The response of lipoprotein lipase to feeding and fasting. Evidence for posttranslational regulation.

The regulation of adipose tissue lipoprotein lipase (LPL) was examined in rats fed or fasted overnight, and was found to be controlled posttranslationally. LPL catalytic activity decreased by 50% after fasting while LPL mRNA levels and rates of synthesis increased nearly 2-fold; enzyme mass remained unchanged. The distribution of LPL within the endoplasmic reticulum (ER) and Golgi/post-Golgi secretory pathway was assessed by differentiating between LPL high mannose and complex forms. After fasting, the majority of LPL is in the high mannose ER form (65%, 0.97 micrograms/g wet weight tissue), whereas the LPL complex form comprises only 35% (or 0.52 micrograms/g). After refeeding, however, the Golgi-derived LPL complex form predominates (65%, 1.03 micrograms/g) over the high mannose ER form (35%, 0.55 micrograms/g). Kinetic analysis suggests that high mannose LPL disappears with a half-life of t0.5 = 40 min in both fed and fasted rats, indicating that the redistribution of LPL mass during feeding/fasting does not arise by differential retention within ER. Instead, the fractional catabolic rate of complex LPL within the Golgi/post-Golgi secretory compartment can be calculated to be 3.5-fold greater in fasting. In heart, changes in LPL activity in response to feeding/fasting are also not due to differences in mRNA levels or rates of synthesis. Based on these findings, a model of LPL posttranslational regulation is proposed and discussed.

The regulation of adipose tissue lipoprotein lipase (LPL) was examined in rats fed or fasted overnight, and was found to be controlled posttranslationally. LPL catalytic activity decreased by 50% after fasting while LPL mRNA levels and rates of synthesis increased nearly %-fold; enzyme mass remained unchanged.
The distribution of LPL within the endoplasmic reticulum (ER) and Golgi/post-Golgi secretory pathway was assessed by differentiating between LPL high mannose and complex forms. After fasting, the majority of LPL is in the high mannose ER form (65%, 0.97 pg/g wet weight tissue), whereas the LPL complex form comprises only 35% (or 0.52 pg/g). After refeeding, however, the Golgi-derived LPL complex form predominates (65%, 1.03 pg/g) over the high mannose ER form (35%, 0.55 rg/g).
Kinetic analysis suggests that high mannose LPL disappears with a half-life of t0.s = 40 min in both fed and fasted rats, indicating that the redistribution of LPL mass during feeding/ fasting does not arise by differential retention within ER. Instead, the fractional catabolic rate of complex LPL within the Golgi/post-Golgi secretory compartment can be calculated to be 3.5-fold greater in fasting. In heart, changes in LPL activity in response to feeding/fasting are also not due to differences in mRNA levels or rates of synthesis.
Based on these findings, a model of LPL posttranslational regulation is proposed and discussed.
Lipoprotein lipase (LPL'; EC 3.1.1.34) is a N-linked glycoprotein secreted by parenchymal cells from a variety of extrahepatic tissues, principally muscle and adipose tissue (1). The secreted enzyme, bound at the surface of capillary endothelium (2) and called "functional LPL," is involved in the hydrolysis of triglycerides from circulating chylomicrons and very low density lipoproteins. Postprandially, LPL activity is elevated in adipose tissue compared with heart and muscle, resulting in the channeling of circulating triglyceride fatty acids into lipid depots. During fasting, the inverse is true; relatively high heart and muscle LPL activities redirect triglyceride fatty acids appropriately into these tissues and away from adipose stores (1,3). Obviously, the coordinated regulation of LPL in adipose tissue and muscle during feeding/ fasting is critical for maintaining triglyceride homeostasis. For example, in obesity, overexpression of adipose tissue LPL may contribute to increased triglyceride deposition (4). In chicken (5) and guinea pig (6), a prolonged 48-h fasting period was observed to reduce, in parallel, adipose tissue LPL activity and mRNA levels. However, insulin, which is believed to be the major effector of LPL postprandially (l), increases LPL activity in 3T3-Ll adipocytes, whereas LPL synthetic rate decreases (7). Indeed, many possible mechanisms of regulating LPL at posttranslational levels have been reported: e.g. catalytic activity may depend on the acquisition and complete processing of LPL's N-linked oligosaccharide chains (8-10); newly synthesized LPL is degraded rather than secreted in the absence of heparin (11-15); and, insulin-enhanced LPL secretion may depend on a phospholipase that releases LPL from a membrane bound glycosyl phosphatidylinositol anchor (16).
