Development 104 Supplement,Printed in Great Britain @ The

rB7 -res (1e88)Company of Biologists Limited 19BB

r87

pressed in a region-specific manner during the forma-tion and differentiation of the embryonic anteropos-terior axis. Although striking patterns of expression ofHox 1.7 and other homeobox genes are seen in overtlysegmented structures of the embryo (i.e. somites,prevertebral elements, neural tube and dorsal spinalganglia) expression is also seen in tissues with noobvious segmental origin. The results suggest thathomeobox genes probably do not play an exclusive rolein segmentation in vertebrates, but are consistent witha role in the assignment of positional identity along theaxis of the embryo.

Key words: homeobox, embryogenesis: gene expression,pattern formation, segmentation.

Expression of homeobox gene Hox 7.7 during mouse embryogenesis

KATHLEEN A. MAHONI, HEINER WESTPHAL1 ANd PETER GRUSS2

rLaboratory of Molecular Genetics, National Institute of Chitd Health and Human Development, National Institutes of Health, Bethesda,Maryland 20892, USA2Max Planck Institute of Biophysical Chemistry, Department of Molecular Cett Biology Am Fassberg, D-3400 Gottingen, FRG

Summary

Many of the genes controlling segmentation and pat-tern formation in Drosophila contain a conserved183 bp sequence known as the homeobox. Homeoboxsequences have been found in a range of metazoanspecies, including the vertebrates mouse and man.This striking conservation suggests that homeoboxgenes may play a fundamental role in developmentalprocesses. If this is the case then it might be expectedthat vertebrate homeobox genes will be differentiallyexpressed during embryogenesis and that the timing oftheir expression will coincide with major morphogen-etic events. Here the spatial and temporal patterns ofexpression of murine homeobox genes will beexplored, concentrating on the Hox 7.7 gene as anexample. Using in situ hybridization to localize RNAtranscripts, it has been found that Hox 7.7 is €X-

lntroduction

Genetic and molecular analyses have revealed thatthe development of the embryonic body plan inDrosophila relies on the proper expression and inter-action of a number of genetic loci (reviewed inGehring & Hiroshi, 1986; Gehring, 1987; Akam,I9B7; and Scott & Carroll, 1987). These genes areinvolved in specifying the body axes and metamericorganization of the embryo, including segment po-larity (segmentation genes) and positional identity(homeotic genes). The important role played by thesegenes has been determined largely through a detailedexamination of the phenotypes of mutant alleles,which in many cases produce quite dramatic alter-ations in the patterning of the embryonic body.

The molecular dissection of several genes from theAntennapedia and Bithorax gene complexes un-covered the existence in each gene of a conserved

183 bp sequence, the homeobox. The homeoboxencodes a putative DNA-binding domain of the helix-turn-helix type (McGinnis et al. I984a,b; Scott &Weiner, 1984), a notion that is supported by the invitro DNA-binding properties of several homeoboxgene products (Desplan et al. 1985), and their nuclearlocalization as established by immunostaining withspecific antibodies against homeodomain peptides(White & Wilcox,1984; Beachy et a|.1985; Carroll &Scott, 1985; Di Nardo et a|.1985). These observationsstrongly support the view that proteins containing thehomeodomain exert their effects by directly regulat-itrg gene expression. The conservation of the homeo-box motif has facilitated the isolation of additionaldevelopmentally important genes from Drosophila.In fact, there appear to be several classes of homeo-box genes that can be distinguished by varianthomeodomain sequences (Gehritrg, 1987; Scott &Carroll, 1987; Rushlow et al. 1987).

188 K. A. Mahon, H. Westphal and P. Gruss

The homeobox conservation extends throughoutthe animal kingdom. It has been found, for example,in the genomes of echinoderffis, annelids and chor-dates (McGinnis et al. I984b; Holland & Hogan,1986). Although to date very little is known about thefunction of homeobox genes in species other thanDrosophila, their presence in such diverse and evol-utionarily distant phyla argues that they play some

conserved role in development.At least L8 genes, the so-called Hox genes (Martin

et Al. I9S7), have been isolated from the mouse thatcontain sequences related to the Antennapedia-typehomeobox. Most, and perhaps all, of these genes ate

located in clusters on at least four different chromo-somes (reviewed in Fienberg et al. 1987; Colberg-Poley et al. 1988; Holland & Hogan, 1988b). Thereappear to be additional genes, such as En-l, and En'2, that have homeoboxes similar in sequence todivergent classes of. Drosophila homeobox genes (inthis case, the Drosophila engrailed gene) (Joyner &Martin , 1987). In addition, several other genes have

been isolated by virtue of homology to conserved

structural motifs present in other genes controllingm.orphogenetic events in Drosophila, such as the'finger domains' from the Krilppel gene (Chowdhuryet al. 1987) and the 'paired box' found in the paired-gooseberry genes (Deutsch et al. 1988). The discoveryof more widespread conservation among several

classes of developmental genes strongly suggests thatsuch homologies are not simply fortuitous and raises

the intriguing possibility that there may be classes ofvertebrate regulatory genes, analogous to those inDrosophila, which are essential for vertebtate devel-opment.

