Ability of Six Different Lipoprotein Fractions to Regulate the Rate of Hepatic Cholesterogenesis in Vito*

Two in uiuo assay procedures were used to study the inhibitory activity of cholesterol carried in three intestinal lymph and three serum lipoprotein fractions on the rate of cholesterol synthesis in the liver. In the first preparation, different lipoproteins were injected intravenously as a bolus into rats at the mid-light phase of the diurnal light cycle, following which they were killed 12 hours later at the mid-dark phase of the cycle. Using this assay, three intestinal lymph lipoprotein fractions of varying S, values all produced a similar degree of inhibition which averaged approximately ll%/mg of cholesterol injected. The serum lipoprotein fractions caused only about one-third this amount of inhibition. Detailed analysis of events occurring within the liver during this 12-hour assay period revealed that there were marked differences in the rate of net cholesterol uptake into the liver and in the rate of net removal of cholesterol esters from the liver following injection of each of these different lipoprotein fractions. The amount of inhibition of sterol synthesis produced by any fraction was proportional to the product of the incremental increase in hepatic cholesterol

. Recently, we have characterized such an assay system in the rat in detail using unfractionated intestinal lipoproteins (4). In these stud-   The incorporation rates of [1-"Cloctanoate into cholesterol and CO, were then corrected for intramitochondrial dilution of the specific activity of the acetyl-CoA pool from the oxidation of endogenous substrates using the specific activity of the total ketones (16,17). Rates of cholesterogenesis or CO, production were calculated as the nanomoles and micromoles, respectively, of acetyl-CoA, i.e. C, units, incorporated into cholesterol and CO, per gram wet weight of liver slices per hour of incubation, i.e. nanomoles g-' hour-' and micromoles g-' hour-'.
These rates of C, flux into cholesterol mirror exactly the rates of HMG-CoA reductase activity in the same livers (17). Aliquots of blood, liver, and the lipoprotein fraction injected were also obtained for determination of total cholesterol and cholesterol ester content (18,19).
Mathematical Treatment of Data-Where appropriate, mean values for groups of data are given + 1 S.E. For correlating two variables, linear regression curves were fitted to the data obtained from individual animals and have the usual form of y = a + bx, where a equals the intercept on the y axis when x equals zero and b is the proportionality constant between the two variables x andy, The values of a and b are given along with + 1 S.D. As has been reported before for cholesterol feeding (2) and for lipoprotein injection (4,15), the rate of hepatic cholesterol synthesis appears to vary in a log-linear fashion with the amount of cholesterol administered to the animal or with the hepatic cholesterol ester content.
Hence, in studies such as those shown in Fig. 2 1. The characteristics of hepatic metabolism in assay animals used for both the bolus injection and continuous infusion of various lipoprotein fractions. Animals weighing 200 to 240 g were subjected to light cycling for 2.5 weeks. Groups of rats were then killed every 6 hours for 36 hours and rates of cholesterogenesis (A), CO, production (B), and ketone synthesis (C) were measured in liver slices. In addition, the dry weight of the gastric contents was measured at the time the animals were killed (D). Simultaneously, other animals were subjected to cannulation of the intestinal lymphatic duct and the rate of cholesterol delivery into the intestinal lymph was measured over a l-hour period. These data also are shown in this figure (E) and are given as the milligrams of lymphatic cholesterol obtained per hour from animals cannulated at various times during the light cycle. Mean values * 1 S.E. are given for three to six animals in each group.
