Biochimica et Biophysica Acta, 1086( 1991) 354-358 © 1991ElsevierScience PublishersB.V. All I'ightsreserved0005-2760/91/$(13.50
354
The regulation of hepatic lipase and cholesteryl ester transfer protein activity in the cholesterol fed rabbit R o d e r i c J. W a r r e n l, D a v i d L E b e r t t, P h i l i p J. B a r t e r 2 a n d A l a n a M i t c h e l l 1 I Baker Medical Research Institute, Prahran, Victoria (Australia) and 2 Wollongong Unh'ersity. Wollongong (Australia)
(Received 12 April 1991)
Key words: Hepatic lipase;Lipase;Cholesteroldiet; Cholesterylester transfer protein; (Rabbit) Hepatic lipase (HL) and cholesteryl ester transfer protein (CETP) activities are both increased in the rabbit by cholesterol feeding. The in vivo regulation of HL and CETP were explored by examining changes in specific steady-state mRNA levels upon cholesterol feeding. On feeding rabbits cholesterol, HL activity increased 3-fold after 2 days and remained at 2.6-times the control value at 28 days. Specific rabbit HL mRNA levels were assessed by dot blot analysis of liver poly (At + RNA hybridized with the h u m a n HL eDNA. No siRn|flcant changes in liver HL mRNA accompanied the increase in activity seen at days 2 and 7. At day 28 a modest rise of 46% was observed. A significant rise in CETP activity, evident 7 days after the commencement of cholesterol feeding, was maintained until day 28 when it was 2A-times the control value. Using the h u m a n CETP eDNA as probe, rabbit liver CETP mRNA was also found to increase by day 7, rising to 3.7-times control by day 28. The strong temporal relationship between the rise in CETP activity and mRNA (r = 0.55, P = 0.02) s u u e s t s that the regulation of CETP may be primarily effected by the levels of specific mRNA. in contrast, the discordance between levels of lipase activity and mRNA suggests that post-transcriptional events may be more important in the regulation of HL in the cholesterol fed rabbit.
Introduction Hepatic lipase (HL) and cholesteryl ester transfer protein (CETP) are two factors that play a fundamental role in determining the composition of high-density lipoproteins (HDL) [1-4]. Their activities can be influenced by a variety of interventions, including diet, steroid administration and exercise [5-12]. Cholesterol feeding is known to increase the activity of both HL and CETP in the rabbit [5,9,10]. Such an increase in activity may occur through an increase in m R N A levels, leading to enhanced synthesis of new protein; by alterations in post-transcrlptional events; or by a combination of mechanisms. Although it has been suggested that an increase in liver CETP m R N A accompa-
nies a rise in CETP protein mass and activity [13], little is known about the regulation of HL activity. To explore the regulation of HL and CETP we examined the changes that occurred in specific liver m R N A levels when the activities of HL and CETP were increased by cholesterol feeding in the rabbit. We report that the low basal HL activity previously reported in the rabbit [14] was increased with cholesterol feeding. There was, however, no accompanying rise in specific liver HL m R N A levels. In contrast, CETP activity and specific liver m R N A levels were both increased by cholesterol feeding and displayed a strong temporal relationship. These findings support the importance of m R N A levels in the regulation of CETP and suggest that the regulation of HL is less dependent on steady-state m R N A levels and is likely to be predominantly post-transcriptlonal.
Abbreviations: HI_.,hepatic lipase; CETP, choleslerylester transfer protein; SDS, sodium dodeeyl sulfate; LPL, lipoprotein lipase; kb, kilobases.
Materials and Methods
Correspondence: R.J. Warren, Baker Medical Research Institute, CommercialRd.. Prahran, Victoria. 3181. Australia.
Unless specified, all reagents were obtained from the Sigma Chemical (St. Louis, MO, U.S.A.).
Chemicals
355
Animals and diet A multi-coloured strain of rabbits maintained at the Baker Medical Research Institute and male SpragueDawley rats (200 g) were used in these studies. For the cholesterol feeding study, female rabbits weighing 2.02.5 kg, previously maintained on commercial rabbit chow (4% fat w/w, no cholesterol, Clark King, Australia), were divided into two groups receiving either commercial chow or the same chow supplemented with 0.5% (w/w) cholesterol Five cholesterol fed animals were killed at each time point (days 2, 7 and 28) while separate control groups were killed at days 0 (n = 4) and 28 (n = 5). The results of all assays did not differ significantly between the two control groups and the averages of the results in these two groups were used for comparisons. The animals were housed individually, received water ad libitt,m and were weighed weekly.
