at SciVerse ScienceDirect

Biochimie 95 (2013) 448e454

Contents lists available

Biochimie

journal homepage: www.elsevier .com/locate/b iochi

Review

Five decades with oxysterols

Ingemar Björkhem*

Department of Laboratory Medicine, Division of Clinical Chemistry, Karolinska Institutet, Karolinska University Hospital Huddinge, C1 741,141 86 Huddinge, Sweden

a r t i c l e i n f o

Article history:Received 30 December 2011Accepted 17 February 2012Available online 1 March 2012

Keywords:OxysterolsCytochrome P-450Bile acid synthesisCerebrotendinous xanthomatosisCholesterol homeostasisIsotope dilution e mass spectrometry

* Tel.: þ46 8 58581235; fax: þ46 8 58581260.E-mail address: ingemar.bjorkhem@karolinska.se.

0300-9084/$ e see front matter � 2012 Elsevier Masdoi:10.1016/j.biochi.2012.02.029

a b s t r a c t

I have been involved in research on oxysterols since 1963 and this review is intended to cover some ofthe most important aspects of this work.

The first project dealed with 7a-hydroxy-4-cholesten-3-one. My successful synthesis of this steroidwith high specific radioactivity allowed a demonstration that it is a bile acid precursor. The mechanism ofconversion of 7a-hydroxycholesterol into 7a-hydroxy-4-cholesten-3-one was investigated and Iconcluded that only one enzyme is required and that no isomerase is involved. Accumulation of7a-hydroxy-4-cholesten-3-one in patients with lack of sterol 27-hydroxylase (Cerebrotendinous xantho-matosis was shown to be an important pathogenetic factor. This disease is characterized by cholestanol-containing xanthomas in tendons and brain and we could show that most of this cholestanol is formedfrom 7a-hydroxy-4-cholesten-3-one. We also showed that 7a-hydroxy-4-cholesten-3-one passes thebloodebrain barrier.

In contrast to cholesterol itself, side-chain oxidized oxysterols have a high capacity to pass lipophilicmembranes. We demonstrated conversion of cholesterol into 27-hydroxycholesterol to be a significantmechanism for elimination of cholesterol from macrophages. We also showed that conversion ofcholesterol into 24S-hydroxycholesterol is important for elimination of cholesterol from the brain.

Side-chain oxidized oxysterols have a high capacity to affect critical genes in cholesterol turnoverin vitro. Most of the published in vitro experiments with oxysteroids are highly unphysiological,however. Mouse models studied in my laboratory with high or low levels of 27-hydroxycholesterol havelittle or no disturbancies in cholesterol homeostasis. 24S-hydroxycholesterol is an efficient ligand to LXRand suggested to be important for cholesterol homeostasis in the brain. We recently developed a mousemodel with markedly increased levels of this oxysterol in circulation and brain. This overexpression hadhowever only a very modest effect on cholesterol turnover.

We concluded that oxysterols are not the master regulators of cholesterol homeostasis in vivo sug-gested previously.

� 2012 Elsevier Masson SAS. All rights reserved.

1. Introduction

My first experience with oxysteroids started in 1963 when I wasaccepted as “research assistant” without salary at the departmentof Medical Chemistry at Karolinska Institutet. My tutor was asso-ciate professor Henry Danielsson, a previous pupil of Prof. SuneBergström who later became a Nobel laureate.

This was before the era of molecular biology and the limitationsin biochemical science were organic and analytical chemistry. Thegreat scientific victories by Sune Bergström and his pupils were inparticular based on organic chemistry: synthesis of new unlabelledand labelled steroids and fatty acids and their analyses by mass

son SAS. All rights reserved.

spectrometry. As a consequence of this the first project I becameinvolved in was synthesis of a steroid that was believed but notproven to be an intermediate in the conversion of cholesterol intobile acids, namely 7a-hydroxy-4-cholesten-3-one.

This specific oxysteroid has followed me during my wholescientific career and I have published a number of studies con-cerned with different aspects on it: its chemical synthesis, themechanism of its enzymatic formation, its metabolism, its role inbile acid biosynthesis, its role in pathogenetic mechanisms, its roleas a marker for bile acid synthesis, quantitative methods for itsanalysis. In this lecture I would like to review some selective partsof this work that range from the early sixties up to now. I will thenreview some specific aspects of oxysteroids: enzymatic and non-enzymatic formation, how they can be used as markers fordifferent pathological states, role of oxysterols as regulators ofcholesterol homeostasis, methodology aspects.

I. Björkhem / Biochimie 95 (2013) 448e454 449

2. 7a-Hydroxy-4-cholesten-3-one

This steroids had previously been synthesized in 1961 by mytutor, Henry Danielsson, by a synthetic route from cholesterylbenzoate and with a yield of only 0.2%. A small amount of tritiumlabelled material had been prepared and injected in a bile fistularat. In contrast to the expectations, only small amounts of thenormal bile acids cholic acid and chenodeoxycholic acid wereformed from the labelled steroid. Henry Danielsson believed,however, that the low yield may be due to the fact that the injectedsteroid had a very low specific radioactivity and that unphysio-logically high levels had been injected. My first project was there-fore to try to synthesize 7a-hydroxy-4-cholesten-3-one with veryhigh specific radioactivity that would allow injections of only traceamounts of the steroid to bile fistula rats.

Because of the fact that I had to work on this project in parallelwith my medical studies, it took me more than a year to find outa suitable synthetic procedure to get materials with high specificradioactivity. The crucial and critical step in the new syntheticprocedure was the introduction of the delta 4 double bond byselenium dioxide. My new method of synthesis used chenodeox-ycholic acid as starting material and the product could be obtainedby only three separate steps with a total yield of about 5%. (Fig. 1).Since it was easy to label chenodeoxycholic acid with tritium itbecame possible to synthesize labelled 7a-hydroxy-4-cholesten-3-one with high specific radioactivity. When trace amounts oftritium labelled 7a-hydroxy-4-cholesten-3-one were injected intoa bile fistula rat, there was a high yield of radioactive cholic acidand chenodeoxycholic acid. When the tritium labelled oxysterolwas diluted with unlabelled oxysterol and injected into a bilefistula rat, a number of other radioactive bile acids were formedthat could not be identified. This study was published in 1965 andbecame the first paper in my thesis that was completed 4 yearslater [1].

