Caspase-3–Dependent Organ Apoptosis Early After Burn Injury

Abstract

Objective

To examine the role played by endotoxin, tumor necrosis factor-alpha (TNF-α), and caspase-3 in the increased apoptosis seen in solid organs in the early period after a burn injury.

Summary Background Data

Burn injury is often associated with immune suppression. Bacterial translocation and systemic endotoxemia have been reported after a burn injury, and caspase-3 activation due to TNF-α and Fas ligand (FasL) are presumed to initiate apoptosis. We hypothesized that endotoxin-induced TNF-α expression and caspase-3 activation could be the stimulus for the apoptosis after burn injury.

Methods

A 20% full-thickness scald burn was used in C57BL/6 mice. Three hours after burn injury, tissue samples were obtained from the thymus, lung, liver, and spleen. Lipopolysaccharide-nonresponsive (C3H/HeJ) and TNFα null B6x129^tnf−/− mice were also used. To detect apoptosis, hematoxylin and eosin stain, in situTUNEL, DNA extraction, and gel electrophoresis were all performed. Caspase-3 activity and TNF-α and FasL mRNA were also measured.

Results

Increased apoptosis and caspase-3 activity were observed in the thymus and spleen 3 hours after burn injury but were not seen in liver or lung. In the thymus and spleen, increased expression of FasL mRNA was also observed, whereas increased TNF-α mRNA was not. Increased apoptosis in thymus and spleen were also observed in C3H/HeJ and B6x129^tnf−/− mice after burn injury. An inhibitor of the caspase-3 (Z-VAD-fmk) reduced apoptosis in both thymus and spleen.

Conclusions

In the early period after a burn injury, increased apoptosis is observed primarily in the lymphoid organs and is independent of endotoxin or TNF-α. The increased caspase-3 activity in thymus and spleen contributes to apoptosis in these organs.

Burn injury is frequently associated with immune suppression and a loss of lymphocyte subpopulations from the blood and lymphoid organs. Increased apoptosis of circulating lymphocytes has been reported after burn injury, ¹ and circulating tumor necrosis factor-alpha (TNF-α) has been occasionally detected in the serum of patients with burns. ² However, the mechanisms and mediators inducing apoptosis in lymphoid and parenchymal organs after burn injury remain unresolved.

Although little is known about apoptotic processes after burn injury, apoptosis is frequently increased during inflammation and can be induced in both lymphoid and parenchymal tissues by cytokines such as TNF-α or Fas ligand (FasL). ³ Increased apoptosis in solid organs has also been reported in sepsis. ^4,5 TNF-α signaling by means of its type I receptor and FasL signaling by CD95/Fas/Apo1 induce concatermization of signal transduction proteins with death domains, and activation of a novel family of cysteine-aspartate proteases (caspases). ^6,7 The caspase-3 family (CPP-32 like) appears to play a dominant role in the downstream signaling of apoptotic death ⁸ and in the apoptotic liver injury secondary to TNF-α ⁹ or FasL. ¹⁰

In this study, we examined the degree of apoptosis in solid organs after a burn injury. Because gut barrier function is maximally impaired early after burn injury, and elevated plasma endotoxin levels have been reported in burned mice, ¹¹ we also used a genetic approach to investigate the role of endotoxin and TNF-α in apoptosis after burn injury. In addition, to establish the signal transduction pathway inducing apoptosis after burn injury, we examined caspase-3 activity and its inhibition on apoptosis. The results suggest that increased apoptosis is seen early after a burn injury in spleen and thymus and is not dependent on either lipopolysaccharide (LPS) or TNF-α, but proceeds through a caspase-3–dependent pathway.

MATERIALS AND METHODS

Animals

Specific free pathogen-free female C57BL/6j, C3H/HeJ (LPS-nonresponsive) and C3H/HeN (LPS-responsive) mice 8 to 10 weeks of age were obtained from Charles River Breeding Laboratory (Wilmington, MA). B6x129 TNF-α null (tnf−/−) and wild-type female (B6x129^tnf+/+) mice were bred and maintained at the University of Florida Health Science Center Animal Resources Department. tnf−/− mice were originally obtained from Amgen, Inc. (Boulder, CO). All mice were housed in a barrier facility with viral pathogen surveillance, sterile cages, food and water, and high-efficiency particle-free air. All protocols were approved by the Institutional Animal Care and Use Committee at the University of Florida College of Medicine before initiation of these studies. Anesthesia and euthanasia protocols were consistent with the recommendations of the American Veterinary Medical Association.

