Proc. Natl. Acad. Sci. USAVol. 90, pp. 9460-9464, October 1993Physiology

Estrogen receptor and progesterone receptor genes are expresseddifferentially in mouse embryos duringpreimplantation developmentQ. HoU AND J. GORSKI*Department of Biochemistry, University of Wisconsin, Madison, WI 53706-1569

Contributed by J. Gorski, July 15, 1993

ABSTRACT Estrogen and progesterone play an importantrole in the development and implantation of preimplantationembryos. However, it is controversial whether these hormonesact directly on the embryos. The effects of these hormonesdepend on the existence of their specific receptors. To deter-mine whether estrogen receptor (ER) and progesterone recep-tor genes are expressed in mouse preimplantation embryos, weexamined RNA from embryos at different stages of preimplan-tation development by reverse transcription-polymerase chainreaction techniques. ER mRNA was found in oocytes andfertilized eggs. The message level began to decline at thetwo-cell stage and reached its lowest level at the five- toeight-cell stage. ER mRNA was not detectable at the morulastage but reappeared at the blastocyst stage. Progesteronereceptor mRNA was not detectable until the blastocyst stage.The embryonic expression of ER and progesterone receptorgenes in the blastocyst suggests a possible functional require-ment for ER and progesterone receptor at this stage of devel-opment. These results provide a basis for determining thedirect role of estrogen and progesterone in preimplantationembryos.

There are no reports of natural mutations that result in eitherdeficient estrogen synthesis or resistance to estrogen action.This contrasts with testosterone synthesis and action, wheresingle-gene mutations that interfere with both processes havebeen characterized in many species. This difference has ledGeorge and Wilson (1) to suggest that such estrogen-relatedmutations may be lethal at an early stage of development.Estrogen and progesterone play key roles in the establish-

ment and maintenance of pregnancy. Elimination of thesehormones from pregnant animals causes deleterious effectson the development and implantation of the embryos. Hy-pophysectomy of pregnant rats was shown to result in thedelayed entry of eggs into the uterus, expulsion of eggs fromthe uterus, and retarded development of eggs. Injection ofestrone and progesterone significantly reversed this effect(2). Delayed implantation has been observed in ovariecto-mized mice treated with progesterone only (3). Dormantblastocysts can be reactivated by the injection of estrogen (4,5). Estrogen and progesterone stimulate the metabolism ofdelayed implantation mouse embryos (6). It appears to beimportant to maintain the concentrations of estrogen andprogesterone at appropriate levels relative to each other.Elevated ratios of estradiol to progesterone were shown toinhibit blastocyst metabolism and implantation (7). Sincethese experiments were performed on ovariectomized ani-mals receiving exogenous hormones, it is not clear whetherthe hormones acted directly on the embryo or indirectlythrough the mother's reproductive tract. Some authorsclaimed that the effects were indirect, since retardation of

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

embryo development and inhibition of implantation wereobserved when embryos were grown in oviductal fluid ob-tained from estrogen-treated, ovariectomized mice, whereasno such effects were observed when embryos were grown indefined medium with estrogen (8). In contrast, studies bySmith (9) showed that embryos treated with estrogen in vitrohad significantly higher implantation rates when transferredto foster mothers receiving only progesterone, suggestingthat estrogen acts directly on the embryo. To obtain anenvironment without the complication of maternal effects,some investigators turned to in vitro culturing of embryos.Under these conditions, estrogen was shown to affect theuptake and incorporation of nucleic acid precursors (10, 11)and amino acids (12) by mouse blastocysts. It was also shownthat antiestrogen CI 628 and aromatase inhibitor 1,4,6-androstatriene-3,17-dione inhibited the development ofmouse embryos (13, 14). This blockage was alleviated by thecoadministration of 17,3estradiol. These results support theidea that estrogen affects the embryo directly.

Estrogen and progesterone elicit their functions by bindingto their specific receptors. The receptors become activatedupon hormone binding and serve as transcription factors thatmodulate the expression of their target genes (15). Forestrogen and progesterone to have any direct effect on theembryo, the receptors for these hormones must be present inthe embryo. A demonstration of the expression of estrogenreceptor (ER) and progesterone receptor (PR) in preimplan-tation embryos would provide a basis for the direct effect ofestrogen and progesterone on the embryo.We examined the levels of both ER and PR mRNA during

preimplantation development by using reverse transcription(RT)-PCR. We found that ER and PR genes were expresseddifferentially in the preimplantation mouse embryo. Theresults for ER contrast with a recent report that preimplan-tation mouse embryos themselves did not express the ERgene even though maternal message was detected in theoocyte (16). Our results suggest that both maternal andembryonic sources of mRNA are involved.

MATERIALS AND METHODSEmbryo Culture. Female C57BL/6 mice (Harlan-Sprague-

Dawley) were superovulated by intraperitoneal injection of 5units of pregnant mare's serum gonadotropin (Diosynth,Chicago), followed by 5 units of human chorionic gonadot-ropin (Organon) 48 hr later. These females were placed withmale mice after human chorionic gonadotropin injection.Mated female mice were sacrificed the following day aroundnoon, and eggs [0.5 day postcoitum (dpc)] were released fromthe excised oviducts. Hyaluronidase (Sigma) at 0.3 mg/mlwas used to remove the cumulus cells. The eggs were washed

Abbreviations: ER, estrogen receptor; PR, progesterone receptor;dpc, days postcoitum; RT, reverse transcription; RPL19, ribosomalprotein L19.*To whom reprint requests should be addressed.

9460

Dow

nloa

ded

by g

uest

on

Janu

ary

31, 2

020

Proc. Natl. Acad. Sci. USA 90 (1993) 9461

through M2 medium (17) and cultured for various times inM16 medium (17) at 37°C in a humidified incubator under 5%C02/95% air. Embryos were harvested at the desired stageand washed three times through M2 medium. The embryoswere either used fresh or frozen on dry ice and stored at-70°C. The final wash of the embryos was also collected toserve as a negative control. In the case ofunfertilized oocytesand one-cell embryos, the zona pellucida was removed withacidic Tyrode's solution (17) before harvesting. The timecourse of embryo development was approximately as fol-lows: two-cell, 1.5 dpc; three- to four-cell, 2 dpc; five- toeight-cell, 2.5 dpc; compacted morula, 3.5 dpc; blastocyst,4.5 dpc. Most blastocysts collected were expanded and somewere hatching.RNA Isolation. Total RNA was isolated from a pool of

about 100 embryos by a micro adaptation of the guanidine/cesium chloride method (18). One hundred fifty microliters of4 M guanidine isothiocyanate/25 mM sodium citrate, pH7/0.5% N-lauroylsarcosine/10 mM 2-mercaptoethanol con-taining about 15 ,g ofEscherichia coli rRNA as a carrier wasused to lyse cells. This lysate was then overlaid onto 150 IlIof 5.7 M CsCl/0.1 M EDTA, pH 7.0, and centrifuged for 2.5hr in a TLA-100 tabletop ultracentrifuge (Beckman) at100,000 rpm, 25°C. The resulting pellet was washed with 70%ethanol and dried. This pellet was suspended, precipitatedwith ethanol, washed twice, and suspended in water. Analiquot was used to determine RNA concentration by absor-bance at 260 nm. The recovery of RNA was calculatedaccording to the recovery of carrier RNA (range, 60-80%).Based on the recovery and the initial number of embryosincluded, adjustment was made so that each sample con-tained RNA from the same number of embryos per volume.RT-PCR. RNA was reverse transcribed at 42°C for 1 hr. A

20-,ul RT mixture contained 50mM KCl, 10mM Tris HCl (pH9.0 at 250C), 0.1% Triton X-100, 5 mM MgCl2, 40 units ofRNasin (Promega), 100 pmol of random hexamer (Pharma-cia), 1 mM each dNTP, and 12 units of avian myeloblastosisvirus reverse transcriptase (Promega). A portion of the RTproduct containing RNA from the same amount of embryoswas then subjected to PCR in a DNA thermal cycler (model480, Perkin-Elmer/Cetus). A 100-,l reaction mixture in-cluded 50 mM KCl, 10 mM Tris HCl (pH 9.0 at 25°C), 0.1%Triton X-100, 4 mM MgCl2, 200 ,uM each deoxynucleotide,2.5 units of Taq DNA polymerase (Promega), and 14 nM eachprimer. After 15-20 cycles, additional primers were added toreach a final concentration of 190 nM, followed by another30-35 cycles. The cycling profile was 94°C, 1 min; cool to55-65°C, 1 min; 55-65°C, 1 min; and 72°C, 1 min. The primersets used were mERa (positions 1354-1374; 5'-GGAGAT-TCTGATGATTGGTCT-3') and mERb (1812-1831; 5'-CATCTCCAGGAGCAGGTCAT-3'), mERo (661-682; 5'-TTCTGACAATCGACGCCAGAAT-3') and mERp (934-955; 5'-CATCATGCCCACTTCGTAACAC-3'), mPRa(2382-2403; 5'-CCCACAGGAGTTTGTCAAACTC-3') andmPRb (2687-2708; 5'-TAACTTCAGACATCATTTCCGG-3'), and RPL19A (5'-CTGAAGGTCAAAGGGAATGTG-3')and RPL19B (5'-GGACAGAGTCTTGATGATCTC-3') (19).The PCR products (15 ,ul) were fractionated in a 2% agarosegel and visualized by ethidium bromide staining.

Southern Analysis. PCR products were separated by aga-rose gel electrophoresis, transferred to nylon membranes(Hybond-N, Amersham) and hybridized to probes that werelabeled with 32P by use ofa random labeling kit (United StatesBiochemical). Hybridization was carried out at 42°C over-night in 50% formamide/6x SSPE (20)/Sx Denhardt's solu-tion (20)/0.5% SDS containing denatured sheared herringsperm DNA (100 ,ug/ml) and the probe (106 cpm/ml). The ERprobe was a mouse ER cDNA (gift of M. Parker; ref. 21). ThePR probe was a 2-kb fragment of mouse PR cDNA (gift of G.Shyamala; ref. 22). The blot was washed sequentially with 2 x

SSC (20)/0.5% SDS at room temperature, 1 x SSC/0.5% SDSat 68°C, and 0.1 x SSC/0.1% SDS at 68°C. The blot was usedto expose Kodak XAR film (Eastman Kodak) at -70°C.

RESULTSPrevious work used immunocytochemistry and immunoblot-ting methods to detect ER in mouse embryos as early as 13days (23) and probably 10 days of gestation age. The sensi-tivity of these procedures is limited; therefore, to extend ourinvestigation to the preimplantation embryo, the more sen-sitive RT-PCR technique was used in this study. Since bothER and PR are expressed in the female reproductive tract, itwas crucial to avoid maternal cell contamination. In vitrocultured embryos were used to minimize this possible con-tamination. Cumulus cells, which also express ER, wereremoved from the eggs and the eggs were washed so that thissource of contamination was avoided (data not shown).Further, unfertilized oocytes and one-cell embryos used foranalysis were freed ofzona pellucida and washed to eliminateany residual cumulus cells that might still have been associ-ated with the zona even after hyaluronidase treatment. RNAfrom the uterus, in which both ER and PR are expressed, wasused as a positive control.ER mRNA Is Present at Both Early and Late Stages of

Preimplantation Development. To determine whether the ERgene is expressed in preimplantation embryos, two sets ofprimers were used in the PCR analysis. One set of primers,mERo and mERp, is specific for the DNA-binding domainand will yield a DNA product of 295 bp. Since this primer setspans an intron, based on human ER genomic organization(24), amplification of genomic DNA will not occur due to thelarge size of the intron (>10 kb). Therefore, a 295-bp bandindicates the presence of ER mRNA in the original sample.Fig. 1A shows the result from PCR using this set of primers.None of the wash control lanes had any detectable signal,suggesting that no contamination occurred. A bright band of295 bp was detected in the lane from the unfertilized oocyte.One-cell- and two-cell-embryo lanes also displayed this band,but the signal was less intense than that seen in the oocytelane. This suggests that one-cell and two-cell embryos alsohave ER mRNA, though not as much as oocytes. However,there were no 295-bp fragments in the three- to four-cell-embryo, the five- to eight-cell-embryo, or the morula lane.A most interesting observation was the reoccurrence of the

295-bp fragment in the blastocyst lane, indicating that the ERgene is expressed at this stage. Southern blot analysis con-firmed that the 295-bp band detected was indeed ofER origin(Fig. 1B).To further confirm these findings, another set of primers,

mERa and mERb, specific for the steroid-binding domain ofthe receptor and spanning three introns, was used. A PCRproduct of 478 bp was expected. Fig. 1C shows the result ofSouthern analysis of the PCR products. The presence of a478-bp band which hybridizes to ER cDNA indicates thepresence of ER mRNA. A 478-bp band of high intensity ispresent in the unfertilized oocyte lane. The band intensityappears to decline from the two-cell embryo through the five-to eight-cell embryo. There was no detectable signal in thelane from the morula. This specific band was detected againin blastocysts. We did not detect any signal in the one-cell-embryo lane in this blot. This is not representative of ourusual results: similar experiments have consistently shown aspecific band exhibiting an intensity close to that shown bythe oocyte.Taken together, the results with the two sets of primers

show an ER mRNA expression pattern demonstrating thatthe ER mRNA level is high in the oocyte. The level dimin-ishes at the two-cell stage and drops to the lowest level in five-to eight-cell embryos. In the morula, the ER mRNA is

Physiology: Hou and Gorski

Dow

nloa

ded

by g

uest

on

Janu

ary

31, 2

020

9462 Physiology: Hou and Gorski

A

:E

506,507 -

396 -344 -298 -

B a

0

v_

Q Q000 0 _4q

04)

-

_

I-,

a) C)C) Iy

C.)4

=

u- 0)

- C4 CM A

u

0)00

Yx

.n

Q

C)

O~C

--

_ >1 >1

00 C)U. W)° g

°o O .5 X

U

co E 8 8 0

O;

Oc Ya3

II

ER-E

-C

.90 0=) 0

O

CU(L

Y C) -

CU4)

tU4

~

(U VN m

4 0

0

essentially not detectable. The decrease in ER mRNA is nota result of dilution of ER mRNA due to cell division, becausetotal RNA from the same number of embryos was assayed.The ER mRNA reappears at the blastocyst stage. Thus farthere is only one known ER gene. Since our primers werechosen both in the DNA-binding domain and the steroid-binding domain, it is unlikely that we have detected forms ofER other than the classical nuclear ER.The Pattern of ER mRNA Expression Is Not a Result of

Selective Sample Degradation. To ensure that all RNA sam-ples were intact, we performed PCR on the same set ofsamples, using primers that amplify ribosomal protein L19(RPL19) cDNA. Because ribosomal proteins should be ex-pressed universally, failure to detect RPL19 mRNA wouldsuggest that the sample was degraded. A band ofthe expectedsize (194 bp) was seen in all the appropriate lanes (Fig. 2),indicating that embryos at all stages had the RPL19 mRNA.The expression of this gene seemed to increase somewhatduring development, consistent with the role of ribosomalproteins in protein synthesis. The consistent detection ofRPL19 gene expression in preimplantation embryos at allstages suggests that all the RNA samples were intact. A faintband in the wash control of the five- to eight-cell embryosuggested that minor contamination of an unknown sourceoccurred in this sample.The PR Gene Is Silent Until the Blastocyst Stage. The

expression of the PR gene was examined by using a set ofprimers specific for the steroid-binding domain. This set ofprimers spans two introns, based on chicken PR genomicorganization (25), and should give a PCR product of 327 bp.No 327-bp band was detected in the lanes from the oocytethrough the morula (Fig. 3A). A band of 327 bp was visibleonly in the lane from the blastocyst. The specificity of theproduct was confirmed by Southern analysis (Fig. 3B). Ap-parently, the PR gene is not expressed until the blastocyststage during preimplantation development. This pattern ofexpression differs from that of the ER gene.

DISCUSSIONThis study examines the expression of two steroid receptorgenes in preimplantation mouse embryos by RT-PCR tech-0c~n

CC)

C)1

CM C

ER - :4 a,,

-1

cl)kl

Q c),l C' 8E ?-- 3 C;

A:;; 5 go

0.0Qo

FIG. 1. Expression of the ER gene. (A) Agarose gel showing thePCR product amplified with primers mERo and mERp. RNA from1.8 embryos was included in each sample. Lanes marked (C) arewash controls. Uterine RNA (50 pg) was used as positive control.The annealing temperature for PCR was 65°C, and 17 plus 33 cycleswere run. (B) Autoradiogram of Southern blot of the gel shown in A.(C) Southern blot analysis of the PCR product amplified with primersmERa and mERb. RNA from 3.6 embryos was included in eachsample. Lanes marked (C) are as in A. Uterine RNA (100 pg) wasused as positive control. The annealing temperature for PCR was62°C, and 20 plus 35 cycles were run. Lane M, molecular sizemarkers (lengths at left in nucleotides).

506,507 \396 :-344 _298220 -_2011547134J-

FIG. 2. Agarose gel showing the PCR product amplified withprimers RPL19A and RPL19B. RNA from 1.8 embryos was includedin each sample. Lanes marked (C) are wash controls. Uterine RNA(50 pg) was used as positive control. The annealing temperature forPCR was 55°C, and 15 plus 30 cycles were run.

Proc. Natl. Acad. Sci. USA 90 (1993)

u

tA

Dow

nloa

ded

by g

uest

on

Janu

ary

31, 2

020

Proc. Natl. Acad. Sci. USA 90 (1993) 9463

A

506,507 A_396 a-3449298 -N-2202011547134'

u

8u

88

-

= =4('D t)

rU'U R

Q4

B 1U

u_,-

" ) 4W)4)4

1-

4)

00 00

c cm

e Z 8 ° Z

;E 2 m m ZD4.)

Iz

._

eO

_- :^ >%

ed C'8

a e c- .

o O X X 22 c-mim >

ju --- ER mRNA

PR miRNA

20-v

X maternal\ mes age

10- \e1g\e embryonicmessage

0-0 1 2 3 4 5

d.p.c.

Oocytematuration

Ovary

v- PR

FIG. 3. Expression of the PR gene. (A) Agarose gel showing thePCR product amplified with primers mPRa and mPRb. RNA from 3.6embryos was included in each sample. Lanes marked (C) are washcontrols. Uterine RNA (100 pg) was used as positive control. Theannealing temperature for PCR was 62°C, and 20 plus 35 cycles wererun. (B) Autoradiogram of Southern blot of the gel shown in A.

niques. Fig. 4 summarizes our results and interpretations. ERmRNA exists in the unfertilized oocyte. The level of ERmRNA declines as the embryo develops. It is greatly dimin-ished at the five- to eight-cell stage and is essentially unde-tectable at the morula stage. The observed decrease in ERmRNA is not due to dilution of the initial mRNA as theembryo proceeds through cell expansion, because the ERmRNA was assayed on a per-embryo basis.

Contrary to a previous report (16), ER mRNA reappearedat the blastocyst stage. In comparison, the mRNA for the PRgene was not detectable in the earlier stages of embryonicdevelopment but did appear at the blastocyst stage. WhileRT-PCR does not provide strictly quantitative data, webelieve that the following discussion is instructive. Based onthe assumption that each uterine cell contains 10 pg ofRNA,we used a ratio of 1.8 embryos to 5 uterine cells for analysis.On the basis of the intensity of the band seen on Southernblots (by /-scan or densitometry; data not shown), a roughestimate of ER mRNA levels in one oocyte is on the order ofthat found in 10 uterine cells. A blastocyst has an ER mRNA

Implantation

Oviduct Uterus

FIG. 4. Summary of mRNA levels of ER and PR in mousepreimplantation embryos. The ER and PR mRNA levels per embryoare shown in relationship to the age of the embryos. The scale isarbitrary. The corresponding developmental stages are depicted. Thedrawing of the embryos at various stages of development wasadapted from Hogan et al. (17) with permission (Cold Spring HarborLaboratory Press, copyright 1986). The origin of the mRNA and itspossible function are also indicated.

level comparable to that found in one uterine cell. The PRmRNA level in a blastocyst is comparable to that in oneuterine cell. Since one blastocyst has 32-64 cells, it is notknown if low levels of ER and PR mRNAs are expressed inmany cells or if moderate levels are expressed in one or a fewcells. This question might be approachable with in situhybridization or in situ PCR. Since RT-PCR assays only forthe mRNA level, we cannot say with certainty that functionalproteins are expressed. Attempts to measure ER protein havebeen limited by the sensitivity of available methods. Forexample, we have calculated that 1000 or more embryoswould be required in order to obtain 20 pg of ER. This couldbe detected by immunoblotting if the background proteinlevel does not interfere. This is not likely with availablemethods.

Origin ofER and PR mRNAs in Preimplantation Embryos.In the mouse, development of the embryo before the middleof the two-cell stage depends largely on protein and RNAsynthesized during oogenesis. The embryonic genome isrelatively inactive until sometime after the first cleavage.Development of early embryos beyond the two-cell stagerequires new mRNA synthesis (17, 26, 27). Since a high levelof ER mRNA was detected in the unfertilized oocyte, it isreasonable to argue that the ER mRNA detected in theone-cell embryo was carried over from the oocyte wherematernally synthesized mRNA is stored. The ER mRNAlevel declined from the two-cell to the five- to eight-cell stage.This is compatible with the general observation of loss ofmaternalRNA in early development (28). Though the embryois capable of synthesizing RNA from the two-cell to the five-to eight-cell stage, the ER mRNA detected at these stages is

__j

Physiology: Hou and Gorski

tuu

Dow

nloa

ded

by g

uest

on

Janu

ary

31, 2

020

9464 Physiology: Hou and Gorski

not likely to be synthesized by the embryo, because theamount continues to decline. At the morula stage, ER mRNAis virtually undetectable. The reappearance ofER mRNA andappearance of PR mRNA in the blastocyst must be a resultof embryonic RNA synthesis. At this stage, embryonictranscription is very active.

Implications for the Development and Implantation of Pre-implantation Embryos. The differential expression of ER andPR genes during preimplantation development suggests thatER and PR may have different roles during this period. It isknown that estrogen stimulates follicular growth in the pro-cess of oocyte maturation (29). That both the cumulus cellsand the oocyte express ER mRNA suggests that estrogen caninfluence both cell types. Blastocyst formation is an impor-tant period in preimplantation development. The embryoimplants at the late blastocyst stage. It is interesting that theembryonic expression of the ER and the PR genes coincideswith this time. Blastocyst-specific genes are expressed in thetransition from morula to blastocyst, and this expression isrequired (30, 31). There are also changes in the synthesis andsecretion of stage-specific proteins from early to late blasto-cysts (5, 32). Estrogen apparently is an important regulator inthis process, since dormant blastocysts reactivated withestrogen undergo the same changes. The presence of estro-gen seems necessary for the development of the embryobecause antiestrogen or aromatase inhibitor can block em-bryo development (13, 14). The function of estrogen might beto stimulate the development of the embryo and prepare it forimplantation. This is consistent with the possible lethality ofan ER mutation (1). The requirement for ER must go beyondthe maternal source, since a homozygous ER mutant fromherterozygous parents would be viable otherwise. Theexpression of the ER gene by the embryo itself must becrucial for survival.

Progesterone is also necessary for implantation in mice.Treatment with progesterone antagonists RU486 andZK98734 prevents normal embryo transport and implanta-tion, but progesterone is apparently not essential for devel-opment (33). The timely expression of the PR gene inblastocysts implies that the effect of progesterone can beexerted on the embryo as well.While this work was in progress, Wu et al. (16) reported

that they could detect ER messages in oocytes and two-cellembryos, which is consistent with our findings, but theyfailed to detect any ER message at the blastocyst stage. Thisdiscrepancy could result from the sensitivity of the PCRmethods, since we used "booster" PCR and a higher numberof cycles, which should give greater sensitivity. They con-cluded that preimplantation embryos do not express ER andthat any effect of estrogen on the embryo acts indirectlythrough the uterus, possibly through the regulation of growthfactors and their receptors. Although the expression ofgrowth factors and growth factor receptors (platelet-derivedgrowth factor A, transforming growth factors a and A3, andepidermal growth factor receptor) in the embryo has beenreported (18, 34, 35), and in some cases, the expression wasregulated by estrogen, it is not clear whether the effect wason the embryo directly or mediated through the uterus. Ourresults clearly demonstrate that the ER gene is expressed bypreimplantation embryos. This finding provides a basis forthe direct effect of estrogen on the embryo. Since both ERand PR genes are expressed at the blastocyst stage, a newperspective of the mechanism of implantation may be inorder. Previous studies have focused largely on uterineresponse to hormones; our findings suggest an embryonictarget for the hormones. Estrogen and progesterone mayinduce changes both in the uterus and in the embryo that leadto successful implantation.

We thank Drs. Rick Monson and Lorraine Leibfried-Rutledge fortheir help in setting up the embryo culture. We thank Dr. SidneyStrictland for helpful discussions and Kathryn A. Holtgraver foreditorial assistance. This work was supported in part by the Collegeof Agricultural and Life Sciences, University ofWisconsin-Madison,and by a grant from the National Institute ofChild Health and HumanDevelopment (HD08192) awarded to J.G.

1. George, F. W. & Wilson, J. D. (1988) in The Physiology ofRepro-duction, eds. Knobil, E., Neil, J. D., Ewing, L. L., Greenwald,G. S., Market, C. L. & Pfaff, D. W. (Raven, New York), Vol. 1, pp.3-27.

2. Wu, J. T., Dickmann, Z. & Johnson, D. C. (1971) J. Endocrinol. 51,569-574.

3. Yoshinaga, K. & Adams, C. E. (1966) J. Reprod. Fertil. 12,593-595.

4. Moore, G. P. & Carter, N. B. (1984) Cell Differ. 14, 19-23.5. Nieder, G. L., Weitlauf, H. M. & Suda, H. M. (1987) Biol. Reprod.

36, 687-699.6. Torbit, C. A. & Weitlauf, H. M. (1975) J. Reprod. Fertil. 42, 45-50.7. Safro, E., O'Neill, C. & Saunders, D. M. (1990) Fertil. Steril. 54,

1150-1153.8. Cline, E. M., Randall, P. A. & Oliphant, G. (1977) Fertil. Steril. 28,

766-771.9. Smith, D. M. (1968) J. Endocrinol. 41, 17-29.

10. Harrer, J. A. & Lee, H. H. (1973) J. Reprod. Fertil. 33, 327-330.11. Lau, N. I. F., Davis, B. K. & Chang, M. C. (1973) Proc. Soc. Exp.

Biol. Med. 144, 333-336.12. Smith, D. M. & Smith, A. E. S. (1971) Biol. Reprod. 4, 66-73.13. Sen Gupta, J., Dey, S. K. & Dickmann, Z. (1977) Steroids 29,

363-369.14. Sengupta, J., Roy, S. K. & Manchanda, S. K. (1982) J. Reprod.

Fertil. 66, 63-66.15. Beato, M. (1989) Cell 56, 335-344.16. Wu, T. C. J., Wang, L. & Wan, Y. J. Y. (1992) Mol. Reprod. Dev.

33, 407-412.17. Hogan, B., Costantini, F. & Lacy, E. (1986) Manipulating the

Mouse Embryo: A Laboratory Manual (Cold Spring Harbor Lab.Press, Plainview, NY).

18. Rappolee, D. A., Brenner, C. N., Schultz, R., Mark, D. & Werb,Z. (1988) Science 241, 1823-1825.

19. Chan, Y.-L., Lin, A., McNally, J., Peleg, D., Meyuhas, 0. & Wool,I. G. (1987) J. Biol. Chem. 262, 1111-1115.

20. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989) MolecularCloning: A Laboratory Manual (Cold Spring Harbor Lab. Press,Plainview, NY).

21. White, R., Lees, J. A., Needham, M., Ham, J. & Parker, M. (1987)Mol. Endocrinol. 1, 735-744.

22. Schott, D. R., Shyamala, G., Schneider, W. & Parry, G. (1991)Biochemistry 30, 7014-7020.

23. Greco, T. L., Furlow, J. D., Duello, T. M. & Gorski, J. (1992)Endocrinology 130, 421-429.

24. Ponglikitmongkol, M., Green, S. & Chambon, P. (1988) EMBO J. 7,3385-3388.

25. Huckaby, C. S., Conneely, 0. M., Beattie, W. G., Dobson,A. D. W., Tsai, M.-J. & O'Malley, B. W. (1987) Proc. Natl. Acad.Sci. USA 84, 8380-8384.

26. Schultz, G. A. (1986) in Experimental Approaches to MammalianEmbryonic Development, eds. Rossant, J. & Pedersen, R. A. (Cam-bridge Univ. Press, Cambridge, U.K.), pp. 239-265.

27. Pedersen, R. A. (1988) in The Physiology of Reproduction, eds.Knobil, E., Neil, J. D., Ewing, L. L., Greenwald, G. S., Market,C. L. & Pfaff, D. W. (Raven, New York), Vol. 1, pp. 198-199.

28. Bachvarova, R. & DeLeon, V. (1980) Dev. Biol. 74, 1-8.29. Austin, C. R. & Short, R. V. (1982) Hormonal Control of Repro-

duction (Cambridge Univ. Press, Cambridge, U.K.).30. Braude, P. R. (1979) Dev. Biol. 68, 440-452.31. Schultz, G. A., Clough, J. R., Braude, P. R., Pelham, H. R. B. &

Johnson, M. H. (1981) in Cellular and Molecular Aspects of Im-plantation, eds. Glasser, S. R. & Bullock, D. W. (Plenum, NewYork), pp. 137-154.

32. Holmes, P. V. & Dickson, A. D. (1973) J. Embryol. Exp. Morphol.29, 639-645.

33. Vinijsanun, A. & Martin, L. (1990) Reprod. Fertil. Dev. 2, 713-727.34. Paria, B. C., Jones, K. L., Flanders, K. C. & Dey, S. K. (1992)

Dev. Biol. 151, 91-104.35. Paria, B. C., Das, S. K., Andrews, G. K. & Dey, S. K. (1993) Proc.

Natl. Acad. Sci. USA 90, 55-59.

Proc. Natl. Acad. Sci. USA 90 (1993)

Dow

nloa

ded

by g

uest

on

Janu

ary

31, 2

020