Homeogene expression patterns and chromosomal imprinting.

In both mouse and Drosophila, Antennapedia-like homeobox-containing genes (homeogenes) display a strict correspondence between the order of genes (3' to 5') along the chromosome and the order of their expression domains (anterior to posterior) in the developing embryo. We show here how this, and other points of similarity, may indicate that both species use a common mechanism of chromosomal imprinting in order to retain cellular memory of homeogene expression patterns throughout embryonic development.

Evidence that Hoxa expression domains are evolutionarily transposed in spinal ganglia, and are established by forward spreading in paraxial mesoderm.

Transposition of anatomical structures along the anteroposterior axis has been a commonly used mechanism for changing body proportions during the course of evolutionary time. Earlier work (Gaunt, S.J., 1994. Conservation in the Hox code during morphological evolution. Int. J. Dev. Biol. 38, 549-552; Burke, A.C., Nelson, C.E., Morgan, B.A., Tabin, C., 1995. Hox genes and the evolution of vertebrate axial morphology. Development 121, 333-346) showed how transposition in mesodermal derivatives (vertebrae) could be attributed to transposition in the expression of Hox genes along the axial series of somites. We now show how transposition in the segmental arrangement of the spinal nerves can also be correlated with shifts in the expression domains of Hox genes. Specifically, we show how the expression domains of Hoxa-7, a-9 and a-10 in spinal ganglia correspond similarly in both mouse and chick with the positions of the brachial and lumbosacral plexuses, and that this is true even though the brachial plexus of chick is shifted posteriorly, relative to mouse, by seven segmental units. In spite of these marked species differences in the boundaries of Hoxa-7 expression, cis regulatory elements located up to 5 kb upstream of the chick Hoxa-7 gene showed much functional and structural conservation with those described in the mouse (Puschel, A.W., Balling, R., Gruss, P., 1991. Separate elements cause lineage restriction and specify boundaries of Hox-1.1 expression. Development 112, 279-287; Knittel, T., Kessel, M., Kim, M.H., Gruss, P., 1995. A conserved enhancer of the human and murine Hoxa-7 gene specifies the anterior boundary of expression during embryonal development. Development 121, 1077-1088). We also show that chick Hoxa-7 and a-10 expression domains spread forward into regions of somites that are initially negative for the expression of these genes. We discuss this as evidence that Hox expression in paraxial mesoderm spreads forward, as earlier found for neurectoderm and lateral plate mesoderm, in a process that occurs independently of cell movement.

Conservation in the Hox code during morphological evolution.

The expression domains in paraxial mesoderm of the chicken embryo are described for Hoxb-3, a-4 and c-6 genes, and these are compared with published expression data for the corresponding genes in the mouse. In both species, it is found that the anterior limits of Hoxb-3 and a-4 expression lie in the upper cervical region, and the anterior limits of Hoxc-6 expression lie in the upper thoracic region. This finding is remarkable because the cervical region, or neck, of the chicken (with fourteen cervical vertebrae) is much longer than that of the mouse (seven cervical vertebrae). The results suggest that the Hox code, at least in the development of homologous axial structures, is conserved between species (Hoxb-3 and a-4, for example, being associated with an anterior cervical phenotype; Hoxc-6 being associated with an anterior thoracic phenotype). The results also suggest that an evolutionary change in body proportions is accomplished by a shift in the relative positions of Hox expression domains during embryonic development.

Evolutionary shifts of vertebrate structures and Hox expression up and down the axial series of segments: a consideration of possible mechanisms.

The term 'transposition' describes how, during vertebrate evolution, anatomical structures have shifted up or down the axial series of segments. For example, the neck/thorax junction and the position of the forelimb in the chicken have shifted posteriorly, relative to mouse, by a distance of seven somites or vertebrae. By examining the expression boundaries of some chick Hox genes not previously described, we provide new evidence that axial shifts in anatomical structures correspond with shifts in Hox expression domains. These shifts occur both in mesodermal components (somites, vertebrae, and lateral plate mesoderm) and neural components (spinal ganglia). We discuss morphogen gradient, timing, spreading, and growth models for the setting of Hoxexpression boundaries, and consider how evolutionary shifts in boundary positions might have been effected in terms of these models.

Homoeobox gene Hox-1.5 expression in mouse embryos: earliest detection by in situ hybridization is during gastrulation.

