- Email: [email protected]
Chronic inflammation inhibits GH secretion and alters the serum insulin-like growth factor system in rats
Life Sciences, Vol 6S, NO. 20. pp. 2049-2060, 1999 Copyight 0 1999 Elsevier Science Inc. Printed in the USA. All tights reserved 0024-3205/99/lsee front matter ELSEVIER
PI1 SOO24-3205(99)00472-S
CHRONIC INFLAMMATION INHIBITS GH SECRETION AND ALTERS THE SERUM INSULIN-LIKE GROWTH FACTOR SYSTEM IN RATS A LOPEZ-CALDERGN, L SOTO, *Al MARTiN Dpt Fisiologia, Fat Medicina, Univ Complutense and *Dpt CC Morfologicas y Fisiologia, Univ Europea, Madrid, Spain. (Received in final form July 2. 1999)
Summary
Adjuvant-induced arthritis in rats is associated with growth failure, hypermetabolism and accelerated protein breakdown. The aim of this work was to study the effects of adjuvant-induced arthritis on GH and insulin-like growth factor-l (IGF-I). Arthritis was induced by an intradermal injection of complete Freund’s adjuvant and rats were killed 18 and 22 days later. IGF-I and GH levels were measured by radioimmunoassay. Pituitary GH mRNA was analyzed by northern blot and IGF binding proteins (IGFBPs) by western blot. Arthritic rats showed a decrease in both serum and hepatic concentrations of IGF-I. On the contrary, arthritis increased the circulating IGFBPs. The serum concentration of IGF-I in the arthritic rats was negatively correlated with the body weight loss observed in these animals. Arthritis decreased the serum concentration of GH and this decrease seems to be due to an inhibition of GH synthesis, since pituitary GH mRNA content was decreased in arthritic rats (~~0.01). These data suggest that the decrease in body weight gain in arthritic rats may be, at least in part, secondary to the decrease in GH and IGF-I secretion. Furthermore, the increased serum IGFBPs may also be involved in the disease process. Key Words: adjuvant-induced arthritis, inflammation, GH, GH mRNA, IGF-I, IGFE3Ps
Adjuvant-induced arthritis (AA) is an animal model used to study rheumatoid arthritis and chronic inflammatory stress. Chronic inflammation, in experimental animal models such as AA, is associated with a decrease in body weight and a loss of body cell mass (1). Similarly, rheumatoid arthritis (RA) patients have cachexia hypermetabolism and accelerated protein breakdown (2). When analyzing the nutritional status, arthritic rats have anorexia (3) but they lose more weight than pair fed controls (1, 4) indicating a catabolic condition, The increased production of cytokines during AA (1) could play an important role in the inhibition of growth,but it seems to be also mediated by disturbances of the endocrine system.
Corresponding author: A Lopez-Calderon. Dpt Fisiologia, Fat Medicina, Univ Complutense. 28040 Madrid. Spain. Tel: 3491-3941491; Fax: 3491-3941628; E-mail: ale @eucmax.sim.es
2050
GH and IGF-I in Chronic Infhmmation
Vol. 65, No. 20, 1999
It is generally accepted that AA is associated with an increased activity of the pituitaryadrenal axis, with an increase in circulating adrenocorticopic hormone (ACTH) and corticosterone as well as in the pituitary proopiomelanocortin (POMC) mRNA content (5). In addition to the modifications of the adrenal axis, a number of neuroendocrine changes has also been described in arthritic animals (5, 6). Regarding anabolic hormones, there is not a clear consensus about the effect of AA on growth hormone (GH) and insulin-like growth factor-l (IGF-I). In humans, patients with rheumatoid arthritis have low serum concentrations of IGF-I and IGF-II (7). Children with RA have shown impaired growth together with normal or low circulating GH levels (8, 9). An increase in prolactin release along with a decrease in circulating GH in AA rats were previously described during the early latency period before the onset of arthritis (6, 10). In contrast, an increased pituitary GH release has also been described in the same animal model when the rats have developed external signs of the disease (11, 12). The aim of this study was to analyze the effect of adjuvant-induced arthritis on the somatotropic axis. To address this question, we analyzed pituitary GH mRNA, serum and hepatic concentrations of IGF-I, and its binding proteins IGFBPs, since they play a crucial role in modulating IGF-I actions.
Methods
Male Sprague-Dawley rats were purchased from Charles River (Barcelona, Spain). The procedures followed the guidelines recommended by the EU for the care and use of laboratory animals. Arthritis was induced in the rats by a intradermal injection of complete Freund’s adjuvant (1 mg heat-inactivated Mycobacterium butyricum in incomplete Freund’s adjuvant, Difco laboratory) at the base of the tail. Control animals were injected with incomplete Freund’s adjuvant. Rats were housed 4-5 per cage under controlled conditions of light (light on from 7:30 to 19:30) and temperature (22 f 2OC). Food and water were available “ad libitum”. Both control and arthritic rats were killed by decapitation18 and 22 days after adjuvant or vehicle injection between 1 I:00 and 12:00 h. Trunk blood was collected in cooled tubes, allowed to clot, centrifuged and the serum was stored at -20 OC until GH, IGF-I, insulin and IGFBPs analysis were performed. Immediately after decapitation the pituitary gland was frozen in a dry ice/acetone mixture and stored at -80 OC until the northern blot analysis. The liver was removed, dissected, frozen and stored at -20 OC until IGF-I assay. The spleen and adrenals were also dissected and weighed. Assessment of arthritis The arthritis index of each animal was scored by grading each paw from 0 to 4, determined as: 0- no erythema or swelling. I- slight erythema or swelling of one or more digits. 2- entire paw swollen. 3- erythema and swelling of the ankle. 4- ankylosis, incapacity to bend the ankle. The severity score was the sum of the clinical scores of each limb, the maximum value being 16 (3). Hormone determination IGF-I concentrations were measured by a double-antibody RIA using the antibody NIDDK UB2-495 a gift from Drs Underwood and Van Wik, and it is distributed by the Hormone Distribution Program of NIDDK through the National Hormone and Pituitary Program. Serum IGF-I binding proteins (IGFBPs) were removed by an acid-ethanol extraction (13). Hepatic IGF-I was extracted as described by Torres-Aleman et a/. (14). Samples were homogenized
Vol. 65, No. 20, 1999
GH and IGF-I in Chronic Inllammation
2051
in 1 N acetic acid, boiled for 20 min, and lyophilized. To confirm the elimination of IGFBPs, extracted serum and liver fractions were incubated with ‘2?-IGF-I at 4OC for 24 hours. The separation of bound and free tracers was carried out with dextran charcoal. Extracted fractions showed 98-100 % of free IGF-I. Levels of IGF-I were expressed in terms of IGF-I A52-EPD-186 standard (Eli Lilly 8 Company). The intra-assay coefficient of variation was 8%; samples from one experiment were run in the same assay. Concentrations of GH were measured by a radioimmunoassay (RIA) using reagents provided by Dr. Parlow of the National Hormone and Pituitary Program. Levels of GH were expressed in terms of NIDDK rat-RP-2 standard. The GH detection level was 10 pg, and the intra-assay coefficient of variation was 3%. Ail necessary comparisons between test and control animals were made within the same assay. Insulin was measured by a radioimmunoassay previously described (15). Protein content was measured by Bradford’s method (16) and serum concentrations of glucose with a commercial kit from Boehringer Mannheim.
