- Email: [email protected]
Prophylactic treatment with growth hormone and insulin-like growth factor I improve systemic bacterial clearance and survival in a murine model of burn-induced gut-derived sepsis
Burns 25 (1999) 425±430
Prophylactic treatment with growth hormone and insulin-like growth factor I improve systemic bacterial clearance and survival in a murine model of burn-induced gut-derived sepsis Ryoji Fukushima a, *, Hideaki Saito c, Tomomi Inoue b, Kazuhiko Fukatsu b, Tsuyoshi Inaba b, Ilsoo Han b, Satoshi Furukawa b, Ming-Tsan Lin b, Tetsuichiro Muto b a
Department of Surgery II, Teikyo University School of Medicine, 2-11-1 Kaga, Itabashi-ku, 173-8605 Tokyo, Japan b Department of Surgery I, The University of Tokyo, Tokyo, Japan c Surgical Center, The University of Tokyo, Tokyo, Japan Accepted 8 December 1998
Abstract The purpose of this investigation was to evaluate the eects of GH and IGF-I administration in a murine model of burninduced gut-derived sepsis. BALB/C mice were treated with 4.8 mg/kg/day of GH, 24 mg/kg/day of IGF-I or saline for 4 days. They were then administered 1010 E. coli by gavage and subjected to 20% full thickness ¯ame burn. All mice received allogeneic blood transfusion 5 days before burn injury to induce mild immunosuppression. Seventy-three mice were observed for survival and 51 mice were sacri®ced at 4 and 20 h postburn. Blood, mesenteric lymph nodes (MLN), spleen and liver were harvested aseptically, and viable bacterial counts in the organs were determined. The small intestine was harvested for the evaluation of villus height and mitoses in the crypts. GH and IGF-I groups showed a signi®cantly better survival than the control group. GH and IGF-I groups had signi®cantly greater villus height and mitoses/crypt than the control group. Translocation of bacteria was not signi®cantly dierent among groups, however, the relation between the numbers of viable bacteria in MLN and blood suggests that both GH and IGF-I reduced systemic spread of translocated bacteria. It is concluded that GH and IGF-I had positive eects on outcome in this model of burn-induced gut-derived sepsis. It appears that GH and IGF-I may have immune-enhancing eects and that administration of these agents may be useful for burn injury. # 1999 Elsevier Science Ltd and ISBI. All rights reserved. Keywords: Bacterial translocation; Growth hormone (GH); Insulin-like growth factor I (IGF-I); Gut-derived sepsis
1. Introduction Recent experimental and clinical studies have suggested that microorganisms may escape from the gut lumen by crossing the intestinal mucosal barrier and eventually spread to systemic compartments via the lymphatics or blood vessels; a process de®ned as bacterial translocation [1, 2]. Various experimental studies have shown the appearance of enteric bacteria in the mesenteric lymph nodes (MLN), liver and spleen * Corresponding author.
at any time from hours to several days after the event in a number of conditions that include burn injury, hemorrhagic shock, pancreatitis, intravenous hyperalimentation and endotoxemia. Many investigators believe that bacterial translocation may play an important role in the development of complicated systemic infection and multiple organ dysfunction syndrome (MODS) in critically ill and injured patients [3]. In recent years, much attention has been focused on hormonal regulation, such as GH and IGF-I as a potential therapeutic strategy to counteract catabolic eects. GH belongs to the somatolactogen family of
0305-4179/99/$20.00+0.00 # 1999 Elsevier Science Ltd and ISBI. All rights reserved. PII: S 0 3 0 5 - 4 1 7 9 ( 9 8 ) 0 0 1 8 8 - 0
426
R. Fukushima et al. / Burns 25 (1998) 425±430
hormones that improve protein metabolism. GH is also the major regulator stimulating the synthesis and secretion of IGF-I from various tissues. The anabolic eects of GH on protein metabolism are mediated mainly by IGF-I. In postoperative or critically ill patients, IGF-I levels are decreased [4, 5]. GH levels can also be reduced in severe stress states [6, 7]. Exogenous GH or IGF-I has been used in these catabolic states to improve impaired protein metabolism [8± 10]. IGF-I has also been shown to induce marked growth of gut tissues, indicating that the gut may be a sensitive IGF-I target tissue. Huang et al. have shown that IGF-I reduces gut atrophy and translocation of bacteria in a rat severe burn model [11]. More recently it has been suggested that GH and IGF-I have immunomodulatory properties [12, 13]. It has also been suggested that administering these agents may improve the outcome in animal models of severe injury or infection. Edwards et al. demonstrated that subcutaneous injection of porcine GH at 500 mg/kg/day for 12 days (6 days of pretreatment followed by 6 days of treatment) improved the survival of rats challenged intraperitoneally with Salmonella typhimurium [14]. Gomezde-Segura et al. have shown that 1 mg/kg GH treatment for 7 days improved survival of irradiated rats [15]. We have previously shown that 6-day pretreatment with sc injection of GH (4.8 mg/kg/day) or IGF-I (24 mg/kg/day) signi®cantly improved survival in the murine E. coli peritonitis model [16]. The purpose of this study was to assess the potential eects of GH and IGF-I on bacterial translocation and outcome (survival) using a gut-derived sepsis model associated with burn injury.
