[Nescafe]

Hi,

Reading into Anticancer drugs and their main targets it seems that most of them target a key player in DNA replication. While this makes sense to me I need someone to explain to me advantages and disadvantages of targeting DNA replication v.s a cytoplasmic protein that regulates it. I get that targeting DNA replication mechanism means a lot of healthy cells dying but wouldn't the advantage be that we are targeting the heart of the problem whereas targeting the protein may or may not be enough to push the balance towards cell death.

 I just want to hear someone else share their thoughts with me, all this stuff is a tangled network of information in my head and I need someone's help to untangle it!

Thanks.

Nescafé.

[fledarmus]

Pick any protein that has anything to do with DNA replication and do a search in PubMed. I can almost guarantee you will find at least one research group trying to target that protein as a means of controlling cancer.

There are two major problems in fighting cancer. One is that cancer cells are so similar to human cells, because that is what they started out as. It is fairly easy to design a chemical that will kill bacterial cells without killing human cells, because it is easy to find proteins bacteria use and humans don't. It is far, far harder to find proteins that cancer cells use that normal human cells don't. The historical answer to this problem is to find pathways that cancer cells use much more than human cells, and shut those pathways off for a time. Theoretically the cancer cells will die faster than the normal cells, and then you can stop the treatment and the remaining normal cells will recover. Since the most obvious difference between cancer cells and normal cells is that cancer cells keep dividing, the obvious answer is to target all the pathways involved in cell replication. Any rapidly dividing cells in the body will be killed off, and once the cancer is gone, and the treatment stopped, the slowly dividing cells will recover. This accounts for many of the typical side effects of cancer drugs, because the human body has many normal cell processes that are fast dividing, mostly in the hair follicles and in the stomach lining. So people undergoing this type of cancer treatment usually lose their hair and have serious problems with digestion and nausea.

The second major problem is that the cancer cells themselves are so different, not only in different types of cancer and in different types of people, but even within the same tumor. Once the mechanisms to prevent or repair mutations have been bypassed or switched off, the cancer cells themselves are following a very rapid Darwinian evolution, and you can get a wide variety of different cancer cells with different processes switched on or switched off to promote cell growth and replication. Targeting any specific pathway may only kill a fraction of the cancer cells in any particular population or even in any particular tumor. Cancer is not just once disease, it is many diseases even in a single person. To be effective against all cancers, a drug would have to target a process required by all cancer cells, and finding one of those that isn't also required by all normal cells is incredibly difficult.

Many solutions currently under development have given up on trying to differentiate between cancer and normal cells - they are simply very toxic processes. Instead, investigators are exploring targeted delivery systems. Even if you have a compound which is extremely toxic to every cell it comes in contact with, it won't do too much damage if you can make sure that it is only distributed to cancer cells. This is the chemical equivalent to surgery. There are some purely biochemical processes that involve recognizing differences in the surface between cancer cells and normal cells, and using antibodies to direct the poisons only to the cancer cells, but again, cancer cells are not only very similar to normal cells, but have a very heterogeneous pattern of differences. There are also some mechanical processes, such as using high-powered magnets to guide the compounds to the tumor site, or implanting the drugs directly into the tumor site.

The hardest point to get across is that cancer is not one disease, even in a single human, and it is very unlikely that a single treatment will ever be discovered. We need to be able to target a broad spectrum of disorders, and we can only do that by targeting an enormous variety of biochemical pathways.

[Nescafe]

Fledarmus,

That was an amazing explanation. You have no idea how much that helped me.


Perhaps you can help me figure out another issue I am having. It is with regards to targeting protein kinases.

In the 3rd edition of introduction to medicinal chemistry it states "All the protein kinase inhibitors of clinical interest have been designed to bind to this region (ATP binding pocket) of the active site rather than the substrate binding site" - Why not target the substrate binding site? I actually have no clue and have not been able to find anything. I thought in order to obtain good selectivity you would want to target the substrate pocket cause the ATP binding pockets would be too similar and therefore your drug less selective. But the book argues that the ATP binding pockets between different kinases have found to be more different from one another than initially thought so it is possible to obtain selectivity by targeting the ATP pockets. I get that, but they never explain why not target the substrate binding pocket :S

Thanks,

Nescafe.

[Nescafe]

Just to add, at the end of the chapter the book writes " It is worth noting that most research has been carried out on ATP mimics, but there is potential in designing inhibitors that bind to the substrate binding region. Many researchers feel that such inhibitors could be more selective and safer to use".

So why start with ATP mimics in the first place? More established? easier to make drugs that mimic it etc?

Nescafe.

[fledarmus]

Excellent questions.

The simple answer is that ATP is a small molecule, while the protein kinase substrates are large proteins. This means that to match the strength of ATP binding to the enzyme, you only need to work with a few isolated and well-defined interactions, while to match the strength of a protein binding to the enzyme you may need to work with many more.

In small molecule drug design, the key to making a strong, selective inhibitor is to build in a few (usually no more than four or five) functional groups that can interact very strongly with specific points on the enzyme, and to organize them geometrically in space so they are in the perfect positions to interact. In this way, the small molecule can mimic the interactions of ATP to the enzyme but interact stronger, out competing the ATP, and have a few other geometrically specific interactions that will ensure that it interacts only with the ATP binding site of a particular enzyme and not with any of the other enzymes that bind ATP.

Protein-protein binding usually has a much larger number of far weaker interactions than protein-small molecule binding. To get a strong enough interaction with the protein substrate binding site of the kinase to out-compete the substrate, you might need to attach  15, 20, or even more specific functional groups in specific geometric locations. Also, small molecule binding sites on enzymes tend to be in pockets or clefts on the enzyme which bring the interacting groups close together, while protein binding sites tend to be spread out across larger sheets. This means you need to build much larger molecules to attain the same binding affinity.

And finally, there is the history of drug discovery. Since ATP binding sites were the easiest sites to find and target, there has been a lot of chemistry done around ATP binding sites in all of the major pharmaceutical companies, and there compound libraries have large numbers of ATP mimics. One of the first things a large pharmaceutical company usually does when it starts a new program is to conduct a high-throughput screen of its compound library against the target enzyme. If the target enzyme has an ATP binding site, the company will find a lot of hits among its ATP binding analogues, which then become leads for new synthetic efforts. Very few companies have a wide variety of large molecules with multiple interacting sites available for high-throughput testing, and for many technical reasons, they are rarely useful drug leads anyway, so the chances of them finding a compound in their libraries which will bind to the substrate binding site are vanishingly small. Consequently, the lead development stage of their research will most likely start with trying to find a selective ATP binding site molecule.

There is a lot of very exciting research going on in drug development against protein-protein binding interactions that is begining to change some of these points. For one thing, fragment screening both in vitro and in silico is maturing to the point where we may be able to isolate just a few critical sites among the many that proteins use to interact, which are critical to the interaction, and which can be bound up strongly with a small molecule of the appropriate geometry. For another, technologies which allow the use of much larger molecules as drugs are beginning to show promise, and may change the definition of "small molecule" drug discovery and the nature of high-throughput libraries. The next ten or fifteen years may completely change the paradigm and lead to a new focus on drug design against protein-protein interactions.

[Nescafe]

Thank you sir! Fruitful, as always.. =)

Nescafe.