Caspase-8, the proximal enzyme in the death-induction pathway of the TNF/nerve growth factor receptor family, is activated upon juxtaposition of its molecules within the receptor complexes and is then self-processed. Caspase-8 also contributes to the regulation of cell survival and growth, but little is known about the similarities or the differences between the mechanisms of these nonapoptotic functions and of the enzyme’s apoptotic activity. In this study, we report that in bacterial artificial chromosome-transgenic mice, in which the aspartate residue upstream of the initial self-processing site in caspase-8 (D387) was replaced by alanine, induction of cell death by Fas is compromised. However, in contrast to caspase-8-deficient mice, which die in utero at mid-gestation, the mice mutated at D387 were born alive and seemed to develop normally. Moreover, mice with the D387A mutation showed normal in vitro growth responses of T lymphocytes to stimulation of their Ag receptor as well as of B lymphocytes to stimulation by LPS, normal differentiation of bone marrow macrophage precursors in response to M-CSF, and normal generation of myeloid colonies by the bone marrow hematopoietic progenitors, all of which are compromised in cells deficient in caspase-8. These finding indicated that self-processing of activated caspase-8 is differentially involved in the different functions of this enzyme: it is needed for the induction of cell death through the extrinsic cell death pathway but not for nonapoptotic functions of caspase-8.
Members of the caspase family of cysteine proteases play a pivotal role in the initiation of programmed cell death in response to a variety of inducing agents. By cleaving a wide set of distinct target proteins at specific substrate sites, they trigger the various functional and structural changes that comprise the apoptotic cell death program (1). The caspases also contribute, by mechanisms that are still poorly understood, to certain nonapoptotic processes (reviewed in Ref. 2).
The different caspases operate in cascades for death induction. Caspase-2, -8, or -9, the “inducer” caspases, which contain a distinct “prodomain” region upstream of their protease moiety, interact through this domain with death-inducing proteins and convey the signal for death to the effector caspases, caspase-3, -6, and -7. These three caspases in turn cleave the various cellular target proteins that initiate the apoptotic death program. Both the inducer and the effector caspases exist in the living cell as inactive zymogens and at the time of death induction are processed to yield the mature caspase molecules. However, the functional consequences of the processing of these two subgroups of caspases differ. The effector caspases attain an enzymatically active conformation only after they have been processed (by the inducer caspases as well as by their own effects on one another). The inducer caspases, in contrast, even before they are processed attain an enzymatically active conformation as a result of oligomerization of these proteins upon binding of their prodomains to death-inducing agents. Eventually, they too are processed (by their own proteolytic activity). However the functional significance of this processing is not known (3).
In the present study, we explored the functional significance of the proteolytic processing of caspase-8. This caspase is known to serve as the initiator caspase in the induction of cell death by receptors of the TNF/nerve growth factor (NGF)4 family (the “extrinsic cell death pathway”) and was also shown to serve various nonapoptotic functions that occur independently of the TNF/NGF receptor family. We report that in bacterial artificial chromosome (BAC)-transgenic mice whose caspase-8 gene is mutated at the site at which the enzyme’s self-processing is initiated, death induction by the death receptor Fas is compromised, whereas the mutation has no effect on a variety of the enzyme’s nonapoptotic functions.
Materials and Methods
All mice used in these experiments had a C57BL/6 background and were between 8 and 10 wk old. C57BL/6 mice were purchased from Harlan. The strains carrying a knocked out caspase-8 allele (Casp−/+) (4) and a conditional caspase-8 allele (Casp8F/+) (5) and their use for tissue-specific deletion of the caspase-8 gene have been previously described (5). For caspase-8 deletion in B lymphocytes (Casp8F/−:CD19-Cre), we used mice expressing Cre under the B cell-specific CD19 promoter (6). Generation of BAC-transgenic mice is described below. All mice were kept in a specific pathogen-free facility and handled according to the criteria outlined in the Guide for the Care and Use of Laboratory Animals prepared by the National Academy of Sciences and published by the National Institutes of Health. All experiments were approved by the institutional animal ethical committee. The mice were infected with the lymphocytic choriomeningitis virus (LMCV) at the animal facility of the Kantonsspital St. Gallen in accordance with Swiss federal and cantonal legislation.
