The finding that RIPK1, an integral protein in cell-death pathways, also functions to preserve the body's epithelial-cell barriers challenges the idea that cell death and survival are regulated by distinct factors. See Letters p.90 & p.95
The saying, “while you do not know life, how can you know death?”, from the great Chinese philosopher Confucius, encapsulates the human desire to understand matters of life and death. Beyond philosophical curiosity, at the cellular level the balance between life and death is essential for health. Mutations in genes that regulate cell death have been found to cause human diseases ranging from inflammation and autoimmunity to cancer. Normally, cell-death genes have dedicated roles in either cell survival or cell death. But in this issue, two groups1,2 show that receptor interacting protein kinase 1 (RIPK1), a factor known to promote cell death, has a paradoxical function in supporting the survival of the epithelial cells that line our body's cavities and skin surfaces.
Since the discovery of RIPK1 almost 20 years ago3, the proposed function of the protein has alternated between that of a cell-death promoter and a cell-death inhibitor. Much of what we know about RIPK1 biology comes from studies of its activation by the tumour-necrosis factor (TNF) receptor, a cell-surface protein that, when bound by TNF molecules, contributes to inflammation, cell differentiation and cell death. Genetically modified mice that lack RIPK1 die shortly after birth4,5,6, and this was thought to be due to failure to stimulate NF-κB, a cellular factor that turns on the expression of survival genes. Consistent with this idea, extensive cell death and inflammation in multiple tissues occur in these RIPK1-deficient mice.
But, in contrast to this survival function, RIPK1 also binds to and activates FADD and caspase 8 — two members of a protein complex that triggers a form of cell death called apoptosis7. To complicate matters further, RIPK1 can also partner with its cousin RIPK3 to induce an inflammatory form of cell death termed necroptosis8. These apparently conflicting functions of RIPK1 have puzzled researchers for years, and we are still struggling to join the dots on how RIPK1 deficiency leads to perinatal lethality.
To try to fill this gap, Dannappel et al.1 and Takahashi et al.2 generated mice lacking RIPK1 specifically in the intestinal epithelium. These mice survived longer after birth than mice with whole-body deficiency of RIPK1, but they developed severe inflammatory bowel disease within the first few weeks of life and died as a result. This colitis was largely due to the heightened sensitivity of the mice to TNF. Histological examination revealed that the intestinal epithelial cells in these animals had undergone extensive apoptosis.
In contrast to previous reports9,10, the activity of the survival factor NF-κB was not significantly affected in the RIPK1-deficient cells. Hence, the sensitivity of these cells to death was not a result of impaired NF-κB. Previous work had shown11,12 that the function of RIPK1 as a kinase enzyme is essential for cell death by apoptosis and necroptosis, but not for its stimulation of NF-κB. However, both Dannappel et al. and Takahashi et al. found that the kinase function of RIPK1 is also not responsible for the severe colitis, because mice engineered to express a kinase-inactive version of RIPK1 developed normally and showed no abnormalities. Nonetheless, the severe colitis was fully reversed when FADD–caspase-8-mediated apoptosis and RIPK3-dependent necroptosis were both inactivated. These results indicate that a yet-to-be defined, kinase-independent function of RIPK1 is responsible for protecting intestinal cells from injury.
Dannappel and colleagues also found that deletion of RIPK1 only in the skin epidermis led to psoriasis-like inflammation. However, unlike in the intestine, blocking necroptosis alone was sufficient to prevent this skin inflammation. Inhibition of FADD–caspase-8-mediated apoptosis and RIPK3-dependent necroptosis has also previously been found to prevent the perinatal lethality of mice with whole-body RIPK1 deficiency4,5,6. Together, these results indicate that RIPK1 promotes cell survival by inhibiting apoptosis and necroptosis and, hence, that the protein has the enigmatic role of both a promoter and an inhibitor of cell death (Fig. 1).
How does RIPK1 orchestrate these diametrically opposing signals? Both groups found that, on TNF stimulation, RIPK1-deficient cells lost expression of the proteins cIAP1, TRAF2 and cFLIP — proteins that switch on the pro-survival factor NF-κB. Moreover, active NF-κB can further increase the levels of these factors. Thus, RIPK1 may protect cells by preserving the integrity of survival proteins such as cIAP1, TRAF2 and cFLIP. In this regard, it is noteworthy that these proteins are regulated by ubiquitination, a process that often tags proteins for degradation by a macromolecular structure called the proteasome. Because the kinase function of RIPK1 is not required for cell survival, it is possible that RIPK1 forms a protective 'scaffold' that shields the survival proteins from ubiquitination and proteasomal degradation. Alternatively, RIPK1 may directly inhibit FADD–caspase-8-mediated apoptosis and RIPK3-dependent necroptosis.
Regardless of the molecular mechanism involved, it is peculiar that, in RIPK1-deficient mice, embryonic development is normal and cell injury and inflammation manifest themselves only at birth. What might be the physiological cue that triggers these events after birth? Newborns encounter many changes in their environment, including altered oxygen levels, exposure to ambient light and colonization of the intestine by commensal bacteria. This bacterial colonization can strongly influence the development of immunity13. Surprisingly, although Takahashi et al. showed that antibiotic treatment to eliminate commensal bacteria protected RIPK1-deficient mice from lethal colitis, Dannappel et al. found that colitis still developed in antibiotic-treated mice and in germ-free mice. The reason behind these discrepant observations is not known. Moreover, although TNF contributes significantly to the colitis in RIPK1-deficient mice, it is not the only driver of the disease. It now remains for biologists to take up the gauntlet to decipher these puzzles and to identify the physiological cues and mechanisms that trigger RIPK1-dependent survival responses.
Dannappel, M. et al. Nature 513, 90–94 (2014).
Takahashi, N. et al. Nature 513, 95–99 (2014).
Stanger, B. Z., Leder, P., Lee, T.-H., Kim, E. & Seed, B. Cell 81, 513–523 (1995).
Rickard, J. A. et al. Cell 157, 1175–1188 (2014).
Dillon, C. P. et al. Cell 157, 1189–1202 (2014).
Kaiser, W. J. et al. Proc. Natl Acad. Sci. USA 111, 7753–7758 (2014).
Edinger, A. L. & Thompson, C. B. Curr. Opin. Cell Biol. 16, 663–669 (2004).
Moriwaki, K. & Chan, F. K.-M. Genes Dev. 27, 1640–1649 (2013).
Kelliher, M. A. et al. Immunity 9, 297–303 (1998).
Ting, A. T., Pimentel-Muiños, F. X. & Seed, B. EMBO J. 15, 6189–6196 (1996).
Chan, F. K.-M. et al. J. Biol. Chem. 278, 51613–51621 (2003).
Wang, L., Du, F. & Wang, X. Cell 133, 693–703 (2008).
Hooper, L. V., Littman, D. R. & Macpherson, A. J. Science 336, 1268–1273 (2012).