WO1995018533A1

WO1995018533A1 - Methods and compounds affecting adipocyte differentiation and obesity

Info

Publication number: WO1995018533A1
Application number: PCT/US1994/014946
Authority: WO
Inventors: Mitchell A. Lazar; Ajay Chawla
Original assignee: The Trustees Of The University Of Pennsylvania
Priority date: 1994-01-04
Filing date: 1994-12-27
Publication date: 1995-07-13

Abstract

Methods of identifying activators and antagonists of peroxisome proliferator activated receptors and activators of retinoic acid receptor η are featured. In addition, the invention features a substantially pure preparation of a naturally occurring peroxisome proliferator activated receptor activator. Also featured are methods of treating obesity in an animal.

Description

METHODS AND COMPOUNDS AFFECTING
ADIPOCYTE DIFFERENTIATION AND OBESITY
Background of the Invention
The field of the invention is obesity and factors which affect obesity.

This invention was made in part with government support under National Institutes of Health Grant Number
DK45586.

Obesity is a serious and prevalent human health problem known to contribute to a variety of illnesses including heart disease and diabetes. Obesity may be genetic and may also arise in individuals as a result of a lipoprotein metabolism disorder. Obesity in humans is known to occur when an individual contains a greater than normal number of adipocytes, which adipocytes are fully differentiated and are capable of storing large quantities of fat.

3T3-L1 cells are cells which have been selected from
Swiss 3T3 fibroblasts for their ability to differentiate into adipocytes following exposure to fetal calf serum (FCS), dexamethasone (Dex), isobutylmethylxanthine (IBMX), and insulin (Green et al., 1974, Cell 3,127-133; Green et al., 1975, Cell 5,19-27. Conversion of preadipocytes to adipocytes involves coordinated expression of a variety of transcription factors prior to morphological change of the cells and concomitant induction of additional cellular structural proteins (Spiegelman, et al., 1993, J. Biol. Chem. 268,6823-6826).

Factors which are induced during adipocyte differentiation include C/EBPa, ss, and 6 (Lin, F.-Y. et al., 1992, Genes and
Dev. 6,533-544; Umek et al., 1991, Science 289,288-292), as well as Jun and Fos (Rauscher et al., 1988, Cell, 52,471-480;
Distel et al., 1987, Cell 49,835-844). In addition, c-myc gene expression is repressed in the mature adipocyte, and overexpression of c-myc inhibits differentiation (Freytag et al., 1992, Science 256,379-382).

Retinoic acid (RA) and peroxisome proliferators are lipophilic substances which act through related nuclear receptors to directly regulate gene expression (Evans, 1988,
Science 240,889-895; Green et al., 1988, Trends in Genet. 4, 309-315; Issemann et al., 1990, Nature 347,645-650; Leid et al., 1992, Trends Biochem Sci. 17,427-433). Both RA receptors (RARs) and peroxisome proliferator activated receptors (PPARs) heterodimerize with retinoid X receptor (RXR) in vitro (Leid et al., 1992, Trends in Biochem Sci. 17,427-433) and, in the case of PPAR, regulation of gene expression is greatest when the activators of both receptors are present (Keller et al., 1993,
Proc. Natl. Acad. Sci. USA 90,2160-2164; Kliewer et al., 1992,
Nature 358,771-774; Gearing et al., 1993, Proc. Natl. Acad.

Sci. USA 90,1440-1444; Bardot et al., 1993, Biochem. Biophys.

Res. Commun. 192,37-45).

RA inhibits 3T3-L1 cell differentiation from preadipocytes to adipocytes (Sato et al., 1980, Biochem.

Biophys. Res. Commun. 95,1839-1945; Kuri-Harcuch, 1982,
Differentiation 23,164-169). Inhibition of differentiation is an unusual effect of RA, which more commonly promotes differentiation of cells (Spornet al., 1983, Cancer Res. 43, 3034-3040). In contrast to the effects of RA, peroxisome proliferators such as clofibrate have been shown to potentiate adipose conversion of 3T3-L1 cells in the presence of Dex, FCS, and insulin (Brandes et al., 1977, Life Sci. 40,935-941).