Experimental Protocols

Mice were randomized into burn (n = 12), sham burn (n = 12), or control (no treatment; C57BL/6j only) groups. Each animal was anesthetized with an intraperitoneal injection of 50 mg/kg pentobarbital sodium, and all dorsal hair was clipped. The clipped skin on the dorsum of the burn group was exposed to steam through an insulated template for 7 seconds, and a 20% total body surface area full-thickness burn was obtained. Animals were immediately resuscitated with 0.08 ml/g of saline intraperitoneally after burn injury. Sham-burned animals underwent the same anesthesia and resuscitation procedures as those in the burn group but did not receive a burn injury. All of the mice were anesthetized and killed by cervical neck dislocation at 3 hours after burn injury. Tissue samples were taken from the thymus, lung, liver, and spleen. Pieces of the organs were immediately homogenized for caspase-3 activity measurements; other parts of each organ were snap-frozen in liquid nitrogen and stored at −80°C for analysis of mRNA, or fixed in 4% phosphate-buffered formalin for histologic analysis. An additional 17 C57BL/6j mice received 15 mg/kg of the caspase-3 inhibitor benzyloxycarbonyl-Val-Ala-Asp-fluoromethylketone (Z-VAD-fmk) (Enzyme Systems Products, Livermore, CA) at 30 minutes before burn and at 0, 1, and 2 hours after burn. Z-VAD-fmk was predissolved in dimethyl sulfoxide (DMSO) at a concentration of 200 mM, diluted with saline, and given in a volume of 300 μl. Seventeen vehicle control animals were given 300 μl of saline with DMSO (1%) at the same time.

Histopathologic Examination

The organs were fixed in 4% buffered formalin and embedded in paraffin. Five-micrometer-thick sections were affixed to slides, deparaffinized, and stained with hematoxylin and eosin to assess morphologic changes. In situ terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) assay was performed using an in situ apoptosis detection kit (Apotag, Oncor Corp., Gaithersburg, MD). All steps were performed according to the supplier’s instructions. Briefly, paraffin-embedded sections were dewaxed and rehydrated and then incubated with proteinase K (20 μg/ml in 100 mM Tris and 50 mM ethylenediaminetetraacetic acid) for 15 minutes at 25°C. After the slides were washed four times with distilled water, the sections were incubated in equilibration buffer for 5 minutes. The sections were then incubated with the labeling solution containing terminal deoxynucleotidyl transferase (TdT) in a humidified chamber for 1 hour at 37°C. The reactions were terminated by rinsing the sections in a stop/wash buffer. The sections were incubated with antidigoxigenin fluorescein for 30 minutes at room temperature and then rinsed three times in phosphate-buffered saline. The fluorescent TUNEL-labeled slides were photographed using a fluorescence microscope. Tissues stained with hematoxylin and eosin were also examined.

DNA Extraction and Qualitative Analysis of DNA Fragmentation by Gel Electrophoresis

DNA extraction and gel electrophoresis were performed for thymus and spleen to confirm apoptosis. Cells were disaggregated mechanically in a 24-well microtiter plate using the flat end of the handle of a sterile tuberculin syringe. Disaggregated cells (1 × 10⁶) were washed and pelleted in phosphate-buffered salt solution twice and lysed in 100 μl lysis buffer (10 mM Tris-HCl, 10 mM EDTA, 0.5% Triton X-100) for 30 minutes at 4°C. After centrifugation at 16,000 rpm for 20 minutes, the aqueous layer was incubated with 2 μl RNase (20 mg/ml) for 1 hour at 37°C, then incubated with 2 μl proteinase K (20 mg/ml) for 1 hour at 37°C. Samples were mixed with 20 μl 5 M NaCl and 120 μl isopropanol and incubated at −20°C overnight. After centrifugation at 16,000 rpm for 15 minutes, the pellet was resuspended in 20 μl Tris-EDTA buffer. The concentration of DNA in each sample was estimated by measurement of optical density at 260 nm; specimens containing 5 μg of DNA were then applied to 1.7% agarose gels and subjected to electrophoresis at 110 V. The gels were stained with ethidium bromide and photographed under ultraviolet illumination.

