- Email: [email protected]
Heparin-binding epidermal growth factor–like growth factor reduces intestinal apoptosis in neonatal rats with necrotizing enterocolitis
Journal of Pediatric Surgery (2006) 41, 742 – 747
www.elsevier.com/locate/jpedsurg
Heparin-binding epidermal growth factor–like growth factor reduces intestinal apoptosis in neonatal rats with necrotizing enterocolitisB Jiexiong Feng, Osama N. El-Assal, Gail E. Besner* Department of Pediatric Surgery, Center for Cell and Vascular Biology, Children’s Research Institute, Columbus, OH 43205, USA and The Ohio State University College of Medicine and Public Health, Columbus, OH 43210, USA Index words: Necrotizing enterocolitis; HB-EGF; Apoptosis; TUNEL; Caspase 3
Abstract Purpose: We have previously demonstrated that enterally administered heparin-binding epidermal growth factor–like growth factor (HB-EGF) decreases the incidence and severity of necrotizing enterocolitis (NEC) in a neonatal rat model. Because apoptosis contributes to gut barrier failure in this model, the aim of this study was to investigate the effect of HB-EGF on apoptosis during the development of NEC. Methods: NEC was induced in neonatal rats by exposure to hypoxia, hypothermia, hypertonic formula feeding (HHHTF) plus enteral administration of lipopolysaccharide (LPS). Fifty-one neonatal rats were randomly divided into the following groups: (1) breast-fed (BF), (2) HHHTF + LPS, and (3) HHHTF + LPS with HB-EGF (600 lg/kg) added to the formula. NEC was evaluated using a standard histological scoring system. Apoptotic cells in intestinal tissues were detected by terminal deoxynucleotidyltransferase-mediated dUTP nick end labeling (TUNEL) and by active caspase 3 immunohistochemical staining. Results: The incidence of NEC in the HHHTF + LPS group was higher than that in the BF group (65% vs 0%, P b .05). With administration of HB-EGF, the incidence of NEC significantly decreased to 23.8% ( P b .05). The median TUNEL and active caspase 3 scores in the HHHTF + LPS group were higher than those in the BF group (1.9 vs 0.9 and 1.75 vs 0.6, respectively, P b .05). The median TUNEL and active caspase 3 scores were significantly decreased in the HHHTF + LPS + HB-EGF group compared with the HHHTF + LPS group (1.24 vs 1.9 and 1.0 vs 1.75, respectively, P b .05). Conclusion: HB-EGF reduces the incidence of NEC in a neonatal rat model in part by decreasing apoptosis. These results support the use of HB-EGF–based clinical regimens for the treatment of NEC. D 2006 Elsevier Inc. All rights reserved.
Presented at the 57th Annual Meeting of the Section on Surgery of the American Academy of Pediatrics, Washington, DC, October 7-9, 2005. B This work was supported in part by GM 61193 from the National Institutes of Health (GEB). * Corresponding author. Department of Surgery, Children’s Hospital, Columbus, OH 43205, USA. Tel.: +1 614 722 3930; fax: +1 614 722 3903. 0022-3468/$ – see front matter D 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.jpedsurg.2005.12.020
Necrotizing enterocolitis (NEC) is the most common acquired intraabdominal emergency in infants and the most common surgical emergency in the neonatal intensive care unit. Despite 3 decades of research, NEC remains a major cause of death for neonates in the United States, causing up to 1000 infant deaths in this country each year [1,2]. Although several lines of evidence suggest that neonatal risk factors, including prematurity, asphyxia, intestinal
HB-EGF reduces intestinal apoptosis in neonatal rats with NEC ischemia, and formula feeding, all contribute to the occurrence of NEC, the pathogenesis of this disease remains unclear [1,2]. Recent studies have shown that apoptosis may account for the initial loss of cells in the apical villi before full development of the disease. Ford et al [3] observed abundant epithelial apoptosis in intestinal villi in histological specimens collected at the time of bowel resection in patients with NEC. Apoptotic enterocytes in the terminal ileum were also found in a rat model of NEC [4]. Moreover, by using terminal deoxynucleotidyltransferase-mediated dUTP nick end labeling (TUNEL) staining, Jilling et al [5] demonstrated that abundant apoptosis of the intestinal epithelium precedes widespread tissue damage in an experimental model of NEC. Taken together, these observations suggest that apoptosis may be associated with the pathogenesis of NEC. Our previous studies had shown that administration of heparin-binding epidermal growth factor–like growth factor (HB-EGF), a member of the EGF family that was initially identified in the conditioned medium of cultured human macrophages [6], decreases the incidence and severity of NEC in a neonatal rat model, with simultaneous preservation of gut barrier integrity [7]. We also demonstrated that HB-EGF down-regulated apoptosis in intestinal epithelial cells exposed to proinflammatory cytokines in vitro [8]. The aim of the current study, therefore, was to investigate whether HB-EGF–mediated protection against experimental NEC is associated with decreased apoptosis.
743
Briefly, pregnant time-dated Sprague-Dawley rats (Harlan Sprague-Dawley, Indianapolis, Ind) were delivered by cesarean section under CO2 anesthesia on day 21.5 of gestation. Newborn rats (average weight 5.2 g) were placed in a neonatal incubator for temperature control. Fifty-one neonatal rats were randomized into 1 of 3 groups. Group 1 (breast-feeding [BF], n = 10) consisted of neonatal rats that were left with their mother to be breast-fed. Group 2 (hypoxia, hypothermia, hypertonic feedings plus lipopolysaccharide [HHHTF + LPS], n = 20) consisted of neonatal rats that were stressed by exposure to 100% nitrogen for 1 minute followed by 48C for 10 minutes twice daily and exposed to intragastric LPS (2 mg/kg) 8 hours after birth. These neonatal rats were gavaged with formula containing 15 g Similac 60/40 (Ross Pediatrics, Columbus, OH, USA) in 75 mL of Esbilac (Pet-Ag, New Hampshire, Ill), a diet that provided 840 kJ/(kgd d) or 200 kcal/kg/d. Feeds were started at 0.1 mL every 4 hours and advanced as tolerated up to a maximum of 0.4 mL per feeding by the fourth day of life. Group 3 [HHHTF + LPS + HB-EGF, n = 21] consisted of neonatal rats that were stressed and formula-fed as in group 2 but treated with HB-EGF (600 lg/kg) added to the formula every 4 hours and administered via the orogastric feeding tube. Animals were monitored until 96 hours of life for clinical signs of NEC. Pups were euthanized by cervical dislocation upon the development of any clinical signs of NEC. After 96 hours, all surviving animals were killed.