transfused via the tail vein to Balb/c mice 5 days prior to burn injury as described previously [17, 18]. 2.3. Growth hormone (GH) and insulin like growth factor I (IGF-I) treatment Recombinant human GH (Genotropin) was supplied by Kabi Pharmacia (Stockholm, Sweden). A 0.8-IU dose of GH is equivalent to 0.4 mg of Genotropin. Recombinant human (IGF-I) was donated by Mitsubishi Chemical Corporation (Tokyo, Japan). After blood transfusion, 4.8 mg/kg/day GH, 24 mg/kg/ day IGF-I or saline was administered subcutaneously, twice a day for 4 days. 2.4. Preparation of bacteria About 40 h before bacterial challenge, culture of E. coli ATCC25922 (courtesy of Clinical Laboratory Department, University of Tokyo) was incubated in brain heart infusion broth in a 378C oscillating water bath for 18 h. The culture was centrifuged at 2,000 rpm for 5 min, and the resulting pellets were washed with 0.9% NaCl. Twenty-four h before gavage, the bacterial suspension was serially diluted and plated. Blood agar plates (Nissui-Seiyaku Co., Tokyo, Japan), consisting of blood agar base supplemented with 4% de®brinated horse blood, were used to count viable bacteria. The plates for aerobic culture were incubated at 378C for 24 h, and viable bacterial colony counts were determined. The bacterial suspension was then kept at 48C in a refrigerator overnight. The suspension was adjusted to a ®nal concentration of 1010 bacteria per ml. 2.5. Gavage and burn procedures
2. Materials and methods 2.1. Animals and animal care Adult female Balb/c mice weighing 19±21 g (Charles River Laboratory, Wilmington, MA) and adult female C3H/HeJ mice 20±25 g (Jackson Laboratory, Bar Harbor, ME) were purchased from Japan SLC Company (Hamamatsu, Japan). In accordance with our institutional guidelines, the animals suered no unnecessary discomfort, pain or injury and received proper care and maintenance. All animals were housed for at least one week before the experiment. They were exposed to constant temperature (248C) and humidity (60%), and were fed standard mouse chow. 2.2. Blood transfusion In order to induce mild immunosuppression, blood was obtained from C3H/HeJ mice and 0.2 ml was
One day before burn injury, the animals had hair of the torso removed by clipping. Food was withheld for 18 h but water was provided ad libitum before gavage of 0.1 ml of the E. coli suspension (1010 E. coli/ml) while the animals were awake. After gavage, they were anesthetized by ether inhalation, and a 20% full thickness ¯ame burn was in¯icted using the technique of Stieritz and Holder [19]. Saline 0.5 ml was given intraperitoneally immediately after burn injury for ¯uid resuscitation and the animals were allowed to recover from anesthesia with free access to food and water. 2.6. Experimental protocol One hundred and twenty-four animals were randomized into GH, IGF-I and saline treatment groups. After 4 days of GH, IGF-I or saline treatment, all animals were administered 1010 E. coli by gavage and then were burned. Seventy-three animals were observed for survival and the rest were sacri®ced 4 (n = 30) and
R. Fukushima et al. / Burns 25 (1998) 425±430
427
test. A w 2 test or Fisher's exact test was used for categorical data. Survival was analyzed by log rank test. Simple linear regression analysis was also conducted. The dierence in regression between groups was tested by z score. A p-value less than 0.05 was considered statistically signi®cant. 3. Results 3.1. Survival
Fig. 1. Survival rate of mice subjected to transfusion, gavage and burn injury, treated for 4 days with GH (4.8 mg/kg/day), IGF-I (24 mg/kg/day) or saline. GH and IGF-I group had signi®cantly better survival than saline controls ( p < 0.05).
20 (n = 21) h after burn injury. The animals were prepped with 70% alcohol, and aseptic techniques were used to remove the mesenteric lymph nodes (MLN), spleen and liver. Blood was also withdrawn by cardiac puncture. The tissues obtained were individually weighed, homogenized with sterile saline, and 100 ml of the homogenate was plated (diluted when appropriate) on blood agar plates (Nissui-Seiyaku Co., Tokyo, Japan) for quantitative determination of bacterial colony counts after 18 h of aerobic incubation at 378C. Viable colony counts were calculated and adjusted to be expressed as bacteria/g or ml of tissue as previously described [20].
Fig. 1 shows the survival curves for animals treated with either GH or IGF-I, and the saline-treated controls. GH and IGF-I signi®cantly improved the survival of animals after burn injury. The best survival was obtained in the IGF-I group (50%), followed by the GH group (46%). Twenty-four percent of saline-control animals survived in this experimental setting. 3.2. Bacterial translocation There was no dierence in the incidence of bacterial translocation 4 h after burn injury among the three
2.7. Intestinal structure Full-thickness samples from three segments of the gastrointestinal tract were obtained 5 cm distal to the pylorus, 5 cm proximal to the terminal ileum and in the mid-portion of the colon, 20 h after burn injury. The samples were ®xed in 10% buered formaldehyde. Paran sections, stained with hematoxylin and eosin, were made from each sample. Images of these sections were acquired with a light microscope connected to a video camera (ITC-370M; Ikegami Tsusinki, Tokyo, Japan). Quantitative morphometric analyses of the villus height were performed on these images, using an Image Command 5098 software package for a personal computer (Nippon Avionics, Tokyo, Japan). For estimating crypt cell proliferation, the number of mitoses per crypt was also determined. All measurements were performed in a blinded fashion on coded slides. 2.8. Statistical analysis Data are expressed as mean 2SEM. Continuous data were analyzed using analysis of variance (ANOVA) followed by Duncan's new multiple range
Fig. 2. Incidence of bacterial translocation in mesenteric lymph nodes (MLN), liver, spleen and blood as assessed by positive culture 4 (n = 10 each) and 20 (n = 7 each) h after burn injury. Dierence among groups showed no statistical signi®cance.
428
R. Fukushima et al. / Burns 25 (1998) 425±430
Table 1 Intestinal structure
3.3. Intestinal structure
Villous height (mm)
GH (n=7) IGF-I (n=7) Control (n=7)
Mitoses/crypt
jejunum
ileum
jejunum
ileum
colon
511230* 498216* 423222
258210 256211 248219
3.120.3* 3.120.2* 2.420.2
2.220.1 2.220.2 2.120.2
2.120.2 2.020.1 1.920.1
GH and IGF-I treatment preserved the intestinal villous height and increased mitoses/crypt in the proximal part of the intestine following burn injury. Values are mean2SEM. *p<0.05 vs control.
groups as assessed by the incidence of positive culture in MLN, liver and spleen (Fig. 2). The incidence of positive blood culture was zero in all groups of animals at 4 h (Fig. 2). Twenty h after burn injury, translocation of bacteria as assessed by the incidence of positive culture in blood, MLN, spleen and liver was lower in both GH- and IGF-I-treated animals than saline controls, however, the dierences did not reach statistical signi®cance (Fig. 2). The vast majority of organisms cultured in MLN, liver, spleen and blood were E. coli. Looking at the relationship between the number of viable bacteria in MLN and blood, the slope of the regression line was much steeper in control animals compared to GH-or IGF-I treated mice ( p < 0.05), which indicates that less bacteria in MLN spread to the systemic circulation in GH- and IGF-I-treated animals compared to controls (Fig. 3).
