Hormonal Regulation of Liver Fatty Acid-Binding

0013-7227/98/$03.00/0 Endocrinology Copyright © 1998 by The Endocrine Society

Vol. 139, No. 6 Printed in U.S.A.

Hormonal Regulation of Liver Fatty Acid-Binding Protein in Vivo and in Vitro: Effects of Growth Hormone and Insulin* LINDA CARLSSON, IDA NILSSON,

AND

JAN OSCARSSON

Department of Physiology, Go¨teborg University, S-405 30 Go¨teborg, Sweden ABSTRACT Liver fatty acid-binding protein (LFABP) is an abundant protein in hepatocytes that binds most of the long chain fatty acids present in the cytosol. It is suggested to be of importance for fatty acid uptake and utilization in the hepatocyte. In the present study, the effects of bovine GH (bGH) and other hormones on the expression of LFABP and its messenger RNA (mRNA) were studied in hypophysectomized rats and in vitro using primary cultures of rat hepatocytes. One injection of bGH increased LFABP mRNA levels about 5-fold after 6 h, but there was no effect of this treatment on LFABP levels. However, 7 days of bGH treatment increased both LFABP mRNA and LFABP protein levels 2- to 5-fold. Female rats had higher levels of LFABP than male rats. Hypophysectomy of female rats, but not that of male rats, decreased LFABP levels markedly. Treatment of hypophysectomized rats with bGH for 7 days as two daily injections or as a continuous infusion increased LFABP levels to a similar degree. This finding indicates that the sex difference in the expression of LFABP is not regulated by the sexually dimorphic secretory pattern of GH.

G

H HAS many effects on lipid and lipoprotein metabolism, including enhanced lipolysis and decreased lipogenesis in adipose tissue (1–3), changes in serum lipoprotein concentrations, and hepatic lipid and lipoprotein metabolism (4 –7). GH enhances hepatic mitochondrial b-oxidation of polyunsaturated fatty acids (8) partly via effects on fatty acid composition of mitochondrial phospholipids (9). GH also decreases acetyl coenzyme A carboxylase activity (10) and fatty acid synthase messenger RNA (mRNA) (11) in the liver, indicating decreased fatty acid synthesis. GH therapy of hypophysectomized (Hx) rats increases triglyceride synthesis, fatty acid oxidation, and very low density lipoprotein assembly and secretion from hepatocytes (12–15). Incubation with GH in vitro increases the activity of phosphatidate phosphohydrolase in primary hepatocyte cultures (16). Therefore, the enhanced triglyceride synthesis and very low density lipoprotein assembly in the liver after GH therapy in vivo must rely on an exogenous fatty acid supply. Liver fatty acid-binding protein (LFABP) belongs to a famReceived October 28, 1997. Address all correspondence and requests for reprints to: Dr. Jan Oscarsson, Department of Physiology, Go¨teborg University, Box 434, Medicinaregatan 1F, S-405 30 Go¨teborg, Sweden. E-mail: [email protected]. * This work was supported by Grants 04P-11010, 8269, and 7142 from the Swedish Medical Research Council; the Novo Nordisk Foundation; the Tore Nilson Foundation, the Åke Wibergs Foundation, and the Magnus Bergvalls Foundation. Presented in part at the Growth Hormone Research Society Conference 1996, London, United Kingdom.

Neither insulin nor insulin-like growth factor I treatment of hypophysectomized rats for 6 –7 days had any effect on LFABP mRNA or LFABP levels. In vitro, bGH dose-dependently increased the expression of LFABP mRNA, but only in the presence of insulin. Insulin alone had a marked dose-dependent effect on LFABP mRNA levels and was of importance for maintaining the expression of LFABP mRNA during the culture. Incubation with bGH increased LFABP mRNA levels within 3 h. GH had no effect on LFABP mRNA levels in the presence of actinomycin D, indicating a transcriptional effect of GH. Incubation with glucagon in vitro decreased LFABP mRNA levels markedly, indicating that glucagon, in contrast to GH, has an effect opposite that of insulin on LFABP mRNA expression. It is concluded that GH is an important regulator of LFABP in vivo and in vitro. In contrast to the effect of GH on insulin-like growth factor I mRNA, the presence of insulin was a prerequisite for the effect of GH on LFABP mRNA expression in vitro. The results emphasize the role of GH in the regulation of hepatic fatty acid metabolism. (Endocrinology 139: 2699 –2709, 1998)

ily of abundant cytosolic small proteins (14 –15 kDa), which includes intestinal, ileal, heart, myelin, and adipocyte FABP as well as retinol- and retinoic acid-binding proteins. These FABPs, except for the last two proteins, are characterized by binding of long chain fatty acids and their coenzyme A esters (17–21). In contrast to the other members of the FABP family, LFABP binds 2 mol fatty acid/mol protein (22). The fatty acids bound to LFABP are preferentially long chain, unsaturated fatty acids. Moreover, LFABP has been shown to bind heme and a number of eicosanoids with high affinity and to bind a large number of other amphipathic ligands with lower affinity, such as bilirubin, bile salts, lysophosphatidylcholine, steroids, and their metabolites (17, 20, 21). Several enzyme activities, especially those involved in microsomal fatty acid metabolism, were stimulated by LFABP (17). LFABP may play a role as an intracellular acceptor of fatty acids, thereby enhancing fatty acid uptake and facilitating intracellular transport (17, 23–27). Rat LFABP complementary DNA (cDNA) was cloned by Gordon and co-workers (28), and the complete nucleotide sequence of the gene was found to contain four exons and three introns (17, 18). LFABP mRNA levels in the liver are developmentally regulated, with a marked increase during the first 24 h after birth and during puberty in the rat (29, 30), whereas LFABP protein levels increase more gradually (31, 32). There is no or very small diurnal variation in LFABP levels (17, 32). LFABP expression is induced by a high fat diet and peroxisome proliferators, such as bezafibrate and clofibric acid (17). LFABP expression has also been shown to be

2699

2700

GH AND LFABP

influenced by hormones, but only in vivo. Female rats have higher levels than male rats (17, 26, 27), and this difference was reversed by a combination of gonadectomy and administration of gonadal steroids (26). Other studies on the hormonal regulation of LFABP have relied on measurements of the binding capacity of the cytosol fraction containing LFABP (33–35). GH has been shown to increase the expression of LFABP mRNA in Hx male rats (30), and the binding capacity of the cytosol fraction containing LFABP was increased in Hx rats given GH compared with that in Hx control rats (35). In the present study the effects of bovine GH (bGH), and other hormones on the expression of LFABP and its mRNA were studied in vivo in Hx rats. Cytosolic LFABP concentrations were measured with an antibody sandwich enzymelinked immunosorbent assay (ELISA). Primary cultures of hepatocytes were used to study the direct effect of bGH and other hormones on the expression of LFABP mRNA. Insulinlike growth factor I (IGF-I) mRNA levels were measured as an indication of GH responsiveness and the characteristics of GH regulation of LFABP mRNA. Rat hepatocytes were cultured on a basement membrane matrix (Matrigel) derived from extraction of the Engelbreth-Holm-Swarm sarcoma (36). This system, which allows serum-free conditions, has been used to study in vitro effects of GH on cytochrome P-450 enzymes and IGF-I mRNA (37). Moreover, LFABP mRNA and LFABP expressions have been shown to be induced by peroxisome proliferators when hepatocytes were cultured on Matrigel, but not when they were maintained on collagencoated plates (38). Materials and Methods In vivo hormonal treatment Sprague-Dawley rats (Mollegaard Breeding Center, Ejby, Denmark) were used. Hypophysectomy was performed at 50 days of age at Mollegaard Breeding Center. Intact normal rats were age matched. The rats were maintained under standardized conditions of temperature (24 –26 C) and humidity (50 – 60%), with lights on between 0500 –1900 h. The rats had free access to standard laboratory chow (rat and mouse standard diet, B&K Universal, Sollentuna, Sweden) and water. Hormonal treatment started 7–10 days after hypophysectomy. If not otherwise stated, all Hx rats were given cortisol phosphate (400 mg/kgzday; Solu-Cortef, Upjohn, Puurs, Belgium) and l-T4 (10 mg/kgzday; Nycomed, Oslo, Norway) diluted in saline as a daily sc injection at 0800 h (39). Recombinant bGH was a gift from American Cyanamide Co. (Princeton, NJ). The hormone was diluted in 0.05 m phosphate buffer, pH 8.6, with 1.6% glycerol and 0.02% sodium azide. bGH (in most experiments 1 mg/ kgzday) was given either continuously by means of an Alzet osmotic minipump 2001 (Alza Corp., Palo Alto, CA) that was implanted sc on the back of the rat or as two daily sc injections at 12-h intervals (0800 and 2000 h) (39). Recombinant human IGF-I was supplied by Genentech (South San Francisco, CA). IGF-I (1.25 mg/kgz day) was diluted in saline and given as a continuous infusion by means of osmotic minipumps (Alzet 2001) (40). Insulin (100 U/ml; Insulatard, Novo Nordisk, Copenhagen, Denmark) was diluted in saline and given as a daily sc injection at 1600 h. The insulin dose was gradually increased (41): days 1 and 2, 1.0 U/day; days 3 and 4, 2.0 U/day; and days 5–7, 3.5 U/day. The treatments continued for 6 –7 days before the rats were decapitated, trunk blood was collected, and the livers were removed. The livers were cut into pieces, immediately frozen in liquid nitrogen, and stored at 270 C until assays.

