Atherosclerosis and Lipoproteins
Studies on the Cholesterol-Free Mouse
Strong Activation of LXR-Regulated Hepatic Genes When Replacing
Cholesterol With Desmosterol
Maura Heverin, Steve Meaney, Anat Brafman, Millicent Shafir, Maria Olin, Marjan Shafaati,
Sara von Bahr, Lilian Larsson, Anita Lövgren-Sandblom, Ulf Diczfalusy, Paolo Parini,
Elena Feinstein, Ingemar Björkhem
Downloaded from http://atvb.ahajournals.org/ by guest on June 17, 2017
Objective—Characterization of cholesterol homeostasis in male mice with a genetic inactivation of 3␤-hydroxysteroid⌬24-reductase, causing replacement of almost all cholesterol with desmosterol.
Methods and Results—There was an increase in hepatic sterol synthesis and markedly increased fecal loss of neutral
sterols. Fecal excretion of bile acids was similar in knockout mice and in controls. The composition of bile acids was
changed, with reduced formation of cholic acid. It was shown that both Cyp7a1 and Cyp27a1 are active toward
desmosterol, consistent with the formation of normal bile acids from this steroid. The levels of plant sterols were
markedly reduced. Hepatic mRNA levels of 3-hydroxy-3-methylglutaryl (HMG) coenzyme A (CoA) reductase,
Srebp-1c, Srebp-2, Cyp7a1, Abcg5, Abcg8, and Fas were all significantly increased.
Conclusions—The changes in hepatic mRNA levels in combination with increased biliary and fecal excretion of neutral
steroids, reduced tissue levels of plant sterols, increased plasma levels of triglyceride-rich VLDL, are consistent with
a strong activation of LXR-targeted genes. The markedly increased fecal loss of neutral sterols may explain the fact that
the Dhcr24⫺/⫺ mice do not accumulate dietary cholesterol. The study illustrates the importance of the integrity of the
cholesterol structure—presence of a double bond in the steroid side-chain is compatible with life but is associated with
serious disturbances in sterol homeostasis. (Arterioscler Thromb Vasc Biol. 2007;27:2191-2197.)
Key Words: desmosterol 䡲 sterol deficiency 䡲 sterol malabsorption 䡲 LXR-activation 䡲 Abcg5
I
displayed about 25% and 40% reduction of total steroids,
respectively. Cholesterol, desmosterol, and total sterol levels
in heterozygous mice were identical to those of the wild-type.
It was considered to be of interest to characterize cholesterol homeostasis in the Dhcr24 knockout mice, in particular
to elucidate the mechanism behind the steroid depletion and
the metabolic consequences of this.
The extra double bond present in the side chain of
desmosterol makes the molecule more polar and affects its
interaction with phospholipids. It has been shown that the
efflux of biosynthetic desmosterol from cells is 3 times more
efficient than that of cholesterol.4 This would be expected to
affect the membrane properties of the steroid and also its
potential to suppress sterol synthesis. On the other hand, it is
known that desmosterol can be sensed by SCAP and induce
a conformational change similar to that caused by
cholesterol.5
Since a ⌬24 double bond is introduced in connection with
normal bile acid biosynthesis, it is possible that normal bile
acids are formed from desmosterol. Whether desmosterol
n a previous work we generated mice with a disruption of
3␤-hydroxysteroid-⌬24-reductase (desmosterol reductase,
Dhcr24), as a result of which cholesterol synthesis is blocked
because of absence of reduction of the C24 –C25 double
bond.1 Surprisingly, mice surviving to full term were viable
and had no gross morphological defects. There were, however, abnormalities in male gonads and in oocyte maturation,
apparent reduction of subcutaneous and mesenterial fat,
apparent decrease of myelination of the central nervous
system, and hypotonia. Because the Dhcr24 knockout mice
are unable to reproduce, they have to be continuously
generated from the corresponding heterozygotes. The relatively mild phenotype is in marked contrast to that of the
human desmosterolosis patients, who have multiple severe
inborn defects.2,3 A plausible reason for the difference could
be the availability of maternal cholesterol during mouse
embryonic development.1
In the adult knockout mouse, desmosterol exceeded cholesterol on a molar basis by 80-fold in the plasma and by
15-fold in the brain. In addition, both brain and plasma
Original received April 17, 2007; final version accepted July 27, 2007.
From the Division of Clinical Chemistry (M.H., S.M., M.O., M.S., S.v.B., L.L., A.L.-S., U.D., P.P., I.B.), Karolinska Institute, Huddinge University
Hospital, Sweden; and Quark Biotech Inc (A.B., M.S., E.F.), Fremont, Calif.
M.H. and S.M. contributed equally to this study.
Correspondence to Dr Ingemar Björkhem, Division of Clinical Chemistry, Karolinska University Hospital, Huddinge, S-14186, Sweden. E-mail
[email protected]
© 2007 American Heart Association, Inc.
Arterioscler Thromb Vasc Biol. is available at http://atvb.ahajournals.org
2191
DOI: 10.1161/ATVBAHA.107.149823
2192
Arterioscler Thromb Vasc Biol.
October 2007
Downloaded from http://atvb.ahajournals.org/ by guest on June 17, 2017
may serve as a substrate in the classical pathway to bile acids
starting with a 7␣-hydroxylation or in the alternative pathway
starting with a 27-hydroxylation is not known.
Very recently it was reported that desmosterol is able to
activate LXR in vitro.6 If this effect is of importance also
under in vivo conditions, one would expect activation of
LXR-targeted genes in our animal model.