The task of defining the function of the homeoboxgenes in vertebrates is a very challenging one, ?S

presently there are no known existing mutations at

any of these loci. Thus, current analyses focus ondetermining the patterns of gene expression in thehope that a possible function can be deduced. A rolein the specification of positional information, sugges-

ted by analogy to Drosophila homeobox genes,

implies that at least two predictions might be true.The Drosophila genes are restricted in their ex-

pression to specific regions of the embryo, generallycorresponding to the regions or segments defective inmutants of these genes (Akam ,1987; Scott & Catroll,1987). If the homeobox genes of vertebrates playsome role in defining the embryonic body plan, thenthey might also display regional specificity in theirexpression during embryogenesis. This prediction has

recently been verified as several murine homeoboxgenes have been found to be differentially expressed

in the central nervous system of midgestation em-

bryos in distinct, yet overlapping, spatial domains(reviewed by Holland & Hogan, 1988b).

The second prediction is that expression shouldcoincide with the time when determinative events inpattern formation take place. In Drosophila, these

decisions occur very early. In fact, considerablepositional information is present in the egg at the timeof fertilization, and most of the zygotically activesegmentation and homeotic genes are expressed be-fore, of shortly after, cellularization (reviewed inAkam , 1987; Scott & Carroll , 1987).

In the mouse, a highly regulative organisffi, deter-minative decisions appear to be made relatively late.For example, there is no obvious morphologicalpolarity in the fertilized egg and individual blasto-meres remain totipotent or pluripotent through initialcleavage events (reviewed in Rossant & Pedersotr,1986). It is through the process of gastrulation, whichbegins in the mouse around 6'5 days post coitum(p.r.) , that the anteroposterior axis is definitivelyestablished (reviewed in Hogan et al. 1985; Hogan et

a|.1986).Prior to gastrulation, the primitive ectoderffi, the

region that gives rise to all embryonic tissue, exists as

a cylindrical layer of cells with no recognizableanteroposterior polarity. In a region of the primitiveectoderm called the primitive streak, some cellsdelaminate and migrate between the primitive ecto-derm and visceral endoderm. These cells migrateboth anteriorly and laterally to form the mesoderm ofthe embryo proper, and posteriorly, to give rise to theextraembryonic mesoderm lining the exocoelom anda structure known as the allantois. The allantoisultimately fuses with the chorion to form the pla-centa. This entire process initiates inductive tissueinteractions leading to the derivation of neural tissueand organogenesis. One of the most striking features,however, involves the development of the axial struc-tures of the embryo, particularly the sequential parti-tioning of paraxial mesoderm into the somites. Apartfrom dividing the mesoderm into metameric units,giving the embryo a segmented appearance, thesomites appear to impose segmentation on the ner-vous system (Hogan et al. 1985). Once formed, thesomites differentiate into three parts: the derma-tome, the myotome and the sclerotome, whicheventually form the dermis of the skin, the muscles ofthe trunk and the axial skeleton, respectively. Boththe initial deposition and subsequent differentiationof the somites occur in a distinct rostrocaudal se-

quence which takes place over a period of severaldays (Rugh , L968; Theiler, 1972).

Experimental embryology of both the mouse andthe chick, which have very similar developmentalstrategies, indicates that the first evidence for restric-tion in developmental potential along the anteropos-terior axis occurs at gastrulation (Kieny et al. 1972;

Snow, 1981, 1985; Beddington, 1982; Bellairs et Al.

1986). It is therefore important to determine thepattern of expression of the homeobox genes duringthis period of immense morphological change whenthe basic body plan of the embryo is established.

This article will describe the spatial pattern ofexpression of the Hox I .1 gene during mouse em-bryogenesis. Although the data will be described inmore detail elsewhere (Mahon et al. 1989), thegeneral conclusions presented here show that Hox7 .7 is expressed in a region-specific manner aroundthe time the axis of the embryo is formed. Takentogether with the expression patterns of other Hoxgenes ) a strong case can be made for a role forhomeobox-containing genes in determining pos-itional identity.

