Sepsis, the leading cause of mortality in surgical and medical intensive care units, is a condition in which the immune system no longer appears to respond to a pathogen adequately, oftentimes leading to multiple organ failure. Although there have been significant advances in supportive care, very few therapeutic advances have been made based on mechanistic processes that underpin the development of septic morbidity/mortality. Our laboratory and others have shown that increased apoptosis of immune cells mediated by death receptor ligation contributes greatly to increased morbidity/mortality in the experimental setting. However, the mechanism by which nonimmune tissue cells are injured/die is less clear.
Receptor interacting protein 1 (RIP1) is a multifunctional adaptor protein that has been implicated in many signaling functions associated with apoptosis as well as inflammation. Receptor interacting protein 1 has three domains: a kinase domain, intermediate domain, and a death domain (1). The death domain of RIP1 has been shown to associate with many receptors, namely, the TNF superfamily, i.e., TNFR1, Fas, and TRAIL (2, 3). Receptor interacting protein 1 has been shown to be involved in switching apoptotic death to a necrotic cell death in L929 cells treated with the pan-caspase inhibitor z-VAD (3). Recently, Cho (4) reported that RIP1 and RIP3 associate to form the basis of a signalosome responsible for initiating programmed necrosis. In studies done with RIP gene-deficient (−/−) fibroblasts, RIP1 was shown to not be necessary to initiate Fas-induced apoptosis. However, RIP1 was necessary for survival of CD4/CD8 double-positive lymphocytes derived from the thymus, indicating that RIP1 can act as a prosurvival signal in some cases (2). Downstream prosurvival signals for RIP1 appear to be mediated through signaling via nuclear factor κB (NF-κB), p38, and/or JNK.
We previously reported in a mouse model of sepsis that immune cell populations in the spleen and thymus show an increase in Fas-mediated apoptotic death. On the other hand, the liver (hepatocytes specifically), while also displaying aspects of an increase in apoptosis, also showed that indices of necrosis were increased as well (5). In light of the above information, we proposed to examine the hypothesis that RIP1 was involved in the alteration of the apoptotic death pathway to "programmed necrosis" in the liver. In the following sections, we describe those experiments designed to address this hypothesis and how our results have led us to consider an alternative role for RIP1 in the pathobiology of sepsis in liver, with a particular focus on its alternative role in cellular survival.
MATERIALS AND METHODS
Mouse model of sepsis
C57BL/6 male mice, 8 to 12 weeks old, were anesthetized using isoflurane (Abbott Laboratories, North Chicago, Ill) (6, 7). A midline incision (1.5-2 cm) was made, and the cecum was isolated, ligated with 4-0 silk, and punctured twice with a 22-gauge needle. The cecum was compressed to extrude a small amount of contents from the cecum. The cecum was returned to the peritoneal cavity, and both layers were closed with 6-0 nylon sutures (Ethicon, Somerville, NJ). Two to three drops of lidocaine were applied to the peritoneum after closure for analgesia. The mice were resuscitated with 1.0 mL of lactated Ringer's solution subcutaneously. Sham controls were subjected to the same surgical procedure except the cecum was neither ligated nor punctured. All animal experiments, described herein, were done in accordance with National Institutes of Health (Bethesda, Md) guidelines and as approved by the local animal use committee at Lifespan-Rhode Island Hospital.
Hepatocytes were isolated using the two-step perfusion method (8). In brief, mice were anesthetized with isoflurane and subjected to ventral midline incision. The hepatic portal vein was isolated and cannulated with a 27-gauge needle. The liver was perfused for 4 min with bicarbonate buffer (Hanks balanced salt solution, EDTA, HEPES, and NaHCO3) using a peristaltic pump (Masterflex; Cole-Palmer, Vernon Hills, Ill). The liver was then perfused with a warm collagenase solution (Leibovitz's modified L-15 medium, glucose, and collagenase IV) for 15 min, teased apart in the collagenase solution, percolated through a 25-mL pipette, and filtered through two layers of gauze. The resultant cells were suspended in RPMI 1640 medium + 5% fetal calf serum and centrifuged at 30g for 4 min, and the hepatocytes were then resuspended in CER I lysis buffer (NE-PER kit; Pierce Biotechnology, Rockford, Ill) to isolate cytosolic as well as the nuclear fractions.