Hepatic lipase activity Hepatic lipase is usually assayed in vivo in post heparin plasma [2]. In recent in vitro studies, however, heparin has been reported to produce marked increases in the content of HL m R N A in Hep G2 cells, albeit after prolonged treatment periods of 22 h or greater [15]. Since beparin treatment may also influence HL m R N A levels in vivo, an alternative method of obtaining HL was used to avoid the in vivo administration of heparin. The method was adapted from that used by Taskinen et al. [16] to elute lipoprotein lipase (LPL) from small samples of adipose tissue. Briefly, 50 mg pieces of fresh liver tissue were cut finely with a razor blade and incubated at 28°C for 40 rain in 600/zl of Kxebs-Ringer buffered with 0.1 M Tris-HC! buffer (pH 8.4), containing I g/100 ml of bovine serum albumin and 2.5 I.U. of heparin. An aliquot of the heparin eluate (150 /zl) was added to 500 /~l of an artificial triolein emulsion, and incubated at 28°C for 60 min in the presence of 1 M NaCI to inhibit LPL, as described by Huttenen et al. [17]. From similarly prepared eluates of rat liver, where the HL activity was much greater, we were able to demonstrate that the assay was linear with respect to both time and enzyme concentration in the range of values obtained from the rabbit liver eluates. For comparison of HL activity between rabbit and rat, activity is expressed as units per gram of liver tissue, with each unit representing i /zmol of labelled free fatty acid released per hour. These results are the mean of triplicate samples for each animal. In the cholesterol feeding study, where only rabbits were assessed, activity is expressed as units per liver since liver weight increased with cholesterol feeding due to lipid accumulation (Day 0 control, 83.3 ± 2.9 g; Day 2, 74.6 + 6.3 g; Day 7, 84.2 + 6.8 g; Day 28, 99.6 + 4.8 g; Day 28 control 77.8 ± 2.7 g). There was no significant change in rabbit weights. As the rat
is known to have high HL activity, a sample of fresh rat liver was included for assay at each time p'~int as a positive control. The results for these samples varied by less than 10%. To confirm the validity of the results of the HL activity from liver eluates, HL activity was also measured in post heparin plasma in a separate group of rabbits after 7 days of cholesterol feeding, and coinpared with the in vitro method. Plasma was obtained 15 rain after the injection of heparin (100 U / k g ) and samples (10 p.I) assayed for HL activity by the method of Huttenen et al. [17]. Activities are expressed as units per ml of post heparin plasma (1 unit = i p.mol free fatty acid released per h).
CETP actit'ity CETP activity was measured as the capacity of lipoprotein-deficient rabbit plasma to facilitate the transfer of radioactively labellod cholesteryl ester from human high-density lipoprotein (HDL) 3 (d = 1.13 g / m l to i.21 g / m l ) to human low-density lipoprotein (LDL) (d = 1.019 g / m l to 1.055 g / m l ) [18]. The source of CETP was the rabbit plasma fraction of d > 1.25 g/ml. Transfer activity is expressed as units per ml plasma, where units are the rate constant k [19] for the transfer of radiolabelled tracer from HDL to LDL multiplied by the time of incubation.
Messenger RNA ~.~'.antification Total RNA from rabbit and rat was isolated from liver tissues by the method of Chomczynski and Sacchi [20] and enriched for poly (A) + RNA by oligo (dT) selection [21]. To measure changes in specific m R N A resulting from cholesterol feeding, 0.5, 1.5 and 3.0/zg aliquots of rabbit liver poly (A) + RIqA were applied to nitrocellulose (Bio-Rad miniblot system) and probed with human HL and human CETP eDNA's. The human HL eDNA was obtained by screening a human liver lambda gt 10 library with rat HL eDNA (provided by Dr. M. Schotz, UCLA). The clone was confirmed as full-length by dideoxy chain termination sequencing [22]. The human CETP cDNA was a gift from Dis. R. Lawn and D. Drayna, Genentech [23]. Both probes were labelled by random priming to a specific activity of 10¢ cpm//zg DNA [24]. On Northern blot analysis, the human HL eDNA hybridized to a discrete rabbit liver m R N A species of 1.6 kb using a final wash of 0.2 x SSC (! × SSC: 15 mM sodium citrate, 150 mM sodium chloride, pH 7.0.), 0.1% sodium dodecyl sulphate (SDS) at 55°C (Fig. I). The human CETP produced a discrete band on Northern blot analysis of rabbit liver poly (A) + RNA using a final wash of 0.5 × SSC, 0.1% SDS at 55°C (Fig. 2). Dot blot autoradiographs were scanned by laser densitometry (UItrascan XL, LKB Bromma, Sweden) to quantify specific m R N A levels. The varying concentrations of
356
Results
Hepatic lipase
1.6 kb -- O
Fig. 1. Northern blot of rabbit poly (A) ~ RNA with the human HL eDNA. Rabbit liver poly (A) ÷ RNA (8 ttg) was size fractionated by gel electrophoresis, transferred to nitrocellulose membrane and hybridized with human HL eDNA to identify a single discrete species of rabbit liver mRNA of 1.6 kb.
m R N A used produced a linear range of hybridization signals and only data in the linear range of the signal response were considered. To correct the data for differences in the m R N A content between dots, the total m R N A content of each sample was normalized by probing the stripped filters with oligo(dT)30, endlabelled with [32p]ATP [24]. The same filters were again stripped and probed with a rat albumin e D N A (provided by Dr. G. Howlett, Melbourne University), labelled to a specific activity of 10 ° c p m / / z g , to ensure that any observed changes in H L m R N A were specific. Results are expressed as changes relative to the values obtained for control animals which were assigned an arbitrary value of 1.0.
Hepatic lipase activity in the heparin eluate of fresh rabbit liver was less than 10% of the corresponding activity from the rat (0.45 + 0.06 u n i t s / g liver, n = 9 vs 5.17 + 0.04, n = 7, P = 0.001). These findings are consistent with previous reports that have shown the rabbit to have low H L activity in post-heparin plasma when compared to other species such as rat and man [14,25]. The high salt lipolytic activity detected in the heparin eluate of fresh rabbit liver was reduced to zero if the animals received a prior injection of heparin (results not shown). This finding suggests that intracellular liver lipases do not contribute significantly to the lipolytic activity measured in the liver eluate assay. Cholesterol feeding produced a significant early and sustained rise in H L activity (Fig. 3), relative to control values which did not alter significantly over the study period (0.36 + 0.10 u n i t s / l i v e r , n = 4 at day 0, and 0.37 + 0.04, n = 5 at day 28). A t day 2 of cholesterol feeding, H L activity was 3.0-times control (1.10 + 0.27 units/liver, P = 0.03). The increase in activity remained at days 7 and 28 (day 7, 0.94 + 0 . 0 9
2.2
kb --
Statistics Data are presented as the mean +_ S.E. Comparison between groups was determined by Student's t-test or the Willcoxon rank sum test for non-parametric data. Correlations were tested by a commercial statistical software package (CSS/Statsoft, U.S.A.). Significance was set at P < 0.05.
