Vol. 6, 647-653, June 1995 Cell Growth & Differentiation 647

3 The abbreviations used are: IL-6, interleukin 6; GM-CSF, granulocyte-macrophage colony-stimulating factor; RT-PCR, reverse transcription PCR.

Regulation of bcl-2, bcl-X1 and bax in the Control ofApoptosis by Hematopoietic Cytokinesand Dexamethasone’

Joseph Lotem and Leo Sachs2

Department of Molecular Genetics and Virology, Weizmann Institute of

Scfence, Rehovot 76100, Israel

Abstract

Treatment of Ml myeloid leukemic cells with interleukin6 (11-6) or dexamethasone (DEX), both of which inducedifferentiation in these cells, down-regulated expressionof the apoptosis-suppressing gene bcl-2 and theapoptosis-promoting gene bax but up-regulatedexpression of the apoptosis-suppressing gene bcI-X�.There was a higher expression of bcI-X� in cells treatedwith DEX or DEX plus IL-6 compared to cells treatedwith 11-6 alone. The alternatively spliced bcl-X gene,bcl-X5, which interferes with the adion of bcl-2, was notexpressed. Treatment with 11-6 increased thesusceptibility of these cells to indudion of apoptosis byAdriamycin or cycloheximide, but treatment with DEXor with 11-6 and DEX did not. Withdrawal of DEX afterup-regulation of bcI-X� resulted in a decrease in bcI-X�expression and a concomitant increase in cellsusceptibility to indudion of apoptosis. Another myeloidleukemia that shows barely detedable expression of bcl-2 also showed up-regulated expression of bcl-XL but nochange in bax after indudion of differentiation withgranulocyte-macrophage colony-stimulating fador, andthis reduced cell susceptibility to indudion of apoptosisby Adriamycin or cycloheximide. The results indicatethat the related apoptosis-regulating genes bcl-2, bcl-XL,and bax are differently regulated and that up-regulationof bcl-XL expression may compensate for down-regulation of bcl-2 in the balance between genes thatcontrol apoptosis.

IntroductionThe program of cell death by apoptosis can be positivelyand negatively regulated by the products of different genes.The gene bcl-2 can suppress induction of apoptosis byvarious cytotoxic treatments (1 ). Some bc/-2-related genes,such as bc/-XL (2, 3), ced-9 in Caenorhabditis e/egans (4),and BHRF-i in Epstein-Barr virus (5), can exert a bc/-2-likeantiapoptotic effect, whereas other bc/-2-related genes,such as an alternatively spliced bcl-X transcript bcl-X5 (2)and the gene bax (6), can antagonize the action of bcl-2 andthus promote apoptosis. There are also other genes thatpromote apoptosis, including wild-type p53 (7-1 0) andc-myc (1 1 -1 3), or that suppress apoptosis, including mutant

Received 10/1/94; revised 3/2/95; accepted 3/17/95.

1 This work was supported by the National Foundation for Cancer Research(Bethesda, MD), the Hermann de Stern Foundation, and the Ebner FamilyBiomedical Research Foundation at the Weizmann Institute of Science inmemory of Alfred and Dolfi Ebner.2 To whom requests for reprints should be addressed.

p53 (1 3), which are not related to bc/-2 (reviewed in Refs.14-16).

It has been shown that the level of bcl-2 expression indifferent myeloid leukemias was associated with the degreeof resistance to induction of apoptosis by cancer chemo-therapy and some other agents (1 3) and that there was abetter response to chemotherapy in leukemia patientswhose leukemic cells expressed a low level of bcl-2 (1 7).Furthermore, down-regulation of bcl-2 expression in Mlmyeloid leukemic cells by treatment with a cytokine such aslL-63 gave an increase in cell susceptibility to induction ofapoptosis by cancer chemotherapy and some other agents(1 8). However, down-regulation of bcl-2 expression by DEXdid not increase cell susceptibility to the induction of ap-optosis, and this steroid suppressed the apoptosis-sensitiz-ing effect of IL-6. This suppression of susceptibility to in-duction of apoptosis by DEX suggests that DEX may activatea bc/-2-independent pathway that can protect cells from theinduction of apoptosis. We show that induction of differ-entiation in two different myeloid leukemias up-regulatedexpression of the apoptosis-suppressing gene, bc/-XL. Thepattern of expression of bc/-2, bcl-X1, and the apoptosis-promoting gene bax in these leukemias indicate that thesegenes can be differently regulated. The results also suggestthat up-regulation of bc/-XL may compensate for down-regulation of bc/-2 in suppressing induction of apoptosis.