Thus we were convinced that 7a-hydroxy-4-cholesten-3-one isan intermediate in the conversion of cholesterol into bile acids. Wecould also show that this oxysterol is formed from 7a-hydrox-ycholesterol in microsomal preparations from rat liver. One of thechallenges was to show if the double bond isomerizes from 5 to 4position first or if the oxidation of the 3b-hydroxy group is the firststep (Fig. 2). I could show convincingly that oxidation of the 3b-hydroxyl group is the first and rate-limiting step. The best evidencefor this was my demonstration of an isotope effect in the oxidationof the 3b-hydroxyl group when the hydrogen atom at 3a wasreplaced with a deuterium or tritium atom [2]. Such an isotope

Fig. 1. Synthesis of 7a-hydr

effect is only possible if the carbonehydrogen bond is broken ina rate-limiting step.

The problem was then to decide if two different enzymes areinvolved in the reaction: a 3b-hydroxysteroid dehydrogenasecatalyzing oxidation of the 3b-hydroxy group and an isomerasecatalyzing the conversion of the delta 5 double bond into the delta 4double bond. Different attempts to isolate the postulated inter-mediate 7a-hydroxy-5-cholesten-3-one in the conversion failed,however. I also failed to synthesize the postulated intermediate inthe conversion, and all attempts resulted in loss of the very labile7a-hydroxyl group. I draw the conclusion that it was most likelythat the dehydrogenase is able to catalyze both the oxidation of the3b-hydroxy group and the isomerization. Thus the enzymaticallyformed intermediate may be protected from elimination of the 7a-hydroxyl group by association to a suitable group on the enzyme. Icould also show that the enzymatic isomerization involveda transfer of hydrogen from the 4b- to the 6b-position.

There were still scientists, however, who believed that theremust be a specific isomerase involved in the conversion of 7a-hydroxycholesterol into 7a-hydroxy-4-cholesten-3-one.

More than 20 years later, I became involved in a study demon-strating beyond all doubts that only one enzyme is involved in boththe oxidation of the 3b-hydroxy group and the isomerization of thedelta 5 double bond [3]. Expression cloning was used to isolatecDNAs encoding the microsomal 3b-hydroxy-delta-5-C27-steroidoxidoreductase. When this cDNA was transfected into culturedcells it encoded an enzyme that was active against a number of 7a-hydroxylated 3b-hydroxy-delta-5-steroids. This was convincingevidence for my suggestion that only one enzyme is required tocatalyze both the oxidation and the isomerization.

The gene coding for the enzyme was shown to be mutated ina patient with urinary excretion of 3b-hydroxy-delta-5-bile acids.About 10 years earlier, we had shown that fibroblasts isolated fromthis patient had a reduced capacity to convert 7a-hydrox-ycholeserol into 7a-hydroxy-4-cholesten-3-one [4].

I would like to review the different metabolic fates of 7a-hydroxy-4-cholesten-3-one. As shown in Fig. 3 this compound canbe 27-hydroxylated, 12a-hydroxylated, converted into the 5b- and5a-saturated steroid and finally dehydrated.

In collaboration with a Norwegian group I had showed that thebasal biochemical defect in patients with the rare inborn diseaseCerebrotendinous xanthomatosis (CTX) is a defect sterol 27-hydroxylase [5]. This disease is characterized by development ofcholestanol and cholesterol-containing xanthomas in the tendonsand in the brain. It was difficult to understand the mechanism

oxy-4-cholesten-3-one.

Fig. 2. Alternative mechanisms for conversion of 7a-hydroxycholesterol into 7a-hydroxy-4-cholesten-3-one.

I. Björkhem / Biochimie 95 (2013) 448e454450

behind the accumulation of cholestanol. We suspected that cho-lestanol could be formed by a 7a-hydroxylated intermediate andtherefore I synthesized 7a-3H-labelled cholesterol. Injection ofa mixture of this cholesterol with 14C-cholesterol in rats and ina healthy human resulted in loss of 20e30% of the tritium in rela-tion to 14C. When this experiment was performed in a patient withCTX, however, more than 70% of the tritium was lost [6]. This isconsistent with a 7a-hydroxylated steroid as intermediate duringthe conversion of cholesterol into cholestanol. There was a markedaccumulation of 7a-hydroxy-4-cholesten-3-one in the liver as wellas in the circulation of CTX patients [7,8], demonstrating that thissteroid could be an intermediate. Furthermore we could demon-strate a conversion of 7a-hydroxy-4-cholesten-3-one into choles-tanol in vitro with cholesta-4,6-dien-3-one as intermediate. Thelevel of the latter steroid was also found to be markedly increasedin the circulation of patients with CTX.

In a recent workwe have demonstrated that there is a flux of 7a-hydroxy-4-cholesten-3-one across the bloodebrain barrier. Micewith a knockout of the sterol 27-hydroxylase were shown toaccumulate cholestanol in the brain in parallel with the accumu-lation of 7a-hydroxy-4-cholesten-3-one [9,10].

With respect to other metabolic pathways for 7a-hydroxy-4-cholesten-3-one this steroid was found to be an efficientsubstrate for the sterol 12a-hydroxylase and may be the mostimportant physiological substrate. The gene coding for the sterol12a-hydroxylase was cloned and characterized by my group incollaboration with a Japanese group [11].

The delta 4 double bond in 7a-hydroxy-4-cholesten-3-one issaturated by a soluble 5b-reductase. This reaction was found to

Fig. 3. Metabolic fate of 7a-hydroxy-4-cholesten-3-one.

involve a trans-diaxial addition of hydrogen and involve a directtransfer of hydrogen from the A position of NADPH to the5b-position of the steroid [12,13]. My group also cloned the geneencoding the human 5b-reductase [14].

The idea came up that 7a-hydroxylated intermediates in bileacid synthesis in the liver could “leak” from the liver and reflect bileacid synthesis. The most logic candidate would be 7a-hydrox-ycholesterol and we could show that the level of this steroid in thecirculation reflects the activity of the cholesterol 7a-hydroxylase inhuman liver [15]. Because of the fact that 7a-hydroxycholesterolalso is formed non-enzymatically, this assay was not optimal.Together with Magnus Axelson we could show that 7a-hydroxy-4-cholesten-3-one is a better alternative as a plasma marker for bileacid synthesis in human liver [16]. This steroid is now generallyused as a marker, both in clinical studies and in studies withexperimental animals. In our laboratory we use a LC-MS methodwith use of a deuterium labelled internal standard [17].