PPAR is of particular interest in adipocytes because it regulates genes involved in lipid metabolism (Tugwood et al., 1992, EMBO J. 11,433-439; Dreyer et al., 1992, Cell 68, 879-887) and is activated not only by peroxisome proliferators, but also by fatty acids (Keller et al., 1993, Proc. Natl. Acad.

Sci. USA 90,2160-2164; Gottlicher et al., 1992, Proc. Natl.

Acad. Sci. USA 89,4653-4657).

Summary of the Invention
The invention features a method of screening candidate activators of PPAR involving the steps of first, inducing a morphological change in a culture of preadipocytes by incubating the preadipocytes in delipidated serum, second, adding to the culture a test compound and third, determining whether the addition of the test compound causes reversal of the morphological change, wherein reversal of the morphological change is an indication that the test compound is a candidate activator of PPAR.

The invention also features a method of screening candidate antagonists of PPAR activation involving the steps of first, inducing a morphological change in a culture of preadipocytes by incubating the preadipocytes in delipidated serum, second, adding to the culture an activator of PPAR in the presence or absence of a test compound, and third, determining whether the presence of the test compound inhibits reversal of the morphological change, wherein inhibition of reversal of the morphological change is an indication that the test compound is a candidate antagonist of PPAR activation.

In addition, the invention features a method of inhibiting differentiation of preadipocytes to adipocytes by adding to a culture of preadipocytes an activator of RARy.

The invention further features a method of treating obesity in an animal, preferably a human, wherein there is administered to the animal a compound in a dose of 0.1 yg to 50 g/kg/day, which compound is either an activator or an antagonist of PPAR, or an activator of RARe, wherein the compound is suspended in a pharmaceutically acceptable carrier.

The type of compound to be administered will depend upon such factors as the age of the animal to be treated.

Also featured in the invention is a substantially pure preparation of a naturally occurring PPAR activator, which activator is normally present in bovine serum. By substantially pure is meant sufficiently separated from the components with which it is normally associated without interfering with its function as an activator. By activator is meant a ligand for PPAR, a compound which can be converted in the cell to a ligand for PPAR, or a compound which is not a ligand for PPAR but which serves to induce the generation of a ligand for PPAR.

Other features and advantages of the invention will be apparent from the following detailed description and from the claims.

Detailed Description
There will now be described the experimental details for the practice of the invention. The invention provides methods of screening candidate compounds which are capable of activating PPARs which are useful in treatment of lipoprotein metabolism disorders, and may eventually be useful for the treatment of obesity. The methods of the invention are the result of the discovery of the presence of a natural activator of PPAR in extracts of serum, the discovery that activators of
PPAR are capable of inducing fat cell differentiation in the absence of Dex, IBMX, fetal calf serum and insulin, and the discovery of the role of RARy1 in inhibition of differentiation of preadipocytes to adipocytes.

The examples given below relate to but are not limited to adipocyte differentiation in mouse cells. As will be evident from the data presented, the methods of the invention are also applicable to human cells and therefore to eventual treatment of humans having lipoprotein metabolism disorders and humans experiencing obesity because of the extensive sequence homology between the mouse and the human RARyl (Zhu et al., 1993, J. Biol. Chem. 268 26817).

Cell Culture Methods. 3T3-L1 cells (American Type
Culture Collection) were cultured in growth medium containing
Dulbecco's modified Eagle's medium (DMEM) and 10k bovine calf serum (Hyclone). The medium was changed every 2 days. Bovine calf serum was delipidated using a modification of the method of Goodman (Goodman, 1958, J. Am. Chem. Soc. 80,3887-3892).

Briefly, serum was extracted two or three times with an equal volume of n-heptane by vigorous stirring at 4 C for 20 hours.

The aqueous phase was separated by centrifugation at 2,000 g for 2 hours. Cells were initially cultured in normal growth medium for 2-3 days (40-50% confluence) and then switched to medium containing 10% delipidated serum.

