Skip to main content

Biological News

Two Drug Companies Rewrite Some Cancer Biochemistry

Mutations in the KRAS gene are a feature of some particularly hard-to-treat cancers, which has made it a target for many research programs over the years. The protein it codes for is way down deep in a lot of very important signaling pathways for cell growth and metabolism – growth factor signaling, glucose transport, activity of key kinase enzymes, and so on. The problem is, the KRAS protein itself has proven pretty much undruggable, as are (frustratingly) several others in that strong-oncology-story category. That’s led, naturally enough, to a number of searches for indirect approaches. These are tricky, though: the ones that are indirect enough to be workable may also be indirect enough not to hit the real target enough (or to be more efficient at affecting other things you didn’t want to touch). We actually end up finding out a lot about biochemical pathways in such projects, in the “Gosh, who knew that that was hooked up to that” sense.

For KRAS, one of these has been the attempt to inhibit the cellular process called autophagy. That’s a sort of recycling system, breaking down various cellular components in an orderly fashion while the cell itself is still very much carrying on. There are several different variations on the process, but it’s often a survival mechanism (for example, under starvation conditions), and it’s used by many cancer cells as yet another means to keep them going at all costs. KRAS mutations have been reported (by many research groups) as having a particular connection to the process of macroautophagy in particular, so its inhibition has been considered a good target.

But perhaps not any more. A joint effort from Novartis and Pfizer has resulted in this recent paper in PNAS, and it aims to break that connection. They were able to take several KRAS-mutant cancer cell lines, delete an enzyme (ATG7) that is widely accepted as being completely essential to macroautophagy, and watch as they carried on unaffected. They grew just as fiercely as before, didn’t seem to be particularly sensitized to radiation or several other chemical agents, and in general behaved as if having their autophagy mechanism shut down was a matter of no consequence at all.

Another interesting (and to many research groups, disturbing) part of the paper was a look at the old antimalarial compound chloroquine. It’s been accepted for some years now as an inhibitor of autophagy, both in vitro and in the clinic, where it’s used along with several chemotherapy regimes. There have been reports that it may not be as connected to that pathway as has been thought, but most of the references to it in the oncology literature have been of the “. . .use of the well-known autophagy modifier chloroquine has shown. . .” variety.

Well, guess what: when the autophagy pathway is taken out of commission in cancer cells by deleting ATG7, chloroquine still works, sensitizing cells to kinase inhibitors in the way that was thought to depend on shutting down autophagy. So the clinical use of chloroquine and its derivatives is certainly still valid, although now we’re not really sure why. Let the target hunts commence!

So how did Pfizer and Novartis end up publishing this paper? According to a press release by Novartis, it started with “a chance encounter at a conference”, the Keystone autophagy meeting in 2014. Both groups were presenting on their accumulated evidence that there was something wrong with the whole KRAS-autophagy story, and put together, they had a convincing case against something that had become a widely accepted fact. They’ve done the whole field a service by getting it all out there in one package.

A final philosophical point: so, what part of NIH-funded research did the rapacious drug companies rip off for this discovery? That’s a facetious question, of course, but a non-facetious one might be to try to figure out if this comes under the heading of “basic” research or “applied”. The lines between those two can get pretty blurry. Novartis and Pfizer were both trying to come up with cancer therapies, drugs to take to the clinic and eventually sell for money, and that sounds pretty applied. But to get there, cellular pathways had to be untangled and questioned, and that sounds like nothing else than basic research. I think that there are certainly cases where the two can be distinguished, but we should also realize that there are cases where they may not be. I’d rather not have the definition of basic research be “stuff that no one’s ever been able to think of a use for”.

Update: AstraZeneca has now published with more on the story.

22 comments on “Two Drug Companies Rewrite Some Cancer Biochemistry”

  1. Anonymous Researcher snaw says:

    Biochemical signaling pathways are what a computer programmer would call “spaghetti code” a poorly organized program in which anything might affect anything else. Wise coders organize their code so that it’s easy to understand months or years later and you say to yourself “what was I thinking when I wrote this function?” Evolution does not care about simplicity.

    1. a noumenon says:

      The machinery of life is not coded by “spaghetti code” in the computer programming sense. Since the coding must be highly adaptable, this suggests its components are actually both simple and robust. It’s the integration of the components, over x billion years, that adds layers of robustness, and complexity – this is what escapes our limited interpretation and understanding of “the code”. Both simplicity and complexity are intertwined in evolution – another way to think about the spaghetti.

    2. Mark Thorson says:

      It’s a mistake to compare biochemical pathways to software. A much closer analogy is to hardware, i.e. circuit design. If you had a diagram of all of these pathways, it would much more closely resemble the circuit diagram of an Intel microprocessor than the software which runs on it.

  2. jbosch says:

    Very interesting indeed, now I need to scratch my head: Identification of an Atg8-Atg3 protein-protein interaction inhibitor from the medicines for Malaria Venture Malaria Box active in blood and liver stage Plasmodium falciparum parasites. PMID: 24786226

    There’s also this paper: Uba1 functions in Atg7- and Atg3-independent autophagy, PMID 23873149 from 2013.

