JOURNAL OF CELLULAR PHYSIOLOGY 198:91–99 (2004)

Effects of Peripheral Benzodiazepine Receptor Ligandson Proliferation and Differentiation of

Human Mesenchymal Stem Cells

D.H. LEE,1 S.K. KANG,1 R.H. LEE,1 J.M. RYU,1 H.Y. PARK,1 H.S. CHOI,1 Y.C. BAE,2

K.T. SUH,3 Y.K. KIM,1 AND JIN SUP JUNG1*1Department of Physiology, College of Medicine, Pusan National University,

Pusan, Korea2Department of Plastic Surgery, College of Medicine,

Pusan National University, Pusan, Korea3Department of Orthopaedic Surgery, College of Medicine,

Pusan National University, Pusan, Korea

The peripheral benzodiazepine receptor (PBR) has been known to have manyfunctions such as a role in cell proliferation, cell differentiation, steroidogenesis,calcium flow, cellular respiration, cellular immunity, malignancy, and apoptosis.However, the presence of PBR has not been examined in mesenchymal stem cells.In this study, we demonstrated the expression of PBR in human bone marrowstromal cells (hBMSCs) and human adipose stromal cells (hATSCs) by RT-PCR andimmunocytochemistry. To determine the roles of PBR in cellular functions ofhumanmesenchymal stem cells (hMSCs), effects of diazepam, PK11195, and Ro5-4864 were examined. Adipose differentiation of hMSCs was decreased by highconcentration of PBR ligands (50 mM), whereas it was increased by lowconcentrations of PBR ligands (<10 mM). PBR ligands showed a biphasic effecton glycerol-3-phosphate dehydrogenase (GPDH) activity. High concentration ofPBR ligands (from 25 to 75 mM) inhibited proliferation of hMSCs. However,clonazepam, which does not have an affinity to PBR, did not affect adiposedifferentiation and proliferation of hMSCs. The PBR ligands did not induce celldeath in hMSCs. PK11195 (50 mM) and Ro5-5864 (50 mM) induced cell cycle arrestin the G2/M phase. These results indicate that PBR ligands play roles in adiposedifferentiation and proliferation of hMSCs. J. Cell. Physiol. 198: 91–99, 2004.� 2003 Wiley-Liss, Inc.

Benzodiazepines have several clinical properties suchas anticonvulsants, anxiolytics, muscle relaxants, andhypnotics. These therapeutic effects are mediated viathe central benzodiazepine receptors located in thecentral nervous system (CNS) coupled with GABAreceptors (Mohler and Okad, 1977; Skolnick and Paul,1982; Bormann, 1988). Benzodiazepine receptors havebeen classified into central-type benzodiazepine recep-tors (CBRs) and peripheral type benzodiazepine recep-tors (PBRs) (Braestrup and Squires, 1977; Tallmanet al., 1980; Schoemaker et al., 1983). CBRs are locatedon the plasma membrane in the CNS and contain aGABA receptor site and chloride ion channel (De Loreyand Olsen, 1992). PBRs are distinct from the CBRs in thepharmacological, anatomical, structural, and physiolo-gical aspects (Pierre et al., 2000). PBRs are expressed invarious peripheral tissues and located on the outermitochondrial membrane (Moshe et al., 1999), and differfrom CBRs in their lack of coupling to GABA receptorsand in their drug specificity.

PBR ligands affect various cellular functions accord-ing to cell types and their concentrations (Stepin et al.,1981; Ikezaki and Black, 1990; Moshe et al., 1992, 1999).

Diazepam and Ro5-4864 stimulated mitosis in pituitarycells, but they inhibited cellular proliferation in mousespleen lymphocytes (Pawlikowski et al., 1987, 1988).Cell proliferation has shown to be inhibited by PBRligands in various tumor cells in micromolar concentra-tions, whereas PBR ligands stimulated mitosis atnanomolar concentrations (Ikezaki and Black, 1990;Moshe et al., 1992). Diazepam induced cell differentia-tion in erythroleukemia cells and Ro5-4864 increasedmelanogenesis in melanoma cells (Clarke and Ryan,1980; Matthew et al., 1981). Recently, it has been

� 2003 WILEY-LISS, INC.

Contract grant sponsor: Ministry of Health; Contract grantnumber: 02-PJ10-PG8-EC01-0018.

*Correspondence to: Jin Sup Jung, Department of Physiology,College of Medicine, Pusan National University 1 Ga, Ami-Dong,Suh-Gu, Pusan 602-739, Korea. E-mail: jsjung@pusan.ac.kr

Received 17 March 2003; Accepted 4 June 2003

DOI: 10.1002/jcp.10391

reported that PBR ligands have various effects on cellsurvial according to cell types. That is, PBR ligands takepart in induction of apoptosis in some cells includingcancer cells (Marchetti et al., 1996a; Tanimoto et al.,1999; Fischer et al., 2001; Decaudin et al., 2002; Sutteret al., 2002), but reveals protective effects on apoptoticcell death in other cells (Carayon et al., 1996; Bono et al.,1999; Stoebner et al., 2001). Beside of effect of PBRligands on proliferation, differentiation, and apoptosis,they affect numerous biological functions, includingimmunodulation, porphyrin transport, heme biosynth-esis, anion transport, and regulation of steroidogenesis(Moshe et al., 1992; Pierre et al., 2000).