Hepatocyte cultures and hormones used in vitro Hepatocytes were prepared by a nonrecirculating collagenase perfusion through vena porta of 200- to 300-g female Sprague-Dawley rats

Endo • 1998 Vol 139 • No 6

essentially as previously described (14). The rats were anesthetized by a combination of xylazine (9 mg/kg; Rompun, Bayer, Lever-Kusen, Germany) and ketamine hydrochloride (77 mg/kg; Ketalar, ParkeDavis, Detroit, MI). The perfusion started with Hanks’ Balanced Salt Solution without calcium and magnesium, pH 7.4 (Life Technologies, Paisley, Scotland) supplemented with 0.6 mm EGTA, 20 mm HEPES, and 10 mm sodium hydrogen carbonate. This solution was followed by perfusion with Williams’ E medium with Glutamax (catalogue no. 32551, Life Technologies) supplemented with penicillin (50,000 IU/liter), streptomycin (50 mg/liter; Life Technologies), 0.28 mm sodium ascorbate (Sigma Chemical Co., St. Louis, MO), 0.1 mm sodium selenite (Sigma Chemical Co.), and 300 – 400 mg/liter collagenase A (Boehringer Mannheim, Mannheim, Germany), or 400 –500 mg/liter collagenase type IV (Sigma Chemical Co.). The first medium was given for 5– 6 min at a flow rate of 40 –50 ml/min, and the second medium was given for 7–9 min at a flow rate of 40 ml/min. The perfusion media were kept at 37 C and infused continuously with 95% oxygen and 5% carbon dioxide. After the perfusion, the cells were filtered through a 250-mm pore size mesh nylon filter followed by a 100-mm pore size mesh nylon filter. The cells were washed by centrifugation at 50 3 g three times for 1 min each time at 4 C in Williams’ E medium with Glutamax supplemented as described above (except for collagenase), but also supplemented with 3 g glucose/ liter and insulin (16 nm 5 ;91 mg/liter 5 ;2600 mU/liter; Actrapid, Novo Nordisk; 28.6 U/mg). The cells were counted in a Burker chamber. The viability was about 90%, as determined by trypan blue exclusion at the start of the experiment and about 98 –100% at the end of the experiments. The cells were seeded at a density of approximately 170,000 cells/cm2 in 10 ml medium on plastic dishes (56.7 cm2; Nunclon, Nalge Nunc International, Copenhagen, Denmark) coated with 500 ml Matrigel (Collaborative Research, Medical Products, Bedford, MA). The cells were plated during the first 16 –18 h in the same medium as that used for washing the cells. After 16 –18 h of culture, the cells were cultured in this medium but with different doses of insulin (Actrapid, Novo Nordisk), bGH (American Cyanamide Co.), or glucagon (Sigma Chemical Co.). Actinomycin D (5 mg/ml; Sigma Chemical Co.) was dissolved in dimethylsulfoxide (DMSO). When actinomycin D was used, all culture dishes were cultured in medium containing 0.15% DMSO. The total culture time was 4 –5 days in all experiments. The medium was changed every day.

Production of antiserum Rabbits (Russian, Mollegaard Breeding Center) were immunized with purified recombinant LFABP, which was a gift from Dr. D. Cistola, Washington University School of Medicine (St. Louis, MO). One hundred micrograms of LFABP were diluted in saline, mixed with equal volumes of Freund’s complete or incomplete adjuvant, and injected im. Blood was collected from the ear vein (20 –30 ml every other week). The serum was shown to contain anti-LFABP antibodies by double immunodiffusion and Western blot. The IgG fraction was isolated by HiTrap protein G affinity columns according to the manufacturer (Pharmacia Biotech, Uppsala Sweden).

Cytosol preparation Cytosol was prepared from rat liver. The tissue was homogenized in 2 ml homogenization buffer/g tissue according to the method of Ockner et al. (27) containing protease inhibitors (Complete Protease inhibitor cocktail tablets, Boehringer Mannheim). The homogenate was centrifuged for 20 min at 12,000 3 g. The supernatant was centrifuged for 1 h at 105,000 3 g. The total protein concentration of liver cytosol was determined by the method of Lowry et al. (42). The cytosol was stored at 270 C until assay.

Western blot Two micrograms of cytosol protein were added to each lane in a buffer containing 62.5 mm Tris-HCl (pH 6.8), 2.3% SDS, 10% glycerol, 0.001% bromophenol blue, and 5% b-mercaptoethanol. The samples were run overnight in 15% polyacrylamide gels containing SDS. The molecular mass standard See Blue (Novex, San Diego, CA) was used. The proteins were transferred to a polyvinyldifluoride membrane (Millipore) by semidry blotting (Multiphore II, Pharmacia). The membrane was then incubated with

GH AND LFABP LFABP antiserum (1:200). Immunoreactive protein was visualized by chemiluminescence using an alkaline phosphatase-conjugated second antibody and AMPPD (disodium 3-(4-methoxyspiro(1,2-dioxetane-3,2’-tricyclo-[3.3.1.13,7]decan)-4-yl)phenyl phosphate) as substrate (Tropix, Bedford, MA). The filters were exposed to ECL film (Amersham, Aylesbury, UK) at room temperature for 5–15 sec, and the films were subsequently developed.

ELISA An antibody sandwich ELISA was developed to measure soluble LFABP in liver cytosol. The purified IgG fraction of the antirat LFABP antiserum was used. Biotinylation of the IgG fraction was performed using the Protein Biotinylation System (Life Technologies, Gaithersburg, MD). The ELISA was standardized using recombinant LFABP dissolved in rat heart cytosol. The protein concentration of recombinant LFABP was determined according to the method of Lowry et al. (42). Rat heart cytosol was used as a carrier because it contains no or very low levels of LFABP. About 1–3 ng/ml or 1 ng/mg of a protein that cross-reacted with the antibody in heart cytosol were detected. The standard liver cytosol was obtained from a female rat (400 g) and contained 351 ng LFABP/ml or 63 ng LFABP/mg protein. The internal standard was obtained from a female rat (200 g) and contained 54 ng LFABP/mg protein. Microtiter plates (Costar, Cambridge, MA) were coated with the IgG fraction of the antiserum at a concentration of 10 mg/ml in 50 ml PBS with 0.05% sodium azide. The plates were incubated at room temperature overnight or at 4 C for at least 2 days before use. The plates were rinsed with distilled water and incubated with 200 ml blocking buffer (PBS with 0.05% Tween-20 and 0.25% BSA, pH 7.2) for 30 min at room temperature and then rinsed with distilled water. The samples were added to the plate in a volume of 50 ml diluted in blocking buffer and incubated for 2 h at room temperature. The plates were washed with distilled water, incubated with 200 ml blocking buffer for 10 min, rinsed again before 50 ml secondary antibody was added (biotinylated IgG, 400 ng/ml dissolved in PBS with 0.25% BSA, pH 7.2), and incubated for another 2 h at room temperature. The washing procedure was repeated, and streptavidin-alkaline phosphatase (Life Technologies) was added at a 1:8000 dilution in PBS with 0.25% BSA, pH 7.2, and incubated for 1.5 h. The plates were washed again and incubated with pNPP (Life Technologies) according to the manufacturer’s directions. After 1 h, the plates were analyzed at 405 nm using a microtiter plate reader (model 450, Bio-Rad, Richmond, CA). The detection limit of the assay was 0.15 ng/ml, or 7.5 ng absolute. The intraassay coefficient of variation (CV) was 5%. The interassay CV was 17% when the plates were assayed on different days, and the CV between plates assayed on the same day was 8%.