Hepatocytes were lysed and incubated with Fas antibody ([M-20] SC-716, lot D1304; Santa Cruz Biotechnology Inc, Santa Cruz, Calif) for an hour. The antibodies were pulled down with protein G-coated agarose beads (SC-2003 lot B1810; Santa Cruz Biotechnology Inc).
Western blot analysis
Protein lysates of mouse hepatocytes were run on 10% Tris-glycine gels (Invitrogen, Carlsbad, Calif). Blotting procedures, chemiluminescent detection, and densitometric analysis were performed as previously described by our laboratory (9). Membranes were probed with either RIP1 polyclonal antibody (catalog 610458, lot 04047; BD Transduction Laboratories, San Diego, Calif) or NF-κB p65 polyclonal antibody (ab7970, lot 870107; Abcam, Cambridge, Mass), and bands were detected at 76 and 65 kd, respectively. GAPDH and histone H2 antibodies were used as loading controls for the cytosolic and nuclear fractions, respectively.
Hydrodynamic siRNA delivery
Mice were injected with 50 μg RIP1 siRNA (5′-GAAUGAGGCUUACAACAG-3′; Dharmacon Inc, Rockford, Ill) in 2 mL saline via the lateral tail vein over an approximately 5-s period, as previously described (9-11). Specificity of RIP1 knockdown was determined by Western blot analysis of hepatocyte lysates derived from mice that had received RIP1 siRNA or scrambled siRNA sequence treatment before the cecal ligation and puncture (CLP) procedure.
RIP1 kinase inhibition
Mice were injected intraperitoneally (i.p.) with 0.125 mg necrostatin 1 (Nec1) (Sigma Aldrich, St Louis, Mo) in 0.2% dimethyl sulfoxide (DMSO) daily for 2 days before CLP. Immediately after CLP, the animals were given another 0.125-mg dose of Nec1. Control animals were injected only with 0.2% DMSO. Subsequent Nec1 and control injections were given daily for the duration of the survival study. Twenty-four hours after CLP, the mouse livers were perfused and harvested for analysis.
Quantification of serum chemokines/cytokines and liver enzymes
Blood was collected into a heparinized syringe via cardiac puncture of control or Nec1-treated mice. The red blood cells were separated by centrifugation, and the serum was removed and stored at −80°C until analysis. IL-10, IL-6, monocyte chemoattractant protein 1 (MCP-1), and TNF-α levels were measured in the serum using a mouse cytometric bead array (BD cytometric bead array mouse inflammation kit; BD Biosciences, San Jose, Calif) according to the manufacturer's instructions. Alanine aminotransferase (ALT) and aspartate aminotransferase (AST) enzyme activities were measured using ALT and AST reagent sets (Biotron Diagnostics, Hemet, Calif).
Periodic acid-Schiff stain of gylcogen stores quantification
Histological slides were stained with periodic acid-Schiff (PAS) and scanned with an Aperio Scan scope-CS (Aperio Technologies, Inc, Vista, Calif), and 5 to 10 fields were selected at 200× magnification for each specimen. Area measurements were performed with iVision software (BioVision Technologies, Exton, Pa). Positive staining was defined through intensity thresholding for total area stained in the field and percent region of interest. A pathologist specializing in gastroenterology, blinded to the groups, reviewed the slides for apoptosis and other morphological differences between samples.
Caspase 3 staining
Mice were injected with Nec1 (as described above) or 0.2% DMSO, as a control. The two groups, Nec1 and DMSO, were subjected to CLP or sham surgery, and the livers were perfused and fixed at 24 h after surgery. These livers were paraffin embedded, and histological slides were cut. Tissue sections 5 μm in thickness were deparaffinized and rehydrated through graded alcohols to water. Tissue sections were subjected to heat-induced antigen retrieval in 10 mM citrate buffer (pH 6.0) for 11 min, blocked using peroxidase block (Dako, Carpinteria, Calif) for 5 min. The slides were then blocked using avidin-biotin block for 10 min for each solution (Vector Labs, Burlingame, Calif). Sections were incubated with caspase 3 antibody overnight at 4°C. The slides were developed using the Envision Plus Dual Link and the Liquid DAB+ substrate kits (Dako). The slides were counterstained in hematoxylin, dehydrated through graded alcohols to xylene, and mounted with Permount (Fisher Scientific, Pittsburgh, Pa). The amount of caspase 3 staining was quantified in the same manner described above.