Mesenchymal stem cells (MSCs) are defined aspluripotent cells capable of replicating extensively andself-maintaining and have demonstrated their differ-entiation into adipocytes, chondrocytes, myoblasts, andosteoblasts (Caplan, 1991; Beresford et al., 1992;Pittenger et al., 1999; Patricia et al., 2001). When MSCsare cultured under the appropriate conditions contain-ing cAMP agonists and induction agents such asisobutyl-methylxanthine (IBMX), indomethacin, insu-lin, and dexamethasone, human mesenchymal stemcells (hMSCs) express several key adipogenic genes in asequence consistent with other models of adipogenesisand develop lipid-containing droplets (Janderova et al.,2003). Therefore, hMSCs provide a unique model tobetter understand early differentiation events becausethey can differentiate into multiple mesenchymal line-ages. Although it has been reported that PBR isexpressed in peripheral tissues such as adrenals,kidney, and heart as well as brain (Davies and Huston,1981; Anholt et al., 1986; Giusti et al., 1994), expressionof PBR and its role has not been examined in MSCs.

This study designated to examine whether hMSCsexpress PBR and to determine the role of PBR onproliferation and differentiation of hMSCs.

MATERIALS AND METHODSReagents

Diazepam, PK11195, and Ro5-4864 were purchasedfrom Sigma Chemical Co. (St. Louis, MO). The PBRligands were dissolved in DMSO. DMSO (at 0.1% as afinal concentration) was added to hMSCs in all controlexperiments.

Cell culture

Human bone marrow stromal cells (hBMSCs) andhuman adipose tissue stromal cells (hATSCs) were ob-tained from primary cultures. The cells were grown inminimum essential medium alpha (a-MEM, Invitrogen,Carlsbad, CA) supplemented with 10% fetal bovineserum (FBS), antibiotics, and antimycoplasma agent at378C in 5% CO2 atmosphere. Jurkat cells were obtain-ed from ATTC. The cells were incubated in RPMI1640 medium (Invitrogen) containing 10% FBS, 2 mML-glutamine, antibiotics, and antimycoplasma agent at378C in 5% CO2 atmosphere.

RT-PCR (reverse transcriptase polymerasechain reaction) and DNA sequencing

Total RNA was isolated from MSCs using RNAsol Bsolution (Tel Test, Inc., Friendswood, TX). Total RNAwas reverse transcribed into first stand cDNA using

oligo dT primer, and PCR reactions were carried out asfollows: PBR was amplified by 30 cycles (948C for 10 sec,588C for 20 sec, and 728C for 30 sec (GeneAmp PCRSystem 2400, Perkin-Elmer Cetus)) using 20 pmoleprimers. The primers for amplification of cDNA weredesigned using the PRIMER program. PBR primers foramplification were: Forward 50-AGGGTCTCCGCTGG-TACGCC-30, Reverse 50-TGGGGCAACCTCTGAAGC-TC-30. The expected size of PBR products by RT-PCRwas 517 bp. The products of the reaction were visualizedon 2% agarose gels by ethidium-bromide staining. ThePBR product were subcloned into PGEM-Teasy vector(Promega, Madison, WI) and sequenced by a Megabasesequencer (Amersharm Pharmacia Biotech Inc., Piscat-away, NJ) using T7 primer.

Immunocytochemistry

Cultured ATSCs and BMSCs were fixed with 4% (v/v)phosphate-buffered paraformaldehyde. Samples wereblocked with 10% nonspecific donkey serum correspond-ing to the species that secondary antibody was gener-ated. AntiPBR primary antibody (R&D Systems,Minneapolis, MN) was diluted to 1:250 in PBS/0.2%Triton X-100/2% bovine serum albumin (BSA), andincubated overnight at 48C. After being rinsed withPBS/0.2% Triton X-100, cultures were incubated for 1 hwith 1:1000 TRITC-conjugated antirabbit secondaryantibodies (Jackson ImmunoResearch Laboratories,West Grove, PA). Fluorescently labeled cells werevisualized using fluorescent microscopy. For a nega-tive control, primary antibodies were replaced withirrelevant IgG.

Induction of differentiation

hMSCs were incubated in a-MEM containing 10%FBS until reaching confluence. After confluence ofhMSCs, the medium was replaced by adipogenic mediathat was DMEM containing 10% FBS, 0.5 mM IBMX,1 mM dexamethasone, 10 mM insulin, 200 mM indo-methacin. The medium was changed three times in aweek. The cell numbers of cells containing lipid dropletsand total cell numbers were measured at 3, 5, 7, and9 day of culture by microscopic counting of one quarter ofeach well in 12-well plate. Measurement was done induplicate at each experiment and experiments wererepeated from three to four times. Because total cellnumbers among different experimental conditions werenot significantly different, data were presented aspercentage of the number of lipid-droplet containingcells in control.