Probe synthesis LFABP. The pJG418 plasmid containing a 515-bp insert in the pBR 322 (28) was a gift from Prof. J. Gordon, Washington University School of Medicine. The fragment obtained by cleavage of the 515-bp insert with PvuII and EcoRI (336 bp) was recloned into the pSP72 vector (Promega, Madison, WI). This new insert was used for the synthesis of a [32P]CTP cDNA probe (Megaprime, Amersham) for Northern blots. For solution hybridization assay, the pSP72 vector was linearized with PvuII, and a [35S]UTP complementary RNA probe (antisense) was generated with T7 RNA polymerase (standard transcription protocol, Promega). The standard (sense) used in the solution hybridization assay was generated with SP6 RNA polymerase (Promega) using the same vector, linearized with EcoRI. IGF-I. A pSP64 vector (Promega) with a 153-bp genomic subclone of mouse IGF-I corresponding to exon 3 (by analogy to human IGF-I) was used (43). The structure of this probe would allow detection of both forms of IGF-I mRNA. The hybridization signal in the solution hybridization assay was compared with that of a synthetic mRNA standard using an another construct in which the 153-bp fragment was inserted in the same vector in the opposite direction (44).

Northern blot Total RNA was prepared according to the method of Chomczynski and Sacchi 1987 (45). Twenty micrograms of RNA were electrophoresed

2701

in an agarose (1%)-formaldehyde (0.66 m) gel. The RNA was transferred to a membrane (Hybond-N, Amersham) with a vacuum transfer system (LKB, Stockholm, Sweden) and baked at 80 C for 3 h. The membranes were prehybridized for 4 h at 42 C in a buffer containing 50% formamide, 25 mm HNa2PO4, 25 mm H2NaPO4, 5 3 SSC (standard citrate solution), 0.1% SDS, 1 mm EDTA, 0.05% BSA, 0.05% Ficoll, 0.05% polyvinylpyrrolidone, 200 mg/ml calf liver RNA, and 200 mg/ml salmon sperm DNA and hybridized for 12–14 h at 42 C in the prehybridization buffer, with the addition of a 32P-labeled LFABP cDNA probe. Filters were washed once in 2 3 SSC-0.2% SDS at 42 C for 30 min and once 0.1 3 SSC-0.2% SDS at 42 C for 30 min. Autoradiography was performed at 270 C using Fuji medical x-ray film (Fuji Photo Film Co., Tokyo, Japan).

Solution hybridization Total nucleic acids were prepared according to the method of Durnam and Palmiter (46), and RNA was prepared according to the method of Chomczynski and Sacchi (45) from frozen liver and hepatocytes. The DNA content in the samples was analyzed as described by Labarca and Paigen (47), and the RNA content was determined spectrophotometrically at 260 nm. Aliquots of the prepared total nucleic acids or RNA samples were mixed with hybridization solution containing the 35S-labeled LFABP or IGF-I complementary RNA probes, 0.6 m NaCl, 22 mm Tris-HCl (pH 7.5), 5 mm EDTA, 0.1% (wt/vol) SDS, 0.75 mm dithiothreitol, 7.5% (wt/vol) transfer RNA, and 25% (vol/vol) formamide in a total incubation volume of 40 ml and hybridized at 70 C overnight (46). The samples were then treated with 40 mg ribonuclease A (RNase A) and 2 mg RNase T1 in the presence of 100 mg herring sperm DNA for 45 min at 37 C in a volume of 1 ml. Protected probe was precipitated with 100 ml 6 m trichloroacetic acid (TCA). The precipitate was collected on glass-fiber filters (GF/C, Whatman International, Maidstone, UK) and counted in a scintillation counter. The signal was compared with a standard curve obtained by hybridization of in vitro transcribed LFABP mRNA or IGF-I mRNA. The intraassay CVs calculated from duplicates were 9% for the LFABP mRNA measurements and 14% for the IGF-I mRNA measurements. The results are expressed as attamoles of LFABP or IGF-I mRNA per mg DNA or RNA.

Other methods Serum insulin concentrations were determined by RIA (Phedabas, Pharmacia), and serum glucose concentrations were determined by the glucose-6-phosphate dehydrogenase method (Merck, Darmstadt, Germany) (40).

Statistics Values are expressed as the mean 6 sem. Comparisons between groups were made using one- or two-way ANOVA; comparisons between individual groups were made using the Student-Newman-Keuls multiple range test. The values were transformed to logarithms when appropriate.

Results Effects of GH in vivo

Northern blot analysis revealed one band with the expected size (;1 kb) of LFABP mRNA in the rat liver (Fig 1). Northern blots of RNA preparations from adipose tissue, skeletal muscle, or heart did not reveal any transcripts, but RNA from the intestine contained a single band of similar size (data not shown). These results were expected from the known tissue distribution of LFABP mRNA (17, 20, 21, 29). Figure 1 shows that liver from a Hx female rat treated for 7 days with T4 and cortisol had less LFABP mRNA than liver from a normal female rat or livers from Hx rats treated with, in addition to T4 and cortisol, bGH as two daily injections or as a continuous infusion. The specificity of the antiserum raised against LFABP was

2702

GH AND LFABP

Endo • 1998 Vol 139 • No 6

FIG. 1. Northern blot analysis of LFABP mRNA in liver. Twenty micrograms of total RNA were electrophoresed, transferred, and hybridized with a 32P-labeled LFABP cDNA probe under the conditions described in Materials and Methods. Lane 1, Normal female rat; lane 2, Hx female rat given T4 and cortisol; lane 3, Hx female rat given T4, cortisol, and bGH as two daily injections; lane 4, Hx female rat given T4, cortisol, and bGH as a continuous infusion. All hormones were given for 7 days.

FIG. 2. Western blot analysis of LFABP in liver cytosol. Two micrograms of protein were electrophoresed in 15% polyacrylamide gels containing SDS and transferred to a polyvinyldifluoride membrane. The membrane was incubated with LFABP antiserum, and the bands were visualized as described in Materials and Methods. The position of the molecular mass standard (lysozyme, 16 kDa; See Blue, Novex) is indicated. Lanes 1 and 2 represent a different blot than lanes 3 and 4. Lane 1, Liver from a normal female rat; lanes 2 and 3, liver from the same Hx female rat given T4 and cortisol; lane 4, liver from a Hx female rat given, in addition to T4 and cortisol, bGH as two daily injections. All hormones were given for 7 days.

tested by Western blotting. The detected band corresponded to the size marker with a molecular mass of 16 kDa (lysozyme; Fig. 2) and migrated as recombinant rat LFABP (data not shown). The expression of LFABP was lower in a Hx rat treated with T4 and cortisol and higher in a Hx rat treated with bGH as two daily injections for 7 days (Fig. 2). To quantify the changes in LFABP mRNA and LFABP levels, a solution hybridization RNase protection assay and an antirat LFABP antibody sandwich ELISA were used. Hx female rats were treated with T4 and cortisol for 3 days and thereafter given a single sc injection of 2 mg/kg bGH. The effects on LFABP mRNA, LFABP protein, and IGF-I mRNA levels were followed after the injection (Fig. 3, A–C). Three hours after bGH injection, LFABP mRNA levels had increased significantly. IGF-I mRNA levels increased in a similar manner as LFABP mRNA, but IGF-I mRNA levels had decreased to basal levels 24 h after the injection, in contrast to LFABP mRNA levels (Fig. 3, A and C). No obvious effect of the bGH injection on the expression of LFABP protein was observed in liver cytosol (Fig. 3B). The changes in LFABP levels were similar when the amount of LFABP was calculated as milligrams of LFABP per g liver (data not shown).