Fig. 2. Northern b;ot of rabbit poly (A) + RNA with human UETP cDNA. Rabbit liver poly (A) + RNA (8 p.g) was size fractionated by gel elcetrophoresis, transferred to nitrocellulose membrane and hybridized with human CETP eDNA to identify a single discrete species of rabbit liver mRNA of 2.2 kb.
357 the increase in H L activity and alteration in m R N A levels with cholesterol feeding of the rabbit.
t251
,
1.o51 ,
,
3
~ °s°l
e ,~
I ~"
o.oo J,0
2
7
28
Cholesterol Feeding (Days)
Fig. 3. The effect of cholesterol feeding on HL activity and mRNA. Hepatic lipase activity (o) is expressed as units per liver with each unit representing one mmol of labelled free fatty acid released per hour. Messenger RNA (e) results are expre~ed as changes relative to the values obtained for control animals, which arc ag~igncd an arbitrary value of 1.0. * P < 0.05.
units/liver, P = 0.005, day 28, 0.95 4- 0.07 un;tg/liver, P -- 0.001). The magnitude of increase in H L activity at day 7, as d e t e r m i n e d by the in vitro elution of HL, was 2.6-times control and was identical to the 2.6 fold-increase in H L activity d e t e r m i n e d in post heparin plasma (day 7, 7.15 + 0.87 u n i t s / m l plasma, n = 4, vs. day 0, 2.66 4- 0.10, n = 4), P ffi 0.001). Hepatic lipase m R N A levels did not change significantly in the two control groups (day 0, !.0 + 0.1 vs. day 28, 1.1 + 0.2 absorbance units) and were not increased in the cholesterol fed animals at days 2 or 7. A n increase of 46% was seen at day 28 (Fig. 3). C h a n g e s in specific H L m R N A upon cholesterol feeding were identical in the rabbits that received in vivo tteparin injection. No correlation was found between
I
° to
01
P
6
i
~
ie
I =;
Cholesterol Feeding (Days) Fig. 4. The effect of cholesterol feeding oa CETP activdy and mRNA. Transfer activity (o) is expressed as units per ml plasma.
where units are the rate constant, k [19] for the transfer of radiolabellud tracer from HDL to LDL multiplied by the time of incubation. Messenger RNA (e) results are expressed as changes relative to the values obtained for control animais which arc gs,signedan arbitrary value of 1.0. * P < 0.05.
Cholesteryl ester trans[er protein CETP activity was increased by cholesterol feeding in the rabbit 2.4-fold above control levels over the 4 week period (29.6 _+ 1.9, vs 12.2 + 1.3 u n i t s / m l plasma, P = 0.001, Fig. 4). Control values were not significantly different ( 10.4 + 1.7 u n i t s / m l plasma, n = 4 at day 0, vs 13.6 + 1.7, n = 5 at day 28, P = NS). With cholesterol feeding, activity was unaltered at day 2, but had risen to 1.75-fold relative to control by day 7 (21.1 +_ i.7, P = 0.01) and was 2.2-fold at day 28. Cholesterol feeding produced a rise in rabbit C E T P m R N A levels that correlated significantly with the observed increase in C E T P activity ( r = 0.55, P = 0.02). As with C E T P activity, there was no significant difference in the two control groups (day O, 1.0 + 0.l vs. day 28, i.I + 0.2 absorbance units) and no significant rise in C E T P m R N A levels at day 2 of cholesterol feeding. By day 7, C E T P m R N A levels had increased significantly (2.2 + 0.5 vs 1.0 _+ 0.2 arbitrary units, P = 0.04). The level of C E T P m R N A continued to rise over the 4 week period to reach nearly 4-times the control value at day 28 (3.7 + 0.8, P = 0.02, Fig. 4). There was no change in albumin m R N A over the four week study period (1.0 _+0.1 vs I.I +_0.1 absorbance units). Discussion This study was designed to investigate the regulation of H L and C E T P in the rabbit by exploring the changes that occurred in specific m R N A levels when the activity of each protein was increased by cholesterol feeding. The rise in HL activity that followed cholesterol feeding was not accompanied by comparable changes in specific H L m R N A levels. This discrepancy between the changes in activity and m R N A levels for H L suggests that changes in steady-state m R N A levels are not crucial for the response of H L to cholesterol feeding in the rabbit. In contrast, we have shown that the increases in C E T P m R N A and activity that occur with cholesterol feeding are closely correlated, consistent with an increase in activity being due, at least in part, to the increase in m R N A levels. We chose to assay H L activity in the heparin eluate of liver tissues collected after the animal was sacrificed to avoid the potential effects of in vivo heparin on m R N A levels, however, H L activity was also assayed by the more usual procedure after release of H L into the plasma by heparin injection. A similar rise in H L activity was observed at day 7 using either assay method, supporting the validity of the heparin eluate assay method. There was no measurable H L activity in the heparin eluate of livers removed from animals after the injection of heparin, suggesting little contribution to
