Otal cell lysates from the transfected cells probed with the antibodies indicated in the Figure. Data represent the mean 6 SD from 3 independent experiments (C) Neonatal rat cardiomyocytes were nucleofected with the vectors indicated in the Figure. In agreement with the findings in A and B, FOG-2-4KR showed stronger MedChemExpress GSK864 repression activity than the wt, while the SUMO-1-FOG-2-4KR fusion demonstrated significantly reduced repression capacity. Of note, co-expression of GATA-4 in cardiomyocytes did not activate the BNP promoter significantly most likely due to quenching by the presence of endogenous GATA-4. In addition, the repression exerted by FOG-2 and FOG-2-4KR reduced luciferase activity to levels lower than the buy GW0742 reporter itself, again suggesting that the promoter alone was activated by endogenous GATA-4 and that FOG-2 repressed this activity. Data represent the arithmetic mean from 23727046 2 independent experiments. IB, immunoblot; ns, not significant; nr, no reporter; S-1, SUMO-1; F2m, FOG-2-4KR; S-1F2m, SUMO-1-FOG-2-4KR. doi:10.1371/journal.pone.0050637.gHaving shown that mutation of the SUMO acceptor lysines in FOG-2 led to enhanced repression capacity, we wished to corroborate the observations using an artificially SUMOylated molecule. It has been shown that mimicking SUMOylation by fusing SUMO to a substrate can recapitulate to a large extent the effects of SUMO modification at the natural target sites [33]. To this end, we fused SUMO-1 at the N terminus of mutant FOG-2 (SUMO-1-FOG-2-4KR) and tested the transcriptional activity of this chimeric construct. Fig. 6B shows that expression of SUMO1-FOG-2-4KR abolished the capacity of FOG-2-4KR to repress GATA-4-mediated transcription, thus implicating SUMOylation in a mechanism that leads FOG-2 to alternate between a repressive and a more permissive transcriptional status. Even though SUMO fusion proteins are artificial and probably exhibit an aberrant level of SUMOylation (the fusion protein is constantly SUMOylated), the fact that SUMO-1-FOG-2-4KR reversed the repression activity of FOG-2-4KR strongly implies that SUMOylation attenuates FOG-2-mediated repression. We next examined whether SUMOylation is relevant for the transcriptional activity of FOG-2 in cardiac cells. AmaxaH nucleofection technology was used to co-transfect the expression vectors indicated in Fig. 6C into neonatal rat cardiomyocytes. The transfection efficiency was determined visually by co-transfection of a GFP expression vector. The data shown in Fig. 6C substantiates the observations in HeLa cells, with FOG-2-4KR demonstrating augmented repression capacity and the SUMO-1FOG-2-4KR chimera neutralizing the repressive competence. Moreover, co-expression of increasing amounts of SUMO-1 in HeLa cells reduced the repression activity of wt FOG-2 but not that of FOG-2-4KR (Fig. 7A). As anticipated from their function, co-expression of the SUMO-specific de-SUMOylating enzymes SENP-1 and SENP-8 resulted in the abrogation of FOG-2 SUMOylation (Fig. 7C, lanes 3 and 4). Notably, co-expression of both SENP-1 and SENP-8 also led to a significant increase in FOG-29s repression capacity in the presence of SUMO-1 (Fig. 7B). Altogether, the data imply that absence of SUMOylation renders FOG-2 a more effective transcriptional repressor.co-expression of increasing amounts of GATA-4 resulted in a corresponding increase in FOG-2 SUMOylation (Fig. 8A, lanes 2 to 4 and Fig. 8B). This is reminiscent of the increase in FOG-1 SUMOylation seen in the p.Otal cell lysates from the transfected cells probed with the antibodies indicated in the Figure. Data represent the mean 6 SD from 3 independent experiments (C) Neonatal rat cardiomyocytes were nucleofected with the vectors indicated in the Figure. In agreement with the findings in A and B, FOG-2-4KR showed stronger repression activity than the wt, while the SUMO-1-FOG-2-4KR fusion demonstrated significantly reduced repression capacity. Of note, co-expression of GATA-4 in cardiomyocytes did not activate the BNP promoter significantly most likely due to quenching by the presence of endogenous GATA-4. In addition, the repression exerted by FOG-2 and FOG-2-4KR reduced luciferase activity to levels lower than the reporter itself, again suggesting that the promoter alone was activated by endogenous GATA-4 and that FOG-2 repressed this activity. Data represent the arithmetic mean from 23727046 2 independent experiments. IB, immunoblot; ns, not significant; nr, no reporter; S-1, SUMO-1; F2m, FOG-2-4KR; S-1F2m, SUMO-1-FOG-2-4KR. doi:10.1371/journal.pone.0050637.gHaving shown that mutation of the SUMO acceptor lysines in FOG-2 led to enhanced repression capacity, we wished to corroborate the observations using an artificially SUMOylated molecule. It has been shown that mimicking SUMOylation by fusing SUMO to a substrate can recapitulate to a large extent the effects of SUMO modification at the natural target sites [33]. To this end, we fused SUMO-1 at the N terminus of mutant FOG-2 (SUMO-1-FOG-2-4KR) and tested the transcriptional activity of this chimeric construct. Fig. 6B shows that expression of SUMO1-FOG-2-4KR abolished the capacity of FOG-2-4KR to repress GATA-4-mediated transcription, thus implicating SUMOylation in a mechanism that leads FOG-2 to alternate between a repressive and a more permissive transcriptional status. Even though SUMO fusion proteins are artificial and probably exhibit an aberrant level of SUMOylation (the fusion protein is constantly SUMOylated), the fact that SUMO-1-FOG-2-4KR reversed the repression activity of FOG-2-4KR strongly implies that SUMOylation attenuates FOG-2-mediated repression. We next examined whether SUMOylation is relevant for the transcriptional activity of FOG-2 in cardiac cells. AmaxaH nucleofection technology was used to co-transfect the expression vectors indicated in Fig. 6C into neonatal rat cardiomyocytes. The transfection efficiency was determined visually by co-transfection of a GFP expression vector. The data shown in Fig. 6C substantiates the observations in HeLa cells, with FOG-2-4KR demonstrating augmented repression capacity and the SUMO-1FOG-2-4KR chimera neutralizing the repressive competence. Moreover, co-expression of increasing amounts of SUMO-1 in HeLa cells reduced the repression activity of wt FOG-2 but not that of FOG-2-4KR (Fig. 7A). As anticipated from their function, co-expression of the SUMO-specific de-SUMOylating enzymes SENP-1 and SENP-8 resulted in the abrogation of FOG-2 SUMOylation (Fig. 7C, lanes 3 and 4). Notably, co-expression of both SENP-1 and SENP-8 also led to a significant increase in FOG-29s repression capacity in the presence of SUMO-1 (Fig. 7B). Altogether, the data imply that absence of SUMOylation renders FOG-2 a more effective transcriptional repressor.co-expression of increasing amounts of GATA-4 resulted in a corresponding increase in FOG-2 SUMOylation (Fig. 8A, lanes 2 to 4 and Fig. 8B). This is reminiscent of the increase in FOG-1 SUMOylation seen in the p.
