Epidermal Growth Factor and 1a,25-Dihydroxyvitamin D3

Suppress Lipogenesis in Hamster Sebaceous Gland CellsIn Vitro

Takashi Sato, Noriyasu Imai, Noriko Akimoto, Takayuki Sakiguchi,* Kenji Kitamura,* and Akira ItoDepartment of Biochemistry, Tokyo University of Pharmacy and Life Science, School of Pharmacy, Hachioji, Tokyo, Japan; *Shiseido Pharmaco

Science Research Laboratories, Tsuzuki, Yokohama, Japan

We have previously reported the establishment of aculture system of hamster auricular sebocytes.Although their morphologic and biochemical proper-ties are very similar to those of human sebocytes, theregulation of lipogenesis is not clear. Therefore, weinvestigated the effect of epidermal growth factor,all-trans retinoic acid, 1a,25-dihydroxyvitamin D3,and androgens such as testosterone and 5a-dihydro-testosterone on lipogenesis in cultured hamstersebocytes. Intracellular lipid droplets detected withOil-Red-O staining were observed in 5 d cultures andincreased in a time-dependent manner; 40.7% 61.11% of 2 wk cultured cells tested lipid-positive by¯ow cytometric analysis. When the hamster sebocyteswere cultured in the presence of epidermal growthfactor, all-trans retinoic acid, or 1a,25-dihydroxyvita-min D3, the intracellular lipid droplets were dimin-ished by all-trans retinoic acid and epidermal growthfactor, and slightly by 1a,25-dihydroxyvitamin D3.The intracellular lipid droplets consisted mainly oftriglycerides (71.8%) and, as minor components,cholesterol (18.0%), wax esters (3.6%), and free fattyacids (6.6%). Epidermal growth factor and all-trans

retinoic acid decreased the intracellular accumulationof triglycerides (92.6% and 96.1% inhibition, respect-ively) and free fatty acids (54.3% and 62.6% inhibi-tion, respectively) in the sebocytes. In addition,1a,25-dihydroxyvitamin D3 decreased the triglyceridelevel (34.3% inhibition), but augmented the accumu-lation of wax esters (30% increase). There was no dif-ference in the level of cholesterol as a result of thesetreatments, however. In contrast, 5a-dihydrotestos-terone augmented the formation of intracellular lipiddroplets along with an increase in the accumulationof triglycerides in hamster sebocytes. Our ®ndingsthat regulation of lipogenesis by all-trans retinoic acidand androgen in hamster sebocytes is identical toregulation in humans suggest that hamster sebocytesare useful for the elucidation of sebaceous function atthe cellular level. Furthermore, this is the ®rst evid-ence that epidermal growth factor and 1a,25-dihy-droxyvitamin D3 may act as suppressors in theregulation of lipogenesis in hamster sebocytes in vitro.Key words: epidermal growth factor/hamster sebocytes/seba-ceous glands. J Invest Dermatol 117:965±970, 2001

Sebaceous glands are important skin appendages and sebumexcretion is considered to be associated with thefunctional maintenance of the cutaneous surface as abiologic barrier (Thody and Shuster, 1989). The develop-ment of sebaceous glands is dependent on androgens and

sebocytic differentiation sequentially occurs with accumulatingabundant cytoplasmic lipids (Sawaya et al, 1988; Akamatsu et al,1992; Zouboulis et al, 1994a; Rosen®eld et al, 1998). An excessincrease in the endogenous level of androgens causes sebaceous-gland disorders such as acne and seborrhoea, which are the mostcommon skin diseases (Harris et al, 1983; PieÂrard et al, 1987;Zouboulis et al, 1998).