Results

Expression of the Apoptosis-suppressing Gene bcI-X� inMyeloid Leukemic Cells. The bc/-2-related gene bcl-X cancode for two proteins, BCL-X1 and BCL-X5, by alternativelyspliced transcripts in the coding region (2). The bcl-X1transcript codes for a protein that is even more effective insuppressing apoptosis than BCL-2 (2, 3), whereas the bcI-X�transcript codes for a protein that antagonizes BCL-2 (2).lL-6 and DEX can induce differentiation in Ml myeloidleukemic cells (Table 1). Differentiation with lL-6 or DEXwas associated with down-regulated expression of bc/-2(Ref. 18; Fig. 1). lL-6 and DEX also down-regulated c-mycand lL-6 but not DEX up-regulated expression of c-jun andjun-B (Fig. 1). Probing the Northern blots of Ml cells withbcl-X cDNA showed a low level expression of an -2.7-kbtranscript in untreated cells, which was increased with timein lL-6- and DEX-treated cells (Fig. 2). At 3 days after theaddition of IL-6 or DEX, bcl-X expression in different ex-periments was increased 2-3-fold and 4.5-6-fold, respec-tively (Fig. 2), and about 9-fold with combined treatment ofIL-6 and DEX (Fig. 3). Since the difference in size betweenbcl-XL and bcl-X5 transcripts is too small to distinguish on aNorthern blot (2), we used RT-PCR to amplify the codingregion of the bcl-X transcripts. Untreated cells or cells

xw

wC )< (0 (0

c-jun

jun-B

648 bcl-2, bcl-X1 , bax and Apoptosis

Table 1 Induction of d

GM-CSE, and DEX

ifferentiation in myeloid le ukemic cells by lL-6,

.Leukemia

MaterialsI

added’

DifferentiatedI,

cells ( /o(

Ml clone 12 None

tL-6

GM-CSE

DEX

IL-6+DEX

GM-CSE + DEX

1 ± 1

71±8

1 ± 1

50±6

97±3

47 ± 7

7-M12 None

IL-6

GM-CSE

DEX

IL-6#{247}DEX

GM-CSE + DEX

1 ± 1

1±1

53±6

10±3

11±3

52 ± 7

.1 Mi clone 1 2 and 7-Mi 2 leukemic cells were cultured for 2 days with no

addition (none), 1 0 ns/mL tL-6, 1 0 ng/mI GM-CSE, 1 �M DEX, lL-6 and DEX,or GM-CSE and DEX.1, Differentiated cells were determined as described in “Materials andMethods.”

treated with IL-6, DEX, or both IL-6 and DEX expressed theamplified coding region of bcl-XL (0.8 kb) but not bc/-X5(0.6 kb; Fig. 4). The results indicate that, unlike expressionof the apoptosis-suppressing gene bc/-2 which is down-regulated after treatment with IL-6 or DEX, the expression ofthe apoptosis-suppressing gene bc/-XL is up-regulated bythese compounds (Figs. 2 and 3).

To determ i ne whether differentiation-associated up-reg-ulation of expression of bcl-XL also occurs in other myeloidleukemic cells, we used clone 7-Mi2 cells derived fromanother myeloid leukemia that can be induced to differen-tiate with GM-CSF but not with IL-6 (19) (Table 1). Un-treated 7-Mi2 leukemic cells show barely detectable ex-

pression of bcl-2 (13) and bcI-XL (Fig. 5). The resultsindicate that induction of differentiation in 7-Mi 2 cells withGM-CSF was also associated with up-regulation of bc/-XLexpression (i9-fold over the control cells), and bc/-X� wasnot expressed (Fig. 5). DEX, which induces differentiationonly weakly in 7-Mi 2 cells (Table i ),induced only a slightincrease in bc/-XL expression in 7-Mi 2 cells that was muchlower than that induced by GM-CSF (Fig. 5). DEX did notchange the level of bcI-XL expression induced by GM-CSF(Fig. 5).

Expression of the Apoptosis-promoting Gene bax in My-eloid Leukemic Cells. The BCL-2 protein heterodimerizeswith the related protein BAX, and this heterodimerizationappears to be required for antiapoptotic activity (6, i6).When BAX is overexpressed, BAX-BAX homodimers ratherthan BAX/BCL-2 heterodimers are formed, and these cellsdisplay an increased susceptibility to apoptosis (6, i 6). Thesusceptibility of cells to induction of apoptosis may, there-fore, be regulated, at least in part, by the ratio of bax/bcl-2.Therefore, we determined the expression of bax in Ml and7-Mi2 myeloid leukemic cells before and after treatmentwith IL-6, DEX, or GM-CSF. The results indicate that bothMi and 7-Mi2 myeloid leukemic cells expressed tworeadily detectable bax transcripts (1 and i .5 kb; Figs. 2and 5). As with the expression of bcl-2, both bax tran-scripts were down-regulated by the induction of differ-entiation of Mi cells with lL-6 or DEX (Fig. 2; 40%decrease of the 1 .5 kb and 60% decrease of the i kb).

c-myc

bcl-2

1�-actin

Fig. 1. Expression of bcl-2, c-myc, jun-B, and c-jun in Mi myeloid leuke-

mic cells (Northern blot). Cells were cultured for 2 days either with noadditions (none) or treated with 1 �M DEX, 10 ng/ml IL-6, or lL-6 plus DEX.