358 measured activity from intracellular lipases in the liver eluate. An increase in H L activity was detected as early as the s~.'cond day after the commencement of cholesterol feeding. However, specific rabbit H L m R N A levels were not increased by day 2 or day 7 and only rose by 46% after 28 days of cholesterol feeding. This discrepancy between the changes in H L activity and m R N A content indicate that steady-state m R N A levels are not a key factor in the regulation of H L in the rabbit and suggest that translational and post-translational events may be more important in this regard. Post-translational regulation may be mediated in a number of ways, including: co-factor activation, alteration in the rate of enzyme degradation, or by changes in the content of stored pools. Such forms of regulation have been proposed for lipoprotein lipase (LPL) which is closely related in structure and function to H L [26]. LPL is known to be activated by a cofactor, apolipoprotein CII [27], though the importance of apolipoproteins to H L activity remains unclear with both activation and inhibition of H L activity by apolipoproteins having been reported [28-30]. A decrease in the rate of degradation of H L may contribute to changes in activity and this mechanism has been proposed as the means by which heparin increases the activity of H L in the rat hepatoma cell line, FuSAH [31] and LPL in cultured avian adipocytes [32]. Current evidence does not support the existence of significant intracellular stores of either LPL in cultured adipocytes [32,33] or H L in rat hepatocytes [34]. Further data supporting post translational regulation of LPL was provided in whole animals by studies performed on fasted rats, where a decrease in LPL activity was not associated with a corresponding decrease in m R N A [35]. For CETP, however, there was a strong correlation between the increases in activity and steady-state m R N A levels with cholesterol feeding. In this study, C E T P activity was increased at day seven, and was nearly three fold greater than control values after 28 days of cholesterol feeding. These findings are similar in time-course and magnitude to the reports of Zilversmit and co-workers [9,10]. The level of specific CETP m R N A had begun to increase at day 2 and was also s;gnificantly increased at day 7. These findings support the study of Quinct et al. [13] who observed a rise in C E T P m R N A that was significant 3 days after initiating cholesterol feeding. Our study combines these observations and demonstrates that the cholesterol induced increases in C E T P activity and m R N A levels are closely related. Data from the present study, in the whole animal,
support evidence from tissue culture that suggests the importance of post-translational events in the regulation of HL. These findings are in contrast to the results from C E T P where, in the same animal model, activity correlates weft with hepatic m R N A content. Acknowledgments This work was supported by a grant from the National Health and Medical Research Council of Australia. The authors would like to acknowledge the invaluable technical assistance of Ms Debra Ramsey. References
1 Jenser. (~.L. et al. 0980)J. Biol. Chem. 255, 1141-1148. 2 Jackson, R.L. (1983) in The Enzymes (Boyer, P.D., ed.), Vol. XVi, p. 141-181, Academic Press, New York. 3 Barter PJ. et al. (1987) Am. Heart J. 113, 538-542. 4 Newnham, H.H. and Barter P.J. (1990) Biochim. Biophys. Acta 1044, 57-64. 5 VanZutphen, L.F.M. et al. (1981) Lab. Animals 15, 61-67. 6 BhaUacharyya,A.K. et aL (1989) Arteriosclerosis. 9, 380-389. 7 Heller, F.R. (1983) Biochim. Biophys. Acta 752, 357-360. 8 Jansen, H. et al. (1989) Biochim. Biophys. Aeta I001, 44-49. 9 Son, Y.C. and Zilversmit, D.B. (1986) Arteriosclerosis 6, 345-351. 10 Quig, D.W. and Zilversmit, D.B. (1988) Atherosclerosis 70, 263271. 11 Tikkanen, M.J. and Nikkil~i,A. (1987) Am. Heart J. i 13, 562-567. 12 Kuusi, T. et al. (1983) Atherosclesosis 41, 209-219. 13 Quinct, E.M. el al. (1990) J. Clin. Invest. 85, 357-363. 14 Clay, M.A. et aL (1989) Biochim. Biophys. Acta 1002,173-181. 15 Busch, [;.J. et al. (1989) J. Biol. Chem. 264, 9527-9532. 16 Taskinen, M.R. et al. (1979) Clin. Chim. Acla 104,107-117. 17 Hununen, J.K. et al. (1975) Clin. Chim. Acta 63, 335-347. 18 Tollefsen, J.H. et al. (19881Am. J. Physiol. 255, E894-002. 19 Pannaik. N.M. et al. (1978) Biochim. Biophys. A¢la 530, 428-438. 20 Chomczynski, P. and Sacchi, N. (1987) Anal. Biochem. 162, 156-159. 21 Arty, H. and Leder, P. (1972) Proc. Natl. Acad. Sci. USA 69, 1408-1412. 22 Martin, G.A. et al. (1987) J. Biol. Chem. 263, 10907-10914. 23 Drayna. D. et al. (19[;7) Nature 327, 632-634. 24 [;ambrook, J., Fritsch, E.F. and Maniatis, T. (1989) in Molecular Cloning. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. 25 Kuusi, T. et al. (1979) FEB[; Len. 104, 384-388. 26 Ben-Zeev, O. et al. (1987) Biochim. Biophys. Acta 919,13-20. 27 Bier, D.M. and Havel, RJ. (1970) J. Lipid Res. 11,565-570. 28 [;hinomiya, M. et al. (19821Biochim. Biophys. Acta 713, 292-299. 29 Kubo, M. el al. (1981) Biochem. Biophys. Res. Commun. 100, 261-266. 30 Jahn, C.E. et al. 0983) Eur. J. Biochem. 131, 25-29. 31 Cisar, L.A. et al. (1989) Biochim. Biophys. Acta 1004,196-204. 32 Cupp, M. el al. (1987) J. Biol. Chem. 262, 6383-0388. 33 [;emb, H. and Olivecrona, T. (1987) Biochim. Biophys. Acta 921, 104-115. 34 Laposata, E.A. et al. (1986)J. Biol. Chem. 262, 5333-5338. 35 Dooliule, M.H. et al. (1990) J. Biol. Chem. 265, 4570-4577.