Spatial patterns of expression of Hox 1.1

The Hox I .1 gene is one of at least seven genes

located in the Hox I cluster on chromosome 6. Itcontains a single intron of 1'1 kb and produces a

2.4kb mRNA transcript. The deduced protein se-

quence encoded by this gene is 229 amino acids long(Kessel et al. 1987). Immunostaining with antibodiesmade against Hox 1.1 peptide indicates that the Hox1 .1 gene product is a nuclear protein (Kessel et al.L987), consistent with the nuclear localization ofmany Drosophila homeobox gene products and theputative DNA-binding properties suggested by pro-tein sequence.

Northern blot analysis of RNA from embryos has

indicated that the peak of. Hox 1./ expression occursat 12.5 days p.c. (Colberg-Poley et al. 1985). Spatialpatterns of expression were analysed by in situhybrid ization of 3ss-1abe11e d Hox I .1 probes to tissuesections of mouse embryos. The results indicated thatHox I .1 is differentially expressed along the antero-posterior axis in I2's-day mouse embryos, primarilyin regions of the developing central and peripheralnervous system and in the axial skeleton.

Hybridization was detectable over the neural tubeand adjacent dorsal spinal ganglia extending from thefourth cervical ganglion (C4) to the lower lumbarregion (Fig. L). The anterior limit of hybridizationappeared quite distinct. However, the intensity ofhybridizationwas variable in both the neural tube andganglia along the longitudinal axis, appearingstrongest in the anterior regions and gradually declin-ing in more posterior regions. Hybridizing regions ofthe neuroectoderm derivatives were largely in align-ment, although the neural tube slightly anterior to C4showed some expression (note arrow in Fig. 1A).

Transcripts also accumulate in condensations ofsclerotome cells that form the primordia of thevertebrae and ribs. As is the case in the neuroecto-

Expression of mouse homeobox gene 189

derm derivatives, the prevertebrae and associated ribprimordia that express Hox I .1 are restricted alongthe anteroposterior axis to a region extending fromthe third thoracic element to the first lumbar element(i.e . I0-20th prevertebrae inclusive) . It is intriguingthat the regions expressing Hox 1.1 tn the neuroecto-derm and in the mesoderm-derived tissues overlapbut are not in complete register (diagrammed inFig. 1A). Interestingly, sclerotome cells immediatelysurrounding the notochord do not express detectablelevels of Hox 7.7 transcripts (Fig. 1BII). In additionto the axial structures, Hox 1.1 is also expressed inthe mesoderm of the stomach at this stage and in themetanephric kidney at later stages.

The expression pattern within both the neural tubeand axial skeleton changes as these tissues differen-tiate. Transcripts in the neural tube, uniformly dis-tributed along the dorsal-ventral axis at I2.5 daysp.c. (Fig. 1), become more prevalent along the dorsalregion in 13.s-day-old embryos (Fig.2A,B), whileexpression in the primordia of the skeletal elementsdiminishes as chondrification centres appear (notshown).

The striking regional specificity characterizing Hox7 .7 expression in these relatively late stages of devel-opment is apparent in earlier stages as well. In 9.5-day embryos, Hox 1.1 transcripts are prominentwithin the neuroectoderm and mesoderm derivativesin the posterior region of the embryo. Expression wasdetectable in the neural tube and neural crest (fromwhich the spinal ganglia and rest of peripheral ner-vous system is derived) as well as in paraxial andlateral plate mesoderm. Expression was also seen inthe mesenchyme of the anterior limb buds. Hybridiz-ation was initially quite strong in the buds at the timewhen they first appear, but became more diffuse byday I2.5 p.c. It is worth noting that the anterior limbbuds originate in a region along the axis spanningseveral somites (somites S-I2) (Theiler, 1972)

within the anteroposterior domain of Hox 7 .7 ex-pression.

Expression of Hox I .1 first becomes apparent inthe posterior ectoderm and mesoderm when theembryo has acquired between 8 and 12 somites(8.0-8.5 days p.c.). However, transcripts weredetected at earlier stages (7.5 days) in extraembry-onic mesoderm (see below). Thus far, transcriptshave not been detected in earlier, preimplantation,embryos.

Tissues expressing Hox 7 .7 in midgestation em-bryos are summaized in Table L. Hox 7.7 is morebroadly expressed in early embryos in both theneuroectoderm and in the mesoderm. Striking, andsomewhat transient, expression patterns evolvewhich ultimately become more restricted as develop-ment proceeds. This is most convincingly evident in

190 K. A. Mahon, H. Westphal and P. Gruss

Fig. L. (A) Schematrc representation of the distribution of" Hox 1.1

transcripts in the dorsal spinal ganglia (g) and sclerotome-derivedprevert ebra (s) in the LZ.5-day-old embryo as determined by an analysis

of serial sections hybridized in situ with 3ss-labelled antisense Hox 1.1

transcripts. Hybridizing regions are indicated by filled and stippled areas.

