0013-7227/97/$03.00/0 Endocrinology Copyright © 1997 by The Endocrine Society
Vol. 138, No. 1 Printed in U.S.A.
Contrasting Effects of Growth Hormone and Insulin-Like Growth Factor I on the Biological Activities of Endotoxin in the Rat* WEI LIAO, MATS RUDLING,
AND
BO ANGELIN
Metabolism Unit, Center for Metabolism and Endocrinology, Department of Medicine, and Molecular Nutrition Unit, Center for Nutrition and Toxicology, NOVUM, Karolinska Institute at Huddinge University Hospital, S-141 57 Huddinge, Sweden ABSTRACT We previously demonstrated that GH potentiates the biological activities of endotoxin in the rat. In the present study, we wanted to determine if the potentiating effects of GH on the biological activities of endotoxin could be reproduced by insulin-like growth factor I (IGFI). Endotoxin (5 mg/kg BW) was injected in rats primed with or without GH or IGF-I for 3 days. As expected, endotoxin administration markedly increased circulating tumor necrosis factor (TNF) and interferon-g (IFNg) and induced organ injury, hypoglycemia, and hyperlipidemia. In GH-primed rats, endotoxin induced a further increase of serum IFNg (but not TNF); and five out of six of those rats
died within 15 h after giving endotoxin. However, little difference between endotoxin-treated rats with and without IGF-I priming could be seen. Furthermore, IGF-I infusion altered blood glucose, urea, and circulating IGF-I levels more than GH infusion. Therefore, IGF-I does not enhance the biological activities of endotoxin in the rat, suggesting that the enhancement of endotoxin effects by GH is via an IGF-Iindependent pathway. Priming rats by GH (but not by IGF-I) induced a further increased response of serum IFNg but not TNF to subsequent endotoxin challenge, suggesting that IFNg rather than TNF is likely to be involved in this process. (Endocrinology 138: 289 –295, 1997)
G
ized by fever, hypotension, disseminated intravascular coagulation, multiple organ failure, and disturbances in the metabolism of carbohydrates and lipoproteins. It is believed that the in vivo biological activities of endotoxin are largely mediated by the induced secretion of inflammatory cytokines, i.e. tumor necrosis factor (TNF), interleukins 1 and 6, and interferon gamma (IFNg). Thus, it is reasonable to speculate that the observed enhancement of the biological activities of endotoxin induced by GH in the rat may be mediated by IGF-I. However, in the present study we showed that, in a striking contrast to GH, IGF-I infusion does not enhance the biological activities of endotoxin in the rat, suggesting that the enhancement of endotoxin effects by GH is via an IGF-I-independent pathway.
H IS AN anabolic hormone, sparing protein stores at the expense of fat in situations of caloric restriction. GH stimulates the synthesis and secretion of insulin-like growth factor I (IGF-I) from various tissues (1), and many GH effects are mediated by IGF-I (2). In addition to its use for the treatment of short-statured children with impaired production or complete lack of GH and for GH deficiency in adults, GH has been used to improve protein metabolism in critical illness. GH has several immunomodulatory effects. GH can prime phagocytes for an enhanced production of reactive oxygen intermediates to subsequent microbial challenges and increase lymphocyte activities (3–7). Thus, GH is required for the normal development and maintenance of important components of the immune system. Because of its immunomodulatory effects, GH may influence the response to microbial challenges. We recently demonstrated that treatment of rats with GH increases the sensitivity to endotoxin in term of endotoxin-induced organ injury and endotoxin-induced disturbances in the metabolism of carbohydrates and lipoproteins (8). Endotoxins are a class of lipopolysaccharide molecules derived from the cell wall of gram-negative bacteria. Endotoxins mediate the gram-negative septic shock syndrome, which is character-
Materials and Methods Materials Endotoxin from Escherichia coli O55B5 was purchased from Sigma Chemical Co. (L-2880, St. Louis, MO). Recombinant human GH (Genotropin, 3 IU/mg; batch no. 62526 –51) and recombinant human IGF-I (Igef, 2 mg/ml; batch no. 59114A51) were kindly provided by Dr. Anna Skottner (Pharmacia, Stockholm, Sweden). Osmotic minipumps (ALZET®, Model 2001 and 2ML1, which deliver solutions at a rate of 1 and 10 ml/h, respectively) were purchased from ALZA corporation (Palo Alto, CA).
Received May 29, 1996. Address all correspondence and requests for reprints to: Wei Liao, M.D., Ph.D., Center for Nutrition and Toxicology, NOVUM, Karolinska Institute, S-141 57 Huddinge, Sweden. E-mail:
[email protected]. * This study was supported by grants from the Karolinska Institute, the Swedish Medical Council (03X-7137), and the Swedish Society of Medicine (565.0 and 619.0), from the Thuring, Widengren, Jeansson, and Lundstro¨m Foundations, and from the Ruth and Richard Julin, the Female Old Servants, the Axelsson Johnson, and the Nordic Insulin Funds.