However, there was no down-regulation of bax expres-sion in 7-M12 cells by induction of differentiation with

GM-CSF or DEX (Fig. 5).Susceptibility to Induction of Apoptosis in Cells Express-

ing Different Levels of bcl-2, bcl-X1, and bax. The un-treated Mi and 7-Mi 2 myeloid leukemic cells expressedlittle or no bcl-XL (Figs. 2 and 5). However, 7-M12 cellsexpressed a barely detectable level of bc/-2 (1 3) and areadily detectable level of bax (Fig. 5), whereas Mi cells

expressed readily detectable levels of both bcl-2 (1 3) andbax (Fig. 2). This indicates that the ratio of bax/bc/-2 ex-pressed in untreated 7-M12 cells is much higher than inuntreated Mi cells, and this can explain the higher suscep-

lL-6 Dexamethasone

0 .5H 1H 3H 6H 1D 2D 3D .5H 1H 3H 6H 1D 2D 3D

bcl-XL , 4

. ,“ � I�.

bax

�-actin � �

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Dexamethasone IL-6

0 2H 6H 24H 0 2H 6H

#{149}f �‘

‘.5. �

IL-6+Dex

0 2H 6H : Time after withdrawal

bcl- XL

bax

bcl- 2

Cell Growth & Differentiation 649

Fig. 2. Time course of bcl-X,and b,ts expression in IL-6- andDEX-treated Ml myeloid leuke-mic cells (Northern blot). Cells

were cultured with no additions(Lane 1 1 or with 1 0 ngJml IL-6(Lanes .?-�8( or 1 psi DEX (Lanes9-15), and RNA was isolated atdifferent times after adding lL-6and DEX. In the cells treated with

IL-h, the thicker band of hcl-XL at6 h is due to the higher amount of

loaded RNA as shown by the

thicker �-actin band.

Time after addition of

�- � - I�- actin

Fig. 3. Reversibility of bcl-X,,

bax, and hcl-2 expression in Mlmyeloid Ieukemic cells (Northern

blot). Ml cells were cultured for 3days with 1 �M DEX, 10 ng/ml

tL-6, or IL-h plus DEX. The cellswere then washed and cultured

without inducers. The level of hcl-XL , bax, and hcl-2 expressed was

determined immediately afterwithdrawal of the inducers (0 h(

and at later times.

r� r��w � w �

tibility of 7-M12 cells to the induction of apoptosis bydifferent agents (i 3), including Adriamycin and cyclohexi-mide (Tables 2 and 3).

Treatment of Mi cells with IL-6, which resulted in down-regulation of bc/-2 and bax expression, was associated withan increased susceptibility to induction of apoptosis byAdriamycin and cycloheximide (Table 2; Fig. 68). Therewas no apoptosis in untreated cells or in cells treated for 2days with IL-6, DEX, or both without inducing apoptosis byadding cycloheximide or Adriamycin (Table 2; Fig. 68).Although there was a similar down-regulation of bc/-2 andbax expression by treatment of Mi cells with IL-6, DEX, orDEX together with IL-6, unlike treatment with lL-6 alone,these other treatments did not increase cell susceptibility tothe induction of apoptosis (Table 2; Fig. 68). This lack ofincreased susceptibility to apoptosis in DEX or in IL-6 plusDEX-treated cells was associated with the expression ofhigher levels of the apoptosis-suppressing gene bcI-X� com-pared to untreated or IL-6 pretreated cells (Figs. 2 and 3).

Withdrawal of DEX from cells that were pretreated for 3days resulted in a decline in the level ofexpression of bcl-XLso that 2 h later, expression of bcI-X� was reduced 2-fold,and at 6 or 24 h, it was reduced 3-S-fold (Fig. 3). This

decline in bcl-XL following withdrawal of DEX was alsoaccompanied by some increase in bax but not in bcl-2

expression (Fig. 3). There was also a decline in bcl-XLexpression following withdrawal of IL-6 or IL-6 plus DEX

from cells pretreated with these inducers (Fig. 3). Whentested for susceptibility to induction of apoptosis by cy-cloheximide, the DEX pretreated Mi cells that had a low

susceptibility to induction of apoptosis by cycloheximide(6 ± 2% apoptotic cells as in untreated Ml cells) showed

an increase in susceptibility to induction of apoptosis by

cycloheximide already at 6 h after withdrawal of DEX

(Table 4). Similarly, cells pretreated with IL-6 and DEX,

which showed the lowest susceptibility to the induction

of apoptosis with cycloheximide (1 ± i’Yoapoptoticcells), also showed an increase in susceptibility 6 h afterwithdrawal of IL-6 and DEX (Table 4). Cells pretreatedwith lL-6 alone, which showed high susceptibility to

induction of apoptosis with cycloheximide, were stillhighly susceptible 6 h after withdrawal of IL-6 (Table 4).The results indicate that the lack of increased sensitivity

to induction of apoptosis after pretreatment of Mi cellswith DEX or IL-6 and DEX, which down-regulated bcl-2

1234567

- � -

,�,.