Caspase-3 Activity Assay

Protein extracts were prepared by homogenization of 20 mg of tissue, and caspase-3 activity was measured as previously described. ^10–12 Briefly, excised organs were homogenized in 25 mM HEPES buffer (pH 7.5) containing 5 mM MgCl₂, 1 mM ethylene glycol-bis (β-aminoethyl ether) N,N,N′,N′-tetraacetic acid (EGTA), 1 mM phenylmethylsulphonyl fluoride, and 1 μg/ml leupeptin and aprotinin. After centrifugation at 15,000 rpm for 10 minutes, the supernatants were collected. Forty micrograms of the extracted proteins was incubated with the synthetic fluorescent substrates benzyloxycarbonyl-Asp-Glu-Val-Asp-7-amino-4-trifluoromethylcoumarin (Z-DEVD-AFC; Enzyme Systems Products) for caspase-3 activity assay at a concentration of 20 mM in 0.1 M HEPES buffer (pH 7.4) containing 2 mM dithiothreitol, 0.1% CHAPS (3-[(3-cholamidopropyl) dimethylammonio]-1-propane-sulfonate) and 10% sucrose. The kinetics of the proteolytic cleavage of the substrates were monitored in a fluorescence microreader using an excitation wavelength of 400 nm and an emission wavelength of 505 nm. The fluorescence intensity was calibrated with standard concentration of AFC, and the caspase-3 activity was calculated from the slope of the recorded fluorescence and expressed in pmol per minute per mg of protein. Protein concentrations in the supernatant were assayed using the Bio-Rad Protein Assay kit (Bio-Rad Laboratories, Hercules, CA).

For the inhibitor studies to confirm the correlation between protease activity and signal detection in vitro, N-acetyl-Asp-Glu-Val-Asp-CHO (Ac-DEVD-CHO) (Pharmingen, San Diego, CA) was added to the supernatant 20 minutes before adding the substrate Z-DEVD-AFC.

Detection of TNF-α and FasL mRNA

Total cellular RNA was isolated, and estimated quantities of TNF-α and FasL mRNA were calculated as previously described. ¹³ Total organ RNA was isolated by guanidinium isothiocyanate and acid-phenol extraction. One microgram of total organ RNA was reverse-transcribed and then amplified using primers for murine TNF-α, FasL, and Cu/Zn superoxide dismutase (SOD) as an internal control. The sequence of the oligonucleotide primers was as follows: 5′-TNF, ATG AGC ACA GAA AGC ATG ATC; 3′-TNF, TAC AGG CTT GTC ACT CGA ATT; 5′-FasL, ATC AGC TCT TCC ACC TGC AGA AGC AAC; 3′-FasL, AGT TCA ACC TCT TCT CCT CCA TTA GCA CC; 5′-SOD, GTC TGC GTG CTG AAG GGC GAC; 3′-SOD, TCT CCT GAG AGT GAG ATC ACA. The polymerase chain reaction was performed using 2.5 U AmpliTaq (Perkin Elmer, Norwalk, CT) for TNF-α; 27 cycles, FasL; 28 cycles, SOD, 21 cycles as follows: 94°C for 1 minute (dissociation), 60°C for 1 minute (annealing), and 72°C for 2 minutes (primer extension). The expected fragment lengths were 276 bp for TNF-α, 390 bp for FasL, and 314 bp for Cu/Zn SOD. Amplicons were visualized using 2% agarose gel electrophoresis. The gels were scanned and the integrated area under the absorbance curves was calculated using a commercial program (SigmaGel, Jandel Scientific, San Rafael, CA). The relative quantities of TNF-α and FasL mRNA are presented as the ratio between the intensity of these bands relative to the intensity of the housekeeping gene, Cu/Zn SOD.

Statistical Analysis

All data are given as mean ± SEM. To determine statistical significance, a one-way analysis of variance with Bonferroni’s t test post hoc comparison was performed. Statistical significance was determined at the 95% confidence interval.