Animals and experimental procedure Male Sprague-Dawley rats (;7 weeks old) were used. They were maintained under standardized conditions with free access to chow and water and allowed to adapt to the environment for at least 1 week before starting the experiment. The light cycle hours were between 0600 h and 1800 h. Seventy-one rats were used in three separate experiments. In each
289
290
GH, IGF-I, AND ENDOTOXIN
Endo • 1997 Vol 138 • No 1
The second experiment was designed exactly the same as described above for the first experiment with the exception that, instead of GH, IGF-I was infused by osmotic minipumps (model 2ML1). IGF-I was infused at a rate of 23 mg/ratzh (2.1 mg/kgzday) for 4 days. The IGF-I dose was chosen, as this dose normalizes plasma IGF-I levels in hypophysectomized rats (14). In the third experiment, rats were infused with IGF-I and then injected with endotoxin as described above. The experiment was terminated 14 h after endotoxin injection by sampling blood into EDTA-containing tubes (Becton Dickinson Vacutainer Systems Europe, B.P. 37-38241, Meylan Cedex, France) via puncture of the abdominal aorta under light ether anesthesia. Plasma was then separated for the assays of renal and liver function, glucose and lipoproteins, etc.
Assays
FIG. 1. Effect of GH priming (A) and IGF-I priming (B) on endotoxininduced cytokine response and lethality. Endotoxin (LPS) was injected to rats with and without GH priming (A), or to rats with and without IGF-I priming (B). Blood was taken 1.5 h after endotoxin injection for assays of TNF and 4.5 h after endotoxin injection for assays of IFNg, glucose, and urea. Values for the assays are mean 6 SEM of six animals with the exception of the control group in the IGF-I priming experiment (n 5 5). ¶, P , 0.025 vs. control; †, P , 0.002 vs. control; ‡, P , 0.001 vs. control; *, P , 0.01 vs. LPS; §, P , 0.005 vs. LPS. experiment, four groups of animals were used, and each group consisted of six animals if not otherwise stated. The study was approved by the institutional Animal Care and Use Committee. In the first experiment, two groups of animals received GH by constant infusion from osmotic minipumps (model 2001) implanted sc under ether anesthesia at the start of the experiment (1500 h on day 0); the other two groups of animals were sham-operated. GH was infused at a rate of 8 mg/ratzh (0.75 mg/kgzday) for 4 days. The GH dose was chosen, as this dose restores the induction of hepatic LDL receptors by estrogen in hypophysectomized rats (9). Our previous studies have also shown that this dose of GH enhances endotoxin activity in the rat (8). At approximately 1900 h on day 3, endotoxin (5 mg/kg BW, using 0.15 m NaCl as vehicle) or the same volume of 0.15 m NaCl was injected ip to GH-primed rats and to sham-operated rats. Food was withdrawn at the time when endotoxin was injected as endotoxin induces anorexia. One and a half hours and 4.5 h after endotoxin injection, blood was taken from the retroorbital plexus by using a capillary tube under light ether anesthesia. Each sampling was approximately 1 ml. Sera were separated for the assays. Samples taken 1.5 h after endotoxin injection were assayed for TNF; and samples taken 4.5 h after endotoxin injection were assayed for IFNg, glucose and urea. The rationale for this was that TNF and IFNg peak around 1.5 h (10 –12) and 4.5 h (13), respectively, after endotoxin challenge. Twenty-four hours after endotoxin injection, rats were again given food ad libitum. Animals were observed for a period of 1 week.
Serum TNF was measured using the mouse ELISA kit (FactorTest-X TNF, Genzyme corporation, Cambridge, MA) according to the instructions by the manufacturer. The kit has full cross-reactivity with rat TNF (determined by Genzyme corporation, Cambridge, MA). The samples were diluted appropriately and assayed in duplicate together with the provided standards. The absorbances were measured at 450 nm by Bio-Rad Microplate Reader (model 450), and the data were analyzed by using Microplate Manager 2.1 software (Bio-Rad Laboratories, Richmond, CA). The intraassay and interassay coefficients of variation of the kit were 3.4 and 7.1%, respectively. The recovery was 90%. The sensitivity was 15 pg/ml. Linear regression analysis of the standard concentration ranging between 35 and 2240 pg/ml resulted in a correlation coefficient larger than 0.995. Serum IFNg were measured using the mouse ELISA kit (InterTest-g IFNg, Genzyme Corp., Cambridge, MA) according to the instructions by the manufacturer. The kit is well cross-reactive with rat IFNg (15). The samples were