Fig. 3. Relationship between number of viable bacteria in MLN and blood. The slope of the regression line was much steeper in control animals compared to GH- or IGF-I-treated mice ( p < 0.05) which indicates that less bacteria in MLN spread to the systemic circulation in GH- and IGF-I-treated animals compared to control. * represents control, q GH and R IGF-I.
The villus height and the number of mitoses per crypt in the proximal small intestine in both GH and IGF-I group were signi®cantly greater than in the saline control group (Table 1). However, the villus height and the number of mitoses per crypt in the distal small intestine and the colon were not signi®cantly dierent among the three groups. 4. Discussion Major burn injury is associated with alterations in the intestinal mucosal barrier and systemic immunity resulting in bacterial translocation to MLN and systemic organs leading to sepsis and sepsis-related MSOF. Two major factors determine the occurrence and the severity of infections associated with bacterial translocation from the gut; the magnitude of the bacterial load and the ability of the host to clear the invading organism. Thus, survival is related to the burden of bacteria (and/or their products) crossing the mucosal barrier and overall competence of the host immune response and antibacterial defenses. An ideal therapeutic modality should decrease the burden of microbes and their products crossing the mucosal barrier and stimulate immune mechanisms to increase bacterial clearance. In the present study, exogenous GH or IGF-I treatment resulted in improved survival in a highly lethal murine burn model. We used a blood transfusion/burn model with gavage of E. coli which is associated with a mortality rate of approximately 70±90% over 10 days of observation [18]. Blood transfusion leads to immunosuppression and increased susceptibility to infections [17]. Blood transfusions are used commonly in surgical patients, burn injury and trauma, and the high mortality rate of this model is useful for the evaluation of therapeutic interventions. The improved survival achieved by GH- and IGF-I in the present study was not directly associated with decreased translocation of viable bacteria to MLN and other organs. However, when looking at the relationship between the number of viable bacteria in MLN and blood, the slope of the regression line was signi®cantly steeper in control animals compared to GH or IGF-I treated mice. It is likely that in GH- and IGF-Itreated animals, less viable bacteria in MLN spread to the systemic circulation compared to controls and this may have contributed to the improved survival. It is becoming increasingly apparent that GH and IGF-I have immunoregulatory roles. In a previous investigation, we have shown that postoperative GH treatment enhances cellular immunity in patients undergoing colonic surgery. Vara-Thorbeck et al.
R. Fukushima et al. / Burns 25 (1998) 425±430
showed that GH reduced the rate of wound infection in postoperative patients [21]. In their study, GH was suggested, though not con®rmed, to enhance human PMN bactericidal capacity. We recently demonstrated that pretreatment with these hormones enhanced host defense and prolonged survival in a murine sepsis model [16]. We also showed that both GH and IGF-I augmented in vitro E. coli-killing by murine peritoneal exudative cells [22]. Fu et al. showed that GH enhanced Oÿ2 secretion and killing of E. coli by rat bone marrow PMN [23]. The morphological evaluation of the small bowel in the current study showed that GH and IGF-I treatment preserved the intestinal villus height following burn injury. The preservation of mucosal structure can be achieved by reduced atrophy, enhanced proliferation of mucosal cells or both reduced atrophy and enhanced proliferation. It is not clear from the present study whether GH and IGF-I actually reduced atrophy of the mucosal cells, however, it is at least apparent that these agents enhanced proliferation as evidenced by the observations of increased mitoses/crypt. The favorable eects of these agents on gut mucosa are in agreement with recent studies of others in which IGF-I administration during burn injury [11], small bowel transplantation [24], and sepsis [25] had positive eects on intestinal structure. Speci®c receptors for IGF-I have been demonstrated in the gastrointestinal tract of both humans and experimental animals, suggesting that the intestine is a major IGF-I target tissue. It is interesting to note, however, that the trophic eect of GH and IGF-I on the gut mucosa was demonstrated exclusively in the proximal part of the small intestine. We found no signi®cant dierence in mucosal structural indices in the distal part of the small intestine or the colon. Similar ®ndings were reported by others [11, 25]. Using a severe burn injury model, Huang et al. showed that the mucosal protein and DNA content in the distal bowel increased less signi®cantly than those in the proximal bowel in IGFI-treated rats [11]. Chen also reported the distal small intestine in septic rats to be less sensitive to IGF-I than the proximal intestine, by histologic study [25]. Lund et al. reported that speci®c mRNA for IGF-I decreases from the duodenum to the colon [26], which suggests that there is a similar distribution for IGF-I receptors, and thus, dierent regions of the bowel show dierent sensitivities to its eects. As opposed to some previous investigations [11, 15], our results showed a tendency but could not demonstrate a signi®cant decrease in the occurrence of bacterial translocation as assessed by the incidence of positive culture in the organs in GH- and IGF-I-treated animals. This may partly be explained by the fact that mucosal integrity was preserved only in the proximal part of the intestine by GH or IGF-I, and that a
429
subtle dierence can not be demonstrated in relatively small numbers of experimental animals. Wells et al. previously reported that translocation of E. coli was most likely to occur in the cecum or colon in their healthy monoassociated mouse model [27]. In addition, there are some reports showing no association between bacterial translocation and gut mucosal height [28]. Although we must always keep in mind that gut barrier function can not always be evaluated by means of culturing viable bacteria in the organs [20], it appeared that GH and IGF-I have greater eects on preventing systemic spread of bacteria rather than on the passage of bacteria across the intestinal mucosa in this experimental setting. In summary, exogenous GH and IGF-I signi®cantly improved survival of animals after burn injury. The bene®cial eects of these hormones can partly be explained by enhancement of host defence as manifested by prevention of systemic spread of translocated bacteria. It appears that administration of GH or IGF-I is of potential bene®t for burn injury.