Transcripts accumulate in the thoracic sclerotomes (T3-L1) and in theganglia beginning at the fourth cervical ganglion (Ca) and extendingcaudally. The rostral limit of hybn dtzatton over the neural tube is slightlyanterior to C4 and is indicated by the arrow. The distribution of grains

over the neural tube is similar to that of. the adjacent ganglia, with thesignal gradually decreasing in more posterior regions (as depicted by thestippled labelling) . Lines I, II and III indi cate planes of sectioning shown

in B.(B) In situ hybridizatton of representative cross-sections along the

planes shown diagrammatically in A. (I.u) Dark-field image of a section

through the midcervical region hybridized to Hox 1.1 antisense

transcripts. Neural tube and spinal ganglia are labelled. (b) An adjacentsection hybridized to Hox 1.1 sense transcripts. No specific labelling isseen. (II) Hybridizatron of Hox 1.1 anti-sense probe to sections taken at

the level of the lung. Theneural tube, spinal gangliaand sclerotome arelabelled. Note thatsclerotome cellsimmediately investing thenotochord are unlabelled(arrow). (a) Dark-fieldillumination. (b) Bright-field illumination.(III) Cross-section throughthe lower thor acic regionhybridized to Hox 1.1 anti-sense probe. In thisregion, the sclerotomes,including rib primordia,are very strongly labelled,whereas there is much less

labelling of the neural tubeand ganglia relative tomore anterior sections.Note that the spinal nerveis unlabelled (arrow).(a) Dark-field illumination.(b) Bright-fieldillumination. Exposuretime: 10 days. s,

sclerotome; S, Eanglion;nt, nettral tub ei ffi,myotome; n, notochord; l,lung; li,ller; h, heart.Mag. 36x.

Expression of mouse homeobox gene 191

Fig.2. Expression of. Hox 1.1 rn the 13.5-day-old embryo, as demonstrated by in situ hybridizatron Midsagittal sectionof a L3.5-day p.c.embryo hybridized with Hox 1.1 probe and photographed under bright-field (A) and dark-field optics(B). The spatial limits of hybrrdtzation along the longitudinal axis are similar to the L2.5 day p.c. embryo. The arrowmarks the most anterior prevertebral element to show hybridizatron Transcripts become restricted to the more dorsalregion of the neural tube at this stage.ht, neural tubei p, prevertebra. Mag. 50x.

Table 1. Embryonic expression of Hox L,1

8-8.5 days p.c. 9.5 days p.c. I2.5 days p.c.

mesoderm H Paraxial -.+somite> lateral plate "--+limb bud

ectodermneural tubeeural crest

-->

sclerotome

-_-_+ limb budstomach

(metanephros*)

--+ neural tube

--+ dorsal signal ganglia

Embryonic germ layers and derivative tissues that express Hox 1.1 at several embryonic stages. Arrows indicate developmentalpathways.

* expressed later than day L2.5.

the mesoderm and its derivatives. In 9.5-day-oldembryos, transcripts are seen in both paraxial andIatercI plate mesoderm. As the somite differentiates,Hox L I expression is progressively turned off in themyotome and dermatome. Expression in the sclero-tome persists until chondrification occurs. Likewise,the initially broad pattern of expression in the lateralplate mesoderm becomes restricted to specific deriva-tives, including the mesoderm of the stomach andmetanephric kidney. It is important to note that thesemodulations in expression occur within a speciflcanteroposterior domain and do not appear to simply

reflect the differentiation state of the axial structures.If this were the case, expression might be expected tocommence, for example, in the most anterior somitesand subsequently proceed like a propagated wavealong the anteroposterior axis over time, in parallelwith the marked rostrocaudal sequence of somitedevelopment.

Anteroposterior domains of homeobox geneexpression

Hox I .1 has a spatial and temporal pattern of ex-

I92 K. A. Mahon, H. Westphal and P. Gruss

pression distinct from that of other homeobox genes.

Transcripts from several genes have been localized byin situ hybridizatron to midgestation (12'5-13'5 dayp.c.) embryos (Awgulewitsch et a|.1986; Gaunt et al.1986; Dony & Gruss, 1987; Gaunt, 1987, 1988; Utset,I9B7; Fains od et al . 1987; Krumlauf et al . 1987 ; Toth et

a|.1987; Holland & Hogzn, I988a; Sharpe et al.1988;Breier et al. 1988; reviewed by Holland & Hogan,1988b). Each gene analysed so far has been expressedin a spatially restricted manner in the developingcentral nervous system (CNS) and, in some cases, in asubset of the prevertebrae and visceral organs. Itappears that expression in the viscera is also restrictedaccording to position of origin along the longitudinalaxis, in a manner consistent with the anteroposteriordomain of expression particular to each gene (Dony& Gruss, 1987; Holland & Hogan, l988a,b; Gaunt,1e88).

