anti-inflammatory functions of the “apoptotic” caspases

Ann. N.Y. Acad. Sci. ISSN 0077-8923

ANNALS OF THE NEW YORK ACADEMY OF SCIENCESIssue: Clearance of Dying Cells in Healthy and Diseased Immune Systems

Anti-inflammatory functions of the “apoptotic” caspases

David Wallach, Tae-Bong Kang, Akhil Rajput, Jin-Chul Kim, Konstantin Bogdanov,Seung-Hoon Yang, and Andrew KovalenkoDepartment of Biological Chemistry, The Weizmann Institute of Science, Rehovot, Israel

Address for correspondence: Dr. David Wallach, Department of Biological Chemistry, The Weizmann Institute of Science,76100 Rehovot, Israel. [email protected]

The two main known functions of the caspases act antagonistically in regulating inflammation. “Inflammatory”caspases trigger inflammation by catalyzing the processing of IL-1β precursors and other proinflammatory cytokines.In contrast, “apoptotic” caspases safeguard against the triggering of inflammation by imposing a cell-death formthat withholds release of alarmins by dying cells and dictates generation of anti-inflammatory mediators. Theseantagonizing functions are exerted by evolution-related mechanisms. Studies of the function of caspase-8, an enzyme-mediating apoptotic cell-death induction in response to TNF-family ligands, reveal that it blocks inflammation inadditional ways. One way is by restricting activation of the RIG-I complex by foreign ribonucleic acid. Chronicskin inflammation in mice with caspase-8–deficient epidermis is associated with constitutive activation of the RIG-Icomplex in keratinocytes. This activation is apparently prompted by nucleic acids released from epidermal cells thatdisintegrate during cornification, and becomes chronic because it is not restricted by caspase-8.

Keywords: apoptosis; caspase; inflammation; IRF3; necrosis; RIG-I

“Inflammatory” and “apoptotic” caspases

The caspase family of cysteine proteases is knownmainly for two distinct activities. One is triggeringof inflammation, which occurs through the process-ing of precursors of inflammatory cytokines, suchas interleukin (IL)-1�, known to be mediated inhumans by caspases 1, 4, and 5 (and antagonizedby caspase-12). The other is induction of apoptoticdeath, with the participation of caspases 2, 3, 6, 7, 8,9, and 10.1,2

Despite the differing spectra of their protein sub-strates and the marked differences in the functionalconsequences of cleavage of their substrates, the ini-tiator caspases of these two groups share similarstructures and similar mechanisms of activation. Inboth groups, activation is initiated at a distinct N-terminal region of the protein—the “prodomain”—which comprises one of two related structural mo-tifs of the “death-fold” group: either the caspase re-cruitment domain (CARD) domain or, in the caseof caspase-8 and -10, the death-effector domain. Inboth groups, these prodomain motifs initiate aggre-

gation and hence activation of the caspases by bind-ing, through homo-association of death-fold mo-tifs, to an oligomer of upstream regulatory proteins.In the case of the activation of caspases that medi-ate death in response to intracellular inducers (theintrinsic cell-death pathway), structural similaritiesto the “inflammatory” caspases extend beyond thecaspases themselves. Apoptotic protease activatingfactor 1 (Apaf1), the adapter protein that im-poses generation of the “apoptosome”—the molec-ular complex that accommodates the activation ofcaspase-9, which initiates the intrinsic cell-deathpathway—is structurally and evolutionarily relatedto the NLR (nucleotide-binding domain, leucinerich) proteins, the adapters that bring together theinflammasome complexes in which the inflamma-tory caspases are activated.1,2

In early essays about the caspases, their two activi-ties known at that time—triggering of inflammationand initiation of apoptosis—were presented as un-related entities, and the fact that in the course ofevolution members of the same enzyme family hadbecome involved in both activities was assumed to

doi: 10.1111/j.1749-6632.2010.05742.xAnn. N.Y. Acad. Sci. 1209 (2010) 17–22 c© 2010 New York Academy of Sciences. 17

Anti-inflammatory roles of “apoptotic” caspases Wallach et al.

be coincidental. With our growing understanding ofthe functional significance of the apoptotic process,however, it has become evident that these two activ-ities serve a common role, namely the regulation ofinflammation, although in antagonistic ways.