Preparations of human Fas ligand (FasL) fused to a leucine zipper and the FLAG tag were generated by transient expression in HEK293T cells, as described for several other ligands of the TNF family (7, 8). Polyclonal Ab against mouse caspase-8 was generated by immunizing rabbits with a peptide corresponding to the C-terminal end of the caspase p18 fragment. The Ab was affinity purified on a column of the immobilized peptide. The following PE- or allophycocyanin-conjugated mAbs, all from eBioscience, were used for FACS analysis: anti-mCD3 (145-2C11), anti-mB220 (RA3-6B2), anti-mCD11b (M1/70), anti-mGr-1 (RB6-8C5), anti-mF4/80 (BM8), anti-mCD4 (L3T4), and anti-mCD8a (53-6.7). Abs used for immunoblotting were anti-caspase-8 (1G12) and anti-cFLIP (Dave2) from Alexis; anti-cleaved caspase-3 (A51), anti-cleaved poly (ADP-ribose) polymerase (PARP; 7C9), anti-phospho-IκBα (14D4), and anti-phospho-p65 (Ser536) from Cell Signaling Technology; anti-caspase-3 (H-277), anti-IκBα (C-21), and anti-p65 (F-6) from Santa Cruz Biotechnology; anti-RIP (clone 38; BD Biosciences), anti-mouse Bid (a gift from Dr. A. Gross, Weizmann Institute, Rehovot, Israel), anti-Fas (7C10; Upstate Biotechnology); and anti-ERK, anti-phospho-ERK, and anti-β-actin from Sigma-Aldrich. Functional grade preparations of anti-mCD3 (145-2C11), anti-mCD28 (37.51; eBioscience), and anti-mouse IgM F(ab′)2 (Jackson ImmunoResearch Laboratories) were used for lymphocyte activation. PE-annexin V was obtained from BD Pharmingen, N-benzyloxycarbonyl-Val-Ala-Asp-(O-methyl) fluoromethane (zVAD-fmk) from Calbiochem, and staurosporine from Sigma-Aldrich.
Modification of a bacterial artificial chromosome encompassing the caspase-8 gene
DH10B bacteria harboring the RP24-238B22 BAC clone (BACPAC Resources) were grown in Luria-Bertani (LB) medium containing 12.5 mg/L chloramphenicol (Chlr). The presence of caspase-8 in the bacterial colonies was verified by PCR using oligonucleotides exon 8F–8R, exon 1F–1R, and 5′-UTRF–R (the oligonucleotides used are listed in Table I). The BAC clone was modified essentially as described elsewhere (9, 10) using the pDelsac shuttle vector both for deletion of the SacB gene from the RP24-238B22 clone and for modification of the caspase-8 gene.
|Primer .||Sequence .||Position .|
|Primer .||Sequence .||Position .|
The following modifications were introduced (see scheme in Fig. 1). 1) To assist in monitoring of transgenic BAC expression, we placed the T7 epitope tag-coding sequence upstream of the stop codon of caspase-8, and the internal ribosome entry site (IRES) sequence and the GFP open reading frame flanked by the two sequences corresponding to the FLP recombinase target (FRT) recombination site downstream of the stop codon of caspase-8. To introduce this sequence into the shuttle vector, the homology arms upstream and downstream of the stop codon (BoxA and BoxB, Fig. 1), cloned from the RP24-238B22 DNA by PCR were introduced into the SalI-EcoRI and SpeI-NotI sites of the pBC FRT-IRES/GFP-FRT vector. The AF-AR2 and BF-BR oligonucleotides were used to generate the caspase-8/T7 tag/FRT-IRES-GFP construct. The composite insert was then transferred to the shuttle vector. Then the T7-coding sequences were introduced into the EcoRI site in BC FRT-IRES/GFP-FRT containing BoxA and BoxB. 2) To facilitate genotyping of the transgenic mice, we exchanged a 20-nucleotide sequence within the first caspase-8 intron with a unique sequence (with equal nucleotide proportions). BoxC and BoxD were amplified using primers CF-CR and DF-DR, respectively. We then fused these two PCR products by using them as templates in a second PCR, which was conducted with the CF and DR primers. The data presented in this study were obtained with mice expressing the BAC with both of the above two modifications. 3) A noncleavable caspase-8 BAC-transgenic mouse was generated by replacement of the aspartic acid at position 387 with alanine (D387A). BoxG and BoxH, which border on this sequence, were amplified and fused by PCR using primers BAC2F-GR and HF-BAC2R. All of the above modification cassettes were cloned into the AscI-NotI sites of the Delsac shuttle vector.
Using a MicroPulser Electroporator (Bio-Rad) at 1.8 kV, the shuttle vector (500 ng) was transformed into 50 μl of electrocompetent BAC-containing bacteria. After electroporation, the cells were incubated for 1 h at 37°C in 1 ml of super optimal broth with catabolite repression medium and selected overnight in LB medium containing Chlr (12.5 mg/L) and ampicillin (30 mg/L). The cells were then diluted and grown on solid LB medium containing Chlr (12.5 mg/L) and ampicillin (50 mg/L). Cointegrates were identified by direct PCR on individual colonies.