All the mice studied in the present work were generated in
2003 in Quark Biotech Inc.1 Some studies on cholesterol
homeostasis in the brain of these animals have been reported.7
More recent attempts to get Dhcr24⫺/⫺ progeny from the
heterozygous breeders in Switzerland,8 Japan,9 in our Swedish laboratory as well as in Quark Biotech have not been
successful, and only very few mice have survived more than
a few days or weeks. The few survivors in the later generations had, however, the same phenotype as the first generations with the same composition of steroids. Why the first
generation of mice survived for a longer time than latter
generations has not been clarified. The limited number of
adult mice possible for us to study has prevented more
detailed investigations including, eg, feeding experiments
with cholesterol. From the observations made thus far on the
unique mice generated in 2003 some important conclusions
can be drawn, however. Because of the limited number of
mice available, and to avoid problems with different genetic
backgrounds, we have only compared Dhcr24⫺/⫺ mice with
Dhcr24⫹/⫺ mice in the present work. In a limited number of
early preliminary experiments (not shown) we showed that
the Dhcr24⫹/⫹ mice are identical to the Dhcr24⫹/⫺ mice with
respect to content of desmosterol and expression of various
genes.
Methods
The detailed methodology is available as a supplement
(http://atvb.ahajournals.org).
Animals
Male mice with a Dhcr24 knockout were generated and maintained
as described previously.1 In most studies the mice had an age of 12
weeks. For further details, see the supplemental materials.
Lipoprotein and Lipid Pattern
Lipoproteins were separated with an automated gel filtration system.
For details, see the supplemental materials.
Sterol Analysis in Plasma and Liver
Levels of cholesterol, desmosterol, and oxysterols in circulation,
plasma, and liver were determined by isotope dilution–mass spectrometry and use of deuterium-labeled internal standards. For details,
see the supplemental materials.
Analyses of Bile Acids
Bile acids in bile were identified after hydrolysis by combined gas
chromatography–mass spectrometry and quantitated either by gas
chromatography or by isotope dilution–mass spectrometry. Unhydrolysed bile acids were analyzed by electrospray-mass spectrometry. For details, see the supplemental materials.
Analyses of Bile Acids and Sterols in Feces
Fecal sterols and bile acids were identified by combined gas
chromatography-mass spectrometry and quantitated by gas chromatography. For details, see the supplemental materials.
Real-Time Polymerase Chain Reaction Analyses
The relative amount of target mRNA was quantitated with use of
TaqMan or SYBR Green systems and the specificity of all the
polymerase chain reaction (PCR) products was confirmed by sequencing. For details, see the supplemental materials.
Statistics
With the exception of the results shown in Figure 4 data are
presented as mean⫾SEM. With one exception (3-hydroxy-3methylglutaryl [HMG] coenzyme A [CoA] reductase activity) the
normal distribution allowed use of Student t test.
Results
Lipoprotein and Lipid Pattern in the Circulation
of the Dhcr24ⴚ/ⴚ Mice
In accordance with the previous work,1 the plasma levels of
3␤-hydroxy-5-unsaturated steroids (99% cholesterol in the
case of Dhcr24⫾mice, about 97% desmosterol in the case of
Dhcr24⫺/⫺ mice) were decreased by almost 50% in the
Dhcr24⫺/⫺ mice as compared with the Dhcr24⫹/⫺ mice. In
contrast, the levels of triglycerides were about double those of
the Dhcr24⫹/⫺ mice.
Figure 1 shows the separation of plasma lipoproteins by
size exclusion chromatography and online determination of
their lipid components in the 2 groups of animals. Most of the
increased triglycerides in plasma of the Dhcr24⫺/⫺ mice was
present in the VLDL fraction. Also total 3␤-hydroxy-5unsaturated steroids and phospholipids were increased in this
fraction. The HDL and LDL fractions of the Dhcr24⫺/⫺ mice
had reduced content of all 3 lipid classes.
Fecal Excretion of Neutral Steroids
The fecal excretion of plant sterols was similar in the 2
different groups of animals (Figure 2A and supplemental
Table II). If taking into account that the knockout mice had a
weight about 25% lower than the heterozygotes and compensating for differences in body weight, the excretion of plant
sterols was about 20% higher from the knockouts.
There was a marked and highly significant (P⬍0.001)
increase in the excretion of C27 neutral steroids (about 4-fold)
from the knockout mice as compared with the heterozygotes
(Figure 2B and supplemental Table III). If compensating for
the lower body weight of the knockout mice, the difference
between the 2 groups was even larger. The fecal excretion
from the knockout mice was mainly desmosterol (about
97%), whereas the excretion from the heterozygotes mainly
consisted of cholesterol (91%).
Composition of Bile Acids in Bile and Feces and
Amounts of Different Bile Acids Excreted in Feces
Using combined gas chromatography–mass spectrometry,
normal bile acids were shown to be present in a hydrolyzed
sample of bile from a knockout mouse. The same bile acids
were found in a bile sample from the wild-type mouse and
from a heterozygote. As shown in Figure 3A, the relative
proportions of the bile acids were different in a sample from
a heterozygote and a knockout mouse. In the bile from the
heterozygote, cholic acid was dominating (54%) followed by
␤-muricholic acid (32%) and ␣-muricholic acid (8%). In the
bile sample from the knockout mouse ␤-muricholic acid was
Heverin et al
Mice With Desmosterolosis
2193
Downloaded from http://atvb.ahajournals.org/ by guest on June 17, 2017
Figure 2. Fecal excretion of plant sterols (A) and neutral sterols (B) from Dhcr24⫺/⫺ and Dhcr24⫹/⫺ mice. Means⫾SEM
from groups of 6 animals, 12 weeks of age, are shown. The
C-27 steroids measured include cholesterol, coprostanone,
epicoprostanol, cholestanol, and desmosterol. The plant sterols measured include sitosterol, sitostanol, campesterol,
methylcoprostanol, ethylcoprostanol, ethylcoprostanone, and
stigmasterol.
