Apolipoprotein gene expression in the rat

We have studied the regulation of rat intestinal and hepatic apolipoprotein gene expression, in vivo, after alterations in thyroid hormone status. When compared to those of chow-fed controls, rates of synthesis of intestinal apoA-I and apoB-48 decreased 6046% in hypothyroid animals and increased threeto fourfold after triiodothyronine (T,) administration. These changes were not accompanied by changes in mRNA abundance. By contrast, intestinal apoA-IV synthesis rates and mRNA abundance were both unaltered over the range of thyroid hormone manipulations tested. Hepatic apoA-I and apoA-IV synthesis rates decreased by 7040% in hypothyroid animals, while synthesis rates and mRNA abundance increased coordinately sixto eightfold when hypothyroid rats were made hyperthyroid. Hepatic apoE synthesis rates increased twofold in hypothyroid rats and decreased sevenfold in hyperthyroid animals. ApoE mRNA abundance, however, was comparable in all groups. Hypothyroid animals had reduced synthesis rates of hepatic apoB100 and apoB-48. After induction of hyperthyroidism, apoB-100 synthesis (studied from 5 to 60 min) was undetectable (< 0.01%) without further change in apoB-48 synthesis and without alterations in either apoB mRNA abundance or transcript size. Despite undetectable hepatic apoB-100 synthesis rates in hyperthyroid animals, total plasma triglyceride secretion rates (after Triton WR1339 injection) were normalized compared to a 50% decrease in hypothyroid rats. UTaken together, the data provide evidence for tissue-specific, independent regulation of apolipoprotein gene expression in vivo. Furthermore, the data suggest that aspects of hepatic triglyceride assembly and secretion and apolipoprotein gene expression may be coordinately responsive to alterations in thyroid hormone status.-Davidson, N. O., R. C. Carlos, M. J. Drewek, and T. G. Parmer. Apolipoprotein gene expression in the rat is regulated in a tissue-specific manner by thyroid hormone. J. Lipid Res. 1988. 29: 151 11522. Supplementary key words apoB-100 apoB-48 triglyceride assembly Alterations in thyroid hormone status produce a wide variety of effects on lipoprotein homeostasis in human subjects and experimental animals (reviewed in reference 1). Many of these alterations have been exploited as a means of manipulating aspects of lipoprotein assembly and secretion. In particular, a combination of cholesterol feeding plus hypothyroidism has been widely used by investigators examining the mechanisms of dietary-induced atherogenic hypercholesterolemia in the rat (2). Observations emerging from studies using this model suggest that hypothyroidism exerts widespread effects on hepatic triglyceride assembly and secretion (3-5), hepatic (5-7) and intestinal cholesteryl ester accumulation (8, 9), and aspects of apolipoprotein metabolism (10, 11). Many of these effects appear to be unique to the hypothyroid state (1 2) and need to be distinguished from the combined effects resulting from hypothyraidism and cholesterol feeding (9, 12). Additionally, hyperthyroidism has been demonstrated in animals to be associated with alterations in hepatic glycerolipid assembly (13, 14), while hyperthyroid human subjects have been shown to have decreased serum low density lipoprotein (LDL) concentrations with alterations in cholesterol synthesis and LDL catabolism ( 1 5, 16). We now report the results of studies in which the effect of altered thyroid hormone status was examined on the tissue-specific accumulation of several rat apolipoprotein mRNAs and in vivo synthesis rates of their primary translation products. Studies were additionally conducted to examine the effects of these alteraAbbreviations: apo, apolipoprotein; CM, chylomicron; VLDL, very low density lipoprotein; LDL, low density lipoprotein; HDL, high density lipoprotein; PBS, phosphate-buffered saline; SDSPAGE, sodium dodecyl sulfate polyacrylamide gel electrophoresis; PMSF, phenylmethylsulfonyluoride; TCA, trichloroacetic acid; SSC (lX), 0.15 M NaCY0.015 M Na citrate, pH 7; EDTA, ethylenediaminetetraacetic acid, Na salt; TLC, thin-layer chromatography; GLC, gas-liquid chromatography; PTU, (2-thio-4-hydroxy-6-npropylpyrimidine); Ts, 3,3’,5-triiodo-~-thyronine. ‘To whom correspondence should be addressed at: Gastroenterology Section, Box 400, Department of Medicine, University of Chicago, 5841 Maryland Avenue, Chicago, IL 60637. Journal of Lipid Research Volume 29,1988 1511 at P E N N S T A T E U N IV E R S IT Y , on F ebuary 2, 2013 w w w .j.org D ow nladed fom tions on aspects of hepatic triglyceride assembly and secretion. The results suggest that thyroid hormone exerts independent, tissue-specific