Protective effects of recombinant human granulocyte colony ...

Pediatr Surg Int (2006) 22:719–723 DOI 10.1007/s00383-006-1728-2

O R I G I N A L A RT I C L E

Protective effects of recombinant human granulocyte colony stimulating factor in a rat model of necrotizing enterocolitis Fuat Emre Canpolat Æ Murat Yurdako¨k Æ ¨ zsoy Æ Rıfkı Hazırog˘lu Æ Aysße Korkmaz S¸ule O

Accepted: 3 July 2006 / Published online: 27 July 2006 Ó Springer-Verlag 2006

Abstract The role of cytokines and growth factors in the pathophysiology of neonatal necrotizing enterocolitis (NEC) is not defined clearly yet. The aim of this study was to determine the effects of recombinant human granulocyte colony stimulating factor (G-CSF) on intestinal cells in hypoxia-induced experimental NEC in rats. The study was experimented on Sprague Dawley rat pups. Group 1 (untreated, n = 7) rats were subjected to hypoxia–reoxygenation (H/O) and then were returned to standard conditions. Group 2 (G-CSF treated, n = 7) rats were subjected to H/O, and then were treated with G-CSF (100 lg/kg enterally) for 5 days. Group 3 was served as nonhypoxic controls. All animals were killed on day five, and histological examination was performed on intestinal samples. There were no histopathological changes in the control group. The histological findings in untreated rats were similar to those seen in neonatal NEC, with destruction of villi and crypts with extension to the muscularis layer. Intestinal damage was mild in group 2 and these histological changes were better than group 1, and worse than group 3. The mean of histologic grade of group 1 was 2.4 (range 2–3), and in the group 2, it was 1.2 (range 0–2). A difference was found when two

F. E. Canpolat (&) Æ M. Yurdako¨k Æ A. Korkmaz _ Neonatology Unit, Ihsan Dog˘ramacı Children’s Hospital, Hacettepe University, 06100 Ankara, Turkey e-mail: [email protected] ¨ zsoy Æ R. Hazırog˘lu S¸ule O Pathology Division, Faculty of Veterinary Medicine, Ankara University, Ankara, Turkey

groups were compared with each other (P < 0.05). In an experimental model of NEC, G-CSF could have a protective effect on intestinal damage. Keywords Necrotizing enterocolitis Æ Granulocyte colony stimulating factor

Introduction Necrotizing enterocolitis (NEC) is the most common gastrointestinal disorder seen in very low birth weight preterm infants with an incidence of 5–10% in neonatal intensive care units. NEC is characterized by various degrees of mucosal or transmural necrosis of the intestinal tissue and is a major cause of morbidity and mortality in preterm infants, occurring in up to 10% of all neonatal intensive care unit admissions [1]. The precise pathophysiological cause of NEC has not been identified, but the predisposing risk factors most consistently associated with NEC are prematurity and enteral feeding [2]. Contributing factors in addition to prematurity include intestinal ischemia, small bowel bacterial colonization, and immunologic immaturity [3, 4]. Very low birth weight infants develop definite NEC at a rate of 5–10% and mortality is around 20% [4]. Platelet activating factor (PAF), erythropoietin (EPO), and epidermal growth factor (EGF) are the most important cytokines studied in experimental animals and clinical trials in NEC [5–7]. The presence of EPO in human milk and demonstration of functional EPO receptors in fetal and postnatal small bowel intestine suggest that EPO and EPO-like growth factors and cytokines may play a role in the growth and development of the gastrointestinal tract [8, 9]. The

123

720

human milk also contains G-CSF, and in the gastrointestinal tract of fetus and neonate there are specific receptors of G-CSF [10]. Our aim was to determine the effects of G-CSF on the intestines of rats with hypoxiainduced NEC.