Figure 1. Chromatography of plasma from Dhcr24⫺/⫺ and
Dhcr24⫹/⫺ mice. The figures show the mean content of triglycerides (A), total 3␤-hydroxy-5-unsaturated steroids (B), free
3␤-hydroxy-5-unsaturated steroids (C), and phospholipids (C) in
the chromatographic fractions corresponding to VLDL, LDL, and
HDL. The plasma samples were obtained from 3 Dhcr24⫺/⫺ mice
(age 6 to 12 weeks, mean 10 weeks) and 4 Dhcr24⫹/⫺ mice (age
12 to 13 weeks, mean 12 weeks). For details, see Experimental
Procedure.
dominating (53%) followed by cholic acid (19%) and
␣-muricholic acid (7%). The alkaline hydrolysis used in the
above analysis does not give cleave sulfate esters and does
not give information about the type of conjugate present in
bile. To obtain qualitative information about this, samples of
unhydrolysed bile were analyzed by electrospray mass spectrometry. In similarity with a previous report on electrospray
mass spectrometry of mouse bile,10 the major peak appearing
both in the analysis of bile from the heterozygote and that
from the knockout mouse had a m/z 514, corresponding to
taurine conjugate of trihydroxy-bile acid (cholic acid or
muricholic acids). A small peak appeared with m/z 498,
corresponding to taurine conjugate of dihydroxy bile acid
(chenodeoxycholic acid and/or deoxycholic acid). No significant peaks corresponding to sulfated bile acids were seen.
Figure 3B summarizes the results of quantitative measurements of fecal excretion of all the different bile acids in
heterozygotes (n⫽6) and gene knockout males (n⫽6). The
total excretion of bile acids was not significantly different in
the 2 different groups. If taking into account the lower body
weight of the knockout animals, the fecal excretion of bile
acids was however about 25% higher in these animals.
Excretion of cholic acid was significantly lower in the
knockout mice than in the heterozygotes (P⬍0.01). There
was also a tendency toward higher excretion of muricholic
acids in the knockout mice.
Relation Between Bile Acids and Neutral
Sterols in Bile
The very high fecal excretion of neutral sterols may in part be
attributable to an increased biliary excretion, and thus it was
considered important to measure the molar ratio between
neutral sterols and bile acids in bile. In bile from 4 heterozygotes this ratio was found to be 0.06⫾0.02, whereas the
corresponding ratio was 0.34⫾0.10 in 4 knockout mice
(mean⫾SEM, P⬍0.01).
Oxysterols in Circulation
In accordance with previous studies,10 7␣-hydroxycholesterol,
24-hydroxycholesterol and 27-hydroxycholesterol were the
major oxysterols in the circulation of the heterozygotes
2194
Arterioscler Thromb Vasc Biol.
October 2007
Table 2. Levels of Plant Sterols in Plasma and Liver of
Heterozygotes and Knockout Mice
Genotype
Campesterol
Sitosterol
Plasma, ␮g/mL
a
Heterozygotes
36⫾13
13⫾4
Knockouts
2.5⫾1.0
1.3⫾0.4
Liver,b ng/mg tissue
a
Heterozygotes
63⫾5
14⫾1
Knockout
4.5⫾0.2
0.5⫾0.2
Measurements in 3 heterozygotes and 4 knockouts about 3 months of age.
Measurements in 4 heterozygotes and 4 knockouts 3 months of age.
b
Measurements of Plant Sterols in Plasma
and Liver
Downloaded from http://atvb.ahajournals.org/ by guest on June 17, 2017
Levels of plant sterols in plasma and liver may be used as a
surrogate marker for absorption of neutral steroids. As shown
in Table 2 there was a marked and highly significant reduction
in the levels of the 2 plant sterols in both compartments. This
reduction was not caused by a reduced intake of plant sterols
as evident from the similar fecal excretion of plant sterols
from heterozygous and knock out mice (Figure 2).
Measurements of Hepatic mRNA Levels
Figure 4 shows hepatic mRNA levels measured by real-time
PCR in 4 male knockout mice (age 6 to 12 weeks) and 6
heterozygous mice (age 5 to 12 weeks). There was a statistically significant upregulation of Srebp-1c, Srebp-2, HMGCoA reductase, HMG-CoA synthase, Cyp7a1, Abcg5,
Abcg8, and fatty acid synthase. There was a slight reduction
of the mRNA levels corresponding to Cyp27 and a very
marked reduction in the mRNA levels of cyp7b1.
The slight changes in the mRNA levels of the LDL
receptor and squalene epoxidase were not statistically significant. Small heterodimer partner (Shp) as well as Squalene
epoxidase, Cyp8b1, and Abca1 were not affected by the gene
disruption.
Similar results as above were obtained also in microarray
analyses (results not shown).
Figure 3. A, Composition of bile acids in Dhcr24⫹/⫺ and
Dhcr24⫺/⫺ mice. Amounts of individual bile acids in the gallbladder bile of a heterozygote and a knockout male mouse 12
weeks of age were determined by gas chromatography/mass
spectrometry11 and are indicated as a percentage of the entire
pool. B, Fecal excretion of bile acids from the same mice as
those used in Figure 2.
(Table 1). The levels of these oxysterols were reduced by
72% to 93% in the knockout mice. In plasma of the knockout
mice an oxysterol was present that had a polarity slightly
higher than that of 27-hydroxycholesterol. This oxysterol had
a mass spectrum identical with that of 27-hydroxylated
desmosterol produced by action of sterol 27-hydroxylase on
desmosterol (supplemental Figure V). The level of 27hydroxydesmosterol was about 7 times higher than the level
of 27-hydroxycholesterol in the heterozygote. The level of
27-hydroxydesmosterol was found to be present in about the
same high concentration also in the circulation of a female
Dhcr24⫺/⫺ mouse. Attempts to find 7␣-hydroxylated desmosterol in plasma failed both in male and female Dhcr24⫺/⫺
mice.
Table 1.
Western Blotting of ABCG5
The ratio between Abcg5 and ␤-Actin protein was 1.5⫾0.3 in
the analysis of membranes from liver of 3 knockout animals.
The corresponding figure for 6 heterozygotes was 0.6⫾0.1
(mean⫾SEM, P⫽0.01). For the Western blottings, See supplemental Figure VI.
Assay of HMG CoA Reductase Activity
HMG CoA reductase activity was measured in liver microsomes from a group of knockout animals about 12 weeks of
age (n⫽4) and a group of heterozygotes of the same age
Plasma Levels of Oxysterols in Wild-Type and Knockout Mice*
Genotype
Heterozygotes, ng/mL
Knockouts, ng/mL
7a-Hydroxy-Cholesterol
24-Hydroxy-Cholesterol
27-Hydroxy-Cholesterol
65
21
53
⬍1
2
4
15
383
*The analyses were made from pools with 3 animals (age 12 weeks) in each group.