Presentation of data and statistical analysis
The results are presented as a mean ± SEM for each group. Mann-Whitney U test was used to analyze the comparisons between two groups. For data comparing more than two groups, a one-way ANOVA was executed followed by a Student-Newman-Keuls test to determine the differences. The survival data were compared using the log-rank test. P ≤ 0.05 was considered statistically significant.
Inhibition of Fas-FasL signal alters septic mouse liver RIP1 gene expression
We have previously reported that Fas-FasL signaling played a role in inducing the onset of apoptosis, most markedly in lymphoid tissues of the spleen (9) and small intestine (12). In addition, we reported that Fas-FasL signaling also seems to affect liver cell survival/damage (9). Although we demonstrated modest evidence of apoptosis, most data for the liver point at necrotic and/or necroapoptotic/nonclassic apoptotic cell death (5). One proposed role of RIP1 was in the regulation of the induction of classic apoptotic versus programmed necrotic cell death (13); we initially tested the hypothesis that RIP1 was involved in the alteration of the apoptotic death pathway leading to "programmed necrosis" in the liver. Therefore, in our initial experiment, we simply attempted to determine the extent to which signaling via Fas-FasL led to the upregulation of RIP1 following the onset of sepsis. We found in doing anti-Fas antibody pull-down of whole-liver homogenates that when probed for the presence of RIP1 that there was a trend toward higher levels RIP1 association in the septic mice as compared with sham mice (Fig. 1).
Knockdown of RIP1 gene expression decreased (not increased) septic mouse survival
Because RIP1 appeared to be associated with Fas and presumably death receptor-induced death in the liver during sepsis, we subsequently chose to determine if RIP1 expression played a role in septic mouse survival. To test this, we knocked down expression of RIP1 via in vivo hydrodynamic delivery of RIP1 siRNA (50 μg/2 mL saline per mouse via the tail vein over 5 s) in a mouse model of sepsis and determined percentage of survival versus control siRNA-injected mice. Our rationale was that by knocking down RIP1 expression the death domain association with the Fas-associated death domain (a Fas-FasL downstream signaling intermediate) would be disrupted, thereby reducing apoptosis (in the liver) and increasing survival. To demonstrate how effective hydrodynamic administration of RIP1 siRNA was in knocking down RIP1 mRNA, we measured protein levels, via Western blot, of RIP1 in whole liver of both RIP1 siRNA and scrambled sequence siRNA (control)-treated mice in response to CLP. We noted that there was a substantial reduction in the amount of RIP1 expression evident at 24 h after RIP1 siRNA administration as compared with control siRNA treatment (Fig. 2). Surprisingly, when we assessed the effect of RIP1 siRNA pretreatment on septic mouse survival, as shown in Figure 3, we found that the RIP1 siRNA-treated animals did poorer, which was the opposite of what our initial hypothesis would have predicted. Over the course of 14 days, the animals that received RIP1 siRNA had a 22.2% survival rate (n = 18) compared with 50.0% survival (n = 18) in the control group. This indicates that RIP1 appears to be necessary for survival in septic injury of C57BL/6 mice.
