Dendritic cell “The immunologist’s dirty little secret” is a secret no longer, and it’s not even so dirty any more now that we have started to understand a little about it.

The “dirty little secret” line was coined by Charlie Janeway, in a 1989 essay1 that was one of the few really revolutionary single papers in immunology. As just about anyone even peripherally involved with immunology knows, Charlie was referring to the puzzle of how immune responses start. At the time, the paradigm was simply that non-self triggered an immune response, while self did not. Yet of course that was not exactly true; to induce a real immune response, experimentally or clinically, immunologists had to include crap2 along with their non-self — the crappier the better, as far as immunity was concerned. Charlie’s proposal was that as well as non-self, an adaptive immune response needed some indication that the non-self was dangerous, and he suggested that the innate immune system must include receptors for pathogen-assocated molecular patterns (PAMPs).3 Triggering the PAMP receptor would signal a dangerous condition, and drive the adaptive response forward.

Janeway’s ideas have been proven brilliantly right, of course. The immune system does recognize PAMPs, via a bunch of receptors including (but not limited to) the dozen-plus toll-like receptors (TLRs), RIG-I, and many others. These receptors bind to various generic aspects of pathogens, like the double-stranded RNA of some viral genomes, the lipopolysaccharide in a gram-negative bacterium, or the unmethylated CpG in a bacterial genome.Charlie Janeway

But Janeway’s PAMPs concept had competition. Polly Matzinger, most famously, proposed a somewhat different “Danger” theory,4 which I’ll refer to as the, um, Danger Theory. According to this theory (and I simplify), adaptive responses happen when normal tissues are damaged — perhaps because the injection had toxic crap along with it, or perhaps simply because of tissue damage associated with the treatment. Detection of factors released by damaged cells would indicate that there was a dangerous situation, and that would trigger the adaptive response. So this is conceptually quite similar to the PAMP explanation, but relies on tissue damage rather than pathogen presence.

I never much liked Matzinger’s Danger idea, and argued against it on and off for reasons which, in hindsight, were a mixed bag; some good,5 some bad, and some merely idiosyncratic. When TLRs were flooding onto the scene, it seemed that her suggestion was going to sink without a trace. But her argument had some solidly-achored observations behind it, and a number of recent findings buoy up the Danger Theory and send it sailing triumphantly home. 6

In fact, the fundamental observation was well established: When cells are damaged, they do release factors that enhance adaptive immune responses. 7 It proved very difficult to purify the specific factor(s) responsible for this, but in 2003 Shi et al. identified one such factor as uric acid.8 I’m not going to talk about that now — it was while I was in the Rock lab, so it was just two benches away from me, a very exciting period — except to note that Yan also identified other fractions in the damaged-cell lysate that had similar activity.

That brings me to a story that has just made Nature Medicine. 9 For a grossly shortened background: It’s been known for some time that high mobility group box protein 1 (HMGB1) can be secreted by macrophages and can act as an inflammatory mediator.10 It binds to a number of TLRs, including TLR4, the receptor responsible for most of the response to LPS. A few years ago it was shown that HMGB1, like uric acid, can leak out of damaged cells and trigger inflammation.11 (I don’t know for sure if this is the high-molecular-weight fraction Shi et al. observed in their fractions, but I am pretty sure it is not.)Apetoh et al, Fig 6

One of Matzinger’s Danger Theory’s strongest points was her argument that “A model of immunity must explain transplants, tumors, and autoimmunity in order to be complete”. Tumors do induce adaptive responses, even in the absence of PAMPs. The simplest explanation is that at least in this context, the Danger Theory is quite right; as tumors outgrow their blood supply, undergo chemotherapy, or otherwise are harmed, the damaged tumor cells release endogenous danger signals that help trigger the adaptive response. Uric acid may be one of those danger signals, but the Apetoh et al. paper in Nature Medicine shows that HMGB1 is a major danger signal in tumors. Remarkably (to me, anyway) they also show that one of the major reasons chemo- and radiotherapy works is exactly this — the therapy damages the tumor and induces it to release danger signals, especially HMGB1, that signal through TLR4; this TLR4-induced inflammation enhances the adaptive response to the tumor:12

Cancer patients and their physicians who receive and apply chemotherapy, respectively, do so in the genuine belief that the prime goal of therapy is to destroy tumor cells. Here, we show for the first time that anticancer chemotherapy has an additional, decisive effect. Dying tumor cells elicit an immune response that is required for the success of therapy.

