GROWTH HORMONE AND PPARα
IN THE REGULATION OF GENES INVOLVED IN HEPATIC LIPID METABOLISM
Caroline Améen
Department of Physiology
and Wallenberg Laboratory for Cardiovascular Research
Sahlgrenska Academy at Göteborg University 2004
A doctoral thesis at a university in Sweden is produced either as a monograph or as a collection of papers. In the latter case, the introductory part constitutes the formal thesis, which summarises the accompanying papers. These papers have already been published or are in manuscripts at various stages (in press, submitted or in manuscript).
ISBN 91-628-6006-2
“The most exciting phrase to hear in science, the one that heralds new discoveries, is not Eureka! (I found it!) but rather, "hmm....
that's funny...."
ISAAC ASIMOV
ABSTRACT
GROWTH HORMONE AND PPARα IN THE REGULATION OF GENES INVOLVED IN HEPATIC LIPID METABOLISM.
Caroline Améen, Department of Physiology and Wallenberg Laboratory for
Cardiovascular Research, Göteborg University, Sahlgrenska University Hospital, SE- 413 45 Göteborg, Sweden (2004).
Growth hormone (GH) plays a key role in the regulation of lipid and lipoprotein metabolism. Its sexually dimorphic secretory pattern regulates many sex-differentiated functions in the liver, such as triglyceride synthesis and VLDL secretion. GH also increases insulin secretion, and the importance of increased insulin levels for the effects of GH in vivo was therefore investigated in hypophysectomised (Hx) rats. GH increased the hepatic triglyceride secretion rate and triglyceride content, as well as fatty acid synthase (FAS), stearoyl-CoA desaturase-1 (SCD-1) and sterol regulatory element binding protein-1c (SREBP-1c) mRNA expression. Insulin suppressed the effect of GH on hepatic triglyceride secretion rate and content, but this was not through changed gene expression of lipogenic enzymes or microsomal triglyceride transfer protein (MTP), the rate-limiting protein in VLDL assembly. The regulation of lipogenic genes and MTP by the sex-differentiated GH secretory pattern was studied both in Hx rats administered GH in a mode that mimics either the female or male plasma pattern of GH, and in intact males feminised with respect to the plasma pattern of GH. SREBP-1c, FAS, glycerol-3- phosphate acyltransferase (GPAT) and MTP levels were higher in females compared to males and specifically upregulated by the female-like GH plasma pattern in Hx rats. The expression of SCD-1 mRNA was not sex-differentiated and increased by GH irrespective of administration mode. Only FAS and GPAT mRNA levels were increased in males with feminised GH plasma pattern, possibly due to decreased insulin sensitivity. Increased expression of FAS, GPAT and MTP could therefore help to explain the previously described stimulatory effects of female sex and GH secretory pattern on VLDL assembly and secretion. Hepatic MTP expression and activity were also increased by the peroxisome proliferator-activated receptor α (PPARα) agonist WY 14,643 (WY), both in vivo (mice and rats) and in vitro (primary mouse and rat hepatocyte cultures). The increase in MTP expression was paralleled by a change in apoB-100 secretion, which shows that the stimulatory effect of WY on apoB-100 secretion could be mediated by MTP.
Key words: Growth hormone, insulin, SREBP-1, lipogenic enzymes, microsomal triglyceride transfer protein, PPARα, LXRα, triglycerides, liver.
LIST OF PUBLICATIONS
This thesis is based on the following papers, which will be referred to in the text by their Roman numerals:
I. Interaction between growth hormone and insulin in the regulation of lipoprotein metabolism in the rat.
Fredrik Frick*, Daniel Lindén*, Caroline Améen, Staffan Edén and Jan Oscarsson.
Am J Physiol Endocrinol Metab 283: E1023-E1031, 2002.
II. Effects of gender and growth hormone secretory pattern on sterol regulatory element binding protein-1c and its downstream genes in rat liver.
Caroline Améen, Daniel Lindén, Britt-Mari Larsson, Agneta Mode, Agneta Holmäng and Jan Oscarsson.
Submitted.
III. Sex difference in hepatic microsomal triglyceride transfer protein expression is determined by the growth hormone secretory pattern in the rat.
Caroline Améen and Jan Oscarsson.
Endocrinology 144(9):3914-3921, 2003.
IV. PPARα activation increases microsomal triglyceride transfer protein expression and activity in the liver.
Caroline Améen, Ulrika Edvardsson, Anna Ljungberg, Lennart Asp, Anna Tuneld, Sven-Olof Olofsson, Daniel Lindén and Jan Oscarsson.
Submitted.
*Both authors contributed equally to this article.
ABBREVIATIONS
ACC acetyl-CoA carboxylase
Apo apolipoprotein
APOBEC-1 apoB mRNA-editing enzyme catalytic peptide-1 bHLH-Zip basic helix-loop-helix leucine zipper
BMI body mass index
CETP cholesterol ester transfer protein
Cis cytokine-inducible SH2-containing protein CPT-I carnitine palmitoyl transferase-I
DGAT diacylglycerol acyltransferase
DR direct repeat
ER endoplasmic reticulum FAS fatty acid synthase FFA free fatty acid
GH growth hormone
GHBP growth hormone-binding protein GHRH growth hormone-releasing hormone GPAT glycerol-3-phosphate acyltransferase HDL high-density lipoprotein
HL hepatic lipase
HMG-CoA 3-hydroxy-3-metylglutaryl-CoA HNF hepatocyte nuclear factor
Hx hypophysectomy/hypophysectomised IDL intermediate-density lipoprotein IGF-I insulin-like growth factor-I IRS insulin receptor substrate
JAK janus kinase
LCAT lecithin-cholesterol acyltransferase LDL low-density lipoprotein
LPL lipoprotein lipase
LRP LDL receptor-related protein receptor
LXR liver X receptor
MAPK mitogen-activated protein kinase
MTP microsomal triglyceride transfer protein PDI protein disulfide isomerase
PLA2 phospholipase A2
PPAR peroxisome proliferator-activated receptor PPRE peroxisome proliferator response element PUFA polyunsaturated fatty acid
RXR retinoic X receptor S1P site-1 protease S2P site-2 protease
SCAP SREBP cleavage-activating protein SCD stearoyl-CoA desaturase
SEM standard error of the mean
SOCS suppressors of cytokine signalling SR-BI scavenger receptor class B type I SRE sterol regulatory element
SREBP sterol regulatory element binding protein STAT signal transducer and activator of transcription
T4 thyroxine
VLDL very-low density lipoprotein
WY WY 14,643
TABLE OF CONTENTS
ABSTRACT………..…..
4LIST OF PUBLICATIONS………...
5ABBREVIATIONS………...…….
6TABLE OF CONTENTS………...…
8INTRODUCTION………..