Comparison of the axial limits of expression of thevarious homeobox genes logically focuses on theCNS, since transcription is invariably detected there,and since many homeobox genes in Drosophila also

show prominent CNS expression (Scott & Carroll,1987). As is the case with Hox I .1 , the anteriorboundaries of hybrrdization in the CNS are typicallyquite sharp and distinct for each gene, whereas theposterior limits are diffuse and hard to define(reviewed by Holland & Hogan, I988b; Holland,1983). Where expression is detected in the mesoder-mal derivatives such as the prevertebral elements, therostrocaudal boundaries are not necessarily coinci-dent with those in the nervous system. It remains a

distinct possibility that the anteroposterior expressiondomains in the neuroectoderm and mesoderm areestablished (or maintained) independently. It is clear,however, that the homeobox genes constitute a class

of genes, hitherto unknowfl, that are expressed inoverlapping but noncoincident spatial domains alongthe embryonic axis.

Less certain is whether, in general, these distinctgene-specific expression domains are apparent at thetime of initial axis formation, &s the expressionpatterns of only a few genes (Hox 1.1,1.5,2.1, and3.1) have been determined at these early embryonicstages (Gaunt,1987, 1988; Sharpe et al.1988; Holland& Hog?n,I988a; Mahon et al.1988). However, in thecases that have been analysed, it appears that thedifferences in region ahzed expression apparent inlater stages are evident during the time of axis

formation in the late-gastrulating embryo (Gaunt,1987, 1988; Holland & Hogan, 1988a; Mahon et al.1988). Transcripts have not been detected prior tomidgastrulation (7.5 days p.c.), but expression atlevels below the limit of detection cannot be ruledout. The timing of the appearance of the transcripts inthese cases correlates well with the anteroposterior

Hox I .1 Hox 1.5

7.5 day

Hox 1 .1

B

Hox 1.5

L2.5 day

Fig. 3. Schematic representation of the domains ofexpression of homeobox genes Hox 1.1 and Hox 1.5 intwo stages of embryogenesis. Shaded and stippled areas

indicate the hybridizing regions and the relative intensityof hybrtdrzatton observed. (A) In the 7.S-day p.c. embryo(approximately 2-3 somites) , Hox I .1 transcripts aredetectable in the allantois (stippled), whereas Hox 1.5transcripts are found in the posterior region of theembryo as well. a, posterior; p, posterior of embryo.(B) The anterior expression boundaries in the CNS ofLL.S-day p.c.embryos extend from the fourth cervicalganglion (Hox 1.1) or myelencephalon (Hox 1.5) to moreposterior regions of the neural tube. Embryos aredepicted as sagittal sections. Hybridization domains weredetermined by hybridizatton of adjacent sections to eitheranti-sense Hox 1.1 or Hox 1.5 transcripts.

domains of expression seen in older embryos (Gaunt,1988; Mahon et al. 1988).

As an example, shown in Fig. 3 is a schematiccomparison of the expression pattern of Hox 1.1 withthat of the Hox 1.5 gene. Localization of Hox 1.5transcripts has been conducted in two laboratories(Gaunt et al. 1986; Gaunt, 1987, 1988; Fainsod et al.1988), and Hox 1.5 probe was hybridized to parallelsections as a control in the in situ hybridtzationexperiments with the Hox I .1 gene probe describedhere (Mahon et a|.1988). Hox 7.5 has a more rostrallimit of expression along the body axis than Hox 1 .1 .

A

In the CNS of the Lz.s-day p.c. embryo, the boundaryof expression of Hox 1.5 is located in the myelen-cephalon and that of. Hox 7.1 is at the fourth cervicalganglion (Fig. 38). In 7.S-day p.c.embryos, Hox 1.5is already expressed in the posterior region of theembryo, including the primitive streak, ectoderm andmesoderm, and to some extent in the allantois(Fig. 3,A.) . In contrast, transcripts of Hox I . I aredetected in the allantois posterior to the embryo, butnot in the embryonic region (Fig. 3D). Transcripts inthe embryonic axis are found slightly later (day 8).

Studies of grafted and explanted embryonic tissueindicate that there is region alization present in theembryos by midgastrulation as anterior and posteriorregions of the primitive streak tend to give rise toanterior and posterior structures, respectively (Snow,1981; Beddington , 1982; Tam & Beddington, 1987). Ithas been postulated that cells acquire positionalinformation as they ingress through the primitivestreak, supporting the view that the anteroposterioraxis is established sequentially over the period ofgastrulation (Hogan et al. 1985). As initially proposedby Gaunt (1987), if there is a rostrocaudal sequenceto cellular determination along the body axis, geneswith more anterior limits of hybridization, such as

Hox 7.5, might be expressed at earlier times indevelopment than those, like Hox 1.1, with moreposterior boundaries. Transcripts from another pos-teriorly expressed gene , Hox 3.1 , appear at approxi-mately the same time and place as Hox 1.1 (Gaunt,1e88).