diluted appropriately and assayed in duplicate together with the provided standards. The absorbances were measured, and the data were analyzed as described above. The intraassay and interassay coefficients of variation of the kit were 4.2 and 6.4%, respectively. The recovery was 95%. The sensitivity was 5 pg/ml. Linear regression analysis of the standard concentration ranging between 20 and 1620 pg/ml resulted in a correlation coefficient larger than 0.997. Plasma IGF-I was measured by RIA using a polyclonal rabbit antihuman IGF-I antiserum. The labeled IGF-I used as tracer is a truncated variant of IGF-I, des(1–3)rhIGHF-I. In this variant, three amino acids are missing from the amino terminal, resulting in a lower affinity for the binding proteins. After acid ethanol extraction of the samples (16) and subsequent overnight incubation with the antibodies and tracer at room temperature, the immune complex was precipitated with a second antibody (polyclonal antirabbit IgG antibody) in the presence of polyethylene glycol. After centrifugation, the pellet was counted in a g counter, using recombinant human IGF-I as standard. The coefficients of variation of the intraassay and interassay were 3.1 and 10%, respectively. The recovery of the assay was 98.5%. The assay has negligible cross-reactivity against proinsulin, insulin, and IGF-II. There is high cross-reactivity between human and rat IGF-I. Plasma urea, creatinine, alanine-amino transferase (ALT), aspartateamino transferase (AST), gamma-glutamyl transferase (g-GT), lactate dehydrogenase (LDH), amylase, albumin, bilirubin, lactate, glucose, total cholesterol, and triglycerides were determined individually by clinical routine techniques as described earlier (8). Analysis of plasma lipoprotein profile was performed by fast protein liquid chromatography (17, 18). Equal volumes of plasma from every rat in each group were pooled (5 ml), and the density was adjusted to 1.21 kg/liter with solid potassium bromide. After ultracentrifugation at 100 3 103 g for 48 h, the supernatant (lipoprotein fraction) was adjusted to 2 ml by addition of the elution solution (0.15 m NaCl, 0.27 mm EDTA, 3 mm sodium azide, pH 7.3). After filtration through a 0.45 mm filter, 1 ml (corresponding to 2.5 ml plasma) was injected onto a 54 3 1.8 cm Superose 6B column; 2 ml fractions were collected at a flow rate of 1 ml/min. Fractions were assayed for total cholesterol and triglycerides.
Statistics Data are presented as means 6 sem and analyzed by using Statistica software (StatSoft, Tulsa, OK). One-way ANOVA was used to evaluate
GH, IGF-I, AND ENDOTOXIN
291
TABLE 1. Effect of IGF-I on biological activities of endotoxin
Body weight (g) A B Urea (mM) Creatinine (mM) ALT (mkat/liter) AST (mkat/liter) g-GT (mkat/liter) LDH (mkat/liter) Amylase (mkat/liter) Albumin (g/liter) Bilirubin (mM) Lactate (mM) Glucose (mM)
Control
IGF-I
LPS
IGF-I1LPS
265 6 5 289 6 6 4.7 6 0.2 43 6 0.8 0.99 6 0.04 1.17 6 0.03 0.017 6 0.002 2.7 6 0.3 197 6 6 15.7 6 0.4 2.3 6 0.4 5.0 6 0.2 9.9 6 0.3
269 6 4 300 6 5 3.0 6 0.2 39 6 0.7 0.86 6 0.04 1.15 6 0.03 0.02 6 0.004 2.9 6 0.4 135 6 9b 12.1 6 0.3b 3.3 6 0.4 4.2 6 0.2 6.3 6 0.5b
266 6 4 290 6 5 17.3 6 3.0a 63 6 8.5c 1.48 6 0.15 2.29 6 0.18 0.082 6 0.019d 4.2 6 0.9 59 6 3b 13.6 6 0.4e 5.8 6 0.8a 6.4 6 0.4e 7.1 6 0.8e
267 6 3 301 6 5 24.7 6 3.5b 59 6 6.4 1.83 6 0.59 3.16 6 0.95e 0.102 6 0.019a 10.0 6 4.9 52 6 4b 11.3 6 0.6b,f 5.3 6 0.3e 6.3 6 0.5 5.7 6 0.6a
Endotoxin (LPS) was injected to the rats with and without IGF-I infusion. Fourteen hours after injection of endotoxin, blood was taken and plasma was separated for the assays indicated. Body weight (mean 6 SEM, n 5 6) was measured the day before IGF-I infusion (A) and at the third day before injection of endotoxin (B). Values for the assays are mean 6 SEM of six animals with the exception of the LPS group (n 5 5) and the IGF-I1LPS group (n 5 4), because after endotoxin injection before sampling one in the LPS group and two rats in the IGF-I 1 LPS group died. a P , 0.002 vs. control. b P , 0.001 vs. control. c P , 0.05 vs. control. d P , 0.01 vs. control. e P , 0.02 vs. control. f P , 0.02 vs. LPS. the presence of significant differences between groups, followed by post-hoc comparisons of the group means according to the method of Tukey. When appropriate, post-hoc comparisons were adjusted for unequal sample sizes by Spjøtvoll and Stoline. Student’s t test was used to evaluate the significances of difference in serum cytokines where only two groups (i.e. endotoxin-treated groups primed with and without GH in the first experiment) were compared.
