Saturday, 11 August 2007

What is the mechanism behind cancer formation?

In the first post on this topic, we discussed what a cancer was. Then, in the topic's second post, we were exposed to the type of things that cause cancer. Now, at last we can turn to the question of how these carcinogens actually cause cancer. Once again, the answer will be an outline of the principles only, and once again I make no apologies for this. In my view, this is the best way to understand something, at least initially. Thereafter, the details can be successfully hung on the hook that this provides.

The first thing to understand is that cancer is a multistep process. No one genetic change will cause a cancer; rather several changes must accumulate over time. There are four major classes of genes that are disrupted in cancer formation. Let's deal with them in turn.

Firstly, we have the proto-oncogenes, which are normal genes that tend to promote cell division (cell 'growth'). When mutated in a fashion that tends to cause cancer (e.g. becoming unregulated, or autonomous), they are named oncogenes (cancer-causing genes).

Usually, cell proliferation is caused by the binding of a growth factor to its specific receptor on the cell membrane. This activates the growth factor receptor, which in turn passes the message all the way to the nucleus via several messengers. At the nucleus, the instructions required to initiate cell division are read off (transcribed) from the DNA, setting off a chain reaction that ultimately culminates in the one cell becoming two. Almost all of the genes that code for the proteins required in this complicated farrago are potential weakpoints; weakpoints that if mutated could become oncogenes.

For instance, a mutation could cause the cell to display far too many growth factor receptors. Since there is usually excess growth factor floating about, the extra receptors would mean extra signals to the cell to divide, divide, divide. (Already sounds neoplastic, doesn't it?). Or perhaps a mutation could transform one of the messangers en route to the nucleus in a way that enhanced the number of 'division signals' issued per growth factor binding.

The second class of genes that, when mutated, can contribute towards cancer formation are the tumour-suppressor genes. Under normal circumstances, this family of genes would apply the 'brakes' to cellular proliferation, limiting the process by keeping it on a short leash. A mutation that disables them would obviously promote neoplasia. An exceptionally common example is the p53 gene. Staggeringly, over 50% of human tumours contain mutations in this gene. If DNA is damaged, p53's products stop the cell's division cycle, and help repair the DNA. If the DNA can't be fixed, p53 orders the cell's self-destruction (apoptosis), rather than risk cancer. Clearly, knocking out this gene is usually a vital step in the 'strategy' of any aspirant cancer.

Next up are the genes regulating apoptosis (programmed cell death). One of the body's most formidable weapons again cancer cells is apoptosis. When a cell becomes cancerous, the changes can often be recognised by some of the surveillance immune cells. They immediately set about to destroy the abberant cell, often by forcefully initiating its own 'self-destuct' sequence. There are normal genes that promote and inhibit this, and so obviously mutations that wreck this balance in the correct way (inhibit apoptosis' promoters, or promote its inhibitors) will tend to be carcinogenic.

Lastly, there are the DNA repair genes. That cancers are actually so rare, when compared with the vast number of of DNA mutations that occur, is partially due to the action of these genes. The DNA repair gene's do just what you'd expect them to, and once again it should come as no surprise that their loss is an incredible blow to our chances of escaping cancers.

So, these are the types of genes that are affected in cancer formation, or carcinogenesis. It is worth emphasising again that malignant tumours arise only after a number of mutations have occurred. In fact, "every human cancer that has been analysed reveals multiple genetic alterations involving activation of several oncogenes and loss of two or more [tumour]-suppressor genes".

OK, so several genes need to be mutated, but surely their order doesn't matter? Actually, research indicates that, at least for some cancers, the order does matter. Certain genes must be mutated in the 'right' way first, setting up the stage for the next round of mutations.

[EDIT 17/05/09: In my opinion, an much better oveview of carcinogenesis was given in a paper by Hanahan and Weinberg.  I've summarised it here.]

Whew! OK, all of this allows us to answer some interesting questions:
  1. Why are most cancers diseases of older people? Well, in this age group, there has simply been enough time to accumulate all the necessary mutations.
  2. Why do some cancers 'run in the family'? In families with a history of certain types of cancer, the usual explanation is that they inherit one (or more) of the mutations necessary for the cancer's development. Let us assume that four mutations are required for breast cancer. People with a familial history of the disease would inherit the first mutation. Thus, they already have one strike against their name, and only need another three for cancer to develop. This would obviously make them more likely to get the cancer than those of us who still need to to get all four mutations.
  3. Can you screen for cancers genetically? In theory, yes. For example, breast cancer is associated with the loss of tumour-suppressor genes BRCA-1 and BRCA-2. It is possible to see if you have mutated versions of these tumour-suppressors, and thus if you need to be more vigilant than average towards breast cancer.

There is much more that can be said on the molecular basis for cancer, but this is more than enough for now. The next post will be shorter and more fun, I promise!


  1. "Robbins Pathological Basis of Disease", 6th edition; Cotran, Kumar, Collins
  2. "Pathology", 2nd edition; Stevens, Lowe

No comments:

Post a Comment