Mystery Rays from Outer Space

Meddling with things mankind is not meant to understand. Also, pictures of my kids

November 29th, 2007

“Darwin’s Surprise” in the New Yorker

Eustace TilleyThe New Yorker has an article on endogenous retroviruses that leaves me with mixed emotions. I’ve subscribed to the New Yorker for 20-odd years now, even hanging on through the lean Tina Brown years, because I love the quality of the non-fiction articles. (And the cartoons.) Their non-fiction is usually well-written and well-researched and usually does a terrific job of bringing across the importance and the interest of a field, whether it’s virology, steel production, or stocking grocery shelves. And in general, I think this article (“Darwin’s Surprise” by Michael Specter) does that as well. There was little in the article that was new to me, but I think that people unfamiliar with the field will have no trouble understanding it.

But here are some quotes from the first page, that made me slap down the magazine and mutter to myself.

Viruses reproduce rapidly and often with violent results, yet they are so rudimentary that many scientists don’t even consider them to be alive.

Because they no longer seem to serve a purpose or cause harm, these remnants have often been referred to as “junk DNA.”

And later on:

… the earliest available information about the history and the course of human diseases, like smallpox and typhus, came from mummies no more than four thousand years old. Evolution cannot be measured in a time span that short.

Those lines are at best lazy, and at worst bullshit.

The whole “viruses aren’t alive” crap isn’t an argument of scientists — in the 20 years I’ve been a virologist, I have never heard a scientist ask the question. Public school teachers, maybe high school biology teachers, maybe philosophers (I don’t know, I’m guessing) might ask the question, but scientists don’t. That’s because it has nothing to do with science. It’s a semantics question. People new to biology think it tells them something profound about What Life Is. It tells you nothing about life; it tells you about “life”, the word. The word “alive” dates back some 800 years. Of course it isn’t suited to explaining viruses. The word “virus” comes from the Latin for “Poison”; that tells us nothing interesting about virus lifestyles.

The whole “junk DNA” has been thrashed out a dozen times (see Genomicron for a good start). The bottom line? If you search Pubmed for the phrase “junk DNA” you will find a total of 80 articles (compare to, say, 985 articles for “endogenous retrovirus”); and a large fraction of those 80 articles only use the phrase to explain what a poor term it is. Scientists don’t use the term “junk DNA’. Lazy journalists use it so they can sneer at scientists (who don’t use it) for using it.

And of course, evolution can easily be measured in four thousand years. For a virus? Eighty years (the history of HIV), 1 or 8 years (West Nile Virus) 2 is plenty.

Aside from those points that bumped my Pet Peeve Bone, the article is a good one. But those just rubbed me the wrong way.

(Having written this, I see that Carl Zimmer at The Loom has beaten me to it.)


  1. Korber, B., Muldoon, M., Theiler, J., Gao, F., Gupta, R., Lapedes, A., Hahn, B. H., Wolinsky, S., and Bhattacharya, T. (2000). Timing the ancestor of the HIV-1 pandemic strains. Science 288, 1789-1796. []
  2. Brault, A. C., Huang, C. Y., Langevin, S. A., Kinney, R. M., Bowen, R. A., Ramey, W. N., Panella, N. A., Holmes, E. C., Powers, A. M., and Miller, B. R. (2007). A single positively selected West Nile viral mutation confers increased virogenesis in American crows. Nat Genet 39, 1162-1166. []
November 29th, 2007

Cancer equilibrium, part II

Colorectal cancer
Blood vessels in a colorectal cancer

As I said in my last post , by the time we can detect a tumor — by the time it’s macroscopic — it’s already been through a long period of selection by the immune system. We see only the survivors of that selection, the cancers that have developed resistance to immune destruction. What happens to the other tumors, that have not yet escaped?

Some of them are eliminated by the immune system. Rarely, if ever, do we observe this; it happens when there are a handful of pre-cancerous cells, a microscopic cluster of a few cells that have made a few steps toward outright carcinogenicity but are not detectable by almost any means we have available today. But if the “long period of selection” is accurate, then there is a third category; as well as those tumors that have escaped control by the immune system and those that are eliminated with it, there must be a set of tumors that are in equilibrium with the immune system.