RESULTS

Increased Apoptosis in Thymus and Spleen Early After Burn

C57BL/6j mice received a burn injury and were killed at 3 hours. The thymus, lung, liver, and spleen were then examined histologically and compared with organs from C57BL/6j mice that did not receive a thermal injury. Histologically, lymphoid cells in the thymus and spleen from burned mice showed characteristic shrinking, chromatin clumping, and nuclear fragmentation consistent with increased apoptosis (Fig. 1). In the thymus and spleen of burned mice, TUNEL staining revealed increased numbers of fluorescent stained nuclei consistent with apoptosis, whereas in the lung, the number of apoptotic nuclei was only modestly increased, and in the liver there was no evidence of increased apoptosis. These apoptotic thymocytes and splenocytes tended to form clusters; however, these clusters of apoptotic cells were not seen in organs from sham-treated animals. In addition, DNA extracted from isolated cells in the thymus and spleen after burn injury revealed a typical 180 base pairs of DNA fragmentation pattern in ladder form consistent with apoptosis (Fig. 2).

Figure 1. (A) Thymus and (B) spleen tissue after burn injury. Hematoxylin and eosin stain showed apoptosis in lymphoid cells in the thymus and spleen after burn injury. Arrows identify cells undergoing pyknosis and karyorrhexis (magnification ×400). (C) Detection of DNA fragmentation in thymus, lung, liver, and spleen using in situ TUNEL method, as described in text. Increased apoptosis was seen in thymus and spleen after burn (×400). Data shown are representative of three independent experiments.

Figure 2. DNA was extracted from disaggregated cells of the thymus and spleen and electrophoresed in agarose gel, as described in text. S, sham group; B, burn group. DNA fragmentation was observed in thymus and spleen of burned animals but not sham-burned mice. Although not shown, DNA laddering was not seen in liver or lung specimens from either burned or sham-burned mice.

Upregulation of FasL mRNA of Thymus and Spleen After Burn

Thymus, lung, liver, and spleen were harvested 3 hours after a burn injury, and total organ RNA was isolated, reverse-transcribed, and amplified by polymerase chain reaction using primers for TNF-α, FasL, and the Cu/Zn SOD as an internal control. As shown in Figure 3, TNF-α mRNA levels were not increased in organs at this early time point, but there was increased FasL mRNA in thymus and spleen after burn injury. There was no increase of expression of FasL mRNA in lung and liver.

Figure 3. Expression of TNF-α and FasL mRNA in thymus, lung, liver, and spleen. (A) Photograph of ethidium bromide-stained amplicons. C, control; S, sham; B, burn. Relative band intensity of TNF-α (B) and FasL (C) after normalization for Cu/Zn SOD mRNA used as an internal control. Gels were scanned as described in text. Upregulation of thymus and spleen FasL mRNA expression was observed after burn injury, whereas increased TNF-α mRNA expression was not observed. Gel data are from a single experiment, whereas the bar graphs represent the mean from three separate experiments.

Upregulation of Caspase-3 Levels in Thymus and Spleen After Burn Injury but not in Z-VAD-fmk Mice

Caspase-3 activity was significantly increased in the spleen and thymus of burned animals (Fig. 4). In contrast, no increased protease activity was observed in lung and liver. Further, the increase in caspase-3 activity observed in the thymus and spleen of burned mice could be blocked by administration of Z-VAD-fmk before and after burn injury, and caspase-3 activity in Z-VAD-fmk mice was comparable to that seen in the sham group (Fig. 5). Treatment of mice with Z-VAD-fmk reduced apoptosis in both thymus and spleen of burned mice, whereas apoptosis remained increased in all thymus and spleen samples of burned mice that did not received Z-VAD-fmk (Fig. 6).

Figure 4. Caspase-3 activity in thymus, lung, liver, and spleen. Caspase-3 activity was determined by measuring the proteolytic cleavage of the caspase-3 substrate Z-DEVD-AFC. Increased caspase-3 activity was observed in thymus and spleen of burned animals. ^*, p < 0.05 vs. control and sham animals. Values represent the mean and standard error of five animals per group.

Figure 5. Caspase-3 activity in thymus and spleen after burn from mice pretreated with Z-VAD-fmk as an inhibitor of caspase-3. Z-VAD-fmk prevented increasing caspase-3 activity in thymus and spleen after burn. ^*, p < 0.05 vs. sham animals and mice treated with Z-VAD-fmk. Values represent the mean and standard error of five animals per group.

Figure 6. In situ TUNEL stain of thymus (A, B) and spleen (C, D) from sham and burn vehicle animals and animals treated with Z-VAD-fmk. The fluorescence microscopic observation showed a reduction in apoptosis in thymus and spleen by Z-VAD-fmk after burn injury (B, D); however, apoptosis was increased in thymus and spleen of the nontreatment group (A, C).