3. Methodological aspects

Sune Bergström had early understood the great potential ofmass spectrometry and this technique was the most importantanalytical tool in our department during the sixties and seventieswhen I was working there. One of Bergström�s previous pupils, JanSjövall, built up a very strong mass spectrometric unit at ourdepartment and also developed new potent separation techniquesthat could be used in our work on steroid metabolism. I becameearly interested in the possibility to use mass spectrometry forquantitative work, in particular in the form of isotope dilution e

mass spectrometry. I was fascinated by the potential of this tech-nique to develop highly accurate methods.

In 1974 I moved from the Department of Medical Chemistry tothe Department of Clinical Chemistry and started my career inClinical Chemistry by setting up highly accurate isotope dilutionmethods for cholesterol, cortisol, testosterone, progesterone,vitamin D, 25-hydroxyvitamin D, 1,25-dihydroxy vitamin D,anabolic steroids and a great number of other compounds Thesereference methods were used to evaluate routine methods andselect the most accurate of them for the analysis [18]. In thisconnection I also used isotope dilution for assay of oxysterols, inparticular 7a-hydroxycholesterol. One of the most importantapplications of this assay was determination of the cholesterol 7a-hydroxylase activity in human liver biopsies [19]. The activity ofthis enzyme in human liver is too low to be assayed by use ofexogenously added labelled cholesterol. In the assay I developedwemeasured the production of 7a-hydroxycholesterol from theendogenous cholesterol present in the microsomes. Since we hadthe possibility to get liver biopsies from patients in connectionwithelective cholecystectomy, and these patients could be pretreatedwith cholesterol and different drugs, we could do some ratherunique studies. These studies helped to define the very importantrole of the cholesterol 7a-hydroxylase for cholesterol homeostasisin humans.

Our first attempt to set up a method for assay of the majoroxysterols in human plasma was made in 1990 together with myPhD student at that time, Olof Breuer [20]. This method was laterfurther expanded and optimized by Ulf Diczfalusy [21]. Of partic-ular importance was his optimization of the hydrolysis procedure.The levels of the oxysterols formed both enzymatically and byautoxidation, such as 7-oxocholesterol, 7b-hydroxycholesterol and5a-cholestane-3b,5a,6b-triol were reported to be lower with themethod developed by Diczfalusy than previously. It has beena trend over the years that the reported levels of these steroids havedecreased, probably as a consequence of more efficient measurestaken to avoid autoxidation during isolation and workup.

Fig. 4. Elimination of cholesterol from macrophages by the sterol 27-hydroxylasemechanism e an alternative to the HDL-dependent reverse cholesterol transport.

I. Björkhem / Biochimie 95 (2013) 448e454 451

Amajor methodological problem in the assay of oxysterols is thefact that some of these steroids may be formed both by autoxida-tion of endogenous cholesterol during isolation and workup and byenzymatic activity. The autoxidation during workup and isolationmay be reduced by addition of antioxidants and handling of thesamples in nitrogen atmosphere. Another important factor is ifEDTA is used for trapping of the metal ions required for the Fentonreaction. After correction for possible autoxidation under isolationand workup, Schroepfer and his group arrived to the conclusionthat little or none of 7b- and 7-oxocholesterol, 5a-cholestan-3b,5a,6b-triol and cholesterol 5,6-epoxide found in plasma areformed in vivo [22]. Enzymatic oxygenation of cholesterol is cata-lyzed by cytochrome P-450 species or a dioxygnase. In both type ofreactions, one of the two oxygen atomes are transferred frommolecular oxygen to the substrate cholesterol. This means thatproduction of an oxysterol in a human exposed to oxygen 18 in theinhalation gas will occur with introduction of 18O in the oysterolmolecule. This will not happen during isolation or workup. Bythis mean we could demonstrate in rats that 7a-, 7b-, and 7-oxocholesterols as well as. 24-, 25-, and 27-hydroxy cholesterolsare all formed enzymatically [23]. Since there was no 18O present inthe epoxides or in 5a-cholestane-3b,5a,6b-triol, it seems likely thatmost of these steroids is formed non-enzymatically. The 18O2-inhalation techniquewas also used for a study on the kinetics in theformation of oxysterols in humans. In a very expensive experimenta human volunteer inhaled 24% 18O2 during 30 min [24]. There wasa rapid incorporation of 18O in 7a-hydroxycholesterol and 27-hydroxycholesterol. It was shown that the half-life of the twooxysterols in the circulation was less than 1 h for both steroids.

I would like to mention papers by two other groups with majormethodological flaws, methods that we have critisized. The groupby Diestel et al. reported levels of 7-oxocholesterol in cerebro-spinal fluid to be more than 3 orders of magnitude higher thanfound by Valerio Leoni in his thesis work [25]. It was not possiblefor Leoni et al. to report the methodological flaw in a letter to theEditor of J. Exp. Med. in which the article by Diestel et al. waspublished. The other example is a paper by Farez et al. in NatureImmunology who reported presence of very high levels of 15-oxygenated steroids in human serum [26]. These levels were re-ported to be increased in patients with multiple sclerosis ina progressive state. We could show convincingly that if present,there are trace amounts only of these steroids in human circula-tion, regardless of presence of multiple sclerosis or not [27]. Wetried during more than a year to publish a letter on this in NatureImmunology without success. After the failure of two othergroups, including the one by William Griffiths, to confirm thefindings by Farez et al., however, it became possible to publish thecritical letter [28].

4. Side-chain oxidized oxysterols as transport forms ofcholesterol

HDL-mediated reversed cholesterol transport is a most impor-tant mechanism in connection with cholesterol homeostasis,allowing a flux of cholesterol from extrahepatic tissues to the liver.Oxidative mechanisms involving formation of side-chain oxidizedoxysterols can be regarded as alternatives to this mechanism. Incontrast to cholesterol, steroid side-chain oxidized cholesterolspecies are able to cross model systems of lipophilic membranes atrates orders of magnitude higher than the rate of transport ofcholesterol [29,30]. Reversed HDL-dependent cholesterol transportrequires specific receptors and transporters to cross the cellmembranes, factors that are under strict metabolic control. Theflux of side-chain oxidized cholesterol species across the cellmembranes appears to be less dependent upon such factors.

Conversion of cholesterol into 27-OHC or 24S-OHC is thusa strategy allowing elimination of excess cholesterol from the cells(Fig. 4). Since the sterol 27-hydroxylase is an evolutionary oldenzyme, it is possible that oxidative mechanisms may have beenmore important for elimination of cholesterol at an earlier stage ofevolution [31].