top of Fig. 1. In the first assay a lipoprotein fraction was injected as a single bolus at the mid-light point of the cycle causing an abrupt increase in the serum cholesterol level that was proportional to the amount of cholesterol administered. The animals were then killed 12 hours later at the mid-dark point of the light cycle. This assay procedure is relatively easy to perform and essentially tests the ability of a given lipoprotein to suppress the expected rise in cholesterol synthetic activity seen with the onset of the dark cycle. The second assay procedure involves the continuous infusion of a lipoprotein fraction for 24 hours beginning at the mid-dark phase of the light cycle. In this procedure the lipoproteins are delivered to the liver in a continuous and, presumably, more physiological manner but the assay is technically more difficult to perform and the results are more complex to interpret since the inhibitory activity of the lipoproteins is superimposed upon a fluctuating base-line level of cholesterol synthetic activity. Representative values obtained using the single bolus injection procedure are shown in Fig. 2 where 0 to 9 mg 100 g-' of cholesterol contained in either the S, >8000 intestinal lipoprotein fraction or the whole serum lipoprotein fraction was injected into the assay animals at the mid-light phase of the light cycle. In the 40 animals injected with the lipoproteins from intestinal lymph the level of cholesterol esters in the liver 12 hours later was a linear function of the amount of cholesterol administered (A). Under these circumstances cholesterol esters increased by 0.108 f 0.011 mg g-' for each milligram 100 g-' cholesterol administered in the S, > 8000 fraction. In contrast, in the 14 rats injected with the same amounts of cholesterol 8707 carried in serum lipoproteins, hepatic cholesterol esters manifested essentially no increase in cholesterol ester content at the end of 12 hours. As seen in Panel B, the rate of synthesis of cholesterol by the liver was markedly suppressed by the injection of the lipoproteins from intestinal lymph but was much less inhibited by the injection of lipoproteins from serum. The slope of the regression curves indicated that the fractional rate of inhibition equaled 0.035 for each milligram of cholesterol injected in the serum lipoproteins but was nearly 3-fold greater (0.106) for cholesterol carried in the S, >8000 fraction. These alterations in cholesterol ester content and rate of cholesterol synthesis occurred under circumstances where the rate of hepatic ketone synthesis (F) and C, flux into CO, (E) did not differ significantly from values found in control animals. Furthermore, the serum cholesterol levels had returned essentially to normal in both groups 12 hours after the bolus injections (D).
The slopes of the regression lines shown in Fig. 2 give the relationships between the amounts of cholesterol injected and the level of hepatic cholesterol esters (A) and rate of hepatic cholesterol synthesis (B), and between the level of hepatic cholesterol esters and the rate of cholesterol synthesis (C) and are entered in Table II along with similar data obtained after injection of two other fractions of intestinal lipoproteins having S, values of 400 to 8000 and 30 to 400. Three points merit emphasis concerning these data. First, as shown in Column B there are striking differences in the level of cholesterol esters achieved 12 hours after injection of the four different lipoprotein fractions: the incremental increase in esters was highest after injection of the large intestinal lipoproteins (0.108 mg g-l) but progressively declined in value where the vehicle for the administration of the cholesterol was the S, 400-8000 (0.071 mg g-l), S, 30-400 (0.031 mg g-l), and serum (0.002 mg g-') lipoprotein fractions. Thus, as shown in Column c* at the time the animals were killed 39.5% of the cholesterol administered in the S, >8000 fraction could be accounted for in hepatic cholesterol esters while, in contrast, only 0.7% of the cholesterol injected in serum lipoproteins was found in the liver. Second, the fraction rate of inhibition of hepatic cholesterogenesis (Column C) was essentially the same for all three lipoprotein fractions obtained from lymph and varied from 0.097 to 0.112 while the degree of inhibition produced by whole serum lipoproteins was much less and equaled only 0.035. Third, the fractional rate of inhibition of cholesterogenesis in the liver associated with an incremental increase in hepatic cholesterol esters of 1.0 mg gg' also varied with the particular lipoprotein fraction injected (Column D). This last result is similar to our previously reported finding where, after the injection of whole intestinal lymph lipoproteins, there was generally a correlation between inhibition of hepatic cholesterogenesis and an increase in hepatocyte ester content but there was no constant quantitative relationship between these two variables (4). Thus, on the basis of these previous results as well as those of the present experiments, it appears that the decrement in cholesterol synthesis associated with a given increment in hepatocyte cholesterol ester content varies with the type of lipoprotein injected, the time frame of the experiment and whether the lipoprotein fraction was administered as a bolus or as a continuous infusion.
However, the values in Table II were all calculated from data points obtained 12 hours after injection of the lipoprotein fractions. In order to examine the possibility that there were marked differences in the clearance of these particles at shorter to these differences in rates of uptake, there were administered in either of the lipoprotein fractions to the assay animals. In addition, the rate of hepatic cholesterogenesis is also plotted as a function of the level of cholesterol esters found in the liver at the time the animals were killed (C). The linear regression curves were fitted to the data by means of the method of least squares for the 40 animals injected with intestinal lipoproteins and the 14 animals injected with whole serum lipoproteins.
The slopes of the lines in A, B, and C have been entered in Table II along with similar data for other lipoprotein fractions.