Increased Apoptosis in B6x129^tnf−/− and LPS Nonresponder Mice After Burn Injury

In thymus and spleen of TNF-α null (B6x129^tnf−/−) and LPS nonresponder (C3H/HeJ) mice, apoptosis was also increased after burn injury (Fig. 7). The magnitude of apoptosis in each burned mouse was the same as that seen in wild-type mice of the same genetic background (B6x129^tnf+/+ and C3H/HeN) (data not shown). Caspase-3 activity in thymus and spleen was also increased in the B6x129^tnf−/− and C3H/HeJ mice, as well as in each genetic background after burn injury (Fig. 8).

Figure 7. In situ TUNEL stain of tissue section of thymus and spleen from TNF-α null (tnf−/−) and LPS nonresponder (C3H/HeJ) mice after burn injury. Increased apoptosis was detected in thymus and spleen after thermal injury. The extent of apoptosis in each organ was comparable to that seen in organs from the respective background mice (data not shown).

Figure 8. Caspase-3 activity in thymus and spleen from B6x129^tnf−/−, C3H/HeJ, and appropriate background (C3H/HeN and B6x129^tnf+/+) mice. Increased caspase-3 activities were observed in thymus and spleen of both strains as compared to appropriate genetic background mice after burn injury. ^*, p < 0.05 vs. and sham animals. Values represent the mean and standard error of five animals per group.

DISCUSSION

In this report, we demonstrate that in the early postburn period, increased apoptosis was observed primarily in lymphoid organs, such as spleen and thymus, but was not seen in the liver or lungs. TNF-α expression did not appear to be increased in this early burn period, and the increased apoptosis was independent of both TNF-α and endotoxin. The increased apoptosis was associated with increased caspase-3 activity in both thymus and spleen of burned mice. Caspase-3 activity in thymus and spleen after burn injury was reduced by Z-VAD-fmk administration to the level seen in nonburned mice, and apoptosis was also markedly reduced, albeit not completely.

We purposefully focused on the early (3 hours) response to a burn injury to dissect the relations between proinflammatory cytokine responses and apoptotic pathways in solid organs. In focusing on the immediate postburn period, secondary responses to organ injury, such as wound infection, anorexia, and nutritional depletion, could be avoided. These acute studies were also amenable to the caspase-3 inhibitor treatment (Z-VAD-fmk), which has a very short biologic half-life. However, the studies do not reveal the more prolonged and adaptive changes that occur in the days and weeks after a burn injury.

A genetic approach was used to identify a possible role for TNF-α and endotoxin in caspase-3 induction and solid organ apoptosis from the burned animals. Bacterial translocation and systemic endotoxemia occur early after a burn injury in the mouse, ¹⁴ and we hypothesized that endotoxin-induced TNF-α expression could be the stimulus for the apoptotic injury. However, the results of our experiments using endotoxin-nonresponding (C3H/HeJ) and TNF-α null (B6x129^tnf−/−) mice suggest that the increased caspase-3 activity and organ apoptosis were independent of either endotoxin or TNF-α. The latter finding is consistent with a lack of any increase in TNF-α mRNA expression from the organs of burned, wild-type mice at this early time point. These findings are also consistent with recent reports ofTNF-α–independent organ apoptosis in a murine sepsis model. ¹⁵

In contrast to TNF-α, FasL expression was increased in spleen and thymus from the burned mice. We could not use the same genetic approach to determine experimentally whether increased FasL expression was contributing to this apoptosis because of phenotypic alterations in the healthy Fas-deficient lpr mouse. Unlike TNF-α null mice, lpr mice are not phenotypically normal but exhibit widespread lymphoproliferative derangements as they age. ¹⁶ More importantly, they respond to an inflammatory challenge, such as Staphylococcus enterotoxin, with an exaggerated compensatory proinflammatory cytokine response, with increased TNF-α, interferon-γ,and interleukin-12 expression. This exaggerated proinflammatory cytokine response makes it difficult to ascertain whether the observed responses are the result of either the absence of a Fas/FasL signaling system, or to increased TNF-α, interferon, or IL-12 production.