The method used to induce differentiation of 3T3-L1 cells and the inhibition of this process by RA is described in
Chawla et al. 1993, J. Biol. Chem. 268,16265-16269. RA was added to the cells in ethanol at a final concentration of 1 x 10-5 M. In experiments wherein PPAR activators were used, 3T3
L1 cells were cultured in normal growth medium until they reached confluence. The medium was then replaced with medium supplemented with varying concentrations of clofibrate (Sigma
Chemical Co.), WY-14,643 (pirinixic acid; Wyeth-Ayerst), or 5, 8,11,14-eicosatetraynoic acid (ETYA; Sigma Chemical Co.), a polyunsaturated fatty acid which also activates PPAR. Each of these compounds was dissolved in ethanol prior to addition to the cells. Control cells were treated with the same volume of ethanol without addition of the compound.

Northern Hybridization Analysis. Northern analyses were performed as described in Chawla et al., 1993, J. Biol.

Chem. 268,16265-16269. The probes which were used included cDNA probes for aP2 (Spiegelman et al., 1983, J. Biol. Chem.

258,10083-10089), C/EBPa (Landschulz et al., 1988, Genes and
Dev. 2,786-800), PPARa (Issemann et al., 1990, Nature 347, 645-650), RXRa, and y (Mangelsdorf et al., 1992, Genes and
Dev. 6,329-344), Nuc-1 (Schmidt et al., 1992, Mol. Endocrinol.

6,1634-1641) c-myc (Stone et al., 1987, 7, 1697-1709), and ssactin (American Type Culture Collection) labeled with sing random hexamers. The RARyl-specific probe (Zelent et al., 1989, Nature, 339,714) was a 478 bp EcoRl/EcoRV fragment contained within the 5'terminus of the gene. Bands representing RNA on autoradiograms were quantitated using a
Molecular Dynamics ImageQuant Densitometer.

Method of Determining Cell Viability. Cells were cultured in 60 mm2 dishes in normal medium until they reached 40-50% confluency at which time the medium was replaced with delipidated medium containing either 1 x 10-5 M RA or an equal volume of ethanol without the addition of RA. Approximately every 24 hours, floating and adherent cells were pooled and cell viability was assessed by trypan blue exclusion. Similar results were observed when cells were exposed to RA for the duration of the experiment or when this compound was added only during the initial 48 hours of the experiment. To assess the
DNA replication capacity of the cells, 3H-thymidine incorporation by the cells was measured. The cells were cultured in 96 well plates and 3H-thymidine (5 yCi/well) was added for 1 hour.

The cells were then harvested and the amount of radioactivity which was incorporated was measured by scintillation counting. To assess the integrity of the cellular DNA, genomic DNA was isolated from adherent and floating cells using a method which selected for fragmented DNA (Debbas et al., 1993, Genes and Dev. 7,546-554). This DNA was then analyzed by electrophoresis through 1.2% agarose gels.

Characteristics of preadipocytes grown in delipidated serum. Preadipocytes were cultured in delipidated serum as described above. Under these conditions, preadipocytes continued to proliferate after control cells had reached confluence such that the number of cells increased approximately 3-fold after 4 days. Incorporation of 3Hthymidine by cells incubated in delipidated serum was approximately 5-fold greater than that of control cells. In the latter case, the amount of 3H-thymidine incorporated decreased by 88% after confluence was achieved. The increased cell number and continued proliferation of the cells was paralleled by the continued expression of c-myc, which expression was dramatically reduced in control preadipocytes shortly after they reached confluence (Freytag et al., 1992,
Science 256,379-382).

An increase in cell density was apparent on inspection of the culture dishes, and the shape of the cells changed in that they became elongated. These morphological changes could be prevented by the addition of a chloroform extract of the material removed from the delipidated serum.

Activators of PPAR prevent and reverse the effects of delipidated serum-containing medium on preadipocytes. The ability of PPAR activators to substitute for the extracted substance (s), i. e., those removed by delipidation, was tested.

Remarkably, treatment with WY-14,643, a peroxisome proliferator and activator of PPAR, prevented and reversed the changes induced by the delipidated serum. Continued exposure of these cells to WY-14,643 ultimately induced adipose conversion.