Thus, regionally localized homeobox gene tran-scripts appear during the period when positionalidentities are thought to be established in the mouseembryo. Expression of homeobox genes has not beendetected prior to midgastrulation. This argues thatthese genes are not involved in initiating majormorphological events in axis formation, such as

gastrulation and neural induction. Rather, these ex-pression patterns are more consistent with a role inregion-specific patterning along the embryonic axis.It will be essential for the spatial and temporalpatterns of expression of other homeobox genes to bedetermined in order to ascertain if this early region-specificity is a general feature of homeobox gene

expression.

Do homeobox genes play a role insegmentation?

The homeobox genes belong to a class of genes whosedefining feature is region-specific expression.Although the Hox genes display cell-type-specificexpression as well, this expression depends primarilyupon position. Thus, two seemingly identical groups

Expression of mouse homeobox gene I93

of cells can differ significantly in the homeobox genetranscripts expressed depending upon their positionsalong the axis.

There are certain general features that the Dros-ophila and mouse homeobox genes share. Both aredifferentially expressed along the anteroposterioraxis during development. At least some genes areexpressed initially in broad patterns which eventuallyoccupy more restricted domains in subsequent devel-opmental stages. Different homeobox genes havedistinct, and sometimes overlapping , re1ionaldomains of expression. This strongly suggests thatthere may be combinatorial interactions amongmouse homeobox gene products similar to thoseproposed for Drosophila.

What is the nature of the role played by homeoboxgenes in vertebrate development? The mouse homeo-box genes, including both Hox and engrailed (Daviset al. 1988) loci, and many Drosophila genes, areconspicuously expressed in the developing CNS,suggesting that they may play a conserved role in thedevelopment andf or diversification of cell types inthe nervous system. However, since many of theDrosophila genes have a clear role in establishing themetameric body plan, the striking expression ofmurine homeobox genes in segmented structures ofthe embryo suggests that they may play a fundamen-tal role in segmentation of the embryo. As has beendiscussed (Holland & Hogan, 1988 a,b; Holland,1988) , zn exclusive role in vertebrate segmentationseems unlikely. First, expression has been detected inseveral visceral tissues with no apparent segmentalorigin. Secoild, none of the Hox genes have beenfound to be expressed in all of the segmented struc-tures along the axis, and consequently cannot beresponsible for generating the segmental pattern. Itseems more likely that homeobox gene productsassign region-specific positional identity among thesegmented structures along the rostrocaudal axis.

The hypothesis that homeobox genes specify pos-itional identity along the anteroposterior axis seemsconsistent with the bulk of the available data. How-ever, some expression patterns, particularly in adulttissues ) are hard to reconcile with this hypothesis(Holland & Hogan, 1988a). Of course, in analysingRNA distribution, important questions concerningthe distribution, longevity and state of modification ofthe protein products remain unanswered and meritfurther study. It seems likely, that the homeoboxgenes may play quite different roles in differentcellular contexts and in combination with differentsets of homeobox gene products. The observationthat embryonic and neural activities of the Dros-ophila ftz gene can be distinguished (Doe et al.1988) suggests that these functions may be verycomplex indeed.

I94 K. A. Mahon, H. Westphal and P. Gruss

References

Arau, M. (1987). The molecular basis for metamericpattern in the Drosophila embryo . Development l0l,1,-22.

AwcurEwITSCH, A., lJtsnr, M. F., He.nt, C. P.,McGINNIs, W. & RuonrE, F. H. (1986). Spatialrestriction in expression of a mouse homoeo box locus

within the central nervous system. Nature, Lond. 320,

328-335.BBacHy, P. A., HnLFAND, S. & HocNBss, D. S. (1985).

Segmental distribution of bithora.r complex proteins

during Drosophila development . Nature, Lond. 313,

545-55r.BpontNcroN, R. S. P. (1982). An autoradiographic

analysis of tissue potency in different regions of the

embryonic ectoderm during gastrulation in the mouse.

J. Embryol. exp. Morph. 69,265-285.Bnrratns, R., EDE, D. A. & Lasu, J. (eds) (1986).

Somites in Developing Embryos. vol. 118. New York:Plenum Press.

BnBrnn, G., DnnssLER, G. R. & Gnuss, P. (1988).

Primary structure and developmental expression

pattern of Hox 3.1, a member of the murine Hox 3'homeobox gene cluster . EMBO J. 7, 1329-1336.