Proliferation and lipogenesis in sebocytes of human and rodentsebaceous glands are known to be augmented by testosterone and5a-dihydrotestosterone (5a-DHT) in vivo and in vitro (Sawaya et al,1988; Akamatsu et al, 1992; Zouboulis et al, 1994a; Rosen®eld et al,1998). In addition to androgen, other hormones and growth factorshave been shown to participate in the development of sebaceousglands. Epidermal growth factor (EGF) stimulates sebaceous glandsin the pinna of hamster to proliferate in vivo (Matias andOrentreich, 1983). In contrast, estrogens reduce lipogenesis inhuman and rat sebaceous glands (Strauss et al, 1962; Ebling andSkinner, 1983). All-trans and 13-cis retinoic acids also prevent lipidformation in human and rodent sebaceous gland cells in vivo andin vitro (Strauss et al, 1980; Jones et al, 1983; Zouboulis et al, 1991;Hommel et al, 1996; Orfanos and Zouboulis, 1998).

Recently, Ito et al (1998) have established a preparation ofhamster sebocytes from the auricle and demonstrated that theproliferation of hamster sebocytes in response to androgens issimilar to that of human sebocytes. In addition, Plewig andLuderschmidt (1977) reported that the hamster sebaceous gland issimilar to the human gland with regard to its size, response toandrogens, and turnover time. Therefore, this culture system is

0022-202X/01/$15.00 ´ Copyright # 2001 by The Society for Investigative Dermatology, Inc.

965

Manuscript received January 11, 2001; revised June 21, 2001; acceptedfor publication June 28, 2001.

Reprint requests to: Dr. Takashi Sato, Department of Biochemistry,Tokyo University of Pharmacy and Life Science, School of Pharmacy,1432-1 Horinouchi, Hachioji, Tokyo 192-0392, Japan. Email: satotak@ps.toyaku.ac.jp

Abbreviations: 1,25(OH)2D3, 1a, 25-dihydroxyvitamin D3; atRA, all-trans retinoic acid; PBS(±), Ca2+ and Mg2+ free phosphate-buffered saline;FSC, forward scatter; SSC, side scatter; TG, triglyceride.

considered to be a useful tool for studying functions of sebaceousglands in vitro. The regulation of lipogenesis in hamster sebocytesremains unclear, however.

In this study, we investigated the effect of growth factor andhormones on lipid formation and demonstrated that testosteroneand 5a-DHT augmented lipogenesis, but EGF, 1a,25-dihydroxy-vitamin D3 [1,25(OH)2D3], and all-trans retinoic acid (atRA)suppressed it in hamster sebocytes in vitro.

MATERIALS AND METHODS

Culture and treatment of hamster sebocytes Hamster sebocyteshave been established from sebaceous glands of auricles of 5-wk-old malegolden hamsters (Ito et al, 1998) and cultured in culture dishes of 60 mmdiameter (1 3 103 cells per cm2) (Becton Dickinson, Tokyo, Japan) inDulbecco's modi®ed Eagle's medium (DMEM)/Ham's F12 medium(1:1) (Life Technologies, Grand Island, MD) supplemented with 6%heat-denatured fetal bovine serum (FBS) (BioWhittaker, Walkersville,MD), 2% human serum (ICN Biochemicals, Costa Mesa, CA), and0.68 mM L-glutamine (Life Technologies) in the presence or absence ofrecombinant human EGF (Progen Biotechnik, Heidelberg, Germany),1,25(OH)2D3, atRA, testosterone or 5a-DHT (Sigma, St. Louis, MO)for up to 14 d.

Hematoxylin and Oil-Red-O staining Cultured hamster sebocyteswere washed once with Ca2+- and Mg2+-free phosphate-buffered saline[PBS(±)] and ®xed with 4% paraformaldehyde (Wako Pure Chemicals,Osaka, Japan) diluted with PBS(±) for 1 h at room temperature. Thecells were stained ®rst with Mayer's hematoxylin solution (Wako) atroom temperature for 5 min and then with 0.3% Oil-Red-O (Sigma) at37°C for 15 min. In addition, the number of cells in six randomlychosen areas per dish was counted under a light microscope at203 magni®cation.