27-Hydroxy-Desmosterol
Heverin et al
Mice With Desmosterolosis
2195
Downloaded from http://atvb.ahajournals.org/ by guest on June 17, 2017
Figure 4. Real-time PCR analyses of hepatic mRNA levels in the Dhcr24⫺/⫺ and Dhcr24⫹/⫺ mice. The relative mRNA levels were calculated according to the Comparative Threshold Cycle method as described in Experimental Procedure and the data shown are the mean
and upper and lower ranges from experiments with 4 Dhcr24⫺/⫺ mice and 6 Dhcr24⫹/⫺ mice. For clarity of presentation the y axis is
depicted with a log scale. The ages of the Dhcr24⫺/⫺ mice ranged from 6 to 12 weeks (mean 8 weeks) and the age of the Dhcr24⫹/⫺
mice from 5 to 12 weeks (mean 9 weeks).
(n⫽4). The activity was almost 5-fold higher in the knockout
mice (n⫽4) than in the heterozygotes (1.5⫾0.5 versus
0.30⫾0.16 nmol/min/mg, P⫽0.02 after log. transformation).
Testing Whether Desmosterol Is a Substrate
for CYP27A1
CYP27A1 was incubated with desmosterol as substrate together with adrenodoxin reductase and adrenodoxin as described previously.11 In a parallel experiment the same
amount of cholesterol was used as substrate. In a typical set
of experiments about 5% of the cholesterol was converted
into 27-hydroxycholesterol. In the parallel experiment with
desmosterol about 13% of this steroid was converted into a
product with a mass spectrum identical to that obtained in
connection with analysis of a serum sample from a Dhcr24⫺/⫺
mouse (see supplemental Figure V).
Testing Whether Desmosterol Is a Substrate
for CYP7A1
CYP7A1 was incubated together with NADPH and NADPH
reductase under the same conditions as previously described.12 In a parallel experiment cholesterol was used as
substrate. In a typical set of experiment the conversion of
cholesterol into 7␣-hydroxycholesterol was about 6%. The
formation of 7␣-hydroxydesmosterol in the parallel experiment was about 5%.
Testing Whether There Is a Formation of
Cholesterol 24,25-Epoxide From Desmosterol
Chloroform extraction of livers of Dhcr24⫺/⫺ as well as livers
of wild-type mice with subsequent work-up, derivatization,
and analysis by combined gas chromatography-mass spectrometry failed to result in detection of the epoxide. As
judged from parallel analysis of the standard compound,
levels of the epoxide ⬎10 ng/g would have been detected
under the conditions used.
Discussion
It is evident from the present animal model that desmosterol
can replace cholesterol in many but not all of its functions. It
has been shown previously that desmosterol can substitute for
cholesterol in cultured fibroblasts without deleterious effects.13 Furthermore most cholesterol in the developing and
early neonatal retina in rats can be replaced with desmosterol,
both structurally and functionally.14
Surprisingly, the ⫺/⫺ mice produced normal bile acids in
about normal amounts. It was shown that desmosterol can
serve as a substrate both in the classical bile acid biosynthesis
pathway starting with a 7␣-hydroxylation and in the alternative pathway starting with a 27-hydroxylation. Interestingly,
the gene coding for the rate-limiting enzyme in the classical
pathway, Cyp7a1, was upregulated. Part of the explanation
for this may be the lower production of cholic acid, the major
suppressor of Cyp7a1 in mice.15 A contributing factor may be
that Cyp7a1 is regulated by a LXR-dependent mechanism.
The gene coding for the rate-limiting enzyme in the
alternative pathway, Cyp27a1, was slightly downregulated by
the gene disruption (Figure 4). That this enzyme is active on
desmosterol in the alternative pathway was evident both from
our in vitro experiment and from the finding that there were
high levels of 27-hydroxylated desmosterol in the circulation.
2196
Arterioscler Thromb Vasc Biol.
October 2007
Downloaded from http://atvb.ahajournals.org/ by guest on June 17, 2017
The fact that we did not find any 7␣-hydroxylated desmosterol in the circulation, but high levels of 27-hydroxylated
desmosterol, is consistent with the possibility that a substantial part of bile acid synthesis in the Dhcr24-deficient mice
may occur by a primary 27-hydroxylation of desmosterol
rather than by a primary 7␣-hydroxylation. In our in vitro
experiments with sterol 27-hydroxylase, desmosterol appeared to be a more efficient substrate than cholesterol. This
is in line with the previous demonstration that the CYP27
activity seems to increase with the polarity of the substrate.11
Because the transport of cholesterol into the inner mitochondrial membranes is a limiting factor for overall sterol 27hydroxylase activity,16 and because desmosterol is able to
pass lipophilic membranes at a considerably higher rate than
cholesterol, this may also be of importance for a high activity
of the sterol 27-hydroxylase. A conversion of cholesterol into
bile acids by the alternative pathway initiated by a 27hydroxylation of cholesterol involves participation of the
oxysterol 7␣-hydroxylase Cyp7b1. The mRNA level of this
enzyme was downregulated in the Dhcr24⫺/⫺ mice. This
finding is in accordance with the recent report by Uppal et al
that activation of LXRs markedly suppresses expression of
Cyp7b1.17 In view of the very high capacity of the oxysterol
7␣-hydroxylase to hydroxylate side-chain oxidized oxysterols,18 the downregulation of the gene may not be critical
for the overall conversion of cholesterol into bile acids by the
alternative pathway.
The alternative pathway for formation of bile acids from
cholesterol is known to give chenodeoxycholic acid and its
metabolites as major products.19 The increased content of
muricholic acids and the reduced content of cholic acid in the
bile of the Dhcr24⫺/⫺ mice is consistent with this.