Western ligand b/of Western blots were prepared as previously described (17). Samples were subjected to 1% SDS-12.5 % acrylamide gels non reducing electrophoresis, and electrotransferred to nitrocellulose membranes (HybondTM -C extra, Amersham, UK). The nitroceiiulose sheets were dried and blocked for 1 h with 5% non-fat dry milk, 0.1 % Tween (Sigma), in Trisbuffered saline, and incubated overnight at 4OC with ‘251-labeled IGF-I (5 x ld cpmlml). The nitroceiiulose sheets were then washed, dried and blots were exposed at -80 OC to X-ray film (Kodak X-Omat AR, Eastman Kodak, Rochester NY) and two intensifying screens for l-5 days according to the signal obtained. Autoradiographs were analyzed by densitometric scanning using a PC-Image VGA24 program for Windows. The density of the IGFBP bands in each lane was expressed as the percentage of the mean density of control sera.
Northern blot of pituitary GH mRNA content The pituitaries were extracted in pairs, total RNA was extracted by the guanidine thiocyanate method using a commercial kit (UltraspecTM RNA, Biotecx Laboratories). Each 5 ug sample of total RNA was denatured, separated by formaldehyde-agarose gel eiectrophoresis, transferred to nylon membranes (Hybond-N+, Amersham, UK) and fixed by uv crosslinking (Fotodyne Hartland, WI, USA). Homogeneity of gel loading was confirmed by the intensity of the ribosomai 18s RNA bands in the transferred membranes stained with ethidium bromide. The rat GH cDNA probe was the Hindlll fragment of the pRGH-1 (18). Prehybridization was performed for 3 h at 42 OC in buffer, (50 % formamide, 5 X SSPE, 5 X Denhart’s reagent, 1% SDS, and 100 ug/mL salmon sperm DNA) followed by hybridization for 16 h at the same temperature with 1-3 x 10’ cpm/mL GH cDNA labeled by random priming (High Prime, Boheringer Mannheim) in the same buffer. The membranes were washed twice with 2 x SSPE, 0.1% SDS at room temperature for 5 min, twice at 42 OC with 2xSSPE, 0.1 % SDS for 30 min and twice with 0.1 x SSPE, 0.1 % SDS at 50 OC also for 30 min. Exposure to X-ray film was performed at -80 OC. The intensities of autoradiogram signal levels and bromide-stained ribosomal 18s RNA in the nylon filters were analyzed by densitometric scanning using a PC-Image VGA24 program for Windows, and were normalized for 18s ribosomal RNA levels. The total GH mRNA content per entire pituitary was then calculated by multiplying the northern hybridization densitometric units of the densitometric analysis by the total amount of the pituitary RNA in micrograms.
Vol. 65, No. 20, 1999
GH and IGF-I in Chronic Inflammation
Statistical analysis Statistical significance was calculated by Duncan Multiple Range test after one-way analysis of variance, and Student’s t test when comparing only two means. Because there were no significant differences between data obtained in the control rats killed 18 or 22 days after vehicle injection, they were used as one control group. Correlation between different variables was calculated by linear regression. Serum GH data were subjected to log transformation since variances showed a log-normal distribution. A p-value less than 0.05 was considered significant.
Results Rats developed signs of inflammation 12-14 days after adjuvant injection, reaching the maximum level on day 21 (Fig 1). Body weight gain in the arthritic rats was not different from controls until day IO when arthritic rats started to gain less body weight than controls until the end of the study (Fig 1).
340 320 300
-8- Control 280 - -o- A r t h r i t i s
16
OI 260
14
E
.% 240 3 &T 220
12
m” 200 180 160 140 120
I 2
. .
l
.
I
I
1
4
6
8
* r.= I I 10 12
I 14
I 16
I 18
I 20
I
22
I
24
Days post-injection
Fig 1. Time course of body weight (in circles) and mean arthritis scores (in triangles) induced by adjuvant injection to rats. Data are expressed as the mean f SEM, *pcO.O5, **p
2053
GH and IGF-I in Chronic Inflammation
Vol. 65, No. 20, 1999
Table 1 Effect of adjuvant-induced arthritis on relative hepatic, adrenal and splenic weights. Data are expressed as the mean f SEM
Control n=20 day 18 n=7 dav 22 n=9 *P
liver g/l OOg 4.51 f 0.1 4.17 f 0.13* 3.84 f 0.04**
adrenal mgll OOg 17 f 0.3 26.9 &I .84** 22.8 f 1.49**
spleen mgfl OOg 258 f 7.4 362 f 18** 389 f 22 **
*PcO.Ol compared to control one-way ANOVA followed by the
Arthritis induced a significant increase in spleen and adrenal weights together with a decrease in hepatic weight (Table l).Arthritis decreased serum concentrations of GH (~~0.05) and IGF-I (p-=0.01) as well as hepatic IGF-I levels (peO.05) as is shown in Fig 2. However, serum concentrations of glucose and insulin were similar in control and in arthritic rats (Table 2).
C
18
22
Fig. 2 Changes in serum concentration of GH and IGF-I and in hepatic concentration of IGF-I, 18 and 22 days after adjuvant injection. Data represent the mean f SEM, * pco.05, ** p
18
22
2054
GH and IGF-I in Chronic Inflammation
Vol. 65, No. 20. 1999
Table 2 Serum concentration of insulin, glucose and proteins in control or in arthritic rats 18 or 22 after adjuvant-injection. Data are expressed as the mean f SEM. insulin pg/mL Control n=20 day 18 n=7 day 22 n=9
701 f 42 649 f 109 669 f 62
glucose mg/dL
proteins mg/mL 42.52 1 44.2 f 1.6 45 f 1.3
127 f 3.8 129 f 7.6 121 f 3.6
0
00
Q
0 0
o”
200
1
5 0.5
0 1.0
1.5
, 2.0
IGF-I pg/mL Fig. 3 Correlation between body weight and serum concentrations of IGF-I in rats, 22 days after adjuvant (solid circles) or vehicle injection (open circles). There was a significant correlation in arthritic rats r=0.924, p
There was a significant correlation between serum concentrations of IGF-I and body weight in arthritic rats both on day 18 (r-=0.78, peO.05) and 22 (r=0.92, p
2055
GH and IGF-I in Chronic Inflammation
Vol. 65, No. 20, 1999
IGFBPs
u) 160
0 control
m day 18
i 140 ([J ’ 120 75 $ 100 8
m day 22
80
24
control
kDa
day 18
Fig. 4 Effect of adjuvant-induced arthritis on serum IGFBPs. The upper panel represents the densitometric analysis of the western ligand blots of 2 uL serum from control or arthritic rats killed 18 or 22 days after adjuvant injection. Analysis of the different bands are expressed relative to that of corresponding IGFPB in control serum. Representative western ligand blots are represented in the lower panels.