Thus, the role of LPL transcription and translation in regulating adipose tissue and heart LPL activity and mass levels was investigated in rats fasted or refed over a period of 12 h. This period of time was chosen since most animals undergo a natural feeding/fasting cycle of approximately l/z day. In addition, to verify posttranslational changes in the cellular distribution and kinetics of adipose tissue LPL in response to feeding and fasting, glycosidase sensitivity was used to distinguish between LPL high mannose and complex forms. These two forms represent the products of a series of N-linked oligosaccharide processing events that are known to occur within the ER and Golgi apparatus, respectively (17); thus, glycosidase sensitivity was used as a means of locating LPL to these organelles (18  densitometry of Northern blots revealed that changes in mRNA levels following feeding or fasting were parallel to changes in LPL synthetic rates, measured by incorporation of ['"Slmethionine (Table I). However, although LPL mRNA levels and synthetic rates increased nearly P-fold after fasting, activity decreased by over 50%. Similarly, enzyme mass did not accompany activity loss after fasting, resulting in a 50% decrease in specific activity (units/mg LPL protein). The nature of the changes mentioned above was very reproducible and a typical response of adipose tissue LPL to 12-h feeding and fasting is shown in Table I ation sites with the consensus sequence Asn-X-Ser(Thr), where X is any amino acid but proline (21,22), and both of these sites are utilized for glycosylation." Likewise, rat LPL contains two glycan groups/protein molecule, as ascertained by partial glycosidase digestion (18,23)  within half that time (Fig. 6A). After 80 min, Endo H cleavage of LPL immunoprecipitates (Fig. 6B) clearly shows that changes observed in LPL activity after feeding or fasting (Table I) Fig. 8).
Turnover Rates of LPL High Mannose and Complex Forms after Feeding and Fasting-During pulse/chase experiments, total adipose tissue LPL mass remained constant. Turnover rates of the LPL high mannose form measured in adipose tissue from fed or fasted rats were identical, with a &, = 40 min or fractional turnover rate (FTR) of 0.75 (h-') (Fig. 7A). Thus, the larger mass of the high mannose form following fasting (0.97 versus 0.55 pg) can be attributed to the proportional increase in the LPL synthetic rate and is not due to selective retention within the ER (see also Fig. 8). In contrast, the kinetics of the complex form are quite different. As shown in Fig. 7B, in fed animals, LPL radioactivity (represented as a fraction of the initial IO-min chase point) peaks approximately 20-30 min after initiation of the chase, after which it disappears steadily. A different kinetic pattern, however, is observed in fasted animals: maximal LPL radioactivity appears to have been reached very rapidly, as evidenced by the lack of a peak prior to the disappearing phase of the curve. Attainment of peak radiospecific activity so early on in these animals suggests that the pool size of the complex form of LPL is reduced by fasting, identical to the conclusion reached above (see Fig. 6B).
The reduced pool size of the intermediate/complex form after fasting appears to arise as a result of an increased fractional catabolic rate (FCR). The FCR of the LPL intermediate/complex form was calculated assuming that steadystate conditions exist; i.e. the influx of LPL from the high mannose form (high mannose pool size x FTR) equals the loss of the LPL intermediate/complex form due to degradation (intermediate/complex pool size X FCR). Thus, the calculated FCR following fasting was 1.40 (h-l) compared with 0.40 (h-l) after refeeding, a difference of 3.5-fold (see Fig. 8).
Evidence for Posttranslational Regulation of Heart LPL-Adipose tissue and heart LPL activities are reciprocally regulated after feeding or fasting (1, 3). Heart LPL activity increased 2-fold after fasting while activity decreased in adipose tissue (Table II). As in adipose tissue, changes in heart LPL activity following feeding/fasting were not accompanied by parallel changes in LPL mRNA levels (Fig. 3, Table II) and rates of lipase synthesis (Table II). However, in contrast to adipose tissue LPL, heart LPL mass changed proportionally with activity, maintaining a constant specific activity.

DISCUSSION
The decrease in rat adipose tissue LPL activity following a 12-h fast and its subsequent increase after refeeding are controlled posttranslationally. Compared with the fasted state, adipose tissue LPL activity increases over 2-fold after refeeding, whereas mRNA levels and rates of synthesis decrease almost by half; enzyme mass remains unchanged. A  Fig. 8 are representations of the LPL mass distribution within the adipose tissue, including the fraction of LPL bound at the endothelial surface under these experimental conditions. It should be noted that the calculated intermediate/complex pool sizes do not take into account endothelial-bound LPL that might have been lost to the medium, although in the absence of heparin, this would constitute a minor fraction (14). A significant finding in our studies is that a major cellular redistribution of adipose tissue LPL occurs in response to feeding/fasting.