We showed earlier (Gaunt, Miller, Powell & Duboule, 1986) that the mouse homoeobox gene Hox-1.5 is expressed in posterior ectoderm and mesoderm of 7 1/2- and 7 3/4-day embryos, and in the 9 1/2-day nervous system posterior to a discrete boundary within the hindbrain. In further in situ hybridization experiments, it is now shown that restriction of Hox-1.5 expression to the posterior regions of the embryo can be detected at stages of development between 7 1/2 and 9 1/2 days. During this period, the intensity of transcription in presomitic and somitic mesoderm declines relative to that in the overlying neural ectoderm, and the transcription boundary within the presumptive hindbrain region sharpens. Hox-1.5 expression posterior to the hindbrain boundary is detected in the 10 1/2- and 12 1/2-day embryo, but this is no longer found in newborn mice. Embryos of ages 3 1/2, 6 1/2 and 7 1/4 days showed no evidence of Hox-1.5 transcripts. It is concluded that embryos undergoing gastrulation (at 7 1/2 days) are the earliest stage at which Hox-1.5 transcripts can be detected by the in situ hybridization technique. In discussion, it is shown how this lies within the period of development during which tissues become determined along the anteroposterior axis of the mouse. Since there may be anterior-to-posterior variation in the time of determination along the body axis, it is suggested that homoeobox genes expressed more posteriorly, such as Hox-3 (Awgulewitsch et al. 1986), might start expression at times later in development.

Mouse homeobox gene transcripts occupy different but overlapping domains in embryonic germ layers and organs: a comparison of Hox-3.1 and Hox-1.5.

By use of in situ hybridization experiments, the mouse homeobox genes Hox-3.1 and Hox-1.5 are compared in the temporal and spatial patterns of their embryonic transcripts. Transcripts of both genes are first detected at about 7 1/2 days, although the appearance of Hox-3.1 transcripts apparently follows Hox-1.5 after a small delay. Hox-3.1 and Hox-1.5 transcripts occupy domains which are different, although overlapping, along the anteroposterior axis of the embryo. The domains are first established within the ectoderm and mesoderm germ layers at 7 1/2-8 days, but subsequently they persist within the nervous system, the prevertebral column and within at least some of the organs (the thyroid, lung, stomach, mesonephric and metanephric kidneys) at 12 1/2 days. In discussion, two different mechanisms are considered by which positional information might first be generated within the germ layers.

Spreading of a sperm surface antigen within the plasma membrane of the egg after fertilization in the rat.

The rat sperm surface antigen 2D6, located over the entire surface of the spermatozoon, is shown by use of a monoclonal antibody in indirect immunofluorescence experiments to spread laterally over the surface of the egg after fusion of sperm and egg plasma membranes at fertilization. Freshly fertilized eggs, obtained from superovulated rats 14 h after hCG injection, showed the 2D6 antigen to have spread in a gradient over a discrete fan-shaped area of the egg surface anterior to the protruding sperm tail. Eggs at a later stage of sperm incorporation, obtained 20 h after hCG injection, showed that the spread of antigen had extended to cover most or all of their surfaces. By 40 h after hCG injection, the approximate time that fertilized eggs cleaved to form 2-cell embryos, most of the 2D6 antigen had been lost from the cell surface. Fertilized eggs, but not unfertilized eggs or 2-cell embryos, were lysed by 2D6 monoclonal antibody in the presence of guinea pig complement. A model for sperm-egg fusion is presented to account for the observed pattern of spreading shown by the 2D6 antigen. The possible role of sperm antigens on the egg surface is discussed.

Expression patterns of mouse Hox genes: clues to an understanding of developmental and evolutionary strategies.

Expression patterns of Antennapedia-like homeogenes in the mouse embryo show many similarities o those of their homologues in Drosophila. It is argued here that homeogenes may regulate development of the body plan in mouse by mechanisms similar to those used in Drosophila. In particular, they may differentially specify positional address of cell groups within lineage compartments along the body axes. In vertebrates, a single ancestral homeogene cluster has become duplicated to give four separate clusters. Comparisons of homeogene expression patterns between different clusters of the mouse suggest ways in which duplication has permitted development of a more complex body plan. Cluster duplication may therefore have provided a selective advantage during vertebrate evolution.