Flow cytometric analysis Hamster sebocytes were maintained for2 wk in DMEM/Ham's F12 medium (1:1) supplemented with 6% heat-denatured FBS, 2% human serum, and 0.68 mM L-glutamine, and thenwere stained with Nile red (Sigma) to detect lipid-positive cells (®nalconcentration of 100 ng per ml) in PBS(±) for 20 min at roomtemperature in the dark. The dye solution was removed and then thecells were washed twice with PBS(±). The cells treated with Nile redwere harvested with 0.25% trypsin and 0.02% ethylenediamine tetraaceticacid (EDTA) in PBS(±) and then resuspended in the culture medium.Cell size and cellular ¯uorescence were measured for 10,000 cells using aFACScan ¯ow cytometer (Becton Dickinson). Excitation laser light was488 nm and the ¯uorescent intensity at 530 6 15 nm was detected. Thecell distribution was expressed as forward and side scatters (FSC andSSC, respectively).

Analysis of intracellular lipid compositions The cells were washedtwice with PBS(±) and then harvested with 0.25% trypsin and 0.02%EDTA in PBS(±). The cells were sonicated and then mixed withchloroform:methanol (2:1) for 5 min at room temperature. The mixturewas centrifuged at 2000 rpm for 5 min at room temperature followingaddition of 0.88% KCl, and then the methanol (the upper phase) wasremoved. Lipids in chloroform (the lower phase) were collected andthen subjected to an automatic thin-layer chromatography Iatroscan(Iatron Laboratories, Tokyo, Japan) (Shantha, 1992; Pisch et al, 1997).Tripalmitin (triglyceride; TG), palmitic acid (free fatty acid), cholesterol,cholesterol palmitate (cholesterol ester) (Doosan Serdary ResearchLaboratories, Englewood Cliffs, NJ), and palmityl palmitate (wax ester)(Nu-Chek-Prep, Elysian, MN) were used as lipid standards in thechromatography. The amounts of lipid components were calculated byan internal control concomitantly performed using authentic cholesterolacetate (2 mg) (Doosan).

Quanti®cation of DNA content The cell lysate (100 ml) preparedabove was coated in triplicate on wells in 24-well multiplates and thenincubated with 3,5-diaminobenzoic acid dihydrochloride (Sigma)(400 mg per ml) at 60°C for 45 min according to the method ofJohnson-Wint and Hollis (1982). After the reaction, the ¯uorescentintensity was measured by excitation at 365 nm and emission at 530 nmfollowing the addition of 1 M hydrochloride. The content ofintracellular DNA was calculated using a standard curve concomitantlyproduced using authentic salmon sperm DNA (6.125±200 mg per ml).

Statistical analysis Data were analyzed by Student's t test; p < 0.05was considered to be statistically signi®cant.

RESULTS

Lipid formation in cultured hamster sebocytes Whenhamster sebocytes were cultured for 14 d, lipid droplets in thecytoplasm were obviously stained with Oil-Red-O (Fig 1A). Thecellular morphology was distorted and nuclei were dif®cult toobserve because the cytoplasm was ®lled with lipid droplets. Inaddition, accumulation of intracellular lipid droplets was observedas early as at 5 d in culture and increased in a time-dependentmanner (Fig 1B). During cultivation, small lipid droplets weredetermined in the cytoplasm within 1 wk by Oil-Red-O stainingand grew to large droplets fusing together (data not shown).Furthermore, a similar histochemical observation was obtained byutilizing the other dye reagent, Nile red (data not shown). On theother hand, we performed ¯ow cytometric analysis to determinethe percentage of lipid-accumulated cells. We ®rst identi®ed anarea with lower FSC and SSC as Gate 1, in which there were morethan 95% of lipid-negative sebocytes induced by EGF (see Fig 3)(Fig 2A, insert). Differentiated sebocytes with intracellular lipidsseemed to be granular so that the cellular population was shifted toan area of larger SSC, termed Gate 2. When sebocytes were labeledwith Nile red reagent (Greenspan et al, 1985) and then subjected to¯ow cytometric analysis, most sebocytes in the 2-wk culture weredistributed in both Gates 1 and 2 (Fig 2A). The ¯uorescenceintensity of cells in Gate 2 was about 40 times higher than that inGate 1 (Fig 2B). We also obtained similar results in the ¯owcytometric analysis when the cells were stained with Oil-Red-O(data not shown), indicating that the cells in Gate 2 were lipid-positive. Furthermore, we calculated that 40.7% 6 1.11% of thecultured sebocytes accumulated intracellular lipid droplets (Fig 2).Thus, these results suggest that hamster sebocytes are able tospontaneously accumulate intracellular lipids under these cultureconditions in vitro.