A

I 3531078

872 ___

603

B

IC.’ � ��“i

Fig. 4. RT-PCR analysis of bcl-X expression in Ml cells. Lane 1, DNA size

markers 0 x 1 74 DNA cleaved with HaeIIt). RT-PCR from untreated cells)Lane2(, cellstreated with 10 ng/ml tL-6 (Lane.)), 1 p�a DEX (Lane4(, or tL-6plus DEX (Lane 5). PCR amplification of the coding region of bcl-X1 and

bcl-X5 from cDNA containing plasmids (positive controls; Lanes 6 and 7,respectively). A, ethidiuni bromide staining. B. hybridization of RT-PCR

products with the ‘2P-laheled cDNA probe for bcl-X5.

a)

+

u_Cl) Cl)

wO 0C ‘ X S

o� W�

A bcl-2

�.. �. bcI-xL

$z.�4�!,I4 . b� ‘�. ax

1�- actin

I 234567

B,__s_.� ,-..- - �.. .� . �_��_j .. ,-) �.-‘

. � - - -� -

. �

Fig. 5. Expression of bcl-2, bcl-XL, and bax in 7-Mi 2 myeloid leukemiccells. 7-Mi 2 cells were cultured for 2 days either with no additions (None)or treated with 10 ng/mI GM-CSE, 1 �M DEX, or GM-CSE plus DEX. A,Northern blot. B, RT-PCR analysis of bcl-X expression in untreated 7-Mi 2cells (Lane 1), in cells treated with GM-CSE (Lane 2), DEX (Lane 3), orGM-CSE plus DEX (Lane 4). PCR amplification of bcl-X5 and bcl-X1 codingregion of cDNA containing plasmids (Lanes S and 6, respectively). Lane 7,

DNA size markers (0 x 1 74 DNA cleaved with Had))). As in Fig. 4, both the0.8 and 0.6 kb fragments hybridized with a radiolabeled bcl-X5 probe.

650 bcl-2, bcl-X, , 1),i.s and Apoptosis

and bax expression, was associated with the increasedbcl-X expression.

Pretreatment of 7-Mi 2 cells with GM-CSF, which gavea i9-fold up-regulated bcl-XL expression, was associatedwith a decrease in susceptibility to induction of apoptosis(Table 3), but pretreatment with DEX that gave only aslight up-regulation of bcl-XL had no such effect. DEXalso did not further decrease the susceptibility of 7-M12cells to the induction of apoptosis beyond the effect of

1353

872

603

GM-CSF (Table 3). The results show that induction ofdifferentiation in different myeloid leukemic cells by IL-6plus DEX or by GM-CSF results in a high induction ofbcl-XL expression, which was associated with decreasedsusceptibility of the cells to the induction of apoptosis.

Discussion

Down-regulated expression of bcl-2 and bax in Mi myeloidleukemic cells by pretreatment with IL-6 was associatedwith increased cell susceptibility to induction of apoptosis(Table 5). However, the protection of Mi leukemic cells

Pretreatment with

IL-6None LL� De� ±fle�

B

1353

872

603

- + - + - + - + : CHX

310

194

113

Fig. 6. Induction of apoptosis in Ml cells by cycloheximide after pretreat-ment with IL-b, DEX, or lL-6 plus DEX. A, morphology of cycloheximide-

induced apoptotic cells (arrows( showing decreased size, condensed (hro-matin and cytoplasm, and nuclear fragmentation. Apoptotic cells had the

same morphology whether the cells were pretreated with lL-6 or DEX. B,DNA fragmentation. Lane 1, DNA size markers (0 x 1 74 DNA cleaved withHaelll(. Cells that were not pretreated (None( or pretreated for 2 (lays with 1 0ng/ml IL-6, 1 �M DEX, or lL-6 plus DEX were washed and incubated for 3 Ii

at 37CC without (-( or with (+( 5 pg/mI cycloheximide (CHX(, and t)NA was

then extracted.

DEX or DEX plus IL-6 showed a stronger increase in bcl-X1expression (6- and 9-fold, respectively). Following DEXwithdrawal from pretreated cells, expression of bcl-X1 de-dined, and susceptibility to the induction of apoptosis wasconcomitantly enhanced. Expression of bcl-XL was alsoinduced by GM-CSF in a different myeloid leukemia (7-Ml 2), which showed barely detectable expression of bcl-2,and this treatment with GM-CSF was also associated withdecreased cell susceptibility to the induction of apoptosis.The results suggest that bcl-XL expression may compensatefor decreased bcl-2 expression in the reduced cell suscep-tibility to the induction of apoptosis. The results also showthat the two apoptosis-suppressing genes, bcl-2 and bcl-X1,can be differently regulated so that bcl-2 expression is

against increased susceptibility to the induction of apopto-sis after pretreatment with DEX or DEX plus IL-6, which alsodown-regulated bcl-2 and bax expression, suggested thatthere may be other genes involved in this protection. Wefound that while IL-6 increased 2-3-fold the expression ofanother apoptosis-suppressing gene, bcl-XL, treatment with