For selection of resolved BACs, two or three positive bacterial colonies of each kind were then spread onto LB plates containing Chlr (12.5 mg/L) and 5% sucrose This was followed by further selection in which we picked ampicillin-sensitive colonies. Positive colonies were validated by direct PCR (TB3F-TB3R and TB4F-TB4R for BAC-T7/IRES/GFP; 1130F-1370R and MutCF-1370R for CD), and by XbaI restriction analysis following PCR, using the GF-HR primer pair, for GH.
Isolation of BAC DNA for microinjection
BAC DNA was isolated by double acetate precipitation and cesium chloride gradient ultracentrifugation. After washing with ethanol, the BAC DNA was dissolved in TE buffer, linearized by digestion with PI-SceI endonuclease (New England Biolabs), and drop-dialyzed for 6 h against microinjection buffer (10 mM Tris (pH 7.5), 0.1 mM EDTA (pH 8.0), and 100 mM NaCl) by floating on a 0.025-μm Millipore membrane filter disc. The quality and quantity of BAC DNA were assessed by pulse-field gel electrophoresis; 5 V/cm, 120° angle, linear ramping time 5–120 s for 24–30 h) in 1% agarose using the CHEF-DR II PFEG system (Bio-Rad). The DNA was diluted to 2 ng/μl in microinjection buffer and mixed with an equal volume of 2× polyamine (60 mM spermine and 140 mM spermidine in injection buffer) for the injection.
Generation and analysis of BAC-transgenic mouse
DNA (1 ng/μl) was injected into the pronuclei of fertilized oocytes derived from CBF1 or C57BL/6 mice. Transgenic mice were identified both by PCR analysis (using primers Mut CF and IN1-1730R) of genomic DNA prepared from tail biopsies and by FACS analysis for GFP expression in peripheral blood leukocytes. To generate BAC-WT or BAC-D387A mice on an endogenous caspase-8 null background (−/−/BAC-WT and −/−/D387A, respectively), the founder mice were crossed with caspase-8+/− mice (on a C57BL/6 background) (4). The caspase-8+/− mice containing the BAC transgene were then selected and further crossed with caspase-8+/− mice. Endogenous caspase-8 was identified by PCR analysis using primers IN1-1130F and IN1-1730R, and the BAC transgene was identified by PCR and FACS analysis as described above.
Immunization and ELISA
Mice were immunized i.p. with 25 μg of trinitrophenyl (TNP)-LPS (Biosearch Technologies) or 100 μg of TNP-OVA (Biosearch Technologies) emulsified with CFA. The mice were bled before immunization and again 7 and 14 days after immunization. Serum Ab specific for TNP was measured by ELISA as follows: plates were coated with 100 μg/ml TNP-BSA (Biosearch Technologies) by overnight incubation at 4°C, then rinsed, and blocked for 2 h at room temperature with 3% BSA in PBS/Tween (PBST). Mouse sera, serially diluted in PBST, were applied to the plates for 1 h at room temperature. Plates were rinsed and the bound Abs were detected by binding of goat F(ab′)2 anti-mouse IgM plus IgG plus IgA (H + L) coupled to HRP (Southern Biotechnology Associates) and then developed with 3,3′,5,5′-tetramethylbenzidine (Southern Biotechnology Associates). The plates were read at 450 nm. Titers of sera were defined by determining the dilution of the serum that yielded 50% of the maximal reading.
Spleens, lymph nodes, and thymi were mechanically disrupted by their passage though a syringe, and single-cell suspensions were obtained by filtration through cell strainers (75 μm; BD Biosciences). Fresh or cultured cells were stained for 20 min at 4°C with the appropriate Ab in 100 μl of staining buffer (2% FBS and 0.05% sodium azide in PBS). A FACScan or a FACSort device (BD Biosciences) was used for flow cytometry and CellQuest software (BD Biosciences) for data analysis. Dead cells were gated out by staining with 7-aminoactinomycin D (1 μg/ml) during data collection. Cell death was measured by propidium iodide staining followed by flow cytometry as described previously (11). Cells were stained simultaneously with PE-annexin V (BD Pharmingen) and Topro-3 (Molecular Probes) according to the manufacturers’ instructions. For flow cytometry, a FACScan machine (BD Biosciences) was used as described previously (5) and the results were analyzed with CellQuest software.