In catheterization experiments on healthy volunteers we couldshow that about 4% of the normal biosynthesis of bile acids involvesextrahepatic 27-hydroxylation of cholesterol and subsequenttransport to the liver [32]. Extrahepatic 27-hydroxylation ofcholesterol seems to be of particular importance in macrophages,and we found high levels of the enzyme in cultured human lungalveolar macrophages [33]. When such cells were exposed to aninhibitor of the sterol 27-hydroxylase, a significant accumulation ofintracellular cholesterol could be demonstrated [32]. The fact thata lack of the sterol 27-hydroxylase causes accumulation of choles-terol in tendon xanthomas may illustrate the importance of thismechanism under in vivo conditions. It seems likely that the sterol27-hydroxylase mechanism is of particular importance underconditionswith low levels of HDL, a condition likely to be present intendons.

Bymeasuring the arteriovenous difference of different oxysterolsbetween the internal jugular vein and an artery in healthy humanvolunteers, we could demonstrate a significant flux of the oxysterol24S-hydroxycholesterol from the brain into the circulation [34].Using an 18O2-inhalation technique, we could demonstrate thatthe flux of 24S-OHC from the rat brain into the circulation corre-sponds to about 2/3 of the rate of cholesterol synthesis in the sameorgan [35]. This was later confirmed in mice with a knockout of thecholesterol 24S-hydroxylase [35]. The brain cholesterol 24S-hydroxylase, denoted CYP46, is thus a critical factor for cholesterolhomeostasis in the brain. This enzyme was later cloned and chaca-terized byDavid Russell andhis group [36]. Under normal conditionsthis enzyme is located to neuronal cells only. I viewof its importancefor cholesterol homeostasis in the brain, onewould expect CYP46 tobe sensitive to a number of regulatory factors. Surprisingly, however,we found it to behighly resistant tomost regulatoryaxes both in vivoand invitro. Oxidative stresswas oneof the very few factors affectingthe expression of the enzyme [37].

Our catheterization experiments demonstrated that there isa net uptake of 27-OHC by the brain, which is of the same magni-tude as the flux of 24OHC from the brain into the circulation [38]. In

I. Björkhem / Biochimie 95 (2013) 448e454452

spite of the similar magnitude of the two fluxes, the levels of 27-OHC in the brain are very low. The low levels are due to anextensive metabolism and we have shown that the most importantmetabolite is 7a-hydroxy-3-oxo-4-cholestenoic acid, and that thereis a net flux of this acid from the brain into the circulation [39].Fig. 5 summarizes the different fluxes of oxysterols across thebloodebrain barrier and the role of the liver for the final conversionof the different oxysterols to bile acids.

5. Side-chain oxidized oxysterols and neurodegeneration

We have discussed the possibility that the balance between theflux of 24OHC from the brain and the flux of 27-OHC into the brainis of importance for the neurodegenerative processes [40]. Underin vitro conditions 24S-OHC has a protective effect on the genera-tion of beta-amyloid in cultured neuroblastoma cells, whereas 27-OHC seems to antagonize this effect [41]. Hypercholesterolaemiais a risk factor for Alzheimer’s disease in spite of the fact thatcholesterol itself does not pass the bloodebrain barrier. Thepossibility must be considered that 27-OHC is mediating the effectof cholesterol on the brain. Feeding mice with cholesterol affectsa number of factors in the brain that are of importance inconnection with neurodegeneration, e.g. the “memory protein”activity regulated cytoskeleton-associated protein (Arc), and therenin angiotensinogen system (RAS) [42,43]. We found that similareffects on the production of these factors could be obtained byexposure of isolated neuronal cells or slices to 27-OHC [42e44].Studies are ongoing in our laboratory to evaluate the importance ofthese effects in vivo with use of mice with a disrupted sterol 27-hydroxylase and no significant levels of 27-OHC in the circulation.

6. Oxysterols as biomarkers for different metabolic pathwaysand pathological conditions

As discussed above, plasma level of 7a-hydroxy-4-cholesten-3-one can be used as a marker for bile acid synthesis in vivo. Levels of27-OHC are markedly reduced in patients with Cerebrotendinousxanthomatosis and markedly increased in patients with hereditaryspastic paresis of the subgroup SPG5 and a mutation in the CYP7B1gene [45]. In view of the fact that 7-oxocholesterol and 7b-

Fig. 5. Cross-talk of oxysterols across the bloodebrain barrier.

hydroxycholesterol are formed in connection with lipid perox-idation, one would expect oxidative stress to be associated withelevated levels of these oxysterols in the circulation. In accordancewith this we found significantly increased levels of 7b-hydrox-ycholesterol in a population of patients with a fast progression ofcarotid atherosclerosis and the level of this oxysterol was amongthe best predictors of those tested for the increase in carotid wallthickness [46]. The low levels of 7b-hydroxycholesterol and the riskfor falsely high levels due to autoxidation of cholesterol duringcollection, extraction and workup limits however the usefulness ofthis oxysterol as a biomarker. Another condition with increasedlipid peroxidation is NiemannePick disease type C. Recently it wasreported by Ory et al. that these patients have high levels of7-oxocholesterol in the circulation [47]. We have confirmed theirresults and the high plasma levels are in accord with our previousfinding that a patient with this disease has a high urinary excretionof a metabolite of 7-oxocholesterol [48].

Ulf Diczfalusy and his group has demonstrated that 4b-hydroxycholesterol may be used as a biomarker for the activity ofthe cytochrome CYP3A4, responsible for metabolism of about 50%of the commonly used drugs [49].

In view of the fact that 24S-OHC is formed almost exclusivelyin neuronal cells, we have investigated the possibility thatplasma level of this oxysterol may be used as a marker for thenumber and activity of neuronal cells. The levels of 24S-OHC areelevated up to the age of 15e20 but are relatively constant atolder ages [50]. The reason for the high levels in infants is mostprobably due to reduced capacity for hepatic elimination [50]. Inaccordance with the contention that destruction of neuronal cellsmay be reflected in plasma levels of 24S-OHC, we found reducedlevels of this oxysterols in patients with advanced Alzheimer’sdisease as well as in patients with other severe neurologicaldisorders [51,53]. The reduction in the levels of 24S-OHC washowever only of the magnitude 15e20%, which is too small to beused in individual cases. The reason for the relatively modestreduction is most probably due to an abnormal expression ofcholesterol 24S-hydroxylase in glial cells in the brain of patientswith Alzheimer’s disease [53]. This abnormal induction may inpart compensate for the reduced production of the oxysterol bythe neuronal cells.