10GENERAL INTRODUCTION………. 10
LIPOPROTEIN METABOLISM………...……….… 10
Lipoproteins………...… 10
Exogenous lipoprotein pathway……….… 11
Endogenous lipoprotein pathway………..… 12
Atherogenic dyslipidemia……… 13
APOB AND VLDL………...… 14
ApoB……… 14
VLDL assembly……….… 15
MICROSOMAL TRIGLYCERIDE TRANSFER PROTEIN……… 16
Regulation of MTP……… 17
STEROL REGULATORY ELEMENT BINDING PROTEINS……… 17
The two-step cleavage process of SREBPs………..… 18
SREBP-1c and regulation of lipogenesis………. 19
LIPOGENIC ENZYMES………... 19
Acetyl-CoA carboxylase……….. 19
Fatty acid synthase………...… 20
Stearoyl-CoA desaturase……….…… 20
Glycerol-3-phosphate acyltransferase………..… 21
PEROXISOME PROLIFERATOR-ACTIVATED RECEPTORS……...… 21
PPARα and lipid metabolism………. 22
GROWTH HORMONE………..…… 22
The GH secretory pattern in females and males……….… 23
GH receptor and GH signalling……….… 24
Influence of the GH secretory pattern on GH signalling………..… 25
Insulin-like and diabetogenic effects of GH……… 25
GH and lipoprotein metabolism in humans………. 26
GH and lipoprotein metabolism in laboratory animals……… 26
AIMS OF THE THESIS………
28METHODOLOGICAL CONSIDERATIONS ………
29ANIMALS………...… 29
Hypophysectomised rats……….. 29
HORMONAL TREATMENT………..… 29
Thyroxine (T4)………... 29
Glucocorticoids……….………… 29
Growth hormone………... 30
Insulin………. 31
Sex steroids……… 31
HEPATOCYTE CULTURES………... 32
SUMMARY OF RESULTS………...
34PAPER I……….. 34
PAPER II………. 35
PAPER III……….….. 36
PAPER IV………... 37
GENERAL DISCUSSION………...
39REGULATION OF LIPOGENESIS AND VLDL SECRETION BY GH…………..……… 39
REGULATION OF LIPOGENESIS AND VLDL SECRETION BY INSULIN………..… 41
REGULATION OF SREBP-1c AND LIPOGENIC ENZYMES……… 43
REGULATION OF MTP - POSSIBLE MECHANISMS OF GH ACTION………44
IMPORTANCE OF CHANGED MTP LEVELS FOR THE EFFECTS OF PPARα……….… 46
MTP AND DIFFERENT REGULATION OF APOB-48 AND APOB-100……….… 47
SUMMARY AND CONCLUSIONS………
49ACKNOWLEDGEMENTS……… …...
51REFERENCES………...
54INTRODUCTION
GENERAL INTRODUCTION
Obesity and type 2 diabetes are major health problems in Western society. They increase in prevalence in both developed and developing countries, and are major risk factors for cardiovascular disease (CVD). One key feature of these conditions is abnormal lipid and lipoprotein metabolism. Growth hormone (GH) is known to play an important role in the regulation of lipoprotein metabolism, and both GH deficiency and GH excess are associated with a 50% increased risk for CVD [1, 2]. The secretion of GH from the pituitary is moreover altered in both obesity and type 2 diabetes [3], indicating changed GH action in these conditions. Knowledge about the importance of changed GH secretion in the regulation of lipid and lipoprotein metabolism is therefore crucial to better understand the development of dyslipidemia and ultimately CVD.
LIPOPROTEIN METABOLISM
Lipoproteins
Lipids play an essential role in energy metabolism, and their transport between tissues is therefore important to maintain energy balance in the body. As lipids are highly hydrophobic molecules with limited water solubility, they are transported in plasma associated with other molecules. Free fatty acids are bound to albumin, while larger lipids are circulating in plasma bound to lipoproteins. Lipoproteins are spherical water- soluble particles consisting of lipids and specialised proteins called apolipoproteins.
Their ability to carry lipids in aqueous surroundings is due to their amphipathic nature, characterised by a hydrophilic outside and a hydrophobic inside. The inner core mainly consists of triglycerides and cholesterol esters, and is surrounded by a monolayer of phospholipids and unesterified cholesterol in which the apolipoproteins are dispersed.
There are many different types of apolipoproteins; the major ones being apolipoprotein (apo) A-I, A-II, apoB-48, apoB-100, apoC-I, apoC-II, apoC-III and apoE. The apolipoproteins are important for the structural integrity of the lipoprotein (e.g. apoB), but they also function as enzyme activators (e.g. apoA-I and apoC-II) and receptor ligands (e.g. apoB-100 and apoE). There are several distinct lipoprotein types with characteristic lipid and apolipoprotein composition, and they are most often classified in terms of their density. In increasing order of density, the major classes of lipoproteins are: chylomicrons, very low-density lipoprotein (VLDL), low-density lipoprotein (LDL), intermediate-density lipoprotein (IDL) and high-density lipoprotein (HDL) particles (Table 1). Since lipids have a lower density than proteins, the density of the lipoprotein particle is inversely related to its lipid content. The two classes of lipoproteins with the lowest density, i.e. chylomicrons and VLDL, are therefore
triglyceride-rich particles, while the lipoprotein particles with higher density, i.e. IDL, LDL and HDL, contain less triglycerides and are mainly carriers of cholesterol esters.
An alternative way of classifying lipoproteins is to divide them into apoB-containing (chylomicrons, VLDL, IDL and LDL) and non-apoB-containing (HDL) lipoproteins.
Table 1. Density classes and apoprotein distribution of lipoprotein subclasses.
Exogenous lipoprotein pathway
Chylomicrons transport dietary fat from the intestine to peripheral tissues, mainly adipose tissue, heart and muscle (Figure 1). In addition to apoB-48, chylomicrons secreted from enterocytes contain apoA-I, A-II and A-IV, while apolipoproteins C and E are soon acquired in the circulation by exchange with HDL. In the capillaries of peripheral tissues, the triglycerides in chylomicrons are hydrolysed by the action of lipoprotein lipase (LPL). This enzymatic reaction is activated by apoC-II and allows the delivery of free fatty acids to the cells, either to be stored as triglycerides or consumed as energy. After hydrolysis of the core triglycerides, the chylomicron is transformed into a particle of smaller size that is enriched in cholesterol esters. These particles are referred to as chylomicron remnants, and they are rapidly taken up by the liver via the LDL receptor or LDL receptor-related protein receptor (LRP) with apoE serving as ligand. In the liver, cholesterol and fatty acids can be used either for VLDL production, or stored as cholesterol esters and triglycerides, respectively. The fatty acids can also be degraded by β-oxidation, while cholesterol can be excreted in the bile as free cholesterol or as bile acids.
Intestine, liver, catabolism of VLDL and chylomicrons A, C-II, E
5-12 1.063–1.21
HDL
*Both apolipoprotein B-48 and apoB-100 are synthesised in rodent liver, while only apoB-100 is synthesised in human liver.
Catabolism of VLDL B-100
18-25 1.019–1.063
LDL 1.019–1.063 18-25 B-100 Catabolism of VLDL
LDL
Catabolism of VLDL and chylomicrons
B-48/B-100, 25-35 E
1.006–1.019
IDL Catabolism of VLDL and
chylomicrons B-48/B-100,
25-35 E 1.006–1.019
IDL
B-48/B-100*,
C, E Liver
30-80 0.95–1.006
VLDL B-48/B-100*,
C, E Liver
30-80 0.95–1.006
VLDL
Sources Major
apoproteins Size
(nm) Density
(g/ml) Lipoprotein
class Major Sources
apoproteins Size
(nm) Density
(g/ml) Lipoprotein
class
Intestine A, B-48, C, E
80-1200
<0.95
Chylomicrons <0.95 80-1200 A, B-48, C, E Intestine Chylomicrons
Figure 1. Exogenous and endogenous lipoprotein pathways. VLDL, very low-density lipoprotein;
HDL, high-density lipoprotein; IDL, intermediate-density lipoprotein; LDL, low-density lipoprotein;
LPL, lipoprotein lipase; HL, hepatic lipase; LDL-R, LDL receptor; LRP, LDL receptor-related protein receptor; FFA, free fatty acids.