Western blot analysis
For Western blotting analysis, cells were lysed by resuspension for 20 min on ice in buffer A (20 mM Tris-HCl (pH 7.4), 150 mM NaCl, 2 mM EDTA, 1 mM NaF, 5 mM sodium vanadate, and 1% (v/v) Nonidet P-40) containing 1× complete protease inhibitor mixture (Roche Diagnostics) and then centrifuged for 15 min at 14,000 × g. Protein concentration in the extract was determined using the BCA kit (Pierce). Protein samples were separated on SDS-PAGE and blotted onto nitrocellulose membrane (Amersham Biosciences). Blots were developed with the ECL kit (Pierce).
Analysis of the death-inducing signaling complex in cultured thymocytes
Thymocytes were treated for the indicated times with FLAG-tagged FasL (20% (v/v) of the ligand-containing HEK293 growth medium) at 37°C, washed twice with cold PBS, and lysed in buffer B (20 mM Tris-HCl (pH 7.4), 150 mM NaCl, 2 mM EDTA, 1 mM NaF, 5 mM sodium vanadate, 1% (v/v) Nonidet P-40, and 0.25% (w/v) sodium deoxycholate) containing complete protease inhibitor mixture (Roche Diagnostics). Total cellular extracts from 108 thymocytes were precleared with agarose beads for 4 h at 4°C. FasL and its associated death-inducing signaling-complex from the extract were immunoprecipitated by incubation for 2.5 h at 4°C with anti-M2 agarose beads (Sigma-Aldrich). After thorough washing with lysis buffer, the bound proteins were eluted by incubation of the beads for 90 min at room temperature with FLAG peptide (1 mg/ml in lysis buffer) and then subjected to SDS-PAGE and Western blot analysis using anti-mouse caspase-8 and anti-mouse Fas. The blot was developed with SuperSignal West Femto Substrate (Pierce).
Injection of mice with anti-Fas and analysis of their livers
Mice were injected i.p. with anti-Fas Ab (Jo-2; BD Biosciences) and their survival was followed for 3 days. At the indicated times, their livers were harvested and either processed for histological analysis by fixation in 10% phosphate-buffered formalin (pH 7.4) followed by embedding in paraffin, cutting into 4-μm sections, and staining with H&E or extracted for 20 min on ice in lysis buffer A supplemented with 1× complete protease inhibitor mixture. Caspase-8 in the liver lysates of the mice was concentrated by preclearing of the lysates with protein G-Sepharose beads for 2 h at 4°C, followed by overnight incubation at 4°C with anti-mouse caspase-8 rabbit Ab and protein G-Sepharose beads. Caspase-8 was then eluted from the beads using the peptide against which the rabbit Ab had been raised (0.5 mg/ml in buffer A). The eluted protein was subjected to Western blotting analysis using rat anti-mouse caspase-8 mAb.
Preparation of peritoneal macrophages and induction of their death
Mice were injected i.p. with 1.5 ml of 3% thioglycolate broth (Life Technologies), and 4 days later peritoneal cells were harvested by lavage with 10 ml of RPMI 1640 medium (supplemented with 10% FBS, 2 mM l-glutamine, 25 U/ml penicillin, and 25 μg/ml streptomycin), washed twice, plated on 96-well plates, and incubated for 3 h. Nonadherent cells were then removed by thorough washing with cold PBS and the adherent macrophages were stimulated for 24 h with LPS (1 μg/ml). After five rinses with culture medium, the cells were treated with FasL for 24 h. Their viability was then assessed by MTT assay (12).
Isolation of B cells and assessment of their proliferation in vitro
B lymphocytes were isolated from mouse spleens using the BD Biosciences IMag Mouse B Lymphocyte Enrichment Set according to the manufacturer’s instructions. More than 92% of these negatively selected cells were found by flow cytometry to stain positively for B220. The purified B cells were plated in triplicate onto 96-well plates at a concentration of 5 × 104 cells/well and then treated for 2 days with anti-IgM F(ab′)2 or LPS (serotype 055:B5, Sigma-Aldrich). Proliferation was assessed by determining thymidine incorporation into the cells.
T cell activation and activation-induced cell death (AICD)
Splenic T cells were isolated using the BD Biosciences IMag Mouse T Lymphocyte Enrichment Set according to the manufacturer’s instructions. T cell purity, as assessed by FACS analysis of CD3 staining, was >94%. For activation in vitro, the cells (1 × 105 cells/well) were stimulated with plate-bound anti-CD3 Ab and anti-CD28 mAbs in 96-well plates and then cultured for 2 or 3 days. Proliferation was then assessed by assay of thymidine incorporation into the cells.