1.2. NEC evaluation
1. Materials and methods 1.1. Experimental design and animal model The experimental protocol was performed in accordance with the guidelines for the ethical treatment of experimental animals and approved by the Institutional Animal Care and Use Committee of the Columbus Children’s Hospital (protocol 04203AR). NEC was induced using a modification of the neonatal rat model of NEC initially described by Barlow et al [9] and further developed on our laboratory [7].
Immediately after killing, the gastrointestinal tract was carefully removed and visually evaluated for typical signs of NEC. Three pieces each of duodenum, jejunum, ileum, and colon in every animal were fixed in 10% formalin for 24 hours, paraffin-embedded, sectioned at 5 lm, and stained with H&E for histological evaluation of the presence and/or degree of NEC using a standard histological scoring system [7,10]. Histological changes in the intestines were graded as follows: grade 0, normal, no damage; Grade 1, epithelial cell lifting or separation; grade 2, sloughing of epithelial
Fig. 1 TUNEL staining score in intestines from neonatal rats. A, Grade 0, solitary positive nuclei; B, grade 1, clusters of positive nuclei in villous tips; C, grade 2, positive nuclei throughout the villi; D, grade 3, apoptotic nuclei penetrating into the crypts (original magnification 100).
744
J. Feng et al.
Table 1
Effect of HB-EGF on the incidence of NEC and intestinal epithelial cell apoptosis in newborn rats
Group
Incidence of NEC (%)
Histological injury score (median)
TUNEL score (median)
Caspase 3 score (median)
BF HHHTF + LPS HHHTF + LPS + HB_EGF
0/10 (0) 13/20 (65)* 5/21 (23.8)**
0 2* 1.0**
0.9 1.9* 1.24**
0.6 1.75* 1.0**
* P b .01 compared with BF. ** P b .05 compared with HHHTF + LPS.
cells to mid villus level; grade 3, necrosis of the entire villus; grade 4, transmural necrosis. All tissues were graded blindly by 2 independent observers. Tissues with histological scores of 2 or higher were designated as positive for NEC.
1.3. TUNEL staining Apoptotic cells were detected by utilizing the ApopTag plus peroxidase in situ apoptosis detection kit (Chemicon, Temecula, Calif). In short, formalin-fixed, paraffin-embedded intestinal tissue sections were deparaffinized and then rehydrated in 100%, 95%, and 70% graded ethanol washes. Protein was digested with proteinase K and endogenous peroxidase activity quenched with 3% hydrogen peroxide. Slides were then incubated with a reaction mixture containing terminal deoxynucleotidyl transferase and its substrate digoxigenin-dUTP at 378C for 1 h. The reaction was terminated and an anti–digoxigenin-peroxidase antibody was applied for 30 minutes. Color was developed by diaminobenzidine for 3 to 6 minutes. Slides were then counterstained with methyl green, washed in 100% butanol, dehydrated with xylene, and mounted. Sections were examined microscopically and assigned a score in a range of 0 to 3 according to an established scoring system [5]. In this system, score 0 was assigned to specimens with only solitary positive nuclei: score 1 was assigned to specimens with clusters of positive nuclei in the villous tips; score 2 was assigned to specimens with apoptotic nuclei throughout the villi with
sparing of the crypts; and score 3 was assigned to specimens with apoptosis penetrating into the crypts (Fig. 1).
1.4. Active caspase 3 staining Active caspase 3 was analyzed by immunohistochemical staining. Briefly, after deparaffinization, rehydration, and incubation in hydrogen peroxide, slides were immersed in a pressure cooker at 1208C for 10 minutes for antigen retrieval. Sections were then blocked with mouse serum and incubated with 10 lg/mL rabbit polyclonal antibody to active caspase 3 (Abcam, Cambridge, Mass) at 48C overnight, followed by UltraTek antipolyvalent and UltraTek HRP (Scytek Laboratories, Logan, Utah). Color was developed by adding a mixture of AEC Chromogen and AEC substrate (Scytek Laboratories) and counterstained with hematoxylin for 1 minute. Negative control sections were treated similarly with the exception that they were incubated with phosphate-buffered saline instead of primary antibody. No immunostaining was observed in these negative control sections. Slides were then evaluated and scored as described above for TUNEL staining.
1.5. Statistical analyses The incidence of NEC was compared between groups using the v 2 test. The severity of NEC, TUNEL, and caspase 3 staining scores were analyzed by the MannWhitney U test. The relationship between TUNEL and
Fig. 2 Summary of TUNEL staining grade distribution in rat pups. Shown are the grades of TUNEL staining for rat pups subjected to BF, HHHTF + LPS, and HHHTF + LPS + HB-EGF.
HB-EGF reduces intestinal apoptosis in neonatal rats with NEC
745
3. Discussion
Fig. 3 Representative active caspase 3 immunostaining in intestines from neonatal rats. A, BF group, with no positive nuclei (grade 0); B, HHHTF + LPS group, with positive nuclei throughout the villi (grade 3); C, HHHTF + LPS + HB-EGF group, with positive nuclei in villous tips only (grade 1) (original magnification 100.
caspase 3 staining was examined by correlation analysis. P b .05 was considered significant.