Acknowledgements Recombinant human GH (Genotropin) was kindly supplied by Kabi Pharmacia Stockholm, Sweden. Recombinant human IGF-I was kindly donated by Mitsubishi Chemical Corporation Tokyo, Japan.
References [1] Berg RD, Garlington AW. Translocation of certain indigenous bacteria from the gastrointestinal tract to the mesenteric lymph nodes and other organs in a gnotobiotic mouse model. Infect. Immun. 1979;23:403±11. [2] Alexander JW, Boyce ST, Babcock GF et al. The process of microbial translocation. Ann. Surg. 1990;212:496±512. [3] Carrico W, Meakins JL, Marschall JC, Fry D, Maier RV. Multiple-organ-failure syndrome. Arch. Surg. 1986;121:196±208. [4] Cotterill AM, Mendel P, Holly JM et al. The dierential regulation of the circulating levels of the insulin-like growth factors and their binding proteins (IGFBP) 1, 2 and 3 after elective abdominal surgery. Clin. Endocrinol. (Oxf) 1996;44:91±101. [5] Abribat T, Brazeau P, Davignon I, Garrel DR. Insulin-like growth factor-I blood levels in severely burned patients: eects of time postinjury, age of patient and severity of burn. Clin. Endocrinol. 1993;39:583±9. [6] Dahn MS, Mitchell RA, Smith S, Lange MP, Whitcomb MP, Kirkpatrick JR. Altered immunologic function and nitrogen metabolism associated with depression of plasma growth hormone. J. Parenter. Enteral. Nutr. 1984;8:690±4. [7] Jeevanandam M, Ramias L, Shamos RF, Schiller WR. Decreased growth hormone levels in the catabolic phase of severe injury. Surgery 1992;111:495±502. [8] Hammarqvist F, Stromberg C, von der Decken A, Vinnars E, Wernerman J. Biosynthetic human growth hormone preserves both muscle protein synthesis and the decrease in muscle-free
430
[9] [10] [11] [12] [13] [14]
[15] [16] [17] [18]
[19]
R. Fukushima et al. / Burns 25 (1998) 425±430 glutamine, and improves whole-body nitrogen economy after operation. Ann. Surg. 1992;216:184±91. Voerman HJ, van Schijndel RJ, Groeneveld AB et al. Eects of recombinant human growth hormone in patients with severe sepsis. Ann. Surg. 1992;216:648±55. Cio WG, Gore DC, Rue LW et al. Insulin-like growth factorI lowers protein oxidation in patients with thermal injury. Ann. Surg. 1994;220:310±9. Huang KF, Chung DH, Herndon DN. Insulin-like growth factor-I (IGF-I) reduces gut atrophy and bacterial translocation after severe burn injury. Arch. Surg. 1993;128:47±53. Kelley KW. The role of growth hormone in modulation of the immune response. Ann. NY Acad. Sci. 1990;594:95±103. Saito H, Inoue T, Fukatsu K et al. Growth hormone and the immune response to bacterial infection. Horm. Res. 1996;45:50± 4. Edwards CK, Yunger LM, Lorence RM, Dantzer R, Kelley KW. The pituitary gland is required for protection against lethal eects of Salmonella typhimurium. Proc. Natl. Acad. Sci. USA 1991;88:22747. Gomez-de-Segura IA, Prieto I, Grande AG et al. Growth hormone reduces mortality and bacterial translocation in irradiated rats. Acta Oncol. 1998;37:179±85. Inoue T, Saito H, Fukushima R et al. Growth hormone and insulinlike growth factor I enhance host defense in a murine sepsis model. Arch. Surg. 1995;130:1115±22. Gianotti L, Pyles T, Alexander JW, Babcock GF, Carey MA. Impact of blood transfusion and burn injury on microbial translocation and bacterial survival. Transfusion 1992;32:312±7. Fukushima R, Gianotti L, Alexander JW, Pyles T. The degree of bacterial translocation is a determinant factor for mortality after burn injury and is improved by prostaglandin analogs. Ann. Surg. 1992;216:438±44 discussion 444-5. Stieritz DD, Holder IA. Experimental studies of the pathogenesis of infections due to Pseudomonas aeruginosa: description of a burned mouse model. J. Infect. Dis. 1975;131:688±91.
[20] Alexander JW, Gianotti L, Pyles T, Carey MA, Babcock GF. Distribution and survival of Escherichia coli translocating from the intestine after thermal injury. Ann. Surg. 1991;213:558±66 discussion 566-7. [21] Vara-Thorbeck R, Guerrero JA, Rosell J, Ruiz RE, Capitan JM. Exogenous growth hormone: eects on the catabolic response to surgically produced acute stress and on postoperative immune function. World J. Surg. 1993;17:530±8. [22] Inoue T, Saito H, Hashiguchi Y et al. Growth hormone and insulin-like growth factor I augment Escherichia coli-killing activity of murine peritoneal exudative cells. Shock 1996;6:345±50. [23] Fu YK, Arkins S, Li YM, Dantzer R, Kelley KW. Reduction in superoxide anion secretion and bactericidal activity of neutrophils from aged rats: reversal by the combination of gamma interferon and growth hormone. Infect. Immun. 1994;62:1±8. [24] Zhang W, Frankel WL, Adamson WT et al. Insulin-like growth factor-I improves mucosal structure and function in transplanted rat small intestine. Transplantation 1995;59:755±61. [25] Chen K, Okuma T, Okamura K, Tabira Y, Kaneko H, Miyauchi Y. Insulin-like growth factor-I prevents gut atrophy and maintains intestinal integrity in septic rats. J. Parenter. Enteral. Nutr. 1995;19:119±24. [26] Lund PK, Ulshen MH, Rountree DB, Selub SE, Buchan AM. Molecular biology of gastrointestinal peptides and growth factors: relevance to intestinal adaptation. Digestion 1990;2:66±73. [27] Wells C, Erlandsen S. Localization of translocating Escherichia coli, Proteus mirabilis and Enterococcus faecalis within cecal and colonic tissues of monoassociated mice. Infect. Immun. 1991;59:4693±7. [28] McCauley RD, Heel KA, Christiansen KJ, Hall JC. The eect of minimum luminal nutrition on bacterial translocation and atrophy of the jejunum during parenteral nutrition. J. Gastroenterol. Hepatol. 1996;11:65±70.