It is noteworthy that expression of Srebp-1c was higher
than that of Srebp-2, probably reflecting the fact that
Srebp-1c in contrast to Srebp-2 is one of the genes activated
by the LXR-mechanism. Both genes may be upregulated as a
consequence of sterol depletion. It is known that Srebp-2
preferentially activates genes of cholesterol metabolism
whereas Srebp-1c preferentially activates genes of fatty acid
and triglyceride metabolism.20 In this connection it is noteworthy that the gene coding for fatty acid synthase was
markedly upregulated in the Dhcr24⫺/⫺ mice. This effect may
be part of the explanation for the increased plasma levels of
triglyceride-rich VLDL particles in the Dhcr⫺/⫺ mice. It was
shown in the recent in vitro study by Yang et al6 that
desmosterol inhibits the Srebp-2 pathway. If such an inhibition is of importance under the present in vivo conditions, it
is evident that this inhibition is not able to overcome the
stimulatory effect of the sterol depletion.
The very marked effect of the gene disruption on the biliary
and fecal excretion of neutral sterols is of particular interest in
relation to the recent demonstration that desmosterol activates
the nuclear receptor LXR.6 Such an activation would be
expected to stimulate the LXR-targeted genes Abcg5 and
Abcg8 with reduced intestinal absorption and increased
biliary excretion of cholesterol and plant sterols as a consequence. It has been reported that treatment of mice with the
LXR agonist GW3965 results in 74% increase in biliary
secretion of cholesterol and a 2-fold increase in fecal neutral
sterol excretion.21 The even higher fecal excretion of neutral
sterols found in our animal model may be attributable to the
chronic state with a combination of reduced absorption and
increased synthesis of neutral sterols.
It should be noted that although the pattern of expression of
the above genes are in accordance with a general stimulation
of the LXR-target genes, the changes were relatively modest
in relation to the very marked effect observed on fecal
excretion of neutral sterols. Surprisingly, the gene coding for
the cholesterol transporter Abca1, which is also stimulated by
LXR, was not increased. It seems possible that replacing
cholesterol with desmosterol in the membranes may modify
the rate of transcription or the stability of the mRNA species
or both.
The changed distribution of lipids in the different lipoproteins in the circulation (Figure 1) are likely to reflect a
mechanism compensating for the reduced levels of 3␤hydroxy-5-unsaturated steroids in the Dhcr24⫺/⫺ mice. The
VLDL-fraction of these mice contained twice as much sterol
as the corresponding fraction from Dhcr24⫹/⫺ mice, whereas
the situation was reversed for HDL. In theory, a LXRdependent upregulation of the fatty acid synthase may be a
driving force for a higher synthesis of triglycerides and
increased secretion of VLDL.
Because epoxidation of desmosterol gives cholesterol
24,25-epoxide, which is a very efficient ligand and activator
of LXR,22 we tested the hypothesis that this compound is
formed from desmosterol in the Dhcr24⫺/⫺ mice. In spite of
extensive efforts, we failed to demonstrate detectable levels
of the epoxide in the plasma and liver of the knockout
animals. Thus it seems likely that it is desmosterol itself
rather than a metabolite that is the direct activator of the LXR
gene.
It is interesting to compare the present animal model with
the mouse model for the Smith-Lemli-Opitz syndrome, with
a mutation in the Dhcr7 gene, causing accumulation of
7-dehydrocholesterol.23,24 These very short-living mice also
have markedly reduced tissue levels of cholesterol and total
sterols. Despite this, the mRNA levels for HMG-CoA reductase, the LDL receptor and Srebp-2 are essentially normal. In
contrast to mRNA, protein levels and activities of HMG-CoA
reductase are markedly reduced and it was shown that
7-dehydrocholesterol accelerates proteolysis of HMG-CoA
reductase.23 In another study neither 7-dehydrocholesterol nor
desmosterol affected reductase ubiquitination and degradation in SV-589 cells.25
To summarize, we have shown that cholesterol homeostasis in Dhcr24 knockout mice is characterized by strong
activation of LXR-targeted genes and by mechanisms compensating for cholesterol depletion. The very high fecal
excretion of neutral sterols, most probably as a consequence
of the LXR activation and upregulation of Abcg5/8 may
explain why the Dhcr24 knockout mice are unable to accumulate dietary cholesterol. The study also illustrates the
importance of the integrity of the structure of cholesterol
molecule. Introduction of a single double bond in the steroid
side chain is thus compatible with life although disturbances
appear in the steroid homeostasis. In contrast a double bond
in the steroid nucleus gives more serious consequences.24,25
Heverin et al
Sources of Funding
This work was supported by grants from the Swedish Science
Council, The Heart-Lung Foundation, and from the Lion⬘s Club.
Disclosures
A.B., M.S., and E.F. are employed by Quark Biotech Inc, the firm
that initially developed the present animal model.
References
Downloaded from http://atvb.ahajournals.org/ by guest on June 17, 2017
1. Wechsler A, Brafman A, Shafir M, Heverin M, Gottlieb H, Damari G,
Gozlan-Kelner S, Spivak I, Moshkin O, Fridman E, Becker Y, Skaliter R,
Einat P, Faerman A, Bjorkhem I, Feinstein E. Generation of viable
cholesterol-free mice. Science. 2003;302:2087.
2. FitzPatrick DR, Keeling JW, Evans MJ, Kan AE, Bell JE, Porteous ME,
Mills K, Winter RM, Clayton PT. Clinical phenotype of desmosterolosis.
Am J Med Genet. 1998;75:145–152.
3. Andersson HC, Kratz L, Kelley R. Desmosterolosis presenting with
multiple congenital anomalies and profound developmental delay. Am J
Med Genet. 2002;113:315–319.
4. Phillips JE, Rodrigueza WV, Johnson WJ. Basis for rapid efflux of
biosynthetic desmosterol from cells. J Lipid Res. 1998;39:2459 –2470.
5. Brown AJ, Sun L, Feramisco JD, Brown MS, Goldstein JL. Cholesterol
addition to ER membranes alters conformation of SCAP, the SREBP
escort protein that regulates cholesterol metabolism. Mol Cell. 2002;10:
237–245.