FIG. 3. The effect of one sc injection of bGH (2 mg/kgzday) on the expression of LFABP mRNA (A), LFABP (B), and IGF-I mRNA (C) in Hx female rats that had been pretreated for 3 days with T4 (10 mg/kgzday) and cortisol (400 mg/kgzday). Control rats were not given bGH, and the other groups of rats were killed 1, 3, 6, or 24 h after the bGH injection. The concentrations of LFABP was determined with ELISA, and LFABP and IGF-I mRNA levels were determined with a solution hybridization assay as described in Materials and Methods. There were four or five rats in each group. Values are the mean 6 SEM. Values with different superscripts are significantly different from each other (P , 0.05, by one-way ANOVA followed by StudentNewman-Keuls test).

Thus, it is not likely that an increased amount of certain major cytosolic proteins as a result of the bGH injection could have blunted an increase in LFABP levels. Next, the effects of 7 days of hormonal treatment of Hx female rats were investigated. Hx rats given no hormonal treatment were compared with Hx rats given combined T4 and cortisol treatment and, in addition, bGH as two daily injections (Table 1). T4 and cortisol treatment resulted in a

GH AND LFABP

2703

TABLE 1. Effects of T4 and cortisol (T4C) and additional treatment with GH on LFABP and its mRNA in Hx female rats Treatment

Wt gain (g/day)

LFABP (ng/mg protein)

LFABP mRNA (amol/mg RNA)

Hx Hx 1 T4C Hx 1 T4C 1 GH

0.1 6 0.1 0.6 6 0.2 5.2 6 0.3

10.5 6 0.4a 14 6 1b 30 6 2c

133 6 13a 136 6 28a 272 6 42b

The effects of replacement therapy with T4 (10 mg/kgzday) and cortisol phosphate (C; 400 mg/kgzday) and the additional treatment with bGH (1 mg/kgzday) of Hx female rats on the expression LFABP and its mRNA were investigated. The rats were Hx at 50 days of age, and hormonal treatment was started 7–10 days later. T4C was given as a daily sc injection; bGH was given as two daily sc injections. The hormonal treatment was given 7 days before death. The concentrations of LFABP and LFABP mRNA were determined as described in Materials and Methods. Values are the mean 6 SEM. There were seven rats in each group. In each column, values with different superscripts are significantly different from each other (P , 0.05, by one-way ANOVA followed by Student-Newman-Keuls test). TABLE 2. Effects of GH with and without additional treatment with T4 and cortisol (T4C) on LFABP and its mRNA in Hx female rats Treatment

Wt gain (g/day)



Hx 1 T4C Hx 1 GH Hx 1 T4C 1 GH

1.0 6 0.1a 5.2 6 0.4b 5.4 6 0.2b

31 6 4a 44 6 4b 62 6 6b

113 6 10a 310 6 64b 287 6 30b

The rats were Hx at 50 days of age, and hormonal treatment was started 7–10 days later. The rats were treated for 7 days with T4 (10 mg/kgzday) and cortisol phosphate (C; 400 mg/kgzday). Bovine GH (1 mg/kgzday) was given as a continuous infusion by means of osmotic minipumps. LFABP concentrations in cytosol and LFABP mRNA levels were determined as described in Materials and Methods. Values are the mean 6 SEM. There were five rats in each group. In each column, values with different superscripts are significantly different from each other (P , 0.05, by one-way ANOVA followed by StudentNewman-Keuls test).

small, but significant, increase in LFABP levels, but had no effect on LFABP mRNA levels. The combination of T4, cortisol, and bGH increased LFABP levels nearly 3-fold and increased its mRNA 2-fold (Table 1). To test the possibility that the effect of bGH on LFABP levels was influenced by treatment with T4 and cortisol, bGH therapy alone was compared with combined T4, cortisol and bGH therapy in Hx female rats (Table 2). GH treatment alone increased LFABP levels 1.5-fold and mRNA levels 2.7-fold, whereas combined T4, cortisol, and bGH treatment increased LFABP levels 2-fold and mRNA levels 2.5-fold (Table 2). In summary, these results show that GH has the capacity to increase LFABP levels without any additional pituitary-dependent hormones, but it cannot be excluded that T4 and cortisol have some influence on LFABP protein levels. Different doses of bGH were given as a continuous infusion to Hx female rats also treated with T4 and cortisol (Fig. 4). The lowest dose given (0.1 mg/kgzday) resulted in a 40% increase in LFABP concentrations. The highest doses (1 and 5 mg/kgzday) resulted in a normalization of the LFABP concentration (Fig. 4A). As shown in Fig. 4B, an increased bGH dose resulted in an increased weight gain of the rats. The next experiment was designed to compare the effects of Hx on cytosolic concentrations of LFABP in male and female rats as well as the effect of different modes of bGH

FIG. 4. The effects of different doses of bGH on the concentration of LFABP in liver cytosol (A) and the daily weight gain (B) of hypophysectomized female rats. Age-matched female rats served as normal controls (N). All Hx rats were treated with T4 (10 mg/kgzday) and cortisol phosphate (C; 400 mg/kgzday). bGH was given as a continuous infusion by means of osmotic minipumps in three doses (0.1, 1, and 5 mg/kgzday) for 7 days. The concentration of LFABP in liver cytosol was determined with ELISA as described in Materials and Methods. There were four or five rats in each group. Values are the mean 6 SEM. Bars with different superscripts are significantly different from each other (P , 0.05, by one-way ANOVA followed by Student-NewmanKeuls test).

administration (Fig. 5). Several sexually dimorphic liver functions have been shown to be regulated by the sexually dimorphic secretory pattern of GH (4, 5, 7, 15, 48). To investigate whether the secretory pattern of GH could influence the levels of LFABP, Hx female rats were given two daily injections of bGH, mimicking the male secretory pattern, or a continuous infusion of GH, mimicking the female secretory pattern. Female rats had 2.3-fold higher levels of LFABP than male rats (Fig. 5). The effect of Hx was marked in female rats, but not in male rats. However, in other experiments (data not shown), a significant decrease in LFABP levels after Hx was observed in male rats. Two daily injections and a continuous infusion of bGH had similar effects on the expression of LFABP, resulting in 4.8- and 3.8-fold increases, respectively (Fig. 5). Effects of GH and insulin in vitro

To study the direct effects of GH on the expression of LFABP mRNA, primary hepatocyte cultures were used. In

2704

GH AND LFABP

FIG. 5. The concentrations of LFABP in liver cytosol of normal male and female rats (N), Hx male and female rats given T4 (10 mg/kgzday) and cortisol (C; 400 mg/kgzday), as well as Hx female rats given, in addition to T4 and cortisol, bGH as two daily injections (GH32) or as a continuous infusion by means of osmotic minipumps (GH c). Hormones were given for 7 days before death. The concentration of LFABP was determined by ELISA as described in Materials and Methods. There were six rats in each group. Values are the mean 6 SEM. Bars with different superscripts are significantly different from each other (P , 0.05, by one-way ANOVA followed by StudentNewman-Keuls test).

the first experiment, the cells were treated with 16 nm insulin during the entire culture and were treated with bGH during the last 24 h. The highest dose of bGH was also given during the last 24 h of culture to cells that had not been treated with insulin the last 3 days of culture (Fig. 6A). Incubation with 100 ng bGH/ml increased the LFABP mRNA level 2.3-fold. However, when no insulin was given during the last 3 days of culture, the level of LFABP mRNA was low despite bGH treatment, indicating an effect of insulin on LFABP mRNA levels (Fig. 6A). In a separate experiment (Fig. 6B), a different dose of bGH was given during the last 24 h to each cell culture dish. All culture dishes were incubated with 16 nm insulin during the entire culture period. The result of this experiment indicated that 100 –200 ng bGH/ml results in a maximal effect on LFABP mRNA expression. Thus, GH has a marked effect on LFABP mRNA levels in vitro, and this effect was dependent on the presence of insulin. In a control experiment, the level of LFABP mRNA in fresh isolated hepatocytes was compared with the level of LFABP mRNA after 16 h and 4 days of culture in the presence or absence of insulin (data not shown). The level of LFABP mRNA declined to about 60% of the level in fresh isolated cells after 16 h of culture in 16 nm insulin. In the presence and absence of 3 nm insulin, the LFABP mRNA levels were 34% and 14%, respectively, of the level in fresh isolated cells after 4 days of culture. Thus, insulin lessens the decline in LFABP mRNA gene expression during culture. To further study the interaction between insulin and GH in the regulation of LFABP mRNA, hepatocytes were cultured during the first 16 –18 h with 16 nm insulin and thereafter were treated with different doses of insulin in the presence or absence of bGH (100 ng/ml) during the next 3 days of culture (Fig. 7, A and B). bGH treatment had no effect on LFABP mRNA levels when the cells were not treated with insulin (Fig. 7A), whereas bGH treatment resulted in 2.5-fold higher IGF-I mRNA levels in the absence of insulin (Fig. 7B). When the data were analyzed with two-way ANOVA, it was

