Modification of proteins with ubiquitin and ubiquitin-like molecules is involved in the regulation of almost every biological process. Historically, each conjugation pathway has its unique set of E1, E2 and E3 enzymes that lead to activation and conjugation of their cognate molecules. Here, we present the unexpected finding that under stress conditions, the ubiquitin E1 enzyme Ube1 mediates conjugation of the ubiquitin-like molecule NEDD8. Inhibition of the 26S proteasome, heat shock and oxidative stress cause a global increase in NEDDylation. Surprisingly, this does not depend on the NEDD8 E1-activating enzyme, but rather on Ube1. A common event in the tested stress conditions is the depletion of "free" ubiquitin. A decrease in "free" ubiquitin levels in the absence of additional stress is sufficient to stimulate NEDDylation through Ube1. Further analysis on the NEDD8 proteome shows that the modified NEDDylated proteins are simultaneously ubiquitinated. Mass spectrometry on the complex proteome under stress reveals the existence of mixed chains between NEDD8 and ubiquitin. We further show that NEDDylation of the p53 tumor suppressor upon stress is mediated mainly through Ube1. Our studies reveal an unprecedented interplay between NEDD8 and ubiquitin pathways operating in diverse cellular stress conditions.

In response to DNA strand breaks in the genome of higher eukaryotes, poly(ADP-ribose)polymerase 1 (PARP-1) catalyses the covalent attachment of ADP-ribose units from NAD(+) to various nuclear acceptor proteins including PARP-1 itself. This post-translational modification affecting proteins involved in chromatin architecture and in DNA repair plays a critical role in cell survival as well as in caspase-independent cell death. Although PARP-1 has been best-studied for its role in genome stability, several recent reports have demonstrated its role in the regulation of transcription. In this study, fluorescence spectroscopy and biochemical techniques are used to investigate the association of the amino-terminal DNA-binding domain of human PARP-1 (hPARP-1 DBD) with various DNA substrates, characterized by different DNA ends and sequence features (5'- or 3'-recessed end, double strands, telomeric repeats, and the palindromic sequence of a Not I restriction site). The correlation between the binding mode of hPARP-1 DBD to the DNA oligoduplexes and the enzymatic activation of hPARP-1 is analyzed. We show that hPARP-1 DBD binds a 5'-recessed DNA end cooperatively with a stoichiometry of two proteins per DNA molecule. In contrast, a 1:1 stoichiometry is found in the presence of a 3'-recessed end and double-strand DNA. A palindromic structure like the Not I restriction site is shown to induce protein dimerization and high enzymatic activation, suggesting that it can represent a recognition element for hPARP-1 in undamaged cells. Protein dimerization is found to be a requisite for high enzymatic activity. Taken together, our data allow further characterization of the features of hPARP-1 recognition in damaged cells and bring additional evidence that hPARP-1 may also play a role in undamaged cells.

Activation of poly(ADP-ribose) polymerase-1 (PARP-1) is an immediate cellular reaction to DNA strand breakage as induced by alkylating agents, ionizing radiation, or oxidants. The resulting formation of protein-bound poly(ADP-ribose) facilitates survival of proliferating cells under conditions of DNA damage probably via its contribution to DNA base excision repair. In this study, we investigated the association of the amino-terminal DNA binding domain of human PARP-1 (hPARP-1 DBD) with a 5' recessed oligonucleotide mimicking a telomeric DNA end. We used the fluorescence of the Trp residues naturally occurring in the zinc finger domain of hPARP-1 DBD. Fluorescence intensity and fluorescence anisotropy measurements consistently show that the binding stoichiometry is two proteins per DNA molecule. hPARP-1 was found to bind the 5' recessed DNA end with a binding constant of approximately 10(14) M(-2) if a cooperative binding model is assumed. These results indicate that hPARP-1 DBD dimerizes during binding to the DNA target site. A footprint experiment shows that hPARP-1 DBD is asymmetrically positioned at the junction between the double-stranded and the single-stranded telomeric repeat. The largest contribution to the stability of the complex is given by nonionic interactions. Moreover, time-resolved fluorescence measurements are in line with the involvement of one Trp residue in the stacking interaction with DNA bases. Taken together, our data open new perspectives for interpretation of the selective binding of hPARP-1 to the junction between double- and single-stranded DNA.