For induction of AICD, the T cells were cultured at a density of 1 × 106 cells/ml with plate-bound anti-CD3 and soluble anti-CD28 mAbs for 2 days. They were then rinsed three times with culture medium, resuspended in medium containing human rIL-2 (100 U/ml), and cultured for 3 days longer. After removal of dead cells by density gradient centrifugation (Lympholyte M; Cedarlane Laboratories), the activated cells were treated for 20 h with anti-CD3 Ab in the presence of IL-2 (50 U/ml). Cell death at that time point was assayed by FACS analysis.
Other tests of cellular functions
CFU-C assay of clonogenic bone marrow hematopoietic progenitors and assessment of the differentiation of bone marrow-derived macrophages were as described elsewhere (5).
Viruses and peptides
WE and ARM strains of lymphocytic choriomeningitis virus (LCMV) were propagated on L929 cells at a low multiplicity of infection and titrated as described previously (13). Mice were infected i.v. with 200 PFU of WE or 2 × 103 PFU of ARM. The LCMV-GP peptides KAVYNFATM (gp33) and FQPGNGQFI (nucleoprotein (np) 396) were purchased from Neosystem.
Intracellular cytokine staining
Spleens were removed 8 days after infection with LCMV. Single-cell suspensions of 1 × 106 splenocytes were incubated for 5 h at 37°C in 96-well round-bottom plates in 200 ml of culture medium containing 25 U/ml IL-2 and 5 mg/ml brefeldin A (Sigma-Aldrich). The cells were stimulated with PMA (50 ng/ml) and ionomycin (500 ng/ml) as a positive control or left untreated as a negative control. Peptide-specific responses were analyzed after cells were stimulated with 10−6 M gp33 or np396 peptides and surface stained as previously described (14). The percentage of CD8+ T cells producing IFN-γ was determined using a FACSCalibur flow cytometer and CellQuest software (BD Biosciences).
Construction of tetrameric class I-peptide complexes and flow cytometry
Class I MHC (H-2Db and H-2Kb) monomers complexed with LCMV-gp33–41 and β-galactosidase-96–103, respectively, were produced as previously described (14) and tetramerized by addition of streptavidin-PE (Molecular Probes). Eight days after infection, single-cell suspensions were prepared from spleens. Cells (5 × 105) were stained by addition of 50 μl of a solution containing tetrameric class I MHC peptides at 37°C for 10 min, followed by staining with anti-CD8-FITC (BD Pharmingen) at 4°C for 20 min. Cells were analyzed by flow cytometry gating on viable leukocytes.
Caspase-8 mutation at the site at which the enzyme’s self-processing is initiated compromises Fas-induced killing of thymocytes
To examine the functional significance of the self-processing of caspase-8, we generated BAC-transgenic mice expressing either the wild-type (WT) caspase-8 transgene or a caspase-8 gene mutated at the site at which processing of the enzyme is initiated (aspartate 387; Fig. 1).
Both groups of transgenic mice were born alive and matured normally. Thymocytes of mice expressing the WT transgene were just as sensitive to the cytotoxic effect of FasL as those of normal mice. The effectiveness of the killing of cells expressing the endogenous or transgenic WT alleles was proportional to the amount of caspase-8 expressed in them, which in turn was proportional to the number of caspase-8 alleles (endogenous or transgenic) in their respective genomes (Fig. 2, A–C). In contrast, thymocytes of mice expressing only the mutant caspase-8 gene (−/−/D387A mice), despite their expression of the mutant caspase-8 protein in amounts comparable to those expressed by the WT enzyme in normal mice (Fig. 2,G), showed no signs of death after incubation with FasL for 3 or even 6 h. At longer incubation periods, we observed some spontaneous death of both the WT and the −/−/D387A thymocytes. Moreover, during those longer periods, the −/−/D387A thymocytes did exhibit some enhancement of their FasL-induced death, although much less than that seen in the WT cells (Fig. 2, D and E).
Analysis of the Fas receptor complexes in the −/−/D387A thymocytes disclosed that, as in the cells of the WT mice, the mutant caspase-8 was recruited to the receptor. However, whereas in the WT cells the receptor complex contained not only intact precursor caspase molecules but also molecules that were partially processed, no such processing was detectable in the cells of the −/−/D387A mice. Moreover, whereas WT caspase-8 occurred only transiently in the receptor complex, apparently because of internalization of the receptors (reportedly prompted by caspase-8 activation (15)) and processing of their associated caspase-8 molecules, the amounts of caspase-8 in the receptor complexes of the −/−/D387A thymocytes kept on increasing for a long time (Fig. 2 F).