Glycero-3-phosphate dehydrogenase(GPDH) activity

GPDH activity was measured by a spectrophotometricmethod. Cells were scraped into ice-cold sucrose buffercontaining 29 mM sucrose, 73 mM Tris, 24 mM EDTA,and 0.02% b-mercaptoethanol. The cell suspensionswere sonicated for 5 sec at 40 W with microtip of sonifier.Homogenate was centrifuged at 100,000g for 1 h at48C. Assay of GPDH was performed according to themethod of Kozak and Jensen (1974). Enzyme activitywas determined in units per mg of cellular protein, 1 Ubeing equal to the oxidation of 1 nmol NADH/min and

92 LEE ET AL.

expressed as a percentage of control cultures. Theprotein content of lysates was determined in paralleledwells using the Bio-Rad protein assay kit (Bio-Radlaboratory, Hercules, CA).

Evaluation of cell proliferation

To determine the proliferation rate, cells weredetached with HBSS containing 0.5% trypsin and0.02% EDTA. hMSCs were plated at low density(5�103 cells/well) in the presence or absence of variousconcentrations (1, 10, 25, 50, and 75 mM) of diazepam,PK11195, and Ro5-4864, and cell number was countedin a hemocytometer after 48 h. For [3H]-thymidineincorporation assay, cells were seeded as describedabove, and incubated with 10 and 50 mM of PBR ligands.Culture medium was replaced by 500 ml mediumcontaining 0.5 mCi [3H]-thymidine and incubated for4 h. Then, 500ml of 0.5% sodium dodecyl sulfate and 50mlof 72% TCA were added. After centrifugation, the pelletswere solubilized in a solution of 50 ml of 1 N NaOH and9 ml of 60 N HCl. The samples were into vials containing4 ml of scintillation cocktail, and counting was per-formed by a scintillation counter (Tricarb 2100TR,Packard, Meridan, CT).

Cell cycle analysis

Cells (5�105 cells/ 10 ml of medium) were treatedwith PK11195 and diazepam at high concentration(50 mM). Then, the cells were incubated at 378C in 5%CO2 atmosphere for 24 h. Both treated and untreatedcells were harvested with trypsin/EDTA, washed withcold PBS containing 1% BSA. The cells were fixed byaddition of cold 70% ethanol for 4 h at 48C. Prior toanalysis, the cells were washed and resuspended in PBScontaining 1% BSA. After washing, the cells werestained with propidium iodide (500 mg/ml) in thepresence of RNase (1 mg/ml) and incubated at 378C for30 min. The stained cells were observed at 488 nm on aBeckman Dickinson FACScan that was equipped withan air-cooled argon ion laser. Cell cycle was analyzedusing FIT software program (DNA analysis packageavailable from Becton Dickinson).

Measurement of cell viability

Cells were grown in culture media containing variousdoses (25, 50, and 75 mM) of PK11195 in 24-well platesfor 48 h. The cells were trypsinized and placed in cellsuspension in phosphate-buffered saline (PBS). Thecells were stained by 0.4% trypan blue (Sigma) and theproportion of dead cells was measured with a hemocyt-ometer. Cell death was determined by the presence ofcytoplasmic trypan blue. This experiment was per-formed in triplicate.

Statistical analysis

Comparison of multiple means was performed withnonparametric ANOVA. Comparison of individual drugtreatments to control treatments was performed withthe unpaired, 2-tailed Mann–Whitney U-test. Only forcell cycle analysis was the unpaired Student’s t-testused. P-values were considered to be significant at<0.05.

RESULTSPBR expression in hMSCs

To investigate the expression of PBR mRNA inhMSCs, RT-PCR analysis was performed. For the nega-tive control, PCR was performed in the samples thatreverse transcription was done without reverse tran-scriptase. Both in hATSCs and in hBMSCs, theamplified product of the PBR cDNA was detected byethidium-bromide staining, and the size of the PCRproduct was exactly matched with the expected size(Fig. 1B). The negative controls of hMSCs did not showany amplified product, indicating that PCR productswere not derived from contaminated genomic DNA. Thesequences of PCR products determined by DNA sequen-cing were exactly matched with that of the human PBRcDNA sequence (data not shown). Expression of PBRprotein was shown by staining with the polyclonal anti-PBR antibody. Permeabilized hATSCs and hBMSCsdisplayed a PBR-specific fluorescence (Fig. 1B), whereascontrols in which primary antibodies were replaced withirrelevant IgG resulted in no detectable staining.

Effects of PBR ligands on celldifferentiation of hMSCs

To test the effects of PBR ligands on differentiation ofhMSCs, we cultured hMSCs until the reaching con-fluence. After then, hMSCs were cultured in adipogenicmedium with or without diazepam, PK11195 and Ro5-4864 for 14 d. Adipogenic differentiation was examinedby counting the number of cells containing lipid dropletsby a phase contrast microscope, and osteogenic differ-entiation was assessed by alizarin staining. Figure 2shows the effect of PK11195 on adipogenic differentia-tion in hATSCs. The low concentration of PK11195(1 mM) increased number of cells containing lipid drop-lets, whereas 50 mM PK11195 decreased it (Fig. 2A). Toquantitate the effect of PBR ligands on adipose differ-entiation, the number of cells containing lipid dropletswas counted and presented as the percentage of controldata. The PBR ligands showed biphasic effects on adipo-genic differentiation of hMSCs (Fig. 2B). In hATSCs andhBMSCs, PBR ligands to be tested (diazepam, PK11195,and Ro5-4864) increased adipose differentiation in lowconcentration range (from 0.1 to 10 mM), whereas theydecreased it at high concentration (50 mM) comparedwith control. Potency of PBR ligands on adipogenicdifferentiation is as follows; diazepam>PK11195>Ro5-4864. To clarify the PBR specificity of the effects,we determined the effect of clonazepam, a structuralanalog of diazepam, of which displays no affinity to thePBR. Clonazepam did not affect adipose differentiationin hMSCs.