Clofibrate as well as ETYA also caused adipose conversion of 3T3-L1 cells. The concentrations required to induce greater than 90% of the cells to differentiate after 7 days were as follows: 3mM clofibrate, 450 yM WY-14,643, and 50 yM ETYA. The relative potencies of these compounds parallel those required for activation of PPAR (Keller et al., 1993,
Proc. Natl. Acad. Sci. USA 90,2160-2164). Adipose conversion induced by PPAR activators was completely inhibited by RA as was adipose conversion which was induced under standard differentiating conditions.

Induction of PPAR and RXR during adipose conversion.

The molecular events involved in adipose conversion by PPAR activators were compared by Northern hybridization analysis with those which occurred during differentiation of the cells under standard conditions. Cells treated with WY-14,643, clofibrate, and ETYA expressed similar levels of the adipocytespecific genes aP2 and C/EBPa, which genes appeared with similar kinetics in these cells as in cells which were differentiated under standard conditions. In addition, PPARa expression increased approximately 10-fold during adipose conversion induced under standard differentiating conditions or by treatment with WY-14,643. The size of the PPAR mRNA was approximately 7.0 kb, larger than that reported by Issemann et al., 1990, Nature 347,645-650), but was similar in size to that found in other mouse tissues, and in rat cells (Gebel et al., 1992, FEBS Letters 309,37-40).

Another PPAR gene, Nuc-1 (Schmidt et al., 1992, Mol. Endocrinol. 6,1634-1641), was also found to be induced by WY-14,643. In addition, WY-14,643 treated cells or cells induced to differentiate under standard conditions also expressed the a, ss, and Y subtypes of RXR.

RXRa-specific mRNA increased within 4 hours after initiation of the standard differentiation protocol; this is the most rapid induction that has been demonstrated for any RXR, and therefore is one of the earliest molecular changes which occurs during adipose conversion. RXRy was also induced, although at a considerably later time than RXRa. The size of the RXRz mRNA was approximately 6 kb, larger than that previously reported (Mangelsdorf et al., 1992, Genes and Dev. 6,329-344) and raises the possibility that an unusual RXRy isoform is expressed in fat tissues.

RA induces apoptosis of preadipocytes cultured in delipidated medium. 3T3-L1 preadipocytes express the RARcf,, and Y subtypes (Kamei et al., 1993, Biochem J. 293,807-812).

Although RA inhibits adipose conversion, it does not cause any appreciable phenotypic change in cells cultured under normal conditions. The effects of RA on cells cultured in delipidated serum were examined because this medium is also depleted of retinoids. Cells incubated in the presence of delipidated serum-containing medium and treated with RA developed cytoplasmic blebbing, vacuolization, and condensation within 34 days of exposure to RA. These changes are characteristic of cells undergoing apoptosis (Wyllie et al., 1980, Int. Rev.

Cytol. 68,251-306). Quantitatively, viable cell number was decreased by greater than 90% 5 days after exposure to RA. In addition, chromosomal DNA obtained from cells undergoing differentiation when analyzed by gel electrophoresis displayed a characteristic periodic (approximately 200 bp) fragmentation pattern frequently seen in apoptotic cells.

RA inhibits adipose conversion during a narrow window of expression of RARyl. The EDso for inhibition of adipocyte conversion by all-trans RA was approximately 5 x 10-'M and 9cis RA, a ligand for RAR as well as RXR, was equally potent.

However, an RXR-specific ligand (Lehmann et al., 1992, Science 258,1944-1946) did not induce adipocyte differentiation thus specifically implicating RARs and not RXRs in this process.

To investigate the mechanism by which RA inhibits adipose conversion, RAR gene expression during adipose conversion was examined. RNA was extracted daily from cells incubated under standard differentiating conditions and
Northern analysis was performed using probes specific for each of the three RAR genes. 3T3-L1 preadipocytes expressed RARt, ss, and y-specific mRNA at Day 0. However, as differentiation progressed, while the levels of RARa and RAR/3 mRNA did not change appreciably during adipocyte conversion, a near complete reduction in the expression of RARy was observed, which reduction occurred prior to the expression of adipocytespecific genes such as aP2 and C/EBPa.

There are two major RARY isoforms, yl and y2 (Kastner et al., 1990, Proc. Natl. Acad. Sci. USA 87,2700-2704). The decline of RARy mRNA levels in differentiating preadipocytes could be almost completely attributed to a decline in the levels of RARyl mRNA in that, levels of RARy1 mRNA were reduced to less than 20% of control levels by 16 hours following induction of differentiation, whereas levels of RARy2, although lower than those of RARz1, did not change significantly during the process of preadipocyte differentiation.