Prophylactic treatment with growth hormone and insulin-like growth factor I improve systemic bacterial clearance and survival in a murine model of burn-induced gut-derived sepsis Ryoji Fukushima a, *, Hideaki Saito c, Tomomi Inoue b, Kazuhiko Fukatsu b, Tsuyoshi Inaba b, Ilsoo Han b, Satoshi Furukawa b, Ming-Tsan Lin b, Tetsuichiro Muto b a
Department of Surgery II, Teikyo University School of Medicine, 2-11-1 Kaga, Itabashi-ku, 173-8605 Tokyo, Japan b Department of Surgery I, The University of Tokyo, Tokyo, Japan c Surgical Center, The University of Tokyo, Tokyo, Japan Accepted 8 December 1998
Abstract The purpose of this investigation was to evaluate the eects of GH and IGF-I administration in a murine model of burninduced gut-derived sepsis. BALB/C mice were treated with 4.8 mg/kg/day of GH, 24 mg/kg/day of IGF-I or saline for 4 days. They were then administered 1010 E. coli by gavage and subjected to 20% full thickness ¯ame burn. All mice received allogeneic blood transfusion 5 days before burn injury to induce mild immunosuppression. Seventy-three mice were observed for survival and 51 mice were sacri®ced at 4 and 20 h postburn. Blood, mesenteric lymph nodes (MLN), spleen and liver were harvested aseptically, and viable bacterial counts in the organs were determined. The small intestine was harvested for the evaluation of villus height and mitoses in the crypts. GH and IGF-I groups showed a signi®cantly better survival than the control group. GH and IGF-I groups had signi®cantly greater villus height and mitoses/crypt than the control group. Translocation of bacteria was not signi®cantly dierent among groups, however, the relation between the numbers of viable bacteria in MLN and blood suggests that both GH and IGF-I reduced systemic spread of translocated bacteria. It is concluded that GH and IGF-I had positive eects on outcome in this model of burn-induced gut-derived sepsis. It appears that GH and IGF-I may have immune-enhancing eects and that administration of these agents may be useful for burn injury. # 1999 Elsevier Science Ltd and ISBI. All rights reserved. Keywords: Bacterial translocation; Growth hormone (GH); Insulin-like growth factor I (IGF-I); Gut-derived sepsis
1. Introduction Recent experimental and clinical studies have suggested that microorganisms may escape from the gut lumen by crossing the intestinal mucosal barrier and eventually spread to systemic compartments via the lymphatics or blood vessels; a process de®ned as bacterial translocation [1, 2]. Various experimental studies have shown the appearance of enteric bacteria in the mesenteric lymph nodes (MLN), liver and spleen * Corresponding author.
at any time from hours to several days after the event in a number of conditions that include burn injury, hemorrhagic shock, pancreatitis, intravenous hyperalimentation and endotoxemia. Many investigators believe that bacterial translocation may play an important role in the development of complicated systemic infection and multiple organ dysfunction syndrome (MODS) in critically ill and injured patients [3]. In recent years, much attention has been focused on hormonal regulation, such as GH and IGF-I as a potential therapeutic strategy to counteract catabolic eects. GH belongs to the somatolactogen family of
0305-4179/99/$20.00+0.00 # 1999 Elsevier Science Ltd and ISBI. All rights reserved. PII: S 0 3 0 5 - 4 1 7 9 ( 9 8 ) 0 0 1 8 8 - 0
426
R. Fukushima et al. / Burns 25 (1998) 425±430
hormones that improve protein metabolism. GH is also the major regulator stimulating the synthesis and secretion of IGF-I from various tissues. The anabolic eects of GH on protein metabolism are mediated mainly by IGF-I. In postoperative or critically ill patients, IGF-I levels are decreased [4, 5]. GH levels can also be reduced in severe stress states [6, 7]. Exogenous GH or IGF-I has been used in these catabolic states to improve impaired protein metabolism [8± 10]. IGF-I has also been shown to induce marked growth of gut tissues, indicating that the gut may be a sensitive IGF-I target tissue. Huang et al. have shown that IGF-I reduces gut atrophy and translocation of bacteria in a rat severe burn model [11]. More recently it has been suggested that GH and IGF-I have immunomodulatory properties [12, 13]. It has also been suggested that administering these agents may improve the outcome in animal models of severe injury or infection. Edwards et al. demonstrated that subcutaneous injection of porcine GH at 500 mg/kg/day for 12 days (6 days of pretreatment followed by 6 days of treatment) improved the survival of rats challenged intraperitoneally with Salmonella typhimurium [14]. Gomezde-Segura et al. have shown that 1 mg/kg GH treatment for 7 days improved survival of irradiated rats [15]. We have previously shown that 6-day pretreatment with sc injection of GH (4.8 mg/kg/day) or IGF-I (24 mg/kg/day) signi®cantly improved survival in the murine E. coli peritonitis model [16]. The purpose of this study was to assess the potential eects of GH and IGF-I on bacterial translocation and outcome (survival) using a gut-derived sepsis model associated with burn injury.
