Yes, you have to correct that picture.
PGC-1 is certainly not antiinlammatory but proinflammatory if associated with inflammation at all!
For example:
PGC-1a and ß are activators of mitochondrial biogenesis and master regulators of muskuloskeletal metabolism.
Recently, they have been suggested to modulate inflammatory responses. Studies were undertaken to unravel the role of PGC-1 isoforms on inflammatory processes. Inflammation was induced by intramuscular injection of lipopolysaccharide (LPS) or tumour necrosis factor a (TNFa) into mice transgenic for PGC-1a or ß. The results show that: 1) Overexpression of PGC-1 exerts distinct inflammatory effects. On the one hand PGC-1 reduces the basal levels of pro-inflammatory p65 (NF?B), while on the other hand it increases the basal levels of pro-inflammatory TNFa. 2) Injection of LPS or TNFa lead to strong inflammation, but the differential pattern already observed at the basal state remains. 3) LPS/TNF stimulate the infiltration of macrophages into muscle tissue more in transgenic than in wild-type mice. We conclude that PGC-1 does exert a mixed effect on inflammation. overexpression of PGC-1 stimulate, but also represses pro-inflammatory markers. The data suggest that no direct interaction between PGC-1 and inflammatory processes occurs. Rather, inflammation seems to occur secondarily and artefactually to the transgene expression. Differences observed after stimulation by LPS/TNFa represent carry-over effects of the artefactual changes in the basal state.
An old paper from 2007 where they used knockout mice for PGC-1 pretended that KO of PGC-1 increases inflammatory markers. But now, the same is found when PGC-1 is overexpressed. Probably, all this is just artefactual, as described above.