We recently identified polynucleotide phosphorylase (PNPase) as a potential binding partner for the TCL1 oncoprotein. Mammalian PNPase exhibits exoribonuclease and poly(A) polymerase activities, and PNPase overexpression inhibits cell growth, induces apoptosis, and stimulates proinflammatory cytokine production. A physiologic connection for these anticancer effects and overexpression is difficult to reconcile with the presumed mitochondrial matrix localization for endogenous PNPase, prompting this study. Here we show that basal and interferon-␤-induced PNPase was efficiently imported into energized mitochondria with coupled processing of the N-terminal targeting sequence. Once imported, PNPase localized to the intermembrane space (IMS) as a peripheral membrane protein in a multimeric complex. Apoptotic stimuli caused PNPase mobilization following cytochrome c release, which supported an IMS localization and provided a potential route for interactions with cytosolic TCL1. Consistent with its IMS localization, PNPase knockdown with RNA interference did not affect mitochondrial RNA levels. However, PNPase reduction impaired mitochondrial electrochemical membrane potential, decreased respiratory chain activity, and was correlated with altered mitochondrial morphology. This resulted in F o F 1 -ATP synthase instability, impaired ATP generation, lactate accumulation, and AMP kinase phosphorylation with reduced cell proliferation. Combined, the data demonstrate an unexpected IMS localization and a key role for PNPase in maintaining mitochondrial homeostasis.

Polynucleotide phosphorylase (PNPase) is an exoribonuclease and poly(A) polymerase postulated to function in the cytosol and mitochondrial matrix. Prior overexpression studies resulted in PNPase localization to both the cytosol and mitochondria, concurrent with cytosolic RNA degradation and pleiotropic cellular effects, including growth inhibition and apoptosis, that may not reflect a physiologic role for endogenous PNPase. We therefore conducted a mechanistic study of PNPase biogenesis in the mitochondrion. Interestingly, PNPase is localized to the intermembrane space by a novel import pathway. PNPase has a typical N-terminal targeting sequence that is cleaved by the matrix processing peptidase when PNPase engaged the TIM23 translocon at the inner membrane. The i-AAA protease Yme1 mediated translocation of PNPase into the intermembrane space but did not degrade PNPase. In a yeast strain deleted for Yme1 and expressing PNPase, nonimported PNPase accumulated in the cytosol, confirming an in vivo role for Yme1 in PNPase maturation. PNPase localization to the mitochondrial intermembrane space suggests a unique role distinct from its highly conserved function in RNA processing in chloroplasts and bacteria. Furthermore, Yme1 has a new function in protein translocation, indicating that the intermembrane space harbors diverse pathways for protein translocation.

No abstract

Sensory hair cells in the cochlea, like most neuronal populations that are postmitotic, terminally differentiated, and non-regenerating, depend on robust mechanisms of self-renewal for lifelong survival. We report that hair cell homeostasis requires a specific sub-branch of the DNA damage nucleotide excision repair pathway, termed transcription-coupled repair (TCR). Cockayne syndrome (CS), caused by defects in TCR, is a rare DNA repair disorder with a broad clinical spectrum that includes sensorineural hearing loss. We tested hearing and analyzed the cellular integrity of the organ of Corti in two mouse models of this disease with mutations in the Csb gene (CSB m/m mice) and Csa gene (Csa Ϫ / Ϫ mice), respectively. Csb m/m and Csa Ϫ / Ϫ mice manifested progressive hearing loss, as measured by an increase in auditory brainstem response thresholds. In contrast to wild-type mice, mutant mice showed reduced or absent otoacoustic emissions, suggesting cochlear outer hair cell impairment. Hearing loss in Csb m/m and Csa Ϫ / Ϫ mice correlated with progressive hair cell loss in the base of the organ of Corti, starting between 6 and 13 weeks of age, which increased by 16 weeks of age in a basal-to-apical gradient, with outer hair cells more severely affected than inner hair cells. Our data indicate that the hearing loss observed in CS patients is reproduced in mouse models of this disease. We hypothesize that accumulating DNA damage, secondary to the loss of TCR, contributes to susceptibility to hearing loss.