Regulation of lipogenesis in hamster sebocytes To clarifythe regulation of lipogenesis in hamster sebocytes, we investigatedthe effect of EGF, 1,25(OH)2D3, and atRA in the accumulation ofintracellular lipids. The level of intracellular lipids was much less inthe EGF-treated sebocytes compared to control culture (Fig 3A,B). In addition, lipid formation no longer occurred in the atRA-treated cells (Fig 3C). Furthermore, 1,25(OH)2D3 decreasedlipogenesis, and even in the Oil-Red-O-positive cells theamount of intracellular lipids became smaller than in theuntreated cells (Fig 3D).

Lipid composition in hamster sebocytes treated with EGF,atRA, and 1,25(OH)2D3 We next quantitatively analyzed thecomposition of intracellular lipids in hamster sebocytes by thin-layer chromatography using a ¯ame ionization detector. As shownin Fig 4B, the intracellular lipids in hamster sebocytes were foundto consist of TG, wax esters, free fatty acids, and cholesterol. Thelipids consisted of TG (71.8%) as the major lipid component, andfree fatty acids (6.6%), cholesterol (18.0%), and wax esters (3.6%)were detected as minor components (Fig 5, control). In addition,we con®rmed the time-dependent increase in total lipidbiosynthesis resulting mostly from the augmentation of the TGlevel in cultured hamster sebocytes (data not shown).

When the sebocytes were treated with EGF, atRA, and1,25(OH)2D3, the intracellular accumulation of TGs was sup-pressed by EGF and atRA (92.6% and 96.1% inhibition, respec-tively). In addition, the level of free fatty acids was similarlydecreased by EGF and atRA (54.3% and 62.6% inhibition,respectively). Furthermore, 1,25(OH)2D3 slightly decreased theTG level (34.3% inhibition) but slightly augmented the accumu-lation of wax esters (30% increase). In contrast, there was nodifference in the level of cholesterol under these treatments (Fig 5).Therefore, these results suggest that the suppression of lipogenesisby EGF, atRA, and 1,25(OH)2D3 is due to a decrease in theaccumulation of TG in hamster sebocytes.

5a-DHT augments lipogenesis in hamster sebocytes 5a-DHT and testosterone have been shown to augment the

966 SATO ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Figure 3. Intracellular lipid formation isinhibited by EGF, atRA, and 1,25(OH)2D3 inhamster sebocytes. Hamster sebocytes at thesecond passage were untreated (A) or treated withEGF (100 ng per ml) (B), atRA (1 mM) (C) and1,25(OH)2D3 (100 nM) (D) every 3 d for 2 wk.After ®xing, the cells were stained with Oil-Red-O and Mayer's hematoxylin as described inMaterials and Methods.

Figure 2. Characterization of lipid-positivehamster sebocytes. Hamster sebocytes at thesecond passage were cultured for 2 wk, and thenstained with Nile red as described in Materials andMethods. The Nile-red-stained cells were subjectedto ¯ow cytometric analysis in SSC and FSC (A)and then the ¯uorescence intensity wasdetermined for characterization of lipid-positiveand lipid-negative cells (B). Inserted panel, EGF-treated sebocytes. Gates 1 and 2 correspond to thelipid-negative and lipid-positive sebocytes,respectively.

Figure 1. Oil-Red-O staining in culturedhamster sebocytes. (A) Hamster sebocytes at thesecond passage were cultured for 2 wk, and thenstained with Oil-Red-O and Mayer's hemat-oxylin. Arrows indicate nuclei. (B) The numberof Oil-Red-O-positive cells was counted at theculture days shown. The data represent the mean6 SD of six individual areas in the dish.