To investigate whether the effect of PBR ligands onadipogenic differentiation in hMSCs is time-dependant,adipogenic differentiation of hMSCs treated by differentconcentrations of PK11195 was determined at 3, 5, 7,and 9 day after the treatment. The effect of adiposedifferentiation in hATSCs by PK11195 was not alteredduring the period of the treatment (Fig. 3). Diazepamand Ro5-4864 showed similar results (data not shown).

Effect of PBR ligands on GPDH activity was deter-mined to further examine the effect of PBR ligands onadipogenic differentiation in hATSCs. Because GPDH

PERIPHERAL BENZODIAZEPINE RECEPTORS IN HUMAN MESENCHYMAL STEM CELLS 93

Fig

.1.

A:E

xp

ress

ion

ofp

erip

her

al

ben

zod

iaze

pin

ere

cep

tor

(PB

R)

mR

NA

inh

um

an

mes

ench

ym

al

stem

cell

s.R

T-P

CR

usi

ng

PB

Rsp

ecifi

cp

rim

ers

an

aly

zed

exp

ress

ion

ofP

BR

mR

NA

.E

xp

ecte

dsi

zeof

PC

Rp

rod

uct

s(5

17

bp

)w

ere

am

pli

fied

inh

um

an

bon

em

arr

owst

rom

al

cell

s(h

BM

SC

s)(l

an

e1)

an

din

hu

man

ad

ipos

est

rom

al

cell

s(h

AT

SC

s)(l

an

e2).

Tot

al

RN

As

from

hB

MS

Cs

(lan

e3)

an

din

hA

TS

Cs

(lan

e4)

wer

ep

erfo

rmed

PC

Rw

ith

out

rever

setr

an

scri

pti

onas

neg

ati

ve

con

trol

.l:

Mol

ecu

lar

wei

gh

tm

ark

erP

GE

M.B

:F

ig.

Imm

un

ocyto

chem

istr

yof

hM

SC

s.C

ult

ure

hA

TS

Cs

an

dh

BM

SC

sw

ere

imm

un

ocyto

chem

icall

yst

ain

edw

ith

an

ti-P

BR

an

tibod

y(r

ed-c

olor

).S

cale

dbar:

20mm

.Rep

rese

nti

ati

ve

fin

din

gs

out

ofth

ree

MS

Cs

are

show

n.

For

vis

uali

zati

onof

sign

al,

TR

ITC

-con

jugate

dan

ti-r

abbit

seco

nd

ary

an

tibod

yw

as

use

d.

Con

trol

s(C

ont)

inw

hic

hp

rim

ary

an

tibod

ies

wer

ere

pla

ced

wit

hir

rele

van

tIg

Gre

sult

edin

no

det

ecta

ble

stain

ing.

94 LEE ET AL.

activity is high in mature adipocytes, it has beenconsidered as a marker of adipose differentiation. Todetermine the GPDH activity, cell lysates were obtainedfrom hATSCs treated with PBR ligands on 7 day.Specific PBR ligands increased GPDH activity in lowconcentration range (from 0.1 to 10 mM), whereas theydecreased GPDH activity at high concentration (50 mM)(Fig. 4).

Effect of PBR ligands on cell proliferationin hMSCs

To determine if PBR ligands affect hMSCs prolifera-tion, changes in cell number of hATSCs and hBMSCswere determined by direct cell counting. The treatmentof PBR ligands for 48 h significantly inhibited prolifera-tion of hATSCs from 25 mM in a dose-dependent manner(Fig. 5A). The results in hBMSCs were similar to that inhATSCs (data not shown). To further determine effectsof PBR ligands on proliferation of hMSCs, the effectof PBR ligands on [3H]-thymidine incorporation wasdetermined in hATSCs. PBR ligands inhibited [3H]-

thymidine incorporation and the potency of the PBRligands to be examined in this study was similar withresults that obtained by cell counting (Fig. 5B).

Induction of cell cycle arrest by PBRligands in hMSCs

To determine the alterations of cell cycle regulation orinduction of apoptosis by PBR ligands, we analyzed cellcycle using a fluorescence-activated cell sorter (FACS).In control experiments, 77.2% of hATSCs was in the G0/G1 phase, 14% in the S phase, and 8.8% in the G2/Mphase. PK11195 (50 mM) and Ro5-4864 (50 mM)increased the proportion of cells in the G2/M phase anddecrease the proportion of cells in the S phase whencompared with control in hATSCs (P< 0.05) (Table 1).