Cell Growth & Differentiation 651

Table 2 Induction of apoptosis in Ml myeloid leukemic cells after

pretreatment svith IL-b and DEX

Pretreatment”Apoptotic cells (0/,,)b

Treatment for apoptosis

None’ Adriamycin Cycloheximide

None <1 2±1 7±2

I[..6 <1 28±5 43±5

DEX <I 4±2 6±2

tL-6#{247}DEX <1 3±1 2±1

.‘ Ml myeloid leukemic cells were cultured for 2 days with no addition(none), 10 ng/ml IL-6, 1 � DEX, or tL.6 and DEX.

S The percentage of apoptotic cells was determined 5 h after adding 5 pg/mIAdriamycin or 3 h after adding 2 pg/mI cycloheximide.‘ As shown in this column, pretreatment with tL-6, DEX. or IL-6 + DEX by

themselves did not induce apoptosis.

Table 3 Induction of apoptosis in 7-Mi 2 myeloid leukemic cells afterpretreatment with GM-CSE and DEX

Pretreatment”Apoptotic cells (%)b

Treatment for apoptosis

None’ Adriamycin Cycloheximide

None 2±1 29±3 46±5

GM-CSE 1 ± 1 10 ± 2 26 ± 4

DEX 2±1 33±4 47±6

GM-CSE + DEX 1 ± 1 1 1 ± 3 23 ± 4

.‘ 7-Mi 2 myeloid leukemic cells were cultured for 2 days with no addition(none), 10 np/mI GM-CSE, 1 p�a DEX, or GM-CSE and DEX.

1� The percentage of apoptotic cells was determined 5 h after adding 0.5pg/mI Adriamycin or 2.5 h after adding 0.5 pg/mI cycloheximide. These

concentrations of Adriamycin and cycloheximide are 10- and 4-fold lowerthan those used for Ml cells (Table 2(.‘ As shown in this column, pretreatment with GM-CSE, DEX, or GM-CSE +

DEX I)y themselves did not induce apoptosis.

Table 4 Susceptibility to induction of apoptosis in DEX and IL-6pretreated Ml cells following withdrawal of DEX and lL-6

Pretreatment”h after

withdrawal of

inducers

Apoptotic cells(%(I’

Cycloheximide

- +

None - 0 6±1

DEX 0 0 6±2

2 0 7±1

6 0 18±3

[-6 0 0 45±4

2 0 52±6

6 0 55±3

lL-6+DEX 0 0 1±1

2 0 2±1

6 0 8±1

., Ml teukemic cells were cultured for 3 days either without (None( or with

1 �M DEX, 10 ng/ml [-6, or DEX and IL-6. Cells were washed and cultured

without inducers for up to 6 h, and then cycloheximide was added (seebelow(.I, The percentage of apoptotic cells was determined 3 h after culture of cells

without (-( or with (+( 5 pg/mI cycloheximide.

652 bcl-2, bcl-xL, bax and Apoptosis

Table 5 Expression of bcl- 2, bcl-XL, and bax in myeloid Ieu kemic cells and susc eptibility to induction of apoptosis

Leukemia TreatmentExpression of�’ Sensitivity

toapoptosis�’bcl-2 bcl-XL bax

Ml None ++ ± ++ +

IL-6

DEX

IL-6+DEX

±

±

±

+

++

+++

+

+

+

+++

+

±

7-Mi 2 None

GM-CSF

DEXGM-CSF + DEX

-

-

-

-

-

+

±

+

++

++

++

++

++++

++

++++

++

‘Levels of expression according to Northern blot analysis. -, not detectable; ±, weakly detectable; + to +++, readily detectable increasing levels of expression.

b Based on data in Tables 2 and 3 on percentage of apoptotic cells induced by cycloheximide or Adriamycin. As shown in Tables 2 and 3, higher concentrations

of cycloheximide and Adriamycin were required to induce apoptosis in Mi cells than in 7-M12 cells.

down-regulated and bcl-XL expression is up-regulated (Ta-ble 5). The availability of an antibody to BCL-XL wouldmake it possible to determine to what extent the increase inbcl-X1 expression is also reflected in the production ofBCL-XL protein. The alternatively spliced bcI-X� transcriptthat can antagonize the activity of bcl-2 and thus promoteapoptosis (2) was not expressed in the leukemic cells eitherbefore or after treatment. This shows that, in these mouseleukemic cells like in other tissues in the developing mouse(20), only the bcl-XL gene is expressed, in contrast to unin-duced human thymocytes or activated T cells that expressonly bcl-X5 (2).