GH and IGF-I in Chronic Inflammation
2056
Vol. 65, No. 20, 1999
Table 3 Pituitary weight, total RNA content and total RNA concentration in control or in rats 22 days after adjuvant injection. Data are expressed as the mean f SEM.
pituitary weight mg/lOO g RNA content ug RNA concentration us/ma
Control n=5 3.36 f 0.1 31.8 f 1.6 3.2 f 0.24
Arthritis n=6 3.67 f 0.32 18 rt 1.34** 2.36 f 0.08**
**PcO.Ol (Student’s t test).
There was a marked decrease in pituitary total RNA content and concentration in arthrtitic rats (Table 3). Fig 5 shows a representative autoradiography of the northern blot of GH mRNA in the pituitary. The densitometric analysis indicated that there was no difference between control and arthritic rats when analyzing the content in 5 ug of RNA. However, taking into account that 5 pg RNA in control rats represent approximately l/6 of the total RNA content, whereas in arthritic rats the same amount represents l/3-1/4 of pituitary RNA content, the GH mRNA content in the pituitary was significantly decreased (p
AA
CAACAAC
GH mRNA
18SrRNA Fig. 5 Autoradiogram of a representative Northern blot hybridization of 5 pg RNA from control (C) or in arthritic rats 22 days after adjuvant injection (AA). The lower autoradiography shows the ethidium bromide signal of 18s ribosomal RNA (rRNA) as load control.
GH and IGF-I in Chronic Inflammation
Vol. 65, No. 20, 1999
2057
150
E : 100
00
2 E
3
** -r
50
:
C
1’1 AA
Fig. 6 Effect of adjuvant-induced arthritis on pituitary GH mRNA content. (C) control rats, (AA) rats killled 22 days after adjuvant injection. Results were expressed as the percentage change in relation to control values f SEM for 5 and 6 different samples. *p
Discussion
After a latency period of lo-14 days, arthritic rats developed signs of inflammatory disease, together with a marked decrease in body weight gain. On days 18 and 22 arthritic rats had a marked reduction in pituitary GH mRNA and, in serum GH. In contrast, as previously reported (4), arthritis did not modify serum concentration of insulin or glucose. The reduced pituitary GH mRNA content in arthritic rats indicates that GH synthesis might be decreased. Therefore, a transcriptional mechanism decreasing GH gene expression seems to be activated during arthritis. These results are in accordance with those reported by others (19) showing that AA induced a decrease in plasma concentrations of GH and pituitary GH mRNA over the 12 days after adjuvant injection. The inhibitory effect of AA on GH supports previous observations concerning the inhibitory effect of the inflammatory response on GH secretion. Lipopolysaccharide (LPS) administration, another experimental model of inflammation, decreased pituitary GH release (20, 21). It is well known that several inflammatory cytokines (IL-l, TNF, IL-2) inhibit pituitary GH release “in vitro” (22, 23). Thus in arthritic rats, the increased concentration of cytokines may inhibit GH secretion by acting directly at the pituitary level. Another possibility is that cytokines inhibit pituitary GH by modifying hypothalamic growth hormone releasing hormone (GHRH) or somatostatin, since IL-I and TNF are able to stimulate somatostatin release and decrease GHRH secretion (24). We have demonstrated decreased levels of IGF-I in serum and in liver together with an increase in serum IGFBPs in arthritic rats. We have also found a strong association between serum concentration of IGF-I and body weight in arthritic rats. In accordance with our data,
20.58
GH and IGF-I in Chronic Inflammation
Vol. 65, No. 20, 1999
a correlation between IGF-I levels and height velocity have been described in juvenile chronic arthritis (JCA) (25). Taking into account that IGF-I stimulates protein synthesis and reduces protein breakdown, the inhibitory effect of chronic inflammation on IGF-I may lead to the catabolic state described not only in rats with AA, but also in humans with RA (1, 2) These data suggest that weight loss during chronic inflammation may result, at least in part, from a lack of anabolic hormones such as IGF-I and GH. Growth retardation is a common complication in juvenile chronic arthritis. However, there is not a clear consensus about the effect of arthritis on GH secretion in humans, since both a decrease in GH secretion and no modifications in circulating GH have been described (8, 9). Nevertheless, GH treatment increases IGF-I plasma levels and growth velocity in children with JCA (26, 27). In addition, in adult patients with newly diagnosed untreated rheumatoid arthritis, the anterior pituitary reserve seems to be unaffected except for GH which showed a blunted response to GHRH stimulation (28). Our data indicate that serum IGF-I levels decrease as result of the decrease in pituitary GH secretion. The parallel changes in serum IGF-I and in hepatic IGF-I suggest that the decrease in serum concentrations of IGF-I reflects changes in synthesis and secretion of IGF-I from the liver. The mechanism by which chronic inflammation decreases liver IGF-I production is not well defined. It has been previously reported that 11-l and TNF inhibit IGF-I synthesis after GH stimulation in hepatocyte cultures (29, 30). Therefore, it is also possible that the increased release of cytokines directly inhibits hepatic IGF-I synthesis. Consequently, the inhibitory effect of chronic inflammation on the GH-IGF axis may be exerted at different levels: at pituitary level decreasing GH secretion, at peripheral level inhibiting hepatic IGF-I synthesis, or both at pituitary and at hepatic level. While serum concentration of IGF-I was decreased in arthritic rats, serum concentrations of IGFBPs were increased. Similarly, LPS administration decreases serum GH and IGF-I concentrations and increases circulating IGFBPs (22, 31). The increase in serum IGFBPs may be due to increased cytokines, because IL-l and TNFcx induce IGFBP-l/2, IGFBP-3 and lGFBP-4 production (32, 33). IGFBPs can buffer growth and metabolic effects of IGF-I by sequestering it in the circulation or by preventing its union to the receptor (34). IGFBP3 has been shown to inhibit both basal growth and IGF-stimulated cell proliferation in human breast cancer cells (35, 35), and to induce apoptosis (37) the latter effect being independent of the IGF-I receptor. Therefore, it is also posible that increased IGFBPs in the serum contribute to the inhibition of growth observed in arthritis. An increase in IGFBP-3 concentration in the synovial fluid in rheumatoid arthritis and in osteoarthritis (OA) patients has been observed (38, 39). Moreover, it has been postulated that the hyporesponsiveness of osteoarthritic chondrocytes to IGF-I may be due to increased chondrocyte synthesis of IGFBP-3 observed in OA (33,40). In another study, IGFBP-2 and IGFBP-3 were elevated both in plasma and in synovial fluid of RA patients, and IGFBP-3 concentrations were positively correlated with the synovial fluid levels of IL-1 and TNF-a (41). IL-I and TNF-a are able to induce IGFBP-3 production (32), this observation may explain the increase in IGFBP-3 levels in arthritis. To our knowlege, this is the first reported study of IGF-I and IGFBPs in adjuvant arthritis. The following endocrine changes have been observed in this experimental model: a decrease in pituitary GH secretion, a larger amount of circulating IGFBPs in arthritic rats together with a decreased concentration of IGF-I in the liver and serum. All of these changes
Vol. 65, No. 20, 1999
GH and IGF-I in Chronic
Inflammation
2059
and the increase in serum concentration of glucocorticoids, reported by others (5) may lead to a catabolic situation with a decrease in protein and fat storage which can explain the decrease in body weight observed in arthritic rats. Acknowledgements The authors are indebted to A. Carmona for technical assistance. We are grateful to Lilly Spain for the IGF-I, and to the NIDDK National Hormone and Pituitary Program for the reagents for GH and IGF-I determinations. This work was supported by DGICYT grant PM950068. References 1. 2.
