Macrophage/microglial cells in the mouse retina during embryonic and postnatal development were studied by immunocytochemistry with Iba1, F4/80, anti-CD45, and anti-CD68 antibodies and by tomato lectin histochemistry. These cells were already present in the retina of embryos aged 11.5 days (E11.5) in association with cell death. At E12.5 some macrophage/microglial cells also appeared in peripheral regions of the retina with no apparent relationship with cell death. Immediately before birth microglial cells were present in the neuroblastic, inner plexiform (IPL), and ganglion cell (GCL) layers, and their distribution suggested that they entered the retina from the ciliary margin and the vitreous. The density of retinal microglial cells strongly decreased at birth, increased during the first postnatal week as a consequence of the entry of microglial precursors into the retina from the vitreous, and subsequently decreased owing to the cessation of microglial entry and the increase in retina size. The mature topographical distribution pattern of microglia emerged during postnatal development of the retina, apparently by radial migration of microglial cells from the vitreal surface in a vitreal-to-scleral direction. Whereas microglial cells were only seen in the GCL and IPL at birth, they progressively appeared in more scleral layers at increasing postnatal ages. Thus, microglial cells were present within all layers of the retina except the outer nuclear layer at the beginning of the second postnatal week. Once microglial cells reached their definitive location, they progressively ramified.

Autophagy is a lysosome-dependent degradative pathway frequently activated in tumor cells treated with chemotherapy or radiation. PARP-1 has been implicated in different pathways leading to cell death and its inhibition potentiates chemotherapy-induced cell death. Whether PARP-1 participates in the cell's decision to commit to autophagy following DNA damage is still not known. To address this issue PARP-1 wild-type and deficient cells have been treated with a dose of doxorubicin that induces autophagy. Electron microscopy examination and GFP-LC3 transfection revealed autophagic vesicles and increased expression of genes involved in autophagy (bnip-3, cathepsin b and l and beclin-1) in wild-type cells treated with doxo but not in parp-1(-/-) cells or cells treated with a PARP inhibitor. Mechanistically the lack of autophagic features in PARP-1 deficient/PARP inhibited cells is attributed to prevention of ATP and NAD(+) depletion and to the activation of the key autophagy regulator mTOR. Pharmacological or genetical inhibition of autophagy results in increased cell death, suggesting a protective role of autophagy induced by doxorubicin. These results suggest that autophagy might be cytoprotective during the response to DNA damage and suggest that PARP-1 activation is involved in the cell's decision to undergo autophagy.

Poly(ADP-ribose) polymerase (PARP)-1, an enzyme that catalyzes the attachment of ADP ribose to target proteins, acts as a component of enhancer/promoter regulatory complexes. In the present study, we show that pharmacologic inhibition of PARP-1 with 3,4-dihydro-5-[4-(1-piperidinyl)butoxyl]-1(2H)-isoquinolinone (DPQ) results in a strong delay in tumor formation and in a dramatic reduction in tumor size and multiplicity during 7,12-dimethylbenz(a)anthracene plus 12-O-tetradecanoylphorbol-13-acetate-induced skin carcinogenesis. This observation was parallel with a reduction in the skin inflammatory infiltrate in DPQ-treated mice and tumor vasculogenesis. Inhibition of PARP also affected activator protein-1 (AP-1) activation but not nuclear factor-KB (NF-KB). Using cDNA expression array analysis, a substantial difference in key tumor-related gene expression was found between chemically induced mice treated or not with PARP inhibitor and also between wild-type and parp-1 knockout mice. Most important differences were found in gene expression for Nfkbiz, S100a9, Hif-1a, and other genes involved in carcinogenesis and inflammation. These results were corroborated by real-time PCR. Moreover, the transcriptional activity of hypoxia-inducible factor-1A (HIF-1A) was compromised by PARP inhibition or in PARP-1-deficient cells, as measured by gene reporter assays and the expression of key target genes for HIF-1A. Tumor vasculature was also strongly inhibited in PARP-1-deficient mice and by DPQ. In summary, this study shows that inhibition of PARP on itself is able to control tumor growth, and PARP inhibition or genetic deletion of PARP-1 prevents from tumor promotion through their ability to cooperate with the activation AP-1, NF-KB, and HIF-1A.

ATM and PARP-1 are two of the most important players in the cell's response to DNA damage. PARP-1 and ATM recognize and bound to both single and double strand DNA breaks in response to different triggers. Here we report that ATM and PARP-1 form a molecular complex in vivo in undamaged cells and this association increases after γ-irradiation. ATM is also modified by PARP-1 during DNA damage. We have also evaluated the impact of PARP-1 absence or inhibition on ATM-kinase activity and have found that while PARP-1 deficient cells display a defective ATM-kinase activity and reduced γ-H2AX foci formation in response to γ-irradiation, PARP inhibition on itself is able to activate ATM-kinase. PARP inhibition induced γ H2AX foci accumulation, in an ATM-dependent manner. Inhibition of PARP also induces DNA double strand breaks which were dependent on the presence of ATM. As consequence ATM deficient cells display an increased sensitivity to PARP inhibition. In summary our results show that while PARP-1 is needed in the response of ATM to gamma irradiation, the inhibition of PARP induces DNA double strand breaks (which are resolved in and ATM-dependent pathway) and activates ATM kinase.