Pediatr Surg Int (2006) 22:719–723

villi; grade 1, villous tip necrosis with preservation of villous crypts; grade 2, necrosis of villous tips and crypts with loss of mucosal and submucosal architecture; grade 3, necrosis extending into the muscularis; and grade 4, transmural necrosis. Statistical analysis

Materials and methods Animals and experimental design This study was performed on 10-day-old Sprague Dawley rat pups (weight range 45–60 g, mean 50 g) that were maintained under standard conditions. All experimental protocols were performed according to the guidelines for the ethical treatment of experimental animals. Local institutional approval for researches was obtained before initiation of the study. Twenty rat pups were divided randomly into three groups. An animal model was used for hypoxia–reoxygenation (H/O) as described previously [11, 12]. Hypoxia was accomplished by placing the pups in an airtight plexiglas (Rohm & Haas, Philadelphia, PA) chamber (Vacunit; Echmann, England), which was perfused with 100% CO2 for 5 min. After hypoxia, the animals were reoxygenated for 5 min with 100% oxygen. After this procedure the animals were kept under standard conditions. Group 1 (untreated, n = 7) rats were served as untreated but administered saline as placebo enterally and intraperitoneally after H/O. Group 2 (G-CSF treated, n = 7) were subjected to H/O, and were treated with G-CSF (NeupogenÒ, 100 lg/kg enteral; 50 lg/ kg intraperitoneal Roche, Basel, Switzerland) for the next 5 days. Group 3 (control, n = 6) were served as nonhypoxic controls. All animals were killed on day five with cervical dislocation. The entire gastrointestinal tract was removed and fixed in 10% formaldehyde. A representative 1-cm-long specimen was taken for histologic study from duodenum, proximal, mid and distal small intestine, cecum and proximal and distal colon. All seven histologic sites were calculated for each case and a mean result statistically compared.

Data were entered and analyzed on a personal computer using SPSSÓ version 11.5. All values are expressed as medians. The data were evaluated using Mann–Whitney test; P values of less than 0.05 were considered significant.

Results Histopathologically, the control rats had no gastrointestinal tract changes (Fig. 1). The untreated animals (group 1) had areas of macroscopic gaseous distension of intestinal loops with localized hemorrhagic lesions that were not seen in the G-CSF treated rats (group 2). The lesions in the untreated rats were similar to neonatal NEC histopathologic findings, with destruction of villi and crypts and in some cases extension to the muscularis (Fig. 2). In contrast, in the rats treated with G-CSF, lesions were limited essentially to the very tips of the villi (Fig. 3). Intestinal injury median score was

Histopathology The specimens were dehydrated and embedded in paraffin wax and stained with H&E. Specimens were taken randomly from each specimen (seven part of one case) and graded microscopically in a blinded fashion on a scale from 0 to 4 according to grading system proposed by Clark et al. [13]. Grade 0 denoted intact

123

Fig. 1 Histologic appearance of normal rat intestine


721

Fig. 2 Destruction of villi and crypts

Fig. 4 Mononuclear cell infiltration

Discussion

Fig. 3 Limited destruction, better than Fig. 2

2.4 (range 2–3) in untreated group and 1.2 (range 0–2) in G-CSF treated group. There was also a mild submucosal mononuclear cellular infiltration in group 2 (Figs. 4, 5). Intestinal damage was significantly less in the G-CSF treated rats than in the untreated rats (P < 0.05; Table 1, Fig. 4).

Although numerous inflammatory mediators have been studied in the pathogenesis of tissue destruction in patients with NEC, their specific roles are still unclear. Intestinal ischemia and the effects of hypoxia have been studied by several animal models. In newborn puppies hypoxia results in mucosal damage by both light microscopy and ultrastructural analysis [14, 15]. In neonatal piglets hypoxia is associated with decreased mucosal blood flow and mucosal ischemic changes. In the pathology of NEC, intestinal ischemia, hypoxia and bacterial colonization have been demonstrated to be the major causative factors leading to a common inflammatory process and tissue damage mediated by several cytokines [16]. Interleukin-10, EGF and recombinant human EPO have been demonstrated to have protective effects on intestine in experimental NEC studies [1, 17, 18]. We used the method described by Okur et al. for H/O as a single insult. Our histopathologic results demonstrated that ischemic lesions had developed

