1. However, heterodimeric GST structures are known to form, particularly by members of the Alpha and Mu classes (Hayes and Pulford, 1995).
However, to our knowledge, this is the first study focusing on the differences in phase II enzyme expression in the corneal layers. This topic is important because the corneal layers exhibit different barrier properties against APIs. The epithelium is the rate-limiting layer for all substances except small lipophilic compounds (Prausnitz and Noonan, 1998). In contrast, the stroma forms a barrier against lipophilic drugs (Prausnitz and Noonan, 1998). Because of these different barrier properties, the transit times of permeating APIs vary in the different corneal layers. Therefore, a closer investigation of the enzymatic activity in the corneal layers will increase the knowledge concerning possible metabolic transformations of drugs occurring during the transcorneal transport process in both ex vivo and in vitro models.
Because the glutathione transferases GSTO1‐1 and GSTP1‐1 were detected at the mRNA level in the Hemicornea construct and human corneal epithelium (Kölln and Reichl, 2012), further investigation of the expression of those enzymes is warranted. Therefore, one aim of the present study was to examine whether GSTO1‐1 and GSTP1‐1 were also expressed on the protein level in the Hemicornea construct and in the commonly used animal models, the porcine and rabbit cornea. Additionally, the functional activity of glutathione transferases in general and the selected enzymes GSTO1‐1 and GSTP1‐1 in particular was investigated in these models.
Materials and methods
Discussion The amino acid sequences of the cytosolic GST enzymes of different classes are <25% identical and those of the enzymes within a class are >40% similar (Hayes et al., 2005). In addition, the conserved residue in the active sites of Alpha, Mu, Pi and Sigma GSTs is tyrosine (Board and Menon, 2013). In contrast, members of the Theta and Zeta classes have a conserved serine in the N-terminal domain (Board and Menon, 2013). Omega GSTs, however, have a cysteine residue (Board, 2011). This explains the overlapping substrate affinities of many GST classes, therefore the total GST activity is determined using CDNB as the substrate. This compound is known to be a substrate of the members of the Alpha, Mu and Pi classes and to be catalyzed at a very low level by members of the Omega class (Board et al., 2000). Therefore, CDNB is suitable for determining whether GST activity is present, as we did using the corneal cell lines, Hemicornea construct, rabbit cornea and porcine cornea (Figs. 4 A and 5 A). GSTO1‐1 activity was determined using the highly specific substrate 4NPG. Board et al. reported that there was no detectable activity of GSTA1‐1, GSTA2‐2, GSTA4‐4, GSTM1‐1, GSTM2‐2, GSTM4‐4, GSTP1‐1, GSTT2‐2 or GSTZ1‐1 in humans. Furthermore, they found no activity of the chloride intracellular ion channel protein (CLIC‐2), a protein that is structurally related to the Omega class GSTs (Board et al., 2008), in humans. However, they found that GSTO2‐2 showed an extremely low level of 4NPG reductase activity (Board et al., 2008). GSTP1‐1 activity was evaluated using the specific substrate ethacrynic acid (Habig and Jakoby, 1981, Mannervik et al., 1985). The specificity of this substrate was confirmed in a study of mice, which in contrast to humans, express two Pi GST members. GSTP1/P2 knockout mice demonstrated no activity against ethacrynic acid (Henderson et al., 1998). However, due to possible overlapping substrate specificity, further studies were required. Therefore, the expression of GSTO1‐1 and GSTP1‐1 was determined in this study using immunofluorescence and western blotting techniques (Fig. 1, Fig. 2, Fig. 3). Only the combination of all of the results obtained can reliably demonstrate GST expression. It should be noted that the antibodies used in this study were directed against human enzymes. Therefore, a negative result obtained using animal corneas, as that obtained for the rabbit cornea using western blotting analysis, does not necessarily indicate the absence of GSTO1‐1 or GSTP1‐1 expression.