2. Results Histological analysis revealed that the HHHTF + LPS group had a significantly higher incidence of NEC than the BF group (65% vs 0%, P b .01), whereas the incidence of NEC decreased to 23.8% in animals exposed to HHHTF + LPS with administration of HB-EGF ( P b .05, Table 1). The degree of intestinal damage was also significantly reduced in animals exposed to HHHTF + LPS + HB-EGF compared with animals exposed to HHHTF + LPS only (grade 1.0 vs 2.0, P b .05, Table 1). HHHTF + LPS pups showed a significant increase in the frequency of nuclei exhibiting DNA fragmentation in the epithelium compared with BF control pups as demonstrated by TUNEL staining (Fig. 2). Using the previously described scoring system, the TUNEL score in HHHTF + LPS animals was significantly higher than that in control animals (1.9 vs 0.9, P b .01, Table 1). The TUNEL score in the HBEGF treatment group decreased significantly to 1.24 ( P b .05, Table 1). Because TUNEL staining has been shown to occasionally produce a false-positive reaction, active caspase 3, a specific marker for apoptotic cells [11], was also used to detect and localize apoptotic changes in intestinal epithelial cells in all experimental groups. The HHHTF + LPS group showed a significant increase in active caspase 3 staining compared with BF control animals (1.75 vs 0.6, P b .01, Table 1). After administration of HB-EGF, the active caspase 3 staining score decreased to 1.0 ( P b .05, Table 1). Representative active caspase 3 immunostaining in intestines from the BF, HHHTF + LPS, and HHHTF + LPS + HB-EGF groups are shown in Fig. 3. Correlation analysis showed that the correlation coefficient between TUNEL and active caspase 3 staining was 0.754, indicating a close relationship between these 2 staining methods.
Our results with both TUNEL and activated caspase 3 staining showed that intestinal epithelial cell apoptosis in newborn rat pups with experimental NEC was increased compared with BF control rat pups. These results are in agreement with previous studies showing that alterations in apoptosis may predispose the intestine to the development of NEC [3-5]. Previous studies from our laboratory demonstrated that HB-EGF could protect enterocytes from proinflammatory cytokine-induced apoptosis in vitro [8]. This is the first report demonstrating that enteral administration of HB-EGF can decrease apoptosis in a rat model of experimental NEC in vivo. These findings are in agreement with a recently reported study showing that EGF decreases apoptosis in a model of asphyxia and cold-induced experimental NEC [11]. Increased expression of inducible nitric oxide synthase (iNOS) and nitric oxide (NO) have been demonstrated in experimental NEC [4,12]. Increased apoptosis in the apical villi of infants with NEC correlated with the degree of nitrotyrosine immunostaining, indicative of increased nitrosative stress in the tissues [3]. Moreover, peroxynitrite (ONOO ) was found to induce apoptosis in a rat intestinal crypt enterocyte cell line (IEC-6 cells) in both a time- and dose-dependent manner [13]. In addition, procaspase 3, a proapoptotic factor, was cleaved and activated in IEC-6 cells after peroxynitrite exposure [14]. On the other hand, administration of aminoguanidine, a selective iNOS inhibitor, decreased the incidence of experimental NEC in rats [12]. Taken together, these findings support the possibility that NO promotes enterocyte apoptosis in intestinal tissue afflicted with NEC. We have shown that HB-EGF can down-regulate the production of NO, an important mediator of cell injury, in intestinal epithelial cells after stimulation with interleukin 1b (IL-1b) and interferon c (IFN-c) in vitro [15,16], and in intestine after exposure to ischemia-reperfusion injury in vivo [17]. We also showed that iNOS messenger RNA and protein levels in intestinal epithelial cells were reduced by the addition of HB-EGF [15]. Because iNOS induction by IL-1 and IFN-c in intestinal epithelial cells has been shown to be dependent upon NF-jB activation, we also investigated whether HB-EGF could block NF-jB activation in cytokine-stimulated intestinal epithelial cells. We showed that HB-EGF not only blocks NF-jB activation but also suppresses its transcriptional activity in intestinal epithelial cells after cytokine stimulation [16]. The ability of HB-EGF to inhibit NF-jB activation was because of its ability to prevent phosphorylation and degradation of IjBa and activation of IKB kinase (IKK) [16]. The ability of HB-EGF to suppress apoptosis in other epithelial cell types has been previously investigated. HBEGF decreases apoptosis in Caco cells exposed to IL-1/ TNF-a/IFN-c [18]. Moreover, inhibition of endogenous HB-EGF expression can increase apoptosis in stromal cells
746 stimulated with TNF-a [19]. In addition, increased levels of endogenous HB-EGF can down-regulate apoptosis in renal epithelial cells during renal injury [20]. The protective role of HB-EGF in acute oxidant-induced apoptosis has been demonstrated in proximal tubule cells [21]. HB-EGF also prevents apoptosis in mouse muscle myoblast cells exposed to hypoxia [22]. We have previously demonstrated that HB-EGF decreases the generation of reactive oxygen species (ROS), known mediators of ischemia-reperfusion– induced apoptosis, in intestinal epithelial cells after intestinal ischemia-reperfusion injury [23]. In summary, our studies indicate that administration of HB-EGF can suppress apoptosis in a neonatal rat model of NEC. These results support the use of HB-EGF as a promising therapeutic treatment of NEC in the future.