Results
In the first experiment, we addressed the question if GH infusion would increase the endotoxin-induced cytokine response. Endotoxin was injected to rats with and without GH priming. One and a half hours after endotoxin injection, blood was taken and serum was separated for TNF assay. As shown in Fig. 1A, controls and animals treated only with GH had no detectable serum TNF. Serum TNF increased markedly after endotoxin injection, reaching levels of approximately 120 ng/ml. Serum TNF in GH-primed rats treated with endotoxin was not higher than that in the animals treated only with endotoxin. Four and a half hours after endotoxin injection, blood was taken and serum was separated for IFNg assay and also for glucose and urea assays. As for TNF, controls and animals treated only with GH had no detectable serum IFNg (Fig. 1A). Endotoxin injection markedly increased serum IFNg. Compared to animals given only endotoxin, serum IFNg was doubled in GH-primed rats treated with endotoxin. Endotoxin injection also reduced blood glucose and increased blood urea. However, GHprimed rats treated with endotoxin showed much more severe hypoglycemia and higher blood urea than the rats given only endotoxin. Animals receiving only GH did not show any visible abnormality. In endotoxin-treated animals, diarrhea occurred, and these animals were clearly less active than normal rats. After endotoxin injection, GH-primed rats became very sick. The general appearance of these rats was similar to what was observed in our previous study (8), i.e. they had severe diarrhea, ruffled furs, and were lethargic. No
animals died among the controls or those treated only with GH; one animal died in the endotoxin-treated group, whereas five of six animals died in GH-primed rats treated with endotoxin (Fig. 1A). All deaths occurred between 5 and 15 h after endotoxin injection. Twenty-four to 48 h after endotoxin injection, the surviving animals recovered. In the second experiment, we addressed the question if IGF-I had an effect similar to GH by using the same experimental design. One and a half hours and 4.5 h after endotoxin injection, blood was taken and serum was separated for assays of TNF and IFNg, respectively. Blood glucose and urea were also measured 4.5 h after endotoxin injection. As shown in Fig. 1B, controls and animals treated only with IGF-I had no detectable circulating TNF and IFNg. Serum TNF and IFNg increased markedly again after endotoxin injection. Serum TNF seemed to be slightly lower in IGF-Iprimed rats treated with endotoxin than in those given only endotoxin, whereas there was no difference in serum IFNg between endotoxin-treated rats with and without IGF-Ipriming. Clearly, IGF-I did not show any synergistic effect with endotoxin on blood glucose. Blood urea was not significantly higher in IGF-I-primed rats treated with endotoxin than in rats given only endotoxin. The general appearance of IGF-I-primed rats treated with endotoxin was not distinguishable from that observed in rats treated only with endotoxin. Animals receiving only IGF-I did not show any visible abnormality. Again, in endotoxin-treated animals, diarrhea occurred and these animals were clearly less active than normal rats. In this experiment, no animals died among the controls, those treated only with IGF-I or endotoxin. Only one of six animals died 24 h after endotoxin injection in the group of IGF-I-primed rats treated with endotoxin (Fig. 1B). In the third experiment, blood was taken 14 h after endotoxin injection to determine if IGF-I-primed rats had an increased organ injury and more severe disturbances in the metabolism of carbohydrates and lipoproteins. IGF-I infu-
292
Endo • 1997 Vol 138 • No 1
GH, IGF-I, AND ENDOTOXIN
sion reduced plasma glucose, urea, amylase, and albumin (Table 1) and slightly increased plasma total cholesterol and triglycerides (Fig. 2A), mainly due to an elevation within the very low density lipoprotein fraction (Fig. 2, B and C). Endotoxin had clear injurious effects, which is consistent with our previous study (8). Thus, plasma levels of urea, creatinine, ALT, AST, g-GT, bilirubin, and lactate were increased after endotoxin injection (Table 1). Endotoxin decreased plasma glucose, amylase and albumin (Table 1). Endotoxin increased plasma levels of total cholesterol and triglycerides (Fig. 2A). The endotoxin-induced increase in plasma cholesterol and triglycerides occurred within the apoB-containing lipoproteins, i.e. very low, intermediate, and low density lipoproteins (Fig. 2, B and C). The injurious effects of endotoxin did not appear to be more pronounced in IGF-I-primed rats. Compared to animals treated only with endotoxin, there