The shaded area in each diagram represents the mean value + 1 S.E. of each parameter found in 10 control animals injected with 0.9% NaCl solution. also differences evident in the rates of clearance of cholesterol esters from the liver.
For cholesterol esters in the liver integrated as a function of time.
That this is the case is strongly suggested by the finding that the ratios of the areas under the three curves shown in Panel B, Fig. 3 (S, >8000 (1.00): S, 30 to 400 (1.11): serum lipoproteins (0.34)) are essentially identical with the ratios of the fractional rates of inhibition (1.00/1.06/0.33) listed in Column C of Table  II for the three lipoprotein fractions. Thus, under the precisely controlled conditions of this assay procedure the amount of inhibition of the cholesterogenic pathway appears to be determined directly by the degree of elevation of the cholesterol ester content produced by a given lipoprotein fraction and the time course over which this elevation takes place.
Another conclusion of physiological importance to be drawn from these studies is that serum lipoproteins appear to be relatively ineffective regulators of hepatic cholesterogenesis. However, in the studies shown in Fig. 3 and Table II the serum had been harvested from rats fed a low cholesterol, low fat chow diet where, as described in detail by Lasser et al. (22) and confirmed in this laboratory, the great majority of the serum cholesterol is carried in high density lipoproteins and there is relatively little cholesterol in the fractions with densities <1.070. In order to test the possibility, therefore, that one of these lower density serum lipoprotein fractions might have much higher inhibitory activity a large volume of rat serum was processed to yield sufficient amounts of lipoproteins with densities of < 1.006 and 1.006 to 1.070 to inject three animals with each of these fractions: the fractional rate of inhibition averaged 0.031 for the former and 0.025 for the latter. While the number of animals in this experiment is too small for statistical analysis, the results are consistent with those found after injection of the whole serum lipoprotein fraction and suggest that all of the major serum lipoproteins are much less effective as inhibitors of hepatic cholesterogenesis than are the lipoproteins of intestinal origin.
We next turned to the second type of assay procedure in order to test the regulatory capacity of these different lipoprotein fractions under the more physiological circumstance where they were administered in low concentrations by continuous intravenous infusion for a 24-hour period. The other points are mean values * 1 S.E. for 4 to 10 animals in each group.

lipoproteins
were utilized in these studies. However, the serum lipoproteins were obtained from rats fed a high cholesterol, high fat diet in order to increase the amount of cholesterol carried in the lower density fractions so that sufficient quantities of cholesterol in serum lipoproteins with densities of <1.006 and 1.006 to 1.070 as well as 1.070 to 1.215 could be obtained to test in these assays.
These studies, summarized in Table III, again show significant differences between the ability of the six lipoproteins tested to cause net increases in cholesterol ester content in the liver and to inhibit the rate of hepatic cholesterol synthesis. As shown in Column B, the observed increase in cholesterol esters was greatest with the S, >8000 fraction (0.053) and decreased as the intestinal lipoprotein fractions with S, values of 400 to 8000 (0.014) and 30 to 400 (0.004) were injected. Furthermore, all three serum lipoprotein fractions were much less effective in causing a net increase in cholesterol ester content than the S, >8000 intestinal lipoprotein fraction, and, because of variation in the data, no significant difference was evident among these three fractions. At the time the animals were killed 19.9% of the administered dose of cholesterol could be accounted for in the ester fraction in the livers of the animals infused with the S, > 8000 fraction whereas significantly lesser amounts were present after injection of the other fractions. These particular results are qualitatively similar to those obtained with the bolus injection assays (Table II). As seen in Column C, however, the fractional rate of inhibition was nearly g-fold greater when the S, >8000 intestinal lipoproteins were infused than when the S, 30 to 400 fractions were administered. However, since the incremental increase in cholesterol esters was disproportionately lower (Column B) for the S, 400 to 8000 and S, 30 to 400 fractions than the fractional rates of inhibition (Column C), the amount of inhibition manifested per mg of incremental increase in cholesterol esters increased from 1.17 (S, >SOOO) to 4.21 (S, 30 to 400) as shown in Column D. Again, the three serum lipoprotein fractions were able to inhibit hepatic cholesterogenesis but at rates that were only about 14 to 35% of that seen with the S, >8000 intestinal lipoproteins, and, as with the bolus injections, there was no significant