J. Feng et al.
[16]
[17]
[18]
[19]
[20]
References [1] Henry MC, Moss RL. Current issues in the management of necrotizing enterocolitis. Semin Perinatol 2004;28:221 - 33. [2] Feng J, El-Assal ON, Besner GE. Heparin-binding EGF-like growth factor (HB-EGF) and necrotizing enterocolitis. Semin Pediatr Surg 2005;14:167 - 74. [3] Ford H, Watkins S, Reblock K, et al. The role of inflammatory cytokines and nitric oxide in the pathogenesis of necrotizing enterocolitis. J Pediatr Surg 1997;32:275 - 82. [4] Nadler EP, Dickinson E, Knisely A, et al. Expression of inducible nitric oxide synthase and interleukin-12 in experimental necrotizing enterocolitis. J Surg Res 2000;92:71 - 7. [5] Jilling T, Lu J, Jackson M, et al. Intestinal epithelial apoptosis initiates gross bowel necrosis in an experimental rat model of neonatal necrotizing enterocolitis. Pediatr Res 2004;55:622 - 9. [6] Besner GE, Higashiyama S, Klagsbrun M. Isolation and characterization of a macrophage-derived heparin-binding growth factor. Cell Regul 1990;1:811 - 9. [7] Feng J, El-Assal ON, Besner GE. Heparin-binding EGF-like growth factor decreases the incidence of necrotizing enterocolitis in neonatal rats. J Pediatr Surg [in press]. [8] Michalsky MP, Kuhn A, Mehta V, et al. Heparin-binding EGF-like growth factor decreases apoptosis in intestinal epithelial cells in vitro. J Pediatr Surg 2001;36:1130 - 5. [9] Barlow B, Santulli TV, Heird WC, et al. An experimental study of acute neonatal enterocolitis—the important of breast milk. J Pediatr Surg 1974;9:587 - 95. [10] Caplan MS, Hedlund E, Adler L, et al. Role of asphyxia and feeding in a neonatal rat model of necrotizing enterocolitis. Pediatr Pathol 1994;14:1017 - 28. [11] Clark JA, Lane RH, Maclennan NK, et al. Epidermal growth factor reduces intestinal apoptosis in an experimental model of necrotizing enterocolitis. Am J Physiol Gastrointest Liver Physiol 2005;288:G755 - 62. [12] Ciftci I, Dilsiz A, Aktan TM, et al. Effects of nitric oxide synthase inhibition on intestinal damage in rats with experimental necrotizing enterocolitis. Eur J Pediatr Surg 2004;14:398 - 403. [13] Potoka DA, Upperman JS, Nadler EP, et al. NF-kappaB inhibition enhances peroxynitrite-induced enterocyte apoptosis. J Surg Res 2002;106:7 - 14. [14] Upperman JS, Nadle EP, Hoffman RA, et al. Peroxynitrite induces enterocyte apoptosis in vitro by modulating mitochondrial membrane potential and PARS cleavage. Surg Forum 1999;81 - 2. [15] Lara-Marquez ML, Mehta V, Michalsky MP, et al. Heparin-binding EGF-like growth factor down regulates proinflammatory cytokine-
[21]
[22]
[23]
induced nitric oxide and inducible nitric oxide synthase production in intestinal epithelial cells. Nitric Oxide 2002;6:142 - 52. Mehta VB, Besner GE. Inhibition of NF-kappa B activation and its target genes by heparin-binding epidermal growth factor-like growth factor. J Immunol 2003;171:6014 - 22. Xia G, Lara-Marquez M, Luquette MH. Heparin-binding EGF-like growth factor decreases inducible nitric oxide synthase and nitric oxide production after intestinal ischemia/reperfusion injury. Antioxid Redox Signal 2001;3:919 - 30. Takenobu H, Yamazaki A, Hirata M, et al. The stress- and inflammatory cytokine-induced ectodomain shedding of heparinbinding epidermal growth factor-like growth factor is mediated by p38 MAPK, distinct from the 12-O-tetradecanoylphorbol-13-acetateand lysophosphatidic acid-induced signaling cascades. J Biol Chem 2003;278:17255 - 62. Sakuda S, Tamura S, Yamada A, et al. NF-kappaB activation in nonparenchymal liver cells after partial hepatectomy in rats: possible involvement in expression of heparin-binding epidermal growth factor-like growth factor. J Hepatol 2002;36:527 - 33. Nguyen HT, Bride SH, Badawy AB, et al. Heparin-binding EGF-like growth factor is up-regulated in the obstructed kidney in a cell- and region-specific manner and acts to inhibit apoptosis. Am J Pathol 2000;156:889 - 98. Bryson JM, Coy PE, Gottlob K, et al. Increased hexokinase activity, of either ectopic or endogenous origin, protects renal epithelial cells against acute oxidant-induced cell death. J Biol Chem 2002;277: 11392 - 400. Horikawa M, Higashiyama S, Nomura S, et al. Upregulation of endogenous heparin-binding EGF-like growth factor and its role as a survival factor in skeletal myotubes. FEBS Lett 1999;459: 100 - 4. Kuhn MA, Xia G, Mehta VB, et al. Heparin-binding EGF-like growth factor (HB-EGF) decreases oxygen free radical production in vitro and in vivo. Antioxid Redox Signal 2002;4:639 - 46.
Discussion Daniel Teitelbaum, MD (Ann Arbor, MI): The results were quite interesting. When you differentiated the location of apoptosis between the villi and the crypt, is the significance of apoptosis occurring in both of these locations the same? In other words, apoptosis that may be occurring in the crypt may be a remodeling of the architecture to facilitate regrowth, whereas apoptosis in the villi may be an injury-related mode? Gail Besner, MD (Columbus, OH): Your question is a really interesting one. We have not investigated whether there are mechanistic differences between apoptosis occurring in the villi versus apoptosis occurring in the crypts, but I do believe that both are stress-related. The rationale behind the grading system that we used is that in the NEC model there is denuding of the villi, so in addition to looking at the number of apoptotic cells present we looked at the depth at which these apoptotic cells are present, with deeper levels of penetration indicative of more severe injury-related apoptosis. But we haven’t specifically looked at differences in mechanisms between
HB-EGF reduces intestinal apoptosis in neonatal rats with NEC apoptosis in the villi versus the crypts. It’s intriguing and we’ll have to think more about that. David Hackam, MD (Pittsburgh, PA):Very beautifully presented, interesting data. You began by stating the hypothesis that a link exists between alterations in apoptosis of intestinal epithelial cells and bacterial translocation. Could you please provide us with some further insights into this link? For instance, I’m wondering whether you’ve measured bacterial translocation in your model, and whether bacterial translocation is affected or reduced by HB-EGF treatment. And secondly, could you share with us any in vitro data that you may have performed in your intestinal cell lines using HB-EGF to see if the same apoptotic pathways are regulated in vitro? This of course would allow you to begin to elucidate the mechanisms underlying this very interesting observation. Gail Besner, MD (Columbus, OH): We have used an intestinal ischemia-reperfusion model with superior mesenteric artery occlusion to show that HB-EGF can
747 suppress bacterial translocation. We have also utilized an intestinal epithelial cell culture model system using dual chamber wells in vitro to show that HB-EGF decreases bacterial translocation through intestinal epithelial cell monolayers. Additionally, Dr Feng has used the rat NEC model to study intestinal permeability, and showed that HB-EGF protects against the increased intestinal permeability that occurs after stress.