The role of death induction by theapoptotic caspases in restrictinginflammation

Whereas our perceptions of the various other formsof cell death, as well as their mechanisms and phys-iological significance, are still in a state of flux, ourconcepts of apoptosis have remained largely un-altered since this form of death was first defined.The main morphological features initially noticedin apoptotic cells—contraction of the dying cell andmaintenance of its membrane integrity, as opposedto the bulging of cells that die in other ways andthe rupture of their membranes3—are still viewedas hallmarks of apoptosis and as crucial determi-nants of its physiological function. Accumulatingevidence supports the notion that these structuralfeatures of the apoptotic process, as well as the ob-served rapid uptake and consumption of apoptoticcells by phagocytic cells, serve to safeguard againstleakage of the contents of the dying cells into theirsurrounding milieu and that if such leakage doesoccur, certain intracellular proteins released fromthe dying cells act as “alarmins” (cytokine-like in-ducers of inflammation).4 There is also persuasiveevidence that apoptosis, in addition to safeguard-ing against induction of inflammation by accidentalspillage from dying cells, actively blocks inflamma-tion through the effects of cell-surface componentsof the apoptotic cells.5

Signaling for necrotic death and itsrestriction by caspases

For some years after scientists became aware of theexistence of the apoptotic form of death, apopto-sis was considered as the only way in which pro-grammed cell death occurred. Necrosis was believedto develop in a totally erratic manner, merely re-flecting the failure of mechanisms underlying vitalprocesses. As it became increasingly evident thatnecrotic death constitutes a prominent trigger ofinflammation and can thus have far-reaching impli-cations for the fate of the organism, more attentionwas given to the possibility that this process too iscontrolled by specific molecules and mechanisms.

Tumor necrosis factor (TNF), a cytokine that caninduce apoptotic death by activation of caspase-8and probably also of caspase-10,6 has long beenknown also to induce necrotic death.7 As with otherapoptotic processes, in the induction of apoptosisby TNF or any of the other major death-inducingligands of the TNF family, activation of the initi-ating caspases presupposes their induced aggrega-tion, which occurs through interactions of death-fold motifs. These include death-effector domains,which are found in the prodomains of the cas-pases as well as in the adapter protein, Media-tor of Receptor-induced Toxicity-1/Fas-Associatedprotein with Death Domain (MORT1/FADD),and death domains, which are also found inMORT1/FADD as well as in two additional adapterproteins—Tumor necrosis factor Receptor type 1-Associated Death Domain (TRADD) and receptorinteracting protein (RIP1)—and in the intracellu-lar domains of the receptors to which the death-inducing ligands bind.6,8 RIP1 also mediates the in-duction by TNF of necrotic death; however, it doesso through its enzymatic activity, serine/threoninephosphorylation, which does not participate in itsinduction of apoptotic death.9,10 The necrotic deathinduced by RIP1 also requires the action of an-other serine/threonine protein kinase, RIP3, which,though lacking a death domain, shares several struc-tural similarities with RIP1 and is capable of asso-ciating with it.11–13 This type of death inductiondoes not require caspase activation; on the con-trary, blocking of caspase action strongly potenti-ates the induction of necrotic death by ligands ofthe TNF family, suggesting that caspases might neg-atively regulate necrosis.14 Since caspase-8 can cleaveRIP1,15 it seems plausible that its inhibitory effecton necrosis occurs by blocking the initiation of thenecrotic process. Mechanisms related to those thatinitiate TNF-induced necrosis also seem to partic-ipate in some other necrotic processes.16 The in-hibitory effects of caspases on such mechanisms are,therefore, likely to affect a spectrum of physiologicaland pathophysiological processes in which inflam-mation is triggered as a result of necrotic cell death.