Gradients and forward spreading of vertebrate Hox gene expression detected by using a Hox/lacZ transgene.

Elucidation of the kinetics with which vertebrate Hox expression patterns develop may help us to choose between various models already proposed to explain this process. The chick Hoxa-7/lacZ transgene, expressed in mouse embryos, changes over time in the distribution of its activity along the developing posterior to anterior axis. During an establishment (E) phase (lasting at least up to 10 days) expression is graded from highest levels posteriorly, to low levels anteriorly. Within the graded domain, the overall level of expression spreads forward with time along both neurectoderm and paraxial mesoderm. Spreading in expression is not due to movement of cells, but to increases in both the proportion of lacZ expressing cells and the intensity of expression per cell. By 10.8 days, embryos have reached a late (L) phase in which an anterior up-regulation in expression, together with a posterior down-regulation, cause the graded nature of the expression to be lost. E and L phases are also seen for Hox gene expression detected by in situ hybridization. The switch from E to L occurs at progressively later times as we move 3' to 5' along the Hox cluster. The results are in keeping with models in which Hox genes become differentially expressed according to a graded concentration of an inducer. Binding motifs for the caudal (cdx) proteins, already proposed as such inducers, are conserved in mouse and chick Hoxa-7 enhancer elements.

In vivo and in vitro cultured mouse preimplantation embryos differ in their display of a teratocarcinoma cell surface antigen: possible binding of an oviduct factor.

By use of a monoclonal antibody, 2B5, in indirect immunofluorescence experiments, it was found that both fertilized and unfertilized mouse eggs obtained directly from the oviduct commenced expression of a cell surface antigen at about 5 h after ovulation. Surface labelling became intense by 16 h after ovulation and persisted over all blastomeres throughout preimplantation development. In contrast, embryos cultured in vitro did not appearance of 2B5 antigen until about 48 h after ovulation, at which time they were at the 2- to 4-cell stage. Antigen expression in vitro commonly began on a single blastomere and did not appear consistently over all blastomeres until the 8-cell stage (72 h after ovulation). Unfertilized eggs maintained for 72 h in culture did not acquire 2B5 antigen. It is postulated that the absence of 2B5 antigen on 1-cell eggs cultured in vitro may be due either to a failure of normal synthesis by eggs as a result of a deficiency in the culture medium, or alternatively, to absence of a soluble oviduct factor which carries the 2B5 antigen, and which normally becomes bound to the surface of eggs after ovulation. The second of these two possibilities was supported by egg transfer experiments which showed that unfertilized eggs within the oviduct became 2B5 antigen-positive even after their prior fixation in glutaraldehyde. By the 2- to 4-cell stage, however, embryos developed their own capacity for synthesis of 2B5 antigen-positive cell surface molecules. This synthesis was inhibited by tunicamycin, suggesting that the antigenic site involved the sugar component of glycoprotein. The range of tissues within the postimplantation embryo and adult reproductive tracts which labelled with 2B5 antibody was found to be very similar to that known for SSEA-1 monoclonal antibody (Solter & Knowles, 1978; Fox et al. 1981; Fox, Damjanov, Knowles & Solter, 1982), and as further evidence of a relationship between 2B5 and SSEA-1 antigens it was found that 125I SSEA-1 antibody could be blocked in its binding to teratocarcinoma cells by preincubation in 2B5 monoclonal antibody.

Multiple forms of ram and bull sperm hyaluronidase revealed by using monoclonal antibodies.

Two monoclonal anti-sperm hyaluronidase-producing cell lines were isolated following inoculation of mice with ram sperm hyaluronidase monomer. Both lines produced antibodies of the IgG1 class; these bound to ram hyaluronidase after 'Western blotting' but did not recognize the native enzyme. Whereas the 1A4 antibody was specific for ram hyaluronidase, and did not react with 'blotted' bull, boar or rabbit hyaluronidase, the 1D6 antibody recognized bull as well as ram hyaluronidase. The antibodies could be used for immunocytochemical localization of hyaluronidase in fixed spermatozoa. However, although some form of denaturation was required to unmask or form the epitopes with which the antibodies reacted, the degree and type of fixation required was critical, for the epitopes were readily destroyed; in particular, they were very sensitive to chemical modification such as glutaraldehyde treatment. It could be demonstrated that, like ram, bull spermatozoa contained an extended oligomeric family of hyaluronidase forms, apparently the result of intermolecular disulphide cross-linking of monomers. In spermatozoa of both species, the enzyme was confined to the anterior acrosomal region of the head.