Surprisingly, neurological diseases were found to be associatedwith increased levels of 24S-OHC in cerebrospinal fluid in spite ofthe reduced levels in the circulation [52]. The increased levels mayreflect the decomposition of the neuronal cells. The diagnosticsensitivity of this cerebrospinal test appeared to be similar to that ofthe established markers for neurodegenerative diseases, tau- andphospho-tau protein and b-amyloid [54]. In addition to 24S-OHC,the levels of 27-OHC are also increased in cerebrospinal fluid ofpatients with neurological diseases [52]. The latter increase isparticularly marked in connection with a reduced integrity of thebloodebrain barrier [55]. The ratio between 27-OHC and 24S-OHCin cerebrospinal fluid was thus found to be increased in patientswith meningitis, polyneuropathy and hemmorage [55].

7. Oxysterols as regulators of cholesterol homeostasis

Oxysterols, in particular those with the extra hydroxyl group inthe steroid side-chain, have a high capacity to affect critical genes incholesterol turnover under in vitro conditions (for reviews, see ref.[56,57]). Because these metabolites are much more potent thancholesterol in this respect, they have been suggested to mediatea great number of cholesterol-induced effects. Due to the unphy-siological levels of free oxysterol used inmost of these experiments,it is not possible to evaluate the role of the oxysterol as a regulatorunder in vivo conditions.

I. Björkhem / Biochimie 95 (2013) 448e454 453

In order to study if side-chain oxidation is of importance for thecapacity of cholesterol to suppress its own synthesis we preparedcholesterol labelled with 2H in the 24- and 27- positions. We couldshow that the hydroxylation of these cholesterol species in vitrowas associated with marked isotope effects. The capacity of thesedeuterated cholesterol species to suppress cholesterol synthesis inmouse feeding experiments was however the same as that ofunlabelled cholesterol [58]. This suggests that there are no 24- or27-hydroxylations involved in the cholesterol-induced suppressionof cholesterol synthesis in mouse liver. Using a similar approach itwas shown that oxidation at C-3, C-5, C-6 or C-7 is not involved inthe cholesterol-induced suppression of cholesterol synthesis inmouse liver [59]. The delta 5 double bond and the stereochemistryof the 3b-hydroxyl group were however found to be structuralrequirement for the suppression.

Transgenic mouse models with highly varying levels of side-chain oxidized oxysterols due to overexpression or lack of thecritical hydroxylase or the major oxysterol-metabolizing enzyme,present very modest changes only in over-all cholesterol homeo-stasis. Side-chain oxidized oxysterols are ligands and activators ofthe nuclear receptor LXR, and reduced or increased levels of theoxysterol would be expected to affect the target genes of LXR. Oneof the most efficient ligands to LXR under in vitro conditions is 24S-OHC. Recently we produced mice with 3e5 fold increasedproduction of 24S-OHC. The increased production had howeververy little effect on the LXR target genes in brain and liver [60],suggesting that 24S-OHC is not important as a regulator of thissystem under in vivo conditions. 27-OH is also a ligand to LXR andin collaborationwith an israelian groupwe have characterizedmicewith knockout as well as overexpression of the sterol 27-hydroxylase [61,62]. The markedly increased and markedlydecreased levels of 24S-OHC in the two mouse models were notassociated with significant changes in cholesterol homeostasis.

Altogether our results support the contention that side-chainoxidized oxysterols are not important as master regulators ofcholesterol homeostasis [56,57]. It should be mentioned, however,that in a recent very elegant experiment Chen et al. demonstratedthat cholesterol feeding of mice with a triple knockout of the 24S-,25- and 27-hydroxylases failed to affect some of the LXR targetgenes that are affected in normal mice [63]. It was suggested thatpart of the cholesterol-induced effects on the LXR target genesweremediated by hydroxylation(s) in the steroid side-chain. Thecomplicated mouse model and the very high dose of cholesterolused in these experiments makes it difficult to evaluate the regu-latory role of the above hydroxylases under normal conditions.

8. General remarks, conclusions and perspectives

It has been fascinating to follow the field of oxysterol researchduring almost 5 decades.

Accurate methods are now available for measurement of oxy-sterols. It is still a significant problem, however, to avoid artefactualformation of oxysterols from endogenous cholesterol duringcollection and workup of the biological samples. The roles of oxy-sterols as intermediates in bile acid synthesis and as transportforms of cholesterol are now well established. Levels of oxysterolsmay be used for estimation of the rate of bile acid synthesis and fordiagnosis of specific diseases such as NiemannePick disease, Cer-ebrotendinous xanthomatosis, and a subform of Hereditary SpasticParesis. Oxysterols can also be used to evaluate progression ofneurodegenerative diseases like Huntington’s disease and Alz-heimer’s disease. It is evident that oxysterols are not obligatory inthe regulation of cholesterol homeostasis but they may play a rolein the fine tuning of this regulation. The enzyme systems producingthe side-chain oxidized oxysterol 24S-hydroxycholesterol and 27-

hydroxycholesterol may be new targets for drugs designed toreduce neurodegeneration. It is impossible, however, to predictwhich aspects in oxysterol research that will dominate in thefuture.

Acknowledgements

The author is grateful to all his PhD students and post docs.Among these the following have made the most significantcontributions to the specific material reviewed here: Olof Breuer,Lionel Bretillon, Ann Båvner, Magnus Hansson, Maura Heverin,Valerio Leoni, Erik Lund, Dieter Lutjohann, Steve Meaney, MarjanShafaati. I am also grateful to my technicians, Manfred Held, AnitaLövgren-Sandblom and Maria Olin. I am also grateful to all mypresent and previous collaborators: Ulla Andersson, Bo Angelin,Magnus Axelsson, Mari Buchmann, Angel Cedazo-Minguez, UlfDiczfalusy, Gösta Eggertssen, Curt Einarsson, Scott Grundy, AndersKallner, Eran Leitersdorf, Kuyichiro Okuda, Yoshikiko Ohyama, UtePanzenboeck, Jan Pedersen, Irina Pikuleva, David Russell, RebeccaSchule, Sverre Skrede, Jan Sjövall. Swedish Science Council hassupported all the work reviewed here. During the last decadesignificant support has also been obtained from the Swedish Heart-Lung Foundation, the Swedish Brain Power and Pfizer.