Although FasL could contribute to the apoptotic injury, glucocorticoids might also play a consequential role in lymphocyte apoptosis. Increased circulating glucocorticoid concentrations have been well described early after burn injury, ¹⁷ and administration of a glucocorticoid receptor antagonist prevented burn-induced thymic apoptosis in rats. ¹⁸ Alam et al ¹⁹ reported increased caspase-3 activity during thymocyte apoptosis induced by dexamethasone. Z-VAD-fmk administration could prevent apoptosis secondary to corticosteroid administration; however, the pathways of steroid-induced apoptosis are not well known in spleen or thymus. ^20,21 A previous study has demonstrated that although both thymus and spleen T cells underwent apoptosis in response to similar concentrations of dexamethasone, a greater proportion of spleen T cells were dexamethasone resistant. ²² Similarly, increased apoptosis was not seen in spleen during polymicrobial sepsis, whereas the increased apoptosis observed in thymus appeared to result from glucocorticoids, ²³ suggesting that the responses and mechanisms determining increased splenocyte apoptosis are probably multifactorial in nature.

In addition to identifying the mediator(s) responsible for initiation of lymphoid tissue apoptosis, our studies were also aimed at dissecting the contribution of caspase-3–mediated signal transduction pathways to this process. Several signal transduction pathways leading to apoptosis have been recently described, although their role in burn injury has not been investigated. These pathways link surface receptors such as Fas or the p55 TNF receptor with a family of caspases responsible for initiation of programmed cell death. TNF and FasL signaling triggers caspase-3 activation, resulting in DNA strand breakage, nuclear condensation, and apoptosis. ²⁴

Regardless of the initiating stimulus, we can demonstrate increased caspase-3 activity in organs containing cells undergoing apoptosis (thymus and spleen) after burn injury, whereas caspase-3 activity was not increased in organs where cells were not undergoing apoptosis (liver and lung). These findings are therefore in agreement with those of Hotchkiss et al, ²⁵ who observed increased thymus and spleen cell apoptosis in mice undergoing cecal ligation and puncture. They also observed increased apoptosis in lymphoid cells recovered from other organs, including lung, liver, and skeletal muscle, a finding we did not observe. In addition, our results are only in partial agreement with the findings of Ayala et al ²³ and Nakanishi et al, ¹⁸ who observed increased apoptosis in thymus, but not spleen, after cecal ligation and puncture, and a thermal injury, respectively.

In a recent report, repeated administration of Z-VAD-fmk was shown to prevent liver apoptosis 6 hours after endotoxin injection. ²⁶ We were able to demonstrate that Z-VAD-fmk treatment completely blocked caspase-3 activity in the spleen and thymus of burned animals, and also reduced cellular apoptosis in both organs. Although caspases represent a common pathway in the apoptotic death of most lymphoid cells, it is unclear whether different lymphocyte populations may use distinct and separate caspase family proteases to activate apoptosis. For example, Sarin et al ²⁷ observed that apoptotic cell death in murine CTLL-2 cells induced by dexamethasone could be inhibited by a broad-acting caspase inhibitor (BD-fmk) but not the more specific CPP32-like (caspase-3) inhibitor Z-VAD-fmk. Although the findings suggest that caspase-3-like proteases may not be fully responsible for inducing apoptosis in all lymphocyte populations, we conclude that primarily caspase-3–dependent pathways may contribute to the increases of lymphoid cell apoptosis after a burn injury in spleen and thymus.

Activated T cells are known to be removed by apoptosis to maintain cellular homeostasis. ²⁸ In this manner, activation- induced cell death may protect the host from excessive immune stimulation. However, during exaggerated inflammatory responses, as may occur after major burn injury, activation-induced cell death may lead to an inappropriate T-cell removal and immunologic dyscrasia. Under these experimental conditions, inhibitors of caspase-3 may prevent immunodeficiency without affecting cytokine-induced signal transduction pathways after burn injury.

Discussion

Dr. John A. Mannick (Boston, Massachusetts): The questions I have for Dr. Mozingo relate to what is dying in his apoptotic organs. The paper quoted “in human” which indicates that T cells circulating in burned humans are undergoing apoptosis. It is a little difficult to interpret because no really rigorous identification of the cell types was ever done in that study. We have tried to reproduce it and simply cannot find evidence of circulating T-cell apoptosis within a couple of weeks of injury after either traumatic injury or burn injury in humans—that is, nonseptic injured humans.