Cannorl, S. B. & Scorr, M. P. (1985). Locahzation ofthe fushi-tarasu protein during Drosophilaembryogenesis . CeIl 43, 47 -57 .

CuowoHURy, K., DeurscH, LJ. & Gnuss, P. (L987). Amultigene family encoding several "finger" Structures is

present and differentially active in mammaliangenomes. Cell 48, 77t-778.

CorsBnc-Porey, A. M., Voss, S. D., CnowDHURY, K.,Srswlnr, C. L., WncNER, E. F. & Gnuss, P. (1985).

Clustered homeo boxes are differentially expressed

during murine development. Cell 43,39-45.CornBnc-Porny, A. M., Voss, S. D. & Gnuss, P. (1988).

Homeobox genes of the mouse. In Oxford Surveys

Eukaryotic Genes, vol. 4 (ed. N. Maclean), pp.92-II5.Oxford: Oxford University Press.

Dlvrs, C. A., Nosrn-Topunna, S. E., RossANr, J. &JovNBn, A. L. (1988). Expression of the homeobox-containing gen e En-2 delineates a specific region of the

developing mouse brain. Genes and Development 2,

361.-371,.

DBsprnN, C., THuS, J. & O'FARRELL, P. H. (1985). The

Drosophila developmental gene , engrailed, encodes a

sequence specific DNA binding activity. I{ature, Lond.318 , 630-635.

DBurscH, U., DnnssLER, G. R. & Gnuss, P. (1988). Pax

I, & member of a paired box homologous murine gene

family, is expressed in segmented structures duringdevelopment . Cell 53, 617-625.

DrNnnoo, S., KuNnn, J. M., THErs , J. & O'FnnnEtt, P.

(1985). Development of the embryonic pattern inDrosophila melanogaster as revealed by accumulationof the nuclear engrailed protein. Cell 43, 59-69.

Dor, C. Q., Hrnovu, Y., GEHnING, W. J. & Go9DMAN,

C. S. (1988). Expression and function of thesegmentation gene fushi tarazu during Drosophilaneurogenesis . Science 239, I70-I75.

DoNv, C. & Gnuss, P. (1987). Specific expression of theHox 1.3 homeo box gene in murine embryonicstructures originating from or induced by themesoderm. EMBO J. 6,2965-2975.

FnrNsoD, A., AwcuI-EwlTsctt, A. & RUDDLE, F. H.(1987). Expression of the murine homeo box gene Hox1.5 during embryogenesis. Devl Biol. 124, I25-I33.

FreNnEnc, A. A., IJtsEr, M. F., Bocan^Lo, L. D., HRnt,C. P., AwcuLEwITScH, A., FnncusoN-SulrH, A.,FnNsoD, A., RnntN,.M. & RuDDLE, F. H. (1987).

Homeo box genes in murine development. Curr.Topics. devl Biol. 23,233-256.

G^q,uNr, S. J. (1987). Homoeobox gene Hox-1'.5

expression in mouse embryos: earliest detection by insitu hybridization is during gastrulation. Development101 , 51-60.

GnuNr, S. J. (1988). Mouse homeobox gene transcriptsoccupy different but overlapping domains in embryonicgerm layers and organs: a comparison of Hox-3.I and

Hox-I.5 . Development 103, 135-1'44.

GnuNr, S. J., MttlEn, J. R., Powett, D. J. & DunoutE,D. (1986). Homeobox gene expression in mouse

embryos varies with position by the primitive streakstage. Nature, Lond. 324, 662-664.

GnnnINc, W. J. (1987). Homeoboxes in the study ofdevelopment . Science 236, 1245-1252.

GpuruNG, W. J. & HInour, Y. (1986). Homeotic genes

and the homeobox. A. Rev. Genetics 20, 1'47-173.

HocnN, B. L. M., CosrANrINI, F. & Lacv, E. (1986).

Manipulating the Mouse Embryo: a LaboratoryManual. New York: Cold Spring Harbor Laboratory.

HocaN, B. L. M., HotrAND, P.W.H. & Scuonntn, P.

N. (1985). How is the mouse segmented? Trends

Genet. l, 67-74.HorrnND, P. W. H. (1988). Homeobox genes and the

vertebrate head. Development 103 Supplement, L7-24.HorraND, P. W. H. & HocaN, B. L. M. (1986).

Phylogenetic distribution of Antennapedia-like homoeoboxes . Nature, Lond. 321', 251,-253 .

HorrnND, P. W. H. & HocnN, B. L. M. (19884).

Spatially restricted patterns of expression of thehomeobox-containing gene Hox 2.1 fiiringembryogenesis . Development 102, 1.59-174.