6. Yang C, McDonald JG, Patel A, Zhang Y, Umetani M, Xu F, Westover
EJ, Covey DF, Mangelsdorf DJ, Cohen JC, Hobbs HH. Sterol intermediates from cholesterol biosynthetic pathway as liver X receptor ligands.
J Biol Chem. 2006;281:27816 –27826.
7. Ohyama Y, Meaney S, Heverin M, Ekstrom L, Brafman A, Shafir M,
Andersson U, Olin M, Eggertsen G, Diczfalusy U, Feinstein E, Björkhem
I. Studies on the transcriptional regulation of cholesterol 24-hydroxylase
(CYP46A1): marked insensitivity toward different regulatory axes. J Biol
Chem. 2006;281:3810 –3820.
8. Crameri A, Biondi E, Kuehnle K, Lutjohann D, Thelen KM, Perga S,
Dotti CG, Nitsch RM, Ledesma MD, Mohajeri MH. The role of seladin1/DHCR24 in cholesterol biosynthesis, APP processing and Abeta generation in vivo. Embo J. 2006;25:432– 443.
9. Mirza R, Hayasaka S, Takagishi Y, Kambe F, Ohmori S, Maki K,
Yamamoto M, Murakami K, Kaji T, Zadworny D, Murata Y, Seo H.
DHCR24 gene knockout mice demonstrate lethal dermopathy with differentiation and maturation defects in the epidermis. J Invest Dermatol.
2006;126:638 – 647.
10. Rosen H, Reshef A, Maeda N, Lippoldt A, Shpizen S, Triger L, Eggertsen
G, Björkhem I, Leitersdorf E. Markedly reduced bile acid synthesis but
maintained levels of cholesterol and vitamin D metabolites in mice with
disrupted sterol 27-hydroxylase gene. J Biol Chem. 1998;273:
14805–14812.
11. Pikuleva IA, Babiker A, Waterman MR, Björkhem I. Activities of recombinant human cytochrome P450c27 (CYP27) which produce intermediates of alternative bile acid biosynthetic pathways. J Biol Chem. 1998;
273:18153–18160.
Mice With Desmosterolosis
2197
12. Norlin M, Toll A, Bjorkhem I, Wikvall K. 24-hydroxycholesterol is a
substrate for hepatic cholesterol 7alpha-hydroxylase (CYP7A). J Lipid
Res. 2000;41:1629 –1639.
13. Rothblat GH, Burns CH, Conner RL, Landrey JR. Desmosterol as the
major sterol in L-cell mouse fibroblasts grown in sterol-free culture
medium. Science. 1970;169:880 – 882.
14. Fliesler SJ, Richards MJ, Miller CY, Cenedella RJ. Cholesterol synthesis
in the vertebrate retina: effects of U18666A on rat retinal structure,
photoreceptor membrane assembly, and sterol metabolism and composition. Lipids. 2000;35:289 –296.
15. Li-Hawkins J, Gåfvels M, Olin M, Lund EG, Andersson U, Schuster G,
Björkhem I, Russell DW, Eggertsen G. Cholic acid mediates negative
feedback regulation of bile acid synthesis in mice. J Clin Invest. 2002;
110:1191–1200.
16. Pandak WM, Ren S, Marques D, Hall E, Redford K, Mallonee D, Bohdan
P, Heuman D, Gil G, Hylemon P. Transport of cholesterol into mitochondria is rate-limiting for bile acid synthesis via the alternative pathway
in primary rat hepatocytes. J Biol Chem. 2002;277:48158 – 48164.
17. Uppal H, Saini SP, Moschetta A, Mu Y, Zhou J, Gong H, Zhai Y, Ren S,
Michalopoulos GK, Mangelsdorf DJ, Xie W. Activation of LXRs
prevents bile acid toxicity and cholestasis in female mice. Hepatology.
2007;45:422– 432.
18. Schwarz M, Lund EG, Lathe R, Björkhem I, Russell DW. Identification
and characterization of a mouse oxysterol 7alpha-hydroxylase cDNA.
J Biol Chem. 1997;272:23995–24001.
19. Russell DW, Setchell KD. Bile acid synthesis. Biochemistry. 1992;31:
4737– 4749.
20. Horton JD, Goldstein JL, Brown MS. SREBPs: activators of the complete
program of cholesterol and fatty acid synthesis in the liver. J Clin Invest.
2002;109:1125–1131.
21. Kruit JK, Plosch T, Havinga R, Boverhof R, Groot PH, Groen AK,
Kuipers F. Increased fecal neutral sterol loss upon liver X receptor
activation is independent of biliary sterol secretion in mice. Gastroenterology. 2005;128:147–156.
22. Lehmann JM, Kliewer SA, Moore LB, Smith-Oliver TA, Oliver BB, Su
JL, Sundseth SS, Winegar DA, Blanchard DE, Spencer TA, Willson TM.
Activation of the nuclear receptor LXR by oxysterols defines a new
hormone response pathway. J Biol Chem. 1997;272:3137–3140.
23. Fitzky BU, Moebius FF, Asaoka H, Waage-Baudet H, Xu L, Xu G,
Maeda N, Kluckman K, Hiller S, Yu H, Batta AK, Shefer S, Chen T,
Salen G, Sulik K, Simoni RD, Ness GC, Glossmann H, Patel SB, Tint GS.
7-Dehydrocholesterol-dependent proteolysis of HMG-CoA reductase
suppresses sterol biosynthesis in a mouse model of Smith-LemliOpitz/RSH syndrome. J Clin Invest. 2001;108:905–915.
24. Wassif CA, Zhu P, Kratz L, Krakowiak PA, Battaile KP, Weight FF,
Grinberg A, Steiner RD, Nwokoro NA, Kelley RI, Stewart RR, Porter
FD. Biochemical, phenotypic and neurophysiological characterization of
a genetic mouse model of RSH/Smith-Lemli-Opitz syndrome. Hum Mol
Genet. 2001;10:555–564.