Endo • 1998 Vol 139 • No 6

FIG. 6. Effects of different doses of bGH on the expression of LFABP mRNA in primary cultures of rat hepatocytes. Hepatocytes were isolated from a female rat and cultured as described in Materials and Methods. A shows an experiment in which three groups of cell culture dishes were cultured in the presence of 16 nM insulin (1) during the entire culture. One group (2) was only treated with insulin during the first 16 –18 h (plating). Thereafter, insulin was withdrawn from the medium. bGH was given for the last 24 h in two different doses (20 and 100 ng/ml). B shows an experiment in which all cell culture dishes was treated with 16 nM insulin during the entire culture, and each culture dish was given a different dose of bGH during the last 24 h of culture (0, 2, 10, 20, 50, 100, 200, and 500 ng/ml). Total nucleic acids were isolated, and LFABP mRNA was measured with a solution hybridization assay as described in Materials and Methods. The values are the mean 6 SEM. In A, there were three dishes in each group. Bars with different superscripts are significantly different from each other (P , 0.05, by one-way ANOVA followed by Student-NewmanKeuls test).

found that both insulin and bGH treatment had a significant effect (P , 0.05) on LFABP mRNA and IGF-I mRNA levels. However, insulin had a marked effect on LFABP mRNA levels, but a small effect on IGF-I mRNA levels. GH, on the other hand, had a more marked effect on IGF-I mRNA than on LFABP mRNA levels. The combined action of 100 ng/ml bGH and 3 nm insulin resulted in a 10-fold increase in LFABP mRNA and a nearly 6-fold increase in IGF-I mRNA levels (Fig. 7, A and B). The time-dependent effects of GH and insulin on LFABP mRNA and IGF-I mRNA levels are shown in Fig. 8, A and B. One group of cultured cells was given 3 nm insulin during the last 3 days of culture. The fourth day of culture, bGH (100 ng/ml) was added to the medium for 1, 3, or 24 h. LFABP and IGF-I mRNA levels increased within 3 h of incubation (Fig. 8). After 3 days of culture without insulin, other cultured cells were given 3 nm insulin for 1, 3, or 24 h. LFABP mRNA levels increased between 3 and 24 h of insulin treatment, whereas IGF-I mRNA levels were not affected by insulin in this experiment (Fig. 8). To determine whether the effect of bGH on LFABP mRNA levels was transcriptional, hepatocytes were cultured in the presence of bGH (100 ng/ml) and actinomycin D for 6 h (Fig. 9). Incubation with actinomycin D alone did not affect LFABP mRNA levels, but the effect of bGH on LFABP mRNA levels was blunted in the presence of actinomycin D. Effects of insulin and IGF-I in vivo

To determine whether insulin has an effect on LFABP mRNA and LFABP expression in vivo, insulin was given in

GH AND LFABP

FIG. 7. Effects of different doses of insulin in the presence (open triangles) or absence (filled squares) of bGH (100 ng/ml) on the expression of LFABP mRNA (A) and IGF-I mRNA (B) in primary cultures of hepatocytes. Hepatocytes were isolated from a female rat and cultured as described in Materials and Methods. All dishes were treated with 16 nM insulin during the first 16 –18 h. The cells were treated for 3 days with the indicated doses of insulin and bGH. Total nucleic acids were isolated, and LFABP mRNA and IGF-I mRNA were measured with a solution hybridization assay as described in Materials and Methods. The values are the mean 6 SEM. There were three dishes in each group. There was a significant effect of insulin and bGH treatment on LFABP and IGF-I mRNA levels (P , 0.05, by two-way ANOVA followed by Student-Newman-Keuls test).

increasing doses for 7 days to Hx female rats (Table 3). Hx rats given T4 and cortisol had lower serum insulin levels than normal rats, and bGH treatment increased serum insulin levels in the Hx rats. The insulin-treated Hx rats had similar or higher serum insulin levels than the Hx rats given bGH alone. The Hx rats given insulin alone had the lowest serum glucose levels. Insulin treatment and bGH treatment had opposite effects on serum glucose levels, and when both

2705

FIG. 8. Time-dependent effects of insulin and GH on the expression of LFABP mRNA (A) and IGF-I mRNA (B) in primary cultures of hepatocytes. Hepatocytes were isolated from a female rat and cultured as described in Materials and Methods. All dishes were treated with 16 nM insulin during the first 16 –18 h. The cells that were given bGH (open triangles; 100 ng/ml) for the indicated periods (1, 3, and 24 h) were treated with 3 nM insulin during the last 3– 4 days of culture. Insulin (filled squares; 3 nM) was given for the indicated periods (1, 3, and 24 h). RNA was isolated, and LFABP mRNA and IGF-I mRNA were measured with a solution hybridization assay as described in Materials and Methods. Values are the mean 6 SEM. There were three or four dishes in each group. Values with different superscripts are significantly different from each other (P , 0.05, by one-way ANOVA followed by Student-Newman-Keuls test).

hormones were given together, intermediate serum glucose levels were observed. Insulin treatment had no effect on LFABP, LFABP mRNA, or IGF-I mRNA levels (Table 3). However, bGH treatment (1 mg/kgzday) had the expected effects on IGF-I mRNA and LFABP expression (Table 3). Although there are few or no IGF type 1 receptors in the liver (49), the effect of recombinant human IGF-I was tested in vivo, because IGF-I serum levels increase markedly as a result of GH therapy and, therefore, might affect other hormonal systems [such as an altered insulin secretion (50, 51)] that could influence LFABP expression (Table 4). However, there was no effect of 6 days of IGF-I treatment of Hx rats on either LFABP or its mRNA in the liver (Table 4).

2706

GH AND LFABP

FIG. 9. The effect of actinomycin D (Act.D) on bGH induction of LFABP mRNA. Hepatocytes from a normal female rat were cultured as described in Materials and Methods. All culture dishes were incubated with 3 nM insulin during the last 3 days of culture. Actinomycin D (5 mg/ml) and bGH (100 ng/ml) were given for 6 h. All culture dishes were cultured in the presence of 0.15% DMSO during the last 6 h of culture. LFABP mRNA was measured with a solution hybridization assay as described in Materials and Methods. Values are the mean 6 SEM. There were three or four dishes in each group. Values with different superscripts are significantly different from each other (P , 0.05, by one-way ANOVA followed by Student-Newman-Keuls test).

Effect of glucagon in vitro

One possible explanation for the lack of effect of insulin treatment in vivo on LFABP expression could be increased serum concentrations of insulin antagonistic hormones, which might have an effect on LFABP mRNA expression opposite that of insulin. One possible candidate for such an effect is glucagon, whose serum concentration is expected to be high when serum glucose levels decrease. Cultured cells were treated during the last 24 h of culture with 100 nm (348 ng/ml) glucagon. This treatment resulted in a marked decrease in LFABP mRNA levels (Fig. 10). Discussion

The lipolytic effect of GH in the rat results in an increased mobilization of fatty acids to the circulation (1, 2). Long chain fatty acids have been shown to function as hormones activating the peroxisome proliferator activated receptor (PPAR) (52, 53). Activation of the PPAR results in enhanced LFABP mRNA and LFABP levels as well as induction of several mitochondrial and peroxisomal enzymes involved in fatty acid oxidation in the liver (38, 54). Thus, GH could potentially increase LFABP expression in vivo via an increased flux of fatty acids to the liver. However, the present results show that GH has a direct effect on LFABP gene expression via its own receptor on hepatocytes. GH rapidly increased LFABP mRNA, but only in the presence of insulin. The lack of effect of GH on LFABP mRNA without insulin treatment was not due to a failure of the cells to respond to GH, as IGF-I mRNA levels increased 2.5-fold with GH alone. These results suggest that the LFABP gene, in contrast to the IGF-I gene, is dependent upon insulin for transcriptional regulation. Apart from the additive effects of insulin and GH on LFABP