Temporal colinearity in expression of anterior Hox genes in developing chick embryos.

Temporal colinearity describes a correspondence between the spatial ordering of Hox genes within their clusters (in the direction 3' to 5') and the time of their first expression (earlier to later) during embryonic development (Izpisúa-Belmonte et al. [1991] EMBO J. 10:2279-2289). It suggests that activation of each Hox gene might be controlled in some way by its position within the cluster. So far, in situ hybridization experiments on vertebrate embryos have provided clear evidence of temporal colinearity only for "posterior" Hox genes (5' located, AbdB related). We now describe a search in the chick embryo for evidence of temporal colinearity in the expression of some anterior Hox genes (Hoxb-1, b-3, b-4, b-6, and a-9). Clear evidence for temporal colinearity was seen in neural tube tissue adjacent to the first few somites. Here, there were delays in the expression of Hoxb-3 following b-1, Hoxb-4 following b-3, and Hoxb-6 following b-4. Temporal colinearity was also detected in anterior primitive streak tissue. Hox gene expression reached both the neural tube and the anterior streak following forward spreading from posteriormost parts of the primitive streak. Overall, therefore, temporal colinearity was seen as sequential waves of Hox genes expression that proceeded forward (3' genes before 5' genes) along the developing chick embryo. Within posterior primitive streak tissue, there was only limited evidence for temporal colinearity. We discuss these results in terms of possible models for the establishment of Hox gene expression patterns.

Forward spreading in the establishment of a vertebrate Hox expression boundary: the expression domain separates into anterior and posterior zones, and the spread occurs across implanted glass barriers.

By use of wholemount in situ hybridization, we show how expression of the chicken homeobox gene Hoxd-4 commences in the posterior part of the primitive streak and then spreads forward, covering most of the primitive streak by the 2 somite stage, covering the entire primitive streak by the 5 somite stage, reaching the somite 1/somite 2 level of the neural tube by the 9 somite stage, and reaching the rhombomere 6/rhombomere 7 junction of the hindbrain by the 15 somite stage. Forward spreading does not depend upon cell migration, as was evidenced by vital dye (DiI) cell marking experiments. Furthermore, forward spreading does not apparently require tissue continuity since it could not be blocked by impermeable (glass) barriers surgically implanted to divide embryonic tissues. As forward spreading of chick Hoxd-4 proceeds, the domain of expression separates, at late primitive streak stages, into "anterior" and "posterior zones," with an intervening "intermediate zone" of weak or non-expression. Clear anterior and posterior zones were also found for Hoxa-3 and a-4 expression in late primitive streak stage mouse embryos. We present evidence that the anterior zone corresponds with the "definitive" domain of Hox gene expression, as has earlier been extensively characterized in midgestation embryos. The posterior zone is transitory, probably persisting only for the duration of the primitive streak, and it is a region of intense Hox expression in primitive streak tissue, Hensen's node, and adjacent regions of neurectoderm and mesoderm. We suggest that the posterior zone marks the source of a morphogen which is the primary activator of Hox gene expression, and we discuss possible models for the mechanism of forward spreading in expression.

Metabolic co-operation between embryonic and embryonal carcinoma cells of the mouse.

Mouse embryonal carcinoma (EC) cells form permeable junctions at their homotypic cell-to-cell contacts which permit intercellular exchange of metabolites (metabolic co-operation). Hooper & Slack (1977) showed how this exchange could be detected by autoradiography as the transfer of [3H]nucleotides between PC13 (a pluripotential EC line) and PC13TG8 (a variant of PC13 which is deficient in hypoxanthine guanine phosphoribosyltransferase). We now show that cells taken from several different tissues of early mouse embryos, that is, from the morula, the inner cell mass of the blastocyst, and the endoderm, mesoderm and embryonic ectoderm of the 8th day egg cylinder, are able to serve as donors of [3H] nucleotides to PC13TG8. In contrast, trophectodermal cells of cultured blastocysts, and the trophectodermal derivatives in the 8th day egg cylinder, that is, extra-embryonic ectoderm and ectoplacental cone cells, showed little or no metabolic co-operation with PC13TG8. With reference to some common properties of EC and embryonic cells, we suggest how our findings may provide insight into cell-to-cell interactions in the early mouse embryo.