123

722


3,5 3,0 2,5

Grade

2,0 1,5 1,0 ,5 0,0 -,5 N=

7

7

1,00

2,00 Groups

6

3,00

Fig. 5 Intestinal injury scores of groups

Table 1 Intestinal injury scores of groups Groups

Number

Histologic grade median (range)

Control Untreated G-CSF treated

6 7 7

0.3 (0–1) 2.4 (2–3)* 1.2 (0–2),

*P < 0.05 for untreated versus control

P < 0.05 for G-CSF treated versus control

P < 0.05 for G-CSF treated versus untreated

after the experiment. The control group (group 3) showed no evidence of any lesions. There was also a cellular infiltration in group 2. We concluded that this infiltration was an effect of G-CSF on submucosal intestinal tissue. Granulocyte colony stimulating factor (G-CSF) is an 18.8-kDa glycoprotein that participates in the regulation of neutrophil production after binding with its specific receptor (G-CSF-Rs) by inhibiting apoptosis of granulocytic progenitors and by supporting their clonal proliferation and differentiation [19, 20]. In addition to its presence on neutrophils and their progenitors, functional G-CSF-Rs are expressed on a variety of nonhematopoietic cells, including enterocytes of the fetus and neonate [10]. In utero, fetal enterocytes are universally exposed to intraluminal G-CSF, which is present in amniotic fluid and continually swallowed by the fetus [21]. Postnatally, infants fed with human milk are also exposed to enteric G-CSF, as human milk

123

contains significant concentrations of G-CSF [10, 22]. Enterocytes have G-CSF receptors and infants fed with human milk have reduced risk of NEC. Hypoxia appears to play an important role in the pathogenesis of NEC. Clinically, the association of perinatal and neonatal hypoxia with NEC has been well established [2–4]. All causative factors for NEC reveals an end of hypoxia and necrosis so that some growth factors may protect from hypoxia or support tissues’ defense mechanisms. Several studies showed that cytokines and growth factors play a role on ischemic and hypoxic changes of intestines in experimental models of NEC [17, 18]. Oxygen-derived free radicals have also been proposed as mediators of the mucosal injury in NEC. There is also a role for (1) oxidants, proteases, or both, produced by resident neutrophils within the mucosa; (2) proinflammatory and anti-inflammatory mediators including tumor necrosis factor interleukin-6, interleukin 1b, PAF, (3) locally produced prostaglandins and leukotrienes; or (4) trophic factors such as EGF in the initiation of mucosal injury [16]. Measurement of these factors in hopes of deriving diagnostic or prognostic information is still controversial. There are some possible mechanisms which could explain the action of G-CSF on destructed mucosal tissue. One of them is that G-CSF may accelerate healing after a mucosal insult as seen with other cytokines or it reduces the extent of injury [23, 24]. The other mechanism of action may be the effect of G-CSF on mucosal cells like hematopoetic cells and the occurrence of the regeneration. A third mechanism is that G-CSF increases circulating neutrophils, neutrophil migration reduces cell debris, other cytokines and scavenges necrotic materials. In summary, the available information, both clinical and experimental, supports the notion that the development of NEC involves multiple factors in the setting of stressed gut with immature protective mechanisms. In our opinion GCSF seems to have had an effect on healing after injury and because of this mechanism we did not measure any mediators (e.g., interleukins, leukotrienes) [25]. This study found that the administration of G-CSF reduced the hypoxic changes of intestines in an experimental model of NEC.

References 1. Edelson MB, Bagwell CE, Rozyeki HJ (1999) Circulating pro-and counter inflammatory cytokine levels and severity in necrotizing enterocolitis. Pediatrics 103:766–771 2. Clark DA, Miller MJ (1990) Intraluminal pathogenesis of necrotizing neterocolitis. J Pediatr 117:S64–S67