transfused via the tail vein to Balb/c mice 5 days prior to burn injury as described previously [17, 18]. 2.3. Growth hormone (GH) and insulin like growth factor I (IGF-I) treatment Recombinant human GH (Genotropin) was supplied by Kabi Pharmacia (Stockholm, Sweden). A 0.8-IU dose of GH is equivalent to 0.4 mg of Genotropin. Recombinant human (IGF-I) was donated by Mitsubishi Chemical Corporation (Tokyo, Japan). After blood transfusion, 4.8 mg/kg/day GH, 24 mg/kg/ day IGF-I or saline was administered subcutaneously, twice a day for 4 days. 2.4. Preparation of bacteria About 40 h before bacterial challenge, culture of E. coli ATCC25922 (courtesy of Clinical Laboratory Department, University of Tokyo) was incubated in brain heart infusion broth in a 378C oscillating water bath for 18 h. The culture was centrifuged at 2,000 rpm for 5 min, and the resulting pellets were washed with 0.9% NaCl. Twenty-four h before gavage, the bacterial suspension was serially diluted and plated. Blood agar plates (Nissui-Seiyaku Co., Tokyo, Japan), consisting of blood agar base supplemented with 4% de®brinated horse blood, were used to count viable bacteria. The plates for aerobic culture were incubated at 378C for 24 h, and viable bacterial colony counts were determined. The bacterial suspension was then kept at 48C in a refrigerator overnight. The suspension was adjusted to a ®nal concentration of 1010 bacteria per ml. 2.5. Gavage and burn procedures
2. Materials and methods 2.1. Animals and animal care Adult female Balb/c mice weighing 19±21 g (Charles River Laboratory, Wilmington, MA) and adult female C3H/HeJ mice 20±25 g (Jackson Laboratory, Bar Harbor, ME) were purchased from Japan SLC Company (Hamamatsu, Japan). In accordance with our institutional guidelines, the animals suered no unnecessary discomfort, pain or injury and received proper care and maintenance. All animals were housed for at least one week before the experiment. They were exposed to constant temperature (248C) and humidity (60%), and were fed standard mouse chow. 2.2. Blood transfusion In order to induce mild immunosuppression, blood was obtained from C3H/HeJ mice and 0.2 ml was
One day before burn injury, the animals had hair of the torso removed by clipping. Food was withheld for 18 h but water was provided ad libitum before gavage of 0.1 ml of the E. coli suspension (1010 E. coli/ml) while the animals were awake. After gavage, they were anesthetized by ether inhalation, and a 20% full thickness ¯ame burn was in¯icted using the technique of Stieritz and Holder [19]. Saline 0.5 ml was given intraperitoneally immediately after burn injury for ¯uid resuscitation and the animals were allowed to recover from anesthesia with free access to food and water. 2.6. Experimental protocol One hundred and twenty-four animals were randomized into GH, IGF-I and saline treatment groups. After 4 days of GH, IGF-I or saline treatment, all animals were administered 1010 E. coli by gavage and then were burned. Seventy-three animals were observed for survival and the rest were sacri®ced 4 (n = 30) and
R. Fukushima et al. / Burns 25 (1998) 425±430
427
test. A w 2 test or Fisher's exact test was used for categorical data. Survival was analyzed by log rank test. Simple linear regression analysis was also conducted. The dierence in regression between groups was tested by z score. A p-value less than 0.05 was considered statistically signi®cant. 3. Results 3.1. Survival
Fig. 1. Survival rate of mice subjected to transfusion, gavage and burn injury, treated for 4 days with GH (4.8 mg/kg/day), IGF-I (24 mg/kg/day) or saline. GH and IGF-I group had signi®cantly better survival than saline controls ( p < 0.05).
20 (n = 21) h after burn injury. The animals were prepped with 70% alcohol, and aseptic techniques were used to remove the mesenteric lymph nodes (MLN), spleen and liver. Blood was also withdrawn by cardiac puncture. The tissues obtained were individually weighed, homogenized with sterile saline, and 100 ml of the homogenate was plated (diluted when appropriate) on blood agar plates (Nissui-Seiyaku Co., Tokyo, Japan) for quantitative determination of bacterial colony counts after 18 h of aerobic incubation at 378C. Viable colony counts were calculated and adjusted to be expressed as bacteria/g or ml of tissue as previously described [20].
Fig. 1 shows the survival curves for animals treated with either GH or IGF-I, and the saline-treated controls. GH and IGF-I signi®cantly improved the survival of animals after burn injury. The best survival was obtained in the IGF-I group (50%), followed by the GH group (46%). Twenty-four percent of saline-control animals survived in this experimental setting. 3.2. Bacterial translocation There was no dierence in the incidence of bacterial translocation 4 h after burn injury among the three
2.7. Intestinal structure Full-thickness samples from three segments of the gastrointestinal tract were obtained 5 cm distal to the pylorus, 5 cm proximal to the terminal ileum and in the mid-portion of the colon, 20 h after burn injury. The samples were ®xed in 10% buered formaldehyde. Paran sections, stained with hematoxylin and eosin, were made from each sample. Images of these sections were acquired with a light microscope connected to a video camera (ITC-370M; Ikegami Tsusinki, Tokyo, Japan). Quantitative morphometric analyses of the villus height were performed on these images, using an Image Command 5098 software package for a personal computer (Nippon Avionics, Tokyo, Japan). For estimating crypt cell proliferation, the number of mitoses per crypt was also determined. All measurements were performed in a blinded fashion on coded slides. 2.8. Statistical analysis Data are expressed as mean 2SEM. Continuous data were analyzed using analysis of variance (ANOVA) followed by Duncan's new multiple range
Fig. 2. Incidence of bacterial translocation in mesenteric lymph nodes (MLN), liver, spleen and blood as assessed by positive culture 4 (n = 10 each) and 20 (n = 7 each) h after burn injury. Dierence among groups showed no statistical signi®cance.
428
R. Fukushima et al. / Burns 25 (1998) 425±430
Table 1 Intestinal structure
3.3. Intestinal structure
Villous height (mm)
GH (n=7) IGF-I (n=7) Control (n=7)
Mitoses/crypt
jejunum
ileum
jejunum
ileum
colon
511230* 498216* 423222
258210 256211 248219
3.120.3* 3.120.2* 2.420.2
2.220.1 2.220.2 2.120.2
2.120.2 2.020.1 1.920.1
GH and IGF-I treatment preserved the intestinal villous height and increased mitoses/crypt in the proximal part of the intestine following burn injury. Values are mean2SEM. *p<0.05 vs control.
groups as assessed by the incidence of positive culture in MLN, liver and spleen (Fig. 2). The incidence of positive blood culture was zero in all groups of animals at 4 h (Fig. 2). Twenty h after burn injury, translocation of bacteria as assessed by the incidence of positive culture in blood, MLN, spleen and liver was lower in both GH- and IGF-I-treated animals than saline controls, however, the dierences did not reach statistical signi®cance (Fig. 2). The vast majority of organisms cultured in MLN, liver, spleen and blood were E. coli. Looking at the relationship between the number of viable bacteria in MLN and blood, the slope of the regression line was much steeper in control animals compared to GH-or IGF-I treated mice ( p < 0.05), which indicates that less bacteria in MLN spread to the systemic circulation in GH- and IGF-I-treated animals compared to controls (Fig. 3).