Effect of PBR ligands on cell viabilityin hMSCs and Jurkat cells

To determine whether PBR ligands affect cell viabilityin hATSCs and Jurkat cells, the various concentrationsof PBR ligands were treated in hATSCs and Jurkat cells

Fig. 2. Effect of PK11195 on adipogenic differentiation of hATSCs.A: hATSCs were grown to confluence and then induced adipogenesisby differentiation medium. Photomicrograph of differentiated hATSCswas taken on 14 days after the treatment of PK11195 in adipogenicmedium. Scaled bar: 20 mm a, Control; b, hATSCs treated with 1 mMPK11195; c, hBMSCs treated with 50 mM PK11195. B: Effect of PBRligands on the adipogenic differentiation of hATSCs (a) and hBMSCs(b). After confluence hATSCs and hBMSCs were cultured in adipo-

genic medium including various PBR ligands (0.1, 1, 10, and 50 mM)for 9 days. The adipose differentiation was determined by direct countof cells containing lipid droplet under the microscopy. The number ofdifferentiated cells was determined in duplicate at the day 9 afterinduction of differentiation. Data were expressed as percentage ofcontrol (a mean�SEM of three different experiments). *P< 0.05compared with control (Cont).

PERIPHERAL BENZODIAZEPINE RECEPTORS IN HUMAN MESENCHYMAL STEM CELLS 95

for 72 h, and then cell viability was assessed by trypanblue dye exclusion assay (Fig. 6). PBR ligands did notinduce morphological changes, and did not affect cellviability in hMSCs, although PBR ligands did inducecell death in Jurkat cells.

DISCUSSION

PBRs were described as proteins that are expressed inperipheral tissues such as adrenals, kidney, heart, andbrain as well as various cells including lymphocytes,glioma, and thymoma cells (Stepin et al., 1981; Ikezakiand Black, 1990). PBRs are mainly located on the outermembrane of the mitochondria and have variouspharmacological, biochemical, pathological, and physio-logical roles (Moshe et al., 1999). However, expression ofPBRs and their function in hMSCs have not beenexamined. In this study, we demonstrated that PBR isexpressed in hMSCs by RT-PCR, and that their ligandsaffect their proliferation and differentiation.

There is a difference of chemical structure betweenPK11195 and Ro5-4864 (Pierre et al., 2000). Based onthe difference of structure, it has been considered thatPK11195 is considered as an antagonist of PBR and Ro5-4864 is as its agonist (Syapin and Skolnic, 1979; Le Furet al., 1983). In fact, PK11195 inhibited the effect ofdiazepam and Ro5-4864 in various cells (Migiwa andMitsuru, 1998; Beinlich et al., 1999). However, otherstudies reported that PK11195 and Ro5-4864 showedsimilar effects in B16 melanoma cells and humancolorectal cancer cells (Laudau et al., 1998; Maaseret al., 2001). Our study showed that PK11195 and Ro5-4864 has similar effects on the proliferation anddifferentiation of hMSCs. That is, PK11195 works asan agonist in hMSCs, not as an antagonist. The data inthis study and other reports indicate that the concept ofantagonists and agonists in PBR ligands should berevised.

Previous studies indicated that PBR ligands inhibitproliferation at micromolar concentration and stimu-lated its nanomolar concentration range in various cells(Stepin et al., 1981; Ikezaki and Black, 1990; Laudauet al., 1998). In our study, PBR ligands also exhibitedbiphasic effect on differentiation of hMSCs, but theeffective dosage of PBR ligands was different. Cellproliferation and cell differentiation of hMSCs were in-hibited by PBR ligands at high concentration (>10 mM).In contrast, PBR ligands increased cell differentiation ofhMSCs in low conceration range (from 0.1 to 10 mM). Inthis study, we could not find any significant stimulatoryeffect of PBR ligands on proliferation of hMSCs. We alsoevaluated potency of PBR ligands on proliferation anddifferentiation of hMSCs. Potency of PBR ligands inadipose differentiation and proliferation is as follow;diazepam>PK11195>Ro5-4864. The PBR ligandsselectively active at the PBRs exhibited different potentantiproliferative activity in different cell types (Laudauet al., 1998; Beinlich et al., 1999; Maaser et al., 2001).The inhibitory potency was PK11195>Ro5-4864>diazepam in B16 melanoma cells, whereas inhibitorypotency was Ro5-4864>PK11195>diazepam in BT 20cells. PBR ligands induce apoptotic cell death in somecells (Marchetti et al., 1996a; Tanimoto et al., 1999;Fischer et al., 2001; Decaudin et al., 2002; Sutter et al.,2002), whereas they prevent cell death in other cell types

Fig. 3. Effect of duration of PK11195 treatment on the adipogenesisof hATSCs. PBR ligands of various concentrations were treated intohATSCs in adipogenic medium. The adipose differentiation wasdetermined by direct counting of cells containing lipid droplets undermicroscopy. On day 3, 5, 7, and 9 the cells containing lipid dropletswere counted in duplicate for determination of adipose differentiation.Data were expressed as a percentage of control value at thecorresponding day (a mean�SEM from three different experiments).

