Fig 2: Sumoylation-deficient Prdx6K122/142 R linked to transduction protein domain (TAT) internalized in cells and exerted enhanced protective activity against oxidative stress and Sumo1 overexpression. (a and b) Confirmation of Sumoylation of mutant TAT-HA-Prdx6K122/142 R and Prdx6WT in vitro and in vivo. The in vitro Sumoylation assay was performed according to the manufacturer's protocol. Briefly, a combination of E1 enzyme, E2 (Ubc9) enzyme, Sumo1WT protein and recombinant Prdx6 protein (TAT-HA-Prdx6) WT or its mutant at K122/142 R were mixed with 20 µl reaction mixture containing Sumoylation buffer, as described in 'Materials and methods'. Reaction products were immunoblotted using anti-Sumo1, anti-HA and anti-Prdx6 polyclonal antibodies as indicated. Sumoylation of recombinant Prdx6WT protein was detected, as shown in figure, lane 1 (* denotes the Sumoylation band). In contrast, His-tagged Prdx6 mutated at K122/142 R did not reveal any detectable band (lane 2). (b) Sumoylation status of TAT-HA-Prdx6 and its mutants transduced into Prdx6-deficient LECs in vivo. Prdx6-/-LECs were transfected with pEGFP-Sumo1, and the transfectants were transduced with TAT-HA-Prdx6 or its mutant TAT-HA-Prdx6K122/142 R as indicated. Cell lysates containing equal amounts of proteins were processed for Sumo1-ELISA assay using anti-HA and antibody specific to Sumo1 as stated in the 'Materials and methods' section. Sumoylated content of Prdx6WT and its mutants proteins are presented as percentages. The data represent mean±SD from three independent experiments (*P<0.001). (c and d) LECs transduced with Sumoylation-deficient protein, TAT-HA-Prdx6K122/142 R showed higher resistance to oxidative stress-Sumo1 induced damage than did Prdx6WT. hLECs overexpressing pHA-Sumo1 were pretreated with TAT-HA-Prdx6 or TAT-HA-Prdx6K122/142 R and then exposed to H2O2. Results of ROS (c) obtained from H2-DCF-DA dye, and cell survival (d) from MTS assay showed that delivery of TAT-HA-Prdx6K122/142 R to cells significantly enhanced protection by efficiently removing ROS. The data represent the mean±SD from three independent experiments (*P<0.001). (e) Transduction of TAT-HA-Prdx6 into cells. An aliquot of 10 µg/ml recombinant protein was added to culture media and transduction of TAT-HA-Prdx6 (lane 2) and TAT-HA-Prdx6K122/142 R (lane 3) was assessed using western analysis by anti-Prdx6 antibody

Fig 3: Disruption of Sumo1 sites K122/142 R did not alter cellular localization of Prdx6. (a) Schematic diagram of the evolutionary conserved Sumo1 motifs of Prdx6 spotted by SUMOplot and Prdx6 mutants fused to GFP plasmid as shown. SUMOplot, a web-based software program (http://sumosp.biocuckoo.org/online.php), was used to examine Sumo1 conjugation motif. Sequence alignment of human, mouse, rat, monkey and cow Prdx6 protein was conducted to identify evolutionary conserved Sumoylation motif (ClustalW). Lysine (K) residues are indicated in red bold letters. (b) Localization of WT and K122R or K142 R or K122/142 R mutant of Prdx6 fused to GFP plasmid. To examine if putative Sumoylation motifs K122 or K142 in Prdx6 protein predicted by Web-based analysis is indeed responsible for Sumo1 conjugation, K residue was changed to R by using SDM and tested for localization pattern for each plasmid. Immunofluorescence images showing localization of Prdx6 and its mutant forms K122R, K142 R and K122/142 R; cells transfected with pGFP-Prdx6WT (WT, upper left panel), mutant pGFP-Prdx6K122R (K122R, upper right panel), pGFP-Prdx6K142R (K142R, lower right panel), pGFP-Prdx6K122/142R (K122/142R, lower right panel) and fluorescence images of live cells were recorded after 24 h of transfection under inverted fluorescence microscope (Nikon Eclipse Ti-U)