Direct regulation of inflammationby caspase-8

As mentioned above, caspase-8 serves as a proximalenzyme in the pathway of apoptotic cell-death in-duction by ligands of the TNF family (the extrinsic

18 Ann. N.Y. Acad. Sci. 1209 (2010) 17–22 c© 2010 New York Academy of Sciences.

Wallach et al. Anti-inflammatory roles of “apoptotic” caspases

Figure 1. Ubiquitous chronic inflammation in mice expressing an enzymatically inactive caspase-8 transgene. Tissue specimens ofmice with one wild-type and one enzymatically inactive caspase-8 allele demonstrate multifocal cellular inflammatory infiltrationin diaphragmatic parietal pleura, lung visceral pleura, and interstitium of pancreas and lung (arrows). Scale bar, 100 �m. FromKovalenko et al.20

cell-death pathway).6 In the mouse, it is apparentlythe only enzyme serving this function, whereas inhumans, caspase-10 might act redundantly with it.When knocking out caspase-8 in mice, we were sur-prised to discover that the knockout was lethal inutero.17 Since no ligand of the TNF family serves anyvital role, this finding was unexpected, and it raisedthe possibility that caspase-8 also serves functionsthat are unrelated to the activity of these ligandsand perhaps distinct from its role in death induc-tion. Subsequent studies in mice with conditionalknockout of caspase-8 confirmed that this enzymeindeed serves various nonapoptotic functions.18

On examining the functional consequences ofcaspase-8 deletion in hepatocytes, we noticed thatpartial hepatectomy in these mice prompted hy-pertrophy of the liver. This response could be as-cribed to the development of a chronic inflamma-tory state.19 Bacterial artificial chromosome (BAC)transgenic mice that express an enzymatically in-active caspase-8 allele (in addition to one normalallele) were also found to develop chronic inflam-mation, which could be discerned in a variety ofinternal organs (Fig. 1) as well as in the skin.20 Wesubsequently found that specific ablation of caspase-8 in the skin triggers a severe chronic inflammatoryskin disorder (Fig. 2).20 Similar inflammation wasobserved in mice with caspase-8 deletion in other ep-ithelia (not shown). Altogether, these findings sug-gested that one of the functions served by caspase-8is to withhold inflammation.

On examining the impact of caspase-8 deficiencyon the functions of various signaling pathways thatcontrol inflammation, we found that in several celltypes caspase-8 deficiency greatly potentiated acti-vation of the transcription factor interferon regula-tory factor 3 (IRF3) by intracellular foreign nucleicacid [transfected DNA (Fig. 3), double-stranded

RNA, or RNA viruses (Rajput, A., A. Kovalenko,K. Bogdanov, S-H. Yang, J-C. Kim, T-B. Kang, andD. Wallach, in preparation)]. As a result, inductionof various IRF3-dependent genes by these insults

Figure 2. Epidermis-specific knockout of caspase-8 triggers afatal chronic skin inflammation. Staining for macrophages (anti-F4/80 antibody), eosinophils (phenol-red uptake), granulocytes(anti-Gr-1 antibody), and T lymphocytes (anti-CD3 antibody)was performed on skin sections from mice whose epidermisat P7 expressed caspase-8 (Casp-8F/+K5-Cre mice) or did notexpress caspase-8 (Casp-8F/−K5-Cre mice). Scale bar, 100 �m.From Kovalenko et al.20