Fig. 4. Effect of PBR ligands on glycerol-3-phosphate dehydrogenase(GPDH) activity. After confluence of hATSCs, PBR ligands of variousconcentrations were added to adipogenic medium for 9 days. Mediumwas changed every other day. Cell lysates were obtained from hATSCstreated with PBR ligands on day 9 for the determination of GPDHactivity. GPDH activity was measured as described in method. Datawere expressed as percentage of control culture (a mean�SEM fromfour different experiments). *P< 0.05 compared with control (Cont).

96 LEE ET AL.

(Carayon et al., 1996; Bono et al., 1999; Stoebner et al.,2001). In the case of hMSCs, any concentration of PBRligands to be determined in the study did not affect cellsurvival, as determined by trypan blue exclusion. Theabove data indicated that the effects and PBR ligands oncellular functions and their effective dosages andaffinities to PBR are variable according to types of cells.

PBR ligands have been reported to induce cell cyclearrest in various cells. A PBR-ligand-induced G1/Sarrest was found for colorectal cancer cells (Maaseret al., 2001) and esophageal cancer cells (Sutter et al.,2002). In breast cancer, however, PBR ligands induced acell cycle arrest at both major restriction points, the G1/S- and the G2/M-junction (Carmel et al., 1999), whereasin lung and melanoma cells an accumulation in the G2/Mphase was observed (Laudau et al., 1998). The cellcycle analysis in this study showed that PK11195(50 mM) and Ro5-5864 (50 mM) induced G2/M arrest in

hMSCs. These differences may reflect tissue-specificPBR signal transduction.

The data in this study showed that proliferation anddifferentiation of hMSCs were reduced significantly byPBR ligands at high concentration. However, the effectsof high concentration of PBR ligands on proliferationand adipose differentiation could not be explained bynonspecific inhibition of cellular functions, because PBRligands did not induce cell death in hMSCs andclonazepam did not affect proliferation and adiposedifferentiation in hATSCs.

Adipogenesis occurs in both the prenatal and post-natal states in humans; in rodents, most fat celldevelopment occurs postnatally. While older literaturesuggested that people are born with all the adipocytesthey will possess in life, there is now convincing evidencethat adipogenesis occurs throughout the lifetime of anorganism. This adipogenesis occurs both as a conse-quence of normal cell turnover, and as a consequence ofthe requirement for additional fat mass that arises withsignificant calorie storage and weight gain (Prins andO’Rahilly, 1997). More convincing evidence has comefrom studies of rats fed a high-calorie diet, where [3H]-thymidine incorporation into new fat cells occursthroughout adulthood (Miller et al., 1984). Whetherthis is true for humans is unknown, but it is clear thatpre-adipocytes purified from the fat pads of even elderlypeople can be induced to differentiate in vitro (Deslexet al., 1987; Kirkland et al., 1990; Entenmann andHauner, 1996). This observation argues for a role for

Fig. 5. Effect of PBR ligands of the proliferation on hMSCs. A: Afterlow density seeding of hBMSCs and hATSCS, cells were grown invarious concentrations of PBR ligands. And proliferation of hATSCswas determined in duplicate by direct cell counting using ahemocytometer at the indicated days after seeding. Data wereexpressed as a percentage of the value in the absence of PBR ligand(a mean�SEM from four experiments). *P< 0.05 compared with the

data in the absence of PBR ligands. B: Effect of PBR ligands on [3H]-thymidine incorporation in hATSCs. hATSCs were incubated until thereaching subconfluence. After 0.5 mCi [3H]-thymidine was treated intoeach sample, the cells were incubated for 4 h and measured by liquidscintillation counter. Data were presented as a percentage of the valuein the absence of PBR ligands (a mean�SEM from three differentexperiments). *P< 0.05 compared with control.

TABLE 1. Effect of peripheral benzodiazepine receptor (PBR) ligandson the cell cycle of human adipose stromal cells (hATSCs)

PBR ligands G0/G1 S G2/M

Control 77.2�6.5 14.0� 1.7 8.8�0.6Ro-5-4864 (50 mM) 79.0�5.4 7.3� 0.3* 13.7� 1.7*PK11195 (50 mM) 78.6�5.7 3.6� 0.2* 17.8� 1.2*

Cell cycle was determined by flow cytometry at 24 h after the treatment of PBRligands. The data are given in percentage of cells as mean�SEM of threedifferent experiments.*P< 0.05 versus control of the same cell cycle phase.

PERIPHERAL BENZODIAZEPINE RECEPTORS IN HUMAN MESENCHYMAL STEM CELLS 97

both adipocyte hypertrophy and hyperplasia in humanobesity. The present data that PBR agonists increaseadipose differentiation may suggest that PBR-inducedadipose differentiation besides of modulation of GABA-induced neurotransmission in the CNS by benzodiaze-pines may be involved in the development of benzodia-zepine-induced obesity.

In summary, we established a model for the study ofthe role of PBR on proliferation and differentiation inhMSCs. In this model, we showed the effects of PBRligands on proliferation and differentiation in hMSCs.However, mechanism of PBR on proliferation anddifferentiation of hMSCs have not been clarified.Further investigations are required to elucidatemechanism of PBRs on proliferation and differentiationin hMSCs.