The question I have, therefore, is in these burned mice early after injury, what cell types are dying, David? We have also looked at this problem a bit in the burned mouse model, which is very similar to the one you are using. And we have been particularly interested in the death of helper lymphocytes, T-helper lymphocytes, and have not found them to be killing themselves.

One thing we have noticed in burned spleens is that there is an influx rather quickly after burn injury of PMNs. Now they congregate around vessels and, as you know, PMNs may not kill themselves in burns as fast as they do in normal animals, but they still have an enthusiasm for suicide that is seen only in 17th century samurai. And I wonder if some of the dying cells may be these PMNs that have gotten into the splenic sinusoids right after injury. So I wonder if you could enlighten us about what is dying in these preparations.

Dr. David N. Herndon (Galveston, Texas): The authors have shown increased apoptosis, fas ligand, and caspase-3 activity 3 hours postburn. Apoptosis in their model was prevented by inhibition of caspase-3 with a specific short half-life antibody. My questions are several.

One, can you use this antibody in patients?

Have you tried to block fas ligand activity?

A study in Science published on October 16 by Flora et al from Geneva showed increased apoptosis in epithelial cells in patients with Stevens-Johnson syndrome, a syndrome similar to burn injury in many ways. This was associated with increased fas ligand lytic activity which could be inhibited in vitro and, very interestingly, in vivo, with pooled human immunoglobulin, which has fas ligand lytic antibody activity. When this was given to these patients in this article on Stevens-Johnson syndrome it actually decreased epithelial cell fas ligand lytic activity and apoptosis and improved clinical outcome in this entity that is very similar to burns in many ways. Have you looked at epithelial cell apoptosis in this model or an immunoglobulin blockade of apoptosis?

My fourth question: gut-associated lymphoid tissue withers morphologically after burn when looked at on autopsy. Did you look for apoptosis in gut-associated lymphoid tissue?

In animal models, others—Dr. Chaudry, who I believe is present—showed glucocorticoid antagonists reversed apoptosis by similar mechanisms. Have you tried this in your model?

Finally, teleologically, why do you think cell death occurs in thymus and spleen postburn? What do you think the signal pathway for apoptosis is? If you can answer that, you’ll probably get the Nobel prize.

Dr. Basil A. Pruitt, Jr. (San Antonio, Texas): In this nice paper, Dr. Mozingo and Dr. Copeland and their colleagues have identified an organ-specific apoptosis as a possible contributor to postburn or postinjury immunosuppression. Since most physiologic changes occur in a dose-proportional fashion, have you studied animals with lesser and greater burns? Or is there some threshold size of burn above which apoptosis in thymus and spleen are inhibited?

You mentioned that steroids can invoke or promote apoptosis but ignore the catechols, which are markedly elevated within the first 3 hours after burn injury. And the catechols are known to stimulate apoptosis, especially in lymphatic tissues.

To define the role of the catechols, have you repeated the experiments with either pre- or postinjury adrenergic blockade? And, if so, does such blockade reduce apoptosis?

Is Z-VAD-fmk given intraperitoneally, and, if so, does it cause peritonitis or exaggerate translocation by inhibiting apoptosis? If Z-VAD-fmk does inhibit apoptosis, is there a risk that it will promote or exaggerate the systemic inflammatory response syndrome state?

In Figure 3, it appears as if the liver and lung produced more SOD under all circumstances than do the thymus and the spleen. Is that another manifestation of organ specificity, or a reflection of greater local need to scavenge oxygen radicals in those organs?

Are these organ-specific changes simply manifestations of the response to injury, with liver cells protected from apoptosis to facilitate its protein synthetic activity and the thymus and spleen simply filtering out leukocytes damaged by the thermal injury per se?

In the TNF-α null mice, it seems possible that fas ligand and even P53 could mediate an increase in apoptosis after injury. Have you examined the effect of a fas ligand neutralizing antibody such as NOK-1 in your model?

Since very few infections occur within 3 hours of injury, what is the clinical significant of your findings, and just how long is thymic or splenic apoptosis increased, and does it correlate with other deficits of lymphocyte function?

Dr. William C. Cioffi, Jr. (Providence, Rhode Island): I have basically four questions for Dr. Mozingo, and they echo some of the previous questioners.

This study was conducted very early following injury; and, although the data is sound, what evidence do you have that there indeed is a proinflammatory state at 3 hours in your model?