was no deterioration in IGF-I-primed rats treated with endotoxin as judged by plasma urea, creatinine, amylase, bilirubin, and lactate. ALT, AST, g-GT, and LDH also were not significantly higher in IGF-I-primed rats treated with endotoxin than in rats treated only with endotoxin (Table 1). IGF-I did not show any synergistic effect with endotoxin on blood glucose. IGF-I-primed rats treated with endotoxin did not develop more pronounced hyperlipidemia than did the animals treated only with endotoxin (Fig. 2). Clearly, in a striking contrast to GH, IGF-I-primed rats did not respond to endotoxin challenge with more severe organ injury and disturbances in the metabolism of plasma glucose and lipoproteins. However, one may ask if the inability of IGF-I to enhance the biological activities of endotoxin was due to an inappropriate dose or a lack of biological activity of IGF-I used. Therefore, we compared plasma levels of IGF-I and indicators of biological function between GH and IGF-I experiments. One such indicator is blood glucose because the reduced levels of blood glucose reflect the hypoglycemic effect of IGF-I. Another indicator is blood urea because the anabolic action of GH and IGF-I involves redistribution of nitrogen from ureagenesis to the extrahepatic anabolic process, and a high anabolic activity should reduce blood urea. The plasma samples for IGF-I assay after GH infusion were obtained from the experiment in which we previously showed that GH markedly enhances the in vivo biological activities of endotoxin (8). After the infusion of GH, plasma IGF-I increased by approximately 20% (Fig. 3A), together with a 10% reduction in plasma glucose but no reduction in plasma urea (see Ref. 8). On the other hand, IGF-I infusion almost doubled the circulating IGF-I levels (Fig. 3B), together with a 36% reduction in both plasma urea and glucose (Table 1). Similar results comparing the effects of GH and IGF-I on blood glucose and urea are shown in Fig. 1. The infusion of GH reduced blood glucose by 13% and had no effect on blood urea (Fig. 1A), whereas the infusion of IGF-I reduced blood glucose by 19% and reduced blood urea by 30% (Fig. 1B). Therefore, IGF-I infusion used in the present study resulted in markedly increased plasma levels of apparently biologically active IGF-I but did not enhance the in vivo biological activities of endotoxin. Furthermore, plasma IGF-I was reduced markedly 14 h after endotoxin injection (Fig. 3), as described by others (19). Clearly, GH treatment did not reverse the endotoxin-induced reduction of plasma IGF-I (Fig. 3A), probably reflecting the resistance to GH in endotoxin shock. Whether IGF-I had any effect on the endotoxin-induced reduction of plasma IGF-I could not be addressed because the assay does not distinguish endogenous from exogenous IGF-I.
FIG. 2. Effect of IGF-I priming on endotoxin-induced hyperlipidemia. Plasma samples were obtained from the experiment described in Table 1. Plasma total cholesterol and triglycerides were measured individually (A). Analysis of lipoprotein profile by fast protein liquid chromatography was performed using pooled plasma of each group (B and C). Values for the assays of plasma total cholesterol and triglycerides are mean 6 SEM of six animals with the exception of the LPS group (n 5 5) and the IGF-I 1 LPS group (n 5 4). HDL, high density lipoprotein; IDL, intermediate density lipoprotein; LDL, low density lipoprotein; VLDL, very low density lipoprotein. †, P , 0.02 vs. control; ‡, P , 0.001 vs. control;
Discussion
In the present study, we have demonstrated that the potentiating effects of GH on the biological activities of endotoxin in the rat, which leads to an increased lethality, cannot be reproduced by IGF-I infusion. While markedly increasing the circulating IGF-I levels and exerting clear biological effects, IGF-I infusion did not potentiate the endotoxin-induced biological activities, as observed in endotoxin-induced death, organ injuries and disturbances in the metabolism of
GH, IGF-I, AND ENDOTOXIN
293
FIG. 3. A, Plasma IGF-I levels after GH infusion and endotoxin injection. Plasma samples were obtained from our previously published study (8). Groups of animals were infused with GH at a rate of 8 mg/ratzh for 4 days. At approximately 2100 h on day 3, food was withdrawn, and endotoxin (LPS, 5 mg/kg BW, using 0.15 M NaCl as vehicle) was injected ip to rats with and without GH infusion. Fourteen hours after injection of endotoxin, blood was taken and plasma separated for assay of IGF-I. Values for the assays are mean 6 SEM of six animals with the exception of the GH 1 LPS group (n 5 5), owing to the death of one animal after endotoxin injection before sampling. B, Plasma IGF-I levels after IGF-I infusion and endotoxin injection. Plasma samples were obtained from the experiment described in Table 1. Values for the assays are mean 6 SEM of six animals with the exception the LPS group (n 5 5) and the IGF-I 1 LPS group (n 5 4). †, P , 0.05 vs. control; *, P , 0.001 vs. control.