Pharmacological inhibition of RIP1 kinase does not improve survival but alters NF-κB translocation in response to sepsis
The observation that inhibition of RIP1 expression actually suppressed survival in septic animals did not support our original hypothesis that; RIP1 expression was involved in septic morbidity/mortality (Fig. 3). Thus, this implies that RIP1's capacity to contribute to the onset of programmed cell death, apoptotically and/or necrotically, may not be its central role in the septic animal. Given the reported ability of RIP1 to increase survival as well as inflammatory cell signaling, via NF-κB activation and so on, in other model systems (1, 13), we alternatively hypothesized that the presence of the RIP1 protein was necessary to propagate signaling involved in supporting the animal's overall survival. To assess this, we attempted to specifically inhibit RIP1 kinase activity (i.e., death induction) while preserving the other RIP1 domains (death domain and the RIP homotypic interaction motif) for use in functions such as NF-κB signaling, and so on (14), via daily in vivo treatment with the RIP1 kinase inhibitor, Nec1 (15, 16). Similar to the study we performed with RIP1 siRNA animals, we assessed the survival of Nec1-treated mice to polymicrobial sepsis (Fig. 4). What we observed was that 25% (n = 16) of the Nec1-treated mice survived versus 50% (n = 16) of the control mice. These results were similar to the RIP1 siRNA study, suggesting that the RIP1 kinase domain is critical in the animal's survival following murine polymicrobial sepsis. To the extent that kinase inhibition via Nec1 treatment altered NF-κB nuclear translocation/activation (something that would not be anticipated 14), we noted a significant change between the degree of NF-κB nuclear translocation seen in the septic DMSO vehicle-treated mice and the Nec1-treated animals (Fig. 5), indicating that the kinase domain is necessary for NF-κB translocation.
Necrostatin 1 treatment increases markers of liver injury and proinflammatory cytokine levels
To determine if Nec1 treatment had any effect on the inflammatory response, we measured serum cytokines by cytometric bead array (Fig. 6). Necrostatin 1 treatment appears to significantly increase TNF-α, MCP-1, IL-10, and IL-6 compared with DMSO CLP group. Furthermore, Nec1 seems to have no apparent effect on liver damage during sepsis as ALT and AST activities remain the same in both Nec1 and control groups.
Necrostatin 1 treatment alters liver glycogen stores
We assessed the effect of in vivo Nec1 pretreatment on liver metabolism following the onset of sepsis. To do this, we examined the morphological changes in the liver after 2 days of pretreatment (0.125 mg Nec1 in 0.2% DMSO i.p. once per day) followed by another injection immediately after CLP. The livers were taken 24 h after CLP, fixed, and stained with PAS stain as a measurement of glycogen stores/production in the liver. We noted obvious changes in glycogen stores, in Nec1-injected CLP mice versus sham and vehicle-treated CLP mice. The sham liver sections are replete with glycogen stores, whereas vehicle-treated CLP sections lost their glycogen stores. Necrostatin 1 pretreatment appeared to preserve glycogen stores in the liver (Fig. 7). These results were confirmed by a pathologist blinded to the treatments of the samples. Both PAS and hematoxylin-eosin (H&E)-stained slides were analyzed by the pathologist. The presence of glycogen in the PAS stained slides did not correspond to any pathology with respect to the H&E slides. In fact, there was not a lot of substantial morphological changes noted in the H&E slides except for mildly reactive hepatocytes, which were seen in roughly equal distribution between all of the experimental groups.
Necrostatin 1 treatment increases activity of caspase 3 in the liver
We then interrogated the level of activated caspase 3 in the same samples we had previously stained for glycogen. Although the pathologist observed minimal signs of apoptosis, caspase 3 activity is significantly higher in Nec1-treated livers than DMSO-treated or sham livers (Fig. 8).
We have shown that mice in which we had knocked down RIP1 gene/protein expression with siRNA treatment exhibit a diminished capability to withstand a septic challenge. Furthermore, use of a pharmacological inhibitor of the RIP 1 kinase domain yields similar results. These two results indicate that the kinase domain of RIP1 is necessary for survival in a murine model of polymicrobial sepsis. We have also demonstrated by Western blot that blocking the RIP1 kinase domain by Nec1 treatment decreases NF-κB p65 nuclear translocation in septic conditions. This result supports reports that RIP1 is integral to inducing NF-κB translocation but suggests that the kinase domain may be involved, contrary to the reports that implicate the intermediate domain (17, 18). We observed these effects in the results of the serum cytokine measurements, namely, IL-6, IL-10, MCP-1, and TNF-α, being all significantly elevated in Nec1-treated mice. Increased cytokines, together with the increased levels of hepatic enzymes ALT and AST, suggests that Nec1 treatment is damaging to hepatocytes.