25. Song BL, Javitt NB, DeBose-Boyd RA. Insig-mediated degradation of
HMG CoA reductase stimulated by lanosterol, an intermediate in the
synthesis of cholesterol. Cell Metab. 2005;1:179 –189.
Downloaded from http://atvb.ahajournals.org/ by guest on June 17, 2017
Studies on the Cholesterol-Free Mouse: Strong Activation of LXR-Regulated Hepatic
Genes When Replacing Cholesterol With Desmosterol
Maura Heverin, Steve Meaney, Anat Brafman, Millicent Shafir, Maria Olin, Marjan Shafaati,
Sara von Bahr, Lilian Larsson, Anita Lövgren-Sandblom, Ulf Diczfalusy, Paolo Parini, Elena
Feinstein and Ingemar Björkhem
Arterioscler Thromb Vasc Biol. 2007;27:2191-2197; originally published online August 30,
2007;
doi: 10.1161/ATVBAHA.107.149823
Arteriosclerosis, Thrombosis, and Vascular Biology is published by the American Heart Association, 7272
Greenville Avenue, Dallas, TX 75231
Copyright © 2007 American Heart Association, Inc. All rights reserved.
Print ISSN: 1079-5642. Online ISSN: 1524-4636
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://atvb.ahajournals.org/content/27/10/2191
Data Supplement (unedited) at:
http://atvb.ahajournals.org/content/suppl/2007/09/20/ATVBAHA.107.149823.DC1
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Arteriosclerosis, Thrombosis, and Vascular Biology can be obtained via RightsLink, a service of the
Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for
which permission is being requested is located, click Request Permissions in the middle column of the Web
page under Services. Further information about this process is available in the Permissions and Rights
Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Arteriosclerosis, Thrombosis, and Vascular Biology is online
at:
http://atvb.ahajournals.org//subscriptions/
Supplementary materials
Methods
Steroids
Labelled and unlabelled steroids were the same as those used in our previous work
(1,7,11,12,13,18).
Animals
Male mice with a Dhcr24 knockout were generated and maintained as described in detail in
the preceeding paper (1). Following the suckling period, the animals were provided with a
commercial rodent diet (Harlan Teklad Rat/Mouse Diet) ad libitum. In most studies the mice
had an age of 12 weeks, but in one study (Fig. 4) there was a range between 5 and 12 weeks.
Analysis of lipoprotein patterns and their content of cholesterol, triglycerides and
phospholipids
Lipoproteins were separated essentially as previously described (10) using a Superose®-& PC
3.2/30 column (Pharmacia Biothech, Uppsala, Sweden).
Sterol analysis in plasma and liver
Levels of cholesterol, desmosterol and oxysterols in circulation, plasma and liver were
determined by isotope dilution – mass spectrometry as described previously (1,12,13). 27Hydroxydesmosterol was identified (cf results) by mass spectrometry as trimethylsilyl ether
and quantitated with use of deuterium labelled 27-hydroxycholesterol as internal standard. In
this assay the total ion current instead of a specific ion was used and the ratio between the
area of the peak corresponding to the internal standard and the area of the peak
corresponding to 27-hydroxydesmosterol was used for the quantitation.
Analyses of bile acids in bile
Bile acids were analysed by gas chromatography under the conditions described previously
(12,13) The samples were analysed by repetitive scanning for identification of all bile acids
present in the samples. In addition, cholic acid, chenodeoxycholic acid and deoxycholic acid
were quantitated by isotope dilution-mass spectrometry with the use of deuterium labeled
internal standards (12,13). The other bile acids were quantitated from the chromatogram (total
ion current) obtained in analysis of material to which no internal standard had been added
(12). Unhydrolysed bile acids were analysed by electrospray mass spectrometry using a
Quattro micro triple quadropole mass spectrometer. The general conditions for pre-treatment
and analysis of the samples were as described previously (12).
Analysis of bile acids and sterols in faeces
Faeces were collected from 4 individual mice belonging to the two different groups as shown
in Table 1 and Fig. 2. The general methodology used was that described by Grundy, Ahrens
and Miettinen (14) for the analysis of faecal bile acids and the method by Miettinen (15) for
faecal neutral sterols.
Real-Time PCR analyses
The relative amount of target mRNA was quantified by singleplex Real Time RT-PCR
analysis on an ABI PRISMR 7000 Sequence Detection System, using m-Hprt (TaqMan or
SYBR Green) as an endogenous control in parallel assays. The primers used are presented in
Table 1. The specificity of all the PCR products was confirmed by sequencing. In all SYBR
Green assays a final concentration of 200 nM or primers were used. For all SYBR Green
assays a melting curve was obtained for each PCR product after each run to confirm that the
signal corresponded to a unique amplification.
Relative mRNA levels were calculated according to the comparative threshold cycle method
(according to User Bulletin #2: ABI PRISM 7700 Sequence Detection System).
Western Blotting of liver membranes
Antibodies towards ABCG5 was a generous gift by Dr Liqing Yu, Dept of Pathology, Section
of Lipid Sciences, Wake Forest University School of Medicine, Winston-Salem, USA.
Assay of HMG CoA reductase activity in liver microsomes
Preparation of liver microsomes and assay of HMG CoA reductase was performed as
described previously (16) using a modification of the method described by Brown and
Goldstein (17).
Incubations with CYP7A1 and CYP27A1
Recombinant CYP27A1 (18), adrenodoxin, adrenodoxin reductase and NADPH were
incubated under the conditions described previously (18) together with the substrate (20 µg
cholesterol or 20 µg desmosterol). Recombinant CYP7A1 (19) together with NADPH
cytochrome P450 reductase and NADPH were incubated under the conditions described
previously (19) together with the substrate (20µg cholesterol or desmosterol). The extracted
product was analyzed by combined gas chromatography-mass spectrometry as described
previously (18,20).
Statistics
With one exception (Fig. 4) all the results are presented as mean +/- SEM in the tables and
figures. With one exception the normal distribution allowed us to use Student´s t-test directly.