LITERATURE CITED

Anholt RRH, Pedersen PL, De Souza EB, Snyder SH. 1986. Theperipheral-type benzodiazepine receptor: Localization to the mito-chondrial outer membrane. J Biol Chem 261:576–583.

Beinlich A, Strohmeier R, Kaufmann M, Kuhl H. 1999. Specificbinding of benzodiazepines to human breats cancer cell lines. LifeSci 65:2099–2108.

Beresford JN, Bennett JH, Devlin C, Leboy PS, Owen ME. 1992.Evidence for an inverse relationship between the differentiation ofadipocytic and osteogenic cells in rat marrow stromal cell cultures.J Cell Sci 102:341–351.

Bono F, Lamarche I, Prabonnaud V, Le Fur G, Hervert JM. 1999.Peripheral benzodiazepine receptor agonists exhibit potent anti-apoptoic activities. Biochem Biolophys Res Commun 265:457–461.

Bormann J. 1988. Electrophysiology of GABAA and GABAB receptorsubtypes. Trends Neurosci 11:112–116.

Braestrup C, Squires RF. 1977. Specific benzodiazepine receptors inrat brain characterized by high-affinity [3H] diazepam binding. ProcNatl Acad Sci USA 74:3805–3809.

Caplan AI. 1991. Mesenchymal stem cells. J Orthop Res 9:641–650.Carayon P, Portier M, Dussossoy D, Bord A, Petitpretre G, Canat X,

Le Fur G, Casellas P. 1996. Involvement of peripheral benzodiaze-pine receptors in the protection of hematopoietic cells againstoxygen radical damage. Blood 87:3170–3178.

Carmel I, Fares FA, Leshiner S, Scherubl H, Weisinger G, Gavish M.1999. Peripheral-type benzodiazepine receptors in the regulation ofproliferation of MCF-7 human breast carcinoma cell line. BiochemPharmacol 58:273–278.

Clarke GD, Ryan PJ. 1980. Tranquilizers can block mitogenesis in 3T3cells and induce differentiation in Friend cells. Nature 287:160–161.

Davies LP, Huston V. 1981. Peripheral benzodiazepine receptorsbinding sites in heart and their interaction with dipyridamole. Eur JPharmacol 73:209–211.

De Lorey TM, Olsen RW. 1992. g-Aminobutyric acidA receptorstructure and function. J Biol Chem 267:16747–16750.

Decaudin D, Castedo M, Nemati F, Beurdeley-Thomas A, De PinieuxG, Caron A, Pouillart P, Wijdenes J, Rouillard D, Kroemer G,Poupon MF. 2002. Peripheral benzodiazepine receptor ligandsreverse apoptosis resistance of cancer cells in vitro and in vivo.Cancer Res 62:1388–1393.

Deslex S, Negrel R, Vannier C, et al. 1987. Differentiation of humanadipocyte precursors in a chemically defined serum-free medium.Int J Obes 11:19–27.

Entenmann G, Hauner H. 1996. Relationship between replication anddifferentiation in cultured human adipocyte precursor cells. Am JPhysiol 270:C1011–C1106.

Fischer R, Schmitt M, Bode JG, Haussinger D. 2001. Expression of theperipheral-type benzodiazepine receptor and apoptosis induction inhepatic stellate cells. Gastroenterology 120:1212–1226.

Giusti L, Trincavelli L, Martini C, Lucachini A. 1994. Characteriza-tion peripheral-type benzodiazepine binding sites from rat and pigpancreas. Biochem Pharmacol 48:583–586.

Ikezaki K, Black KL. 1990. Stimulation of cell growth and DNAsynthesis by peripheral benzodiazepines. Cancer Lett 49:115–120.

Janderova L, Mcneil M, Murrell AN, Mynatt RL, Smith SR. 2003.Human mesenchymal stem cells as an in vitro model for humanadipogenesis. Obes Res 11:65–74.

Kirkland JL, Hollenberg CH, Gillon WS. 1990. Age, anatomic site, andthe replication and differentiation of adipocyte precursors. Am JPhysiol 258:C206–C210.

Kozak LP, Jensen JT. 1974. Genetic and developmental control ofmultiple forms of L-glycerol 3-phosphate dehydrogenase. J BiolChem 249:7775–7781.

Laudau M, Weizman A, Zoref-Shani E, Beery E, Wassenman L,Landau O, Gavish M, Brenner S, Nordenverg J. 1998. Antiproli-ferative and differentiating effects of benxodiazepine receptorligands on B16 melanoma cells. Biochem Pharmacol 56:1029–1034.

Le Fur G, Perrier ML, Vaucher N, Imbault F, Flamier A, Benavides J,Uzan A, Renault C, Dubroeucq MC, Gueremy C. 1983. Peripheralbenzodiazepine binding sites: Effect of PK11195, 1-(2-chlorophe-nyl)-N-(1-methylpropyl)-3-isoquinoline carboxzmide. Life Sci 32:1839–1847.

Maaser K, Hopfner M, Jansen A, Weisinger G, Gavish M, KozikowskAP, Weizman A, Carayon P, Riecken E-O, Zeiz M, Scherubl H. 2001.Specific ligands of the peripheral benzodiazepine receptor induceapoptosis and cell cycle arrest in human colorectal cancer cells. Br JCancer 85:1771–1780.