These equilibrium tumors, again, probably consist only of a handful of cells. The immune system recognizes them and destroys them, but because of some genetic or epigenetic changes in the tumor1 the fledgling tumor is not destroyed; neither can it grow out to be detectable, until at some point it develops further changes that allow it to escape immune control altogether. It’s quite likely that most of the tumors that we eventually see, spend most of their lifespan in the equilibrium stage.

That’s the theory, anyway. As I mentioned last post, there’s more solid evidence for the “Elimination” and “Escape” stages, but there is some circumstantial evidence for “Equilibrium”. Now, a very recent paper by Robert Schreiber’s group looks at the equilibrium stage directly.

Catherine M. Koebel, William Vermi, Jeremy B. Swann, Nadeen Zerafa, Scott J. Rodig, Lloyd J. Old, Mark J. Smyth, Robert D. Schreiber (2007). Adaptive immunity maintains occult cancer in an equilibrium state Nature, 450 (7171), 903-907 DOI: 10.1038/nature06309

The problem with testing the theory is (among other things) that the tumors in equilibrium are, pretty much by definition, undetectable. If they were detectable, we’d say they had escaped. What’s more, another part of the “equilibrium” definition is that it’s a long-term interaction. So not only must you detect the undetectable, you must do this over a long period.

Blogging on Peer-Reviewed ResearchKoebell et al overcame this, at the cost of making the system more artificial, by treating mice with a low dose of a carcinogen. Some mice developed tumors, but eventually after a couple of hundred days new tumors stopped appearing. Many mice were still tumor-free. 2 The question then is this: Are those mice truly tumor-free, or so they actually have tiny, undetectable tumors that are in equilibrium with their immune systems?

Koebell et al, Figure 1 To test this, Koebell et al now shut down the immune systems of the tumor-free mice. Sure enough, tumors abruptly started to grow out in previously clean mice (the red traces in the figure to the right).

This (along with some further controls I won’t go in to) argues that the mice were harboring tumors in equilibrium with the immune system. Knowing this, could these proto-tumors be detected?

In fact, the sites of carcinogen injection often did have tiny, but non-progressing, lumps, a few millimeters around. I would guess that such tiny firm lumps would usually be dismissed as scar tissue from the injection, and so indeed some proved to be; but some of them turned out to look like cancerous cells. And these cells formed progressive tumors in immune-deficient mice — but not mice with an immune system; even though normal immune systems had not been able to eliminate them previously.

The paradox that stable masses from our MCA-treated immunocompetent mice often contained transformed cells but did not increase in size in vivo suggested that net tumour cell expansion was being immunologically restrained.

Is this equilibrium stage a dead end for the tumor? Do unsuccessful tumors get shunted into this stage and either smolder indefinitely, or eventually get eliminated? At least in this model — and very likely in naturally-occurring cancer, though that remains to be proven — equilibrium seems to be a true intermediate stage that can lead to outright tumor growth. Although most of the spontaneous, detectable tumors appeared early on after carcinogen treatment, and after 200 days or so it usually took immune suppression for the tumors to grow out, in a few mice even after several hundred days tumors spontaneously appeared. And these tumors, that spontaneously left the equilibrium stage in normal mice — these tumors were not rejected by normal mice, unlike (many of) the cells from the equilibrium-stage tumors. 3

We show that equilibrium is indeed a component of cancer immunoediting because tumour cells in equilibrium are highly immunogenic (unedited), whereas those spontaneously exiting equilibrium that become growing tumours have attenuated immunogenicity (edited)-results that place this process temporally between elimination and escape.

This has lots of interesting implications. Are tumors in equilibrium potential targets for cancer prevention — say, cranking up immune responses to shift the balance toward elimination? Do tumors in equilibrium have some kind of molecular signature that will make their detection feasible in natural situations? Precisely what changes allow the tumors to progress from equilibrium to escape — and can those changes be targeted, to give the immune system a hand?


  1. Or perhaps because of pure luck and location[]
  2. Actually, many had “small stable masses” where they had been injected with the carcinogen, so they weren’t quite tumor free — see below — but they had no progressive tumors.[]
  3. I notice that although wild-type mice usually rejected the cells from the stable equilibrium tumors, a significant minority actually did go on to form tumors. It’s not clear to me from the figure whether these were progressive tumors or tumors in equilibrium — but in either case, either in the rejection or in the take, something changed in the balance when the equilibrium tumor cells were transplanted into new hosts, even genetically identical ones. I am very interested in what that change in balance is.[]
|