In addition to caspase-8 itself, several other cellular proteins are believed to be direct targets of the proteolytic activity of this enzyme. They include cFLIP-L, which at high cellular concentrations inhibits activation of the extrinsic cell death pathway (16); RIP, a signaling protein which, through activation of the NF-κB transcription factors, also antagonizes activation of this death pathway (17); Bid, a Bcl2 family member whose product of processing by caspases (tBid) triggers mitochondrial death mechanisms (reviewed in Ref. 18); and caspase-3. Western blot analysis of thymocyte extracts revealed a marked decrease in the rate of Fas-induced processing of cFLIP-L (16), RIP (17), and Bid (18) in the −/−/D387A cells. We did, however, discern in them some low and delayed generation of processed caspase-3 (Fig. 2,G), as well as generation of processed Bid and poly(ADP-ribose) polymerase (PARP; Fig. 2,H), in line with our finding that after prolonged treatment with FasL the −/−/D387A thymocytes displayed some induced death (Fig. 2,E). Consistently with the association of this protracted death with cleavage of caspase substrates, death could be effectively blocked by the caspase inhibitor zVAD-fmk (Fig. 2,I) and most of the dying cells were stained with annexin V but not with Topro-3 (Fig. 2 J), indicating that this death, like that of the WT cells, was apoptotic.
The D387A mutation in caspase-8 also compromises death induction by Fas in other cells, but does not affect Fas-induced nonapoptotic signaling
Like the thymocytes of the −/−/D387A mice, the splenic B and T lymphocytes as well as the peritoneal macrophages of these mice showed markedly reduced sensitivity to Fas cytotoxicity. However, they were just as sensitive as the WT cells to death induction by staurosporine, indicating that the effect of the D387A mutation on death induction does not occur through the intrinsic cell death pathway (Fig. 3, A–C).
FasL-induced phosphorylation of IκBα, p65 NF-κB protein, and ERK, as well as its induction of degradation of IκBα, occurred normally in the −/−/D387A peritoneal macrophages (Fig. 2 D). Thus, whereas the D387A mutation compromised the induction of death by Fas, it seemed not to affect the Fas-induced activation of some nonapoptotic signaling pathways.
To examine the effect of the D387A mutation on induction of the extrinsic cell death pathway in vivo, we injected the mice with an anti-Fas Ab that triggers Fas signaling. In WT mice, triggering of Fas expressed in the hepatocytes resulted in their massive apoptosis and led to death of the mice within a few hours. The −/−/D387A mice, on the other hand, were resistant to such death induction, and their livers showed no discernible signs of cell death induction (Fig. 3,, E and F). Western blot analysis of caspase-8 in liver lysates confirmed that the enzyme’s initial cleavage in response to Fas triggering was compromised by the D387A mutation (Fig. 3,G), and whereas in the WT liver Fas triggering induced processing of caspase-3 as well as of its substrate PARP, no such processing was detectable in livers of the −/−/D387A mice (Fig. 3 H).
The D387A mutation of caspase-8 does not affect various nonapoptotic caspase-8 functions
Knockout of caspase-8 is lethal in utero (4). The fact that the −/−/D387A mice were born alive indicated that the function of caspase-8 whose ablation has a fatal effect is, unlike the cytotoxic effect of Fas, independent of caspase-8 processing. Except for the fact that the vital function of this enzyme seems, at least partly, to be exerted in endothelial cells (5) and to be required for proper remodeling of capillaries in the yolk sac (5, 19), we still have no clear knowledge of its nature. However, since none of the receptors of the TNF/NGF family is required for the survival of mice in utero, it seems likely that this vital function of caspase-8 is independent of the extrinsic cell pathway and might not reflect cell death induction at all.
Using cells derived from the −/−/D387A mice to assess several nonapoptotic cellular functions that were previously shown to depend on caspase-8, we found that these functions were not affected by the D387A mutation. Thus, whereas B lymphocytes in which caspase-8 was conditionally deleted were unresponsive to the mitogenic effect of LPS (Fig. 4,A and Ref. 20), B lymphocytes isolated from the −/−/D387A mice showed effective LPS-induced growth (Fig. 4,A). Also not affected by the D387A mutation were two other cellular functions previously shown to depend on caspase-8: M-CSF-induced differentiation of bone marrow macrophages and generation of colonies in the CFU-C assay of clonogenic bone marrow hematopoietic progenitors (Ref. 5 and Fig. 4, B and C).
Stimulation of the Ag receptors on T cells (TCRs) triggers their proliferation. However, restimulation of these cells in the absence of costimulation results in induction of their death. The latter, termed AICD, is believed to serve the function of restricting the duration of an immune response induced by limited exposure to a foreign Ag. Both the increase in the number of T cells and their death in response to TCR stimulation reportedly involve caspase-8 function (reviewed in Refs. 21 and 22). Using T lymphocytes of the −/−/D387A mice, we examined the requirement for caspase-8 processing in the case of these two opposing caspase-8 functions.