RARz1 is highly regulated during mouse embryogenesis (Kastner et al., 1990, Proc. Natl. Acad. Sci. USA 87,27002704) and is capable of inhibiting transcriptional activation by other RARs (Nagpal et al., 1992, Cell 70,1007-1019; Husmann et al., 1991, Mol. Cell. Biol. 11,4097-4103). To assess the functional consequences of the decline in RARy1 mRNA levels, RA was added to preadipocytes at various times after exposure of the cells to standard differentiating medium. At 7 days posttreatment, the level of expression of C/EPBa was assessed as a measure of differentiation. When RA was added as late as 24 hours following induction of differentiation it was only mildly effective in preventing adipocyte conversion. Significantly, if RA was added at times later than 24 hours, the cells were almost completely refractory to the effects of this compound.

Concomitantly, the levels of RARy1 in cells cultured under standard differentiating conditions were measured. The kinetics of inhibition of differentiation by RA and of downregulation of RARyl transcription are closely correlated, suggesting that RARy1 mediates the inhibitory effects of RA.

Constitutive expression of RARyl inhibits differentiation of preadipocytes to adipocytes. The RARyl and the RARe2 gene were independently cloned into a plasmid encoding the neomycin resistance gene such that expression of either RARy1 or RARE2 was linked to expression of the neomycin resistance gene and was driven by the mouse moloney sarcoma virus long terminal repeat promoter/enhancer sequences (Miller et al., 1989, Biotechniques, 7,980-990 ; Adam et al., 1991, J.

Virol. 65,4985-4990). Cells were transfected with either
RARy-containing plasmid or with a control plasmid encoding the neomycin resistance gene whose transcription was also regulated by the mouse moloney sarcoma virus promoter. Stable transfectants were selected following incubation in the presence of G418. In transient expression assays, the plasmids were shown to express functional products.

Stable transfectants containing either the control or
RARyl encoding plasmid were then cultured in normal medium, differentiating medium or medium containing RA and the level of expression of RARyl-specific mRNA was assessed in each cell type. The endogenous RARyl mRNA was down-regulated during adipocyte conversion induced by DM, which down-regulation is inhibited by treatment with RA. Expression of RARE1 in two independently isolated RARy1-transfected clones was examined and the results were identical. Two forms of RARyl-specific mRNA were observed in these cells, the endogenous form and the polycistronic form comprising the RARy1-specific mRNA covalently linked to the neomycin resistance-specific mRNA.

The expression of either mRNA was not down-regulated when the cells were subjected to differentiating conditions regardless of whether or not RA was present in the cultures.

Next, transfected clones were examined for their ability to express C/EBPa as a measure of their ability to differentiate when treated with differentiating medium.

Inhibition of this process by RA was also assessed. In clones transfected with control plasmid C/EBPa-specific mRNA was expressed following adipocyte conversion induced by differentiating medium, which expression was inhibited by RA.

However, in two independent clones stably transfected with
RARyl expression of C/EBPa was markedly absent providing evidence for the crucial role of RARyl in preventing adipocyte conversion. Moreover, it is clear that it is RARz1 and not
RARz2 which plays a role in adipocyte conversion because in three independent clones stably transfected with RARy2 the pattern of C/EBPa expression was identical to that in control plasmid transfected cells. C/EBPa expression was absent in cells incubated in normal medium; C/EBPa expression was evident in cells incubated in differentiating medium, which expression was inhibited in RA treated cells. Similar results were obtained when aP2 instead of C/EBPa was used as a marker of differentiation in RARyl transfected cells.

It is not necessary to characterize differentiation of the cells using molecular markers. Rather, the differentiation state of the cells can be determined by simple morphological analysis. Cells which were stably transfected with RARyl were morphologically identical to normal undifferentiated cells and remained so with time in culture.

In contrast, cells which were stably transfected with RARy2 proceeded to differentiate with time in culture and acquired the morphology characteristic of fully differentiated adipocytes after several days.