Endo • 1998 Vol 139 • No 6

mRNA, the different time course for the induction of LFABP mRNA indicates different mechanisms of action of these hormones on LFABP mRNA expression. The normalization of LFABP levels after GH therapy of Hx rats suggests that GH is a major regulator of LFABP expression. The increased concentration of LFABP in liver cytosol after GH therapy seems to at least in part be dependent on an increased amount of LFABP mRNA, as the mRNA and protein levels increased in parallel after a week of GH treatment. There was, however, no increase in the expression of LFABP 24 h after a single dose of bGH, although there was a marked effect on LFABP mRNA levels. Thus, an effect of GH on LFABP expression is dependent upon the presence of GH for a longer period of time. Induction of LFABP by peroxisome proliferators has also been shown to have a slow onset (17, 38). As GH treatment results in increased insulin secretion (55, 56) and increased serum concentrations of insulin, the in vivo effect of GH on LFABP levels may be the result of the additive effects of GH and insulin on LFABP mRNA expression. The lack of effect of insulin in vivo argues against this possibility. However, the rats given insulin had lower serum glucose concentrations, indicating increased serum concentrations of nonpituitary-dependent insulin antagonists such as glucagon. The inhibitory effect of glucagon in vitro indicates that the lack of effect of insulin in vivo could at least partly be explained by increased serum glucagon levels. Diabetic rats have been shown to have decreased cytosolic levels of LFABP, which were partly reversed by insulin treatment (33). Thus, it is possible that insulin also has effects on LFABP expression in vivo, but only in insulin-deficient rats. The insulin concentrations in portal blood have been shown to be 0.34 nm (;50 mU/liter) in the fasting rat (57). This finding indicates that the portal levels of insulin are markedly higher than those in peripheral blood and that the doses used in the hepatocyte cultures were within or near physiological levels. The GH doses used were in the physiological range, as indicated by the normal mean plasma levels of GH (50 –100 ng/ml) and the mean pulse heights of GH in plasma (150 – 300 ng/ml) in the adult rat (58). In the rat, GH secretion has been shown to decrease as a result of food deprivation for a period of 24 h or longer (59). Together with low levels of insulin and high levels of glucagon in serum, decreased GH secretion would result in decreased hepatic LFABP levels. However, parallel decreases in the content of LFABP and total cytosol protein have been observed after prolonged fasting (17), indicating that LFABP is not specifically decreased by fasting. The continuous and irregular secretion of GH in the adult female rat and the regular episodic secretion of GH in the adult male rat have been mimicked in several studies by sc administration of GH as a continuous infusion and two daily injections of GH, respectively (4, 5, 15, 39, 48). In contrast to many other sexually dimorphic hepatic functions in the rat, the higher levels of LFABP in female rats were not dependent upon the secretory pattern of GH. Therefore, the sex difference in LFABP levels could be dependent on direct effects of the gonadal steroids on the hepatocyte or, alternatively, occur via indirect effects on the A and B cells of the pancreatic islets. Thus, it has been shown that estradiol increases the

GH AND LFABP

2707

TABLE 3. Effects of GH and insulin on LFABP, LFABP mRNA, and IGF-I mRNA in Hx female rats Treatment

Wt gain (g/day)

Insulin (mU/liter)

Glucose (mmol/liter)



IGF-I mRNA (amol/mg RNA)

N HxT4C Hx 1 T4C 1 insulin Hx 1 T4C 1 GH Hx 1 T4C 1 insulin 1 GH

3.1 6 1.9a 20.3 6 0.1c 0.9 6 0.1d 5.5 6 0.1b 5.5 6 0.2b

18.8 6 1.4a 6.0 6 0.4c 59.8 6 25.0a,b 30.0 6 9.1a 104.0 6 55.5b

8.1 6 0.3b 6.7 6 0.2b 3.4 6 0.4c 6.7 6 0.2b 4.8 6 0.8a

79 6 7b 30 6 3a 29 6 2a 77 6 10b 58 6 4b

590 6 61b 106 6 6c 130 6 21c 351 6 30a 284 6 30a

20 6 1b 2 6 0.3a 4 6 2a 18 6 1.5b 15 6 2b

Female rats were Hx at 50 days of age. Age-matched female rats served as normal controls (N). After 7–10 days of observation, the rats were hormonally treated for 7 days. All Hx rats were given T4 (10 mg/kgzday) and cortisol phosphate (C; 400 mg/kgzday) as a daily sc injection. Bovine GH (1 mg/kgzday) was given as two daily sc injections. Insulin (Insulatard) was given as a daily sc injection at 1600 h. The dose was gradually increased: days 1 and 2, 1.0 U/day; days 3 and 4, 2.0 U/day; and days 5–7, 3.5 U/day. The analyses were performed as described in Materials and Methods. There were five or six rats in each group. Values are the mean 6 SEM. Values with different superscripts are significantly different from each other (P , 0.05, by one-way ANOVA followed by Student-Newman-Keuls test). TABLE 4. Effects of recombinant human IGF-I on LFABP and its mRNA in Hx female rats Treatment

Wt gain (g/day)



Hx 1 T4C Hx 1 T4C 1 IGF-I

21.2 6 0.2a 1.4 6 0.2b

16 6 1 13 6 1

156 6 18 126 6 13

Hypophysectomy was performed when the rats were 50 days old. Hormonal treatment started 7–10 days later. All Hx rats were treated with T4 (10 mg/kgzday) and cortisol phosphate (C; 400 mg/kgzday). IGF-I (1.25 mg/kgzday) was given sc by means of osmotic minipumps. After 6 days of hormonal treatment, the rats were killed, and the livers were excised. LFABP concentrations in cytosol and LFABP mRNA levels were determined as described in Materials and Methods. There were seven rats in each group. Values are the mean 6 SEM. Values with different superscripts are significantly different from each other (P , 0.05, by one-way ANOVA followed by StudentNewman-Keuls test).

FIG. 10. The effect of glucagon (100 nM) on the expression of LFABP mRNA in primary cultures of hepatocytes. Hepatocytes were isolated from a female rat and cultured as described in Materials and Methods. All dishes were treated with 16 nM insulin the first 16 –18 h. Thereafter, hormones were withdrawn for 2 days, and glucagon was given during the last 24 h of culture. RNA was isolated, and LFABP mRNA was determined with a solution hybridization assay as described in Materials and Methods. Values are the mean 6 SEM. There were four dishes in each group. Values with different superscripts are significantly different from each other (P , 0.05, by Student’s t test).

relative insulin to glucagon molar ratio in the portal vein (57). In line with the present results, this effect of estradiol would result in increased expression of LFABP mRNA in the liver. The effect of GH on LFABP mRNA was compared with the effect on IGF-I mRNA, as GH is a known regulator of IGF-I mRNA in hepatocytes in vivo and in vitro (37, 43). Insulin has

also been shown to play a role in the regulation of IGF-I mRNA in vitro (37, 60). In line with these studies, we observed a 2-fold potentiation of the GH-induced expression of IGF-I mRNA by insulin treatment. We observed small or no effects of insulin alone on the expression of IGF-I mRNA. The half-maximal effective dose of insulin observed by others (0.47 nm) (60) is in line with our results. The lack of effect of insulin therapy of Hx rats on IGF-I mRNA could be due to a relatively small effect of insulin on IGF-I gene expression and also to the fact that the rats were not insulin deficient. The mechanism of induction of LFABP mRNA by GH is at least partly an increased transcription rate of the gene, as indicated by the lack of effect of GH in the presence of actinomycin D. IGF-I mRNA levels increase after GH therapy as a result of an increased transcription rate (43, 61). The induction of IGF-I mRNA in vivo by a single dose of GH was shown to be transient, in contrast to the induction of LFABP mRNA. This finding may reflect a transient increase in serum GH levels, resulting in an increased transcription followed by degradation of IGF-I mRNA. The maintained LFABP mRNA levels in the same experiment indicates slower turnover of LFABP mRNA than of IGF-I mRNA. The slow induction of LFABP mRNA by insulin treatment will be of interest to study further. It may be that the increase in LFABP mRNA levels caused by insulin treatment reflects an increased stability of the transcript, and the permissive action of insulin on the GH induction of LFABP mRNA may be due to this effect of insulin. Cytosolic LFABP levels have previously been measured by immunological methods using radial immunodiffusion (17, 27) and direct noncompetitive ELISA (32). Similar amounts of LFABP in the liver of mature (;60 days of age) male and female rats were observed in this study and in other studies using immunological methods (17, 27). We used the Lowry procedure for determination of protein concentrations. Overestimation of LFABP levels has been reported to occur when purified LFABP is measured according to the Lowry procedure (32). Thus, it cannot be excluded that we overestimated the amount of LFABP, especially if there was no overestimation of the total protein concentration in cytosol. Nonimmunological measurements of the amount of LFABP in liver cytosol have relied on the property of LFABP to bind various ligands. There are sometimes discrepancies between the nonimmunological and immunological measurements of the amount LFABP (32). Nonimmunological methods have several disadvantages compared with the immunological meth-