In one case (HMG CoA reductase) this test had to be used after a log-transformation. For
clarity of presentation, the Y-axis in Fig. 4 is depicted with a log scale to allow a better
evaluation of both increase and decrease of the mRNA levels. In view of the fact that the
calculation of the mRNA levels has to include the variance of both the target gene and the
house-keeping gene we present the variations of the mRNA levels shown in Fig. 4 in the form
of a range rather than SEM.
Tables
Table 1. Sequences of Taq-Man primers and probes and SYBR Green primers used for
quantitations by real time PCR
Gene
Cyp 7a1
Hmgcr
Cyp27a1
Cyp8b1
Hmgcs
Shp
Fas
Abca1
Ldlr
Abcg5
Abcg8
Srebp1c
Srebp2
Sqle
Cyp7b1
Hprt
Sequences of forward and reverse primers
5´-CCATGATGCAAAACCTCCAAT-3´
5´-ACCCAGACAGCGCTCTTTGA-3´
5´-CCGGCAACAACAAGATCTGTG-3´
5´-ATGTACAGGATGGCGATGCA-3´
5´-GCCTTGCACAAGGAAGTGACT-3
5´-CGCAGGGTCTCCTTAATCACA-3´
5´-GAACTCAACCAGGCCATGCT-3´
5´-GGCACCCAGACTCGAACCT-3´
5´-CTCTGTCTATGGTTCCCTGGCT-3´
5´-TCCAATCCTCTTCCCTGCC-3´
5´-AAGGGCACGATCCTCTTCAA-3´
5´-CTGTTGCAGGTGTGCGATGT-3´
5´-GGCATCATTGGGCACTCCTT-3´
5´-GCTGCAAGCACAGCCTCTCT-3´
5´-CCCAGAGCAAAAAGCGACTC-3´
5´- GGTCATCATCACTTTGGTCCTTG-3´
5´-GCATCAGCTTGGACAAGGTGT-3´
5´-GGGAACAGCCACCATTGTTG-3´
5´-TGGATCCAACACCTCTATGCTAAA-3´
5´-GGCAGGTTTTCTCGATGAACTG-3´
5´-TGCCCACCTTCCACATGTC-3´
5´-ATGAAGCCGGCAGTAAGGTAGA-3´
5´-GGAGCCATGGATTGCACATT-3´
5´-GGCCCGGGAAGTCACTGT-3´
5´-GCGTTCTGGAGACCATGGA-3´
5´-ACAAAGTTGCTCTGAAAACAAATC-3´
5´-CATGAGTCTCCGGAAAGCAG-3´
5´-TGAAGCACAACACCTTCTATAAACTT-3´
5´-CCTCTTTCCTCCACTCATACACAA-3´
5´-GAACCGATCGAACCTAAATTCCTT-3´
5´-GGTGAAAAGGACCTCTCGAAGTG-3´
5´-ATAGTCAAGGGCATATCCAACAACA-3´
Sequences of probes
5´-TGTCATGAGACCTCCGGGCCTTC-3´
5´-TGTCGCTGCTCAGCACGTCCTCTTC-3´
5´-CCCTTCGGGAAGGTGCCCCAG-3´
5´-ACAGCCTATCCTTGGTGATGCTAGGGCC-3´
5´-TGTCCTGGCACAGTACTCACCTCAGCA-3´
5´-ATGTGCCAGGCCTCCGTGCC-3´
5´-CCATCTGCATAGCCACAGGCAACCTC-3´
5´-AGACTACTCTGTCTCTCAGACAACACTTGACCAAG-3´
5´-CACTCCTTGATGGGCTCATCCGACC-3´
5´-CCAGACTTTGTTGGATTTGAAATTCCAGACAA-3´
Table 2. Faecal excretion of plant sterols from Dhcr24-/- and Dhcr24+/- micea
__________________________________________________________________________
Methyl Ethyl- Ethyl
Sitosterol Sitostanol Total
copro- copro- coproplant
stanol
stanol, stanone
sterolsc
Campesterolb
___________________________________________________________________________
µg/24h
Dhcr24+/-
52±20
681±75 257±23
1638±119 623±51
3526±284
Dhcr24-/-
<5
473±38 198±11
1350±85 514±33
2860±165
a
Measurements in 6 heterozygots and 6 knockouts 3 months of age. The figures given are
mean ±SEM.
b
Mixture of ethylcoprostanol and campesterol that was quantitated together.
c
The quantitation includes two plant sterols (C28 and C29) that was never identified and
present in minor amounts.
Table 3. Faecal excretion of C27 neutral sterols from Dhcr42+/- and Dhcr24-/- micea
___________________________________________________________________________
Coprostanol
Cholesterol+ cholestanol
Desmosterol
µg/24h
Dhcr24+/-
72±17
692±48
Dhcr24-/-
<5
91±24
<5
3740±274
___________________________________________________________________________
a
Measurements in 6 heterozygotes and 6 knockouts 3 months of age. Figures are given as
mean±SEM.
Legends to Figures
Fig. 5. Mass spectrum of trimethylsilyl derivative of 27-hydroxydesmosterol present in
plasma of a male Dhcr24-/- mouse (A). An identical spectrum was obtained after incubation
of desmosterol with CYP27 (B). Characteristic fragments are seen at m/z 544 (M), m/z 454
(M-90). The fragments at m/z 373 (loss of steroid side-chain), m/z 371, m/z 253 m/z 213 and
m/z 129 are seen also in the mass spectrum of trimethylsilyl ether of desmosterol.
Fig. 6. Western blotting of membrane preparations from liver of one Dhcr24-/-, two +/- and
one +/+ mouse using antibodies towards Abcg5 and Actin.
Abundance
A
143
10000
9000
8000
7000
6000
5000
343
4000
3000
253
2000
173
283
213
1000
371
415
309
0
150
200
250
300
350
400
454
450
544
500
550
600
m/z
Abundance
143
B
18000
16000
14000
12000
10000
343
8000
6000
253
4000
173
283
213
2000
371
454
100
150
200
250
300
544
415
309
0
350
m/z
400
450
500
550
600
+/+
-/-
+/-
-/-
+/kDa
Abcg5
75
Actin
37