Moreover, when RARy1 and RAR2 transfected cells were treated with either clofibrate or WY-14,643, only RARy2 transfected cells proceeded to differentiate to an adipocyte phenotype. RARyl transfected cells were refractory to the effects of these agents.

The results indicate that down-regulation of RARy1 is essential for normal adipose conversion. Moreover, overexpression of RARy2 cannot substitute for RARy1-induced inhibition of differentiation indicating that the A domain of
RARyl is the functional part of the molecule with respect to down-regulation during differentiation. Nagpal et al. (1992,
Cell 70 1007-1019) have shown that the A domain of the RARy isoforms function differently in transactivation, and Husmann et al. (1991, Mol. Cell. Biol. 11 4097-4103,1991) have reported that RAR-yl is capable of antagonizing the effects of other RARs including RARyl with regard to their ability to transactivate selected genes including the RARES gene RARE.

These effects of RARs require specific recognition of target gene DNA sequences, which recognition is enhanced by interaction with RXRs. RXRa andRXRy are also highly regulated during adipocyte cell differentiation, particularly RXRa which is induced within 4 hours after exposure to DM.

Collectively, the results of these experiments demonstrate that (i) activators of PPAR are capable of potentiating differentiation of preadipocytes to adipocytes in the absence of Dex, IBMX, fetal calf serum and insulin; (ii) a natural activator of PPAR can be demonstrated in extracts of serum; (iii) down-regulation of RAR 1 is essential for adipocyte conversion; and, (iv) the strong correlation between
RARyl expression and the effectiveness of RA in inhibiting adipocyte conversion suggests that RARE1 mediates this phenomenon. Each of these observations can be exploited in order to provide the compositions and methods of the invention as follows.

Methods of identifying natural or synthetic PPAR activators. PPAR is a transcription factor whose activity is known to be involved in lipoprotein metabolism in addition to fat cell differentiation as described above. Known PPAR activators such as clofibrate (Physician's Desk Reference, 1993) and WY-14,643 (Keller et al., 1993, Toxicol. Appl.

Pharmacol, 119,52-58) are lipophilic, small molecules which can be absorbed with ease from the gastrointestinal tract and diffuse into cells. Other PPAR activators such as ETYA can be administered intravenously and are also absorbed into cells (Anderson et al., 1988, Agents Actions, 24 8-19).

Current methods useful for screening additional compounds for their ability to activate PPARs are cumbersome and difficult to perform in that they involve multiple transient transfection assays designed to measure transcription of PPARs following addition of a test compound to transfected cells (Issemann et al., 1990, Nature 347,645-650). In contrast, the above-described observed effects of heptaneextracted serum on preadipocytes provide a rapid and simple screening assay for the identification of additional PPAR activators, which activators may eventually prove useful in the treatment of the disorders noted above.

Essentially, the method involves the following steps.

First, aliquots of preadipocytes are cultured in the presence of heptane-extracted serum in order to induce morphological changes in the cells. As a control, a known PPAR activator such as clofibrate which is dissolved in a diluent, is added to one aliquot of cells so changed, while a test compound dissolved in the same diluent is added to another aliquot of the same cells. An additional set of controls will include cells incubated in normal medium and cells incubated in medium to which only the diluent has been added. Reversal of the morphological changes in cells following the addition of the known PPAR activator is an indication that the assay is functioning correctly, whereas reversal of the morphological changes in cells incubated with the test compound in an indication that the test compound is a candidate PPAR activator.

Such a candidate compound can be further tested for its ability to activate PPARs in transient transfection assays (Issemann et al., 1990, Nature, 347 645-650) and its ability to bind PPAR can be assessed in a simple ligand binding assay the performance of which is known to those skilled in the art. For example, Schatchard analysis or other methods may be performed to determine ligand binding such as that commonly used to analyze binding of ligands to thyroid hormone and other receptors (Allenby et al., 1993, Proc. Natl. Acad. Sci. USA, 90 30-34 ; Banner et al., 1992, Anal. Biochem. 200 163-170).