3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.
R. ROUBENOFF, L.M. FREEMAN, D.E. SMITH, L.W. ABAD, C.A. DINARELLO, J.J. KEHAYIAS, Arthritis Rheum. 40 534-539 (1997) R. ROUBENOFF, R.A. ROUBENOFF, J. G. CANNON, J.J. KEHAYIAS, H. ZHUANG, B. DAWSON-HUGHES, C.A. DINARELLO, I.H. ROSENBERG, J. Clin. Invest. 93 379-2386 (1994) H. TANAKA, Y. UETA, U. YAMASHITA, H. KANNAN, H. YAMASHITA, Brain Res. Bull. 39 33-37 (1991) A. M. ROFE, M.W. WHITEHOUSE, C.S. BOURGEOIS, D. R. HAYNES, 8. VERNON-ROBERTS, Immunol. Cell. Biol. 68 63-69 (1990) A. STEPHANOUS, N. J. SARLIS, R. A. KNIGHT, S. L. LIGHTMAN, H. S. CHOWDREY, J. Mol. Endocrinol. 9 273-281(1992) M. NEIDHART, D.F. LARSON, J. Neuroimmunol. 26 97-105 (1990) A. G. JOHANSSON, D. J. BAYLINK, E. EKENSTAM, E. LINDH, S. MOHAN, S. LJUNGHALL, Bone. Miner. 24 25-21 (1994) 0. BUTENANDT, A. KELCH, E. RAJMAN, Eur. J. Pediatr. 118 53-62 1974 P. M. WOO, Clin. Exp. Rheumatol. 12 Suppl 10 pS87-90 (1994) L. SELGAS, A. ARCE, A. I. ESQUIFINO, D. P. CARDINALI, Chronobiol. Int. 14 253265 (1997) B. CALVINO, J. BESSON, F. MOUNIER, C. KORDON, M. BLUET-PAJOT, Pain 49 27-32 (1992) M. T. BLUET-PAJOT, F. MOUNIER, A. SLAMA, C. VIDEAU, C. KORDON, J. EPELBAUM, B. CALVINO, Neuroendocrinol. 63 85-92 (1996) W. H. DAUGHADAY, P. ROTWEIN, J. Clin. Endocrinol. Metab. 51 781-788 (1980) I. TORRES-ALEMAN, S. PONS, F. F. SANTOS-BENITO, Eur. J. Neurosci. 4 864869 (1992) E. VARA, J. TAMARIT-RODRIGUEZ, Acta Endocrinol. 118 429-436 (1988) M.M. BRADFORD, Anal. Biochem. 72 248-254 (1976) P. HOSSENLOP, D. SEURIN, B. SEGOVIA-QUINSON, S. HARDOUIN, M. BINOUX, Anal. Biochem. 154 138-143 (1986) H. S. PETER, J. SHINE, J. A. MARTIAL, J. D. BAXTER, H. M. GOODMAN, Nature 270 486-494 (1977) M. NEIDHART, E. W. FLUCKIGER, Immunol. 77 449-455 (1992) N. W. KASTING, J. B. MARTIN, Am. J. Physiol. 43 E332-E337 (1982) J. FAN, P. E. MOLINA, M. C. GELATO, C. H. LANG, Endocrinology 134 1685-1692 (1994) E. D. BERNTON, J. E. BEACH, J. W. HOLADAY, R. C. SMALLRIDGE, H. C. FEIN, Science 238 519-521 (1987)
2060
23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41.
GH and IGF-I in Chronic Inflammation
Vol. 65, No. 20, 1999
P. E. WALTON, M. J. CRONM, Endocrinology 125 925-929 (1989) J. N. PEISEN, K. J. MCDONNELL, S. E. MULRONEY, M. D. LUMPKIN, Endocrinology 136 3378-3390 (1995) T. J. AITMAN, R. G. PALMER, J. LOFTUS, B. M. ANSELL, J. P. ROYSTON, J. D. TEALE, R. N. CLAYTON, Clin. Exp. Rheumatol. 7 557-561 (1989) U. M. DAVIES, M. ROONEY, M. A. PREECE, B. M. ANSELL, P. WOO, J. Rheumatology. 21153-I 58 (1994) G. TOUATI, A.M. PRIEUR, J.C. RUIZ, M. NOEL, P. CZERNICHOW, J Ctin Endocrinot and Metab 83 403-409 (1998) E. TEMPL, M. KOELLER, M. RIEDL, 0. WAGNER, W. GRANINGER, A. LUGER, Brit J Rheumatot 35 350-356 (1996) M. WOLF, S. BBHM, M. BRAND, G. KREYMANN, Europ J Endocrinot 135 729-737 (1996) J.P. THISSEN, J. VERNIERS, Endocrinology 138 1078-1084 (1997) L. SOTO, A.I. MART/N, S. MILtiN, E. VARA, A. LOPEZ-CALDERON, J Endocrinot 159 239-245 (1998) J. FAN, D. CHAR, G.J. BAGBY, M.C. GELATO, C.H. LANG, Amer J Physiot 269 1204-1212 (1995) R.C. OLNEY, D.M. WILSON, M. MOHTAI, P.J. FIELDER, R.L. SMITH, J Endocrinot 146 279-286 (1995) M.M. RECHLER, D.R. CLEMMONS, Trends Endocrinot and Metabot 9 176-183 (1998) J.L. MARTIN, J.A. COVERLEY, ST. PATTISON, R.C. BAXTER, Endocrinology 136 2119-26 (1995) H. HUYNH, X. YANG, M. POLLAK, J Biol and Chem 271 1016-1021 (1996) R. RAJAH, B. VALENTINIS, P. COHEN, J Biot Chem 272 12181-12188 (1997) J.K. FERNIHOUGH, M.E. BILLINGHAM, S. CWYFAN-HUGHES, J.M. HOLLY, Arthritis Rheum 39 1556-l 565 (1996) T. MATSUMOTO, SE. GARGOSKY, K. IWASAKI, R.G. ROSENFELD, J Ctin Endocrinol Metab 81 150-I 55 (1996) G. TARDIF, P. REBOUL, J.P. PELLETIER, C. GENG, J.M. CLOUTIER, J. MARTELPELLETIER, Arthritis Rheum 39 968-978 (1996) J. NEIDEL, W.F. BLUM, H.J. SCHAEFFER, M. SCHULZE, E. SCHONAU, J. LINDSCHAU, J. Fc)LL, Rheumatot tnt 17 29-37 (1997)
PI1 SOO24-3205(99)00472-S
CHRONIC INFLAMMATION INHIBITS GH SECRETION AND ALTERS THE SERUM INSULIN-LIKE GROWTH FACTOR SYSTEM IN RATS A LOPEZ-CALDERGN, L SOTO, *Al MARTiN Dpt Fisiologia, Fat Medicina, Univ Complutense and *Dpt CC Morfologicas y Fisiologia, Univ Europea, Madrid, Spain. (Received in final form July 2. 1999)
Summary
Adjuvant-induced arthritis in rats is associated with growth failure, hypermetabolism and accelerated protein breakdown. The aim of this work was to study the effects of adjuvant-induced arthritis on GH and insulin-like growth factor-l (IGF-I). Arthritis was induced by an intradermal injection of complete Freund’s adjuvant and rats were killed 18 and 22 days later. IGF-I and GH levels were measured by radioimmunoassay. Pituitary GH mRNA was analyzed by northern blot and IGF binding proteins (IGFBPs) by western blot. Arthritic rats showed a decrease in both serum and hepatic concentrations of IGF-I. On the contrary, arthritis increased the circulating IGFBPs. The serum concentration of IGF-I in the arthritic rats was negatively correlated with the body weight loss observed in these animals. Arthritis decreased the serum concentration of GH and this decrease seems to be due to an inhibition of GH synthesis, since pituitary GH mRNA content was decreased in arthritic rats (~~0.01). These data suggest that the decrease in body weight gain in arthritic rats may be, at least in part, secondary to the decrease in GH and IGF-I secretion. Furthermore, the increased serum IGFBPs may also be involved in the disease process. Key Words: adjuvant-induced arthritis, inflammation, GH, GH mRNA, IGF-I, IGFE3Ps