Through our histological studies of the livers of RIP1 gene knockdown mice after CLP, we observed an altered distribution of glycogen in the hepatocytes. Similarly, inhibition of RIP1 kinase by Nec1 also displays alteration in glycogen distribution. When contrasted with the levels of glycogen in scrambled sequence and vehicle controls, these results suggest that the kinase domain of RIP1 may contribute to cell signaling that alters cellular energy consumption during sepsis. These data are in keeping with Zhang et al. (19), who showed that RIP1 (through its homotypic interaction with RIP3) was necessary for regulation of glycogen phosphorylase activity. They further showed that RIP1 and RIP3 were necessary for activation of the metabolic enzyme, glycogen phosphorylase, in vivo. Furthermore, another group recently published that the RIP1 kinase domain is necessary for the stable formation of complex IIb (4). Based on the aforementioned published results, we suspect what might be happening in our system is that Nec1 has inhibited the kinase domain of RIP1; therefore, RIP1 and RIP3 can no longer autophosphorylate each other to regulate glycogen phosphorylase (19). As a result, more glycogen remains in the liver during the initial systemic inflammatory response, which, we presume, provides energy to activate caspases (Fig. 8). It is important to note that, RIP1, in addition to its versatile signaling capability, after initiator caspases are activated is cleaved by caspase 8, thus eliminating the necroptotic pathway. Once RIP1 is cleaved, it becomes ubiquitinylated and subsequently degraded by the proteosome (17). As we have demonstrated here, RIP1 is essential for necroptotic death through Fas, and others have reported that RIP1 is a critical glycogen metabolism regulator by its interaction with RIP3 (19, 20). Perhaps the continuous inhibition of RIP1 kinase by Nec1 during sepsis leads to a condition where, once the glycogen is consumed in the liver, apoptosis will cease. As such, as long as there is energy present, i.e., glycogen, the effect of Nec1 on the ability of RIP1 and RIP3 to associate and form a functional complex IIb might then be hidden. Once the energy is consumed, we begin to see differences in the survival of Nec1-treated animals versus control. Necrostatin 1-treated mice displayed significantly lower NF-κB translocation 24 h after CLP (Fig. 5). We would expect NF-κB translocation to be an early event peaking 2 to 4 h after CLP (21, 22). Coupled with the significant increase in systemic levels of IL-10, IL-6, MCP-1, and TNF-α at 24 h after CLP in the Nec1-treated group (Fig. 6), this suggests that the cytokines are from other sources and not exclusively from the hepatocytes.
Nevertheless, the data reported here suggest several possible roles for RIP1 in the liver during septic injury. More questions remain to be answered, such as: What is the effect of long-term Nec1 administration in these animals? As shown in the Nec1 survival study, these animals have a modest early survival benefit but go on to die in a sustained manner over time, whereas the control group survival curve plateaus. It would be interesting to examine the livers at multiple time points, i.e., 6 h and 7 days, to delineate any other changes in death markers, serum cytokines, and architecture in the liver.
Finally, recent publications have linked IKK (a regulator of NF-κB translocation), to autophagy, whereby IKK is sufficient to induce autophagy in nutritionally deprived cells (23). Given the role of RIP1 in energy metabolism, as depicted by the effect of RIP1 silencing/Nec1 treatment on liver glycogen levels, and as a member of the IKK complex, it may be worth examining the markers of autophagy after Nec1 treatment in the liver.
In conclusion, the data presented here suggest that the kinase domain of RIP1 plays a role in the alterations seen in cell survival and death in the liver during sepsis. We speculate that these confounding results are due to the dual-signaling responsibilities of RIP1. Receptor interacting protein 1 through its kinase domain can complex with RIP3 to form complex IIB, inducing necrotic death through Fas ligation. However, RIP1 can also initiate prosurvival signaling and the maintenance of a competent immune response through NF-κB activation, which may account for its effect on septic animal survival.
The authors thank Mr Paul Monfils, Ms Rosemarie Tavares, and Ms Virginia Hovanesian, Core Research Laboratories, Rhode Island Hospital, for assistance with histology and the digital morphometrics.
