Hepatic cholesterol metabolism and resistance to dietary cholesterol in LXRbeta-deficient mice

Abstract

The nuclear oxysterol-receptor paralogues LXRalpha and LXRbeta share a high degree of amino acid identity and bind endogenous oxysterol ligands with similar affinities. While LXRalpha has been established as an important regulator of cholesterol catabolism in cholesterol-fed mice, little is known about the function of LXRbeta in vivo. We have generated mouse lines with targeted disruptions of each of these LXR receptors and have compared their responses to dietary cholesterol. Serum and hepatic cholesterol levels and lipoprotein profiles of cholesterol-fed animals revealed no significant differences between LXRbeta(-/-) and wild-type mice. Steady-state mRNA levels of 3-hydroxy-3-methylglutaryl coenzyme A reductase, farnesyl diphosphate synthase, and squalene synthase were increased in LXRbeta(-/-) mice compared with LXRbeta(+/+) mice, when fed standard chow. The mRNA levels for cholesterol 7alpha-hydroxylase, oxysterol 7alpha-hydroxylase, sterol 12alpha-hydroxylase, and sterol 27-hydroxylase, respectively, were comparable in these strains, both on standard and 2% cholesterol chow. Our results indicate that LXRbeta(-/-) mice - in contrast to LXRalpha(-/-) mice - maintain their resistance to dietary cholesterol, despite subtle effects on the expression of genes coding for enzymes involved in lipid metabolism. Thus, our data indicate that LXRbeta has no complete overlapping function compared with LXRalpha in the liver.

Figure 1

Strategy for the creation of LXRβ- and LXRα-deficient mice using a Cre/loxP system. (a and b) The genomic organization of LXRβ and LXRα, the corresponding targeting constructs containing the loxP sites, the predicted homologous recombinant alleles, and the LXR locus after Cre-mediated deletion are shown. Filled boxes indicate exons, arrows loxP sites, open boxes neomycin resistance and thymidine kinase genes. Filled bars indicate probes used for Southern blot analyses. H, HindIII; B, BglII; DBD, DNA-binding domain; ATG, translational start codon; TAG/TGA, stop codons. (c) Southern blot analysis. DNA was prepared from mouse tails and digested with BglII or HindIII as described in Methods. (d) Northern blot analysis. Poly A(+)-enriched RNA was extracted from a pool of livers (five females) from each respective group. Hybridization was performed by using full-length LXRα and LXRβ cDNAs as probes.

Figure 2

Macroscopic appearance of livers from LXRα^+/+, LXRα^–/–, and LXRβ^–/– mice. Animals were fed a standard rodent diet or a 2% cholesterol-enriched diet for 28 days. Livers were removed immediately after the animals were sacrificed.

Figure 3

(a and b) Total cholesterol content in hepatic lipid extracts and in serum. Animals were fed a standard rodent diet or 2% cholesterol-enriched diet for 28 days. Livers were removed immediately after the animals were sacrificed. Blood was obtained by cardiac puncture and serum was collected by centrifugation. Total cholesterol was determined enzymatically in hepatic lipid extracts and serum from ten animals per experimental group. All values are expressed as mean + SEM. P values are versus wild-type. The results from two independent analyses are shown in a and b. NS, not significant.

Figure 4

Serum lipoprotein profiles after separation by FPLC in LXRβ^–/– (top) and LXRα^–/– (bottom) mice. Animals were fed standard rodent diet (left) or challenged with 2% cholesterol (right) for 28 days. Ten microliters of pooled serum from each group was directly separated on a FPLC Superose 6B column. Cholesterol content was determined using the commercially available kit MPR 21442350 (Roche Molecular Biochemicals, Indianapolis, Indiana, USA) that was mixed online with the separated lipoproteins at a flow rate of 40 + 40 μl/min. The mixture was passed over a 37°C reaction coil and adsorbance was monitored every 20 seconds at 500 nm. The adsorbance profiles from two independent analyses are shown in the upper and lower panels.

Figure 5

Measurement of 24-hydroxycholesterol (24-OH-Chol) (a) and of 27-hydroxycholesterol (27-OH-Chol) (b) in serum of LXRα^–/– or LXRβ^–/– mice and their respective wild-type controls. All values were obtained from pooled serum of the same animals as in Figure 3.

Figure 6

Expression of various lipid-regulating genes and of genes coding for enzymes in the neutral and acidic bile acid pathway in LXRβ^–/– mice (a), LXRα^–/– mice (b), and corresponding wild-type controls assayed by Northern blot analysis. Used were 2 or 5 μg/lane poly A⁺-enriched RNA, extracted from a pool of livers from five female mice fed as indicated. Filters were hybridized with [α-³²P] dCTP-labeled cDNA probes, generated as described in Methods. Signals were standardized to the corresponding GAPDH expression. Data show the fold change compared with the respective wild-type control group fed the same diet. The results are representative of two or more independent determinations.