We found that in uninduced Ml and 7-Mi 2 cells, therewas a high level expression of the apoptosis-promotinggene bax, which was down-regulated by the induction ofdifferentiation with IL-6 and DEX in Mi cells but was notdown-regulated by GM-CSF or DEX in 7-Ml 2 cells(Table 5). The results indicate that expression of the threerelated genes bcl-2, bax, and bcl-XL can be differentlyregulated (Table 5) and suggest that the response to apop-tosis-inducing treatments is controlled by the balance be-tween expression of these and possibly other apoptosis-inducing and apoptosis-suppressing genes.

Materials and Methods

Cells and Cell Culture. Ml clone 1 2 myeloid leukemiccells (21) and the 7-Ml2 myeloid leukemia cells (19) weregrown in suspension in DMEM (H-2i ; GIBCO, Grand Is-land, NY) with iO% inactivated (56#{176}Cfor 30 mm) horseserum (GIBCO). Both clones of leukemic cells multiplied asmyeloblasts to promyelocytes, but only Ml cells expressedreadily detectable bcl-2 mRNA (13, 18). To induce differ-entiation and differentiation-associated decrease in bcl-2expression (1 3, i 8), Ml cells were incubated for 2 days with1 0 ng/mI recombinant mouse IL-6 (obtained from Dr. J. VanSnick), 0.4 pg/mI DEX (1 pM; Sigma Chemical Co., St. Louis,MO), or with both compounds together. The 7-Ml 2 leuke-mic cells were induced to differentiate with 10 ng/ml re-combinant mouse GM-CSF (obtained from Dr. S. Gillis), 0.4pg/mI DEX, or with both compounds together. The percent-age of undifferentiated (myeloblasts and promyelocytes)and more differentiated cells (myelocytes, metamyelocytes,monocyte-like cells, and macrophages) was determined onMay-Grunwald Giemsa-stained cytospin preparations bycounting 400 cells. Three experiments were carried out,and the data are given as mean ± SEM.

Assay for DNA Fragmentation and Morphological Anal-ysis of Apoptotic Cells. To assay for DNA fragmentation,DNA from 2 x 1 0� cells was isolated. Cells were lysed in0.5 ml lysis buffer [1 0 mM Tris (pH 7.5), 0.6% SDS, and 10mM EDTAJ; then RNAse solution (Ambion) was added to aconcentration of 15 pg/mI, and the lysate was incubated for20 mm at 37#{176}C.NaCI was then added to 1 M, and tubeswere incubated at 4#{176}Cfor 2 h. The tubes were centrifuged(1 4,000 X g; 4#{176}Cfor 30 mm), and DNA in the supernatantwas extracted by phenol-chloroform and ethanol precipi-tated at -20#{176}Covernight. Following 1 4,000 X g centrifu-gation (4#{176}Cfor 30 mm), the DNA pellet was air dried anddissolved in 20 p1 TE buffer (1 0 mt�i Tris-1 0 mtvt EDTA).Equal amounts of DNA (3 rig/lane) were electrophoresed

(3V/cm, 4 h) in 1.5% agarose gel containing 0.5 pg/mIethidium bromide and visualized under UV light. 0 x 174DNA digested with HaelIl (New England Biolabs, Beverly,MA) was used as fragment size markers. Morphologicalanalysis of the percentage of apoptotic cells was performedon May-Gru nwald G iemsa-stai ned cytospi n preparationsby counting 400 cells. Apoptotic cells were smaller than theviable cells and had a condensed cytoplasm and chromatinand fragmented nuclear chromatin bodies (7, 1 3, 18, 22).Three experiments were carried out, and the data are givenas the mean ± SEM.

Northern Blot Analysis. Cytoplasmic RNA was isolatedfrom myeloid leukemic cells that were cultured with noadditions or from cells induced to differentiate with IL-6,DEX, or lL-6 together with DEX in Ml cells, or withGM-CSF, DEX, or GM-CSF together with DEX in 7-Mi 2cells. Cells were lysed in buffer containing 0.65% NP4O,and nuclei were removed by centrifugation. The superna-tant was collected, 7 M urea and 1 % SDS were added, andRNA was extracted with phenol/chloroform and precipi-tated by ethanol. Twenty pg RNA were loaded in each laneof a 1 % agarose gel containing 3% formaldehyde. Gelelectrophoresis, RNA blotting to Gene Screen Plus mem-branes, and hybridization under stringent conditions withnick-translated (23) cDNA probes for mouse bcl-2 (24),

bcl-X (2), bax (6), c-myc (25), c-jun (26), jun-B (27), and ratgenomic probe for cytoplasmic �-actin (28) was carried outas described previously (18). Blots were exposed to KodakX-Omat film with an intensifying screen at -80#{176}Candscanned in a BAS 1 000 Biolmaging analyzer (Fuji PhotoFilm Co., Tokyo, Japan). The intensity oftranscript signals inblots of treated or untreated cells were normalized accord-

Cell Growth & Differentiation 653

ing to the intensity of the J3 actin signals serving as ameasure of the relative amounts of RNA/lane.