Fig. 3. Relationship between number of viable bacteria in MLN and blood. The slope of the regression line was much steeper in control animals compared to GH- or IGF-I-treated mice ( p < 0.05) which indicates that less bacteria in MLN spread to the systemic circulation in GH- and IGF-I-treated animals compared to control. * represents control, q GH and R IGF-I.
The villus height and the number of mitoses per crypt in the proximal small intestine in both GH and IGF-I group were signi®cantly greater than in the saline control group (Table 1). However, the villus height and the number of mitoses per crypt in the distal small intestine and the colon were not signi®cantly dierent among the three groups. 4. Discussion Major burn injury is associated with alterations in the intestinal mucosal barrier and systemic immunity resulting in bacterial translocation to MLN and systemic organs leading to sepsis and sepsis-related MSOF. Two major factors determine the occurrence and the severity of infections associated with bacterial translocation from the gut; the magnitude of the bacterial load and the ability of the host to clear the invading organism. Thus, survival is related to the burden of bacteria (and/or their products) crossing the mucosal barrier and overall competence of the host immune response and antibacterial defenses. An ideal therapeutic modality should decrease the burden of microbes and their products crossing the mucosal barrier and stimulate immune mechanisms to increase bacterial clearance. In the present study, exogenous GH or IGF-I treatment resulted in improved survival in a highly lethal murine burn model. We used a blood transfusion/burn model with gavage of E. coli which is associated with a mortality rate of approximately 70±90% over 10 days of observation [18]. Blood transfusion leads to immunosuppression and increased susceptibility to infections [17]. Blood transfusions are used commonly in surgical patients, burn injury and trauma, and the high mortality rate of this model is useful for the evaluation of therapeutic interventions. The improved survival achieved by GH- and IGF-I in the present study was not directly associated with decreased translocation of viable bacteria to MLN and other organs. However, when looking at the relationship between the number of viable bacteria in MLN and blood, the slope of the regression line was signi®cantly steeper in control animals compared to GH or IGF-I treated mice. It is likely that in GH- and IGF-Itreated animals, less viable bacteria in MLN spread to the systemic circulation compared to controls and this may have contributed to the improved survival. It is becoming increasingly apparent that GH and IGF-I have immunoregulatory roles. In a previous investigation, we have shown that postoperative GH treatment enhances cellular immunity in patients undergoing colonic surgery. Vara-Thorbeck et al.
R. Fukushima et al. / Burns 25 (1998) 425±430
showed that GH reduced the rate of wound infection in postoperative patients [21]. In their study, GH was suggested, though not con®rmed, to enhance human PMN bactericidal capacity. We recently demonstrated that pretreatment with these hormones enhanced host defense and prolonged survival in a murine sepsis model [16]. We also showed that both GH and IGF-I augmented in vitro E. coli-killing by murine peritoneal exudative cells [22]. Fu et al. showed that GH enhanced Oÿ2 secretion and killing of E. coli by rat bone marrow PMN [23]. The morphological evaluation of the small bowel in the current study showed that GH and IGF-I treatment preserved the intestinal villus height following burn injury. The preservation of mucosal structure can be achieved by reduced atrophy, enhanced proliferation of mucosal cells or both reduced atrophy and enhanced proliferation. It is not clear from the present study whether GH and IGF-I actually reduced atrophy of the mucosal cells, however, it is at least apparent that these agents enhanced proliferation as evidenced by the observations of increased mitoses/crypt. The favorable eects of these agents on gut mucosa are in agreement with recent studies of others in which IGF-I administration during burn injury [11], small bowel transplantation [24], and sepsis [25] had positive eects on intestinal structure. Speci®c receptors for IGF-I have been demonstrated in the gastrointestinal tract of both humans and experimental animals, suggesting that the intestine is a major IGF-I target tissue. It is interesting to note, however, that the trophic eect of GH and IGF-I on the gut mucosa was demonstrated exclusively in the proximal part of the small intestine. We found no signi®cant dierence in mucosal structural indices in the distal part of the small intestine or the colon. Similar ®ndings were reported by others [11, 25]. Using a severe burn injury model, Huang et al. showed that the mucosal protein and DNA content in the distal bowel increased less signi®cantly than those in the proximal bowel in IGFI-treated rats [11]. Chen also reported the distal small intestine in septic rats to be less sensitive to IGF-I than the proximal intestine, by histologic study [25]. Lund et al. reported that speci®c mRNA for IGF-I decreases from the duodenum to the colon [26], which suggests that there is a similar distribution for IGF-I receptors, and thus, dierent regions of the bowel show dierent sensitivities to its eects. As opposed to some previous investigations [11, 15], our results showed a tendency but could not demonstrate a signi®cant decrease in the occurrence of bacterial translocation as assessed by the incidence of positive culture in the organs in GH- and IGF-I-treated animals. This may partly be explained by the fact that mucosal integrity was preserved only in the proximal part of the intestine by GH or IGF-I, and that a
429
subtle dierence can not be demonstrated in relatively small numbers of experimental animals. Wells et al. previously reported that translocation of E. coli was most likely to occur in the cecum or colon in their healthy monoassociated mouse model [27]. In addition, there are some reports showing no association between bacterial translocation and gut mucosal height [28]. Although we must always keep in mind that gut barrier function can not always be evaluated by means of culturing viable bacteria in the organs [20], it appeared that GH and IGF-I have greater eects on preventing systemic spread of bacteria rather than on the passage of bacteria across the intestinal mucosa in this experimental setting. In summary, exogenous GH and IGF-I signi®cantly improved survival of animals after burn injury. The bene®cial eects of these hormones can partly be explained by enhancement of host defence as manifested by prevention of systemic spread of translocated bacteria. It appears that administration of GH or IGF-I is of potential bene®t for burn injury.
Acknowledgements Recombinant human GH (Genotropin) was kindly supplied by Kabi Pharmacia Stockholm, Sweden. Recombinant human IGF-I was kindly donated by Mitsubishi Chemical Corporation Tokyo, Japan.