Fig 4: (a and b). Prdx6-deficient LECs displayed insignificantly low levels of Phospholipase A2 as well as lower GSH peroxidase activities compared with Prdx6+/+. Prdx6+/+ and Prdx6−/− LECs cultured in identical conditions as described in 'Materials and methods'. Cells were harvested and total extracts containing equal amounts of proteins were processed to measure PLA2 (a) and glutathione peroxidase activity (b) following the company's protocols. Black bars show significantly reduced PLA2 and GSH peroxidase activities in Prdx6-deficient cells. The data represent the mean±SD from three independent experiments (*P<0.001). Upper panel, a schematic illustration of active sites responsible for PLA2 (S32/H26/D140) and GSH peroxidase (C47) activities. (c and d). Disruption of Sumoylation motif K122/142 R in Prdx6 protein promoted PLA2 and glutathione peroxidase activities. Prdx6−/−LECs were transfected with pEGFP-Vector, pGFP-Prdx6WT and its mutants, pGFP-Prdx6K122R, pGFP-Prdx6K142 R and pGFP-Prdx6K122/142R fused to GFP plasmids. After 48 h, total lysates containing equal amounts of proteins were processed for PLA2 (c) and GSH peroxidase (d) activities through Enzchek PLA2 and GSH peroxidase assay kits (Invitrogen), respectively. (Prdx6WT versus mutants; *P<0.001; **P<0.05). (e and f). Recombinant mutant Prdx6K122/142 R protein had increased PLA2 and GSH peroxidase activities compared with Prdx6WT. Prdx6−/−LECs were transduced with TAT-HA-Prdx6 and its mutant TAT-HA-Prdx6K122/142 R. After 24 h, total protein was isolated and assays were performed for PLA2 (e) and Glutathione peroxidase (f) activities as described in ‘Materials and methods' section. The data represent the mean±SD from three independent experiments (*P<0.001). (g and h). Sumo1 overexpression diminished phospholipase PLA2 and GSH peroxidase activities. Prdx6−/−LECs were co-expressed with pEGFP-Sumo1 along with pEGFP-Vector or pGFP-Prdx6WT or its mutant K122/142 R, 48 h later total cell lysates containing equal amounts of proteins were utilized to measure PLA2 (g) and glutathione peroxidase (h) activities; results are presented as nmol/min/mg protein and units/min/mg protein, respectively. The data represent the mean±SD from three independent experiments (*P<0.001)

Fig 5: Sumo1 was conjugated to lysine K122 and K142 of Prdx6 in vivo. (A) Top panel, a diagrammatic illustration of Prdx6WT and its mutant plasmids. hLECs were transfected with pHA-Sumo1 along with pEGFP-Vector, pGFP-Prdx6, pGFP-Prdx6K122R, pGFP-Prdx6K142 R or pGFP-Prdx6K122/142 R plasmids. Cell lysates containing equal amounts of proteins were processed for IP using anti-Prdx6 monoclonal Ab and immunoblotted with anti-Sumo1, anti-HA (A, a), anti-Prdx6 or anti-GFP (A, b) polyclonal antibodies as described in 'Materials and methods' section. Input, visualized with anti-Sumo1 and anti-Prdx6 antibodies as shown. IP with Prdx6 monoclonal antibody shows a single-exogenous Sumoylated band at ~80 kDa (lane 4). pHA-Sumo1 co-transfected with pEGFP-Vector or pGFP-Prdx6K122/142 R does not show a Sumoylated band (Lane 5), whereas pHA-Sumo1 co-transfected with pGFP-Prdx6K122R or pGFP-Prdx6K142 R shows reduced Prdx6 Sumoylation. Data indicate both K122 and K142 in Prdx6 sumo1 motifs were targets for Sumo1 modification. (B) Confirmation of in vivo Sumoylation of Prdx6 at lysine K122 and K142. hLECs were transfected with pEGFP-Sumo1 along with Prdx6WT or its mutant K122/142 R (mutated at both sites) plasmid linked to GFP or pEGFP-Vector as indicated. Prdx6 was immunoprecipitated from cell lysates containing equal amount of proteins, and its Sumoylation was measured with anti-Prdx6 polyclonal antibody (B, a) and antibody specific to Sumo1 (B, b) as indicated. Cell lysates were prepared and subjected to IP using anti-Prdx6 monoclonal antibody. IP with Prdx6 monoclonal antibody shows single-exogenous Sumoylated band at ~100 kDa (lane 2, pEGFP-Sumo1+GFP-Prdx6). No Sumoylation band could be detected in cell extracts of pEGFP-Sumo1+pEGFP-Vector or pEGFP-Sumo1+pGFP-Prdx6K122/142 R linked GFP transfected cells (B, a and b; lanes 1 and 3) (C) Assessment of conjugation efficiency of Sumoylation motifs of Prdx6 and its mutants to Sumo1 in hLECs and Prdx6-/-LECs. (C,a) Conjugation efficiency of Prdx6 and its mutants mutated at only one and both sumo1-binding motifs to Sumo1. hLECs were transfected with pHA-Sumo1 along with pGFP-Prdx6 or its mutants K122R or K142 R or K122/142 R fused to GFP plasmid as shown. Total cell lysates containing equal amount of proteins were processed for Sumo1-ELISA assays to assess the relative efficiency of Sumoylation of Prdx6 and its mutant proteins. The data represent mean±SD from three independent experiments (*P<0.001). (C,b) Extent of Sumoylation of Prdx6WT and its mutants in vivo. Prdx6-/-LECs were co-transfected with pEGFP-Sumo1 with pGFP-Prdx6WT, pGFP-Prdx6K122R, pGFP-Prdx6K142R, pGFP-Prdx6K122/142R as indicated. Cell lysates consisting of equal amounts of proteins were processed by Sumo1-ELISA assays as described in 'Materials and methods' and as described earlier. Sumoylated content of Prdx6WT and its mutant proteins are presented in percentage. The data represent mean±SD from three independent experiments (*P<0.001)