Expression of the mouse goosecoid gene during mid-embryogenesis may mark mesenchymal cell lineages in the developing head, limbs and body wall.

After an earlier, transient phase of expression in the developing primitive streak of 6.4- to 6.8-day mouse embryos, the homeobox gene goosecoid is now shown to be expressed in a later phase of mouse development, from 10.5 days onwards. The later, spatially restricted domains of goosecoid expression are detected in the head, limbs and ventrolateral body wall. At all sites, the domains of expression are first detected in undifferentiated tissue, and then expression persists as these tissues undergo subsequent morphogenesis. For example, goosecoid expression is noted in the first branchial arch at 10.5 days, and then expression persists as this tissue undergoes morphogenesis to form the lower jaw and the body of the tongue. Expression in tissues around the first branchial cleft persists as these undergo morphogenesis to form the base of the auditory meatus and eustachian tube. Expression in tissues around the newly formed nasal pits persists as these elongate to form the nasal chambers. Expression in the ventral epithelial lining of the otic vesicle persists as this eventually gives rise to the non-sensory epithelium of the cochlea. Expression in the proximal limb buds and ventrolateral body wall persists as these tissues undergo morphogenesis to form proximal limb structures and ventral ribs respectively. Our findings lead us to suggest that the goosecoid gene product plays a role in spatial programming within discrete embryonic fields, and possibly lineage compartments, during organogenesis stages of mouse development.

The mouse has a Polycomb-like chromobox gene.

The Drosophila gene Polycomb (Pc) has been implicated in the clonal inheritance of determined states and is a trans-regulator of the Antennapedia-like homeobox genes. Pc shares a region of homology (the chromobox) with the Drosophila gene Heterochromatin Protein 1 (HP1), a component of heterochromatin. The Pc chromobox has been used to isolate a mouse chromobox gene, M33, which encodes a predicted 519 amino acid protein. The M33 chromodomain is more similar to that in the Pc protein, than that in the HP1 protein. In addition to the chromodomain, the M33 and Pc proteins also share a region of homology at their C termini. The temporal and spatial expression patterns of M33 have been studied by in situ hybridization and northern analysis. During the final 10 days of embryonic development, M33 expression mirrors that of the cell-cycle-specific cyclin B gene. It is therefore suggested that the rate of cellular proliferation controls M33 expression. From comparisons of the characteristics of M33 with those of Pc it is proposed that M33 is a Pc-like chromobox gene. The roles of M33 and Pc in models of cellular memory are examined and implications of the memory models addressed.

Identification of mobile and fixed antigens on the plasma membrane of rat spermatozoa using monoclonal antibodies.

We have used monoclonal antibodies to study the mobility and distribution of three different antigens on the cell surface of rat spermatozoa. We classified two of the antigens (designated 2B1 and 2D6) as 'mobile', since when detected by indirect immunofluorescence they were situated over the entire sperm flagellum and were susceptible to antibody-induced patching. Patching was critically dependent upon antibody concentrations and was much reduced at 4 degrees C. Patching of the 2B1 antigen was not induced by the 2B1 monoclonal antibody alone. Thus, 2B1 antibody labelled directly with fluorescein bound with a uniform distribution over the sperm flagellum, but this uniform fluorescence was made patchy on subsequent incubation in an unlabelled second antibody layer of anti-mouse IgG anti-serum. By 'Western blotting', the 2B1 antigen was found to be located to a 40 kD molecular weight polypeptide. The remaining 'fixed' antigen (designated 1B6) was not susceptible to antibody-induced patching, and was restricted to a discrete domain on the post-acrosomal region of the sperm surface. We discuss the relationship between mobility of sperm surface antigens and their segregation to discrete domains on the plasma membrane.

Antigens on rat spermatozoa with a potential role in fertilization.

