PerspectiveImmunology

Dangers In and Out

See allHide authors and affiliations

Science  27 Mar 2009:
Vol. 323, Issue 5922, pp. 1683-1684
DOI: 10.1126/science.1172794

Every organism faces a bewildering array of threats, pathogens foremost among them. But how does the immune system distinguish between an infection and trauma? Both elicit similar inflammatory immune responses. On page 1722 of this issue, Chen et al. (1) explain why and how the immune system responds appropriately in either scenario.

All mammals have an impressive arsenal of molecules and cells specialized to fight pathogens. Adaptive immunity, in the form of antibody production by B cells and the instruction of “killer” or cytotoxic T cells, is a critical component of the body's defenses. However, this is only a second line of defense that selectively recognizes microbes attacking for a second time. Without a first line of defense—innate immunity—mammals would succumb to pathogens still unrecognized by B and T cells.

The broad outlines of our current understanding were first sketched 20 years ago by Charles Janeway (2), starting from the idea that the immune system cannot recognize pathogens individually, because the information required is huge and would rapidly become obsolete. Pathogens continually evolve, confounding the ability of the immune system to recognize them. Instead, immune cells recognize broad molecular patterns rather than detailed features of specific pathogens. Such pathogen-associated molecular patterns (PAMPs) comprise molecular structures that are found in microbes but not in host tissues. Moreover, PAMPs are essential for the survival or the pathogenicity of microbes; thus, they cannot simply do away with PAMPs to evade recognition by the immune system.

Toll-like receptor 4 (TLR4) was the first receptor to be identified that recognizes PAMPs (3). It recognizes lipopolysaccharide, a component of the outer membrane of Gram-negative bacteria. As the name implies, it is related to Toll, a receptor involved in pathogen recognition in the fly Drosophila melanogaster. Thus, Toll-like receptors are evolutionarily ancient, although their number in mammals has grown to about one dozen.

When cells of the innate immune system—such as macrophages, mast cells, natural killer cells, and dendritic cells—encounter a PAMP, they secrete cytokines and chemokines, soluble molecules that signal “danger” to other cells. The most immediate response is inflammation at the site of infection, and the recruitment of additional immune cells, including neutrophils. Importantly, neutrophils “shoot at first sight,” releasing reactive oxygen species and proteases, thereby causing extensive collateral damage to the host tissue. Usually, the intruding pathogens are eliminated, and the adaptive immune system “remembers” their identity in case the organism is reinfected by the same pathogens; eventually, the tissue is reconstructed and healed. Inflammation, therefore, is a damaging but essential response, and becomes a problem only when it is excessive, or persists (chronic inflammation).

This picture of pathogen infection is complicated by the fact that physical trauma (such as a wound or a broken bone) causes many of the same effects as invading pathogens, including inflammation. Indeed, pathogens are thought to be initially recognized by the immune system precisely because they cause tissue damage (4). How, then, does the immune system recognize tissue damage? Damage-associated molecular patterns (DAMPs) were postulated as the counterparts to PAMPs, with the important distinction that DAMPs should be endogenous—the body's own molecules—just as the PAMPs should be pathogen-borne and thus exogenous. For example, the molecule high mobility group box 1 (HMGB1) fits the hypothetical description of an endogenous danger signal (5) and instructs adaptive immunity in ways similar to those elicited by exogenous danger signals (6). HMGB1 is a component of chromatin, the DNA-protein complex that makes up chromosomes, and thus normally resides in the cell nucleus. The release of HMGB1 by cells that have died as a result of tissue trauma signals “danger” to neighboring cells and to the immune system. Importantly, HMGB1 is also recognized by the pattern recognition receptors TLR2, TLR4, and TLR9, as well as by the receptor of advanced glycation endproducts, another “danger” receptor (79). Thus, trauma and pathogens (DAMPs and PAMPs) engage the same immune cell receptors, neatly explaining why they elicit the same inflammatory responses, although the molecular details are still largely unknown.

Yet, some outcomes must be different between the two scenarios. Chen et al. explored how these partly different outcomes occur by inducing liver necrosis in mice with an excess of acetaminophen. This treatment causes the release of HMGB1, and thus leads to inflammation in the absence of any pathogen. The authors found that mice lacking CD24, a membrane protein expressed by immune and stem cells, developed an inflammatory response that was more powerful and lethal than that in wild-type mice. In fact, HMGB1 was found to directly associate with CD24 and Siglec-G, a member of the sialic acid-binding immunoglobulin-like lectin family. Mice lacking Siglec-G also were sensitive to inflammation due to acetaminophen-induced liver necrosis. CD24 does not contain a cytosolic domain, and signals through Siglec-G, which contains an immune receptor tyrosine-based inhibitory motif (ITIM). ITIMs are cytosolic domains that reduce activation of nuclear factor κB), which is a transcription factor activated by both DAMPs (10) and PAMPs (11) and is essential for many aspects of the inflammatory response, including the secretion of cytokines and chemokines. The CD24-Siglec complex also recognizes heat shock proteins, another class of endogenous danger signals, but does not respond to lipopolysaccharide or poly-(dI:dC), two exogenous danger signals.

Signaling pathways negatively regulate Toll-like receptor responses to PAMPs as a control for excessive inflammation during infection. It now appears that endogenous danger signals activate a different “braking circuit” that is specific for DAMPs (see the figure). This dampens the immune response to injury and limits collateral damage to the tissue. Interestingly, both HMGB1 and Toll-like receptors appeared early in evolution, but only more modern vertebrates have CD24 and Siglecs. It appears that this particular braking circuit is an add-on to the ancient activating system.

Danger signals.

Exogenous and endogenous signals, such as bacterial and host cell molecules, respectively, elicit the inflammatory response through the same Toll-like receptors on immune cells. However, a specific signaling pathway limits the response to endogenous signals. This may prevent a runaway immune response to injuries.

CREDIT: C. BICKEL/SCIENCE

Can the braking circuit also moderate adaptive immunity (which first appeared with fishes), so as to avoid autoimmune responses? CD24 has already been implicated in autoimmunity, and genetic variations in CD24 influence the susceptibility to autoimmune diseases, including multiple sclerosis and lupus (12). The CD24-Siglec system might also respond to the various complexes that HMGB1 forms with lipopolysaccahride, interleukin-1β, single-stranded DNA, and nucleosomes. At least one of these complexes, nucleosome-bound HMGB1, is implicated in dendritic cell activation and the production of autoantibodies (13). Can the HMGB1-CD24-Siglec system limit the more severe forms of sterile inflammation, such as sepsis? The growing insight into how the immune system distinguishes between internal and external danger is likely to have a substantial impact on therapeutic approaches.

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
View Abstract

Stay Connected to Science

Navigate This Article