Cellular senescence and cancer

The proliferative lifespan of normal mammalian cells is limited by intrinsic controls, which desensitize the cell‐cycle machinery to extrinsic stimulation after a given number of cell divisions. One underlying clock driving this process of ‘replicative senescence’ is the progressive erosion of chromosome telomeres, which occurs with each round of DNA replication. This appears to trigger growth inhibition via activation of the tumour suppressor gene (TSG) product, p53, and the consequent up‐regulation of the cell‐cycle inhibitor p21 WAF1 . Other inhibitory pathways are also activated (possibly by additional clocks), including the TSG p16 INK4a and the less well‐defined complementation group genes. Loss of one pathway can be compensated, after a limited extension of lifespan, by further up‐regulation of the others, so that to escape mortality a developing tumour must overcome multiple ‘proliferative lifespan barriers’ (PLBs) by successive genetic events, each conferring a new wave of clonal expansion. This provides one explanation for the existence of multiple genetic abnormalities in human cancers; furthermore, the diversity in the nature and timing of these PLBs between different cell types may explain the variation in the spectrum of abnormalities observed between the corresponding cancers. Even if all senescence pathways are inactivated, immortalization can only be achieved if erosion of telomeres is halted, before their end‐protecting function is lost. This usually requires either activation of telomerase during tumour development, if the cell of origin is telomerase‐negative, or up‐regulation if the normal cell already has some activity, but not enough to prevent erosion. In either case, cancers often maintain near‐critical telomere lengths; hence pharmacological inhibition of telomerase remains an attractive approach to the selective killing of tumour cells. Copyright © 1999 John Wiley & Sons, Ltd.