Ann. N.Y. Acad. Sci. 1209 (2010) 17–22 c© 2010 New York Academy of Sciences. 19


Figure 3. Caspase-8–deficient keratinocytes display an en-hanced response to transfected DNA. (A and B) Effect on geneactivation. (A) Effect of caspase-8 knockdown. Relative expres-sion of the IFN-� gene (ratio of the IFN-� transcript in caspase-8 knockdown keratinocytes to the same transcripts in mock-transfected cells) assayed in keratinocytes from one pair of P1littermates that were transfected in duplicate, 1 day after plat-ing, with either control small interfering RNA (siRNA) (blue)or caspase-8 siRNA (maroon) and, 72 h later, transfected withpoly(dA-dT) for the indicated times. Shown are mean expressionratios ± SD for the respective duplicate transfections. (B) Effectof caspase-8 knockout. Expression of the indicated transcripts atdifferent times after DNA transfection of keratinocyte culturesestablished from epidermis of two pairs of P1 Casp-8F/+K5-Creand Casp-8F/−K5-Cre littermates, 1 day after plating. Cells weretransfected either with poly(dA-dT) or with calf thymus DNA,as indicated. (C) Effect on Isg15 expression. Western blot anal-ysis of the expression of Isg15 protein at different times afterpoly(dA-dT) transfection of keratinocytes from Casp-8F/+K5-Cre and Casp-8F/−K5-Cre epidermis from two pairs of P1 litter-mates, 1 day after plating, as in A. Analysis of the expression of�-actin served as a loading control. From Kovalenko et al.20

was strongly enhanced. Analysis of the mechanismsaccounting for this enhancement pointed to theretinoic acid-inducible gene I (RIG-I) signalingcomplex as one of the targets for caspase-8 action.This signaling complex mediates the activation ofseveral transcription factors, including IRF3, in re-sponse to foreign ribonucleic acid21 [and indirectly,also in response to foreign deoxyribonucleic acid(DNA)21–23]. We found that once this complex is as-sembled, caspase-8 is recruited to it and suppressesits activity (Rajput, A., A. Kovalenko, K. Bogdanov,S-H. Yang, J-C. Kim, T-B. Kang, and D. Wallach,in preparation). It seems, therefore, that cells defi-cient in caspase-8 manifest excessive activation ofthis complex, which might prompt inflammation.

Notably, whereas in various established cell lines,caspase-8 deficiency facilitated the activation ofIRF3 by RIG-I-activating agents such as Sendai virusbut had no effect on basal IRF3 activity in their ab-sence, in the caspase-8–deficient epidermis, as wellas in cultures of keratinocytes freshly derived fromthis epidermis, activation of IRF3, and of IRF3-dependent genes occurred spontaneously (Fig. 4and data not shown). The spontaneous activationof IRF3-dependent genes in the caspase-8–deficient

Figure 4. IRF3 is constitutively activated in the caspase-8–deficient epidermis. The figure depicts Western blot analysisof epidermal lysates derived from two Casp-8F/+K5-Cre andtwo Casp-8F/−K5-Cre littermates at P3 for phosphorylation ofTANK-binding kinase-1 (TBK1) (ph-TBK1, enriched by im-munoprecipitation with anti-TBK antibody) and of IRF3 (ph-IRF3), as well as for expression of IRF3 and the IRF3-inducedprotein Isg15, and for conjugation of Isg15 to the cellular pro-teins. From Kovalenko et al.20


Wallach et al. Anti-inflammatory roles of “apoptotic” caspases

keratinocytes could be arrested by knocking downthe expression of RIG-I in them, suggesting the ex-istence of some endogenous activators of the RIG-Icomplex in these cells (data not shown). When weapplied in situ RNA hybridization to define the sitewithin the caspase-8–deficient epidermis where ex-pression of the affected genes occurs, we found thatit takes place in the granular layer of the epidermis,in which structural changes characteristic of corni-fication become apparent.20 This finding raised thepossibility that a byproduct of the terminally dif-ferentiating keratinocytes provides the stimulus forIRF3 activation in the epidermis. One of the IRF3-activating stimuli whose effects was potentiated bycaspase-8 deficiency was transfected DNA (Fig. 3).It, therefore, seems plausible that the spontaneousactivation of IRF3 in the caspase-8–deficient epi-dermis occurs in a manner similar to that observedin macrophages of mice deficient in DNAase-II,24

namely, in response to DNA fragments that accu-mulate in the granular layer as a result of the massivedisintegration of cells in this region.