2708

GH AND LFABP

ods, such as competition with endogenous ligands, nonspecific binding, binding stoichiometry of LFABP, and the possibility that the ligands bind to other proteins in the low mol wt protein fraction. However, the results obtained using a ligand binding assay to determine the amount of LFABP in liver cytosol after hypophysectomy and GH therapy were similar to the results obtained in this study (35). In summary, GH was shown to have a marked effect on LFABP regulation. Small or no effects of thyroid hormones, glucocorticoids, insulin, and IGF-I could be detected in vivo using Hx rats. In contrast to other sexually dimorphic hepatic proteins, LFABP was not differently regulated by the maleand female-specific secretory patterns of GH. The induction of LFABP by GH in vivo was accompanied by an increase in LFABP mRNA, indicating a pretranslational regulation of LFABP expression by GH. However, a posttranslational regulation cannot be excluded. The in vitro studies indicate a direct hormonal regulation of LFABP mRNA expression involving physiological concentrations of insulin and GH. Thus, in addition to the known in vitro activation of LFABP via the PPARa receptor and its presumed ligands, long chain fatty acids (19, 38, 52–54), insulin, and GH receptors are involved in the regulation of LFABP gene expression. We cannot exclude the possibility that fatty acids mobilized as a result of GH treatment activate the PPARa receptor and augment the direct effect of GH on the hepatocyte on LFABP expression. The role of LFABP in the diverse effects of GH on hepatic lipid and lipoprotein metabolism (1, 2, 4 –16) is still unclear. It could be speculated that GH increases intracellular trafficking of fatty acids in hepatocytes at least partly via a direct effect on the expression of LFABP. Acknowledgments We thank Barbro Basta for excellent technical assistance. We also thank Prof. Jeffery Gordon and Dr. Dave Cistola at Washington University School of Medicine (St. Louis, MO) for their kind supply of the pJG 418 plasmid and recombinant LFABP. We thank Dr. Agnetha Mode for her advice regarding long term cultures of primary rat hepatocytes.

11. 12. 13. 14.

15.

16. 17. 18. 19. 20.

21. 22. 23. 24. 25. 26. 27.

References

28.

1. Goodman H, Schwartz J 1974 Growth hormone and lipid metabolism. In: Knobil E, Sawyer WH (eds) Handbook in Physiology. American Physiological Society, Washington DC, vol 4:211–232 2. Davidson MB 1987 Effect of growth hormone on carbohydrate and lipid metabolism. Endocr Rev 8:115–131 3. Etherton TD, Louveau I, Tang Sorensen M, Chaudhuri S 1993 Mechanisms by which somatotropin decreases adipose tissue growth. Am J Clin Nutr [Suppl] 58:287S–295S 4. Edeń S, Jansson J-O, Oscarsson J 1987 Sexual dimorphism of growth hormone secretion. In: Isaksson O, Binder C, Hall K, Ho¨kfelt B (eds) Growth Hormone: Basic and Clinical Aspects. Elsevier, Amsterdam, pp 129 –151 5. Oscarsson J, Edeń S 1988 Sex differences in fatty acid composition of rat liver phosphatidylcholine are regulated by the plasma pattern of growth hormone. Biochim Biophys Acta 959:280 –287 6. Angelin B, Rudling M 1994 Growth hormone and hepatic lipoprotein metabolism. Curr Opin Lipid 5:160 –165 7. Edeń S, Oscarsson J 1995 Lipids and growth hormone. In: Saggese G, Stanhope R (eds) Recent Advances in Growth and Growth Hormone Therapy. Freund, London, pp 253–266 8. Clejan S, Schulz H 1986 Effect of growth hormone on fatty acid oxidation: growth hormone increases the activity of 2:4-dienoyl-CoA reductase in mitochondria. Arch Biochem Biophys 246:820 – 828 9. Maddaiah VT, Clejan S 1986 Growth hormone and liver mitochondria: time course of effects on respiration and fatty acid composition in hypophysectomized rats. Endocrinology 119:250 –252 10. Bornstein J, Ng FM, Heng D, Wong KP 1983 Metabolic actions of pituitary growth hormone. I. Inhibition of acetyl CoA carboxylase by human growth

29. 30. 31. 32.

33. 34. 35. 36. 37.

Endo • 1998 Vol 139 • No 6

hormone and a carboxyl terminal part sequence acting through a second messenger. Acta Endocrinol (Copenh) 103:479 – 486 Donkin SS, McNall AD, Swencki BS, Peters JL, Etherton TD 1996 The growth hormone dependent decrease in hepatic fatty acid synthase mRNA is the result of a decrease in gene transcription. J Mol Endocrinol 16:151–158 Elam MB, Simkevitch CP, Solomon SS, Wilcox HG, Heimberg M 1988 Stimulation of in vitro triglyceride synthesis in the rat hepatocyte by growth hormone treatment in vivo. Endocrinology 122:1397–1402 Elam MB, Wilcox HG, Solomon S, Heimberg M 1992 In vivo growth hormone treatment stimulates secretion of very low density lipoprotein by the isolated perfused rat liver. Endocrinology 131:2717–2722 Sjo¨berg A, Oscarsson J, Bostro¨m K, Innerarity TL, Edeń S, Olofsson S-O 1992 Effects of growth hormone on apolipoprotein B (apoB) messenger ribonucleic acid editing, and apoB 48 and apoB 100 synthesis and secretion in the rat liver. Endocrinology 130:3356 –3364 Sjo¨berg A, Oscarsson J, Boreń J, Edeń S, Olofsson S-O 1996 Mode of growth hormone administration influences triacylglycerol synthesis and assembly of apolipoproteinB-containing lipoproteins in cultured rat hepatocytes. J Lipid Res 37:275–289 Pittner RA, Bracken P, Fears R, Brindley DN 1986 Insulin antagonises the growth hormone-mediated increase in the activity of phosphatidate phosphohydrolase in isolated rat hepatocytes. FEBS Lett 202:133–136 Bass NM 1988 The cellular fatty acid binding proteins: aspects of structure, regulation and function. Int Rev Cytol 3:143–184 Materese V, Stone RL, Waggoner DW, Bernlohr DA 1989 Intracellular fatty acid trafficking and the role of cytosolic lipid binding proteins. Prog Lipid Res 28:245–272 Kaikaus RM, Chan WK, Ortiz de Montellano PR, Bass NM 1993 Mechanisms of regulation of liver fatty acid binding protein. Mol Cell Biochem 123:93–100 Glatz JFC, Vork MM, Cistola DP, van der Vusse GJ 1993 Cytoplasmic fatty acid binding protein: significance for intracellular transport of fatty acids and putative role on signal transduction pathways. Prostaglandins Leukotrienes Essent Fatty Acids 48:33– 41 Veerkamp JH, van Kuppevelt HMSM, Maatman RGHJ, Prinsen CFM 1993 Structural and functional aspects of cytosolic fatty acid binding proteins. Prostaglandins Leukotrienes Essent Fatty Acids 49:887–906 Thompson J, Winter N, Terwey D, Bratt J, Banaszak L 1997 The crystal structure of liver fatty acid-binding protein. A complex with two bound oleates. J Biol Chem 272:7140 –7150 Luxon BA, Weisiger RA 1993 Sex difference in intracellular fatty acid transport: role of cytoplasmic binding proteins. Am J Physiol 265:G831–G841 Prows DR, Murphy EJ, Schroeder F 1995 Intestinal and liver fatty acid binding proteins differentially affect fatty acid uptake and esterification in L-cells. Lipids 30:907–910 Murphy EJ, Prows DR, Jefferson JR, Schroeder F 1996 Liver fatty acid binding protein expression in transfected fibroblasts stimulates fatty acid uptake and metabolism. Biochim Biophys Acta 1996:191–198 Ockner RK, Lysenko N, Manning JA, Monroe SE, Burnett DA 1980 Sex steroid modulation of fatty acid utilization and fatty acid binding protein concentration in rat liver. J Clin Invest 65:1013–1023 Ockner RK, Manning JA, Kane JP 1982 Fatty acid binding protein: isolation from rat liver, characterization, and immunological quantification. J Biol Chem 257:7872–7878 Gordon JI, Alpers DH, Ockner RK, Strauss AW 1983 The nucleotide sequence of rat liver fatty acid binding protein mRNA. J Biol Chem 258:3356 –3363 Gordon JI, Elshourbagy N, Lowe JB, Liao WS, Alpers DH, Taylor JM 1985 Tissue specific expression and devolopmental regulation of two genes coding for rat fatty acid binding protens. J Biol Chem 260:1995–1998 Berry SA, Yoon J-B, List J, Seelig S 1993 Hepatic fatty acid binding protein mRNA is regulated by growth hormone. J Am Coll Nutr 6:638 – 642 Sheridan M, Wilkinson TCI, Wilton DC 1987 Studies on fatty acid binding proteins: changes in the concentration of hepatic fatty acid binding protein during development in the rat. Biochem J 242:919 –922 Paulussen RJA, Geelen MJH, Beynen AC, Veerkamp JH 1989 Immunological quantitation of fatty acid binding proteins. I. Tissue and intracellular distribution, postnatal development and influence of physiological conditions on rat heart and liver FABP. Biochim Biophys Acta 1001:201–209 Brandes R, Arad R 1983 Liver cytosolic fatty acid binding proteins: effect of diabetes and starvation. Biochim Biophys Acta 750:334 –339 Nakagawa S, Kawashima Y, Hirose A, Kozuka H 1994 Regulation of hepatic level of fatty acid binding protein by hormones and clofibric acid in the rat. Biochem J 297:581–584 Singer SS, Henkels K, Deucher A, Barker M, Singer J Trulzsch DV 1996 Growth hormone and aging change rat liver fatty acid binding protein levels. J Am Coll Nutr 15:169 –174 Bissell DM, Arenson DM, Maher JJ, Roll FJ 1987 Support of cultured hepatocytes by a laminin-rich gel: evidence for functionally significant subendothelial matrix in normal rat liver. J Clin Invest 79:801– 812 Tollet P, Enberg N, Mode A 1990 Growth hormone (GH) regulation of cytochrome P-450IIC12, insulin-like growth factor-I (IGF-I), and GH receptor messenger RNA expression in primary rat hepatocytes: a hormonal interplay with insulin, IGF-I, and thyroid hormone. Mol Endocrinol 4:1934 –1942