Whereas T cells deficient in caspase-8 reportedly display reduced activation-induced proliferation (23), T lymphocytes of −/−/D387A mice responded normally to such stimulation (Fig. 4,D). Consistently, as opposed to the marked drop in peripheral T lymphocyte count in mice whose T cells were deficient in caspase-8 (23), no such decrease, nor any aberration in the relative amounts of the other major kinds of leukocytes, was detectable in the lymphoid organs of the −/−/D387A mice (Fig. 4, E–H). Histological analysis of their spleens also revealed no abnormalities of B or T follicle structure (data not shown).
On the other hand, the extent of the cell death induced in the −/−/D387A mice when T lymphocytes were restimulated in the absence of costimulation was dramatically reduced (Fig. 4 I).
The D387 mutation does not affect the humoral response of the mice to foreign Ags or the clonal burst of their CTL upon LCMV infection
To examine the impact of the D387A mutation on lymphocyte function in vivo, we first assessed the humoral response of the mice to specific Ags. We found that the −/−/D387A mice exhibited normal Ab responses to both a T cell-dependent and a T cell-independent Ag (Fig. 5).
Since mice with caspase-8-deficient T cells are reportedly incapable of exhibiting a T cell-dependent immune response to LCMV (23), we also examined the response of the D387A mice to infection by this virus. In the course of LCMV infection, the virus-specific CTLs are efficiently stimulated to proliferate, with maximum expansion of the population around day 8 postinfection (24, 25). This accumulation of effector CTLs during the acute phase of LCMV is critical for purging of the virus from infected organs.
Unlike in mice with caspase-8-deficient T cells, in which the infection is persistent (23), we found that the virus was cleared from peripheral organs (kidney and liver) of the −/−/D387A mice on day 8 postinfection (data not shown) and that viral titers in their spleens were comparable to those of WT controls (Fig. 6,A), suggesting that −/−/D387A mice can exhibit normal CTL responses to viral infection. Analysis with MHC class I tetramers complexed with the immunodominant CTL epitope gp33 derived from the LCMV glycoprotein (tet-gp33) indeed revealed that the mice efficiently generated antiviral CTL responses (Fig. 6, B and C). Normal clonal burst was also apparent when we assessed CTL function in terms of IFN-γ production following short-term in vitro restimulation (Fig. 6 D). Taken together, these data show that the D387 mutation does not affect the reactivity of either B or T lymphocytes to foreign Ags.
Whereas the effector caspases possess enzymatic activity only in their processed form, caspase-8, like other inducer caspase, acquires enzymatic activity before being processed, upon its recruitment to signaling complexes (3). The results of this study show, however, that despite such activation, the extrinsic cell death pathway cannot be effectively initiated by caspase-8 unless the enzyme is processed. On the other hand, it seems that processing of the enzyme is not required for mediation of its various nonapoptotic functions.
The observed dependence of the extrinsic cell death pathway on the self-processing of caspase-8 is consistent with the fact that in cells induced to die by ligands of the TNF family this processing occurs before any signs of cell death are visible. However, the mechanistic basis for this requirement remains to be defined. Being enzymatically active, the unprocessed enzyme molecules should be capable of cleaving substrate proteins. Our analysis of thymocytes of the mice in which caspase-8 was mutated at the site where its self-processing is initiated indeed revealed that Fas treatment induces some processing of caspase-3 in these cells. The extent of this processing, however, was significantly lower than that in the WT cells. Three other proteins believed to be targeted by caspase-8, namely, cFLIP-L, RIP, and Bid, whose processing also facilitates death induction, were cleaved only ineffectively in the −/−/D387A thymocytes.
There are several possible reasons why death induction by an initiator caspase is dependent on its processing. First, in the unprocessed form, the activated caspase molecules probably remain associated (through their prodomain) with the signaling complexes that triggered their activation; thus, their processing, which results in release of the caspase molecules from these complexes, is likely to increase their accessibility to target molecules located at cellular sites remote from these complexes. Such release, moreover, may allow new caspase zymogen molecules to bind to the same signaling complex and, thus, by facilitating repetitive cycles of recruitment, activation, and processing, greatly enhance the generation of active caspase-8 molecules. There is also some evidence to suggest that the processing of caspase-8 enhances the stability of its catalytic conformation (26).