Adjuvant-induced arthritis (AA) is an animal model used to study rheumatoid arthritis and chronic inflammatory stress. Chronic inflammation, in experimental animal models such as AA, is associated with a decrease in body weight and a loss of body cell mass (1). Similarly, rheumatoid arthritis (RA) patients have cachexia hypermetabolism and accelerated protein breakdown (2). When analyzing the nutritional status, arthritic rats have anorexia (3) but they lose more weight than pair fed controls (1, 4) indicating a catabolic condition, The increased production of cytokines during AA (1) could play an important role in the inhibition of growth,but it seems to be also mediated by disturbances of the endocrine system.
Corresponding author: A Lopez-Calderon. Dpt Fisiologia, Fat Medicina, Univ Complutense. 28040 Madrid. Spain. Tel: 3491-3941491; Fax: 3491-3941628; E-mail: ale @eucmax.sim.es
2050
GH and IGF-I in Chronic Infhmmation
Vol. 65, No. 20, 1999
It is generally accepted that AA is associated with an increased activity of the pituitaryadrenal axis, with an increase in circulating adrenocorticopic hormone (ACTH) and corticosterone as well as in the pituitary proopiomelanocortin (POMC) mRNA content (5). In addition to the modifications of the adrenal axis, a number of neuroendocrine changes has also been described in arthritic animals (5, 6). Regarding anabolic hormones, there is not a clear consensus about the effect of AA on growth hormone (GH) and insulin-like growth factor-l (IGF-I). In humans, patients with rheumatoid arthritis have low serum concentrations of IGF-I and IGF-II (7). Children with RA have shown impaired growth together with normal or low circulating GH levels (8, 9). An increase in prolactin release along with a decrease in circulating GH in AA rats were previously described during the early latency period before the onset of arthritis (6, 10). In contrast, an increased pituitary GH release has also been described in the same animal model when the rats have developed external signs of the disease (11, 12). The aim of this study was to analyze the effect of adjuvant-induced arthritis on the somatotropic axis. To address this question, we analyzed pituitary GH mRNA, serum and hepatic concentrations of IGF-I, and its binding proteins IGFBPs, since they play a crucial role in modulating IGF-I actions.
Methods
Male Sprague-Dawley rats were purchased from Charles River (Barcelona, Spain). The procedures followed the guidelines recommended by the EU for the care and use of laboratory animals. Arthritis was induced in the rats by a intradermal injection of complete Freund’s adjuvant (1 mg heat-inactivated Mycobacterium butyricum in incomplete Freund’s adjuvant, Difco laboratory) at the base of the tail. Control animals were injected with incomplete Freund’s adjuvant. Rats were housed 4-5 per cage under controlled conditions of light (light on from 7:30 to 19:30) and temperature (22 f 2OC). Food and water were available “ad libitum”. Both control and arthritic rats were killed by decapitation18 and 22 days after adjuvant or vehicle injection between 1 I:00 and 12:00 h. Trunk blood was collected in cooled tubes, allowed to clot, centrifuged and the serum was stored at -20 OC until GH, IGF-I, insulin and IGFBPs analysis were performed. Immediately after decapitation the pituitary gland was frozen in a dry ice/acetone mixture and stored at -80 OC until the northern blot analysis. The liver was removed, dissected, frozen and stored at -20 OC until IGF-I assay. The spleen and adrenals were also dissected and weighed. Assessment of arthritis The arthritis index of each animal was scored by grading each paw from 0 to 4, determined as: 0- no erythema or swelling. I- slight erythema or swelling of one or more digits. 2- entire paw swollen. 3- erythema and swelling of the ankle. 4- ankylosis, incapacity to bend the ankle. The severity score was the sum of the clinical scores of each limb, the maximum value being 16 (3). Hormone determination IGF-I concentrations were measured by a double-antibody RIA using the antibody NIDDK UB2-495 a gift from Drs Underwood and Van Wik, and it is distributed by the Hormone Distribution Program of NIDDK through the National Hormone and Pituitary Program. Serum IGF-I binding proteins (IGFBPs) were removed by an acid-ethanol extraction (13). Hepatic IGF-I was extracted as described by Torres-Aleman et a/. (14). Samples were homogenized
Vol. 65, No. 20, 1999
GH and IGF-I in Chronic Inllammation
2051
in 1 N acetic acid, boiled for 20 min, and lyophilized. To confirm the elimination of IGFBPs, extracted serum and liver fractions were incubated with ‘2?-IGF-I at 4OC for 24 hours. The separation of bound and free tracers was carried out with dextran charcoal. Extracted fractions showed 98-100 % of free IGF-I. Levels of IGF-I were expressed in terms of IGF-I A52-EPD-186 standard (Eli Lilly 8 Company). The intra-assay coefficient of variation was 8%; samples from one experiment were run in the same assay. Concentrations of GH were measured by a radioimmunoassay (RIA) using reagents provided by Dr. Parlow of the National Hormone and Pituitary Program. Levels of GH were expressed in terms of NIDDK rat-RP-2 standard. The GH detection level was 10 pg, and the intra-assay coefficient of variation was 3%. Ail necessary comparisons between test and control animals were made within the same assay. Insulin was measured by a radioimmunoassay previously described (15). Protein content was measured by Bradford’s method (16) and serum concentrations of glucose with a commercial kit from Boehringer Mannheim.