GH AND LFABP 38. Brandes R, Kaikaus RM, Lysenko N, Ockner RK, Bass NM 1990 Induction of fatty acid binding protein by peroxisome proliferators in primary hepatocyte cultures and its relationship to the induction of peroxisomal b-oxidation. Biochim Biophys Acta 1034:53– 61 39. Oscarsson J, Olofsson S-O, Bondjers G, Edeń S 1989 Differential effects of continuous vs. intermittent administration of growth hormone to hypophysectomized female rats on serum lipoproteins and their apoproteins. Endocrinology 125:1638 –1649 40. Sjo¨berg A, Oscarsson J, Olofsson S-O, Edeń S 1994 Insulin-like growth factor-I and growth hormone have different effects on serum lipoproteins and secretion of lipoproteins from cultured rat hepatocytes. Endocrinology 135:1415–1421 41. Gause I, Isaksson O, Lindahl A, Edeń S 1985 Effect of insulin treatment of hypophysectomized rats on adipose tissue responsiveness to insulin and growth hormone. Endocrinology 116:945–951 42. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ 1951 Protein measurement with the Folin phenol reagent. J Biol Chem 193:265–275 43. Mathews LS, Norstedt G, Palmiter R 1986 Regulation of insulin-like growth factor I gene expression by growth hormone. Proc Natl Acad Sci USA 83:9343–9347 44. Isgaard J, Mo¨ller C, Isaksson OGP, Nilsson A, Mathews LS, Norstedt G 1988 Regulation of insulin-like growth factor messenger ribonucleic acid in rat growth plate by growth hormone. Endocrinology 122:1515–1520 45. Chomczynski P, Sacchi N 1987 Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162:156 –159 46. Durnam D, Palmiter R 1983 A practical approach for quantitating specific mRNAs by solution hybridization. Anal Biochem 131:385–393 47. Labarca C, Paigen K 1980 A simple, rapid, and sensitive DNA assay procedure. Anal Biochem 102:344 –352 48. Waxman DJ, Pampori NA, Ram PA, Agrawal AK, Shapiro BH 1991 Interpulse interval in circulating growth hormone patterns regulates sexually dimorphic expression of hepatic cytochrome P450. Proc Natl Acad Sci USA 88:6868 – 6872 49. Coro JF, Poulos J, Ittop O, Pories WJ, Flickinger EG, Sinha MK 1988 Insulinlike growth factor I binding in hepatocytes from human liver, human hepatoma, and normal, regenerating, and fetal rat liver. J Clin Invest 81:976 –981 50. Sieradzki J, Fleck H, Chatterjee AK, Schatz H 1988 Stimulatory effect of

51. 52. 53. 54. 55. 56. 57.

58. 59. 60. 61.

2709

insulin-like growth factor-I on 3H-thymidine incorporation, DNA content and insulin biosynthesis and secretion of isolated pancreatic rat islets. J Endocrinol 177:59 – 62 Leahy JL, Vandekerkhove KM 1990 Insulin-like growth factor I at physiological concentrations is a potent inhibitor of insulin secretion. Endocrinology 126:1593–1598 Go¨ttlisher M, Widmark E, Li Q, Gustafsson J-Å 1992 Fatty acids activate a chimera of the clofibric acid-activated receptor and the glucocorticoid receptor. Proc Natl Acad Sci USA 89:4653– 4657 Forman BM. Chen J, Evans RM 1997 Hypolipidemic drugs, polyunsaturated fatty acids, and eicosanoids are ligands for peroxisome proliferator-activated receptors a and d. Proc Natl Acad Sci USA 94:4312– 4317 Wahli W, Braissant O, Desvergne B 1995 Peroxisome proliferator activated receptors: transcriptional regulators of adipogenesis, lipid metabolism and more. Chem Biol 2:261–266 Pierluissi J, Pierluissi R, Ashcroft SJH 1980 Effects of growth hormone on insulin release in the rat. Diabetologia 19:391–396 Nielsen JH 1982 Effects of growth hormone, prolactin, and placental lactogen on insulin content and release, and deoxyribonucleic acid synthesis in cultured pancreatic islets. Endocrinology 110:600 – 606 Mandour T, Kissebah AH, Wynn V 1977 Mechanism of oestrogen and progesterone effects on lipid and carbohydrate metabolism: alteration in the insulin:glucagon molar ratio and hepatic enzyme activity. Eur J Clin Invest 7:181–187 Jansson J-O, Ekberg S, Isaksson OGP, Edeń S 1984 Influence of gonadal steroids on age- and sex-related secretory patterns of growth hormone. Endocrinology 114:1287–1294 Tannenbaum GS, Rorstad O, Brazeau P 1979 Effects of prolonged food deprivation on the ultradian growth hormone rhythm and immunoreactive somatostatin tissue levels in the rat. Endocrinology 104:1733–1738 Bo¨ni-Schnetzler M, Schmid C, Meier PJ, Froesch ER 1991 Insulin regulates insulin-like growth factor I mRNA in rat hepatocytes. Am J Physiol 260:E846 –E851 Johnson TR, Rudin SD, Blossey BK, Ilan J, Ilan J 1991 Newly synthesized RNA: simultaneous measurement in intact cells of transcription rates and RNA stability of insulin-like growth factor I, actin, and albumin in growth hormone stimulated hepatocytes. Proc Natl Acad Sci USA 88:5287–5291