effects on intestinal and hepatic apolipoprotein gene expression. Furthermore, the combination of quantitative approaches suggests that regulation may involve both preand posttranslational mechanisms. Finally, evidence is presented that links several of these alterations to aspects of hepatic triglyceride assembly and secretion. MATERIALS AND METHODS Animals and treatment protocol Male Sprague-Dawley rats were obtained in the weight range of 150 g from Charles River, Wilmington, MA. Animals were housed four per cage and segregated according to specific treatment protocol. Control animals were fed Purina rat chow (RalstonPurina, St. Louis, MO) ad libitum for 2 1-28 days prior to study. The remaining animals were randomly assigned to one of three experimental groups and received chow supplemented with 0.1 % (wtw) propylthiouracil (2-thio-4-hydroxy-6-n-propylpyrimidine, Sigma, St. Louis, MO) as previously described (9) for 21 days. All treated animals were thereby rendered initially hypothyroid, as confirmed by radioimmunoassay (performed by the University of Michigan Veterinary Laboratory) of T4, T,, free Tq, and free T, levels in serum (hypothyroid: T4, 0.10 & 0.08 ng/ml; T,, 0.43 -t 0.01 ng/ml; free T4, 0.17 ? 0.29 pg/ml; free T,, 0.52 +0.36 pg/ml vs. control: T4, 16.7 ? 2.5 ng/ml; Ts, 0.63 +0.06 ng/ml; free T4, 15.1 5 1.6 pg/ ml; free Ts, 2.2 & 0.47 pg/ml). Groups of hypothyroid rats were made chemically euthyroid by seven daily intraperitoneal injections of 3,3,’5-triiodo-~-thyronine (T,), 0.5 pg/lOO g body weight. Other groups were made hyperthyroid by alternate-day injections of 50 pg Td100 g body weight for 7 days. Both groups of T,-treated animals continued to consume the chowPTU diet. Untreated hypothyroid rats were fed chow0.1 % PTU for 28 days. Animals were studied, as detailed below, following a 16-20-hr fast. This period, in the T,-treated euthyroid and hyperthyroid groups, began following the last injection of T,. At the time of killing, control rats weighed 280-300 g while hypothyroid, T,-treated euthyroid and hyperthyroid rats weighed 200-230 g with no differences noted among the three experimental groups. 5 cm from ligament of Treitz) was isolated and pulselabeled with ~-[4,5-’N]leucine (500 Fci, sp act 120 Cil mmol, Amersham, Arlington Heights, IL) for 9 min. This time point effectively precludes any apparent hepatic contribution to the newly synthesized intestinal apolipoprotein pool ( 17). Following exsanguination, the loop was removed and enterocytes were isolated using citrate-EDTA chelation buffers containing sequentially 20,10, and 5 mM leucine as detailed (17) to prevent isotope reutilization. The final washed cell pellet was homogenized on ice in buffer H (PBS1% Triton-2 mM leucine-1 mM PMSF-1 mM benzamidine, pH 7.4) and a 105,000 g supernatant was prepared. These conditions have been previously shown to optimize apolipoprotein recovery from intestinal cells (17). Aliquots of homogenate were saved for measurement of total protein concentration (18) and trichloroacetic acid (TCA)-insoluble radioactivity. Samples of the final 105,000 g supernatant were frozen at 80°C prior to immunoprecipitation with specific antisera as described below. Determination of hepatic apolipoprotein synthesis rates Animals were anesthetized with sodium pentobarbital and received 1 .O mCi ~-[4,5-~H]leucine via intraportal vein injection. At the time intervals detailed in the legends to the figures, animals were exsanguinated via the abdominal aorta, the liver was perfused in situ for 10 min with 75 ml iced PBS-PO mM leucine, and pieces were taken from all lobes for homogenization in five volumes of buffer H. A 225,000 g supernatant was prepared and stored at -80°C prior to immunoprecipitation. Aliquots of liver homogenate were processed as described above. Quantitative immunoprecipitation of intestinal and hepatic apolipoproteins was performed as described ( 17) using monospecific polyclonal antisera directed against rat apoA-I (19), apoA-IV (20), and apoB (17). Anti-albumin antibodies were obtained commercially (ICN, Costa Mesa, CA), and anti-apoE antibodies were a gift from Dr. R. Hay, University of Chicago. Aliquots of cytosolic supernatant were reacted with excess antiserum (as defined below), and the immune complexes were precipitated by addition of washed S. aurm cells (IgG-sorb Enzyme Center, Boston, MA). After extensive washing, the immune complexes were characterized by denaturing SDS-PAGE and radioactivity incorporation into the specific apolipoproteins was determined by liquid scintillation spectrometry following addition of 3% Protosol-Econofluor (NEN, Boston, MA) to gel slices. In all instances, antibody excess stoichiometry