VOL. 117, NO. 4 OCTOBER 2001 SUPPRESSION OF LIPOGENESIS IN HAMSTER SEBOCYTES 967

biosynthesis of lipids in human and rat sebaceous glands in vivo andin vitro (Sawaya et al, 1988; Akamatsu et al, 1992; Zouboulis et al,1994a; Rosen®eld et al, 1998). Ito et al (1998) also reported that theproliferation of hamster sebocytes is androgen-dependent. Wedemonstrated that the accumulation of intracellular lipids incytoplasm was augmented in 5a-DHT-treated hamster sebocytes(Fig 6), and that the lipid-positive cells in 5a-DHT treatmentamounted to 52.1% 6 1.58% and increased 1.3-fold in comparisonto untreated cells (p < 0.001) (data not shown). Furthermore, asshown in Fig 7, 5a-DHT increased the amount of TG whereasthe levels of wax esters, free fatty acids, and cholesterol were notin¯uenced, indicating that the increased lipid formation was mostlydue to the increase in TG level. Therefore, these results suggestthat, as in human and rat sebaceous glands, androgens act asinducers for lipogenesis in hamster sebocytes in vitro.

DISCUSSION

Sebocytes are epithelial cells. Accumulation of lipid droplets in theircytoplasm re¯ects terminal differentiation (Zouboulis et al, 1994b).So far, many investigators have reported the function andregulation of sebum under physiologic and pathologic conditions(Harris et al, 1983; PieÂrard et al, 1987; Thody and Shuster, 1989;Zouboulis et al, 1998). The regulatory mechanism of lipogenesis insebaceous glands is not fully understood, however. Therefore, it isof interest to clarify physiologic candidate(s), which negativelyregulate(s) lipid formation in sebaceous glands. In this study, wedemonstrated that EGF, 1,25(OH)2D3, and atRA inhibitedlipogenesis in cultured hamster sebocytes.

It has been reported that organ culture systems of sebaceousglands are established in human and rodents (Strauss et al, 1962;1980; Ebling and Skinner, 1983; Jones et al, 1983; Sawaya et al,1988). Recently, in vitro culture systems of human and rat sebocyteshave also been reported, and the accumulation of intracellular lipidsis augmented by androgens such as testosterone and 5a-DHT andinhibited by retinoic acids (Jones et al, 1983; Zouboulis et al, 1994a;Rosen®eld et al, 1998). The hamster sebaceous gland is similar tothe human gland with regard to its size, response to androgens, andturnover time in vivo (Plewig and Luderschmidt, 1977). Ito et al(1998) have established a culture method of hamster sebocytes, and,like human sebocytes, the proliferation of hamster sebocytes is

androgen-dependent. Furthermore, the accumulated intracellularlipids consist of TG, wax esters, cholesterol, and free fatty acids;these are identical to accumulated intracellular lipids in humansebocytes except for squalene and cholesterol esters (Ito et al, 1998).We demonstrated that the accumulation of intracellular lipids inhamster sebocytes was augmented by both testosterone (data notshown) and 5a-DHT and suppressed by atRA. Therefore, theseresults suggest that the cell proliferation and lipogenesis of hamstersebocytes in response to androgens and retinoic acids are similar tothose of human sebocytes.

We characterized lipid-positive hamster sebocytes as amountingto 40.7% 6 1.11% of the cultured cells, and 5a-DHT treatmentfurther augmented this percentage (30% increase) (data not shown).Rosen®eld et al (1999) also reported that in rat sebocytes only 5% ofthe cells were lipid-positive, but 5a-DHT treatment increased thepercentage to approximately 15% in culture. Although the cultureconditions and analytical procedures for evaluating lipid-positivecells are different for hamster and rat sebocytes, the differentialpotency in sebocytes is likely to be higher in hamster than in rat.