RT-PCR Analysis of bcl-X1 and bcl-X5. First-strand cDNAwas prepared from approximately 2 pg cytoplasmic RNAfrom untreated or treated cells in a volume of 20 p1 usingMoloney murine leukemia virus reverse transcriptase (BRL,Inc., Gaithersburg, MD) and oligo-dT as a primer at 37#{176}Cfor1 h. PCR amplification was carried out in a volume of 40 p1containing 4 p1 of the RT reaction products, 0.2 pg each ofthe 5’ and 3’ primers from the bcl-X coding region (seebelow). and 1 unit ofTaq E DNA polymerase (Promega Co.,Madison, WI) for 35 cycles of i mm at 94#{176}C,2 mm at 55#{176}C,and 3 mm at 72#{176}C.Positive controls for bcl-XL and bcl-X5cDNA amplification in the PCR reaction contained 20 ng ofthe bcl-XL or bcl-X5 plasmids. The bcl-X primers used were5’-GTGAGTGGACGGTCAGTG-3’ and 5’-TTGGACAATG-GACTGGTTGA-3’. Twenty p1 of the PCR products wereelectrophoresed in 1 .5% agarose (4V/cm for 3 h), and thegel was stained with ethidium bromide. The bcl-XL andbcl-X5 PCR products used as positive controls gave theexpected size bands of 0.8 and 0.6 kb, respectively (2). Toensure that the ethidium bromide-stained bands correspondto bcl-X amplified sequences in the PCR reaction, the gelwas blotted and hybridized with the nick-translated cDNAprobe for bcI-X� as above.

AcknowledgmentsWe thank Dr. Jacques Van Snick (Ludwig Institute for Cancer Research,

Brussels, Belgium) for recombinant mouse IL-6; Dr. Steven Gillis (ImmunexCorporation, Seattle, WA) for recombinant mouse GM-CSE; Drs. Craig B.Thompson and Lawrence H. Boise (Howard Hughes Medical Institute Re-search Laboratories, University of Chicago, Chicago, IL) for mouse bcl-X1and bcl-X5 plasmids; Dr. Stanley J. Korsmeyer (Howard Hughes MedicalInstitute, Washington University School of Medicine, St. Louis, MO) formouse bax and bcl-2 plasmids; and Dr. Kenneth B. Marcu (State Universityof New York, Stony Brook, NY), Dr. Daniel Nathans (Department of Molec-ular Biology and Genetics, Johns Hopkins University School of Medicine,Baltimore, MD), and Dr. Un Nudel (Weizmann Institute of Science) for theplasmids containing mouse c-myc, c-jun. jun-B, and rat �-actin, respec-

tively. We thank Rachel Kama, Nurit Dorevitch, and Dr. Lydia Cohen forskillful technical assistance.

References1 . Korsmeyer, S. J. bcl-2 initiates a new category of oncogenes: regulators ofcell death. Blood, 80: 879-886, 1992.

2. Boise, L. H., Gonzalez-Garcia, M., Postema, C. E., Ding, L., Lindsten, T.,Turka, L. A., Mao, x., Nu#{241}ez,G., and Thompson, C. B. bcl-x, a bcl-2-related

gene that functions as a dominant regulator of apoptotic cell death. Cell, 74:

597-608, 1993.

3. Gottschalk, A. R., Boise, L. H., Thompson, C. B., and Quintans, J. Iden-tification of immunosuppressant-induced apoptosis in a murine B-cell line

and its prevention by bcl-X but not bcl-2. Proc. NatI. Acad. Sd. USA, 91:7350-7354, 1994.

4. Hengartner, M. 0., Ellis, R., and Horvitz, H. R. Caenorhabditis elegansgene ced-9 protects cells from programmed cell death. Nature (Lond.), 356:494-499, 1992.

5. Henderson, S., Huen, D., Rowe, M., Dawson, C., Johnson, G., andRickinson, A. Epstein-Barr virus coded BHRE 1 protein, a viral homolog ofBcl-2, protects human B cells from programmed cell death. Proc. NatI. Acad.

Sd. USA, 90:8479-8483, 1993.

6. Oltvai, Z., Milliman, C. L., and Korsmeyer, S. J. BcI-2 heterodimerizes invivo with a conserved homolog, Bax, that accelerates programmed cell

death. Cell, 74: 609-619, 1993.

7. Yonish-Rouach, E., Resnitzky, D., Lotem, I., Sachs, L., Kimchi, A., andOren, M. Wild-type p53 induces apoptosis of myeloid leukemic cells that isinhibited by interleukin 6. Nature (Lond.), 352: 345-347, 1991.

8. Lotem, J., and Sachs, L. Hematopoietic cells from mice deficient inwild-type p53 are more resistant to induction of apoptosis by some agents.

Blood, 82: 1092-1096, 1993.