was established by exhaustive reprecipitation. Determination of intestinal apolipoprotein synthesis rates Animals were anesthetized with sodium pentobarbital and a 10-cm loop of .jejunum (proximal portion 1512 Journal of Lipid Research Volume 29, 1988 at P E N N S T A T E U N IV E R S IT Y , on F ebuary 2, 2013 w w w .j.org D ow nladed fom Apolipoprotein synthesis rates are expressed as a fraction of total protein synthesis (9, 17). Each value represents the mean of two to six separate assays corrected for nonspecific and background radioactivity. RNA extraction and analysis Total cellular RNA was extracted and pooled (up to four animals per pool) from the proximal (jejunum) small intestine using 8 M guanidine-HC1 as previously described (21). Yields of total cellular RNA averaged 3-6 mg/g wet weight mucosa. Hepatic RNA was similarly extracted, with comparable yield. All preparations of RNA were determined to be intact following analytical methylmercury agarose gel electrophoresis For quantitation of apolipoprotein mRN A abundance, serially diluted aliquots of total cellular RNA (0.5-3.0 pg) were applied to nitrocellulose filters using a commercial template (23). In addition, samples of intestinal and hepatic total cellular RNA (30 ng-3 pg) were run as internal standards for each filter. Filters were probed with various cDNAs (as detailed below) labeled with s2P to a specific activity of 108-109 cpm/ pg (24). Hybridization solutions (pH 7.0) contained 50% (volhol) formamide, 6 X SSC, 50 mM sodium phosphate, 1 mM EDTA, 1 X Denhardt's solution, 50 pg/ml sheared, single-stranded salmon sperm DNA, and 10% (wdvol) dextran sulfate. Following a 24-hr incubation at 42"C, filters were washed twice in 0.1% (wdvol) NaDodSoJl X SSC at 25°C and four times (15 min each) in 0.1 x SSC at 50°C prior to autoradiography. mRNA abundance was calculated by quantitative scanning densitometry using an LKB Laser Densitometer (Ultroscan LX, LKB, Gaithersburg, MD). Data are expressed as absorbance units per pg RNA, each value representing the mean -+ SD of three or four representative pools. Relative abundance was calculated by reference to a standard curve constructed from the signals of hepatic and intestinal RNA standards, thereby allowing comparison between different films. Only signals in the linear range of film sensitivity were used (23). Northern blots of 20p.g total hepatic RNA were prepared as described (25) following fractionation through 6% formaldehyde/0.75% agarose gels. These blots were hybridized as described above with a rat 3' apoB cDNA (26). cDNAs used in these studies include a 2.9 Kb 3' rat apoB fragment (26) (a gift from Drs. J. Elovson and A. Lusis, UCLA Medical Center, CA); rat apoA-I (27), apoA-IV (28). and rat liver fatty acid binding protein (29), (gifts from Dr. J. Gordon, Washington University, St. Louis); rat apoE (30), (agift from Dr. J. Taylor, Gladstone Foundation Laboratories, San Francisco, CA); and human beta actin (31) (a gift from Dr. P. (22). Gunning, VA Medical Center, Stanford University, CA) . Hepatic microsome preparation and analysis After the rats were exsanguinated, portions of the liver were removed and finely minced prior to homogenization in five volumes of buffer I (0.25 M sucrose10 mM Tris1 mM EDTA, pH 7.4) using a loose-fitting Teflon-glass homogenizer. The homogenate was centrifuged at 2000 g for 10 min and the resulting supernatant was centrifuged at 25,000 g for 10 min, both at 4°C. This supernatant was then centrifuged at 100,000 g for 60 min at 4°C and the pellet was suspended in 1 ml 0.5 M KC1-0.25 M sucrose. Following recentrifugation at 100,000 g, the final, washed microsome pellet was resuspended in 2 ml buffer I, aliquoted, and frozen at 80°C. Once thawed, samples were not reused. Samples were submitted to lipid extraction according to the method of Folch et al. (32). Total lipid classes were separated by thin-layer chromatography (TLC) using silica gel G in a solvent of petroleum ether-ethyl ether-glacial acetic acid 80:20: 1 (v/v/v). Triglyceride and free fatty acid bands were identified by comparison to standards, scraped into Teflon-lined screwcapped tubes, and transmethylated directly using 14% BF!, in methanol, following addition of heptadecanoic acid as an internal standard. The derivatized fatty acids were assayed using a Perkin Elmer model 84 10 gas-liquid chromatograph equipped with a 6 ft x 2 mm ID column packed with 10% SP-2330 on 100/120 mesh Chromasorb (Supelco, Bellefonte, PA). Authentic fatty acid methyl esters (Nu-Chek-Prep, Elysian, MN) were used to identify peaks based on their relative retention times. Values are presented as pg fatty acid, normalized to protein content.