Martin Keller, MD (St Louis, MO): Did this model produce NEC in just the small intestine and if not, was there a difference in what you saw in the large versus the small intestine? Jeixiong Feng, MD: Yes, injury occurs in the colon as well as in the small bowel. Martin Keller, MD (St Louis, MO): Was there any difference? Jeixiong Feng, MD: The degree of injury is greater in the small bowel than in the colon.
www.elsevier.com/locate/jpedsurg
Heparin-binding epidermal growth factor–like growth factor reduces intestinal apoptosis in neonatal rats with necrotizing enterocolitisB Jiexiong Feng, Osama N. El-Assal, Gail E. Besner* Department of Pediatric Surgery, Center for Cell and Vascular Biology, Children’s Research Institute, Columbus, OH 43205, USA and The Ohio State University College of Medicine and Public Health, Columbus, OH 43210, USA Index words: Necrotizing enterocolitis; HB-EGF; Apoptosis; TUNEL; Caspase 3
Abstract Purpose: We have previously demonstrated that enterally administered heparin-binding epidermal growth factor–like growth factor (HB-EGF) decreases the incidence and severity of necrotizing enterocolitis (NEC) in a neonatal rat model. Because apoptosis contributes to gut barrier failure in this model, the aim of this study was to investigate the effect of HB-EGF on apoptosis during the development of NEC. Methods: NEC was induced in neonatal rats by exposure to hypoxia, hypothermia, hypertonic formula feeding (HHHTF) plus enteral administration of lipopolysaccharide (LPS). Fifty-one neonatal rats were randomly divided into the following groups: (1) breast-fed (BF), (2) HHHTF + LPS, and (3) HHHTF + LPS with HB-EGF (600 lg/kg) added to the formula. NEC was evaluated using a standard histological scoring system. Apoptotic cells in intestinal tissues were detected by terminal deoxynucleotidyltransferase-mediated dUTP nick end labeling (TUNEL) and by active caspase 3 immunohistochemical staining. Results: The incidence of NEC in the HHHTF + LPS group was higher than that in the BF group (65% vs 0%, P b .05). With administration of HB-EGF, the incidence of NEC significantly decreased to 23.8% ( P b .05). The median TUNEL and active caspase 3 scores in the HHHTF + LPS group were higher than those in the BF group (1.9 vs 0.9 and 1.75 vs 0.6, respectively, P b .05). The median TUNEL and active caspase 3 scores were significantly decreased in the HHHTF + LPS + HB-EGF group compared with the HHHTF + LPS group (1.24 vs 1.9 and 1.0 vs 1.75, respectively, P b .05). Conclusion: HB-EGF reduces the incidence of NEC in a neonatal rat model in part by decreasing apoptosis. These results support the use of HB-EGF–based clinical regimens for the treatment of NEC. D 2006 Elsevier Inc. All rights reserved.
Presented at the 57th Annual Meeting of the Section on Surgery of the American Academy of Pediatrics, Washington, DC, October 7-9, 2005. B This work was supported in part by GM 61193 from the National Institutes of Health (GEB). * Corresponding author. Department of Surgery, Children’s Hospital, Columbus, OH 43205, USA. Tel.: +1 614 722 3930; fax: +1 614 722 3903. 0022-3468/$ – see front matter D 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.jpedsurg.2005.12.020
Necrotizing enterocolitis (NEC) is the most common acquired intraabdominal emergency in infants and the most common surgical emergency in the neonatal intensive care unit. Despite 3 decades of research, NEC remains a major cause of death for neonates in the United States, causing up to 1000 infant deaths in this country each year [1,2]. Although several lines of evidence suggest that neonatal risk factors, including prematurity, asphyxia, intestinal
HB-EGF reduces intestinal apoptosis in neonatal rats with NEC ischemia, and formula feeding, all contribute to the occurrence of NEC, the pathogenesis of this disease remains unclear [1,2]. Recent studies have shown that apoptosis may account for the initial loss of cells in the apical villi before full development of the disease. Ford et al [3] observed abundant epithelial apoptosis in intestinal villi in histological specimens collected at the time of bowel resection in patients with NEC. Apoptotic enterocytes in the terminal ileum were also found in a rat model of NEC [4]. Moreover, by using terminal deoxynucleotidyltransferase-mediated dUTP nick end labeling (TUNEL) staining, Jilling et al [5] demonstrated that abundant apoptosis of the intestinal epithelium precedes widespread tissue damage in an experimental model of NEC. Taken together, these observations suggest that apoptosis may be associated with the pathogenesis of NEC. Our previous studies had shown that administration of heparin-binding epidermal growth factor–like growth factor (HB-EGF), a member of the EGF family that was initially identified in the conditioned medium of cultured human macrophages [6], decreases the incidence and severity of NEC in a neonatal rat model, with simultaneous preservation of gut barrier integrity [7]. We also demonstrated that HB-EGF down-regulated apoptosis in intestinal epithelial cells exposed to proinflammatory cytokines in vitro [8]. The aim of the current study, therefore, was to investigate whether HB-EGF–mediated protection against experimental NEC is associated with decreased apoptosis.
743
Briefly, pregnant time-dated Sprague-Dawley rats (Harlan Sprague-Dawley, Indianapolis, Ind) were delivered by cesarean section under CO2 anesthesia on day 21.5 of gestation. Newborn rats (average weight 5.2 g) were placed in a neonatal incubator for temperature control. Fifty-one neonatal rats were randomized into 1 of 3 groups. Group 1 (breast-feeding [BF], n = 10) consisted of neonatal rats that were left with their mother to be breast-fed. Group 2 (hypoxia, hypothermia, hypertonic feedings plus lipopolysaccharide [HHHTF + LPS], n = 20) consisted of neonatal rats that were stressed by exposure to 100% nitrogen for 1 minute followed by 48C for 10 minutes twice daily and exposed to intragastric LPS (2 mg/kg) 8 hours after birth. These neonatal rats were gavaged with formula containing 15 g Similac 60/40 (Ross Pediatrics, Columbus, OH, USA) in 75 mL of Esbilac (Pet-Ag, New Hampshire, Ill), a diet that provided 840 kJ/(kgd d) or 200 kcal/kg/d. Feeds were started at 0.1 mL every 4 hours and advanced as tolerated up to a maximum of 0.4 mL per feeding by the fourth day of life. Group 3 [HHHTF + LPS + HB-EGF, n = 21] consisted of neonatal rats that were stressed and formula-fed as in group 2 but treated with HB-EGF (600 lg/kg) added to the formula every 4 hours and administered via the orogastric feeding tube. Animals were monitored until 96 hours of life for clinical signs of NEC. Pups were euthanized by cervical dislocation upon the development of any clinical signs of NEC. After 96 hours, all surviving animals were killed.