References [1] Berg RD, Garlington AW. Translocation of certain indigenous bacteria from the gastrointestinal tract to the mesenteric lymph nodes and other organs in a gnotobiotic mouse model. Infect. Immun. 1979;23:403±11. [2] Alexander JW, Boyce ST, Babcock GF et al. The process of microbial translocation. Ann. Surg. 1990;212:496±512. [3] Carrico W, Meakins JL, Marschall JC, Fry D, Maier RV. Multiple-organ-failure syndrome. Arch. Surg. 1986;121:196±208. [4] Cotterill AM, Mendel P, Holly JM et al. The dierential regulation of the circulating levels of the insulin-like growth factors and their binding proteins (IGFBP) 1, 2 and 3 after elective abdominal surgery. Clin. Endocrinol. (Oxf) 1996;44:91±101. [5] Abribat T, Brazeau P, Davignon I, Garrel DR. Insulin-like growth factor-I blood levels in severely burned patients: eects of time postinjury, age of patient and severity of burn. Clin. Endocrinol. 1993;39:583±9. [6] Dahn MS, Mitchell RA, Smith S, Lange MP, Whitcomb MP, Kirkpatrick JR. Altered immunologic function and nitrogen metabolism associated with depression of plasma growth hormone. J. Parenter. Enteral. Nutr. 1984;8:690±4. [7] Jeevanandam M, Ramias L, Shamos RF, Schiller WR. Decreased growth hormone levels in the catabolic phase of severe injury. Surgery 1992;111:495±502. [8] Hammarqvist F, Stromberg C, von der Decken A, Vinnars E, Wernerman J. Biosynthetic human growth hormone preserves both muscle protein synthesis and the decrease in muscle-free
430
[9] [10] [11] [12] [13] [14]
[15] [16] [17] [18]
[19]
R. Fukushima et al. / Burns 25 (1998) 425±430 glutamine, and improves whole-body nitrogen economy after operation. Ann. Surg. 1992;216:184±91. Voerman HJ, van Schijndel RJ, Groeneveld AB et al. Eects of recombinant human growth hormone in patients with severe sepsis. Ann. Surg. 1992;216:648±55. Cio WG, Gore DC, Rue LW et al. Insulin-like growth factorI lowers protein oxidation in patients with thermal injury. Ann. Surg. 1994;220:310±9. Huang KF, Chung DH, Herndon DN. Insulin-like growth factor-I (IGF-I) reduces gut atrophy and bacterial translocation after severe burn injury. Arch. Surg. 1993;128:47±53. Kelley KW. The role of growth hormone in modulation of the immune response. Ann. NY Acad. Sci. 1990;594:95±103. Saito H, Inoue T, Fukatsu K et al. Growth hormone and the immune response to bacterial infection. Horm. Res. 1996;45:50± 4. Edwards CK, Yunger LM, Lorence RM, Dantzer R, Kelley KW. The pituitary gland is required for protection against lethal eects of Salmonella typhimurium. Proc. Natl. Acad. Sci. USA 1991;88:22747. Gomez-de-Segura IA, Prieto I, Grande AG et al. Growth hormone reduces mortality and bacterial translocation in irradiated rats. Acta Oncol. 1998;37:179±85. Inoue T, Saito H, Fukushima R et al. Growth hormone and insulinlike growth factor I enhance host defense in a murine sepsis model. Arch. Surg. 1995;130:1115±22. Gianotti L, Pyles T, Alexander JW, Babcock GF, Carey MA. Impact of blood transfusion and burn injury on microbial translocation and bacterial survival. Transfusion 1992;32:312±7. Fukushima R, Gianotti L, Alexander JW, Pyles T. The degree of bacterial translocation is a determinant factor for mortality after burn injury and is improved by prostaglandin analogs. Ann. Surg. 1992;216:438±44 discussion 444-5. Stieritz DD, Holder IA. Experimental studies of the pathogenesis of infections due to Pseudomonas aeruginosa: description of a burned mouse model. J. Infect. Dis. 1975;131:688±91.
[20] Alexander JW, Gianotti L, Pyles T, Carey MA, Babcock GF. Distribution and survival of Escherichia coli translocating from the intestine after thermal injury. Ann. Surg. 1991;213:558±66 discussion 566-7. [21] Vara-Thorbeck R, Guerrero JA, Rosell J, Ruiz RE, Capitan JM. Exogenous growth hormone: eects on the catabolic response to surgically produced acute stress and on postoperative immune function. World J. Surg. 1993;17:530±8. [22] Inoue T, Saito H, Hashiguchi Y et al. Growth hormone and insulin-like growth factor I augment Escherichia coli-killing activity of murine peritoneal exudative cells. Shock 1996;6:345±50. [23] Fu YK, Arkins S, Li YM, Dantzer R, Kelley KW. Reduction in superoxide anion secretion and bactericidal activity of neutrophils from aged rats: reversal by the combination of gamma interferon and growth hormone. Infect. Immun. 1994;62:1±8. [24] Zhang W, Frankel WL, Adamson WT et al. Insulin-like growth factor-I improves mucosal structure and function in transplanted rat small intestine. Transplantation 1995;59:755±61. [25] Chen K, Okuma T, Okamura K, Tabira Y, Kaneko H, Miyauchi Y. Insulin-like growth factor-I prevents gut atrophy and maintains intestinal integrity in septic rats. J. Parenter. Enteral. Nutr. 1995;19:119±24. [26] Lund PK, Ulshen MH, Rountree DB, Selub SE, Buchan AM. Molecular biology of gastrointestinal peptides and growth factors: relevance to intestinal adaptation. Digestion 1990;2:66±73. [27] Wells C, Erlandsen S. Localization of translocating Escherichia coli, Proteus mirabilis and Enterococcus faecalis within cecal and colonic tissues of monoassociated mice. Infect. Immun. 1991;59:4693±7. [28] McCauley RD, Heel KA, Christiansen KJ, Hall JC. The eect of minimum luminal nutrition on bacterial translocation and atrophy of the jejunum during parenteral nutrition. J. Gastroenterol. Hepatol. 1996;11:65±70.