Methods of identifying compounds capable of inhibiting
PPAR activation. The data presented above demonstrate that compounds which are known to activate PPAR are also capable of inducing adipocyte differentiation. Moreover, such compounds are capable of reversing the effects of delipidated serum on preadipocytes. It therefore follows that compounds which inhibit PPAR activation will also be capable of inhibiting adipocyte differentiation. A simple screening assay designed to identify candidate compounds which inhibit PPAR activation involves culturing preadipocytes in the presence of a PPAR activator (under the conditions described above) in the presence or absence of a test compound. Cells which are incubated in the presence of the PPAR activator alone will proceed to differentiate, a process which can be easily assessed using any of the methods described above.

However, if the test compound is an antagonist of PPAR activation, differentiation of cells incubated in the presence of both the
PPAR activator and the test compound will be inhibited.

Alternatively, the screening method can be designed to take advantage of the observed morphological changes which preadipocytes undergo following the addition of delipidated serum, which changes are reversed upon addition of a PPAR activator. Thus, an assay can be designed as follows. Cells are first incubated in the presence of delipidated serum.

Following the induction of morphological changes in the cultures, a compound known to reverse these changes is added to the cultures in the presence or absence of a test compound. A lack of reversal of morphological changes in the cells in the presence of the test compound and reversal of such changes in the absence of the test compound is an indication that the test compound is a candidate antagonist of PPAR activation and therefore adipocyte differentiation. Candidate antagonists of
PPAR activation can be further examined for their ability to inhibit PPAR activation in transient transfection assays (Issemann et al., 1990, Nature, 347,645-650).

The methods of screening for candidate agonists or antagonists of PPAR activation are useful for the identification of candidate synthetic and candidate natural compounds capable of potentiating or inhibiting PPAR activation. Natural compounds include any compound normally found in biological fluids or tissues of either healthy or diseased animals or humans, and also include extracts of plants, bacteria and the like. By synthetic compounds is meant any compound which has been chemically synthesized. This term is also meant to include chemical modifications of natural compounds.

Isolation of the natural activator of PPAR. The data presented above indicate that a natural activator of PPAR is present in bovine calf serum and moreover, can be removed from the serum following extraction with n-heptane. This activator can be further purified by extraction of the n-hep constitutive expression of RARyl inhibits differentiation of these cells. It follows that compounds which inhibit downregulation of RARyl, i. e., those which induce constitutive expression of RARyl may also inhibit adipocyte differentiation.

Candidate compounds capable of inhibiting downregulation of RARrl can be identified by adding a test compound to cultures of preadipocytes incubated under standard differentiating conditions. One set of cultures is monitored for their ability to differentiate using any of the methods described above, while the level of expression of RARyl is assessed in a parallel set of cultures by Northern analysis.

Compounds which act through RARyl to inhibit adipocyte differentiation are likely to have significant advantages over
RA as potential anti-obesity agents because they are likely to act much more specifically than RA thereby minimizing the plethora of side effects induced by RA (Hill et al., 1992, Ann.

Rev. Nutr. 12 161-181). Thus, prevention of fat cell formation may be accomplished by oral administration of lipophilic compounds which would be expected to produce far less toxicity than RA.

There is an approximately 99k homology between mouse and human RART (Kastner et al., 1990, Proc. NAtl. Acad. Sci.

USA, 87 2700-2704). The specific involvement of mouse RARyl in adipocyte differentiation suggests that overexpression of this gene in cells is a useful method for preventing adipocyte development in humans in vivo. The promoter region of RARy1 is known (Lehmann et al., 1991, Nucl. Acids Res. 1991, 19 573-578) and in addition, specific activators of RARy have been identified (Bernard et al., 1992, Biochem. Biophys. Res. Comm.

1992,186 977-983). Since this invention discloses the discovery of the association between RARy1 and adipocyte differentiation, activators or ligands of RARyl can now be tested in the animal models described above for their ability to affect obesity.

It is known that mature differentiated adipocytes arise from undifferentiated stem cells and it is also known that stem cells are capable of cell division. Thus, inhibition of adipocyte formation from stem cells should reduce the number of adipocytes capable of fat storage and thereby reduce obesity. In this case, compounds which inhibit adipocyte formation should function as anti-obesity agents. In accordance with the present invention, such compounds are inhibitors of PPAR and activators of RARy1. However, depending upon the age of the animal at which treatment is initiated, the converse situation may also reduce obesity. In other words, compounds which promote stem cell differentiation to mature adipocytes should reduce the number of stem cells capable of cell division, which divided cells presumably will eventually differentiate into adipocytes and further cause obesity.