1.2. NEC evaluation
1. Materials and methods 1.1. Experimental design and animal model The experimental protocol was performed in accordance with the guidelines for the ethical treatment of experimental animals and approved by the Institutional Animal Care and Use Committee of the Columbus Children’s Hospital (protocol 04203AR). NEC was induced using a modification of the neonatal rat model of NEC initially described by Barlow et al [9] and further developed on our laboratory [7].
Immediately after killing, the gastrointestinal tract was carefully removed and visually evaluated for typical signs of NEC. Three pieces each of duodenum, jejunum, ileum, and colon in every animal were fixed in 10% formalin for 24 hours, paraffin-embedded, sectioned at 5 lm, and stained with H&E for histological evaluation of the presence and/or degree of NEC using a standard histological scoring system [7,10]. Histological changes in the intestines were graded as follows: grade 0, normal, no damage; Grade 1, epithelial cell lifting or separation; grade 2, sloughing of epithelial
Fig. 1 TUNEL staining score in intestines from neonatal rats. A, Grade 0, solitary positive nuclei; B, grade 1, clusters of positive nuclei in villous tips; C, grade 2, positive nuclei throughout the villi; D, grade 3, apoptotic nuclei penetrating into the crypts (original magnification 100).
744
J. Feng et al.
Table 1
Effect of HB-EGF on the incidence of NEC and intestinal epithelial cell apoptosis in newborn rats
Group
Incidence of NEC (%)
Histological injury score (median)
TUNEL score (median)
Caspase 3 score (median)
BF HHHTF + LPS HHHTF + LPS + HB_EGF
0/10 (0) 13/20 (65)* 5/21 (23.8)**
0 2* 1.0**
0.9 1.9* 1.24**
0.6 1.75* 1.0**
* P b .01 compared with BF. ** P b .05 compared with HHHTF + LPS.
cells to mid villus level; grade 3, necrosis of the entire villus; grade 4, transmural necrosis. All tissues were graded blindly by 2 independent observers. Tissues with histological scores of 2 or higher were designated as positive for NEC.
1.3. TUNEL staining Apoptotic cells were detected by utilizing the ApopTag plus peroxidase in situ apoptosis detection kit (Chemicon, Temecula, Calif). In short, formalin-fixed, paraffin-embedded intestinal tissue sections were deparaffinized and then rehydrated in 100%, 95%, and 70% graded ethanol washes. Protein was digested with proteinase K and endogenous peroxidase activity quenched with 3% hydrogen peroxide. Slides were then incubated with a reaction mixture containing terminal deoxynucleotidyl transferase and its substrate digoxigenin-dUTP at 378C for 1 h. The reaction was terminated and an anti–digoxigenin-peroxidase antibody was applied for 30 minutes. Color was developed by diaminobenzidine for 3 to 6 minutes. Slides were then counterstained with methyl green, washed in 100% butanol, dehydrated with xylene, and mounted. Sections were examined microscopically and assigned a score in a range of 0 to 3 according to an established scoring system [5]. In this system, score 0 was assigned to specimens with only solitary positive nuclei: score 1 was assigned to specimens with clusters of positive nuclei in the villous tips; score 2 was assigned to specimens with apoptotic nuclei throughout the villi with
sparing of the crypts; and score 3 was assigned to specimens with apoptosis penetrating into the crypts (Fig. 1).
1.4. Active caspase 3 staining Active caspase 3 was analyzed by immunohistochemical staining. Briefly, after deparaffinization, rehydration, and incubation in hydrogen peroxide, slides were immersed in a pressure cooker at 1208C for 10 minutes for antigen retrieval. Sections were then blocked with mouse serum and incubated with 10 lg/mL rabbit polyclonal antibody to active caspase 3 (Abcam, Cambridge, Mass) at 48C overnight, followed by UltraTek antipolyvalent and UltraTek HRP (Scytek Laboratories, Logan, Utah). Color was developed by adding a mixture of AEC Chromogen and AEC substrate (Scytek Laboratories) and counterstained with hematoxylin for 1 minute. Negative control sections were treated similarly with the exception that they were incubated with phosphate-buffered saline instead of primary antibody. No immunostaining was observed in these negative control sections. Slides were then evaluated and scored as described above for TUNEL staining.
1.5. Statistical analyses The incidence of NEC was compared between groups using the v 2 test. The severity of NEC, TUNEL, and caspase 3 staining scores were analyzed by the MannWhitney U test. The relationship between TUNEL and
Fig. 2 Summary of TUNEL staining grade distribution in rat pups. Shown are the grades of TUNEL staining for rat pups subjected to BF, HHHTF + LPS, and HHHTF + LPS + HB-EGF.
HB-EGF reduces intestinal apoptosis in neonatal rats with NEC
745
3. Discussion
Fig. 3 Representative active caspase 3 immunostaining in intestines from neonatal rats. A, BF group, with no positive nuclei (grade 0); B, HHHTF + LPS group, with positive nuclei throughout the villi (grade 3); C, HHHTF + LPS + HB-EGF group, with positive nuclei in villous tips only (grade 1) (original magnification 100.
caspase 3 staining was examined by correlation analysis. P b .05 was considered significant.