Various endogenous factors such as growth factors and hormonesare known to participate in maintaining skin homeostasis in vivo.EGF plays an important role in maintaining cutaneous events suchas the proliferation and keratinization of the epidermis (Frati et al,1977; Green et al, 1983; Bhora et al, 1995) and the remodeling ofextracellular matrix (McDonnell et al, 1990; McCawley et al, 1998).It has been reported that EGF receptors exist on sebaceous glands inrat (Green et al, 1983). In addition, Matias and Orentreich (1983)reported that administration of EGF into the pinna of hamsterincreases the number of cells per sebaceous gland unit in vivo. Wedemonstrated that the cellular DNA content in EGF-treatedhamster sebocytes was 2.6-fold higher than in untreated cells (datanot shown). In contrast, Zouboulis et al (1998) reported that EGFdoes not affect the proliferation of human sebocytes in vitro.Therefore, it is likely that the involvement of EGF in sebaceousproliferation may be dependent on cell species (Nikkari, 1974) andat least different between hamster and human sebocytes. On theother hand, we demonstrated that the accumulation of intracellularlipids was inhibited in EGF-treated hamster sebocytes. Similar

Figure 4. Characterization of intracellular lipids in hamstersebocytes. Lipids were analyzed by automatic thin-layer chromato-graphy as described in Materials and Methods. (A, B) The chromato-graphic analyses of lipid standards and lipids derived from hamstersebocytes at the third passage, respectively. W, wax esters; CE,cholesterol esters; InC, internal control; T, triglycerides; F, free fattyacids; CH, cholesterol.

Figure 5. Alteration of lipid composition in EGF, atRA, and1,25(OH)2D3-treated hamster sebocytes. Hamster sebocytes at thesecond passage were untreated or treated with EGF (100 ng per ml),1,25(OH)2D3 (100 nM), and atRA (1 mM) as described in Fig 3. Thecomposition of intracellular lipids extracted from the cells was analyzedby thin-layer chromatography as described in Materials and Methods. Datarepresent the mean 6 SD of triplicate dishes. *, **, ***, Signi®cantlydifferent from each composition in control (p < 0.05, 0.01, and 0.001,respectively). Total, sum of each composition; TGs, triglycerides; WAXs,wax esters; FFAs, free fatty acids; CHO, cholesterol.

968 SATO ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

observations have been reported in organ cultures of the humansebaceous gland, in which lipogenesis is increased with EGFdepletion from the culture medium (Guy and Kealey, 1998). Theseresults therefore suggest that EGF acts not only as a mitogenicfactor but also as an antilipogenic factor in hamster sebocytes.

1,25(OH)2D3 has been shown to play an important role inepidermal differentiation. 1,25(OH)2D3 inhibits the proliferationbut augments the differentiation of human keratinocytes (Gibbset al, 1996; Sorensen et al, 1997). In addition, the administration of1,25(OH)2D3 to psoriasis patients is ef®cacious for repairing theorganization of epidermis in vivo (van de Kerkhof, 1998; Peters et al,2000). We demonstrated that 1,25(OH)2D3 suppressed the accu-mulation of intracellular lipid droplets in hamster sebocytes, but theextent of suppression was less than that of EGF or atRA treatment.In addition, like EGF and atRA, 1,25(OH)2D3 decreased the levelof TG but slightly augmented that of wax esters, whereas thecellular DNA content was not altered in 1,25(OH)2D3-treatedhamster sebocytes. Thus, this is the ®rst evidence that 1,25(OH)2D3

may be involved in the suppressive regulation of lipogenesis inhamster sebocytes.

Brind et al (1986) showed that sterol esters are major componentsof ear sebum from Syrian hamster in vivo. Concerning thediscrepancy between that report and our ®ndings, we suggest a

possibility that the lipid composition in sebaceous glands in vivo mayresult from the modulation of lipogenesis by various growth factorsand hormones. In this study, the level of cholesterol in EGF- oratRA-treated sebocytes was higher than that of TG. In addition,the level of wax ester was almost equal to that of TG under thesetreatments. Therefore, we speculate that the substantial com-position of hamster sebum may be similar in vitro and in vivo.Further experiments will be required to address the hypothesis.

In conclusion, we demonstrated that, as in human and ratsebocytes, lipogenesis in hamster sebocytes is augmented byandrogens and inhibited by atRA. Taken together with previousreports (Ito et al, 1998), this suggests that the hamster sebocyteculture system will be useful for the elucidation of sebaceousfunction at the cellular level. Furthermore, this is the ®rst evidencethat EGF and 1,25(OH)2D3 may act as suppressors to controlsebum excretion in hamster sebaceous glands.