carbohydrates and lipoproteins. Accordingly, other studies have shown that GH primes phagocytes for an increased production of reactive oxygen intermediates (4, 5) and stimulates hepatic LDL receptor expression via pathways not involving IGF-I (14, 20), even though IGF-I may mimic the effects of GH in some of these situations (4, 20). Our results thus suggest that the potentiating effects of GH on the biological activities of endotoxin is independent of IGF-I. Priming rats by GH induced an increased response of IFNg but not TNF to subsequent endotoxin challenge, suggesting that IFNg rather than TNF is likely to be involved in this process. IGF-I-primed rats subsequently treated with endotoxin did not show any further increase in serum TNF and IFNg, suggesting that, in vivo, IGF-I does not prime macrophages and lymphocytes for an increased synthesis and release of inflammatory cytokines. It is believed that TNF, interleukins 1 and 6, and IFNg play important roles in the mediation of the in vivo biological activities of endotoxin. TNF administration to animals mimics endotoxic shock (21–23); and passive immunization against TNF prevents the lethal effect of endotoxin and gramnegative bacteremia in animals (24, 25). TNF also induces the synthesis and secretion of other important inflammatory cytokines, such as interleukins 1 and 6 (26), thus establishing the central role of TNF in the mediation of the biological activities of endotoxin. However, we found that GH infusion to rats increased endotoxin-induced lethality without an increased TNF response to endotoxin, suggesting that the enhancement of the in vivo biological activities of endotoxin by GH is not likely to be due to GH priming of macrophages for an increased synthesis and secretion of TNF. It has been shown that in vitro treatment of human monocytes with GH had either no effect or inhibited the endotoxin-induced pro-
duction of TNF and interleukin 1 (27, 28). In vivo, GH pretreatment blunts the plasma increase in TNF and interleukins 1 and 6 in response to subsequent challenge by endotoxin in the calf (29) or by Escherichia coli bacteria in the mouse (30). Taken together, these studies support the notion that GH treatment does not enhance the responses of TNF (and interleukins 1 and 6) to subsequent challenge with inflammatory stimuli. However, another study showed that in vivo treatment of hypophysectomized rats with GH induces an enhanced synthesis of TNF by macrophages in response to subsequent endotoxin challenge in vitro (31). This may suggest that GH-treated hypophysectomized rats are different from GH-treated normal rats in response to endotoxin challenge, or alternatively, the macrophages isolated from GHtreated rats respond to endotoxin challenge differently from the in vivo setting. The lymphocyte-derived cytokine, IFNg, also plays important role in the mediation of endotoxin effects in vivo. Anti-IFNg antibodies can prevent the endotoxin-induced Shwartzman reaction (32) and reduce the lethality from endotoxic shock (33) and gram-negative bacteria (34). IFNg is synergistic with TNF in inducing lethality in animals (35). Hence, endogenous IFNg plays an important role in the pathogenesis of events leading to endotoxic shock. In the present study, we showed that GH-primed rats treated with endotoxin had a further increase in serum IFNg and an increased lethality, suggesting an important role of IFNg as a mediator of the GH-induced potentiation of endotoxin effects. In contrast to our study, GH adminstration to hypophysectomized rats actually enhances the resistance to experimental Salmonella typhimurium infection (36), presumably by priming macrophages for an increased production of reac-
294
GH, IGF-I, AND ENDOTOXIN
tive oxygen intermediates (37) and TNF (31). GH administration to mice also reduces the death from subsequent challenge with Escherichia coli bacteria without increasing plasma cytokines (TNF, and interleukins 1 and 6) (30). The differences in choosing the animal models, the inflammatory stimuli, the GH administration modes (infusion or bolus injections), and the doses and periods of GH exposure may be in part accounted for the differences among these studies. Thus, GH therapy in critical patients with infection and endotoxemia could have either beneficial or detrimental consequences. Although it may presently only be speculated on, our results strengthen the notion that caution should be exerted when considering GH treatment in patients with catabolic conditions complicated with endotoxemia and infections (8), and may imply that IGF-I treatment could be preferable in such situations. There exist marked species differences in the sensitivity to endotoxin. Rats and mice are rather resistant to endotoxin, as compared with primates (including humans), sheep, pigs, and rabbits (38). Endotoxin administration to rats results in a dramatic (by 80 –90% after 1– 4 h) but transient (normalized after 24 h) reduction of plasma GH levels (19), whereas it increases plasma GH levels in humans (39) and sheep (40). TNF inhibits basal and GH-releasing hormone-stimulated GH secretion in vitro from cultured rat anterior pituitary cells (41). TNF reduces plasma levels GH and IGF-I in the rat (42). However, anti-TNF antibody attenuates endotoxin-induced reduction of IGF-I but not GH (42). IFNg also inhibits GH secretion in vitro from cultured rat anterior pituitary cells (43). Thus, TNF and IFNg may play important roles in mediation of endotoxin-induced reduction of plasma GH. The reduction of plasma GH may attenuate the endotoxic effects and thus provide an important protective mechanism against endotoxin in the rat. In summary, GH enhances endotoxin effects in the rat, thereby increasing the endotoxin-induced lethality. IGF-I is not likely to be the mediator because IGF-I does not enhance the biological activities of endotoxin. GH-primed rats had markedly increased response in serum IFNg but not TNF to endotoxin challenge, suggesting that priming lymphocytes induced by GH for an enhanced synthesis and secretion of IFNg to subsequent endotoxin challenge is likely to be involved in this potentiating effect. Further studies are needed to fully establish the critical importance of IFNg in this situation.
5. 6.
7. 8. 9. 10. 11. 12.
13. 14. 15. 16.
17. 18. 19. 20. 21. 22.
23.
24.
Acknowledgments We thank Dr. Anna Skottner for providing GH and IGF-I, Dr. Paolo Parini for help with statistical analyses, Mr. Roland Eklo¨f for arranging the routine assays, and Ms. Vibeke Ta¨pp for measuring IGF-I.