Then, in this case, compounds which promote adipocyte differentiation may function as anti-obesity agents. In accordance with the present invention, such compounds are activators of PPAR and inhibitors of RARe1.

Methods of testing candidate compounds in vivo.

Compounds which are identified as potential anti-obesity agents in the assays described above can be further tested for their ability to affect obesity in any of the well known animal models available for the study of obesity and described, for example, in York (Genetic Models of Animal Obesity) and
Sclafani (Dietary Obesity Models) both published in Obesity,
Bjorntorp and Brodoff eds. J. B. Lippincott Company, 1992.

In order to determine whether a candidate compound is capable of preventing obesity in vivo in an animal, the compound can be administered to a suitable animal bearing a fetus which is genetically susceptible to obesity. By monitoring the animal for obesity after birth, it can be determined whether the test compound has an effect in preventing obesity. Similarly, to determine whether a candidate compound is useful for the treatment of obesity, the compound can be administered to an obese animal and the effect of the compound on obesity can be determined.

A candidate compound can be administered to an animal in a pharmaceutically acceptable carrier in one of the traditional modes (e. g., orally, parenterally, transdermally or transmucosally), in a sustained release formulation using a biodegradable biopolymer, or by on-site delivery using micelles, gels and liposomes, or rectally (e. g., by suppository or enema). The compound can be administered to the animal in a dosage of 0.1 y/kg/day to 50 g/kg/day, either daily or at intervals sufficient to prevent or treat obesity. Precise formulations and dosages may be determined using standard techniques by a pharmacologist of ordinary skill in the art.

Compounds which are determined to be effective for the prevention or treatment of obesity in animals may also be useful in treatment of obesity in humans. Those skilled in the art of treating obesity in humans will know, based upon the data obtained in animal studies, the dosage and route of administration of the compound to humans. In general, the dosage and route of administration in humans is expected to the similar to that in animals.

While this invention has been disclosed with reference to specific embodiments, it is apparent that other embodiments and variations of this invention may be devised by others skilled in the art without departing from the true spirit and scope of the invention. The appended claims are intended to be construed to include all such embodiments and equivalent variations.

Claims

What is claimed is: 1. A method of identifying an activator of peroxisome proliferator activated receptor comprising: incubating a culture of preadipocytes in delipidated serum to induce a morphological change; adding to said culture of preadipocytes a test compound; and determining whether addition of said test compound causes reversal of said morphological change, reversal of said morphological change indicating that said test compound is a candidate activator of peroxisome proliferator activated receptor.

2. A method of identifying antagonists of peroxisome proliferator activated receptor activation comprising: incubating a culture of preadipocytes in delipidated serum to induce a morphological change; adding to said culture of preadipocytes an activator of peroxisome proliferator activated receptor in the presence or absence of a test compound; and determining whether the presence of said test compound inhibits reversal of said morphological change, inhibition of reversal of said morphological change indicating that said test compound is a candidate antagonist of peroxisome proliferator activated receptor activation.

3. A method of inhibiting differentiation of preadipocytes to adipocytes comprising adding to a culture of preadipocytes an activator of retinoic acid receptor.

4. A method of treating obesity in an animal comprising administering to said animal an activator of peroxisomal proliferator activated receptor in a dose of 0.1 yg to 50 g/kg/day wherein said activator is suspended in a pharmaceutically acceptable carrier.

5. A method of treating obesity in an animal comprising administering to said animal an inhibitor of peroxisome proliferator activated receptor in a dose of 0.1 yg to 50 g/kg/day wherein said inhibitor is suspended in a pharmaceutically acceptable carrier.

6. A method of treating obesity in an animal comprising administering to said animal an inhibitor of retinoic acid receptory in a dose of 0.1 yg to 50 g/kg/day wherein said inhibitor is suspended in a pharmaceutically acceptable carrier.

7. A substantially pure preparation of a naturally occurring peroxisome proliferator activated receptor activator.