REFERENCES

Akamatsu H, Zouboulis ChC, Orfanos CE: Control of human sebocyte proliferationin vitro by testosterone and 5-alpha-dihydrotestosterone is dependent on thelocalization of the sebaceous glands. J Invest Dermatol 99:509±511, 1992

Bhora FY, Dunkin BJ, Batzri S, Aly HM, Bass BL, Sidawy AN, Harmon JW: Effectof growth factors on cell proliferation and epithelialization in human skin. J SurgRes 59:236±244, 1995

Brind JL, Alani E, Wheatley VR, Orentreich N: Analysis of ear sebum of the Syrianhamster (Mesocricetus auratus) reveals pronounced sexual dimorphism. CompBiochem Physiol B 84:403±407, 1986

Ebling FJ, Skinner J: The local effects of topically applied estradiol, cyproteroneacetate, and ethanol on sebaceous secretion in intact male rats. J Invest Dermatol81:448±451, 1983

Frati L, D'Armiento M, Gulletta E, Verna R, Covelli I: The control of epidermisproliferation by epidermal growth factor (EGF). Relationship with cyclicnucleotides systems. Pharmacol Res Commun 9:815±822, 1977

Gibbs S, Backendorf C, Ponec M: Regulation of keratinocyte proliferation anddifferentiation by all-trans-retinoic acid, 9-cis-retinoic acid and 1,25-dihydroxyvitamin D3. Arch Dermatol Res 288:729±738, 1996

Green MR, Basketter DA, Couchman JR, Rees DA: Distribution and number ofepidermal growth factor receptors in skin is related to epithelial cell growth.Dev Biol 100:506±512, 1983

Greenspan P, Mayer EP, Fowler SD: Nile red: a selective ¯uorescent stain forintracellular lipid droplets. J Cell Biol 100:965±973, 1985

Guy R, Kealey T: The organ-maintained human sebaceous gland. Dermatology196:16±20, 1998

Harris HH, Downing DT, Stewart ME, Strauss JS: Sustainable rates of sebumsecretion in acne patients and matched normal control subjects. J Am AcadDermatol 8:200±203, 1983

Hommel L, Geiger J-M, Harms M, Saurat J-H: Sebum excretion rate in subjectstreated with oral all-trans-retinoic acid. Dermatology 193:127±130, 1996

Ito A, Sakiguchi T, Kitamura K, Akamatsu H, Horio T: Establishment of a tissueculture system for hamster sebaceous gland cells. Dermatology 197:238±244,1998

Johnson-Wint B, Hollis S: A rapid in situ deoxyribonucleic acid assay for determiningcell number in culture and tissue. Anal Biochem 122:338±344, 1982

Jones DH, King K, Miller AJ, Cunliffe WJ: A dose±response study of 13-cis-retinoicacid in acne vulgaris. Br J Dermatol 108:333±345, 1983

van de Kerkhof PC: An update on vitamin D3 analogues in the treatment of psoriasis.Skin Pharmacol Appl Skin Physiol 11:2±10, 1998

Matias JR, Orentreich N: Stimulation of hamster sebaceous glands by epidermalgrowth factor. J Invest Dermatol 80:516±519, 1983

McCawley LJ, O'Brien P, Hudson LG: Epidermal growth factor (EGF)- and scatterfactor/hepatocyte growth factor (SF/HGF)-mediated keratinocyte migration is

Figure 6. Augmentation of intracellular lipidformation in 5a-DHT-treated hamstersebocytes. Hamster sebocytes at the secondpassage were untreated (A) or treated with 5a-DHT (10 nM) (B) as described in Fig 3. After®xing, the cells were stained with Oil-Red-O andMayer's hematoxylin as described in Materials andMethods.

Figure 7. 5a-DHT increases intracellular TG level in hamstersebocytes. Hamster sebocytes at the second passage were untreated ortreated with 5a-DHT (10 nM) and then the extracted lipids wereanalyzed as described in Fig 5. Data represent the mean 6 SD oftriplicate dishes. *Signi®cantly different from control (p < 0.05). Total,sum of each composition; TGs, triglycerides; WAXs, wax esters; FFAs,free fatty acids; CHO, cholesterol.