References

[1] I. Björkhem, H. Danielsson, H.C. Issidorides, A. Kallner, On the synthesis andmetabolism of cholest-4-en-7a-ol-3-one, Acta Chem. Scand. 19 (1965)2151e2154.

[2] I. Björkhem, On the mechanism of the enzymatic conversion of cholest-5-ene-3b,7a-diol into 7a-hydroxycholest-4-en-3-one, Eur. J. Biochem. 8 (1969)337e344.

[3] M. Schwartz, A.C. Wright, D.L. David, H. Nazer, I. Björkhem, D.W. Russell, Thebile acid synthetic gene 3b-hydroxy-delta-5-C27-steroid oxidoreductase inmutated in progressive intrahepatic cholestasis, J. Clin. Invest. 106 (2000)1175e1184.

[4] M.S. Buchmann, E.A. Kvittingen, H. Nazer, T. Gunasekaran, P.T. Clayton,J. Sjövall, I. Björkhem, Lack of 3b-hydroxy-delta-5-steroid dehydrogenase/isomerase in fibroblasts from a child with urinary excretion of 3b-hydroxy-delta-5-bile acids. A new inborn error of metabolism, J. Clin. Invest. 86 (1990)2034e2037.

[5] H. Oftebro, I. Björkhem, S. Skrede, A. Schreiner, J.I. Pedersen, Cerebrotendinousxanthomatosis. A defect in mitochondrial 26-hydroxylation required fornormal synthesis of cholic acid, J. Clin. Invest. 65 (1980) 1418e1430.

[6] S. Skrede, I. Björkhem, M.S. Buchmann, H. Hopen, O. Fausa, A novel pathwayfor biosynthesis of cholestanol from 7a-hydroxylated C27-steroids as inter-mediates and its importance for the accumulation of cholestanol in Cere-brotendinous xanthomatosis, J. Clin. Invest. 75 (1985) 448e455.

[7] I. Björkhem, H. Oftebro, S. Skrede, J.I. Pederesen, Assay of intermediates in bileacid biosynthesis using isotope dilution e mass spectrometry: hepatic levelsin the normal state and in Cerebrotendinous xanthomatosis, J. Lipid. Res. 22(1981) 191e200.

[8] I. Björkhem, S. Skrede, M.S. Buchmann, C. Eas, S. Grundy, Accumulation of7a-hydroxy-4-cholesten-3-one and cholesta-4,6-dien-3-one in patients withCerebrotendinous xanthomatosis: effect of treatment with chenodeoxycholicacid, Gastroenterology 7 (1987) 266e271.

[9] U. Panzenboeck, U. Andersson, W. Sattler, S. Meaney, I. Björkhem, On themechanism of cerebral accumulation of cholestanol in patients with Cere-brotendinous xanthomatosis, J. Lipid Res. 48 (2007) 1167e1174.

[10] A. Båvner, M. Shafaati, M. Hansson, M. Olin, S. Shpitzen, V. Meiner,E. Leitersdorf, I. Björkhem, On the mechanism of accumulation of cholestanolin the brain of mice with a disruption of sterol 27-hydroxylase, J. Lipid Res. 51(2010) 2722e2730.

[11] G. Eggertsen, M. Olin, U. Andersson, H. Ishida, S. Kubota, U. Hellman, K. Okuda,I. Björkhem, Molecular cloning and expression of rabbit sterol 12a-hydroxy-lase, J. Biol. Chem. 271 (1996) 278e285.

[12] O. Berseus, I. Björkhem, Enzymatic conversion of a delta-4-3-ketosteroid intoa 3b-hydroxy-5b-steroid: mechanism and stereochemistry of hydrogentransfer from NADPH, Eur. J. Biochem. 2 (1967) 503e507.

[13] I. Björkhem, Stereochemistry of the enzymatic conversion of a delta-4-3-oxosteroid into a 3-oxo-5b-steroid, Eur. J. Biochem. 7 (1969) 413e417.

[14] K.H. Kondo, M.H. Kai, Y. Setoguchi, G. Eggertsen, P. Sjöblom, T. Setoguchi,K. Okuda, I. Björkhem, Cloning and expression of cDNA of human delta-4-3-oxosteroid 5beta-reductase and substrate specificity of the expressedenzyme, Eur. J. Biochem. 219 (1994) 357e363.

[15] I. Björkhem, E. Reihnér, B. Angelin, S. Ewerth, J.-E. Åkerlund, K. Einarsson, Onthe possible use of the serum level of 7a-hydroxycholesterol as a marker for

I. Björkhem / Biochimie 95 (2013) 448e454454

increased activity of the cholesterol 7a-hydroxylase in humans, J. Lipid Res. 28(1987) 889e894.

[16] M. Axelson, I. Björkhem, E. Reihnér, K. Einarsson, The plasma level of7a-hydroxy-4-cholesten-3-one reflects the activity of hepatic cholesterol7a-hydroxylase in man, FEBS Lett. 284 (1991) 216e218.

[17] A. Lövgren-Sandblom, M. Heverin, H. Larsson, E. Lundström, J. Wahren,U. Diczfalusy, I. Björkhem, Novel LC-MS/MS method for assay of 7a-hydroxy-4-cholesten-3-one in human plasma. Evidence for a significant extrahepaticmetabolism, J. Chromatogr. B. 856 (2007) 15e19.

[18] I. Björkhem, R. Blomstrand, O. Lantto, L. Svensson, G. Öhman, Towardsabsolute methods in clinical chemistry: application of mass frag-mentography to high accuracy analyses, Clin. Chem. 11 (1976)1789e1801.

[19] K. Einarsson, B. Angelin, S. Ewerth, K. Nilsell, I. Björkhem, Bile acid synthesis inman: assay of microsomal cholesterol 7a-hydroxylase activity by isotopedilution mass spectrometry, J. Lipid Res. 27 (1986) 82e88.

[20] O. Breuer, I. Björkhem, Simultaneous quantification of several cholesterol andmonohydroxylated products by isotope-dilution mass spectrometry, Steroids55 (1990) 185e192.

[21] S. Dzeletovic, O. Breuer, E. Lund, U. Diczfalusy, Determination of cholesteroloxidation products in human plasma by isotope dilution mass spectrometry,Anal. Biochem. 225 (1995) 73e80.