2. Results Histological analysis revealed that the HHHTF + LPS group had a significantly higher incidence of NEC than the BF group (65% vs 0%, P b .01), whereas the incidence of NEC decreased to 23.8% in animals exposed to HHHTF + LPS with administration of HB-EGF ( P b .05, Table 1). The degree of intestinal damage was also significantly reduced in animals exposed to HHHTF + LPS + HB-EGF compared with animals exposed to HHHTF + LPS only (grade 1.0 vs 2.0, P b .05, Table 1). HHHTF + LPS pups showed a significant increase in the frequency of nuclei exhibiting DNA fragmentation in the epithelium compared with BF control pups as demonstrated by TUNEL staining (Fig. 2). Using the previously described scoring system, the TUNEL score in HHHTF + LPS animals was significantly higher than that in control animals (1.9 vs 0.9, P b .01, Table 1). The TUNEL score in the HBEGF treatment group decreased significantly to 1.24 ( P b .05, Table 1). Because TUNEL staining has been shown to occasionally produce a false-positive reaction, active caspase 3, a specific marker for apoptotic cells [11], was also used to detect and localize apoptotic changes in intestinal epithelial cells in all experimental groups. The HHHTF + LPS group showed a significant increase in active caspase 3 staining compared with BF control animals (1.75 vs 0.6, P b .01, Table 1). After administration of HB-EGF, the active caspase 3 staining score decreased to 1.0 ( P b .05, Table 1). Representative active caspase 3 immunostaining in intestines from the BF, HHHTF + LPS, and HHHTF + LPS + HB-EGF groups are shown in Fig. 3. Correlation analysis showed that the correlation coefficient between TUNEL and active caspase 3 staining was 0.754, indicating a close relationship between these 2 staining methods.
Our results with both TUNEL and activated caspase 3 staining showed that intestinal epithelial cell apoptosis in newborn rat pups with experimental NEC was increased compared with BF control rat pups. These results are in agreement with previous studies showing that alterations in apoptosis may predispose the intestine to the development of NEC [3-5]. Previous studies from our laboratory demonstrated that HB-EGF could protect enterocytes from proinflammatory cytokine-induced apoptosis in vitro [8]. This is the first report demonstrating that enteral administration of HB-EGF can decrease apoptosis in a rat model of experimental NEC in vivo. These findings are in agreement with a recently reported study showing that EGF decreases apoptosis in a model of asphyxia and cold-induced experimental NEC [11]. Increased expression of inducible nitric oxide synthase (iNOS) and nitric oxide (NO) have been demonstrated in experimental NEC [4,12]. Increased apoptosis in the apical villi of infants with NEC correlated with the degree of nitrotyrosine immunostaining, indicative of increased nitrosative stress in the tissues [3]. Moreover, peroxynitrite (ONOO ) was found to induce apoptosis in a rat intestinal crypt enterocyte cell line (IEC-6 cells) in both a time- and dose-dependent manner [13]. In addition, procaspase 3, a proapoptotic factor, was cleaved and activated in IEC-6 cells after peroxynitrite exposure [14]. On the other hand, administration of aminoguanidine, a selective iNOS inhibitor, decreased the incidence of experimental NEC in rats [12]. Taken together, these findings support the possibility that NO promotes enterocyte apoptosis in intestinal tissue afflicted with NEC. We have shown that HB-EGF can down-regulate the production of NO, an important mediator of cell injury, in intestinal epithelial cells after stimulation with interleukin 1b (IL-1b) and interferon c (IFN-c) in vitro [15,16], and in intestine after exposure to ischemia-reperfusion injury in vivo [17]. We also showed that iNOS messenger RNA and protein levels in intestinal epithelial cells were reduced by the addition of HB-EGF [15]. Because iNOS induction by IL-1 and IFN-c in intestinal epithelial cells has been shown to be dependent upon NF-jB activation, we also investigated whether HB-EGF could block NF-jB activation in cytokine-stimulated intestinal epithelial cells. We showed that HB-EGF not only blocks NF-jB activation but also suppresses its transcriptional activity in intestinal epithelial cells after cytokine stimulation [16]. The ability of HB-EGF to inhibit NF-jB activation was because of its ability to prevent phosphorylation and degradation of IjBa and activation of IKB kinase (IKK) [16]. The ability of HB-EGF to suppress apoptosis in other epithelial cell types has been previously investigated. HBEGF decreases apoptosis in Caco cells exposed to IL-1/ TNF-a/IFN-c [18]. Moreover, inhibition of endogenous HB-EGF expression can increase apoptosis in stromal cells
746 stimulated with TNF-a [19]. In addition, increased levels of endogenous HB-EGF can down-regulate apoptosis in renal epithelial cells during renal injury [20]. The protective role of HB-EGF in acute oxidant-induced apoptosis has been demonstrated in proximal tubule cells [21]. HB-EGF also prevents apoptosis in mouse muscle myoblast cells exposed to hypoxia [22]. We have previously demonstrated that HB-EGF decreases the generation of reactive oxygen species (ROS), known mediators of ischemia-reperfusion– induced apoptosis, in intestinal epithelial cells after intestinal ischemia-reperfusion injury [23]. In summary, our studies indicate that administration of HB-EGF can suppress apoptosis in a neonatal rat model of NEC. These results support the use of HB-EGF as a promising therapeutic treatment of NEC in the future.
