Mitochondria are devoid of poly(ADP-ribose)polymerase-1, but harbor its product oligo(ADP-ribose)

There are conflicting data about localization of poly(ADP-ribose)polymerase-1 and its product poly(ADP-ribose) in mitochondria. To finally clarify the discussion, we investigated with biochemical and cell biological methods the potential presence of poly(ADP-ribose) polymerase-1 in these organelles. Our data show that endogenous and overexpressed poly(ADP-ribose)polymerase 1 is only localized to the nucleus with a clear exclusion of cytosolic compartments. In addition, highly purified mitochondria devoid of nuclear contaminations do not contain poly(ADP-ribose)polymerase-1. Although no poly(ADP-ribose)polymerase-1 enzyme is detectable in mitochondria, a shorter variant of its product poly(ADP-ribose) is present, associated specifically with a small subset of mitochondrial proteins as revealed by immunoprecipitation and protein fingerprint analysis. These proteins are located at key-points of the Krebs-cycle, are chaperones involved in mitochondrial functionality and quality-control, and are RNA-binding proteins important for transcript stability, respectively. Of note, despite the fact that especially poly(ADP-ribose)polymerase-1 is its own major target for modification, we could not detect this enzyme by mass spectrometry in these organelles. These data suggests a new way of targeted nuclear-mitochondrial signaling, mediated by nuclear poly(ADP-ribosyl)ation dependent on poly(ADP-ribose)polymerase-1.