[22] K. Kudo, G.T. Emmons, E.W. Casserly, D.P. Via, L.C. Smith, S.J. Pyrek,G.J. Schroepfer, Inhibitors of sterol synthesis. Chromatography of acetatederivatives of oxygenated sterols, J. Lipid Res. 30 (1989) 1097e1111.

[23] O. Breuer, I. Björkhem, Use of an 18O2 inhalation technique and mass iso-topomer distribution analysis to study oxygenation of cholesterol in rats.Evidence for in vivo formation of 7-oxo-, 7b-hydroxy-, 24-hydroxy and 25-hydroxycholeserol, J. Biol. Chem. 270 (1995) 20278e20284.

[24] S. Meaney, M. Hassan, A. Sakinis, D. Lutjohann, K. von Bergmann,Å. Wennmalm, I. Björkhem, Evidence that the major oxysterols in humancirculation originate from distinct pools of cholesterol: a stable isotope study,J. Lipid Res. 42 (2001) 70e78.

[25] V. Leoni, D. Lutjohan, T. Masterman, Levels of 7-oxocholesterol in cerebro-spinal fluid are more than one thousand times lower than reported inmultiple sclerosis, J. Lipid Res. 46 (2005) 191e195.

[26] M. Farez, F.J. Quintana, R. Gandhi, M. Izquierdo, M. Lucas, H.L. Weiner, Toll-likereceptor 2 and poly (ADP-ribose) polymerase 1 promote central nervoussystem neuroinflammation in progressive EAE, Nat. Immun. 10 (2009)958e964.

[27] I. Björkhem, A. Lövgren-Sandblom, F. Piehl, M. Khademi, H. Pettersson,V. Leoni, T. Olsson, U. Diczfalusy, High levels of 15-oxygenated steroids incirculation of patients with multiple sclerosis: fact of fiction? J. Lipid Res. 52(2011) 170e174.

[28] I. Björkhem, U. Diczfalusy, T. Olsson, D.W. Russell, J.G. McDonald, Y. Wang,W.J. Griffiths, Detecting oxysterols in the human circulation, Nat. Immun. 12(2011) 577.

[29] Y. Lange, J. Ye, F. Strebel, Movement of 25-hydroxycholesterol from theplasma membrane to the rough endoplasmic reticulum in cultured hepatomacells, J. Lipid Res. 36 (1995) 1092e1097.

[30] S. Meaney, K. Bodin, U. Diczfalusy, I. Björkhem, On the rate of translocationin vitro and kinetics in vivo of the major oxysterols in human circulation:critical importance of the position of the oxygen function, J. Lipid Res. 44(2002) 2130e2135.

[31] U. Diczfalusy, E. Lund, D. Lütjohann, I. Björkhem, Novel pathways for elimi-nation of cholesterol by extrahepatic formation of side-chain oxidized oxy-sterols, Scand. J. Clin. Lab. Invest. 56 (1996) 9e17.

[32] E. Lund, O. Andersson, J. Zhang, A. Babiker, G. Ahlborg, U. Diczfalusy,K. Einarsson, J. Sjövall, I. Björkhem, Importance of a novel oxidative mecha-nism for elimination of intracellular cholesterol in humans, Arteriosclerosis 16(1996) 208e212.

[33] I. Björkhem, O. Andersson, U. Diczfalusy, B. Sevastik, R.-J. Xiu, C. Duan, E. Lund,Atherosclerosis and sterol 27-hydroxylase: evidence for a role of this enzymein elimination of cholesterol from human macrophages, Proc. Natl. Acad. Sci.91 (1994) 8592e8596.

[34] D. Lütjohann, O. Breuer, G. Ahlborg, I. Nennesmo, Å. Sidén, U. Diczfalusy,I. Björkhem, Cholesterol homeostasis in human brain: evidence for an age-dependent flux of 24S-hydroxycholesterol from the brain into the circula-tion, Proc. Natl. Acad. Sci. (USA) 93 (1996) 9799e9804.

[35] I. Björkhem, D. Lütjohann, O. Breuer, A. Sakinis, Å. Wennmalm, Importance ofa novel oxidative mechanism for elimination of brain cholesterol, J. Biol.Chem. 272 (1997) 30178e30184.

[36] E.G. Lund, J.M. Guileyardo, D.W. Russell, CDNA cloning of the cholesterol 24-hydroxylase, a mediator of cholesterol homeostasis in the brain, Proc. Natl.Acad. Sci. USA 96 (1999) 7238e7243.

[37] Y. Ohyama, S. Meaney, M. Heverin, L. Ekström, A. Brafman, U. Andersson,M. Olin, G. Eggertsen, U. Diczfalusy, E. Feinstein, I. Björkhem, Studies on thetranspcriptional regulation of cholesterol 24-hydroxylase (CYP46A1): markedinsensitivity towards different regulatory axes, J. Biol. Chem. 281 (2006)3810e3820.

[38] M. Heverin, S. Meaney, D. Lutjohann, U. Diczfalusy, J. Wahren, I. Björkhem,Crossing the barrier: net flux of 27-hydroxycholesterol into the human brainand possible consequences for cerebral cholesterol homeostasis, J. Lipid Res.46 (2005) 1047e1052.

[39] S. Meaney, M. Heverin, U. Panzenboeck, L. Ekström, Axelson, U. Andersson,U. Diczfalusy, I. Pikuleva, J. Wahren, I. Björkhem, Novel route for the elimi-nation of brain oxysterols across the blood-brain barrier: conversion into7a-hydroxy-3-oxo-4-cholestenoic acid, J. Lipid Res. 52 (2007) 450e454.

[40] I. Björkhem, V. Leoni, A. Cedazo-Minguez, S. Meaney, Oxysterols and neuro-degeneration, Mol. Aspects Med. 30 (2009) 171e179.

[41] D. Famer, S. Meaney, M. Mousavi, I. Björkhem, M. Crisby, Regulation of a-and b-secretase activity by oxysterols: cerebrosterol stimulates processing of APP viathe a-secretase pathway, Biochem. Biophys. Res. Commun. 359 (2007) 46e50.

[42] L. Mateos, S. Akterin, F.J. Gil-Bea, S. Spulber, A. Rahman, I. Björkhem,M. Schultzberg, A. Flores-Morales, A. Cedazo-Minguez, Activity-regulatedcytoskeleton-associated protein in rodent brain is down-regulated by high fatdiet invivo andby27-hydroxycholesterol invitro, BrainPathol. 19 (2009)69e80.