Previous findings have suggested that processing of caspase-8 might not be required for its reported contribution to the proliferation of T lymphocytes in response to TCR stimulation (27). Our study confirms that notion and extends it to several additional nonapoptotic functions of caspase-8. We found that the D387A mutation did not interfere with bacterial endotoxin-stimulated B lymphocyte proliferation nor with M-CSF-induced macrophage differentiation, both of which have been shown to depend on caspase-8 (5, 20). Moreover, unlike complete deletion of the enzyme, the D387A mutation had no effect on the formation of CFU-C colonies by bone marrow hematopoietic progenitors, nor did it result in intrauterine death of the mice, an effect of caspase-8 deletion that apparently reflects the enzyme’s involvement in endothelial cell function (5).
It seems likely that the triggering of nonapoptotic functions of caspase-8 is initiated, like the induction of cell death by this enzyme, by association of its prodomain with some inducing proteins. FADD/MORT1, the adapter protein to which caspase-8 binds for the induction of its cytotoxic effect, indeed seems also to be required for at least some of the nonapoptotic functions of caspase-8 (see Ref. 21). These nonapoptotic effects may thus also necessitate binding of caspase-8 to this protein. In view of our finding that nonapoptotic functions of caspase-8 do not depend on its self-processing, it also seems likely that, in contrast to the way in which the caspase-8 apoptotic effect is induced, in the induction of its nonapoptotic effects the enzyme remains associated with the adapter protein and the signaling complex that initiate this induction. This would imply that, unlike death induction through the extrinsic cell death pathway, the nonapoptotic functions of caspase-8 are exerted in a locally restricted manner and by the limited number of caspase molecules activated in a single activation cycle.
Notably, in a detailed kinetic analysis of the response of the −/−/D387A mice to FasL, we observed that although these cells were completely resistant to the cytotoxic effect of the ligand during the first few hours of treatment, unlike cells deficient in caspase-8, they did show some induced apoptotic death later on. It thus seems that there is also a caspase-8-dependent mechanism for death induction by ligands of the TNF family that does not depend on the enzyme’s self-processing. Such death occurs only after prolonged incubation of the thymocytes in culture, when some spontaneous death has already begun. The data presented in Fig. 2, H and I, indicate that both this spontaneous late death and its enhancement by FasL are associated with and dependent on caspase activation. This caspase dependence raises the possibility that the delayed Fas-induced death in the −/−/D387A cells is triggered by the effects on the Fas-activated caspase-8 molecules exerted by mediators (such as processed caspase-3 and caspase-6) that are generated via spontaneous activation of the intrinsic cell death pathway.
While the requirement for self-processing seems to be distinctive for cell death induction by caspase-8, there might also be certain structural features of caspase-8 that are specifically required for its involvement in nonapoptotic functions and perhaps even some that distinguish between different kinds of such functions. Further analysis of structure/function relationships for the various activities of caspase-8 in transgenic mouse models should facilitate better classification of the enzyme’s cellular activities, provide clues to their molecular mechanisms, and assist definition of their relative contributions to specific in vivo processes. Furthermore, in view of the evidence that mutation of caspase-8 in humans may result in immune deficiency (28), and that caspase-8 mutations and polymorphisms in humans are linked to cancer (see e.g., Refs. 29, 30, 31), further studies of the pathological consequences of specific mutations of caspase-8 in mice can be expected to extend our knowledge of the genetic basis of human diseases.
We thank Drs. Nathaniel Heintz, Shiaoching Gong, Gérard Eberl, Nathalie Uyttersprot, and Yael Pewzner-Jung for advice and help with BAC transgenesis; Karen Laviv for assistance with BAC DNA modification; Golda Damari for BAC DNA microinjection; Inna Kolesnik, Tatiana Shalevich, and Dvir Mintz for genotyping the mice; and Shirley Smith for scientific editing. Special thanks to Prof. Israel Vlodavsky for assistance.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported in part by grants from Ares Trading S.A., Switzerland (to T.B.K., G.S.O., A.K., and D.W.), a Center of Excellence Grant from the Flight Attendant Medical Research Institute, and the Kekst Family Center for Medical Genetics at Weizmann Institute of Science. The work of E.S., B.B., and B.L. was supported by the Kanton of St. Gallen.
Abbreviations used in this paper: NGF, nerve growth factor; AICD, activation-induced cell death; BAC, bacterial artificial chromosome; Chlr, chloramphenicol; LCMV, lymphocytic choriomeningitis virus; FasL, Fas ligand; PARP, poly(ADP-ribose) polymerase; LB, Luria-Bertani; IRES, internal ribosome entry site; TNP, trinitrophenyl; zVAD-fmk, N-benzyloxycarbonyl-Val-Ala-Asp-(O-methyl) fluoromethane; WT, wild type; FRT, FLP recombinase target.