Concluding remarks

In the course of evolution, conserved biomolec-ular structures often adopt functions that areunrelated to each other.25 However, there are alsostructures whose patterns of effects seem to havemaintained their relationships to one commonphysiological function.26 To some extent, this seemsto be true for the caspases. Their functions are het-erogeneous, and some—for example, the regulationof certain aspects of epidermal differentiation bycaspase-1427—might indeed prove to have no rel-evance to the others. As described above, however,their spectrum of functions appears to be dominatedby those that concern one particular physiologicalprocess, namely inflammation. Study of the func-tions of the caspases provides an intriguing point ofview of molecular “decisionmaking” with regard tothe nature of regulation of inflammation. Althoughactivated by similar mechanisms, inflammatory andapoptotic caspases mediate practically opposing ef-fects: the former induce inflammation, whereas thelatter prevent inflammation by sacrificing the in-jured cell in a way that withholds the release of itscontents and prompts its uptake by phagocytic cells.The findings described above concerning the anti-inflammatory functions of caspase-8 reveal that theapoptotic caspases can also act to restrict inflamma-

tion in ways that do not require cell sacrifice. Thus,a cell’s molecular “decision” on whether to respondto an insult by initiating inflammation through ac-tivation of inflammatory caspases or by preventinginflammation via activation of apoptotic caspasesrequires a further “decision” about whether activa-tion of the latter response will lead to the demise ofthe cell or spare its life. Knowledge of the molecularbasis for these “decisions” is only now beginning toemerge.

Acknowledgments

Work cited from the authors’ laboratory was sup-ported in part by grants from Ares Trading SA,Switzerland, a Center of Excellence Grant fromthe Flight Attendant Medical Research Institute(FAMRI), the Kekst Family Center for Medical Ge-netics, and the Shapell Family Center for GeneticDisorders Research at the Weizmann Institute ofScience.

Conflict of interest

The authors declare no conflict of interest.

References

1. Yeretssian, G., K. Labbe & M. Saleh. 2008. Molecular reg-ulation of inflammation and cell death. Cytokine 43: 380–390.

2. Pop, C. & G.S. Salvesen. 2009. Human caspases: activation,specificity, and regulation. J. Biol. Chem. 284: 21777–21781.

3. Kerr, J.F., A.H. Wyllie & A.R. Currie. 1972. Apoptosis: a basicbiological phenomenon with wide-ranging implications intissue kinetics. Br. J. Cancer 26: 239–257.

4. Zitvogel, L., O. Kepp & G. Kroemer. 2010. Decoding celldeath signals in inflammation and immunity. Cell 140: 798–804.

5. Fadok, V.A., D.L. Bratton & P.M. Henson. 2001. Phagocytereceptors for apoptotic cells: recognition, uptake, and con-sequences. J. Clin. Invest. 108: 957–962.

6. Wallach, D., E.E. Varfolomeev, N.L. Malinin, et al. 1999. Tu-mor necrosis factor receptor and Fas signaling mechanisms.Annu. Rev. Immunol. 17: 331–367.

7. Laster, S.M., J.G. Wood & L.R. Gooding. 1988. Tumor necro-sis factor can induce both apoptic and necrotic forms of celllysis. J. Immunol. 141: 2629–2634.

8. Wang, L., F. Du & X. Wang. 2008. TNF-alpha inducestwo distinct caspase-8 activation pathways. Cell 133: 693–703.

9. Holler, N., R. Zaru, O. Micheau, et al. 2000. Fas triggers analternative, caspase-8-independent cell death pathway usingthe kinase RIP as effector molecule. Nat. Immunol. 1: 489–495.