354
The regulation of hepatic lipase and cholesteryl ester transfer protein activity in the cholesterol fed rabbit R o d e r i c J. W a r r e n l, D a v i d L E b e r t t, P h i l i p J. B a r t e r 2 a n d A l a n a M i t c h e l l 1 I Baker Medical Research Institute, Prahran, Victoria (Australia) and 2 Wollongong Unh'ersity. Wollongong (Australia)
(Received 12 April 1991)
Key words: Hepatic lipase;Lipase;Cholesteroldiet; Cholesterylester transfer protein; (Rabbit) Hepatic lipase (HL) and cholesteryl ester transfer protein (CETP) activities are both increased in the rabbit by cholesterol feeding. The in vivo regulation of HL and CETP were explored by examining changes in specific steady-state mRNA levels upon cholesterol feeding. On feeding rabbits cholesterol, HL activity increased 3-fold after 2 days and remained at 2.6-times the control value at 28 days. Specific rabbit HL mRNA levels were assessed by dot blot analysis of liver poly (At + RNA hybridized with the h u m a n HL eDNA. No siRn|flcant changes in liver HL mRNA accompanied the increase in activity seen at days 2 and 7. At day 28 a modest rise of 46% was observed. A significant rise in CETP activity, evident 7 days after the commencement of cholesterol feeding, was maintained until day 28 when it was 2A-times the control value. Using the h u m a n CETP eDNA as probe, rabbit liver CETP mRNA was also found to increase by day 7, rising to 3.7-times control by day 28. The strong temporal relationship between the rise in CETP activity and mRNA (r = 0.55, P = 0.02) s u u e s t s that the regulation of CETP may be primarily effected by the levels of specific mRNA. in contrast, the discordance between levels of lipase activity and mRNA suggests that post-transcriptional events may be more important in the regulation of HL in the cholesterol fed rabbit.
Introduction Hepatic lipase (HL) and cholesteryl ester transfer protein (CETP) are two factors that play a fundamental role in determining the composition of high-density lipoproteins (HDL) [1-4]. Their activities can be influenced by a variety of interventions, including diet, steroid administration and exercise [5-12]. Cholesterol feeding is known to increase the activity of both HL and CETP in the rabbit [5,9,10]. Such an increase in activity may occur through an increase in m R N A levels, leading to enhanced synthesis of new protein; by alterations in post-transcrlptional events; or by a combination of mechanisms. Although it has been suggested that an increase in liver CETP m R N A accompa-
nies a rise in CETP protein mass and activity [13], little is known about the regulation of HL activity. To explore the regulation of HL and CETP we examined the changes that occurred in specific liver m R N A levels when the activities of HL and CETP were increased by cholesterol feeding in the rabbit. We report that the low basal HL activity previously reported in the rabbit [14] was increased with cholesterol feeding. There was, however, no accompanying rise in specific liver HL m R N A levels. In contrast, CETP activity and specific liver m R N A levels were both increased by cholesterol feeding and displayed a strong temporal relationship. These findings support the importance of m R N A levels in the regulation of CETP and suggest that the regulation of HL is less dependent on steady-state m R N A levels and is likely to be predominantly post-transcriptlonal.
Abbreviations: HI_.,hepatic lipase; CETP, choleslerylester transfer protein; SDS, sodium dodeeyl sulfate; LPL, lipoprotein lipase; kb, kilobases.
Materials and Methods
Correspondence: R.J. Warren, Baker Medical Research Institute, CommercialRd.. Prahran, Victoria. 3181. Australia.
Unless specified, all reagents were obtained from the Sigma Chemical (St. Louis, MO, U.S.A.).
Chemicals
355
Animals and diet A multi-coloured strain of rabbits maintained at the Baker Medical Research Institute and male SpragueDawley rats (200 g) were used in these studies. For the cholesterol feeding study, female rabbits weighing 2.02.5 kg, previously maintained on commercial rabbit chow (4% fat w/w, no cholesterol, Clark King, Australia), were divided into two groups receiving either commercial chow or the same chow supplemented with 0.5% (w/w) cholesterol Five cholesterol fed animals were killed at each time point (days 2, 7 and 28) while separate control groups were killed at days 0 (n = 4) and 28 (n = 5). The results of all assays did not differ significantly between the two control groups and the averages of the results in these two groups were used for comparisons. The animals were housed individually, received water ad libitt,m and were weighed weekly.