1. Stanger BZ, Leder P, Lee TH, Kim E, Seed B: RIP: a novel protein containing a death domain that interacts with Fas/APO-1 (CD95) in yeast and causes cell death. Cell
2. Declerq W: RIP kinases at the crossroads of cell death and survival. Cell
3. Silke J, Brink R: Regulation of TNFRSF and innate immune signaling complexes by TRAFs and cIAPs. Cell Death Differ
4. Cho YS: Phosphorylation driven assembly of the RIP1-RIP3 complex regulated programmed necrosis and virus induced inflammation. Cell
5. Ayala A, Evans TA, Chaudry IH: Does hepatocellular injury in sepsis involve apoptosis? J Surg Res
6. Baker CC, Chaudry IH, Gaines HO, Baue AE: Evaluation of factors affecting mortality rate after sepsis in murine cecal ligation and puncture model. Surgery
7. Ayala A, Herdon CD, Lehman DL, Ayala CA, Chaudry IH: Differential induction of apoptosis in lymphoid tissues during sepsis: variation in onset, frequency, and the nature of the mediators. Blood
8. Jiang X, Gregory SH, Wing EJ: Immune CD8+
T lymphocytes lyse Listeria
monocytogenes-infected hepatocytes by a classical MHC class I-restricted mechanism. J Immunol
9. Wesche-Soldato DE, Chung CS, Lomas-Neira JL, Doughty LA, Gregory SH, Ayala A: In vivo
delivery of caspase 8 or Fas siRNA improves the survival of septic mice. Blood
10. Song E, Lee S-K, Wang J, Ince N, Ouyang N, Min J, Chen J, Shankar P, Lieberman J: RNA interference targeting Fas protects mice from fulminant hepatitis. Nat Med
11. Wesche-Soldato DE, Chung CS, Gregory SH, Salazar-Mather TP, Ayala CA, Ayala A: CD8+
T cells promote inflammation and apoptosis in the liver after sepsis: role of Fas-FasL. Am J Pathol
12. Stromberg PE, Woolsey CA, Clark AT, Clark JA, Turnbull IR, McConnell KW, Chang KC, Chung CS, Ayala A, Buchman TG, et al: CD4+
lymphocytes control gut epithelial apoptosis and mediate survival in sepsis. FASEB J
13. Grimm S, Stanger BZ, Leder P: RIP and FADD: two "death domain"-containing proteins can induce apoptosis by convergent, but dissociable, pathways. Proc Natl Acad Sci
14. Holler N, Zaru R, Micheau O, Thome M, Attinger A, Valitutti S, Bodmer JL, Schneider P, Seed B, Tschopp J: Fas triggers an alternative, caspase-8-independent cell death pathway using the kinase RIP as effector molecule. Nat Immunol
15. Vandenabeele P, Declercq W, Vanden Berghe T: Necrotic cell death and `necrostatins': now we can control cellular explosion. Trends Biochem Sci
16. Degterev A, Hitomi J, Germscheid M, Ch'en IL, Korkina O, Teng X, Abbott D, Cuny GD, Yuan C, Wagner G, Hedrick SM, Gerber SA, Lugovskoy A, Yuan J: Identification of RIP1 kinase as a specific cellular target of necrostatins. Nat Chem Biol
17. Xu EA: Ubiquitin specific peptidase 21 (usb21) inhibits tumor necrosis factor a (TNFa)-induced nuclear factor kappa b (NF-kB) activation via binding to and deubiquitinating receptor interacting protein 1 (RIP1). J Biol Chem
18. Zhang D-W, Shao J, Lin J, Zhang N, Lu B-Z, Lin S-C, Dong M-Q, Han J: RIP3 an energy metabolism regulator that switches TNF induced cell death from apoptosis to necrosis. Science
19. Lim SY, Davidson SM, Mocanu MM, Yellon DM, Smith CC: The cardioprotective effect of necrostatin requires the cyclophilin-D component of the mitochondrial permeability transition pore. Cardiovasc Drugs Ther
20. Wertz E, Dixit VM: Regulation of death receptor signaling by the ubiquitin system. Cell Death Differ
21. Vandenabeele P: The role of kinases RIP1 and RIP3 in TNF induced necrosis. Sci Signal
22. Zingarelli B, Piraino G, Hake PW, O'Connor M, Denenberg A, Fan H, Cook JA: Peroxisome proliferator-activated receptor δ regulates inflammation via NFκB signaling in polymicrobial sepsis. Am J Pathol
23. Criollo A: The IKK complex contributes to the induction of autophagy. EMBOJ