Western ligand b/of Western blots were prepared as previously described (17). Samples were subjected to 1% SDS-12.5 % acrylamide gels non reducing electrophoresis, and electrotransferred to nitrocellulose membranes (HybondTM -C extra, Amersham, UK). The nitroceiiulose sheets were dried and blocked for 1 h with 5% non-fat dry milk, 0.1 % Tween (Sigma), in Trisbuffered saline, and incubated overnight at 4OC with ‘251-labeled IGF-I (5 x ld cpmlml). The nitroceiiulose sheets were then washed, dried and blots were exposed at -80 OC to X-ray film (Kodak X-Omat AR, Eastman Kodak, Rochester NY) and two intensifying screens for l-5 days according to the signal obtained. Autoradiographs were analyzed by densitometric scanning using a PC-Image VGA24 program for Windows. The density of the IGFBP bands in each lane was expressed as the percentage of the mean density of control sera.
Northern blot of pituitary GH mRNA content The pituitaries were extracted in pairs, total RNA was extracted by the guanidine thiocyanate method using a commercial kit (UltraspecTM RNA, Biotecx Laboratories). Each 5 ug sample of total RNA was denatured, separated by formaldehyde-agarose gel eiectrophoresis, transferred to nylon membranes (Hybond-N+, Amersham, UK) and fixed by uv crosslinking (Fotodyne Hartland, WI, USA). Homogeneity of gel loading was confirmed by the intensity of the ribosomai 18s RNA bands in the transferred membranes stained with ethidium bromide. The rat GH cDNA probe was the Hindlll fragment of the pRGH-1 (18). Prehybridization was performed for 3 h at 42 OC in buffer, (50 % formamide, 5 X SSPE, 5 X Denhart’s reagent, 1% SDS, and 100 ug/mL salmon sperm DNA) followed by hybridization for 16 h at the same temperature with 1-3 x 10’ cpm/mL GH cDNA labeled by random priming (High Prime, Boheringer Mannheim) in the same buffer. The membranes were washed twice with 2 x SSPE, 0.1% SDS at room temperature for 5 min, twice at 42 OC with 2xSSPE, 0.1 % SDS for 30 min and twice with 0.1 x SSPE, 0.1 % SDS at 50 OC also for 30 min. Exposure to X-ray film was performed at -80 OC. The intensities of autoradiogram signal levels and bromide-stained ribosomal 18s RNA in the nylon filters were analyzed by densitometric scanning using a PC-Image VGA24 program for Windows, and were normalized for 18s ribosomal RNA levels. The total GH mRNA content per entire pituitary was then calculated by multiplying the northern hybridization densitometric units of the densitometric analysis by the total amount of the pituitary RNA in micrograms.
Vol. 65, No. 20, 1999
GH and IGF-I in Chronic Inflammation
Statistical analysis Statistical significance was calculated by Duncan Multiple Range test after one-way analysis of variance, and Student’s t test when comparing only two means. Because there were no significant differences between data obtained in the control rats killed 18 or 22 days after vehicle injection, they were used as one control group. Correlation between different variables was calculated by linear regression. Serum GH data were subjected to log transformation since variances showed a log-normal distribution. A p-value less than 0.05 was considered significant.
Results Rats developed signs of inflammation 12-14 days after adjuvant injection, reaching the maximum level on day 21 (Fig 1). Body weight gain in the arthritic rats was not different from controls until day IO when arthritic rats started to gain less body weight than controls until the end of the study (Fig 1).
340 320 300
-8- Control 280 - -o- A r t h r i t i s
16
OI 260
14
E
.% 240 3 &T 220
12
m” 200 180 160 140 120
I 2
. .
l
.
I
I
1
4
6
8
* r.= I I 10 12
I 14
I 16
I 18
I 20
I
22
I
24
Days post-injection
Fig 1. Time course of body weight (in circles) and mean arthritis scores (in triangles) induced by adjuvant injection to rats. Data are expressed as the mean f SEM, *pcO.O5, **p
2053
GH and IGF-I in Chronic Inflammation
Vol. 65, No. 20, 1999
Table 1 Effect of adjuvant-induced arthritis on relative hepatic, adrenal and splenic weights. Data are expressed as the mean f SEM
Control n=20 day 18 n=7 dav 22 n=9 *P
liver g/l OOg 4.51 f 0.1 4.17 f 0.13* 3.84 f 0.04**
adrenal mgll OOg 17 f 0.3 26.9 &I .84** 22.8 f 1.49**
spleen mgfl OOg 258 f 7.4 362 f 18** 389 f 22 **
*PcO.Ol compared to control one-way ANOVA followed by the
Arthritis induced a significant increase in spleen and adrenal weights together with a decrease in hepatic weight (Table l).Arthritis decreased serum concentrations of GH (~~0.05) and IGF-I (p-=0.01) as well as hepatic IGF-I levels (peO.05) as is shown in Fig 2. However, serum concentrations of glucose and insulin were similar in control and in arthritic rats (Table 2).
C
18
22
Fig. 2 Changes in serum concentration of GH and IGF-I and in hepatic concentration of IGF-I, 18 and 22 days after adjuvant injection. Data represent the mean f SEM, * pco.05, ** p
18
22
2054
GH and IGF-I in Chronic Inflammation
Vol. 65, No. 20. 1999
Table 2 Serum concentration of insulin, glucose and proteins in control or in arthritic rats 18 or 22 after adjuvant-injection. Data are expressed as the mean f SEM. insulin pg/mL Control n=20 day 18 n=7 day 22 n=9
701 f 42 649 f 109 669 f 62
glucose mg/dL
proteins mg/mL 42.52 1 44.2 f 1.6 45 f 1.3
127 f 3.8 129 f 7.6 121 f 3.6
0
00
Q
0 0
o”
200
1
5 0.5
0 1.0
1.5
, 2.0
IGF-I pg/mL Fig. 3 Correlation between body weight and serum concentrations of IGF-I in rats, 22 days after adjuvant (solid circles) or vehicle injection (open circles). There was a significant correlation in arthritic rats r=0.924, p
There was a significant correlation between serum concentrations of IGF-I and body weight in arthritic rats both on day 18 (r-=0.78, peO.05) and 22 (r=0.92, p
2055
GH and IGF-I in Chronic Inflammation
Vol. 65, No. 20, 1999
IGFBPs
u) 160
0 control
m day 18
i 140 ([J ’ 120 75 $ 100 8
m day 22
80
24
control
kDa
day 18
Fig. 4 Effect of adjuvant-induced arthritis on serum IGFBPs. The upper panel represents the densitometric analysis of the western ligand blots of 2 uL serum from control or arthritic rats killed 18 or 22 days after adjuvant injection. Analysis of the different bands are expressed relative to that of corresponding IGFPB in control serum. Representative western ligand blots are represented in the lower panels.