Pediatr Surg Int (2006) 22:719–723 3. Stoll BJ (1994) Epidemiology of necrotizing enterocolitis. Clin Perinatol 2:205–218 4. Kliegman RM, Fanarof AA (1984) Necrotizing enterocolitis. N Engl J Med 310:1093–1103 5. MacKendrick W, Hill N, Hsuch W, Caplan M (1993) Increase in plasma platelet-activating factor levels in enterally fed preterm infants. Biol Neonate 64:89–95 6. Ledbetter DJ, Juul SE (2000) Erythropoietin and the incidence of necrotizing enterocolitis in infants with very low birth weight. J Pediatr Surg 35:178–182 7. Warner BW (2002) NEC, EGF, Milk and Spit. Gastroenterology 123:383–384 8. Kling PJ, Sullivan TM, Roberts RA et al (1998) Human milk as a potential enteral source of erythropoietin. Pediatr Res 43:216–221 9. Juul SE, Yachnis AT, Rojiani AM et al (1998) The tissue distribution of Epo and Epo-R in the developing human fetus. Early Hum Dev 52:235–249 10. Calhoun DA, Lunøe M, Du Y, Christensen RD (2000) Granulocyte colony-stimulating factor is present in human milk and its receptor is present in human fetal intestine. Pediatrics 105(1):e7 11. Okur H, Ku¨çu¨kaydın N, Ko¨se K et al (1995) Hypoxia-induced necrotizing enterocolitis in the immature rat: the role of lipid peroxidation and management by vitamin E. J Pediatr Surg 30:1416–1419 12. Kazez A, Ku¨çu¨kaydın N, Ku¨çu¨kaydın M et al (1997) A model of hypoxia-induced necrotizing enterocolitis: the role of distension. J. Pediatr Surg 32:1466–1469 13. Clark DA, Farnabaio DM, McNeill H et al (1988) Contribution of oxygen-derived free radicals to experimental necrotizing enterocolitis. Am J Pathol 130:537–542 14. Hansbrough F, Priebe CJ, Falterman KW et al (1983) Pathogenesis of early necrotizing enterocolitis in the hypoxic neonatal dog. Am J Surg 145:169–175 15. Harrison MW, Connell RS, Campbell JR et al (1975) Microsirculatory changes in gastrointestinal tract of the hypoxic puppy: an electron microscope study. J Pediatr Surg 10:599–608

723 16. Kliegman RM (1990) Models of the pathogenesis of necrotizing enterocolitis. J Pediatr 117:S2–S5 17. Akisu M, Kullahcioglu Girgin F, Baka M et al (2001) The role of recombinant human erythropoietin in lipid peroxidation and platelet-activating factor generation in a rat model of necrotizing enterocolitis. Eur J Pediatr Surg 11:167–172 _ O ¨ ztu¨rk H, Dokucu AI, ¨ gun C, Buÿu¨kbayram H (2002) 18. O Protective effects of recombinant human interleukin-10 on intestines of hypoxia-induced necrotizing enterocolitis in immature rats. J Pediatr Surg 37:1330–1333 19. Welte K, Platzer E, Lu L et al (1985) Purification and biochemical characterization for human pluripotent hematopoietic colony-stimulating factor. Proc Natl Acad Sci USA 82:1526–1530 20. Calhoun DA, Christensen RD (2000) Human developmental biology of granulocyte colony-stimulating factor. Clin Perinatol 27:559–576 21. Calhoun DA, Chegini H, Polliotti BM et al (2001) Granulocyte colony-stimulating factor in preterm and term pregnancy, parturition, and intra-amniotic infection. Obstet Gynecol 97:229–234 22. Calhoun DA, Lunøe M, Du Y et al (1999) Concentrations of granulocyte colony-stimulating factor in human milk after in vitro simulations of digestion. Pediatr Res 46:767–771 23. Maheshwari A (2004) Role of cytokines in human intestinal villous development. Clin Perinatol 31:143–155 24. Ido A, Numata M, Kodama M, Tsubouchi H (2005) Mucosal repair and growth factors: recombinant human hepatocyte growth factor as an innovative therapy for inflammatory bowel disease. J Gastroenterol 40:925–931 25. Canpolat FE, Yurdakok M, Korkmaz A, Yigit S, Tekinalp G (2006) Enteral granulocyte colony-stimulating factor for the treatment of mild (stage I) necrotizing enterocolitis: a placebo-controlled pilot study. J Pediatr Surg 41:1134–1138

123