10. Degterev, A., J. Hitomi, M. Germscheid, et al. 2008. Identi-fication of RIP1 kinase as a specific cellular target of necro-statins. Nat. Chem. Biol. 4: 313–321.

Ann. N.Y. Acad. Sci. 1209 (2010) 17–22 c© 2010 New York Academy of Sciences. 21


11. Zhang, D.W., J. Shao, J. Lin, et al. 2009. RIP3, an en-ergy metabolism regulator that switches TNF-inducedcell death from apoptosis to necrosis. Science 325: 332–336.

12. Cho, Y.S., S. Challa, D. Moquin, et al. 2009. Phosphorylation-driven assembly of the RIP1-RIP3 complex regulates pro-grammed necrosis and virus-induced inflammation. Cell137: 1112–1123.

13. He, S., L. Wang, L. Miao, et al. 2009. Receptor interactingprotein kinase-3 determines cellular necrotic response toTNF-alpha. Cell 137: 1100–1111.

14. Festjens, N., T. Vanden Berghe & P. Vandenabeele. 2006.Necrosis, a well-orchestrated form of cell demise: sig-nalling cascades, important mediators and concomitantimmune response. Biochim. Biophys. Acta 1757: 1371–1387.

15. Lin, Y., A. Devin & Y. Rodriguez, et al. 1999. Cleavage ofthe death domain kinase RIP by caspase-8 prompts TNF-induced apoptosis. Genes Dev. 13: 2514–2526.

16. Upton, J.W., W.J. Kaisr & E.S. Mocarski. 2010. Virus in-hibition of RIP3-dependent necrosis. Cell Host Microbe 7:302–313.

17. Varfolomeev, E.E., M. Schuchmann, V. Luria, et al. 1998.Targeted disruption of the mouse Caspase 8 gene ablates celldeath induction by the TNF receptors, Fas/Apo1, and DR3and is lethal prenatally. Immunity 9: 267–276.

18. Ben Moshe, T., T.-B. Kang, A. Kovalenko, et al. 2008.

Cell-autonomous and non-cell-autonomous functions ofcaspase-8. Cytokine Growth Factors Rev. 19: 209–217.

19. Ben Moshe, T., H. Barash, T.B. Kang, et al. 2007. Role ofcaspase-8 in hepatocyte response to infection and injury inmice. Hepatology 45: 1014–1024.

20. Kovalenko, A., J.C. Kim, T.B. Kang, et al. 2009. Caspase-8 de-ficiency in epidermal keratinocytes triggers an inflammatoryskin disease. J. Exp. Med. 206: 2161–2177.

21. Takeuchi, O. & S. Akira. 2010. Pattern recognition receptorsand inflammation. Cell 140: 805–820.

22. Chiu, Y.H., J.B. Macmillan & Z.J. Chen. 2009. RNA poly-merase III detects cytosolic DNA and induces type I inter-ferons through the RIG-I pathway. Cell 138: 576–591.

23. Ablasser, A., F. Bauernfeind, G. Hartmann, et al. 2009. RIG-I-dependent sensing of poly(dA:dT) through the inductionof an RNA polymerase III-transcribed RNA intermediate.Nat. Immunol. 10: 1065–1072.

24. Okabe, Y., K. Kawane & S. Nagata. 2008. IFN regulatoryfactor (IRF) 3/7-dependent and -independent gene induc-tion by mammalian DNA that escapes degradation. Eur. J.Immunol. 38: 3150–3158.

25. Jacob, F. 1977. Evolution and tinkering. Science 196: 1161–1166.

26. Tomkins, G.M. 1975. The metabolic code. Science 189: 760–763.

27. Denecker, G., P. Ovaere, P. Vandenabeele, et al. 2008.Caspase-14 reveals its secrets. J. Cell Biol. 180: 451–458.


anti-inflammatory functions of the “apoptotic” caspases

Documents