Transcriptomic Profiles of CD47 in Breast Tumors Predict Outcome and Are Associated with Immune Activation.

Targeting the innate immune system has attracted attention with the development of anti- CD47 antibodies. Anti-CD47 antibodies block the inhibition of the phagocytic activity of macrophages caused by the up-regulation of CD47 on tumor cells. In this study, public genomic data was used to identify genes highly expressed in breast tumors with elevated CD47 expression and analyzed the association between the presence of tumor immune infiltrates and the expression of the selected genes. We found that 142 genes positively correlated with CD47, of which 83 predicted favorable and 32 detrimental relapse-free survival (RFS). From those associated with favorable RFS, we selected the genes with immunologic biological functions and defined a CD47-immune signature composed of PTPRC, HLA-E, TGFBR2, PTGER4, ETS1, and OPTN. In the basal-like and HER2+ breast cancer subtypes, the expression of the CD47-immune signature predicted favorable outcome, correlated with the presence of tumor immune infiltrates, and with gene expression signatures of T cell activation. Moreover, CD47 up-regulated genes associated with favorable survival correlated with pro-tumoral macrophages. In summary, we described a CD47-immune gene signature composed of 6 genes associated with favorable prognosis, T cell activation, and pro-tumoral macrophages in breast cancer tumors expressing high levels of CD47.

Exo1 phosphorylation inhibits exonuclease activity and prevents fork collapse in rad53 mutants independently of the 14-3-3 proteins.

The S phase checkpoint is crucial to maintain genome stability under conditions that threaten DNA replication. One of its critical functions is to prevent Exo1-dependent fork degradation, and Exo1 is phosphorylated in response to different genotoxic agents. Exo1 seemed to be regulated by several post-translational modifications in the presence of replicative stress, but the specific contribution of checkpoint-dependent phosphorylation to Exo1 control and fork stability is not clear. We show here that Exo1 phosphorylation is Dun1-independent and Rad53-dependent in response to DNA damage or dNTP depletion, and in both situations Exo1 is similarly phosphorylated at multiple sites. To investigate the correlation between Exo1 phosphorylation and fork stability, we have generated phospho-mimic exo1 alleles that rescue fork collapse in rad53 mutants as efficiently as exo1-nuclease dead mutants or the absence of Exo1, arguing that Rad53-dependent phosphorylation is the mayor requirement to preserve fork stability. We have also shown that this rescue is Bmh1-2 independent, arguing that the 14-3-3 proteins are dispensable for fork stabilization, at least when Exo1 is downregulated. Importantly, our results indicated that phosphorylation specifically inhibits the 5' to 3'exo-nuclease activity, suggesting that this activity of Exo1 and not the flap-endonuclease, is the enzymatic activity responsible of the collapse of stalled replication forks in checkpoint mutants.

Checkpoint inhibition of origin firing prevents DNA topological stress.

A universal feature of DNA damage and replication stress in eukaryotes is the activation of a checkpoint-kinase response. In S-phase, the checkpoint inhibits replication initiation, yet the function of this global block to origin firing remains unknown. To establish the physiological roles of this arm of the checkpoint, we analyzed separation of function mutants in the budding yeast Saccharomyces cerevisiae that allow global origin firing upon replication stress, despite an otherwise normal checkpoint response. Using genetic screens, we show that lack of the checkpoint-block to origin firing results in a dependence on pathways required for the resolution of topological problems. Failure to inhibit replication initiation indeed causes increased DNA catenation, resulting in DNA damage and chromosome loss. We further show that such topological stress is not only a consequence of a failed checkpoint response but also occurs in an unperturbed S-phase when too many origins fire simultaneously. Together we reveal that the role of limiting the number of replication initiation events is to prevent DNA topological problems, which may be relevant for the treatment of cancer with both topoisomerase and checkpoint inhibitors.

Checkpoint-dependent RNR induction promotes fork restart after replicative stress.

The checkpoint kinase Rad53 is crucial to regulate DNA replication in the presence of replicative stress. Under conditions that interfere with the progression of replication forks, Rad53 prevents Exo1-dependent fork degradation. However, although EXO1 deletion avoids fork degradation in rad53 mutants, it does not suppress their sensitivity to the ribonucleotide reductase (RNR) inhibitor hydroxyurea (HU). In this case, the inability to restart stalled forks is likely to account for the lethality of rad53 mutant cells after replication blocks. Here we show that Rad53 regulates replication restart through the checkpoint-dependent transcriptional response, and more specifically, through RNR induction. Thus, in addition to preventing fork degradation, Rad53 prevents cell death in the presence of HU by regulating RNR-expression and localization. When RNR is induced in the absence of Exo1 and RNR negative regulators, cell viability of rad53 mutants treated with HU is increased and the ability of replication forks to restart after replicative stress is restored.

Different nucleosomal architectures at early and late replicating origins in Saccharomyces cerevisiae.

BACKGROUND: Eukaryotic genomes are replicated during S phase according to a temporal program. Several determinants control the timing of origin firing, including the chromatin environment and epigenetic modifications. However, how chromatin structure influences the timing of the activation of specific origins is still poorly understood. RESULTS: By performing high-resolution analysis of genome-wide nucleosome positioning we have identified different chromatin architectures at early and late replication origins. These different patterns are already established in G1 and are tightly correlated with the organization of adjacent transcription units. Moreover, specific early and late nucleosomal patterns are fixed robustly, even in rpd3 mutants in which histone acetylation and origin timing have been significantly altered. Nevertheless, higher histone acetylation levels correlate with the local modulation of chromatin structure, leading to increased origin accessibility. In addition, we conducted parallel analyses of replication and nucleosome dynamics that revealed that chromatin structure at origins is modulated during origin activation. CONCLUSIONS: Our results show that early and late replication origins present distinctive nucleosomal configurations, which are preferentially associated to different genomic regions. Our data also reveal that origin structure is dynamic and can be locally modulated by histone deacetylation, as well as by origin activation. These data offer novel insight into the contribution of chromatin structure to origin selection and firing in budding yeast.