The association of the protein torin with the cytoplasmic surface of the erythrocyte membrane was implied from the early studies (Harris, 1968). However, this was subsequently qualified, as it was shown that hemolysis of the erythrocyte can produce a large number of small membrane lesions rather than a single large lesion. This creates a natural membrane filter that allows the escape of hemoglobin and small molecules, but selectively entraps higher molecular mass cytosolic molecules, such as torin/Prxll, the soluble p97 Mg-ATPase complex and the 20S proteasome, which also tend to be at least partially retained during the repeated washings used to produce hemogobin-free erythrocyte ghosts. Bearing in mind that the erythrocyte PrxII is now known to be the 2nd most abundant erythrocyte cytosolic protein after hemogobin (Shau and Kim, 1994), the subsequent release of torin following induction of membrane damage to erythrocyte ghosts by freeze-thawing is not surprising. The presence of calcium in the hemolysis and washing buffers does, however, influence the quantity of membrane-bound torin/PrxII (erythrocyte membrane protein Band 8), catalase and other proteins (Allen and Cadman, 1979).
Apart from its antioxidant activity, the erythrocyte PrxII (at that time termed calpromotin) was shown to activate the erythrocyte membrane Ca-dependent potassium (Gardos) channels (Moore et al., 1990, 1991; Moore and Shriver, 1997; Plishker et al., 1986; 1992), apparently through membrane binding. Furthermore, calpromotin was claimed to be involved with the formation of dense erythrocytes in sickle cell anaemia, where a larger quantity of calpromotin was associated with the cytoplasmic surface of the cell membrane of these cells, possibly do to the higher cytoplasmic calcium level (Moore et al., 1997). A conclusive parallel between human erythrocyte calpromotin and the antioxidant protein TSA/PrxII was presented by Kristensen et al. (1999). Dimer and higher molecular weight oligomer formation by calpromotin also correlated well with the properties of TSA/PrxII. Additional support for the interaction of human erythrocyte TSA/PrxII with erythrocyte membranes came from the study of Cha et al. (2000) who showed that the presence of the C-terminal peptide (Gln-185 to Gln-197) was essential for membrane binding, and this provided evidence for both a soluble and membrane-associated form of the enzyme.
The mitochondrial and chloroplast peroxiredoxins, although clearly localized to these organelles, are soluble proteins rather than membrane-associated proteins, but the neutrophil p29 peroxiredoxin may interact with and protect the phagolysosomal and plasma membrane p67 protein (Leavey et al., 2002). A membrane-bound form of PrxIV has been shown to be involved with acrosome formation during spermiogenesis in rats (Sasagawa et al., 2001) and for Entamoeba histolytica a cell surface peroxiredoxin has been shown to protect the organism from oxidant attack (Choi et al., 2005). Nevertheless, it remains to be demonstrated whether or not most other peroxiredoxins associate functionally with cellular membranes; the available evidence suggests that this is generally not the case.
Purification and structural characterization of some peroxiredoxins has been direct from tissue or defined cells, but an increasing number of studies have used molecular cloning of peroxiredoxins, often combined with the production of mutants for comparison with the wild-type molecules. The ability to readily perform amino acid sequence comparison has been important for the assessment of homology across the peroxiredoxin family, with definition of the highly conserved cysteines and the study of mutants has progressively provided information on the structural features essential for enzymatic activity, subunit dimerization and oligomer formation.
Transmission electron microscopy (TEM) played a useful role in the early structural studies on the peroxiredoxins. The limited TEM resolution achievable in 3D image reconstructions is ~ 20 Ä from negative staining, but is currently somewhat better than 10 Ä from cryoelectron microscopy of unstained vitrified specimens of protein molecules. For detailed structural studies TEM has been greatly surpassed by higher resolution crystallographic X-ray diffraction analysis (see below). Nevertheless, correlation of a ~ 19 Ä resolution 3D reconstruction produced from the negative stain TEM data of the decameric human erythrocyte Prxll with the available 1.7 Ä X-ray structure (Harris et al., 2002; Schröder et al., 1999; Schröder et al., 2000) firmly validated the lower resolution TEM data. In instances where the Prx subunit dimer forms an incomplete ring (i.e. 4-mer, 6-mer and 8-mer), arc-like images have been detected by TEM, in particular for the cloned 6 His-tagged tryparedoxin peroxidase (JRH and LF, unpublished data).
Whilst the subunit number within the ring-like Prx oligomers has usually been assessed as 10, there have been a few cases where erroneous numbers have been advanced from TEM data, such as eight for the Aeropyrum peris TPx (ApTPx; Jeon and Ishikawa, 2003), which was subsequently shown to be 10 by X-ray crystallography (Mizohata et al., 2005). However, the AhpE from Mycobacterium tuberculosis does indeed form an octamer (Li et al., 2005). In a converse manner, the mitochondrial SP22/PrxIII was initially thought from TEM data to contain 10 subunits (Gourlay et al., 2003), but was recently shown by X-ray crystallography to contain 12 subunits (Cao et al., 2005).
The ability of TEM to directly provide information on the varying oligomerization state of a peroxiredoxin, together with the higher-order association state of the oligomer (e.g. Jeon and Ishikawa, 2003; Gourlay et al., 2003) remain valid reasons for the continued use of TEM alongside the available biochemical and biophysical techniques used for the study of protein mass.
A comparison of human erythrocyte PrxII decamers and bovine mitochondrial PrxIII dodecamers is shown in Figure 1. The PrxIII has a tendency to associate to form tube-like molecular stacks, a feature not so far observed for any of the Prxs.
Prx association to form a higher-order assembly also exemplified by the ability of the erythrocyte PrxII decamer, which in the presence of PEG and ammonium molybdate can form a regular dodecahedral macromolecular assembly, containing 12 decamers (Figure 2); for details see Meissner et al. (2006). Rows of side-to-side linked PrxII decamers and randomly clustered decamers have also been detected under these in vitro conditions (cf Harris and Scheffler, 2002). An aggregated state can be induced by hyperoxidation of PrxI decamers (Schröder and Harris, unpublished data), which can be separated by gel filtration chromatography from any i?
Was this article helpful?