br Mammalian DGKs current knowledge Despite our
Mammalian DGKs: current knowledge Despite our understanding of prokaryotic DGKs, our limited understanding of the structure of mammalian DGKs is a major gap in our knowledge. Importantly, major differences exist between prokaryotic and mammalian enzymes often revealing the fact that structural features in the mammalian enzymes have evolved for specific functions, localizations, and regulation. This is reflected in the fact that there are 10 mammalian isoforms not including additional splice variants for some isoforms (Imai et al., 2005; Luo et al., 2004). Our knowledge of the structure of these isoforms is primarily limited to similarities in linear amino Xylazine HCl structure sequences. While limited, there is probably more structural information regarding DGK-ε than any of the other mammalian DGKs. CD spectral analysis of this enzyme indicates it is composed of 29% α-helices and 22% β-strands (Jennings et al., 2017). Additionally, a hydrophobic segment located at N-terminal residues 20–40 appears to be essential for targeting this enzyme to the endoplasmic reticulum (Matsui et al., 2014; Nakano et al., 2016). With respect to DGK-θ, there are significant differences between this enzyme and DGK-ε. DGK-ε is inhibited by PtdOH and PtdSer (Shulga et al., 2011a) while DGK-θ is activated by these lipids (Tu-Sekine and Raben, 2012). It is noteworthy that PtdOH competitively inhibits DGK-ε suggesting this lipids binds to a site in DGK-θ which is distinct from the binding site in DGK-ε. DGK-ε is also inhibited by PtdIns(4,5)P2 while this lipid has not been shown to affect DGK-θ. Further, DGK-ε is the smallest of the mammalian DGKs (approximately 64 kD), is predominantly membrane-associated, and is the only mammalian DGK that shows a fairly strong substrate preference (Lung et al., 2009). DGK-θ (approximately 110 kD) is both soluble and membrane-bound and a substrate preference has not been observed. This isoform is also activated by proteins containing a polybasic region but the endogenous activator(s) has not been identified. While linear sequences are important, they must be viewed with some caution. One noteworthy example is the C1 domains which have attracted some interest as they are known to bind phorbol esters and DAG (reviewed in (Das and Rahman, 2014)). In this, it was not unreasonable to suspect that one or more of the three C1 domains in DGK-θ is likely involved in binding catalytic DAG. The presence of these motifs, however, does not establish catalytic DAG binding. Hurley et al. analyzed 54 C1 domains including six DGKs (α, β, γ, ε, δ and ζ) and suggested that of these DGKs, only DGK-β and DGK-γ contained a C1 domain that fit a profile for phorbol ester binding (Hurley et al., 1997). Importantly, the C1 domains of DGK-θ, as well as DGKδ and η, do not bind DAG (Sakane et al., 1996; Shindo et al., 2001, 2003). The involvement of these domains in catalysis is further questioned by the observation that the drosophila DGK1 does not contain any C1 domains (Masai et al., 1992), and porcine DGK-α lacking its C1 domains retains catalytic activity with a DAG Km similar to the wild type enzyme (Sakane et al., 1996). As a result, while the C1 domains may contribute to membrane localization for some DGKs, the catalytic DAG binding site remains unresolved. While these domains are not likely to be involved in binding catalytic DAG, they may bind other lipids or participate in protein-protein interactions as suggested by Shulga et al. (2011a). In support of this notion, the C1 domain of DGKζ has been shown to mediate interactions with β-arrestins (Nelson et al., 2007) and Rac1 (Yakubchyk et al., 2005). Finally, it is important to note that there is some evidence suggesting this domain may be involved in membrane association following the activation of some G-protein coupled receptors (van Baal et al., 2005). The RA and PR domains are also poorly understood. Binding energies derived from in silico analyses suggests that the RA domain of DGK-θ does not bind Ras (Kiel et al., 2005). DGK-θ binds to, and is inhibited by the GTP-bound form of another small GTPase, RhoA, but it is not clear whether RhoA specifically binds to the RA domain. As PR domain contains a pXPXXP motif (Yu et al., 1994), it is also tempting to suggest that they bind SH3-domain containing proteins.