Marchetti P, Hirsch T, Zamzami N, Castedo M, Decaudin D, Susin SA,Masse B, Kroemer G. 1996a. Mitochondiral permeability transitiontriggers lymphocyte apoptosis. J Immunol 157:4830–4836.

Matthew E, Laskin JD, Zimmerman EA, Weinstein IG, Hsu KD,Englehardt DL. 1981. Benzodiazepines have high-affinity bindingsites and incuce melanogenesis in B16/C3 melanoma cells. Proc NatlAcad Sci USA 78:3935–3939.

Migiwa O, Mitsuru K. 1998. Inhibitory reculation of amylase releasein rat parotid acinar cells by benzodiazepine receptors. Eur JPharmacol 359:243–249.

Miller WH, Jr., Faust IM, Hirsch J. 1984. Demonstration of de novoproduction of adipocytes in adult rats by biochemical and radio-autographic techniques. J Lipid Res 25:336–347.

Moshe G, Yeshayahu K, Shalom BA, Ronit W. 1992. Biochemical,physiological, and pathological aspects of the peripheral benzodia-zepine receptor. J Neurochem 58:1589–1601.

Moshe G, Idit B, Rami S, Yeshayahu K, Leo V, Gary W, Abraham W.1999. Enigma of the peripheral benzodiazepine receptor. PharmacolRev 51:629–650.

Fig. 6. Effect of PK11195 on the cell viability of hATSCs and Jurkatcells. The PBR ligands of various concentrations were added into theculture medium for 48 h. Cell viability was measured with trypan blueexclusion. PK11195 did not induce cell death in hBMSCs, but inducedcell death in Jurkat cell. Data represent mean of triplicate experi-ments. The control contained DMSO at a concentration equivalent tothat introduced with PBR ligands and the final concentration ofDMSO was 0.1%. Data represent percentage of viable cells (amean�SEM from four different experiments). *P<0.05 comparedwith the data in the absence of PK11195.

98 LEE ET AL.

Mohler H, Okad T. 1977. Benzodiazepine receptor: Demonstration inthe central nervous system. Science 198:849–851.

Patricia AZ, Min Z, Hiroshi M, Jerry H, William FJ, Adam JK, ProsperB, Peter LH, Hedrick MH. 2001. Multilineage cells from humanadipose tissue: Implications for cell-based therapies. Tissue Eng7:211–227.

Pawlikowski M, Pawlikowska A, Stepien H. 1987. Effects ofbenzodiazepines on anterior pituitary cell proliferation. Neuroen-docrinol Lett 9:43–49.

Pawlikowski M, Lyson K, Kunert-Radek J, Stepien H. 1988. Effect ofbenzodiazepines on the proliferation of mouse spleen lymphocytesin vitro. J Neural Transm 73:161–166.

Pierre C, Sylvaine G, Anthony SB. 2000. Peripheral benzodiazepinereceptors and mitochondrial function. Neurochem Int 40:475–486.

Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, Douglas R, MoscaJD, Moorman MA, Simonetti DW, Craig S, Marshak DR. 1999.Multilineage potential of adult human mesenchymal stem cells.Science 284:143–147.

Prins JB, O’Rahilly S. 1997. Regulation of adipose cell number in man.Clin Sci (London) 92:3–11.

Schoemaker H, Boles RG, Horst G, Yamamura HI. 1983. Specific high-affinity binding sites for [3H] Ro 5-4864 in rat brain and kidney.J Pharmacol Exp Ther 225:61–69.

Skolnick P, Paul SM. 1982. Benzodiazepine receptors in the centralnervous system. Int Rev Neurobiol 23:103–140.

Stepin H, Pawlikowski A, Pawlikowski M. 1981. Effects of benzodia-zepines on thymus cell proliferation. Thymus 12:3935–3939.

Stoebner PE, Carayon P, Casellas P, Portier M, Lavabre-Bertrand T,Cuq P, Cano JP, Meynadier J, Meunier L. 2001. Transientprotection by peripheral benzodiazepine receptors during the earlyevents of ultraviolet light-induced apoptosis. Cell Death Differ 8:747–753.

Sutter AP, Maaser K, Hopfner M, Barthel B, Grabowski P, Faiss S,Carayon P, Zeitz M, Scherubl H. 2002. Specific ligands of theperipheral benzodiazepine receptor induce apoptosis and cell cyclearrest in human esophageal cancer cells. Int J Cancer 102:318–327.

Syapin PJ, Skolnic P. 1979. Characterization of benzodiazepinebinding sites in cultured cells of neural origin. J Neurochem32:1047–1051.

Tallman JF, Paul SM, Skolnick P, Gallager DW. 1980. Receptors forthe age of anxiety: Pharmacology of benzodiazepines. Science 207:274–281.

Tanimoto Y, Onishi Y, Sato Y, Kizaki H. 1999. Benzodiazepinereceptor agionist modulate thymocyte apoptosis through reductionof the mitochondrial transmembrane potential. Jpn J Pharmacol 79:177–183.

PERIPHERAL BENZODIAZEPINE RECEPTORS IN HUMAN MESENCHYMAL STEM CELLS 99