HonnND, P. W. H. & HocnN, B. L. M. (1988b).

Expression of homeobox genes during mouse

development: a review. Genes and Development 2,

773-782.JovNsn, A. L. & MnnrtN, G. R. (1987) . En-L and En-2,

two mouse genes with sequence homology to theDrosophila engrailed gene: expression duringembryogenesis . Genes and Development l, 29-38.

KBssBr, M., ScuuLZE, F., FIu, M. & Gnuss, P. (1987).

Primary structure and nuclear locahzation of a murinehomeo domain protein. Proc. natn. Acad. Scl. U.S.A.84, 5306-5310.

KreNy, M., MnucsR, A. & SsNcEt, P. (1972). Earlyregionahzation of the somitic mesoderm as studied bythe development of the axial skeleton of the chickembryo. Devl Biol. 28, 142-150.

KnuurAUF, R., HottAND, P. W. H., McVpv, J. H. &HocnN, B. L. M. (1987). Developmental and spatial

patterns of expression of the mouse homeobox gene

Hox 2.1,. Development 99, 603-617 .

MauoN, K. A., ScHULZE, F., Wpsrpuat, H. & Gnuss, P.

(1989). Temporal and spatial patterns of expression ofhomeobox gene Hox 1.1 during mouse embryogenesis.(submitted).

ManrrN, G. R., BoNCINELII, E., DunouLE, D., GRUss,

P., JncKSoN , L, KnuurAUF, R., LoNRI, P., McGINNIS,

W., RuoprE, F. & WorcEMUrH, D. (1987).

Nomenclature for homeo-box-containing genes . Nature,Lond. 325,2I-22.

McGrNNrs, W., G,qngBR, R. L., Wrxz, J., KunoIwA, A. &GsunrNc, W. J. (I984b). A homologous protein codingsequence in Drosophila homeotic genes and itsconservation in other metazoans . Cell 37 , 403-409.

McGrNNrs, W., LBvtNs, M. S., HnrBN, E., KunoIwA, A.& GBHRING, W. J. (1984a). A conserved DNAsequence in homoeotic genes of the DrosophilaAntennapedia and bithora.r complexes . Nature, Lond.308 , 428-433.

RossnNr, J. & PnoERSEN, R. A. (eds) (1986).

Experimental Approaches to Mammalian EmbryonicDevelopment. Cambridge: Cambridge University Press.

RucH, R. (1968) . The Mouse: Its Reproduction andDevelopment. Minneapolis : Burgess.

RusHrow, C., DoyLE, H., Hony, T. & LEvrNE, M.(1987). Molecular characterrzation of the zerknillltregion of the Antennapedia gene complex inDrosophila. Genes and Dev.l, t268-I279.

Scorr, M. P. & CIRRoLL, S. B. (1987). The segmentationand homeotic gene network in early Drosophiladevelopment . Cell 5L, 689-698.

Scorr, M. P. & WBINER, A. J. (1984). Structural

Expression of mouse homeobox gene 195

relationships among genes that control development:sequence homology between the Antennapedia,Ultrabithorax and fushi tarazu loci of Drosophila. Proc.natn. Acad. Sci. U.S.A. 81, 41.I5-4II9.

Ssanpn, P. T., MrnnR, J. R., EvRNS, E. P.,BunrsNsHAw, M. D. & GnuNr, S. J. (1988). Isolationand expression of a new mouse homeobox gene.Development 102,, 397 -407 .

SNow, M. H. L. (1981). Autonomous development ofparts isolated from primitive-streak-stage mouseembryos. Is development clonal? J. Embryol. exp.Morph. 65 Supplement, 269-287.

SNow, M. H. L. (1985). The embryonic cell lineage ofmammals and the emergence of the basic body plan. InMolecular Determinants of Animal Form, (ed. G. M.Edelmar), pp.73-98. New York: Alan R. Liss Inc.

Tnu, P. P. L. & BpoDrNGroN, R. S. P. (1987). Theformation of mesodermal tissues in the mouse embryoduring gastrulation and early organogenesis.D ev elopment 99, I09 -126.

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

Toru,L. E., SrnwrN, K. L., PrNrAR, J. E. & NcuyEN-Huu, M. C. (1987). Region-specific expression ofmouse homeobox genes in the embryonic mesodermand central nervous system . Proc. natn. Acad. Sci.

u.s.A. 84, 6790-6794.Llrsnr, M. F., AwculEwrrscH, A., Ruoore, F. H. &

McGrNNrs, W. (1987). Region-specific expression oftwo mouse homeo box genes. Science 235, 1379-1382.

Wutre, R. A. H. & Wltcox, M. (1984). Protein productsof the bithora.r complex in Drosophila. Cell 391,63-17r.