Activity versus peroxisomal targeting of PerCR.

Pig heart peroxisomal carbonyl reductase (PerCR) belongs to the short chain dehydrogenases/reductases (SDRs) family, a large and diverse family of enzymes found in bacteria, yeast, and multicellular animals (Kallberg et al., 2002). The discovery that enzymes that catalyzed the synthesis or inactivation of steroid hormones, such as estradiol, testosterone, and cortisol were SDRs, stimulated initial interest in this protein family (Baker, 2001; Wu et al., 2007). Currently, the PDB contains over 200 crystal structures of wild-type and mutant SDRs cocrystallized with physiological and synthetic substrates, and much is known about the mechanism of action of SDRs (Benach et al., 1998; Tanaka et al., 1996; Wu et al., 2007). SDRs have a catalytically active tyrosine, which forms a triad with a highly conserved lysine and serine at the catalytic site. Of structural and functional importance is the presence of the catalytically active tyrosine in a helix F, which has a hydrophobic exterior surface, and along with a helix E, forms the stabilizing intersubunit interface (Benach et al., 1998; Tanaka et al., 1996; Tsigelny and Baker, 1995) in SDR dimers and tetramers. With one exception (Ghosh et al., 2001; Tsigelny and Baker, 1995), all SDRs are active as either dimers or tetramers. Moreover, the exception contains an extra segment with a hydrophobic a helix that forms an ‘‘internal dimer interface’’ with a helix F (Ghosh et al., 2001). It makes sense that SDR monomers would be catalytically inactive because exposure of the hydrophobic surface of a helix F to water would disrupt its structure and the configuration of the essential tyrosine in the catalytic site (Tsigelny and Baker, 1995). It would appear that with such an extensive body of information about SDR structure and function that there is not much more to learn from these enzymes. In this issue, however, studies by Tanaka et al. (Tanaka et al., 2008) on pig heart PerCR demonstrate that SDRs have more to teach us about the relationship between structure and biological mechanisms, such as protein transport to organelles. Moreover, Tanaka et al.’s report is an excellent demonstration of the value and, in this case, the necessity of having the crystal structure of pig heart PerCR to explain puzzling biochemical data, and in the process uncover a novel mechanism for regulating the trafficking of oligomeric proteins to peroxisomes. Pig heart PerCR localizes to peroxisomes, as expected in view of the SRL sequence at the carboxyl terminus (Tanaka et al., 2008). SRL is a type 1 peroxisomal targeting sequence (PTS1) and is a variant of SKL, the canonical PTS1 (Leon et al., 2006). SHL, another PTS1 variant, is at the carboxyl terminus of dog liver PerCR. Tanaka et al. (2008) undertook straightforward biochemical and molecular studies to confirm that SRL functioned as a PTS1 in pig PerCR by constructing SLL and SL mutants, which are known to lack PTS1 function (Maynard and Berg, 2007; Swinkels et al., 1992). Transfection of HeLa cells with cDNA for pig PerCR, or mutants with SLL and SL and with SKL and SHL as controls, gave the expected results. The SLL and SL mutants were not targeted to peroxisomes and were enzymatically inactive. Only the SKL and SHL mutants were targeted to peroxisomes. Both mutants were enzymatically active, but the SKL variant was less stable. A distinguishing feature of protein transport across the peroxisomal membrane is that folded and oligomeric proteins are transported across this membrane, with or without noncovalently-bound cofactors (Subramani, 2002). In fact, many peroxisomal proteins are multimeric and several are known to contain bound cofactors. If oligomeric proteins are imported into peroxisomes, then it follows that their PTSs must be available for binding to the PTS receptors in the oligomeric state. This is borne out by binding studies showing that dimeric dihydroxyacetone synthase DHAS (which has the C-terminal sequence, NKL) can indeed interact with Pex5 (PTS1 receptor) from Hansenula polymorpha (Faber et al., 2002). Unexpectedly, introduction of the wildtype pig PerCR tetramer directly into HeLa cells did not lead to peroxisomal localization (Tanaka et al., 2008). What prevented the transport of pig PerCR to the peroxisome? To answer this question, Tanaka et al. (2008) crystallized pig PerCR with NADPH. It is a homotetramer, with a structure that resembles that of other SDRs including a dimer interface consisting of a helices E and F from each subunit. Pig PerCR contains the triad of tyrosine, lysine, and serine at the catalytic site. Binding of the coenzyme NADPH to pig PerCR resembles that of other SDRs. It is the structure of the C-terminal SRL that explains the perplexing observation that introduction of pig PerCR protein into HeLa cells does not lead to peroxisomal localization. The 3D structure shows that SRL is at the tetramer interface, where it is extensively structured via hydrogen bonding, presumably shielding the PTS1 from interactions with Pex5 (Tanaka et al., 2008). Tanaka et al. (2008) propose that SRL is exposed to solvent in the PerCR monomer, which would allow binding to Pex5 and transport into the