    Eng et al. LC3-1/LC3-II Western look very convincing, although some of the blots look also as if LC3 (uncleaved) is migrating with LC3-I.
    Would be interesting to know the proteome profile of those clonal lines compared to the parental line.

    1. anonymous says:

      “now I need to scratch my head: Identification of an Atg8-Atg3 protein-protein interaction inhibitor from the medicines for Malaria Venture Malaria Box active in blood and liver stage Plasmodium falciparum parasites. PMID: 24786226”
      Note that the class of small molecules reported in this paper has recently been identified as a PAINS scaffold. “Promiscuous 2-aminothiazoles (PrATs): a frequent hitting scaffold. J Med Chem. 2015, 12;58(3):1205-14. PMID: 25559643.”

      1. jbosch says:

        I know that it was flagged as a PAINS, but I don’t believe that in this particular case it is one as we’ve used multiple orthogonal screening methods plus we have additional unpublished data that are in agreement with the mechanism of action. But thanks for your comment nevertheless.

        1. sgcox says:

          Fig 3a and Fig7b in malaria paper are worrisome. There is no saturation. SPR data plotted in log(conc.) should show a classical sigmoidal curve for the specific 1;1 binding event.

      2. jbosch says:

        Would you classify 4-Bromopyrazole or the 4-Iodopyrazole also as PAINS? Pretty promiscuous binder, but seems to be selective nevertheless.

  3. ScientistSailor says:

    Off Topic, but how about some non-holiday recipe blogging? Derek, Do you have a recipe handy for Fesenjan?

  4. Dr. Manhattan says:

    THis finding neatly harks back to the recent discussion on “cloud biology” and is an illustration of just how complex cellular biochemical pathways can be and how difficult it can be to unravel them. With all the information from past papers on chloroquin and autophagy fed into the cloud computing system, you would come up with an answer, but it would most probably be wrong in light of this new experimental evidence.

    1. Phil says:

      Yep. That would be called a data integrity problem. Garbage in, garbage out.

  5. Neuroman says:

    There are many very smart chemists on this blog. Could someone offer an opinion on Cu atsm for ALS? There is a large amount of excitement about this in the entire online neurodegenerative community. The word cure is appearing.

    1. Me says:

      Looks like a play on the ‘protein misfolding due to aberrant homeostasis of metals’ theory. You’ll find people proposing this for all of the major neurodegenerative diseases. A few studies on mouse models look reasonably promising, except for the fact that they’re animal models. The good news is there are some examples of human use with the product, so a clinical trial will be relatively easy to set up – then you’ll know for certain.

  6. Daniel says:

    Sounds to me like the sort of paper that needs to get reproduced by different groups with different methods. There’s certainly something important going on, but I wouldn’t be so sure it’s what they think it is. Maybe ATG7 isn’t as essential to autophagy as people thought? Some widely-held belief is wrong (unless they didn’t actually manage to deprive the cells of ATG7 for some strange reason), but I’d wait until another essential gene also doesn’t turn out to matter before concluding that the process isn’t autophagy but chloroquine is effective anyway.

  7. Oblarg says:

    I’d certainly call this basic research, and is the type of thing we *should* be doing more of.

  8. EB says:

    We need more biologists on this blog 😉 Chloroquine is an inhibitor of lysosomal function. Lysosomes are vesicles that are downstream of the autophagy process. In some contexts inhibition of lysosomal function may lead to inhibition of autophagic vesicle formation if a simple feedback mechanism was operating. But most of the time life is not simple. Autophagy is regulated by nutrient, bacterial, lipid and other sensing pathways. So whether or not chloroquine inhibits autophagy depends on what else is going on in the cell.

    1. Matt King says:

      I’m a cell biology postdoc at AstraZeneca, investigating the role of autophagy in the resistance of cancers to targeted therapeutics. We’ve just reported in Oncogene (, that synergy between inhibitors of the mTOR pathway – which is associated with the activation of autophagy – and chloroquine, is independent of autophagy. Instead, we show that mTOR pathway inhibitors suppress cholesterol metabolism, thereby rendering lysosomes more vulnerable to osmotic stress induced by chloroquine and accelerating lysosomal cell death. We also find no specific requirement for autophagy in RAS mutant vs WT cell lines.

  9. kk says:

    Would inhibiting autophagy be a good treatment for radiation sickness ?

  10. Ernesto (Palo) says:

    Two Pharma scientist get together in a Keystone meeting (95% NIH scientists) and add a piece to thousand pieces of the autophagy puzzle, and that is an argument against the “mostly NIH-driven drug discovery” position? You’ve been very successful at attacking that position, not with this one.

    1. simpl says:

      I dont think the attack is directed at NIT scientists, but filling in some dots for quick-fix politicos who conclude, “NIH is the source of all drugs, we subsidise them, ……. and therefore drug prices should be zero”

      1. Nile says:

        You get that kind of trite “We subsidise it, we should get it free” sophistry too?

        Best answer I heard today is: “Trucks and taxis should be run for free, they run on roads we paid for!”

        A stock of quick rebuttals is a useful toolkit for anyone who needs to communicate the benefits of science to a crowd containing small but vocal clots of dangerously clever idiots.

Comments are closed.