GH and IGF-I in Chronic Inflammation
2056
Vol. 65, No. 20, 1999
Table 3 Pituitary weight, total RNA content and total RNA concentration in control or in rats 22 days after adjuvant injection. Data are expressed as the mean f SEM.
pituitary weight mg/lOO g RNA content ug RNA concentration us/ma
Control n=5 3.36 f 0.1 31.8 f 1.6 3.2 f 0.24
Arthritis n=6 3.67 f 0.32 18 rt 1.34** 2.36 f 0.08**
**PcO.Ol (Student’s t test).
There was a marked decrease in pituitary total RNA content and concentration in arthrtitic rats (Table 3). Fig 5 shows a representative autoradiography of the northern blot of GH mRNA in the pituitary. The densitometric analysis indicated that there was no difference between control and arthritic rats when analyzing the content in 5 ug of RNA. However, taking into account that 5 pg RNA in control rats represent approximately l/6 of the total RNA content, whereas in arthritic rats the same amount represents l/3-1/4 of pituitary RNA content, the GH mRNA content in the pituitary was significantly decreased (p
AA
CAACAAC
GH mRNA
18SrRNA Fig. 5 Autoradiogram of a representative Northern blot hybridization of 5 pg RNA from control (C) or in arthritic rats 22 days after adjuvant injection (AA). The lower autoradiography shows the ethidium bromide signal of 18s ribosomal RNA (rRNA) as load control.
GH and IGF-I in Chronic Inflammation
Vol. 65, No. 20, 1999
2057
150
E : 100
00
2 E
3
** -r
50
:
C
1’1 AA
Fig. 6 Effect of adjuvant-induced arthritis on pituitary GH mRNA content. (C) control rats, (AA) rats killled 22 days after adjuvant injection. Results were expressed as the percentage change in relation to control values f SEM for 5 and 6 different samples. *p
Discussion
After a latency period of lo-14 days, arthritic rats developed signs of inflammatory disease, together with a marked decrease in body weight gain. On days 18 and 22 arthritic rats had a marked reduction in pituitary GH mRNA and, in serum GH. In contrast, as previously reported (4), arthritis did not modify serum concentration of insulin or glucose. The reduced pituitary GH mRNA content in arthritic rats indicates that GH synthesis might be decreased. Therefore, a transcriptional mechanism decreasing GH gene expression seems to be activated during arthritis. These results are in accordance with those reported by others (19) showing that AA induced a decrease in plasma concentrations of GH and pituitary GH mRNA over the 12 days after adjuvant injection. The inhibitory effect of AA on GH supports previous observations concerning the inhibitory effect of the inflammatory response on GH secretion. Lipopolysaccharide (LPS) administration, another experimental model of inflammation, decreased pituitary GH release (20, 21). It is well known that several inflammatory cytokines (IL-l, TNF, IL-2) inhibit pituitary GH release “in vitro” (22, 23). Thus in arthritic rats, the increased concentration of cytokines may inhibit GH secretion by acting directly at the pituitary level. Another possibility is that cytokines inhibit pituitary GH by modifying hypothalamic growth hormone releasing hormone (GHRH) or somatostatin, since IL-I and TNF are able to stimulate somatostatin release and decrease GHRH secretion (24). We have demonstrated decreased levels of IGF-I in serum and in liver together with an increase in serum IGFBPs in arthritic rats. We have also found a strong association between serum concentration of IGF-I and body weight in arthritic rats. In accordance with our data,
20.58
GH and IGF-I in Chronic Inflammation
Vol. 65, No. 20, 1999
a correlation between IGF-I levels and height velocity have been described in juvenile chronic arthritis (JCA) (25). Taking into account that IGF-I stimulates protein synthesis and reduces protein breakdown, the inhibitory effect of chronic inflammation on IGF-I may lead to the catabolic state described not only in rats with AA, but also in humans with RA (1, 2) These data suggest that weight loss during chronic inflammation may result, at least in part, from a lack of anabolic hormones such as IGF-I and GH. Growth retardation is a common complication in juvenile chronic arthritis. However, there is not a clear consensus about the effect of arthritis on GH secretion in humans, since both a decrease in GH secretion and no modifications in circulating GH have been described (8, 9). Nevertheless, GH treatment increases IGF-I plasma levels and growth velocity in children with JCA (26, 27). In addition, in adult patients with newly diagnosed untreated rheumatoid arthritis, the anterior pituitary reserve seems to be unaffected except for GH which showed a blunted response to GHRH stimulation (28). Our data indicate that serum IGF-I levels decrease as result of the decrease in pituitary GH secretion. The parallel changes in serum IGF-I and in hepatic IGF-I suggest that the decrease in serum concentrations of IGF-I reflects changes in synthesis and secretion of IGF-I from the liver. The mechanism by which chronic inflammation decreases liver IGF-I production is not well defined. It has been previously reported that 11-l and TNF inhibit IGF-I synthesis after GH stimulation in hepatocyte cultures (29, 30). Therefore, it is also possible that the increased release of cytokines directly inhibits hepatic IGF-I synthesis. Consequently, the inhibitory effect of chronic inflammation on the GH-IGF axis may be exerted at different levels: at pituitary level decreasing GH secretion, at peripheral level inhibiting hepatic IGF-I synthesis, or both at pituitary and at hepatic level. While serum concentration of IGF-I was decreased in arthritic rats, serum concentrations of IGFBPs were increased. Similarly, LPS administration decreases serum GH and IGF-I concentrations and increases circulating IGFBPs (22, 31). The increase in serum IGFBPs may be due to increased cytokines, because IL-l and TNFcx induce IGFBP-l/2, IGFBP-3 and lGFBP-4 production (32, 33). IGFBPs can buffer growth and metabolic effects of IGF-I by sequestering it in the circulation or by preventing its union to the receptor (34). IGFBP3 has been shown to inhibit both basal growth and IGF-stimulated cell proliferation in human breast cancer cells (35, 35), and to induce apoptosis (37) the latter effect being independent of the IGF-I receptor. Therefore, it is also posible that increased IGFBPs in the serum contribute to the inhibition of growth observed in arthritis. An increase in IGFBP-3 concentration in the synovial fluid in rheumatoid arthritis and in osteoarthritis (OA) patients has been observed (38, 39). Moreover, it has been postulated that the hyporesponsiveness of osteoarthritic chondrocytes to IGF-I may be due to increased chondrocyte synthesis of IGFBP-3 observed in OA (33,40). In another study, IGFBP-2 and IGFBP-3 were elevated both in plasma and in synovial fluid of RA patients, and IGFBP-3 concentrations were positively correlated with the synovial fluid levels of IL-1 and TNF-a (41). IL-I and TNF-a are able to induce IGFBP-3 production (32), this observation may explain the increase in IGFBP-3 levels in arthritis. To our knowlege, this is the first reported study of IGF-I and IGFBPs in adjuvant arthritis. The following endocrine changes have been observed in this experimental model: a decrease in pituitary GH secretion, a larger amount of circulating IGFBPs in arthritic rats together with a decreased concentration of IGF-I in the liver and serum. All of these changes
Vol. 65, No. 20, 1999
GH and IGF-I in Chronic
Inflammation
2059
and the increase in serum concentration of glucocorticoids, reported by others (5) may lead to a catabolic situation with a decrease in protein and fat storage which can explain the decrease in body weight observed in arthritic rats. Acknowledgements The authors are indebted to A. Carmona for technical assistance. We are grateful to Lilly Spain for the IGF-I, and to the NIDDK National Hormone and Pituitary Program for the reagents for GH and IGF-I determinations. This work was supported by DGICYT grant PM950068. References 1. 2.
