Upon the aberrant activation of oncogenes, normal cells can enter the cellular senescence program, a state of stable cell-cycle arrest, which represents an important barrier against tumour development in vivo. Senescent cells communicate with their environment by secreting various cytokines and growth factors, and it was reported that this 'secretory phenotype' can have pro- as well as anti-tumorigenic effects. Here we show that oncogene-induced senescence occurs in otherwise normal murine hepatocytes in vivo. Pre-malignant senescent hepatocytes secrete chemo- and cytokines and are subject to immune-mediated clearance (designated as 'senescence surveillance'), which depends on an intact CD4(+) T-cell-mediated adaptive immune response. Impaired immune surveillance of pre-malignant senescent hepatocytes results in the development of murine hepatocellular carcinomas (HCCs), thus showing that senescence surveillance is important for tumour suppression in vivo. In accordance with these observations, ras-specific Th1 lymphocytes could be detected in mice, in which oncogene-induced senescence had been triggered by hepatic expression of Nras(G12V). We also found that CD4(+) T cells require monocytes/macrophages to execute the clearance of senescent hepatocytes. Our study indicates that senescence surveillance represents an important extrinsic component of the senescence anti-tumour barrier, and illustrates how the cellular senescence program is involved in tumour immune surveillance by mounting specific immune responses against antigens expressed in pre-malignant senescent cells.

Microglia are crucial for immune responses in the brain. Although their origin from the yolk sac has been recognized for some time, their precise precursors and the transcription program that is used are not known. We found that mouse microglia were derived from primitive c-kit(+) erythromyeloid precursors that were detected in the yolk sac as early as 8 d post conception. These precursors developed into CD45(+) c-kit(lo) CX(3)CR1(-) immature (A1) cells and matured into CD45(+) c-kit(-) CX(3)CR1(+) (A2) cells, as evidenced by the downregulation of CD31 and concomitant upregulation of F4/80 and macrophage colony stimulating factor receptor (MCSF-R). Proliferating A2 cells became microglia and invaded the developing brain using specific matrix metalloproteinases. Notably, microgliogenesis was not only dependent on the transcription factor Pu.1 (also known as Sfpi), but also required Irf8, which was vital for the development of the A2 population, whereas Myb, Id2, Batf3 and Klf4 were not required. Our data provide cellular and molecular insights into the origin and development of microglia.

Microglia are crucially important myeloid cells in the CNS and constitute the first immunological barrier against pathogens and environmental insults. The factors controlling microglia recruitment from the blood remain elusive and the direct circulating microglia precursor has not yet been identified in vivo. Using a panel of bone marrow chimeric and adoptive transfer experiments, we found that circulating Ly-6C(hi)CCR2(+) monocytes were preferentially recruited to the lesioned brain and differentiated into microglia. Notably, microglia engraftment in CNS pathologies, which are not associated with overt blood-brain barrier disruption, required previous conditioning of brain (for example, by direct tissue irradiation). Our results identify Ly-6C(hi)CCR2(+) monocytes as direct precursors of microglia in the adult brain and establish the importance of local factors in the adult CNS for microglia engraftment.

No abstract