(My emphasis.) Not only is this true in mice (where you can test chemotherapy in TLR4 knockout animals: APetoh et al. showed that chemo- and radiotherapy was much less effective in these knockouts), it’s very likely true in humans as well:

The clinical relevance of these findings is underscored by the observation that the Asp299Gly TLR4 mutation, which affects the binding of HMGB1 to the receptor, has a negative prognostic impact on human patients with breast cancer.

(The figure to the right, above, shows survival of breast cancer patients with a version of TLR4 that does not bind to HMGB1 [the dark trace] compared to those with the “wild-type” TLR4, which does bind HMGB1 [grey trace]).

This is not the whole story — they make a case for calreticulin playing a part as well, for example — but if true this is not only very cool, but also a little unsettling; it kind of makes chemotherapy sound like cargo-cult medicine, or witch-doctory, trying to cause cell death even though that is only peripheral to the true therapeutic effect of priming adaptive immunity. Focusing on triggering cross-priming rather than cell death may offer new approaches to cancer therapy. The overall observation, of course, is further encouragement to the concept of cancer vaccines.


  1. Approaching the asymptote? Evolution and revolution in immunology.Janeway, C.A.Jr.. Cold Spring Harb. Symp. Quant. Biol. 54, 1-13 (1989) []
  2. Austerely called “adjuvant”, but when you start grinding up tuberculosis bacteria you lose some of your detachment from the wheel of being, you know what I’m saying?[]
  3. The crap in the needle[]
  4. Tolerance, danger, and the extended family. Matzinger P. Annu Rev Immunol. 1994;12:991-1045. []
  5. I still think her original proposal was way over-extended: “the immune system does not care about self and non-self” is hard to champion[]
  6. Much like this metaphor, in fact.[]
  7. For example, Shi, Y., Zheng, W., and Rock, K. L. (2000). Cell injury releases endogenous adjuvants that stimulate cytotoxic T cell responses. Proc Natl Acad Sci U S A 97, 14590-14595. []
  8. Shi, Y., Evans, J. E., and Rock, K. L. (2003). Molecular identification of a danger signal that alerts the immune system to dying cells. Nature 425, 516-521. []
  9. Toll-like receptor 4-dependent contribution of the immune system to anticancer chemotherapy and radiotherapy. Apetoh, L., Ghiringhelli, F., Tesniere, A., Obeid, M., Ortiz, C., Criollo, A., Mignot, G., Maiuri, M. C., Ullrich, E., Saulnier, P., Yang, H., Amigorena, S., Ryffel, B., Barrat, F. J., Saftig, P., Levi, F., Lidereau, R., Nogues, C., Mira, J. P., Chompret, A., Joulin, V., Clavel-Chapelon, F., Bourhis, J., Andre, F., Delaloge, S., Tursz, T., Kroemer, G., and Zitvogel, L. (2007). Nat Med 13, 1050 – 1059. []
  10. Wang, H., Bloom, O., Zhang, M., Vishnubhakat, J. M., Ombrellino, M., Che, J., Frazier, A., Yang, H., Ivanova, S., Borovikova, L., Manogue, K. R., Faist, E., Abraham, E., Andersson, J., Andersson, U., Molina, P. E., Abumrad, N. N., Sama, A., and Tracey, K. J. (1999). HMG-1 as a late mediator of endotoxin lethality in mice. Science 285, 248-251. []
  11. Scaffidi, P., Misteli, T., and Bianchi, M. E. (2002). Release of chromatin protein HMGB1 by necrotic cells triggers inflammation. Nature 418, 191-195. []
  12. via cross-presentation, which I really have to talk about some day[]