Hepatic lipase activity Hepatic lipase is usually assayed in vivo in post heparin plasma [2]. In recent in vitro studies, however, heparin has been reported to produce marked increases in the content of HL m R N A in Hep G2 cells, albeit after prolonged treatment periods of 22 h or greater [15]. Since beparin treatment may also influence HL m R N A levels in vivo, an alternative method of obtaining HL was used to avoid the in vivo administration of heparin. The method was adapted from that used by Taskinen et al. [16] to elute lipoprotein lipase (LPL) from small samples of adipose tissue. Briefly, 50 mg pieces of fresh liver tissue were cut finely with a razor blade and incubated at 28°C for 40 rain in 600/zl of Kxebs-Ringer buffered with 0.1 M Tris-HC! buffer (pH 8.4), containing I g/100 ml of bovine serum albumin and 2.5 I.U. of heparin. An aliquot of the heparin eluate (150 /zl) was added to 500 /~l of an artificial triolein emulsion, and incubated at 28°C for 60 min in the presence of 1 M NaCI to inhibit LPL, as described by Huttenen et al. [17]. From similarly prepared eluates of rat liver, where the HL activity was much greater, we were able to demonstrate that the assay was linear with respect to both time and enzyme concentration in the range of values obtained from the rabbit liver eluates. For comparison of HL activity between rabbit and rat, activity is expressed as units per gram of liver tissue, with each unit representing i /zmol of labelled free fatty acid released per hour. These results are the mean of triplicate samples for each animal. In the cholesterol feeding study, where only rabbits were assessed, activity is expressed as units per liver since liver weight increased with cholesterol feeding due to lipid accumulation (Day 0 control, 83.3 ± 2.9 g; Day 2, 74.6 + 6.3 g; Day 7, 84.2 + 6.8 g; Day 28, 99.6 + 4.8 g; Day 28 control 77.8 ± 2.7 g). There was no significant change in rabbit weights. As the rat
is known to have high HL activity, a sample of fresh rat liver was included for assay at each time p'~int as a positive control. The results for these samples varied by less than 10%. To confirm the validity of the results of the HL activity from liver eluates, HL activity was also measured in post heparin plasma in a separate group of rabbits after 7 days of cholesterol feeding, and coinpared with the in vitro method. Plasma was obtained 15 rain after the injection of heparin (100 U / k g ) and samples (10 p.I) assayed for HL activity by the method of Huttenen et al. [17]. Activities are expressed as units per ml of post heparin plasma (1 unit = i p.mol free fatty acid released per h).
CETP actit'ity CETP activity was measured as the capacity of lipoprotein-deficient rabbit plasma to facilitate the transfer of radioactively labellod cholesteryl ester from human high-density lipoprotein (HDL) 3 (d = 1.13 g / m l to i.21 g / m l ) to human low-density lipoprotein (LDL) (d = 1.019 g / m l to 1.055 g / m l ) [18]. The source of CETP was the rabbit plasma fraction of d > 1.25 g/ml. Transfer activity is expressed as units per ml plasma, where units are the rate constant k [19] for the transfer of radiolabelled tracer from HDL to LDL multiplied by the time of incubation.
Messenger RNA ~.~'.antification Total RNA from rabbit and rat was isolated from liver tissues by the method of Chomczynski and Sacchi [20] and enriched for poly (A) + RNA by oligo (dT) selection [21]. To measure changes in specific m R N A resulting from cholesterol feeding, 0.5, 1.5 and 3.0/zg aliquots of rabbit liver poly (A) + RIqA were applied to nitrocellulose (Bio-Rad miniblot system) and probed with human HL and human CETP eDNA's. The human HL eDNA was obtained by screening a human liver lambda gt 10 library with rat HL eDNA (provided by Dr. M. Schotz, UCLA). The clone was confirmed as full-length by dideoxy chain termination sequencing [22]. The human CETP cDNA was a gift from Dis. R. Lawn and D. Drayna, Genentech [23]. Both probes were labelled by random priming to a specific activity of 10¢ cpm//zg DNA [24]. On Northern blot analysis, the human HL eDNA hybridized to a discrete rabbit liver m R N A species of 1.6 kb using a final wash of 0.2 x SSC (! × SSC: 15 mM sodium citrate, 150 mM sodium chloride, pH 7.0.), 0.1% sodium dodecyl sulphate (SDS) at 55°C (Fig. I). The human CETP produced a discrete band on Northern blot analysis of rabbit liver poly (A) + RNA using a final wash of 0.5 × SSC, 0.1% SDS at 55°C (Fig. 2). Dot blot autoradiographs were scanned by laser densitometry (UItrascan XL, LKB Bromma, Sweden) to quantify specific m R N A levels. The varying concentrations of
356
Results
Hepatic lipase
1.6 kb -- O
Fig. 1. Northern blot of rabbit poly (A) ~ RNA with the human HL eDNA. Rabbit liver poly (A) ÷ RNA (8 ttg) was size fractionated by gel electrophoresis, transferred to nitrocellulose membrane and hybridized with human HL eDNA to identify a single discrete species of rabbit liver mRNA of 1.6 kb.
m R N A used produced a linear range of hybridization signals and only data in the linear range of the signal response were considered. To correct the data for differences in the m R N A content between dots, the total m R N A content of each sample was normalized by probing the stripped filters with oligo(dT)30, endlabelled with [32p]ATP [24]. The same filters were again stripped and probed with a rat albumin e D N A (provided by Dr. G. Howlett, Melbourne University), labelled to a specific activity of 10 ° c p m / / z g , to ensure that any observed changes in H L m R N A were specific. Results are expressed as changes relative to the values obtained for control animals which were assigned an arbitrary value of 1.0.
Hepatic lipase activity in the heparin eluate of fresh rabbit liver was less than 10% of the corresponding activity from the rat (0.45 + 0.06 u n i t s / g liver, n = 9 vs 5.17 + 0.04, n = 7, P = 0.001). These findings are consistent with previous reports that have shown the rabbit to have low H L activity in post-heparin plasma when compared to other species such as rat and man [14,25]. The high salt lipolytic activity detected in the heparin eluate of fresh rabbit liver was reduced to zero if the animals received a prior injection of heparin (results not shown). This finding suggests that intracellular liver lipases do not contribute significantly to the lipolytic activity measured in the liver eluate assay. Cholesterol feeding produced a significant early and sustained rise in H L activity (Fig. 3), relative to control values which did not alter significantly over the study period (0.36 + 0.10 u n i t s / l i v e r , n = 4 at day 0, and 0.37 + 0.04, n = 5 at day 28). A t day 2 of cholesterol feeding, H L activity was 3.0-times control (1.10 + 0.27 units/liver, P = 0.03). The increase in activity remained at days 7 and 28 (day 7, 0.94 + 0 . 0 9
2.2
kb --
Statistics Data are presented as the mean +_ S.E. Comparison between groups was determined by Student's t-test or the Willcoxon rank sum test for non-parametric data. Correlations were tested by a commercial statistical software package (CSS/Statsoft, U.S.A.). Significance was set at P < 0.05.