[43] L. Mateos, M.A. Ismail, F.J. Gil-Bea, V. Leoni, B. Winblad, I. Björkhem,A. Cedazo-Minguez, Upregulation of brain renin angiotensin system by 27-hydroxycholesterol in Alzheimer�s disease, J.Alzheimer�s Dis. 23 (2011) 1e11.

[44] L. Mateos, I. Muhammad-Al-Mustafa, F.J. Gil-Bea, R. Schule, L. Schöls,R. Folkesson, I. Björkhem, A. Cedazo-Minguez, Side-chain oxidized oxysterolsregulate the brain renineangiotensin system through a liver X receptordepending mechanism, J. Biol. Chem. 286 (2011) 25574e25585.

[45] R. Schule, T. Siddique, H.X. Deng, Y. Yang, S. Donkervoort, M. Hansson,R.E. Madrid, N. Siddique, L. Schöls, I. Björkhem, Marked accumulation of 27-hydroxycholesterol in patients with hereditary spastic paresis (SPG5) anda mutation in the gene coding for the oxysterol 7 alpha hydroxylase (CYP7B1).Is this accumulation pathogenetic factor? J. Lipid Res. 51 (2010) 819e823.

[46] J.T. Salonen, K. Nyyssönen, R. Salonen, E. Porkkala-Sarataho, T.P. Tuomainen,U. Diczfalusy, I. Björkhem, Lipid oxidation and progression of carotidatherosclerosis, Circulation 18 (1997) 840e845.

[47] D. Ory, F. Porter, D. Scherrer, M. Lanier, S. Langmade, V. Molugu, S. Gale,D. Olzeski, Cholesterol oxidation products next term are sensitive and specificblood-based biomarkers for Niemann-Pick disease, type C, Mol. Genet. Metab.99 (2010) S28.

[48] G. Alvelius, O. Hjalmarsson, W.J. Griffiths, I. Björkhem, J. Sjövall, Identificationof unusual 7-oxygenated bile acid sulfates in a patient with Niemann-Pickdisease type C, J. Lipid. Res. 42 (2001) 1571e1577.

[49] U. Diczfalusy, H. Nylen, L. Bertilsson, 4b-hydroxycholesterol, an endogenousmarker ofCYP3A4/5 activity in humans, Br. J. Pharmacol. 71 (2011) 183e189.

[50] L. Brétillon, D. Lütjohann, L. Ståhle, T. Widhe, L. Bindl, G. Eggertsen,U. Diczfalusy, I. Björkhem, Plasma levels of 24S-hydroxycholesterol reflect thebalance between cerebral production and hepatic metabolism and areinversely related to body surface, J. Lipid Res. 41 (2000) 840e845.

[51] L. Bretillon, Å. Sidén, L. Wahlund, D. Lütjohann, M. Crisby, J. Hillert, C.-G. Groth,Diczfalusy, I. Björkhem, Plasma levels of 24S-hydroxycholesterol in neurolog-ical patients, Neurosci. Lett. 293 (2000) 87e90.

[52] V. Leoni, T. Masterman, F. Mousavi, B. Wretlind, L-.O. Wahlund, U. Diczfalusy,J. Hillert, I. Björkhem, Diagnostic use of cerebral and extracerebral oxysterols,Clin. Chem. Lab.Med 2 (2004) 186e191.

[53] N. Bogdanovic, L. Bretillon, E.G. Lund, U. Diczfalusy, L. Lannfelt, B. Winblad,D.W. Russell, I. Björkhem, Induction of cholesterol catabolic enzymecholesterol24S-hydroxylase (CYP46) in Alzheimer brain, Neurosci. Lett. 314(2001) 45e48.

[54] V. Leoni, M. Shaafati, A. Solomon, M. Kivipelto, I. Björkhem, L.-O. Wahlund, Arethe CSF levels of 24S-hydroxycholesterol a suitable biomarker for mildcognitive impairment? Neurosci. Lett. 397 (2006) 83e87.

[55] V. Leoni, T. Masterman, P. Patel, S. Meaney, U. Diczfalusy, I. Björkhem, Side-chain oxidized oxysterols in cerebrospinal fluid and integrity of the blood-brain and bloodecerebrospinal fluid barriers, J. Lipid Res. 44 (2003) 793e799.

[56] I. Björkhem, Do oxysterols control cholesterol homeostasis? J. Clin. Invest. 110(2002) 725e730.

[57] I. Björkhem, Are side-chain oxidized oxysterols regulators also in vivo? J. LipidRes. 50 (2009) 213e218.

[58] E. Lund, O. Breuer, I. Björkhem, Evidence that 24- and 27-hydroxylation arenot involved in the cholesterol-induced down-regulation of HMG CoAreductase in mouse liver, J. Biol. Chem. 267 (1992) 25092e25097.

[59] E. Lund, I. Björkhem, Downregulation of hepatic HMG CoA reductase in miceby dietary cholesterol. Importance of the delta 5 double bond and evidencethat oxidation at C-3, C-5, C-6, or C-7 is not involved, Biochemistry 33 (1994)291e297.

[60] M. Shafaati, M. Olin, A. Båvner, H. Pettersson, B. Rozell, S. Meaney, P. Parini,I. Björkhem, Enhanced production of the endogenous LXR ligand 24S-hydroxycholesterolis not sufficient to drive LXR-dependent gene expressionin vivo, J. Int. Med. 270 (2011) 377e387.

[61] H. Rosen, A. Reshef, N. Maeda, A. Lippoldt, S. Shpizen, L. Triger, G. Eggertsen,I. Björkhem, E. Leitersdorf, Markedly reduced bile acid synthesis but main-tained levels of cholesterol and vitamin D metabolites in mice with a dis-rupted sterol 27-hydroxylase gene, J. Biol. Chem. 273 (1998) 14805e14812.

[62] K. Meir, D. Kitsberg, I. Alkalay, F. Shkedy, H. Rosen, S. Shpizen, L. Ben-Avi,B.Y. Staels, B.Y. Fievet, V. Meiner, I. Björkhem, E. Leitersdorf, Human sterol 27-hydroxylase (CYP27) overexpression transgenic mouse model. Evidenceagainst 27-hydroxycholesterol as acritical regulator of cholesterol homeo-stasis, J. Biol. Chem. 277 (2002) 34036e34041.

[63] W. Chen, D. Chen, D.L. Head, D. Mangelsdorf, D.-W. Russell, Enzymaticreduction of oxysterols impairs LXR signalling in cultured cells and in mice,Cell Metab. 5 (2007) 73e79.