More importantly, as Dr. Pruitt alluded to, immune cell dysfunction occurs much later following injury. Are the processes of early apoptosis and immune dysfunction related?

I would echo Dr. Mannick’s question of exactly what cell populations in the thymus and spleen are undergoing apoptosis, and what percentage of various T cells are undergoing apoptosis? You have nicely executed TNF and endotoxin as mediators, but you did note the increase in fas ligand. Given that you used the genetic approach that you did, why did you not use fas-ligand–deficient GLD mice to examine this? The lymphoproliferative disorder that they develop occurs late in life, and you could, obviously, run controls to protect against that.

Finally, 3 hours is a very short time period to see classic induction of activation-induced cell death. Similar to what Dr. Pruitt asked, what other potential mechanisms could explain your findings? Glucocorticoids, catechols, et cetera? But this seems an incredibly short time period to see what has classically been described in the literature.

Dr. David W. Mozingo (Closing Discussion): I’d like to thank my discussants for their insightful comments. First of all, Dr. Mannick, you asked what is dying in the apoptotic organs. And we are currently looking at surface markers of the cells that we are identifying to be apoptotic, and I don’t have the whole story at the moment, but we do feel that they are probably mostly immature T cells. We have not broken down the subtypes, but they do seem to be of that origin.

As far as polymorphonuclear leukocytes being involved in this, we do see some early PMN infiltration into the thymus but not into the spleen. But we have particularly chosen an early time period, which I will discuss a little more later, to, hopefully, not have too much effect of other white cells.

Dr. Herndon, caspase inhibition probably can be used in humans. The caspase inhibitor is a small polypeptide that is a component of the site that the caspase recognizes that is similar between species. And it is a small polypeptide that has been targeted by some pharmaceutical companies as a probable product.

As far as blocking fas ligand, this will answer a question from, I think, Dr. Pruitt, and Dr. Cioffi as well. It is that though we did not use any genetically engineered mice because of their problems with abnormal immunodeficiency, we did use a fas ligand fusion protein, which is similar to what has been alluded to by some of the questioners, that when we gave this fas fusion protein, which is an immunoadhesant to fas ligand, it did not block apoptosis at all. So it appears that fas ligand was not involved in this early signal.

We did not examine epithelial cell apoptosis other than to look at the margins of the burn wound at 3 hours and did not see any apoptosis in those areas.

We did not examine gut-associated lymphoid tissue. We have examined the role of glucocorticoids in apoptosis by administering the glucocorticoid antagonist RU-486 and found that this completely blocks the apoptotic response at 3 hours in the thymus but incompletely blocks the response in the spleen.

For the purely illogical Nobel Prize question, we looked at this at 3 hours, so we were looking at the earliest changes that would not be influenced by colonization or infection in the burn wound with bacteria. It would not be related to the inflammatory cell infiltrate that occurs in the wound and the less systemic activation of white cells. And as it turns out, with the lack of these proinflammatory cytokines and other mediators, lack of their ability to induce apoptosis and the fact that we are seeing that glucocorticoid and probably—in response to Dr. Pruitt’s question—epinephrine will probably do the same thing, inducing apoptosis. We may be looking at the very early changes due to the neurohumoral response and not to a cellular mediated response.

Dr. Pruitt, we did not look at different burn sizes, though we did do a few studies in an early excision and immediate closure, the wound that did not block the apoptosis. The Z-VAD was administered intravenously, so the presence or absence of translocation would probably not affect that.

The liver cells—we are not stating that they are protected from apoptosis necessarily. There are many less lymphoid cells through the liver, and this early change—probably, as I said, mediated by a neurohumoral mechanism—is not related to the other studies that have induced apoptosis in the liver, like the hepatitis studies and the sepsis studies, in which you see more liver apoptosis that is probably white-cell–mediated.

As far as the duration, we have also measured this at 6 hours, and that seems to be the time of peak apoptosis, and it does persist, but diminish at 12 to 24 hours. Again, 3 hours was chosen so that we diminished the other white cell responses.

Dr. Cioffi asked is there a proinflammatory state at 3 hours. I think, again, the neurohumoral response predominates and the inflammatory response probably is more profound at 6 to 12 hours after this injury, as indicated by some previous findings of Dr. Herndon.

We did not use the fas-ligand–deficient GLD mice, as we had the fas fusion protein available and could do it pharmacologically.