9. Lowe, S. W., Schmitt, E. M., Smith, S. W., Osborne, B. A., and Jacks, 1.

p53 is required for radiation-induced apoptosis in mouse thymocytes. Nature(Lond.), 362:847-849, 1993.

10. Clarke, A. R., Purdie, C. A., Harrison, D. J., Morris, R. G., Bird, C. C.,Hooper, M. L., and Wyllie, A. H. Thymocyte apoptosis induced by p53-dependent and independent pathways. Nature (Lond.), 362: 849-852,1993.

1 1 . Askew, D. S., Ashmun, R. A., Simmons, B. C., and Cleveland, J. 1.

Constitutive c-myc expression in an IL-3-dependent myeloid cell line sup-presses cell cycle arrest and accelerates apoptosis. Oncogene, 6:

1915-1922, 1991.

12. Evan, G. I., Wyllie, A. H., Gilbert, C. S., Littlewood, T. D., Land, H.,Brooks, M., Waters, C. M., Penn, L. Z., and Hancock, D. C. Induction ofapoptosis in fibroblasts by c-myc protein. Cell, 69: 119-128, 1992.

13. Lotem, I., and Sachs, L. Regulation by bcl-2, c-myc and p53 of suscep-

tibility to induction of apoptosis by heat shock and cancer chemotherapycompounds in differentiation-competent and defective myeloid leukemiccells. Cell Growth & Differ., 4:41-47, 1993.

14. Sachs, L., and Lotem, J. Control of programmed cell death in normal andleukemic cells: new implications for therapy. Blood, 82: 1 5-21 , 1993.

15. Sachs, L., and Lotem, J. Apoptosis in the hemopoietic system. In: C. D.

Gregory (ed), Apoptosis and the Immune System, pp. 371-403. New York:

John Wiley, 1995.

16. Oltvai, Z. N., and Korsmeyer, S. J. Checkpoints of duelling dimers foil

death wishes. Cell, 79: 189-192, 1994.

17. Campos, L., Rouault, i-P., Sabido, 0., Oriol, P., Roubi, N., Vasselon, C.,Archimbaud, E., Magaud, J-P., and Guyotat D. High expression of bcl-2protein in acute myeloid leukemic cells is associated with poor response to

chemotherapy. Blood, 81: 3091-3096, 1993.

18. Lotem, J., and Sachs, L. Control of sensitivity to induction of apoptosis inmyeloid leukemic cells by differentiation and bcl-2 dependent and indepen-dent pathways. Cell Growth & Differ., 5: 321-327, 1994.

19. Lotem, I., and Sachs, L. Genetic dissection of the control of normaldifferentiation of myeloid leukemic cells. Proc. NatI. Acad. Sci. USA, 74:5554-5558, 1977.

20. Gonzalez-Garcia, M., P#{233}rez-Ballestero, R., Ding, L., Duan, L., Boise, L.H., Thompson, C. B., and Nu#{241}ez,G. bcl-x1 is the major bci-x mRNA form

expressed during murine development and its product localizes to mitochon-dna. Development (Camb.), 120: 3033-3042, 1994.

21 . Eibach, E., Hayashi, M., and Sachs, L. Control of normal differentiationof myeloid leukemic cells to macrophages and granulocytes. Proc. NatI.

Acad. Sci. USA, 70: 343-346, 1973.

22. Lotem, J., Cragoe, E. J., Jr., and Sachs, L. Rescue from programmed celldeath in leukemic and normal myeloid cells. Blood, 78: 953-960, 1991.

23. Rigby, P. W. I., Deikmann, M., Rhodes, C., and Berg, P. Labelling

deoxyribonucleic acid to high specific activity in in vitro by nick translationwith DNA polymerase. J. Mol. Biol., 1 13: 237-251, 1977.

24. Nu#{241}ez,G., London, L., Hockenbery, D., Alexander, M., McKearn, J. P.,and Korsmeyer, S. J. Deregulated bcl-2 gene expression selectively prolongs

survival of growth factor deprived hemopoietic cell lines. I. Immunol., 144:

3602-3610, 1990.

25. Stanton, L. W., Watt, R., and Marcu, K. B. Translocation, breakage andtruncated transcripts of c-myc oncogene in murine plasmacytomas. Nature(Lond.), 303:401-406, 1983.

26. Ryder, K., and Nathans, D. Induction of proto-oncogene c-jun by serumgrowth factors. Proc. NatI. Acad. Sci. USA, 85: 8464-8467, 1988.

27. Ryder, K., Lau, F., and Nathans, D. A gene activated by growth factorsis related to the oncogene v-jun. Proc. NatI. Acad. Sd. USA, 85: 1487-1 491,1988.

28. Nudel, U., Zakut, R., Shani, M., Neuman, S., Levy, Z., and Yaffe, D. Thenucleotide sequence of the rat cytoplasmic 13actin gene. Nucleic Acids Res.,11:1759-1771, 1983.