Fig. 2. Northern b;ot of rabbit poly (A) + RNA with human UETP cDNA. Rabbit liver poly (A) + RNA (8 p.g) was size fractionated by gel elcetrophoresis, transferred to nitrocellulose membrane and hybridized with human CETP eDNA to identify a single discrete species of rabbit liver mRNA of 2.2 kb.
357 the increase in H L activity and alteration in m R N A levels with cholesterol feeding of the rabbit.
t251
,
1.o51 ,
,
3
~ °s°l
e ,~
I ~"
o.oo J,0
2
7
28
Cholesterol Feeding (Days)
Fig. 3. The effect of cholesterol feeding on HL activity and mRNA. Hepatic lipase activity (o) is expressed as units per liver with each unit representing one mmol of labelled free fatty acid released per hour. Messenger RNA (e) results are expre~ed as changes relative to the values obtained for control animals, which arc ag~igncd an arbitrary value of 1.0. * P < 0.05.
units/liver, P = 0.005, day 28, 0.95 4- 0.07 un;tg/liver, P -- 0.001). The magnitude of increase in H L activity at day 7, as d e t e r m i n e d by the in vitro elution of HL, was 2.6-times control and was identical to the 2.6 fold-increase in H L activity d e t e r m i n e d in post heparin plasma (day 7, 7.15 + 0.87 u n i t s / m l plasma, n = 4, vs. day 0, 2.66 4- 0.10, n = 4), P ffi 0.001). Hepatic lipase m R N A levels did not change significantly in the two control groups (day 0, !.0 + 0.1 vs. day 28, 1.1 + 0.2 absorbance units) and were not increased in the cholesterol fed animals at days 2 or 7. A n increase of 46% was seen at day 28 (Fig. 3). C h a n g e s in specific H L m R N A upon cholesterol feeding were identical in the rabbits that received in vivo tteparin injection. No correlation was found between
I
° to
01
P
6
i
~
ie
I =;
Cholesterol Feeding (Days) Fig. 4. The effect of cholesterol feeding oa CETP activdy and mRNA. Transfer activity (o) is expressed as units per ml plasma.
where units are the rate constant, k [19] for the transfer of radiolabellud tracer from HDL to LDL multiplied by the time of incubation. Messenger RNA (e) results are expressed as changes relative to the values obtained for control animais which arc gs,signedan arbitrary value of 1.0. * P < 0.05.
Cholesteryl ester trans[er protein CETP activity was increased by cholesterol feeding in the rabbit 2.4-fold above control levels over the 4 week period (29.6 _+ 1.9, vs 12.2 + 1.3 u n i t s / m l plasma, P = 0.001, Fig. 4). Control values were not significantly different ( 10.4 + 1.7 u n i t s / m l plasma, n = 4 at day 0, vs 13.6 + 1.7, n = 5 at day 28, P = NS). With cholesterol feeding, activity was unaltered at day 2, but had risen to 1.75-fold relative to control by day 7 (21.1 +_ i.7, P = 0.01) and was 2.2-fold at day 28. Cholesterol feeding produced a rise in rabbit C E T P m R N A levels that correlated significantly with the observed increase in C E T P activity ( r = 0.55, P = 0.02). As with C E T P activity, there was no significant difference in the two control groups (day O, 1.0 + 0.l vs. day 28, i.I + 0.2 absorbance units) and no significant rise in C E T P m R N A levels at day 2 of cholesterol feeding. By day 7, C E T P m R N A levels had increased significantly (2.2 + 0.5 vs 1.0 _+ 0.2 arbitrary units, P = 0.04). The level of C E T P m R N A continued to rise over the 4 week period to reach nearly 4-times the control value at day 28 (3.7 + 0.8, P = 0.02, Fig. 4). There was no change in albumin m R N A over the four week study period (1.0 _+0.1 vs I.I +_0.1 absorbance units). Discussion This study was designed to investigate the regulation of H L and C E T P in the rabbit by exploring the changes that occurred in specific m R N A levels when the activity of each protein was increased by cholesterol feeding. The rise in HL activity that followed cholesterol feeding was not accompanied by comparable changes in specific H L m R N A levels. This discrepancy between the changes in activity and m R N A levels for H L suggests that changes in steady-state m R N A levels are not crucial for the response of H L to cholesterol feeding in the rabbit. In contrast, we have shown that the increases in C E T P m R N A and activity that occur with cholesterol feeding are closely correlated, consistent with an increase in activity being due, at least in part, to the increase in m R N A levels. We chose to assay H L activity in the heparin eluate of liver tissues collected after the animal was sacrificed to avoid the potential effects of in vivo heparin on m R N A levels, however, H L activity was also assayed by the more usual procedure after release of H L into the plasma by heparin injection. A similar rise in H L activity was observed at day 7 using either assay method, supporting the validity of the heparin eluate assay method. There was no measurable H L activity in the heparin eluate of livers removed from animals after the injection of heparin, suggesting little contribution to