References 1. D’Ercole AJ, Stiles AD, Underwood LE 1984 Tissue concentrations of somatomedin C: further evidence for multiple sites of synthesis and paracrine or autocrine mechanism of action. Proc Natl Acad Sci USA 81:935–939 2. Chwals WJ, Bistrian BR 1991 Role of exogenous growth hormone and insulinlike growth factor I in malnutrition and acute metabolic stress: a hypothesis. Crit Care Med 19:1317–1322 3. Edwards CK, Ghiasuddin SM, Schepper JM, Yunger LM, Kelley KW 1988 A newly defined property of somatotropin: priming of macrophages for production of superoxide anion. Science 239:769 –771 4. Fu YK, Arkins S, Wang BS, Kelley KW 1991 A novel role of growth hormone
25. 26. 27.
28. 29. 30.
Endo • 1997 Vol 138 • No 1
and insulin-like growth factor-I. Priming neutrophils for superoxide anion secretion. J Immunol 146:1602–1608 Warwick-Davies J, Lowrie DB, Cole PJ 1995 Growth hormone is a human macrophage activating factor. Priming of human monocytes for enhanced release of H2O2. J Immunol 154:1909 –1918 Fu YK, Arkins S, Li YM, Dantzer R, Kelley KW 1994 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 62:1– 8 Kelley KW 1990 The role of growth hormone in modulation of the immune response. Ann NY Acad Sci 594:95–103 Liao W, Rudling M, Angelin B 1996 Growth hormone potentiates the in vivo biological activities of endotoxin in the rat. Eur J Clin Invest 26:254 –258 Rudling M, Norstedt G, Olivecrona H, Reihne´r E, Gustafsson JÅ, Angelin B 1992 Importance of growth hormone for the induction of hepatic low density lipoprotein receptors. Proc Natl Acad Sci USA 89:6983– 6987 Beutler BA, Milsark IW, Cerami A 1985 Cachectin/tumor necrosis factor: production, distribution, and metabolic fate in vivo. J Immunol 135:3972–3977 Michie HR, Manogue KR, Spriggs DR, Revhaug A, O’Dwyer S, Dinarello CA, Cerami A, Wolff SM, Wilmore DW 1988 Detection of circulating tumor necrosis factor after endotoxin administration. N Engl J Med 318:1481–1486 Cannon JG, Tompkins RG, Gelfland GA, Michie HR, Stanford GG, van der Meer JWM, Endres S, Lonnemann G, Corsetti J, Chernow B, Wilmore DW, Wolff SM, Burke JF, Dinarello CA 1990 Circulating interleukin-1 and tumor necrosis factor in septic shock and experimental endotoxin fever. J Infect Dis 161:79 – 84 Kiener PA, Marek F, Rodgers G, Lin PF, Warr G, Desiderio J 1988 Induction of tumor necrosis factor, IFN-g, and acute lethality in mice by toxic and non-toxic forms of lipid A. J Immunol 141:870 – 874 Rudling M, Olivecrona H, Eggertsen G, Angelin B 1996 Regulation of rat hepatic low density lipoprotein receptors. In vivo stimulation by growth hormone is not mediated by insulin-like growth factor I. J Clin Invest 97:292–299 Teti G, Mancuso G, Tomasello F 1993 Cytokine appearance and effects of anti-tumor necrosis factor alpha antibodies in a neonatal rat model of group B streptococcal infection. Infect Immun 61:227–235 Daughaday WH, Mariz IK, Blethen SL 1980 Inhibition of access of bound somatomedin to membrane receptor and immunobinding sites: a comparison of radioreceptor and radioimmunoassay of somatomedin in native and acidethanol-extracted serum. J Clin Endocrinol Metab 51:781–788 Ha YC, Barter PJ 1985 Rapid separation of plasma lipoproteins by gel permeation chromatography on agarose gel Superose 6B. J Chromatogr 341:154 –159 Rudling M 1992 Hepatic mRNA levels for the LDL receptor and HMG-CoA reductase show coordinate regulation in vivo. J Lipid Res 33:493–501 Fan J, Molina PE, Gelato MC, Lang CH 1994 Differential tissue regulation of insulin-like growth factor-I content and binding proteins after endotoxin. Endocrinology 134:1685–1692 Parini P, Angelin B, Lobie PE, Norstedt G, Rudling M 1995 Growth hormone specifically stimulates the expression of low density lipoprotein receptors in human hepatoma cells. Endocrinology 136:3767–3773 Tracey KJ, Beutler B, Lowry SF, Merryweather J, Wolpe S, Milsark IW, Hariri RJ, Fahey III TJ, Zentella A, Albert JD, Shires GT, Cerami A 1986 Shock and tissue injury induced by recombinant human cachectin. Science 234:470 – 474 Tracey KJ, Lowry SF, Fahey III TJ, Albert JD, Fong Y, Hesse D, Beutler B, Manogue KR, Calvano S, Wei H, Cerami A, Shires GT 1987 Cachectin/tumor necrosis factor induces lethal shock and stress hormone