Eight monoclonal antibodies (McAbs), directed against antigens on rat cauda epididymal spermatozoa, were tested for their capacity to interfere with fertilization in vitro as a means of identifying molecules with a potential role in sperm-egg recognition and fusion. Antigens recognized by the McAbs were visualized on live spermatozoa by indirect immunofluorescence (IIF) and characterized by immunoblotting. Five McAbs (designated 1B5, 2C4, 4B5, 5B1, and 8C4) recognized antigens specifically on the sperm acrosome and three (designated 2B1, 2D6, and 6B2) bound to the flagellum. Of the eight McAbs investigated, three (2B1, 2C4, and 6B2) were effective in blocking fertilization in vitro when added as culture supernatants to mixtures of sperm and eggs. McAb 6B2 was inhibitory due to its ability to agglutinate spermatozoa. McAbs 2B1 and 2C4 did not agglutinate capacitated spermatozoa, had no observable effect on motility, and yet blocked fertilization in a dose-dependent manner. McAb 2C4 did not give a reaction on immunoblots, but the 2B1 antigen was identified as an Mr 40 kD glycoprotein. McAb 2B1 appeared to block fertilization at the level of zona binding, whereas the effects of 2C4 were directed more against zona penetration and/or fusion with the vitellus. When sperm-egg complexes were stained with 2C4 or 2B1 McAbs and viewed by IIF, all spermatozoa that were attached to the zona showed fluorescence on the head. These results suggest that different antigens on the rat sperm head participate in different aspects of the fertilization process and that during capacitation there is either exposure of these antigens or else they migrate to their site of action from the flagellum.

Mouse homeo-genes within a subfamily, Hox-1.4, -2.6 and -5.1, display similar anteroposterior domains of expression in the embryo, but show stage- and tissue-dependent differences in their regulation.

By use of in situ hybridization experiments on mouse embryo sections, we compare the transcript patterns of three homeo-genes from the Hox-1.4 subfamily (Hox-1.4, -2.6 and -5.1). Genes within a subfamily are true homologues, present in the genome as a result of duplication of an ancestral homeo-gene cluster. We show that Hox-1.4, -2.6 and -5.1 are similar, although apparently not identical, in the limits of their transcript domains along the anteroposterior axis. Within the prevertebral column of the 12 1/2 day embryo, for example, the anterior boundary of transcripts for each of the three genes was most obvious at the junction of the first and second prevertebrae. Similarly, all three genes showed an anterior boundary of transcripts within the central nervous system that was located in the mid-myelencephalon of the hindbrain. Both in the prevertebral column and hindbrain, however, Hox-2.6 and Hox-5.1 transcripts extended slightly anterior to the anteriormost limits detected for Hox-1.4. In spite of close similarities in the positions of their transcript domains, Hox-1.4, -2.6 and -5.1 displayed striking stage- and tissue-dependent differences in the relative abundance of their transcripts. For example, Hox-5.1 transcripts were abundant within mesoderm and ectoderm of early stages (8 1/2 and 9 1/2 days), yet were detected only weakly in mesodermal components of the lung and stomach at 10 1/2 days, and were apparently absent from these tissues at 12 1/2 days. In contrast, Hox-1.4 and Hox-2.6 transcripts were relatively weakly detected at 8 1/2 and 9 1/2 days, but were abundant within the lung and stomach at 12 1/2 days. Our findings suggest, but do not prove, that genes within the Hox-1.4 subfamily might be coordinately regulated in their expression. We discuss the patterns of mouse homeo-gene expression now observed in terms of models originally devised for Drosophila. We also propose how our new findings may help to explain any selective advantage to the vertebrates of homeo-gene duplication to form subfamilies.

Homoeobox gene expression in mouse embryos varies with position by the primitive streak stage.

Pattern formation in animal development requires that genes be expressed differentially according to position in the sheets of cells that make up the early embryo. The homoeobox-containing genes of Drosophila are control genes active both in the establishment of a segmentation pattern and in the specification of segment identity. In situ hybridization experiments confirm that these genes are expressed in a segmentally-restricted manner and that their expression presages morphological differentiation of segmental structures. Homoeobox genes have recently been isolated from the mouse and have been shown to be expressed during mouse development. Using in situ hybridization, we show here that expression of the mouse homoeobox gene Mo-10 (ref. 7) is spatially restricted in the developing embryo and that localization of expression is already evident within the germ layers before their morphological differentiation. These findings support the suggestion that the homoeobox genes of mammals, like those of Drosophila, may be important in pattern formation.