CANCER AND CELL SENESCENCE
Cells can progress to a state of senescence that has been considered to be an irreversible, essentially vegetative, state. It is only in the last few years that senescence has been recognised as a cancer protection mechanism in animals. Unsurprisingly, p53 and RB1 play central roles in promoting senescence in vivo. Thus aberrant signals in pre-malignant cells activate either the ARF-p53 or the INK4A-RB1 pathway to force cells into senescence and prevent cancer progression. Notwithstanding, when the disease develops senescence is unblocked and proliferation activated: expression of INK4A and other senescence markers is lost.
In response to oncogenic signals (DNA damage, over-expression of RAS, short telomeres) signalling pathways can be activated that lead to the up-regulation of ARF and INK4A, the combined upshot of which is that p53 and un-phosphorylated RB1 combine to drive cells into senescence (Fig. 1).
1. Regulation of p53 activity to promote senescence.
Thus is revealed the central importance of ARF, INK4A, p53 and RB1 in our cancer defences that are present as functional genes we have two very effective strategies: apoptosis and senescence. Knock out even one and tumour development is almost inevitable. Encouragingly, recent transgenic mouse experiments reveal another route to senescence, and therefore cancer protection, that is independent of ARF and p53. This involves the cyclin-dependent kinases (CDKs) that regulate the cell cycle. We’ve seen that one way in which cell division is controlled is by the sequential synthesis and breakdown of key proteins – the cyclins and also the inhibitors of CDKs. Proteins to be broken down are tagged with ubiquitin by an enzyme complex (SKP2-SCF). Abnormally high expression of SKP2-SCF is found in some human cancers, constituting incriminating evidence of cancer promoting. Knocking out SKP2-SCF activity might, therefore, be a useful therapeutic strategy. In transgenic mice, at least, this works: SKP2-SCF- defective mice are resistant to tumour development driven by oncogenic RAS or by the loss of the tumour suppressor PTEN. Instead, SKP2- SCF inactivation triggers senescence: the mice have raised levels of the CDK inhibitors WAF1 and KIP1. The really startling result, however, is that this happens even if the ARF-p53 pathway is knocked out. In other words, even in the presence of very common cancer-causing mutations, interfering with the machinery of the cell cycle can protect against cancer. A slightly different approach generated CDK2-deficient mice and showed that they too are sensitised to senescence. So senescence, like apoptosis, is a protective mechanism which response to the level at which key proteins are expressed. Even more exciting is the fact that drugs blocking the action of SKP2 or CDK2 can cause senescence in established mouse tumours, leading to regression.
Four types of DNA virus are associated with human cancers: human papillomaviruses (HPVs), hepatitis B virus (HBV), hepatitis C virus (HCV) and Epstein–Barr virus (EBV). Of the more than 100 types of HPV, 15 are oncogenic, of which the most important are types 16 and 18 that cause 70% of cervical cancers worldwide. They encode two proteins (E6 and E7) that target p53 and RB1 for degradation, thereby ablating two central tumour suppressors (Fig. 2).
2. The central axis of cancer signalling.
Hepatitis B virus, responsible for the majority of liver cancers that claim one million lives a year, also makes proteins that perturb normal cell signalling and promote genetic instability. The mechanism of HCV is unknown but it may interfere with DNA repair.
Burkitt’s lymphoma, the most common childhood cancer in equatorial Africa, arises in B cells of the lymphatic system and commonly involves the jaw or other facial bones. It is caused by the Epstein–Barr virus (EBV, also called human herpesvirus 4 (HHV-4)). The EBV genome encodes a number of proteins that can ‘immortalise’ B cells in vitro, which is thought to reflect the way in which it promotes the human lymphoma. It can also cause nasopharyngeal cancer.
Having discussed the core pathways that determine cancer susceptibility, we should now take a deep breath and face the fact that we mentioned earlier: RAS-MAPK is not the only signal pathway that can contribute to cancer. Figure 6.4 overlays the other major pathways from cell surface receptors that play important roles in cancer. They’re important because, as for RAS-MAPK, all their components have tumourigenic potential and many have been identified as oncoproteins or tumour suppressors. The about these pathways is that they work in exactly the same way as our MAPK model: they’re protein relays – it’s just that the performers are different. The salient feature of the map is that these pathways converge on the central axis. Thus, as long as the critical components of cancer defence are working, aberrant signalling by one of these pathways should activate a protection mechanism. It’s only when a core control is lost that an abnormal signal is likely to lead to a tumour cell.
Somewhat paradoxically, although signalling overall is convergent, the pathways are frequently divergent – that is, they branch and ‘cross-talk’ with other pathways. A common means of divergence is via proteins that act as ‘scaffolds’ – they have multiple binding sites for other proteins and thus provide initiation points for branching routes. We’ve already met scaffolds as the cytosolic domains of RTKs and (Fig.3) shows that they can also activate phospholipase Cγ (PLCγ), SRC and JAKs.
4. Major intracellular signalling pathways that can be involved in cancer.
However, a striking point about RAS is that it initiates four more major pathways in addition to RAF-MAPK. RAS is extraordinary in that at least ten proteins can interact with a small domain in a not very large protein, making it a major point of pathway divergence. The pathways initiated from RAS regulate essentially every aspect of cellular behaviour: transcription, apoptosis, cytoskeletal structure and the cell cycle.
In addition to RAF-MAPK, the major pathways emanating from activated RAS are PI3K (which can also be activated directly by RTKs and controls survival), mTOR (which regulates proliferation, survival, metabolism and the cytoskeleton), JNK (stress-response pathway) and CDC42 (which with other small GTPases regulates the cytoskeleton). This feature of proteins acting as multi-pathway activators is widespread in cancer signalling and we shall see in the examples of major pathways that, although the scaffolds differ, the essential mantra of cancer signalling emerges as ‘everyone talks to everyone else’.
The activity of many of these signalling pathways may be abnormal in cancers although the MAPK and the PI3K pathways appear to be the most frequently activated in a wide range of primary human tumours. To explain the principles clearly we have thus far looked only at the outlines of pathways and focused on a small number of genes and their encoded proteins as examples of the type of specific mutation that can affect pathway activity. The following summaries provide greater detail for the major pathways, together with a table listing the major ‘cancer genes’ or ‘drivers’ in which either germline mutations have been identified. It should be borne in mind that these are a small selection of over 500 oncogenes and 100 tumour suppressor genes that have been identified. However, as we’ve noted, many of these appear to be either ‘passengers’ or possibly to exert a ‘driving’ capacity only in very specific cancers.
Finally, there is another group of genes in which mutations have not been identified but have been shown to be abnormally expressed in some tumour samples. Such data are usually obtained by immunostaining tissue sections but it is often difficult to obtain control tissues for comparison and quantitation of the amounts of protein present is technically challenging. With those reservations, many proteins have been reported to be abnormally expressed in a wide range of primary and secondary tumours and cell lines. Two notable examples are the adaptor protein GRB2 and the exchange factor SOS to which it couples to regulate the activity of RAS. These proteins lie at the heart of one of the major ‘cancer pathways’ but, somewhat counter-intuitively, neither GRB2 nor SOS has been shown to undergo mutation in cancers although they have been reported to be independently over-expressed in some types of a tumour cell.