response in the dog. Surg Gynecol Obstet 164:415– 422 Natanson C, Eichenholz PW, Danner RL, Eichhacker PQ, Hoffman WD, Kuo GC, Banks SM, MacVittie TJ, Parrillo JE 1989 Endotoxin and tumor necrosis factor challenges in dogs stimulate the cardiovascular profile of human septic shock. J Exp Med 169:823– 832 Tracey K, Fong Y, Hesse DG, Manogue KR, Lee AT, Kuo GC, Lowry SF, Cerami A 1987 Anti-cachectin/TNF monoclonal antibodies prevent septic shock during lethal bacteremia. Nature 330:662– 664 Beutler B, Milsark IW, Cerami A 1985 Passive immunization against cachectin/tumor necrosis factor protects mice from lethal effect of endotoxin. Science 229:869 – 871 Fony Y, Tracey KJ, Moldawer LL, Hesse DG, Manogue KR, Kenney JS 1989 Antibodies to cachectin/TNF reduce interleukin-1b and interleukin-6 appearance during lethal bacteremia. J Exp Med 170:1627–1633 Warwick-Davies J, Lowrie DB, Cole PJ 1995 Growth hormone activation of human monocytes for superoxide production but not tumor necrosis factor production, cell adherence, or action against Mycobacterium tuberculosis. Infect Immun 63:4312– 4316 Kappel M, Hansen MB, Diamant M, Pedersen BK 1994 In vitro effects of human growth hormone on the proliferative responses and cytokine production of blood mononuclear cells. Horm Metab Res 26:612– 614 Elsasser TH, Fayer R, Rumsey TS, Hartnell GF 1994 Recombinant bovine somatotropin blunts plasma tumor necrosis factor-a, cortisol, and thromboxane-B2 responses to endotoxin in vivo. Endocrinology 134:1082–1088 Inoue T, Saito H, Fukushima R, Inaba T, Lin MT, Fukatsu K, Muto T 1995
GH, IGF-I, AND ENDOTOXIN
31.
32. 33. 34. 35. 36. 37.
Growth hormone and insulin-like growth factor I enhance host defense in a murine sepsis model. Arch Surg 130:1115–1122 Edwards CK, Lorence RM, Dunham DM, Arkins S, Yunger LM, Greager JA, Walter RJ, Dantzer R, Kelley KW 1991 Hypophysectomy inhibits the synthesis of tumor necrosis factor a by rat macrophages: partial restoration by exogenous growth hormone or interferon g. Endocrinology 128:989 –996 Heremans H, Van Damme J, Dillen C, Dijkmans R, Billiau A 1990 Interferon g, a mediator of lethal lipopolysaccharide-induced Shwartzman-like shock reactions in mice. J Exp Med 171:1853–1869 Heinzel FP 1990 The role of IFN-g in the pathology of experimental endotoxemia. J Immunol 145:2920 –2924 Silva AT, Cohen J 1992 Role of interferon-g in experimental gram-negative sepsis. J Infect Dis 166:331–335 Doherty GM, Lange JR, Langstein HN, Alexander HR, Buresh CM, Norton JA 1992 Evidence for IFN-g as a mediator of the lethality of endotoxin and tumor necrosis factor-a. J Immunol 149:1666 –1670 Edwards CK, Yunger LM, Dantzer R, Kelley KW 1991 The pituitary gland is required for protection against lethal effects of Salmonella typhimurium. Proc Natl Acad Sci USA 88:2274 –2277 Edwards CK, Ghiasuddin SM, Yunger LM, Lorence RM, Arkins S, Dantzer R, Kelley KW 1992 In vivo administration of recombinant growth hormone or
38. 39.
40.
41. 42.
43.
295
gamma interferon activate macrophages: enhanced resistance to experimental of Salmonella typhimurium infection is correlated with generation of reactive oxygen intermediates. Infect Immun 60:2514 –2521 Redl H, Bahrami S, Schlag G, Traber DL 1993 Clinical detection of LPS and animal models of endotoxemia. Immunobiology 187:330 –345 Elin RJ, Wolff SM, McAdam KP, Chedid L, Audibert F, Bernard C, Oberling F 1981 Properties of reference Escherichia coli endotoxin and its phthalylated derivative in humans. J Infect Dis 144:329 –336 Coleman ES, Elsasser TH, Kemppainen RJ, Coleman DA, Sartin JL 1993 Effect of endotoxin on pituitary hormone secretion in sheep. Neuroendocrinology 58:111–122 Walton PE, Cronin MJ 1989 Tumor necrosis factor-a inhibits growth hormone secretion from anterior pituitary cells. Endocrinology 125:925–929 Fan J, Char D, Bagby GJ, Gelato MC, Lang CH 1995 Regulation of insulin-like growth factor-I (IGF-I) and IGF-binding proteins by tumor necrosis factor. Am J Physiol 38:R1204 –R1212 Vankelecom H, Carmeliet P, Heremans H, Van Damme J, Dijkmans R, Billiau A, Denef C 1990 Interferon-gamma inhibits stimulated adrenocorticotropin, prolactin, and growth hormone secretion in normal rat anterior pituitary cell cultures. Endocrinology 126:2919 –2926