J. Feng et al.
[16]
[17]
[18]
[19]
[20]
References [1] Henry MC, Moss RL. Current issues in the management of necrotizing enterocolitis. Semin Perinatol 2004;28:221 - 33. [2] Feng J, El-Assal ON, Besner GE. Heparin-binding EGF-like growth factor (HB-EGF) and necrotizing enterocolitis. Semin Pediatr Surg 2005;14:167 - 74. [3] Ford H, Watkins S, Reblock K, et al. The role of inflammatory cytokines and nitric oxide in the pathogenesis of necrotizing enterocolitis. J Pediatr Surg 1997;32:275 - 82. [4] Nadler EP, Dickinson E, Knisely A, et al. Expression of inducible nitric oxide synthase and interleukin-12 in experimental necrotizing enterocolitis. J Surg Res 2000;92:71 - 7. [5] Jilling T, Lu J, Jackson M, et al. Intestinal epithelial apoptosis initiates gross bowel necrosis in an experimental rat model of neonatal necrotizing enterocolitis. Pediatr Res 2004;55:622 - 9. [6] Besner GE, Higashiyama S, Klagsbrun M. Isolation and characterization of a macrophage-derived heparin-binding growth factor. Cell Regul 1990;1:811 - 9. [7] Feng J, El-Assal ON, Besner GE. Heparin-binding EGF-like growth factor decreases the incidence of necrotizing enterocolitis in neonatal rats. J Pediatr Surg [in press]. [8] Michalsky MP, Kuhn A, Mehta V, et al. Heparin-binding EGF-like growth factor decreases apoptosis in intestinal epithelial cells in vitro. J Pediatr Surg 2001;36:1130 - 5. [9] Barlow B, Santulli TV, Heird WC, et al. An experimental study of acute neonatal enterocolitis—the important of breast milk. J Pediatr Surg 1974;9:587 - 95. [10] Caplan MS, Hedlund E, Adler L, et al. Role of asphyxia and feeding in a neonatal rat model of necrotizing enterocolitis. Pediatr Pathol 1994;14:1017 - 28. [11] Clark JA, Lane RH, Maclennan NK, et al. Epidermal growth factor reduces intestinal apoptosis in an experimental model of necrotizing enterocolitis. Am J Physiol Gastrointest Liver Physiol 2005;288:G755 - 62. [12] Ciftci I, Dilsiz A, Aktan TM, et al. Effects of nitric oxide synthase inhibition on intestinal damage in rats with experimental necrotizing enterocolitis. Eur J Pediatr Surg 2004;14:398 - 403. [13] Potoka DA, Upperman JS, Nadler EP, et al. NF-kappaB inhibition enhances peroxynitrite-induced enterocyte apoptosis. J Surg Res 2002;106:7 - 14. [14] Upperman JS, Nadle EP, Hoffman RA, et al. Peroxynitrite induces enterocyte apoptosis in vitro by modulating mitochondrial membrane potential and PARS cleavage. Surg Forum 1999;81 - 2. [15] Lara-Marquez ML, Mehta V, Michalsky MP, et al. Heparin-binding EGF-like growth factor down regulates proinflammatory cytokine-
[21]
[22]
[23]
induced nitric oxide and inducible nitric oxide synthase production in intestinal epithelial cells. Nitric Oxide 2002;6:142 - 52. Mehta VB, Besner GE. Inhibition of NF-kappa B activation and its target genes by heparin-binding epidermal growth factor-like growth factor. J Immunol 2003;171:6014 - 22. Xia G, Lara-Marquez M, Luquette MH. Heparin-binding EGF-like growth factor decreases inducible nitric oxide synthase and nitric oxide production after intestinal ischemia/reperfusion injury. Antioxid Redox Signal 2001;3:919 - 30. Takenobu H, Yamazaki A, Hirata M, et al. The stress- and inflammatory cytokine-induced ectodomain shedding of heparinbinding epidermal growth factor-like growth factor is mediated by p38 MAPK, distinct from the 12-O-tetradecanoylphorbol-13-acetateand lysophosphatidic acid-induced signaling cascades. J Biol Chem 2003;278:17255 - 62. Sakuda S, Tamura S, Yamada A, et al. NF-kappaB activation in nonparenchymal liver cells after partial hepatectomy in rats: possible involvement in expression of heparin-binding epidermal growth factor-like growth factor. J Hepatol 2002;36:527 - 33. Nguyen HT, Bride SH, Badawy AB, et al. Heparin-binding EGF-like growth factor is up-regulated in the obstructed kidney in a cell- and region-specific manner and acts to inhibit apoptosis. Am J Pathol 2000;156:889 - 98. Bryson JM, Coy PE, Gottlob K, et al. Increased hexokinase activity, of either ectopic or endogenous origin, protects renal epithelial cells against acute oxidant-induced cell death. J Biol Chem 2002;277: 11392 - 400. Horikawa M, Higashiyama S, Nomura S, et al. Upregulation of endogenous heparin-binding EGF-like growth factor and its role as a survival factor in skeletal myotubes. FEBS Lett 1999;459: 100 - 4. Kuhn MA, Xia G, Mehta VB, et al. Heparin-binding EGF-like growth factor (HB-EGF) decreases oxygen free radical production in vitro and in vivo. Antioxid Redox Signal 2002;4:639 - 46.
Discussion Daniel Teitelbaum, MD (Ann Arbor, MI): The results were quite interesting. When you differentiated the location of apoptosis between the villi and the crypt, is the significance of apoptosis occurring in both of these locations the same? In other words, apoptosis that may be occurring in the crypt may be a remodeling of the architecture to facilitate regrowth, whereas apoptosis in the villi may be an injury-related mode? Gail Besner, MD (Columbus, OH): Your question is a really interesting one. We have not investigated whether there are mechanistic differences between apoptosis occurring in the villi versus apoptosis occurring in the crypts, but I do believe that both are stress-related. The rationale behind the grading system that we used is that in the NEC model there is denuding of the villi, so in addition to looking at the number of apoptotic cells present we looked at the depth at which these apoptotic cells are present, with deeper levels of penetration indicative of more severe injury-related apoptosis. But we haven’t specifically looked at differences in mechanisms between
HB-EGF reduces intestinal apoptosis in neonatal rats with NEC apoptosis in the villi versus the crypts. It’s intriguing and we’ll have to think more about that. David Hackam, MD (Pittsburgh, PA):Very beautifully presented, interesting data. You began by stating the hypothesis that a link exists between alterations in apoptosis of intestinal epithelial cells and bacterial translocation. Could you please provide us with some further insights into this link? For instance, I’m wondering whether you’ve measured bacterial translocation in your model, and whether bacterial translocation is affected or reduced by HB-EGF treatment. And secondly, could you share with us any in vitro data that you may have performed in your intestinal cell lines using HB-EGF to see if the same apoptotic pathways are regulated in vitro? This of course would allow you to begin to elucidate the mechanisms underlying this very interesting observation. Gail Besner, MD (Columbus, OH): We have used an intestinal ischemia-reperfusion model with superior mesenteric artery occlusion to show that HB-EGF can
747 suppress bacterial translocation. We have also utilized an intestinal epithelial cell culture model system using dual chamber wells in vitro to show that HB-EGF decreases bacterial translocation through intestinal epithelial cell monolayers. Additionally, Dr Feng has used the rat NEC model to study intestinal permeability, and showed that HB-EGF protects against the increased intestinal permeability that occurs after stress.
Martin Keller, MD (St Louis, MO): Did this model produce NEC in just the small intestine and if not, was there a difference in what you saw in the large versus the small intestine? Jeixiong Feng, MD: Yes, injury occurs in the colon as well as in the small bowel. Martin Keller, MD (St Louis, MO): Was there any difference? Jeixiong Feng, MD: The degree of injury is greater in the small bowel than in the colon.