358 measured activity from intracellular lipases in the liver eluate. An increase in H L activity was detected as early as the s~.'cond day after the commencement of cholesterol feeding. However, specific rabbit H L m R N A levels were not increased by day 2 or day 7 and only rose by 46% after 28 days of cholesterol feeding. This discrepancy between the changes in H L activity and m R N A content indicate that steady-state m R N A levels are not a key factor in the regulation of H L in the rabbit and suggest that translational and post-translational events may be more important in this regard. Post-translational regulation may be mediated in a number of ways, including: co-factor activation, alteration in the rate of enzyme degradation, or by changes in the content of stored pools. Such forms of regulation have been proposed for lipoprotein lipase (LPL) which is closely related in structure and function to H L [26]. LPL is known to be activated by a cofactor, apolipoprotein CII [27], though the importance of apolipoproteins to H L activity remains unclear with both activation and inhibition of H L activity by apolipoproteins having been reported [28-30]. A decrease in the rate of degradation of H L may contribute to changes in activity and this mechanism has been proposed as the means by which heparin increases the activity of H L in the rat hepatoma cell line, FuSAH [31] and LPL in cultured avian adipocytes [32]. Current evidence does not support the existence of significant intracellular stores of either LPL in cultured adipocytes [32,33] or H L in rat hepatocytes [34]. Further data supporting post translational regulation of LPL was provided in whole animals by studies performed on fasted rats, where a decrease in LPL activity was not associated with a corresponding decrease in m R N A [35]. For CETP, however, there was a strong correlation between the increases in activity and steady-state m R N A levels with cholesterol feeding. In this study, C E T P activity was increased at day seven, and was nearly three fold greater than control values after 28 days of cholesterol feeding. These findings are similar in time-course and magnitude to the reports of Zilversmit and co-workers [9,10]. The level of specific CETP m R N A had begun to increase at day 2 and was also s;gnificantly increased at day 7. These findings support the study of Quinct et al. [13] who observed a rise in C E T P m R N A that was significant 3 days after initiating cholesterol feeding. Our study combines these observations and demonstrates that the cholesterol induced increases in C E T P activity and m R N A levels are closely related. Data from the present study, in the whole animal,
support evidence from tissue culture that suggests the importance of post-translational events in the regulation of HL. These findings are in contrast to the results from C E T P where, in the same animal model, activity correlates weft with hepatic m R N A content. Acknowledgments This work was supported by a grant from the National Health and Medical Research Council of Australia. The authors would like to acknowledge the invaluable technical assistance of Ms Debra Ramsey. References
1 Jenser. (~.L. et al. 0980)J. Biol. Chem. 255, 1141-1148. 2 Jackson, R.L. (1983) in The Enzymes (Boyer, P.D., ed.), Vol. XVi, p. 141-181, Academic Press, New York. 3 Barter PJ. et al. (1987) Am. Heart J. 113, 538-542. 4 Newnham, H.H. and Barter P.J. (1990) Biochim. Biophys. Acta 1044, 57-64. 5 VanZutphen, L.F.M. et al. (1981) Lab. Animals 15, 61-67. 6 BhaUacharyya,A.K. et aL (1989) Arteriosclerosis. 9, 380-389. 7 Heller, F.R. (1983) Biochim. Biophys. Acta 752, 357-360. 8 Jansen, H. et al. (1989) Biochim. Biophys. Aeta I001, 44-49. 9 Son, Y.C. and Zilversmit, D.B. (1986) Arteriosclerosis 6, 345-351. 10 Quig, D.W. and Zilversmit, D.B. (1988) Atherosclerosis 70, 263271. 11 Tikkanen, M.J. and Nikkil~i,A. (1987) Am. Heart J. i 13, 562-567. 12 Kuusi, T. et al. (1983) Atherosclesosis 41, 209-219. 13 Quinct, E.M. el al. (1990) J. Clin. Invest. 85, 357-363. 14 Clay, M.A. et aL (1989) Biochim. Biophys. Acta 1002,173-181. 15 Busch, [;.J. et al. (1989) J. Biol. Chem. 264, 9527-9532. 16 Taskinen, M.R. et al. (1979) Clin. Chim. Acla 104,107-117. 17 Hununen, J.K. et al. (1975) Clin. Chim. Acta 63, 335-347. 18 Tollefsen, J.H. et al. (19881Am. J. Physiol. 255, E894-002. 19 Pannaik. N.M. et al. (1978) Biochim. Biophys. A¢la 530, 428-438. 20 Chomczynski, P. and Sacchi, N. (1987) Anal. Biochem. 162, 156-159. 21 Arty, H. and Leder, P. (1972) Proc. Natl. Acad. Sci. USA 69, 1408-1412. 22 Martin, G.A. et al. (1987) J. Biol. Chem. 263, 10907-10914. 23 Drayna. D. et al. (19[;7) Nature 327, 632-634. 24 [;ambrook, J., Fritsch, E.F. and Maniatis, T. (1989) in Molecular Cloning. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. 25 Kuusi, T. et al. (1979) FEB[; Len. 104, 384-388. 26 Ben-Zeev, O. et al. (1987) Biochim. Biophys. Acta 919,13-20. 27 Bier, D.M. and Havel, RJ. (1970) J. Lipid Res. 11,565-570. 28 [;hinomiya, M. et al. (19821Biochim. Biophys. Acta 713, 292-299. 29 Kubo, M. el al. (1981) Biochem. Biophys. Res. Commun. 100, 261-266. 30 Jahn, C.E. et al. 0983) Eur. J. Biochem. 131, 25-29. 31 Cisar, L.A. et al. (1989) Biochim. Biophys. Acta 1004,196-204. 32 Cupp, M. el al. (1987) J. Biol. Chem. 262, 6383-0388. 33 [;emb, H. and Olivecrona, T. (1987) Biochim. Biophys. Acta 921, 104-115. 34 Laposata, E.A. et al. (1986)J. Biol. Chem. 262, 5333-5338. 35 Dooliule, M.H. et al. (1990) J. Biol. Chem. 265, 4570-4577.