Endogenous lipoprotein pathway
VLDL has a similar role as chylomicrons, but transports triglycerides that are synthesised in the liver (Figure 1). The size of the VLDL particles that are secreted varies, mostly due to the amount of triglycerides in the particle. The main subclasses of VLDL particles are large lipid-rich VLDL 1 and the smaller and denser VLDL 2 that contains less triglycerides. Just like chylomicrons, triglycerides in VLDL particles are hydrolysed by LPL in the capillaries to provide tissues with energy or to build up a triglyceride depot. After LPL-induced lipolysis, the VLDL particle is converted into IDL. IDL can either be taken up by the liver via apoE-recognising receptors, or further metabolised into LDL via a lipolytic enzyme in the liver called hepatic lipase (HL).
LDL particles are taken up by LDL receptors that are present on the cell surface and LPL
LPL
LDL-R
SR-BI
FFA
Adipose tissue Muscle
Intestine
Remnants
Chylomicrons VLDL
IDL LDL
HDL
HL
LRP/ HL
LDL-R Liver
LRP/
LDL-R
LPL LPL
bind lipoproteins containing apoB-100 and apoE. The LDL receptor is expressed on all cells, but the major amount is present in the liver where at least 50% of all LDL catabolism occurs. The LDL receptor is continuously internalised from the cell surface by endocytosis, with or without bound lipoproteins, and then recycled back to the cell surface. The expression of the LDL receptor is controlled by the cholesterol level inside the cell, i.e. when the cell is in need of cholesterol the LDL receptor is upregulated and vice versa. HDL is produced in the liver and small intestine, but can also be formed from surface excess lipids and apolipoproteins during catabolism of VLDL and chylomicrons. Nascent HDL can moreover be formed in the periphery by the action of ABCA1-transporters that pump out phospholipids and cholesterol of the cells to circulating apoA-1, which are acceptors of these lipids. The combined action of apoA-1 and ABCA1-transporters is the first step in a process called reverse cholesterol transport that is responsible for the transfer of cholesterol from peripheral tissues to the liver. The enzyme lecithin-cholesterol acyltransferase (LCAT) present on the surface of HDL converts phosphatidylcholine (lecithin) and cholesterol to cholesterol esters, which enter the core of the HDL particle. Cholesterol-loaded HDL can then be transported to the liver, where it is taken up via scavenger receptor B-I (SR-BI). In humans, cholesterol esters can also be transferred from HDL to apoB- containing lipoproteins in exchange of triglycerides by the action of cholesterol ester transfer protein (CETP). CETP is not present in rodents, however, which contributes to the HDL based lipoprotein profile in rodents as compared to the more LDL based lipoprotein characteristic of humans.
Atherogenic dyslipidemia
Disturbances of lipoprotein metabolism can result in the development of atherosclerosis and subsequently coronary heart disease, which is a major cause of deaths in the Western society. There is a fundamental notion that an elevated serum LDL cholesterol level is a major risk factor for CVD, reflecting an unbalance in forward and reverse cholesterol transport. For example, subjects with familial hypercholesterolemia have raised LDL cholesterol levels due to a mutation in the LDL receptor, resulting in premature atherosclerosis. Although LDL indeed is an important atherogenic lipoprotein, looking only at LDL levels may not be sufficient to identify the risk for CVD. In subjects with insulin resistance or with disorders such as metabolic syndrome and type 2 diabetes mellitus [4], elevated plasma triglycerides (in particular large VLDL 1 particles) together with an increased number of small dense LDL particles and a low level of HDL cholesterol [5, 6] have emerged as a significant lipid risk profile for CVD. These three abnormalities are closely related metabolically as the increase in triglyceride-rich particles generates small dense LDL and low HDL cholesterol levels. This is due to the fact that increased VLDL 1 levels favour the exchange of core lipids between VLDL 1 and LDL and HDL. In this reaction, LDL
and HDL particles are depleted in cholesterol esters and enriched in triglycerides, which makes them good substrates for HL. Core triglycerides in LDL and HDL particles are hydrolysed by HL and small dense LDL and HDL particles are thus formed. Small dense LDL particles are particularly atherogenic as they readily bind to the proteoglycans of the arterial intima and are susceptible to oxidation, a modification that enhance foam cell formation that in turn initiates the development of atherosclerotic plaques. Smaller HDL particles, on the other hand, have an enhanced catabolic rate and the number of circulating HDL particles will therefore be reduced.
Thus, raised VLDL triglycerides levels, small dense LDL and low HDL levels constitute an atherogenic lipoprotein profile. Together with raised LDL cholesterol, these alterations in lipid metabolism represent a fourfold entity that is characteristic of atherogenic dyslipidemia [5].
All these parameters are potential targets of therapy to reduce the risk for CVD. Statins and fibrates are the main drugs that are in clinical use, either by themselves or in combination. Statins are inhibitors of 3-hydroxy-3-methylglutaryl coenzyme A (HMG- CoA) reductase, the rate-limiting enzyme in cholesterol synthesis. As LDL receptors are controlled by the cholesterol level in the cell, statins indirectly upregulate the LDL receptor and are therefore highly effective in lowering LDL cholesterol level. Fibrates, on the other hand, are very efficient agents in lowering the plasma concentration of triglycerides and, as a consequence, also reduce plasma levels of small dense LDL [7].
These compounds are agonists of specific nuclear receptors called peroxisome proliferator-activated receptors (PPARs) and they both inhibit the secretion of VLDL triglycerides and accelerate VLDL clearance (for reviews see [4, 8, 9]).
APOB AND VLDL
ApoB
A single apoB molecule is an integral part of each VLDL particle and is a prerequisite for VLDL production. The apoB structure is unusual in comparison with other lipoproteins, having amphipathic β-strands of importance for lipid binding. ApoB is one of the largest single chain mammalian polypeptides and exists in two distinct forms termed apoB-100 and apoB-48. ApoB-100 represents the full-length protein containing 4536 amino acids (512 kDa), whereas apoB-48 constitutes the amino terminal 2152 amino acids of the full-length form (250 kDa). The molecular weight of apoB-48 is approximately 48% of the full length apoB-100, thereby its name. ApoB- 48 is synthesised from the same primary transcript as apoB-100 by a post- transcriptional modification of a single mRNA nucleotide called mRNA editing. In this editing process, a cytidine at nucleotide position 6666 is deaminated to form uridine, which changes the codon for glutamine (CAA) to a stop codon (UAA) [10,
11]. The translation is therefore terminated at this site and apoB-48 is formed. The apoB mRNA editing process is catalysed by a multicomponent enzyme complex consisting of the catalytic subunit APOBEC-1 (apoB mRNA-editing enzyme catalytic peptide 1) [12] in addition to auxiliary proteins [13]. Editing of apoB mRNA occurs in the small intestine of all mammals, producing apoB-48 molecules that associate with chylomicrons. In the liver, however, there is no editing in humans in contrast to rodents. This results in the exclusive production of the apoB-100 form in human liver, whereas rodents synthesise and secrete both apoB-48 and apoB-100 from the liver (reviewed in [14-16]).
Figure 2. The two-step assembly of apoB-containing particles. MTP, microsomal triglyceride transfer protein; TG, triglycerides; VLDL, very low-density lipoprotein.
VLDL assembly
ApoB-containing VLDL particles are synthesised in hepatocytes in a two-step process [16-18] (Figure 2). In the first step, a partially lipidated apoB molecule called pre- VLDL is formed. This occurs by the addition of small amounts of lipids to apoB during its translation and coincident translocation into the lumen of the rough endoplasmic reticulum (ER) [18-20]. This reaction is catalysed by a lipid transfer protein called microsomal triglyceride transfer protein (MTP), further described below.
If the lipidation is incorrect, apoB will not fold properly and is directed to degradation.
This occurs mainly through retrograde translocation of misfolded apoB from the ER lumen back to the cytosol, where apoB is ubiquitinated and subsequently degraded by the proteasome [21, 22]. Although proteasomal degradation appears to be the major
apoB
MTP TG
MTP
TG
Rough ER TG membrane
Mature VLDL Pre-VLDL
Lipid droplet Smooth ER
membrane
TG
route for degradation of apoB, there are indications that lysosomal enzymes also may be involved [23]. In the second step in VLDL assembly, the mature VLDL particle is formed by fusion of the pre-VLDL particle with a preformed apoB-free lipid droplet.
The lipid droplets are produced in the smooth ER, while the fusion process is thought to occur in the junction between rough and smooth ER [16, 17]. Once the mature VLDL particle is formed, it is transported to the Golgi apparatus and secreted from the cell. The assembly of apoB-containing lipoproteins in the intestine, i.e. chylomicrons, is thought to occur by similar mechanisms but have not been studied in detail.
MICROSOMAL TRIGLYCERIDE TRANSFER PROTEIN
Microsomal triglyceride transfer protein (MTP) is essential for assembly of apoB- containing lipoproteins. This 97-kDa lipid transfer protein heterodimerises with a 58- kDa multifunctional chaperone called protein disulfide isomerase (PDI). MTP is primarily found in the lumen of microsomes in major apoB-secreting organs, i.e. liver and small intestine [24], while PDI is expressed ubiquitously [25, 26]. MTP is responsible for the transport of neutral lipids, preferentially triglycerides and cholesterol esters, to developing apoB molecules in the lumen of the ER [27]. This occurs via physical interaction between apoB and MTP [28, 29], which is likely to place the lipid-binding cavity of MTP close to the lipid-binding sites in apoB [30]. The MTP-mediated transfer of lipids has a stabilising effect on nascent apoB molecules [31-35]. The role for PDI in apoB secretion is not clear, but it has been suggested that it is necessary to maintain MTP within the ER [36, 37] and/or to maintain the stability of MTP [38]. The level of MTP expression has been shown to determine the secretion rate of apoB-containing lipoproteins [32, 39-41]. In humans, mutations in MTP cause abetalipoproteinemia, a rare disorder that results in an inability to secrete apoB- containing lipoproteins from the liver and the intestine [42]. Together these observations show that the secretion of both VLDL and chylomicrons is critically dependent on the presence of MTP. MTP also has a role in the second step in VLDL assembly, in which the major amount of lipids is added to the primordial VLDL by fusion with a preformed triglyceride-rich droplet. Electron microscopy studies have shown that VLDL-sized lipoproteins, i.e. the apoB-free lipid droplets, are absent in the Golgi stacks in mice lacking hepatic MTP expression [43]. Thus, this demonstrates that the formation of these lipid droplets requires the presence of MTP. However, the fusion process between the pre-VLDL particle and the lipid droplet appears to be MTP- independent [44].
Regulation of MTP
To date, only a few factors have been shown to regulate MTP expression. Insulin and high concentrations of glucose reduce the level of MTP mRNA in HepG2 cells [45], which is in line with the finding of a negative insulin response element in the human MTP promoter [46]. In insulin resistance models, such as fructose-fed Syrian hamsters [47, 48] and obese diabetic ob/ob mice [49], the hepatic expression of MTP is conversely increased together with increased secretion of triglyceride-rich apoB- containing lipoproteins. Diets enriched in triglycerides have also been found to upregulate MTP expression in hamster [50, 51] and rat [52]. Similarly, cholesterol increases hepatic concentrations of MTP mRNA in hamsters [53] and in HepG2 cells [46], which may contribute to a coordinated response to hepatic cholesterol accumulation leading to increased VLDL secretion. Conversely, cholesterol depletion of HepG2 cells lowers the level of MTP mRNA and protein [54], which is due to the upregulation of sterol regulatory element binding proteins (SREBPs) [54]. However, in some models the levels of both SREBP and MTP are increased [49, 55, 56], indicating that SREBP instead upregulates MTP.
STEROL REGULATORY ELEMENT BINDING PROTEINS
Sterol regulatory element binding proteins (SREBPs) belong to the basic helix-loop- helix leucine zipper (bHLH-Zip) family of transcription factors and regulate the biosynthesis of cholesterol, fatty acids and triglycerides. Three isoforms of SREBPs have been identified, designated SREBP-1a, SREBP-1c and SREBP-2 [57]. SREBP-1a and SREBP-1c are encoded from the same gene through the use of alternate promoters and differ only in their first exon [58]. This makes the length of the amino-terminal transactivation domain shorter in SREBP-1c, giving rise to a less potent transcriptional activator than SREBP-1a [59]. SREBP-2 is transcribed from a separate gene and has about 50% identity with the SREBP-1 amino acid sequence [58]. Most organs, including liver and adipose tissue, predominantly express the SREBP-1c and SREBP-2 isoforms. In contrast, most cell lines mainly express SREBP-1a in addition to SREBP- 2 [60]. Studies with transgenic mice that overexpress SREBP-1a, SREBP-1c or SREBP-2 in the liver have demonstrated distinct roles for the different SREBP isoforms [55, 61-64]. Each type of SREBP-overexpressing animal presented a different pattern of synthesis and accumulation of fatty acids and/or cholesterol in the liver. These data suggest that SREBP-1c is more selective in activating genes involved in fatty acid and triglyceride synthesis, while SREBP-2 is more specific for genes involved in cholesterol metabolism. SREBP-1a, on the other hand, is a regulator of genes involved both in fatty acid, triglyceride and cholesterol metabolism [61].
The two-step cleavage process of SREBPs
SREBPs are synthesised as inactive precursors bound to the ER membrane and nuclear envelope [57]. Each SREBP precursor is structurally composed of three domains: 1) an amino-terminal domain that contains the DNA-binding bHLH-Zip region as well as a transactivation region, 2) a central domain that contains two transmembrane segments linked by a short loop projecting into the ER lumen, and 3) a carboxyl-terminal regulatory domain. In order to activate target genes, the SREBP precursor must undergo a two-step proteolytic process to release the amino-terminal domain so it can function as a transcription factor (Figure 3) [65, 66]. This cleavage process requires the presence of SREBP cleavage-activating protein (SCAP) in addition to two proteases termed Site-1 protease (S1P) and Site-2 protease (S2P). SCAP is a sensor of cholesterol and upon cholesterol depletion it transports the SREBP precursor to the Golgi apparatus where the two proteases reside. In contrast, when the cholesterol content of cells increases, the SCAP-SREBP complex is retained in the ER. In the Golgi apparatus, SP1 cleaves SREBP in the luminal loop between its two membrane- spanning regions, generating two membrane-bound segments. The amino-terminal of SREBP is then liberated via a second cleavage mediated by S2P, producing the mature form of SREBP that enters the nucleus and activates transcription. Like other members of the bHLH-Zip family, SREBPs recognise palindromic sequences called E-boxes in the promoter region of SREBP target genes. However, due to the presence of a unique tyrosine residue in the DNA-binding domain, SREBPs can also bind to sterol response elements (SREs) or related sites [67]. This dual binding property makes it possible for SREBPs to bind a large variety of target genes in both lipogenic and cholesterol synthesis pathways.
Figure 3. Model for the sterol-mediated cleavage of membrane-bound SREBPs. Reg, regulatory domain; TA, transactivating domain; SCAP, SREBP cleavage-activating protein; S1P, Site-1 protease; S2P, Site-2 protease; SRE, sterol response element.
ER
Re
g TA
SCAP
- Sterols
Golgi SCAP
S1P S2P Re
g TA
Nucleus
Transcription
SRE
Re g Re
g TATA
SREBP-1c has been shown to activate transcription even in the presence of cholesterol, suggesting that the sterol-regulated proteolytic cleavage system is not as specific for SREBP-1c as for SREBP-2 [68]. The induction of SREBP-1c has instead been shown to be mainly at the mRNA level [68-71], which correlates with both the precursor and mature form of SREBP-1 [72]. Thus, while SREBP-2 controls cholesterol synthesis almost completely at the cleavage level, SREBP-1c largely regulates fatty acid synthesis by changing its own transcription level.
SREBP-1c and regulation of lipogenesis
The liver is the principal organ for lipogenesis, i.e. the production of fatty acids and triglycerides from excess dietary carbohydrate. Under lipogenic conditions, glucose in the cell is converted to pyruvate via the glycolytic pathway. Pyruvate is next converted into acetyl-CoA, which is used as a building block in the synthesis of long chain fatty acids. The produced fatty acids can in turn be used for esterification of glycerol-3- phosphate to generate triglycerides. Thus, in this pathway carbohydrates are converted to fat.
The enzymes catalysing the lipogenic reactions, thus designated lipogenic enzymes, are mostly regulated at the transcriptional level during different nutritional and hormonal states. One of the classic actions of insulin is to induce the entire lipogenic program and this is particularly observed during fasting-refeeding treatments to rodents. Moreover, several studies show that SREBP-1c mediates the stimulatory effect of insulin on lipogenic enzymes [55, 69, 70, 73, 74]. In contrast, polyunsaturated fatty acids (PUFAs) are negative regulators of hepatic lipogenesis through suppression of SREBP-1 [71].
LIPOGENIC ENZYMES
The major enzymes in the lipogenic pathway include not only genes for fatty acid synthesis, such as acetyl-CoA carboxylase (ACC), fatty acid synthase (FAS) and stearoyl-CoA desaturase-1 (SCD-1), but also a gene involved in triglyceride synthesis;
glycerol-3-phosphate acyltransferase (GPAT) (Figure 4). E-boxes/SRE-like sequences have been identified in the promoters of all these genes through which the mature SREBP-1c protein can exert it transcriptional activation [67, 75-77].
Acetyl-CoA carboxylase
Acetyl-CoA carboxylase (ACC) is responsible for the first committed step in fatty acid synthesis by catalysing the ATP-dependent formation of malonyl-CoA from acetyl- CoA and bicarbonate [78]. Malonyl-CoA is both a substrate for the next enzyme in the lipogenic pathway (fatty acid synthase) and regulates the oxidation of fatty acids by
inhibiting carnitine palmitoyl transferase-I (CPT- I), an enzyme responsible for the transport of fatty acyl-CoA into the mitochondria. Thus, ACC has important regulatory functions both in fatty acid synthesis and fatty acid oxidation. ACC exists in two isoforms, ACC-1 and ACC-2, and it has been speculated that ACC-1 is mostly involved in the regulation of fatty acid synthesis, while ACC-2 is suggested to control β-oxidation. Apart from being transcriptionally regulated, ACC is also regulated in the short-term at the posttranslational level by covalent modifications, such as phosphorylation/dephosphorylation [78]. Insulin has been shown to increase ACC activity by dephosphorylation, while glucagon and epinephrine decrease the activity by phosphorylation [78].
Fatty acid synthase
Fatty acid synthase (FAS) is a multifactorial enzyme responsible for the next step in the elongation of fatty acids by catalysing the formation of palmitate (16:0) from malonyl-CoA and acetyl-CoA. FAS is predominantly expressed in the liver and adipose tissue, i.e. tissues with a high degree of fatty acid synthesis. A high level of glucose has been shown to be required for insulin to induce FAS transcription [79], and glucose is also important for the stabilisation of FAS mRNA [80].
Stearoyl-CoA desaturase
Stearoyl-CoA desaturase (SCD) is a microsomal enzyme that catalyses the production of monounsaturated fatty acids by insertion of a double bond into saturated fatty acids [81]. The preferred substrates for SCD are palmitoyl (16:0) CoA and stearoyl (18:0) CoA, which are converted to palmitoleoyl (16:1) CoA and oleoyl (18:1) CoA, respectively. There are two rat SCD isoforms termed SCD-1 and SCD-2 [63]. Most tissues express both isoforms, but the liver is exceptional in that it only expresses SCD-1. Observations in mouse strains that have a natural mutation in the SCD-1 gene (asebia mice) [82] or a targeted disruption of the SCD-1 gene [83] demonstrate that the
Acetyl-CoA
Malonyl-CoA
Palmitic acid
Fatty acyl-CoA
Monoacylglycerol 3-phosphate
Triglycerides and phospholipids
Mono- unsaturated
fatty acids Acetyl-CoA
carboxylase (ACC)
Fatty acid synthase (FAS)
Stearoyl-CoA desaturase (SCD)
Glycerol-3-phospate acyl transferase (GPAT)
Figure 4. The lipogenic pathway
oleoyl-CoA and palmitoleoyl-CoA produced by SCD-1 are necessary to synthesise sufficient amounts of cholesterol esters and triglycerides in the liver. The regulation of SCD is also physiologically important because it changes the ratio between saturated and monounsaturated fatty acids in the cell. This in turn affects membrane fluidity that is known to be critical for many cellular processes [81].
Glycerol-3-phosphate acyltransferase
Glycerol-3-phosphate acyltransferase (GPAT) catalyses the first committed and probably rate-limiting step in triglyceride and phospholipid biosynthesis (reviewed in [84]). There are two isoforms of GPAT, one located in the mitochondrial outer membrane and the other on the ER. These two isoforms have different substrate preferences, where mitochondrial GPAT preferentially uses saturated fatty acyl-CoA, while microsomal GPAT uses saturated and unsaturated fatty acyl-CoAs equally well.
Moreover, only mitochondrial GPAT is cloned and also seems to be more extensively regulated by the nutritional and hormonal status. The mitochondrial GPAT is expressed mainly in lipogenic tissues, and mice deficient in this form of GPAT have a lower hepatic content and plasma concentration of triglycerides [85]. These findings indicate an important role of mitochondrial GPAT in the synthesis of triglycerides.
PEROXISOME PROLIFERATOR-ACTIVATED RECEPTORS
The peroxisome proliferator-activated receptors (PPARs) are nuclear receptors that exert a general regulatory effect on lipid homeostasis. There are three different types of PPARs, termed PPARα, γ and δ, which have distinct functions and tissue distribution. PPARα is predominantly expressed in liver, heart, and kidney, where it controls the transcription of genes that participate in fatty acid catabolism, most notably those involved in peroxisomal and mitochondrial β-oxidation [86]. PPARγ target genes are mostly implicated in lipogenic pathways in the adipose tissue, where PPARγ also is highly expressed. PPARδ is abundantly expressed in most tissues, with the exception of its very low expression in liver [87]. Several fatty acids and their derivatives are natural ligands for PPARs, although they bind with different affinity.
PPARα has a clear preference for binding long chain unsaturated fatty acids. To be active as transcription factors, PPARs must heterodimerise with retinoic X receptor (RXR) α, another nuclear receptor. This complex binds PPAR-response elements (PPREs) in the promoter region of target genes and activates transcription upon ligand binding. The PPRE consists of a direct repeat of the nuclear receptor hexameric DNA recognition motif (AGGTCA) that is separated by one nucleotide, thus called DR-1.
PPARα and lipid metabolism
PPARα is a central regulator of hepatic fatty acid metabolism by controlling genes involved in several fatty acid pathways, including uptake by the cells, intracellular binding as well as oxidation. The importance of PPARα is demonstrated during conditions when efficient fatty acid oxidation is required. For example, PPARα expression and activation are markedly induced during fasting in order to stimulate β- oxidation and production of ketone bodies. When PPARα-deficient mice are fasted, they accumulate massive amounts of lipids in their liver, reflecting the impaired expression of fasting-induced PPARα target genes in these mice [88, 89].
Fibrates are clinically used hypolipidemic drugs that lower triglycerides and increase HDL cholesterol concentrations through PPARα activation. PPARα mediates fibrate action on HDL levels in humans by inducing transcription of the major HDL apolipoproteins, apoA-I and apoA-II. However, in contrast to humans, fibrates do not increase HDL concentrations in rats due to sequence differences in the rat and human apoA-I gene promoters [90]. Fibrates increase the LPL expression and reduce the apoC-III expression, which leads to increased lipolysis of triglyceride-rich lipoproteins and thus accelerated clearance of these particles from the circulation [91]. In the liver, increased β-oxidation as well as reduced triglyceride synthesis [92] may further contribute to the lipid-lowering effect of fibrates by decreasing substrates for VLDL secretion. Thus, the triglyceride-lowering effect of fibrates occurs as a consequence of both enhanced catabolism of plasma triglyceride-rich lipoproteins and reduced secretion of VLDL triglycerides from the liver.
GROWTH HORMONE
Growth hormone (GH) is produced by the somatotropic cells in the anterior pituitary.
In humans, GH consists of a single peptide of 191 amino acids that is crosslinked by two disulfide bridges. Rat GH is one amino acid shorter and has 66% homology with its human counterpart [93]. The secretion of GH into the circulation is primarily regulated by two peptide hormones with opposite effects: GH-releasing hormone (GHRH), which stimulates the secretion of GH, and somatostatin, which inhibits the secretion of GH. These hormones are produced in the hypothalamus and reach the pituitary via the portal vascular system. The GH secretion is also regulated by a number of other factors, such as insulin-like growth factor-I (IGF-I), thyroid hormone, glucocorticoids, fasting and type 1 and type 2 diabetes mellitus (for review see [94]), indicating the complexity of the regulation of GH secretion.
The GH secretory pattern in females and males
There is a marked sex difference in the pattern of GH secretion in most adult species, including rodents [95] (Figure 5) and man [96]. Male rats secrete GH regularly in large pulses every 3-3.3 hours with low or undetectable levels between the peaks [95, 97]. Female rats, on the other hand, secrete GH in a near continuous fashion with lower pulse amplitudes and higher baseline levels than male rats [95]. The mean plasma level of GH, however, is similar in both sexes [98]. The sexually dimorphic secretory pattern of GH in humans is similar to that in rodents but is less pronounced [96]. The average daily GH concentration, however, is higher in women compared to men [99, 100].
Figure 5. The sex-differentiated GH secretory pattern in rats.
The difference in the secretory pattern of GH does not become apparent until after the onset of puberty. This suggests that the GH secretory pattern is highly controlled by gonadal steroids. Both neonatal and prepubertal gonadectomy of male rats result in elevated baseline GH levels during adult life, which can be completely reversed by sustained testosterone treatment [98]. However, GH pulse height is only decreased after neonatal gonadectomy, with unchanged pulse heights after prepubertal gonadectomy [98]. This suggests that a neonatal testosterone surge is needed to maintain normal GH pulse amplitudes in adult male rats, while a continuous presence of testosterone is necessary for preserving low GH basal levels in adult male rats.
Neonatal gonadectomy of female rats only mildly affects the female GH secretory pattern [98]. However, the importance of estradiol in maintaining the sexually dimorphic GH secretory is shown by the feminisation of the GH secretory pattern in intact males after estradiol treatment, i.e. higher baseline GH levels, lower pulse heights and more frequent GH bursts [101, 102].
The mechanism of sex steroid action is probably by modulation of the GHRH and somatostatin levels, as well as by direct effects on GH secretion in the pituitary. The
GH GH
Time Time
GH GH
Time Time
greater peaks that are observed in male rats compared to female rats may be due to the androgen-induced increase of both GHRH in the hypothalamus and GH secretion from the pituitary [103]. Some studies also show that androgens might enhance the somatostatin levels between peaks, which could account for the lower GH baseline level in male rats. Moreover, GHRH inhibits its own secretion and increases the secretion of somatostatin [104]. These properties in male rats are therefore likely to account for the cyclic GH secretory pattern through feedback mechanisms. In female rats, there is no such cyclic variation in somatostatin levels due to the inhibition by estradiol [105], which explains why female rats secrete GH more continuously than male rats.
GH receptor and GH signalling
The GH receptor is a member of the large cytokine receptor superfamily. This family includes receptors for more than 25 ligands, such as prolactin, multiple interleukins, leptin and erythropoietin [106]. Cytokine receptors are generally composed of an extracellular region, a single transmembrane domain and an intracellular region. A soluble form of the extracellular region of the GH receptor is found in plasma. This glycoprotein is called growth hormone-binding protein (GHBP) and it binds up to 60%
of circulating GH [107].
The initial step in GH signalling is the sequential binding of GH to two GH receptors that results in receptor dimerisation [108]. This moves the receptors in close proximity of each other, which increases the affinity of the GH receptor for the tyrosine kinase JAK2. As the GH receptor itself lacks tyrosine kinase activity, each JAK2 transphosphorylates the other JAK2 molecule and they are both thereby activated [109]. Activated JAK2 proteins subsequently phosphorylate the GH receptor at several tyrosine residues, converting them to docking sites for other signalling molecules [110]. One such molecule is the signal transducer and activator of transcription 5 (STAT5), which upon binding becomes phosphorylated by the action of JAK2 [111, 112]. This phosphorylation event activates STAT5 proteins, triggering their dimerisation, nuclear translocation and activation of gene transcription [112, 113].
There are two closely related isoforms of STAT5, termed STAT5a and STAT5b, of which STAT5b appears to be most important in GH signalling [114]. Other signalling pathways that are activated by GH include the mitogen-activated protein kinase (MAPK) pathway, the insulin receptor substrates (IRS)-1 and IRS-2, and protein kinase C.
The activation phase is usually transient, which means that effective shut off mechanisms by which the GH signalling pathway is inhibited must occur in the cell.
The most important negative regulators of GH signalling are members of the cytokine-
inducible gene family, termed suppressors of cytokine signalling (SOCS) [115]. These proteins are induced in response to GH or other cytokines by increased transcription via JAK-STAT activation. The SOCS proteins subsequently inhibit GH signalling by reducing the kinase activity of JAKs [115]. As the SOCS switch off the signalling pathway that initially led to its production, these proteins are involved in a classical negative feedback loop.
Influence of the GH secretory pattern on GH signalling
The second messenger system of intracellular signalling is activated differently by the intermittent and continuous GH stimulus. This is demonstrated by the activation of STAT5b by the intermittent GH secretory pattern in contrast to the continuous GH secretory pattern [116]. STAT5b has therefore been suggested to play a key role in the regulation of the sexual dimorphic gene expression in the liver that is induced by the male pulsatile GH secretory pattern. The hepatic expression of Cis (cytokine-inducible SH2-containing protein), a member of the SOCS family, is higher in female rats compared to males due to the continuous GH secretory pattern of females [117]. As Cis is responsible for the desensitisation of GH-induced STAT5b signalling, a more pronounced expression of Cis could result in less expression of male-characteristic genes in female rats. There is not much known about signalling that is specifically induced by the female secretory pattern of GH. However, one study shows that incubation of rat hepatocytes with GH stimulates female-specific CYP2C12 expression via upregulation of phospholipase A2 (PLA2). This effect is dependent on the subsequent P450-catalysed formation of an arachidonic acid metabolite [118] that may function as an intracellular second messenger.
Insulin-like and diabetogenic effects of GH
GH exerts both insulin-like and diabetogenic (antiinsulin-like) effects in adipose tissue and skeletal muscle (for review see [119]). The insulin-like effects occur soon after GH exposure and involve increased glucose utilisation and decreased lipolysis. When tissues are exposed to GH for a longer time, responsive cells are turned into unresponsive cells towards insulin-like actions, which is termed the refractory effect of GH. As GH is secreted endogenously throughout the day, causing a constant refractory state, the insulin-like GH effects probably have no physiological role. The late diabetogenic effects of GH that occur after prolonged GH exposure are therefore considered to better reflect the physiological situation. These effects include impaired glucose utilisation, hyperglycemia, stimulation of lipolysis, and induction and maintenance of the refractory state to insulin-like effects. The mechanism by which GH induces refractoriness to the insulin-like effects is thought to occur via upregulation of SOCS-3 and thereby blocking JAK2 activation [120].
GH and lipoprotein metabolism in humans
GH regulates a number of important functions in cholesterol and lipoprotein metabolism (Table 2). This is particularly emphasised during conditions of GH deficiency and GH excess (acromegaly), which are both associated with an abnormal lipid profile. GH-deficient subjects have elevated levels of total cholesterol, LDL cholesterol and triglycerides, and a reduced HDL cholesterol level [2]. This serum lipid profile is atherogenic and GH-deficient patients indeed have an increased risk for CVD [121]. GH replacement therapy in these patients has been shown to be favourable with respect to the changed plasma lipoprotein profile. Most notably, GH administration lowers the LDL cholesterol and increases HDL cholesterol, while plasma triglyceride levels are principally unchanged [122-124]. GH therapy stimulates the secretion of VLDL-apoB from the liver [125], but despite this the plasma concentration of apoB and LDL cholesterol decreases [122, 123]. This is probably due to the increased clearance of these particles after GH treatment [125], which might at least partly be explained by upregulation of LDL receptors [126, 127]. Moreover, GH treatment reduces the activity of CETP in GH-deficient patients, which increases the HDL cholesterol level and further contributes to the decreased LDL cholesterol level [128]. However, GH therapy has also been reported to increase the level of the atherogenic lipoprotein (a) [123, 124], which may contribute to the increased risk for CVD in acromegaly [1]. These patients also have increased triglyceride levels, decreased HDL levels and are often insulin resistant [129, 130], which might further contribute to the increased prevalence of CVD in acromegaly.
GH and lipoprotein metabolism in laboratory animals
GH plays an important role in cholesterol and lipoprotein metabolism also in rodents (Table 2). GH is known to stimulate lipolysis in adipose tissue, which increases the flux of free fatty acids to the liver and other tissues such as skeletal muscle. A continuous GH infusion to Hx rats has also a stimulatory effect on some hepatic lipogenic enzymes in vivo [131-134] and on triglyceride synthesis in hepatic cultures derived from these rats [135, 136]. Although this results in a stimulated secretion of VLDL [136, 137], no increase in serum levels of VLDL has been found in GH-treated Hx rats [138], which is in line with the findings in humans. This indicates that GH increases both production and clearance of VLDL. The increased turnover of VLDL in rats can partly be explained by the fact that GH enhances the editing of apoB mRNA.
This results in an increased proportion of secreted apoB-48-containing VLDL particles [139], known to have a considerably shorter half-life than particles containing apoB- 100 [140]. GH also has a stimulatory effect on LPL activity in skeletal muscle, which may further contribute to a more rapid catabolism of secreted VLDL particles [141, 142]. In addition, the secretion of apoE is increased in hepatocytes isolated from Hx rats treated with a continuous infusion of GH [143]. As apoB-48-containing
lipoproteins can be removed via interaction of apoE with LRP and LDL receptors, an increased content of apoE may add to the enhanced turnover of lipoproteins in response to GH. HDL cholesterol levels are increased by GH treatment in both rats [138] and mice [144], which in rats has been shown to be due to the female GH secretory pattern [138]. The LDL cholesterol level is conversely decreased by GH treatment in hypophysectomised rodents [138]. This is probably mainly due to the stimulatory effect that GH exerts on hepatic LDL receptor expression [126, 127], but increased editing of apoB mRNA [139] and increased HL activity [142] may also be involved. In conclusion, even though it is clear that GH increases VLDL secretion from the liver, the mechanisms behind the stimulatory effect of GH on VLDL assembly and secretion is not known.
Table 2. Effects of GH on lipoprotein metabolism in humans and laboratory animals. *Hx rats.
Humans Lab. animals*
HDL cholesterol
↑ ↑
LDL cholesterol
↓ ↓
TG
↔
or↑ ↔
or↑
ApoB
↓ ↓
ApoE
↑* ↑*
Lp(a)
↑*
Hepatic apoB mRNA editing
↑
Hepatic TG synthesis
↑*
VLDL-secretion
↑ ↑*
LDL receptor
↑ ↑
CETP
↓
LPL in adipose tissue
↓ ↔
LPL in skeletal muscle
↔ ↑
*Effect dependent on mode of GH administration
AIMS OF THE THESIS
The general aim of this thesis was to investigate the effects of GH and PPARα on key genes of importance for hepatic lipogenesis and VLDL assembly.
The specific aims of Paper I-IV were:
• To study the role of increased insulin levels for the effects of GH on lipoprotein metabolism in vivo
• To study the effects of gender and the sex-differentiated GH secretory pattern on the mRNA expression of SREBP-1c and lipogenic enzymes
• To study the effects of gender and the sex-differentiated GH secretory pattern on MTP expression
• To study whether PPARα activation increases MTP expression and activity
METHODOLOGICAL CONSIDERATIONS
Detailed descriptions of the assays that have been used are given in each paper and references therein. In this section, additional considerations concerning the in vivo and in vitro models are commented upon.
ANIMALS
Hypophysectomised rats
Sprague Dawley rats were hypophysectomised (Hx) by the temporal approach at 50 (Paper I and III) or 42-43 (Paper II and III) days of age (Møllegaard Breeding Center Ltd, Ejby, Denmark). The pituitary deficiency in hypophysectomised animals is total, meaning that the endogenous secretion of all hormones normally released from the pituitary is abolished. The hormonal treatment was initiated 10-14 days after hypophysectomy. The body weight gain was monitored for about one week during this period to determine the completeness of hypophysectomy. A body weight gain of more than 0.5 g/day was considered as incomplete hypophysectomy and used as an exclusion criterion. An alternative way of controlling the completeness of Hx is by examining the sella turcica for remaining GH activity using RIA. This has previously been done and the results of these measurements indicate that monitoring weight gain is a sensitive alternative of investigating completeness of Hx.
Gonadectomised (Gx) rats (Paper III) and PPARα deficient mice (Paper IV) were also used as described in the respective papers.
HORMONAL TREATMENT
Thyroxine (T4)
The plasma level of thyroxine is very low two weeks after hypophysectomy [145] due to the lost action of thyroid-stimulating hormone. This was substituted for by giving the hypophysectomised rats a daily subcutaneous injection of 10 µg L- thyroxine/kg/day (Nycomed, Oslo, Norway) diluted in saline. This dose results in somewhat higher thyroxine levels than in normal rats [139, 146], but has been shown to be within the physiological range as measured by the effect of different thyroxine doses on longitudinal bone growth [147]. However, the level of the active form of thyroxine, i.e. T3, has not been measured after this substitution.
Glucocorticoids
To substitute for the lack of adrenocorticotropic hormone, all Hx rats were given a daily subcutaneous injection of 400 µg cortisol phosphate/kg/day (Solo-Cortef,
Upjohn, Puurs, Belgium) diluted in saline. A replacement dose of 500 µg cortisone/kg/day has been shown to be within the physiological range with respect to body growth and longitudinal bone growth [148] as well as GH binding to adipocytes [149]. The dose of cortisol was adjusted to 400 µg/kg/day due to the higher potency of cortisol than cortisone.
In the in vitro studies, 1 nM dexamethasone was added to the medium since glucocorticoids exert a permissive action on some GH effects [150, 151]. The dose of 1 nM dexamethasone, or 0.39 ng/ml, corresponds to 32.5 ng/ml corticosterone after correction for the different potencies of these glucocorticoids. This dose of dexamethasone is somewhat below the physiological range of corticosterone concentration in serum (50-150 ng/ml) [152].
Growth hormone
Recombinant bovine GH was given in a dose of 0.5-1.5 mg/kg/day diluted in 0.05 M phosphate buffer (pH 8.6) containing 1.6% glycerol and 0.02% sodium azide. Bovine GH was chosen rather than human GH, since bovine GH only binds to the GH receptor in contrast to human GH that also binds the prolactin receptor. Compared to rat GH, bovine GH is more stable and also much more available on the market.
GH were administered to the hypophysectomised rats either as a continuous infusion via an osmotic mini-pump (Alza Corp., Palo Alto, CA, USA) implanted subcutaneously between the scapulae, or as two daily subcutaneous injections at 12-h intervals (0800 and 2000 h). These modes of GH administration have been shown to experimentally imitate the female and the male GH secretory pattern, respectively, with respect to feminisation and masculinisation of P450 enzyme levels in rat liver [101, 153]. GH injections, however, will result in fluctuations of GH plasma levels during the day with the possible outcome that GH effects on mRNA species with high turnover are not detected. The hepatic expression was therefore analysed both at 2 and 6 hours after the last GH injection.
The normal secretory rate of GH in 50-60 days old female rats is 1.3 mg/kg/day as calculated from the GH clearance rate (1.19 ml/min) [154] and the normal mean plasma level of GH (135 ng/ml) [155]. Thus, both the in vitro dose of GH (100 ng/ml, Paper II) and the continuous infusion of 1.5 mg GH/kg/day to Hx rats (Paper I) are within the physiological range. When the hypophysectomy model was used to investigate the regulatory role of the feminine and masculine GH secretory pattern, a lower dose of GH (0.7 mg/kg/day) was administered (Paper II and III). This dose was chosen to assure very low or undetectable GH levels between the GH injections analogous to the GH secretory pattern of the male. Although lower than the normal
GH secretion rate, this GH restitution is sufficient as indicated by the increased final body weight and body weight gain in hypophysectomised rats ([156] and Paper II).
To assess the role of the different GH secretory patterns without giving GH injections that result in slow diurnal variations in GH concentrations, a low dose of GH (0.5 mg/kg/day) was administered as a continuous infusion to intact rats. This mode of GH administration will feminise male rats with respect to the GH secretory pattern without any major changes in the total mean plasma level of GH [157, 158]. The final body weight, body weight gain and IGF-I mRNA expression did not change by this GH treatment in Paper II and III, indicating the total GH exposure was not particularly affected in our studies.
Insulin
A slow-release form of insulin (Insulatard, 100 IU/ml, Novo Nordisk A/S, Denmark) diluted in saline was given as a daily subcutaneous injection at 1600 h to the Hx animals in Paper I. To avoid insulin-induced fatal hypoglycemia, the dose of insulin was gradually increased from 1 IU to 2 IU/day. This insulin treatment has been shown to result in serum insulin levels similar to those in sham-operated animals [159].
In the in vitro studies, the cells were plated in a medium containing 16 nM insulin and then cultured in the presence of 3 nM insulin. The insulin concentration in the portal blood of fasting rats is 0.34 nM [160] and is several-fold higher in fed rats. The in vitro doses of insulin are therefore near or above the physiological range.
Sex steroids
Testosterone was diluted in propylene glycol and given at a dose of 0.5 mg/kg/day as a daily subcutaneous injection to Gx male rats (Paper III). A similar dose of testosterone has been shown to increase the level of plasma testosterone in Gx rats to that of intact male rats [161], and also to masculinise the secretory pattern of GH [155]. 17β- estradiol was diluted in propylene glycol and administered at a dose of 0.1 mg/kg/day as a daily subcutaneous injection to female Gx rats (Paper III). The serum level of estradiol in intact female rats varies between 2 and 50 pg/ml throughout the cycle [162]. As treatment of intact rats with 0.01 mg estradiol/kg/day increases the plasma estradiol level to approximately 51 pg/ml [163], our higher dose of estradiol (0.1 mg/kg/day) would therefore probably result in supraphysiological serum levels of estradiol in Gx rats.
HEPATOCYTE CULTURES
Hepatocytes were isolated by nonrecirculating collagenase perfusion through the vena porta of anaesthetised female Sprague Dawley rats or PPARα null and wild type mice [151, 164, 165]. During the first 5-6 minutes, the liver was perfused with a medium containing EGTA at a flow rate of 40-50 ml/min for rats and 20-30 ml/min for mice.
EGTA has a Ca2+-chelating function and therefore facilitates the disruption of Ca2+- dependent cell-to-cell interactions. For the next 7-9 minutes, the liver was perfused with a second medium supplemented with collagenase IV. Collagenase is a metalloprotease that degrades collagen and this further helps separating the cells inside the liver capsule. After the perfusion, the liver capsule containing the liver cells was excised. The cells were filtered trough a 250 µm pore size mesh nylon filter followed by a 100 µm pore size mesh nylon filter to remove undigested part of the liver.
Thereafter, the cell suspension was repeatedly washed in plating medium in order to remove collagenase and cells with lower density, such as Kuppfer cells and fat-storing Ito cells. The cells were seeded at ~120 000 cells/cm2 in petri dishes coated with a layer of Matrigel (Figure 6). The Matrigel contains large amounts of laminin, proteoglycan and collagenase type IV, which support the hepatocytes in the subendothelial space of normal liver. In comparison with primary hepatocytes that are cultured on plastic or collagen, hepatocytes cultured on Matrigel have been shown to better maintain their characteristic properties [166, 167], such as expression of functional GH receptors [168]. In experiments where protein concentrations were measured, the matrigel was first removed by adding PBS supplemented with 5 mM EDTA in order to eliminate contaminating proteins from the Matrigel. During the first night of culturing, the medium contained a high dose of insulin (16 nM) as this has been shown to enhance plating efficiency and formation of dispersed monolayers [169]. This medium also contains
glucose (28 nM), which has been shown to maintain the expression of GH receptors [170]. After plating, the cells were cultured in a medium without insulin or with a lower insulin concentration (3 nM), and with the hormone/substance of interest supplemented. Dexamethasone (1 nM) was also added to the culture medium due to its permissive actions on GH effects [151]. The hepatocytes were incubated during the next 3 days
unless otherwise stated. The constant Figure 6. Primary rat hepatocytes in culture
presence of GH in the culture medium is similar to the continuous GH infusion in vivo that mimics the feminine GH secretory pattern. The in vitro results of GH incubation can therefore be used as an indication of the direct or indirect nature of the continuous GH secretory pattern on hepatic functions in vivo. The effects in the hepatocytes cultures were related to the DNA content in each culture dish to correct for differences in cell number (Paper III and IV).