Although the high mannose form comprises the majority of LPL following fasting, the intermediate/ complex form predominates after refeeding (Fig. 6); this strongly suggests a shift of LPL out of ER and into the post- The LPL mass of each compartment, located within the stippled boxes, is microgram LPL/g wet weight adipose tissue. The kinetic parameters are: relative rate of synthesis, defined as the percentage of [35S]methionine incorporated into LPL relative to total protein synthesis; FTR, defined as the fraction of the pool of the high mannose form of LPL turned over each hour with the units (hours-'); and FCR, defined as the fraction of the pool of the intermediate/complex form of LPL leaving the system by degradation each hour with the units (hours-'). The relative rate of synthesis and FTR were determined experimentally; the FCR was calculated from the pool sizes of the high mannose and intermediate/ complex forms of LPL (micrograms of LPL/g of tissue), and the FTR assuming the system was at steady state. For details see "Experimental Procedures" and "Results."  Figs. 9 and 10, proposing a possible mechanism of regulating the amount of endothelialbound functional LPL in adipose tissue during feeding and fasting, taken into account the findings of this study; i.e. a change in the activity and cellular distribution of LPL without a change in total LPL mass (Fig. 9). It is the endothelialbound fraction of LPL that is critical in determining the rate of triglyceride fatty acid influx into adipose stores during feeding and fasting. Fig. 10 proposes that the redistribution of LPL mass could arise by diverting newly synthesized LPL into pathways leading either to secretion (during feeding) or The cross-hatched and stippled rectangular areas represent the distribution of mass within the high mannose and intermediate/complex forms of LPL, as determined in Fig.  6 and summarized in Fig. 8. Total LPL mass remains constant after feeding or fasting, as indicated in Table I Details of the models are in the text. The relative areas of the circles are meant to approximate LPL mass within each compartment. As in the legend to Fig. 9, the distribution of LPL mass residing in the functional pool uersus the Golgi/post-Golgi compartments is at present unknown. Heavy arrows represent accelerated rates. The cross-hatched and open boxes refer to possible blocks within the secretory and degradation pathways of LPL in adipose tissue.
Other investigators (11-X), using isolated adipocytes, have also shown that newly synthesized LPL is diverted into a lysosomal degradative pathway in the absence of a secretagogue (heparin). In an alternative model, shown by the dashed lines and question marks in Fig.  10, secretion is constitutive, and the amount of functional LPL is regulated by turnover at the endothelial surface. However, endothelium does not appear to be the major site of LPL degradation, since exogenously added LPL is degraded much slower by endothelial cells than adipocytes (25). Nevertheless, it is still possible that LPL is removed from endothelium by triglyceride-rich lipoproteins and degraded in liver (26,27).
The decrease in adipose tissue LPL specific activity noted after fasting (Table I) most likely arises from the presence of inactive mass. Coincidently, in fasting, the majority of LPL is within ER, which may be inactive. For example, studies utilizing tunicamycin (9-11) and glucose deprivation (28) in adipocytes indicate that LPL glycosylation is required for secretion and that lipase retained intracellularly is inactive. Carbonyl cyanide m-chlorophenylhydrazone, which blocks the energy dependent budding involved in the formation of transport vesicles from ER to Golgi, causes retention of LPL within ER and loss of activity (8, 12). Combined lipase deficiency in mice (29,30), a genetic defect located on a chromosome separate from the LPL gene (31) and apparently affecting LPL translocation from ER to Golgi,4 again is characterized by inactive mass. However, our laboratory has recently shown by site-directed mutagenesis5 that unglycosylated hepatic lipase, which is structurally and functionally very similar to LPL (20, 32), is catalytically active. Thus, loss of catalytic activity within ER is probably not a consequence of "immature" glycoslyation. Indeed, the high mannose form of LPL in guinea pig adipocytes is active against a triolein substrate (18), although specific activity was not reported. Therefore, the precise intracellular location and nature of the inactive LPL mass arising in rat adipose tissue following fasting remains to be determined.
LPL activity in heart is also regulated posttranslationally. Although LPL activity is modulated several fold after feeding/ fasting, LPL mRNA levels and rates of synthesis are not affected proportionally. Since pulse/chase studies in perfused hearts are technically difficult, the mechanism of LPL regulation in this tissue has not been determined. It would seem a priori that the regulation in heart would be similar to that proposed in adipose tissue; i.e. a redistribution of LPL in response to feeding/fasting. In cells isolated from newborn rat hearts (33), an increase in heparin-releasable LPL activity in response to the P-adrenergic agent isoproterenol or dibutyryl-CAMP did, in fact, coincide with an apparent redistribution of LPL from the high mannose to complex forms. However, unlike adipose tissue, LPL mass did change in parallel with activity following feeding or fasting (Table II). Whether this indicates that LPL is regulated by a posttranslational mechanism different in heart than in adipose tissue is unknown.
LPL activity levels are most likely regulated during feeding and fasting by a complex array of factors, such as insulin, glucagon, and glucocorticoid levels as well as sympathetic innervation. This study examined adipose and heart LPL regulation at only one time point (12 h) after fasting and refeeding. It is certaintly possible that LPL activity levels could be regulated by different mechanisms and to different degrees at various time points after refeeding, e.g. when circulating insulin levels are maximal. Unquestionably under different conditions and possible in different species, LPL activity levels are also regulated at the level of transcription. For example, guinea pig (6) and chicken (5) adipose tissue LPL mRNA levels and rates of synthesis do respond in parallel to activity changes in a prolonged (48 h) fast, which is four times as long as the fasting period used in our studies. In addition, tumor necrosis factor (6) and insulin (34) also regulate LPL activity in adipocytes at the level of transcription and translation. Thus, LPL can be regulated by several mechanisms, most likely dependent on the time scale of the required response and the effector involved. Indeed, as evidence mounts, the adipocyte, myocyte, and most likely other LPL producing cell types appear to have the capacity to regulate LPL activity at a variety of levels.