Effect Measure

Important new flu paper in Cell: part II

In our previous post we set the stage for discussing the results of a significant new paper by Imai et al. and colleagues on the mechanism of lung damage from diverse pathogens, including SARS, bird flu H5N1, 1918 H1N1 flu, inhalational anthrax and Monkeypox. If this work is verified it is a major step forward in our understanding of how the devastating consequences of Acute Respiratory Distress Syndrome (ARDS) and Acute Lung Injury (ALI) come about and may well provide clues about how to treat what is still an essentially untreatable and catastrophic medical condition.

There are two main parts to their elegant results. The first has to do with TLR4. Imai et al. first showed that the rapid impairment of lung function in mice caused by breathing an acid chemical depended on TLR4. They did this by exposing various inbred mouse strains and finding that one, the C3H/HeJ mouse, didn’t react in the same way. It was known that this mouse also showed resistance to LPS because of a mutation in the TLR4 gene. Interesting. But the dissection of the mechanism was just beginning.

We return to the idea that TLR4 is part of cell signaling (see our previous post). While some kinds of cell signaling is relatively simple, much depends on pathways of protein – protein interactions. The receptor button gets pushed by something binding to it, say at the surface of the cell (that’s the signal). Now we need a response. The binding of the receptor by the signaling molecule causes the receptor to modify another protein which then interacts with yet another protein, etc., etc., in a potentially long pathway of interactions. The path may also branch into two or more pathways. Branching allows complicated interactions and feedbacks to produce dynamics not possible with a single straight line bucket brigade style signal.

There is a branch right after TLR4 is activated. One branch starts with another protein called MD88 which then interacts with still others until it winds up activating an important cellular protein called NFkappaB. The other branch starts with a protein called TRIF. The MyD88 branch is the best studied. Surprisingly, though, when a mutant mouse deficient in MyDD88 mice was used it showed no resistance to ALI. That suggested that the ALI (remember that means Acute Lung Injury) was being mediated by the other branch, through TRIF. Sure enough, abolishing TRIF in another strain of mutant mice restored the resistance to ALI on exposure to acid mist.

They then looked at cytokine production. Cytokines are chemicals that are used for local signaling. They are released in the immune response during inflammation (among other things) and if the release is not well regulated a runaway inflammatory response can result. Local production of cytokines is characteristic of ALI. But there are a lot of cytokines and unraveling how they all interact and what part each plays is a difficult problem. In this case Imai et al. noticed that one particular cytokine, IL-6 (interleukin 6) was related to TLR4 signaling or lack of signaling. Sure enough, mice deficient in IL-6 were much more resistant to acid mist exposure than mice capable of producing IL-6. This doesn’t mean other cytokines aren’t also involved but it does point to a role for IL-6.

If TLR4 signaling is involved, what is pushing the TLR4 button? One thought is that because lung cells are open to oxygen, maybe oxidative damage was the signal (I am skipping their experiments to rule out any hidden LPS involvement. Suffice to say there isn’t.) Cells of the innate immune system, like macrophages (present as wandering scavengers and killers of pathogens in the lung), use the production of highly reactive oxygen species to kill pathogens. They do this with a “respiratory burst,” again something that has to be controlled because the same reactive species that kills pathogens can cause damage to the host cell. TLR4 is embedded in the outside of the cell membrane (the bag that holds the cell contents). We discussed the make-up of the cell membrane in some earlier posts, connecting it specifically to the receptor the flu virus uses to glom onto cells. You can find those earlier posts here, here, here, here, here). Suffice it to say one possible consequence of a respiratory burst is some oxidative damage to the cell membrane, which is made up of a lipid bilayer composed of phospolipids (see the earlier posts for a complete explanation of phospholipids). The cell membrane is not the only place where there are abundant phospholipids in the lung, however. There is a soap-like agent in the lung, surfactant, whose main function is to reduce surface tension and make it easier for the lung to keep expanded. It is made up of 8-% to 90% phospholipid. With all these potentially oxidizable materials around, Imai et al. wondered if oxidative damage to the phospholipids might be the trigger setting TLR4 signaling in motion.

The answer is a qualified “yes.” We’ll discuss the evidence and the connection to H5N1, SARS and some other agents in the last post, next.


  1. #1 Marissa
    April 23, 2008

    “With all these potentially oxidizable materials around, Imai et al. wondered if oxidative damage to the phospholipids might be the trigger setting TLR4 signaling in motion.” One can think of many other situations in which accidental cell membrane oxidative damage could result in setting off TLR-4 signaling cascades. Oxidized polyunsaturated fatty acids incorporated into cell membranes, for instance (i.e., peroxides of linoleic acid). Then what? IL6 and TNF alpha go up. If the agent is persistent, next thing you know, you’ve got inflammation, rheumatoid arthritis or atherosclerosis.

  2. #2 revere
    April 23, 2008

    marissa: Quite right. TLR4 is at the center of research on inflammation and atherosclerosis.