VOL. 117, NO. 4 OCTOBER 2001 SUPPRESSION OF LIPOGENESIS IN HAMSTER SEBOCYTES 969

coincident with induction of matrix metalloproteinase (MMP) -9. J Cell Physiol176:255±265, 1998

McDonnell SE, Kerr LD, Matrisian LM: Epidermal growth factor stimulation ofstromelysin mRNA in rat ®broblasts requires induction of proto-oncogenesc-fos and c-jun and activation of protein kinase C. Mol Cell Biol 10:4284±4293,1990

Nikkari T: Comparative chemistry of sebum. J Invest Dermatol 62:257±267, 1974Orfanos CE, Zouboulis ChC: Oral retinoids in the treatment of seborrhoea and acne.

Dermatology 196:140±147, 1998Peters BP, Weissman FG, Gill MA: Pathophysiology and treatment of psoriasis. Am J

Health Syst Pharm 57:645±659, 2000PieÂrard GE, Pierard-Franchimont C, Le T: Seborrhoea in acne-prone and acne-free

patients. Dermatologica 175:5±9, 1987Pisch S, Bornscheuer UT, Meyer HH, Schmid RD: Properties of unusual

phospholipids IV. Chemienzymatic synthesis of phospholipids bearingacetylenic fatty acids. Tetrahedron 53:14627±14634, 1997

Plewig G, Luderschmidt C: Hamster ear model for sebaceous glands. J Invest Dermatol68:171±176, 1977

Rosen®eld RL, Eplewski D, Kentsis A, Ciletti N: Mechanisms of androgeninduction of sebocyte differentiation. Dermatology 196:43±46, 1998

Rosen®eld RL, Kentsis A, Deplewski D, Ciletti N: Rat preputial sebocytedifferentiation involves peroxisome proliferator-activated receptors. J InvestDermatol 112:226±232, 1999

Sawaya ME, Honig LS, Hsia SL: Increased androgen binding capacity in sebaceousglands in scalp of male-pattern baldness. J Invest Dermatol 92:91±95, 1988

Shantha NC: Thin-layer chromatography-¯ame ionization detection Iatroscansystem. J Chromatogr 624:21, 1992

Sorensen S, Solvsten H, Politi Y, Kragballe K: Effects of vitamin D3 on keratinocyteproliferation and differentiation in vitro: modulation by ligands for retinoic acidand retinoid X receptors. Skin Pharmacol 10:144±152, 1997

Strauss JS, Kligman AM, Pochi PE: Effects of androgens and estrogens on humansebaceous glands. J Invest Dermatol 39:139±155, 1962

Strauss JS, Stranieri AM, Farrell LN, Downing DT: The effect of marked inhibitionof sebum production with 13-cis-retinoic acid on skin surface lipidcomposition. J Invest Dermatol 74:66±67, 1980

Thody AJ, Shuster S: Control and function of sebaceous glands. Physiol Rev 69:383±416, 1989

Zouboulis ChC, Korge B, Akamatsu H, Xia L, Schiller S, Gollnick H, Orfanos CE:Effects of 13-cis-retinoic acid, all-trans-retinoic acid and acitretin on theproliferation, lipid synthesis and keratin expression of cultured humansebocytes in vitro. J Invest Dermatol 96:792±797, 1991

Zouboulis ChC, Akamatsu H, Stephanek K, Orfanos CE: Androgens affect theactivity of human sebocytes in culture in a manner dependent on thelocalization of the sebaceous glands and their effect is antagonized byspironolactone. Skin Pharmacol 7:33±40, 1994a

Zouboulis ChC, Krieter A, Gollnick H, Mischke D, Orfanos CE: Progressivedifferentiation of human sebocytes in vitro is characterized by increased cell sizeand altered antigenic expression and is regulated by culture duration andretinoids. Exp Dermatol 3:151±160, 1994b

Zouboulis ChC, Xia L, Akamatsu H, et al: The human sebocyte culture modelprovides new insights into development and management of seborrhoea andacne. Dermatology 196:21±31, 1998

970 SATO ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY