Detrimental effects of two active X chromosomes on early ... · difference between male and female. The complete lack of any known mutation that disturbs X-inactivation suggests that

Development 109, 189-201 (1990)Printed in Great Britain © The Company of Biologists Limited 1990

189

Detrimental effects of two active X chromosomes on early mouse

development

NOBUO TAKAGI and KENJI ABE

Research Center for Molecular Genetics, and Chromosome Research Unit, Faculty of Science, Hokkaido University, North 10, West 7,Sapporo 060, Japan

Summary

Matings between female mice carrying Searle's translo-cation, T(X;16)16H, and normal males give rise tochromosomally unbalanced zygotes with two completesets of autosomes, one normal X chromosome and oneX16 translocation chromosome (XnX16 embryos). SinceX chromosome inactivation does not occur in theseembryos, probably due to the lack of the inactivationcenter on X16, XnX16 embryos are functionally disomicfor the proximal 63 % of the X chromosome and trisomicfor the distal segment of chromosome 16. Developmentalabnormalities found in XnX16 embryos include: (1)growth retardation detected as early as stage 9, (2)continual loss of embryonic ectoderm cells either bydeath or by expulsion into the proamniotic cavity, (3)

underdevelopment of the ectoplacental cone throughoutthe course of development, (4) very limited, if any,mesoderm formation, (5) failure in early organogenesisincluding the embryo, amnion, chorion and yolk sac.Death occurred at 10 days p.c. Since the combination ofXO and trisomy 16 does not severely affect early mousedevelopment, it is likely that regulatory mechanismsessential for early embryogenesis do not function cor-rectly in XnX16 embryos due to activity of the extra Xchromosome segment of X16.

Key words: X-chromosome inactivation, mousedevelopment, ectoplacental cone, mesoderm, Searle'stranslocation, inactivation center, functional disomy X.

Introduction

In early embryogenesis of female mammals, one of twoX chromosomes is genetically inactivated (Lyon, 1961),which results in compensation of X chromosome dosagedifference between male and female. The complete lackof any known mutation that disturbs X-inactivationsuggests that it is vital to mammals and its failure leadsto death of affected individuals as shown in Drosophila(Belote and Lucchesi, 1980; Lucchesi and Skripsky,1981). However, no direct observation for this view hasever been reported in any species of mammals.

One way to inquire into the significance and functionof X chromosome inactivation and X chromosome itselfwould be to examine development of mouse embryos inwhich one entire X chromosome and an additional Xchromosome segment are active (partial functionaldisomy X). It is believed that X-inactivation is con-trolled by a c«-acting X chromosomal site called Xchromosome controlling element {Xce) or inactivationcenter (Cattanach, 1975), loss of which probably ren-ders the chromosome unresponsive to inactivation sig-nals. We set out to examine developmental effects ofpartial functional disomy X making use of Searle'sX-autosome translocation, T(X;16)16H (Searle, 1962;Lyon et al. 1964). In this translocation, the proximal Xchromosome segment, corresponding to 63% of the

entire X, has lost Xce, and gained a distal chromosome16 segment in exchange (X ). The other translocationchromosome (16X) consists of the proximal chromo-some 16 segment and the distal X chromosome segmentcarrying Xce (Fig. 1). Matings of female translocationheterozygotes with chromosomally normal males regu-larly give rise to unbalanced zygotes with two completesets of autosomes, one normal X chromosome fromfather and X16 from mother. These karyotypicallyunbalanced XnX16 embryos consistently suffered fromsevere growth retardation in early postimplantationstages and had no cytogenetically identifiable inactive Xchromosome (Takagi, 1980).

The present paper follows the development of suchembryos in vivo. Growth retardation was evident inXnX embryos as early as 6.5 days p.c , probably atstage 9, and death occurred at 10.5 days p.c.Interestingly, these XnX16 embryos maintained thebasic structure of 6- to 7-day embryos until 10.5 daysp.c. without apparent differentiation of the chorion,yolk sac and embryo proper. Sporadic XO embryostrisomic for chromosome 16 consistently showed onlyslight growth retardation at 8.5 days p.c. Partial mono-somy X or partial trisomy for chromosome 16 or bothare not directly responsible for the severe developmen-tal arrest observed in XnX16 embryos. It seems likelythat regulatory mechanisms essential to early embryo-

190 N. Takagi and K. Abe

genesis do not function correctly in XnX16 embryos dueto activity of the extra X chromosome segment of X16.We may conclude that functional disomy for the entireX chromosome causes much severer developmentalabnormality than that found in XnX16 embryos andX-inactivation is indeed indispensable for normal em-bryogenesis.

Materials and methods

T(X;16)16H translocationThe breakpoint in the X chromosome is at D band and inchromosome 16 is at B5 band. Thus the X16 chromosomeroughly corresponds in length to the intact X (Xn) and 16X tointact chromosome 16 (Fig. 1). Chromosome 16X is assumedto carry the inactivation center or Xce (Rastan, 1983), sincethis chromosome can be inactivated (Takagi, 1980; McMahonand Monk, 1983).

Recovery and BrdU labelling of embryosSpontaneously ovulating T16H/+ females were caged withkaryotypically normal males hemizygous for Tabby, and werechecked daily for vaginal plugs early in the morning. The daywhen the vaginal plug was found was taken as day 0 ofpregnancy. Embryos were usually recovered from decidualswellings at 9 to 10 a.m. from day 6 to 10 of pregnancy. Theywill be designated as embryos at 6.5-10.5 days post coitum(p.c). The staging system proposed by Theiler (1972) wasused whenever applicable, because developmental variationbetween litters was evident in early postimplantation stages.Stages of XnX16 embryos were determined from those ofnormally grown littermates.

Recovered embryos were photographed and incubated inEagle's minimum essential medium (MEM) supplemented

X16 16*

B

Fig. 1. Schematic representation of (A) T(X;16)Htranslocation and (B) the unbalanced karyotype ofXn,X16,16,16.

with 10% fetal calf serum and 100 jig ml"1

5-bromo-2-deoxyuridine (BrdU) at 37° in an atmosphere of5% CO2 in air. The duration of incubation was 6h (6.5-dayembryos) to 9h (9.5-day embryos) including the last hour inthe presence of lf/giril"1 Colcemid. After hypotonic treat-ment with 1% sodium citrate, embryos were fixed with 3:1methanol: acetic acid. Chromosome slides were preparedaccording to a modification of the air-drying method de-scribed earlier (Wroblewska and Dyban, 1969). Slides werestained with freshly prepared acridine orange and examinedunder a fluorescence microscope.

Histological examination of embryosEmbryos isolated from decidual swellings were fixed withBouin's fixative, stained with hematoxylin, embedded inparaffin wax and 7/tm sections were cut and stained withhematoxylin and eosin (HE). In certain cases, the entireimplantation sites were sectioned with surrounding uterinetissue. In order to ascertain identification of X"X16 embryos,embryos freed of decidual tissue and Reichert's membranewere cut into the extraembryonic and the embryonic regionwith the aid of fine glass needles, and the former was used forhistological examination and the latter for karyotyping asdescribed above. For detailed histological examination, iso-lated embryos were fixed with 2.5% glutaraldehyde in caco-dylate buffer, postfixed with 1 % osmium tetroxide, dehy-drated with acetone and embedded in Epon 812. Sections cutat 1-2 /.an were stained with 1% toluidine blue.

Staining with lectinsFITC-labelled concanavalin A (Con A), soybean agglutinin(SBA), peanut agglutinin (PNA) and wheat germ agglutinin(WGA) were purchased from Vector Laboratories, Burl-ingame, CA. Dewaxed tissue sections were incubated in PBSfor Ih and then in a dilute solution of a lectin (100 ^g ml"1 inPBS) for 1 h at room temperature in a moist chamber. Afterincubation, sections were washed throughly in PBS, mountedin fresh PBS and observed under a fluorescence microscope.

Culture of X"X16 embryonic ectoderm in vitroThe embryonic region was removed from 8.5-day presump-tive XnX embryos after treatment with 2.5% pancreatinwith the aid of tungsten needles. The embryonic ectodermdeprived of the embryonic visceral endoderm was cultured ina 30 mm tissue culture dish containing Eagle's minimumessential medium containing 10% fetal calf serum.

Results

Identification of X"X16 embryosIn order to detect morphological traits adequate for theidentification of XnX embryos, karyotypes were stud-ied in 351 embryos from 86 females heterozygous forT16H mated with chromosomally normal males at 6.5to 9.5 days p.c. In contradiction to our previous study(Takagi, 1980), the frequency of XnX16 embryos rela-tive to the total embryos produced by alternate oradjacent 1 disjunction remained about 20% during thisperiod of pregnancy (Table 1). The previous conclusionthat XnX embryos were lost by 8.5 days p.c. was basedon a smaller number of embryos and could have beenpremature.

Developmental variation was evident among littersrecovered at day 6 of pregnancy. It was not possible to

Two active X chromosomes in mouse development 191

Table 1. Karyotypes of embryos recovered from T16H heterozygous females mated with chromosomally normalmales

Karyotype

2:2 normal disjunction40(Xn,Xn,16,16)40(Xn,Y,16,16)40(Xn X16,16\16)4O(X1(5,Y,16",16)40(Xn,X'fi,16,16)Subtotal

2:2 adjacent 2 disjunction40(Xn,Xn X'6,16)40(Xn,X"\Y,16)40(Xn, 16,16,16)

3:1 disjunction41(Xn,Xn,16\16,16)41(Xn,Y,16\16,16)41(Xn,Xn,16,16,16)41(Xn,Y,16,16,16)41(Xn,X'6,Y,l6x,16)39(Xn,16,16)

Triploid

Total

Abbreviations: Ms, monosomy; Ts, trisomy; d.

Geneticimbalance

nonenonenonenone

MsXd;Tsl6d

TsXp;Msl6pMsXd;Msl6p

MsX;Tsl6

TsXd;Tsl6pMsXp;Tsl6p

Tsl6Tsl6XXYMsX

segment distal to the break point;

No.

6

2012232621

102

o o

to

300100

2

110

embryos recoveredgestation days

7

152021151586

011

111000

0

91

8

2720241923

113

235

100011

2

128

p, segment proximal to the

at

9

61553

20

0 ^

i 1000000

0

22

break

Total

6853736562

321

• 1 0

4

351

point.

distinguish XnX16 embryos based on gross phenotypicdifferences among litters probably at stage 8. Oc-casional X"X16 embryos were as big as or bigger thantheir littermates having a normal or a balanced karyo-type. However, growth retardation of XnX16 embryoswas striking in remaining litters at stage 9 with theaverage long axis 53.6% (range, 42%-69%) that ofthe normally growing littermates (Fig. 2a-e). XnX16

embryos could not, however, be identified with cer-tainty on the growth retardation alone. Embryos withvarious karyotypic abnormalities other than XnX16

were usually retarded in development. Furthermore,growth retardation was quite common among chromo-somally normal or balanced littermates. This wasfurther confirmed by a histological study of 6.5-dayembryos in situ. 20 out of 65 viable embryos weresmaller than their littermates and had an underde-veloped ectoplacental cone (see below). All of theseembryos could not be XnX16, in view of the much lowerfrequency of XnX16 embryos at 6.5 days p.c.

Gross morphology of the 7.5-day XnX embryos, onthe other hand, was striking enough for their overtidentification in most cases. In addition to the increasedgrowth retardation, XnX16 embryos had a small andthin embryonic region, undulated visceral endodermfacing the yolk cavity particularly at the extraembryonicregion (Fig. 2f-j). The mean long axis of 7.5-day(stages 10, 11) XnX16 embryos was 54.2% that ofnormally grown littermates, but the wide range of 30 %to 65 % shows that the manifestation of the unbalancedkaryotype is by no means uniform.

™ embryos are trisomic for the distal segment of

chromosome 16 and monosomic for the distal segmentof the X chromosome. XO embryos trisomic forchromosome 16, resulting from the adjacent 2 disjunc-tion of T16H quadrivalent, were occasionally obtainedin this study. Those embryos were well-balanced,though slightly retarded, at 8.5 and 9.5 days p.c.(Fig. 3) in accord with the observation that embryostrisomic for the entire chromosome 16 survive untilimmediately before or even shortly beyond birth (Miya-bara et al. 1983), and XO mice are phenotypicallynormal and fertile female. Thus, the severe develop-mental anomalies found in XnX embryos may not beexplained without postulating imbalance associatedwith X chromosome inactivation.

In order to characterize XnX16 embryos more fully,we carried out combined histological and cytogeneticalanalyses on 7.5-day embryos. 15 out of 86 embryosstudied proved to have the unbalanced karyotype. Inaddition to marked growth retardation, these embryoswere characterized by a very poorly developed, if any,ectoplacental cone irrespective of developmental vari-ation between litters. It was also noted that the dorsalspace of the embryo bound by the visceral endodermwas occupied by densely packed cells. It is suggestedthat the cells destined to form the ectoplacental conewere prevented from invading the decidual cavity andforced to remain in the embryo.

Eight intact 7.5-day embryos presumed to have theunbalanced karyotype were then subject to histologicalexamination. They resembled pregastrulation embryosat stage 9, but their embryonic ectoderm was small andthe proamniotic cavity was filled with cell debris.


Furthermore, in occasional embryos, a number of foldswere formed by the visceral endoderm in the ventralside of the extraembryonic region. Thus, small size,underdeveloped ectoplacental cone, packed cells at thesite of the ectoplacental cavity, poorly developed em-bryonic ectoderm with a number of dead cells in theproamniotic cavity, lack or shortage of mesoderm andabnormal visceral endoderm appeared features specificto 7.5-day XnX16 embryos.

Morphological abnormalities characteristic to 7.5-day XnX16 embryos were expressed in more exagger-ated form in 8.5-day embryos. They were easily recog-nized under the dissection microscope because of theirruggedly outlined yolk sac and the tiny underdevelopedembryonic region. Later, the outline of the embryobecame smooth with the expansion of the extraembry-onic region. Now growth of these embryos variedextensively and the average long axis of XnX16 embryosvaried from 14% to 60% that of normally grownlittermates with the mean value of 38.6% (Fig. 2k-q).Most 9.5-day XnX16 conceptuses were round in shapesometimes with a structure resembling an embryoproper (Fig. 2r-w). The size and shape of putativeXnX16 embryos remained unchanged at 10.5 days p.c.

Fig. 2. Photomicrographs of intactembryos recovered at 6.5 (a-e), 7.5 (f-i),8.5 (k-q) and 9.5 (r-w) days p.c. XnX'*embryos are shown on the right andnormally grown littermates on the left forcomparison. Bar=0.1mm.

Developmental defects in X"X16 embryosBecause of higher resolution, plastic embedding andultramicrotomy rather than a paraffin embeddingmethod was applied to the detailed histological study ofXnX16 embryos identified by gross morphology.

6-day embryos at Stage 9Although we failed to find a clear histological featurespecific to 6.5-day XnX16 embryos in HE-stained sec-tions, we noticed that the proamniotic cavity of oc-casional retarded embryos was filled with clumps ofcells. They could be XnX16 embryos, because thepresence of such cells in the proamniotic cavity is one ofthe most consistent characteristics of XnX16 embryos inlater stages. Compared with a normally grown embryo(Fig. 4B), one of such putative XnX16 embryos shownin Fig. 4A had an underdeveloped ectoplacental coneand free cells in the proamniotic cavity.

7-day embryos at stage 10In putative XnX16 embryos at stage 10, the embryonicectoderm obviously everted or evaginated into theproamniotic cavity near the ventral end (Figs 4C,10B).This phenomenon was observed in three consecutive


Fig. 3. Photomicrographs of XO embryos trisomic forchromosome 16 at 8.5 (c,f) and 9.5 (h) days p.c. withnormally grown littermates. Embryo proper enclosed by theamnion is evident in these karyotypically unbalancedconceptuses. Bar=0.1mm.

embryos. Mitotic cells were abundant in the evaginatedtissue. Ectoplacental cone was underdeveloped in com-parison with normally grown littermates (Fig. 3D).Furthermore, the extraembryonic visceral endodermtended to be thicker than expected.

7-day embryos at stage 11Histological findings made in HE-stained preparationwere confirmed in 10 embryos examined. Folds formedby the visceral endoderm were conspicuous in theembryo shown in Fig. 5B, but not in the embryo shownin Fig. 5A. In the latter, however, the extraembryonicvisceral endoderm was thickened considerably. In bothembryos, embryonic ectoderm was disproportionatelysmall and mesodermal component was still almostabsent. It is possible that the lack of mesodermal cellsthat usually line the extraembryonic visceral endodermforming yolk sac led the free inner surfaces of the

visceral endoderm to stick together (cf. Fig. 5E). Apacked cell mass at the dorsal end and underdevelop-ment of the ectoplacental cone were evident inFig. 5A,B,D.

8-day embryos at stages 12-13A total of 10 putative XnX16 embryos were studied. At afirst glance, the basic structure of a well-developedembryo shown in Fig. 6A resembled that of normalembryos at stage 11 (Fig. 5C). The poorly developedembryonic region still retained the morphology of theembryonic ectoderm and mesoderm cells were sparseboth in the embryonic and extraembryonic region. Inthe cavity lined with the embryonic ectoderm, deadcells were abundant. Observation of serial sectionsrevealed clear differences between XnX16 at stage 13and normal embryos at stage 11. In the latter, threecavities are present: most ventral of these is theamniotic cavity lined with embryonic ectoderm; in themiddle is the exocoelom lined with mesoderm; andfinally the ectoplacental cavity lined with extraembry-onic ectoderm. One of the most remarkable features inXnX16 embryos at stage 13 was that embryonic ecto-derm was still continuous with extraembryonic ecto-derm in a complicated way (Fig. 6B). The cavityapparently corresponding to the exocoelom was linedwith a thin single cell layer. Mesoderm was producedmuch less extensively in XnX16 embryos grown poorly.

9-day embryos at stages 14-15In spite of apparent morphological variability, the basicstructure of 7 putative XnX embryos examined atthese stages was relatively uniform. The sphericalembryo consisted of two cavities, one was lined with theembryonic and extraembryonic ectoderm layer and theother with a thin mesodermal layer (Fig. 7D). Findingsin 8-day embryos suggested that the former correspondsto the proamniotic cavity and the latter to the abnormalexocoelom or yolk sac with scanty mesodermal cells. Insome embryos, the boundary between these cavitieswas complicated. The structure resembling an embryoproper in Fig. 2j was a part of such proamniotic cavitylined with a cell layer made up of the embryonic and theextraembryonic ectoderm (Figs 7C, 8G). The atrophicembryonic region was identified because of its mor-phology resembling the embryonic ectoderm of earlypostimplantation stages and accompanying dead cells(Fig. 10D). No mesodermal component was found inthe smaller embryo shown in Fig. 7A. There was nosign of maternal blood circulation at the ectoplacentalregion in this case.

10-day embryos at stages 16-17Putative XnX16 embryos were identified because of thegeneral morphology common to embryos at previousstages. A histological study in 3 embryos showed thatthe basic structure was exactly the same as that of XnX16

embryos at 9.5 days p.c., but signs of degeneration wereevident (Fig. 8A,B). The yolk cavity was filled withdead endoderm cells (Fig. 8D). Degeneration was alsofound in embryonic ectoderm and adjoining visceral


pc

ee

endoderm (Fig. 8C,E). Embryonic ectoderm was lessaffected in other regions (Fig. 8F,G). A large clump ofdead or dying cells was found at the dorsal end of theembryo (Fig. 8B). This embryo looked healthier inother regions.

Expression of lectin receptors in X"X16 embryosAbnormal development of XnX16 embryos preventedus from identifying various tissues for certain on thebasis of morphology alone. Distribution of lectin recep-tors may help interpreting the abnormal developmentof XnX16 embryos. Deparaffinized sections of normally

Fig. 4. Thick-section light micrographs of(A) putative 6-day XnX16 embryo at stage 9,(B) normal 6-day embryo at stage 9,(C) putative 7-day XnX'6 embryos at stage 10,(D) normal 7-day embryo at stage 10. Thearrow in C indicates examination of theembryonic ectoderm, ee, embryonicectoderm; ep, ectoplacental cone; ex,extraembryonic ectoderm; m, mesoderm; pc,proamniotic cavity, ve, visceral endoderm.Bar=0.1mm.

grown embryos at stage 11 and XnXL6 embryos at stage13 were stained with either of FITC-labelled Con A,SBA, PNA and WGA.

Con A receptors were expressed on various tissues ofnormal embryos. Fluorescence was strongest on thefree surface of the extraembryonic visceral endodermfacing yolk cavity, followed by the ectoplacental coneand cells around it, the inner surface of the embryonicectoderm facing the amniotic cavity and the mesodermincluding the allantois. Con A binding was similar inXnX16 embryo. However, these embryos did not havetissues or groups of cells having a mode of reaction

Two active X chromosomes in mouse development

ep

Fig. 5. Thick-section light micrographs of putative 7-day embryos at stages 12-13: (A) well-grown X"X16 embryo,(B) severely retarded XnX16 embryo, (C) normal 8-day embryo, (D) high magnification of the dorsal half of the embryo inB, (E) high magnification of the ventral half of the embryo in B. The arrow indicates packed cells at the dorsal end of theembryo in A, and the absence of ectoplacental cone in D. a, amniotic cavity; al, allantois; am, amnion; ec, ectoplacentalcavity; ee, embryonic ectoderm; ep, ectoplacental cone; ex, extraembryonic ectoderm; ve, visceral endoderm. Bar=0.1mm(A,B,C),0.05mm (D,E).

corresponding to the mesoderm of normal embryos(data not shown).

As reported earlier, PNA receptors were expressedon various tissues of normal embryos (Hamada et al.1983). Only the extraembryonic ectoderm expressedPNA receptors on both the inner and the outer surfacesfacing the exocoelom and the ectoplacental cavity. InXnX embryos, the cell layer adjoining the embryonicectoderm (data not shown) showed a similar pattern of

expression. This seems to support the finding that theembryonic ectoderm and the extraembryonic ectodermfailed to separate in XnX16 embryos.

Expression of WGA receptors was also in agreementwith the above finding. WGA reacted with the ectopla-cental cone in normal embryos and cells packed at thedorsal end of X"X16 embryos showed a similar patternof expression. Expression of Con A and PNA receptorswas consistent with this observation. Only cell surface


ve

Fig. 6. Thick-section lightmicrographs of welldeveloped putative 8-dayXnX16 embryos: (A) saggitalsection, (B) frontal section,ee, embryonic ectoderm; eo,exocoelom; ep, ectoplacentalcone; ex, extraembryonicectoderm; m, mesoderm; ve,visceral endoderm.Bar=0.1mm.

of the extraembryonic visceral endoderm facing theyolk cavity expressed SBA receptors in normal andXnX16 embryos.

Proliferative potential of embryonic ectoderm fromX"Xi6 embryos in vitroIn order to examine the proliferative capability ofXnX16 embryonic ectoderm cells, the embryonic regionwas dissected from five putative XnX16 embryos atstages 12-13 and cultured in MEM supplemented with10% fetal calf serum. Within 24 h a sheet of cells withepithelial characters spread around the explant (Fig. 9).Outward spreading of endodermal cells continued with-out appreciable increase in cell number for the next twodays followed by gradual deterioration. The embryonicectoderm cell clump at the center flattened and each cellspread on the substratum within 48 h, but the cellsapparently began to die off 2 to 3 days later withoutundergoing active cell division. No healthy cells wereleft 7 days after the initiation of culture.

Big halos of spindle-shaped and epithelial cells devel-oped around embryonic ectoderm fragments dissectedfrom normally grown littermates within 24 h. Thesecells were mitotically very active and groups of pulsat-ing cells were found frequently. Apparently, the ca-pacity of outgrowth did not depend on the size ofexplants. Thus, it is very likely that XnX16 embryonicectoderm either did not have proliferative potential invitro or had lost it by 8.5 days p.c. The yolk sac was alsocultured in vitro. Cells grew out vigorously from normalyolk sac and stayed mitotically active at least for 2weeks. No cell outgrowth was found, on the other hand,

in the case of XnX16 embryos for several days afterexplantation, which probably reflects the shortage ofmesoderm in the explanted tissue. The XnX sexchromosome constitution was ascertained in culturedextraembryonic halves of 3 embryos used for this study.Karyotype could not be determined in remaining 2embryos because of the lack of good metaphasespreads.

Discussion

In the previous study, we concluded that X-inactivationdoes not occur in XnX16 embryos, because we couldnever find an X chromosome showing out-of-phasereplication (Takagi, 1980). During the course of thisstudy, we examined a total of 516 XnX16 metaphasecells labelled with BrdU mainly from 6.5-day embryos,and in no case could we detect an asynchronouslyreplicating X chromosome. There is no a priori reasonto suppose complete selection against Xt6-inactive cells,if they ever occurred, since such cells are geneticallybalanced or trisomic only for the distal region ofchromosome 16. Inactivation of X", on the other hand,makes XnX16 cells partially nullisomic for the Xchromosome and trisomic for the distal segment ofchromosome 16, an imbalance probably severe enoughfor rigorous selection against such cells. Continualdeath of the embryonic ectoderm cells in XnX16 em-bryos (Fig. 10) may favor selection against such Xn-inactive cells. However, our study in normal embryosindicated that X-inactivation in the embryonic ecto-derm had finished within 24 h (Takagi et al. 1982).


Fig. 7. Thick-sectionlight micrographs ofthree putative 9-dayXnXltf embryos. B and Cwere from the sameembryo, eo, exocoelom;ep, ectoplacental cone;pc, proamniotic cavity.Bar=0.1mm.

Given the same situation in XnX16 embryos, we have toconclude that, after X-inactivation, selection did notoccur rapidly in every cell and cell death continued overa long period of time from 6 to 10 days p.c.: XnX16 cellswith a late-replicating X chromosome would have beendetected during that period. It is very likely thatX-inactivation has not occurred in XnX16 embryos,because X16 chromosome had lost the Xce or theinactivation center as a result of T16H translocation.Current models of X-inactivation (Gartler and Riggs,1983) predict that such cells behave as if they had onlyone X chromosome. It follows from what has beendiscussed above that most developmental abnormalitiesfound in XnX16 embryos are attributable to partialfunctional disomy X.

Developmental abnormalities found in XnX16 em-bryos are: (1) growth retardation detected as early as

stage 9, (2) continual loss of embryonic ectoderm cellseither by death or by expulsion into the proamnioticcavity, (3) underdevelopment of the ectoplacental conethroughout the course of development, (4) very limited,if any, mesoderm formation, (5) failure in early organo-genesis including the embryo, amnion, chorion andyolk sac. It is hard for any mutation affecting earlydevelopment to identify the specific cell type in whichthe gene first acts (McLaren, 1976), and the defectivegene product on the basis of phenotype (Magnuson,1986), primarily because of the number and complexityof steps involving close interdependence between dif-ferent cell types in developing embryos. Functionaldisomy X studied here is no exception. The difficulty ismuch greater in this case, because genes on 63 % of theX chromosome rather than a single gene have to betaken into account.


Fig. 8. Thick-section light micrographs of a degenerating putative XnX16 embryo at 10 days p.c. C,D,E were magnified fromthe section shown in B. F was magnified from the section shown in A. ee, embryonic ectoderm; eo, exocoelom; ex,extraembryonic ectoderm; pc, proamniotic cavity; pe, parietal endoderm; ve, visceral endoderm; yc, yolk cavity.Bar=0.1mm (A,B), 0.02mm (C-G).

Developmental processes are often a result of inter-actions between different cell types as seen in theclassical example of embryonic induction in Amphib-ians. It is conceivable that the normal course of devel-opment is perturbed critically in mice by the absence or

shortage of one of definitive germ layers. The absenceof mesodermal components seems to explain variousaspects of abnormal development in XnX embryos.

In normal development, mesoderm moves into theextraembryonic region and pushes the junctional region


A '

{*%•" U\-v-\t

•t

D ?;

Fig. 9. Phase-contrast micrographs of X"X16 embryonic ectoderm grown in vitro for (A) 24h, (B) 48h, (C) 72 h, (D) 120h.Bar=0.05mm.

between the embryonic and extraembryonic region intofolds at stage 10. The amniotic folds thus formed bulgeinto the proamniotic cavity and eventually fuse at stage11. Coalescence of lacunae formed within the amnioticfolds leads to the formation of the exocoelom which islined by mesoderm and separates the extraembryonicectoderm from the embryonic ectoderm. The failure toform amniotic folds and hence the amnion itself inXnX16 embryos seems to be attributed to deficientmesoderm, which, in turn, accounts for the failure toseparate the extraembryonic from the embryonic ecto-derm.

Decreasing sizes of the embryonic ectoderm togetherwith the occurrence of cell debris associated with itindicate that the embryonic ectoderm cells die continu-ally in XnX16 embryos. It is tempting to postulate thatcertain signals from mesodermal tissue(s) are necessaryfor proliferation of the embryonic ectoderm and forneural induction. With the shortage of such signals, theembryonic ectoderm cells must have been unable to

undergo differentiation in vivo and lost proliferativepotential in vivo and in vitro. Alternatively, death ofXnX16 embryonic ectoderm cells are due to functionaldisomy X. This is, however, apparently at variance withthe fact that certain types of cells are viable with twofunctional X chromosomes (Epstein, 1969; Gardnerand Lyon, 1971; Martin et al. 1978; McBurney andStrutt, 1980). Furthermore, Hockey et al. (1989)showed that clones of C86 embryonal carcinoma couldform differentiated cells with two apparently active Xchromosomes.

Inadequate nutrition due to the poorly developedectoplacental cone, and possibly also inadequate differ-entiation of trophoblasts, may explain the early growthretardation. Blastocyst reconstitution experimentsshowed that poor development of trophoblast signifi-cantly impaired development of the embryo proper(Barton et al. 1985). The large number of cells packed atthe dorsal end of the XnX16 embryo seem to correspondto diploid ectoplacental cone cells that failed to invade


Fig. 10. Embryonic region of putative XnX16 embryos showing continual loss of embryonic ectoderm cells by cell death orrepulsion into the proamniotic cavity from (A) 7-day embryo at stage 9, (8) 7-day embryo at stage 10, (C) 8-day embryo,(D) 9-day embryo. Bar=0.05mm.

the decidual cavity. It may be speculated that thecrowded extraembryonic region put pressure on thegrowing embryonic ectoderm and compelled it to evagi-nate into the proamniotic cavity. Alternatively, attenu-ated cell-cell contact due to cell death or altered cellsurface property could explain evagination or expulsionof embryonic ectoderm cells. Putative diploid ectopla-cental cone cells retained at the dorsal end of theembryo also suggest alteration in cell surface propertyor cell motility.

It is evident that neither X chromosome of XnX16

•embryos carries any mutant genes that gravely affectembryonic development, e.g. albino deletions (Lewis etal. 1976; Niswander et al. 1988, 1989) and mutations(Bennett and Dunn, 1958; Spiegelman etal. 1976), sinceXn and X16 plus 16" were fully expressed in reproducti-vely competent father and mother, respectively. Dupli-cation of the transcriptionally active X chromosomedosage is apparently responsible for the defective cellu-lar differentiation, and, hence, abnormal developmentof XnX16 embryos, which is consistent with the fact thatX-chromosome differentiation occurs at different timesin different cell lineages, and is associated with thedeparture, or differentiation of cells from the stem line(Monk, 1981). In the paucity of data concerning earlydevelopment of autosomal trisomics and tetrasomics, itis difficult to determine whether the X chromosome is

involved in early development and cell differentiationmore closely than average autosomes.

In normal female mouse embryos, the paternallyderived X chromosome (Xp) is inactivated in thetrophectoderm and the primitive endoderm (Takagiand Sasaki, 1975; Takagi etal. 1982). The polar troph-ectoderm cell lineage fails to develop in mouse embryoscarrying an extra maternally derived X chromosome(Xm) such as XmXmXp and XmXmY (Shao and Takagi,unpublished observation). X chromosome inactivationpatterns observed in triploid mouse embryos (Endo andTakagi, 1982) predict that two X^s have remainedactive in extraembryonic regions of such embryosprobably due to imprinting of Xm (Lyon and Rastan,1984). This, if indeed the case, would probably supportthe assumption that the X chromosome is involved inearly mouse development and two doses of active Xchromosome derange regulatory mechanisms. In hu-mans, where XmXmY and possibly XmXmXp are com-patible with survival, imprinting, if any, may not lastuntil the time of X-inactivation in extraembryonictissues.

We thank Professor Tadashi Yamamoto and Mr YoshioTakakuwa, Department of Zoology, Faculty of Science,Hokkaido University for kind demonstration of histologicaltechniques. We also thank to Professor Hiroshi Hori for the


loan of a glass knife maker. Our grateful thanks are due toMiss Etsuko Obara for skilful and painstaking help in paraffinsectioning, photography and animal care. Mice used for thisstudy were raised at the Center for Experimental Plants andAnimals, Hokkaido University. This study was partiallysupported by the Grant-in-Aid for General Scientific Re-search, Grant-in-Aid for Special Project Research and Grant-in-Aid for Scientific Research on Priority Area from Ministryof Education, Science and Culture, Japan.

References

BARTON, S. C , ADAMS, C. A., NORRIS, M. L. AND SURANI, M. A.

H. (1985). Development of gynogenetic and parthenogeneticinner cell mass and trophectoderm tissues in reconstitutedblastocysts in the mouse. J. Embryol. exp. Morph. 90, 267-285.

BELOTE, J. M. AND LUCCHESI, J. C. (1980). Male-specific lethalmutations of Drosoplula melanogaster. Genetics 96, 165-186.

BENNETT, D. AND DUNN, L. C. (1958). Effects on embryonicdevelopment of a group of genetically similar lethal allelesderived from different populations of wild house mice. J. Morph.103, 135-157.

CATTANACH, B. M. (1975). Control of chromosome inactivation.Ann. Rev. Genet. 9, 1-18.

ENDO, S., TAKAGI, N. AND SASAKI, M. (1982). The late-replicating

X chromosome in digynous mouse triploid embryos. Develop.Genet. 3, 165-176.

EPSTEIN, C. J. (1969). Mammalian oocytes: X-chromosome activity.Science 163, 1078-1079.

GARDNER, R. L. AND LYON, M. F. (1971). X-chromosomeinactivation studied by injection of a single cell into the mouseblastocyst. Nature 231, 385-386.

GARTLER, S. M. AND RIGGS, A. D. (1983). Mammalian X-chromosome inactivation. Ann. Rev. Genet. 17, 155-190.

HAMADA, H., SATO, M., MURATA, F. AND MURAMATSU, T. (1983).

Differential expression of lectin receptors in germ layers of themouse egg cylinder and teratocarcinoma. Expl Cell Res. 144,489-495.

HOCKEY, A. J., ADRA, C. N. AND MCBURNEY, M. W. (1989).Reactivation of hprt on the inactive X chromosome with DNAdemethylating agents. Somat. Cell Genet. IS, 421-434.

LEWIS, S. E., TURCHIN, H. A. AND GLUECKSOHN-WAELSCH, S.

(1976). The developmental analysis of an embryonic lethal (c6H)in the mouse. J. Embryol. exp. Morph. 36, 363-371.

LUCCHESI, J. C. AND SKRIPSKY, T. (1981). The link between dosagecompensation and sex determination in Drosoplula melanogaster.Chromosoma 82, 217-227.

LYON, M. F. (1961). Gene action in the X-chromosome of themouse (Mus musculus L.). Nature 190, 372-373.

LYON, M. F. AND RASTAN, S. (1984). Parental source ofchromosome imprinting and its relevance for X-chromosomeactivation. Differentiation 26, 63-67.

LYON, M. F., SEARLE, A. G., FORD, C. E. AND OHNO, S. (1964). A

mouse translocation suppressing sex-linked variegation.Cytogenetics 3, 306-323.

MAGNUSON, T. (1986). Mutations and chromosomal abnormalities:How are they usuful for studying genetic control of earlymammalian development. In Experimental Approaches to

Mammalian Embryonic Development (ed. J. Rossant and R. A.Pedersen), pp. 437-474. Cambridge: Cambridge Univ. Press.

MARTIN, G. R., EPSTEIN, C. J., TRAVIS, B., TUCKER, G., YATZIV,

S., MARTIN, D. W., JR, CLIFT, S. AND COHEN, S. (1978). X-chromosome inactivation during differentiation of femaleteratocarcinoma stem cells in vitro. Nature 271, 329-333.

MCBURNEY, M. W. AND STRUTT, B. J. (1980). Genetic activity of Xchromosomes in pluripotent female teratocarcinoma cells andtheir differentiated progeny. Cell 21, 357-364.

MCLAREN, A. (1976). Genetics of the early mouse embryo. Ann.Rev. Genet. 10, 361-388.

MCMAHON, A. AND MONK, M. (1983). X-chromosome activity infemale mouse embryos heterozygous for Pgk-I and Searle'stranslocation, T(X;16)16H. Genet. Res. 41, 69-83.

MIYABARA, S., GROPP, A. AND WINKING, H. (1982). Trisomy 16 in

the mouse fetus associated with generalized edema andcardiovascular and urinary tract anomalies. Teratology 25,369-380.

MONK, M. (1981). A stem-line model for celluar and chromosomaldifferentiation in early mouse development. Differentiation 19,71-76.

NISWANDER, L., YEE, D., RINCHIK, E. M., RUSSELL, L. B. ANDMAGNUSON, T. (1988). The albino deletion complex and earlypostimplantation survival in the mouse. Development 102, 45-53.

NISWANDER, L., YEE, D., RINCHIK, E. M., RUSSELL, L. B. AND

MAGNUSON, T. (1989). The albino-deletion complex in the mousedefines genes necessary for development of embryonic andextraembryonic ectoderm. Development 105, 175-182.

RASTAN, S. (1983). Non-random X chromosome inactivation inmouse X-autosome translocation embryos - location of theinactivation centre. J. Embryol. exp. Morph. 78, 1-22.

SEARLE, A. G. (1962). Is sex-linked Tabby really recessive in themouse? Heredity 17, 297.

SEARLE, A. G., BEECHY, C. V., EVANS, E. P. AND KIRK, M. (1983).Two new X-autosome translocations in the mouse. Cytogenet.Cell Genet. 35, 279-292.

SPIEGELMAN, M., ARTZT, K. AND BENNETT, D. (1976).Embryological study of a T/t locus mutation (tw73) affectingtrophectoderm development. J. Embryol. exp. Morph. 36,373-381.

TAKAGI, N. (1980). Primary and secondary nonrandom Xchromosome inactivation in early female mouse embryos carryingSearle's translocation T(X;16)16H. Chromosoma 81, 439-459.

TAKAGI, N. AND SASAKI, M. (1975). Preferential inactivation of thepaternally derived X chromosome in the extraembryonicmembranes of the mouse. Nature 256, 640-642.

TAKAGI, N., SUGAWARA, D. AND SASAKI, M. (1982). Regional and

temporal changes in the pattern of X-chromosome replicationduring the early postimplantation development of the femalemouse. Chromosoma 85, 275-286.

THEILER, K. (1972). The House Mouse. Springer-Verlag (Berlin,Heidelberg, New York).

WROBLEWSKA, J. AND DYBAN, A. P. (1969). Chromosomepreparation from embryos during early organogenesis:dissociation after fixation, followed by air drying. Stain Technol.44, 145-150.

(Accepted 12 February 1990)

Documents

Detrimental effects of two active X chromosomes on early ... · difference between male and female. The complete lack of any known mutation that disturbs X-inactivation suggests that