3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.
R. ROUBENOFF, L.M. FREEMAN, D.E. SMITH, L.W. ABAD, C.A. DINARELLO, J.J. KEHAYIAS, Arthritis Rheum. 40 534-539 (1997) R. ROUBENOFF, R.A. ROUBENOFF, J. G. CANNON, J.J. KEHAYIAS, H. ZHUANG, B. DAWSON-HUGHES, C.A. DINARELLO, I.H. ROSENBERG, J. Clin. Invest. 93 379-2386 (1994) H. TANAKA, Y. UETA, U. YAMASHITA, H. KANNAN, H. YAMASHITA, Brain Res. Bull. 39 33-37 (1991) A. M. ROFE, M.W. WHITEHOUSE, C.S. BOURGEOIS, D. R. HAYNES, 8. VERNON-ROBERTS, Immunol. Cell. Biol. 68 63-69 (1990) A. STEPHANOUS, N. J. SARLIS, R. A. KNIGHT, S. L. LIGHTMAN, H. S. CHOWDREY, J. Mol. Endocrinol. 9 273-281(1992) M. NEIDHART, D.F. LARSON, J. Neuroimmunol. 26 97-105 (1990) A. G. JOHANSSON, D. J. BAYLINK, E. EKENSTAM, E. LINDH, S. MOHAN, S. LJUNGHALL, Bone. Miner. 24 25-21 (1994) 0. BUTENANDT, A. KELCH, E. RAJMAN, Eur. J. Pediatr. 118 53-62 1974 P. M. WOO, Clin. Exp. Rheumatol. 12 Suppl 10 pS87-90 (1994) L. SELGAS, A. ARCE, A. I. ESQUIFINO, D. P. CARDINALI, Chronobiol. Int. 14 253265 (1997) B. CALVINO, J. BESSON, F. MOUNIER, C. KORDON, M. BLUET-PAJOT, Pain 49 27-32 (1992) M. T. BLUET-PAJOT, F. MOUNIER, A. SLAMA, C. VIDEAU, C. KORDON, J. EPELBAUM, B. CALVINO, Neuroendocrinol. 63 85-92 (1996) W. H. DAUGHADAY, P. ROTWEIN, J. Clin. Endocrinol. Metab. 51 781-788 (1980) I. TORRES-ALEMAN, S. PONS, F. F. SANTOS-BENITO, Eur. J. Neurosci. 4 864869 (1992) E. VARA, J. TAMARIT-RODRIGUEZ, Acta Endocrinol. 118 429-436 (1988) M.M. BRADFORD, Anal. Biochem. 72 248-254 (1976) P. HOSSENLOP, D. SEURIN, B. SEGOVIA-QUINSON, S. HARDOUIN, M. BINOUX, Anal. Biochem. 154 138-143 (1986) H. S. PETER, J. SHINE, J. A. MARTIAL, J. D. BAXTER, H. M. GOODMAN, Nature 270 486-494 (1977) M. NEIDHART, E. W. FLUCKIGER, Immunol. 77 449-455 (1992) N. W. KASTING, J. B. MARTIN, Am. J. Physiol. 43 E332-E337 (1982) J. FAN, P. E. MOLINA, M. C. GELATO, C. H. LANG, Endocrinology 134 1685-1692 (1994) E. D. BERNTON, J. E. BEACH, J. W. HOLADAY, R. C. SMALLRIDGE, H. C. FEIN, Science 238 519-521 (1987)
2060
23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41.
GH and IGF-I in Chronic Inflammation
Vol. 65, No. 20, 1999
P. E. WALTON, M. J. CRONM, Endocrinology 125 925-929 (1989) J. N. PEISEN, K. J. MCDONNELL, S. E. MULRONEY, M. D. LUMPKIN, Endocrinology 136 3378-3390 (1995) T. J. AITMAN, R. G. PALMER, J. LOFTUS, B. M. ANSELL, J. P. ROYSTON, J. D. TEALE, R. N. CLAYTON, Clin. Exp. Rheumatol. 7 557-561 (1989) U. M. DAVIES, M. ROONEY, M. A. PREECE, B. M. ANSELL, P. WOO, J. Rheumatology. 21153-I 58 (1994) G. TOUATI, A.M. PRIEUR, J.C. RUIZ, M. NOEL, P. CZERNICHOW, J Ctin Endocrinot and Metab 83 403-409 (1998) E. TEMPL, M. KOELLER, M. RIEDL, 0. WAGNER, W. GRANINGER, A. LUGER, Brit J Rheumatot 35 350-356 (1996) M. WOLF, S. BBHM, M. BRAND, G. KREYMANN, Europ J Endocrinot 135 729-737 (1996) J.P. THISSEN, J. VERNIERS, Endocrinology 138 1078-1084 (1997) L. SOTO, A.I. MART/N, S. MILtiN, E. VARA, A. LOPEZ-CALDERON, J Endocrinot 159 239-245 (1998) J. FAN, D. CHAR, G.J. BAGBY, M.C. GELATO, C.H. LANG, Amer J Physiot 269 1204-1212 (1995) R.C. OLNEY, D.M. WILSON, M. MOHTAI, P.J. FIELDER, R.L. SMITH, J Endocrinot 146 279-286 (1995) M.M. RECHLER, D.R. CLEMMONS, Trends Endocrinot and Metabot 9 176-183 (1998) J.L. MARTIN, J.A. COVERLEY, ST. PATTISON, R.C. BAXTER, Endocrinology 136 2119-26 (1995) H. HUYNH, X. YANG, M. POLLAK, J Biol and Chem 271 1016-1021 (1996) R. RAJAH, B. VALENTINIS, P. COHEN, J Biot Chem 272 12181-12188 (1997) J.K. FERNIHOUGH, M.E. BILLINGHAM, S. CWYFAN-HUGHES, J.M. HOLLY, Arthritis Rheum 39 1556-l 565 (1996) T. MATSUMOTO, SE. GARGOSKY, K. IWASAKI, R.G. ROSENFELD, J Ctin Endocrinol Metab 81 150-I 55 (1996) G. TARDIF, P. REBOUL, J.P. PELLETIER, C. GENG, J.M. CLOUTIER, J. MARTELPELLETIER, Arthritis Rheum 39 968-978 (1996) J. NEIDEL, W.F. BLUM, H.J. SCHAEFFER, M. SCHULZE, E. SCHONAU, J. LINDSCHAU, J. Fc)LL, Rheumatot tnt 17 29-37 (1997)