DEVELOPMENTAL
BIOLOGY
151,204-212 (19%)
Distinct Effects of Ectopic Expression of Writ-I, Activin B, and bFGF on Gap Junctional Permeability in 32-cell Xenopus Embryos DANIELJ.OLSONANDRANDALL T. MOON Department
of Pharmacology,
University
of Washington School of Medicine, Seattle, Washington 98195
Accepted January 16, 1992
A polarity in gap junctional permeability normally exists in 32-cell stage Xenow embryos, in that dorsal cells are relatively more coupled than ventral cells, as measured by transfer of Lucifer yellow dye. The current study extends our analysis of whether gap junctional permeability at this stage can be modulated by secreted factors, and whether the polarity in gap junctional permeability correlates with the effects of ectopic expression of these secreted factors on the subsequent phenotype of the developing embryo. Following ectopic expression of activin B or Writ-1, but not bFGF,the transfer of Lucifer yellow between ventral animal pole cells is detected in a greater percentage of 32-cell stage embryos. This increased incidence of dye transfer between ventral cells correlates with axial duplications later in development. However, there are differences in the extent of Lucifer yellow transfer between animal and vegetal hemisphere blastomeres which is dependent on whether activin B or Wnt-1 RNA had previously been injected. These results suggest that enhanced gap junctional permeability between ventral cells of 32-cell Xenoms embryos correlates with subsequent defects in the dorsoventral axis, although there are at present no direct data demonstrating a role for gap junctions in establishment or maintenance of this axis. Moreover, while both activin B and bFGF are mesoderm-inducing growth factors, only activin B has effects on gap junctional permeability in 32-cell embryos following ectopic expression, demonstrating an interesting difference in physiological responses to expression of these factors. o 1992 Academic Press, Inc.
by bFGF and activin (Christian et ah, 1991), and ectopic expression of Xwnt 8 modifies the response of Xenopus Mesodermal induction in Xenopus embryos appears to blastula caps to bFGF (Christian et al., 1992). Directly involve peptide growth factors which belong to the fibro- supporting a role for a Wnt signalling pathway in early blast growth factor (FGF) family and to the transformdevelopment, early delocalized expression of some ing growth factor B (TGF-B) family (reviewed by Smith, members of the Wnt gene family leads to effects on the 1989). Maternally expressed basic FGF (Kimelman et embryonic axis (McMahon and Moon, 1989; Christian et ah, 1988; Slack and Isaacs, 1989) may induce ventrolatal., 1991; Sokol et aZ., 1991). Ectopic expression of activin era1 mesoderm (Slack et al, 1987; Kimelman and B also has effects on the embryonic axis (Thomsen et ab, Kirschner, 1987), giving rise to mesenchyme, blood, and 1990), but only Wnts are able to induce formation of a some posterior structures, although recent evidence new Spemann organizer (Sokol et aC, 1991). In contrast, suggests that bFGF can induce both dorsal and ventral microinjection of bFGF RNA into fertilized eggs has no mesoderm (Christian et ab, 1992; Kimelman and Maas, effect on the phenotype of intact embryos (Kimelman 1992). The TGF-B-related factors, activin A and activin and Maas, 1992). The specific mechanisms and pathways B, are capable of inducing dorsal mesoderm, leading to by which ectopic expression of some Wnts and activin B secondary induction of neural structures (Smith et ab, affect axis formation are unknown. 1990a; Sokol et al, 1990; Thomsen et al, 1990). Recent Since ectopic expression of some Wnts (McMahon and evidence has also been presented that a maternal acti- Moon, 1989; Christian et al, 1991; Sokol et ak, 1991) and vin may be active during early development (Asashima activins (Thomsen et al., 1990), but not bFGF (Kimelet al., 1991). man and Maas, 1992), have effects on the embryonic The Wnt gene family encodes secreted polypeptides axis, it is reasonable to postulate that there may be which are expressed in distinct regions of vertebrate measureable cell physiological changes which precede embryos, and at distinct times during embryonic devel- these overt phenotypic effects. We recently reported an in vivo assay which demonstrated changes in gap juncopment, and which may play roles in pattern formation tional permeability, monitored by Lucifer yellow dye (reviewed in McMahon, 1991). In considering the physiotransfer (Guthrie, 1984; Guthrie et al., 1988) following logical actions of mesoderm-inducing growth factors in ectopic expression of some members of the Wnt gene development, it is interesting to also note the potential involvement of Wnt signalling pathways. For example, family (Olson et al., 1991). We reported that the ectopic expression of Xwnt-8 is induced in isolated blastula caps expression of Wnt-1 and Xwnt-8 enhances the incidence INTRODUCTION
0012-1606/92 $5.00 Copyright All rights
0 1992 by Academic Press, Inc. of reproduction in any form reserved.
204
OLSONANDMOON
Gap Junctional
of transfer of Lucifer yellow via gap junctions in ventral animal pole blastomeres at the 32-cell stage of Xenopus development, and that this correlates with the bifurcation of the embryonic axis later in development. As ectopic expression of activin (Thomsen et ah, 1990), but not bFGF (Kimelman and Maas, 1992), also affects the embryonic axis, this raises the question of whether ectopic expression of either of these mesoderm-inducing growth factors also leads to measurable changes in gap junctional permeability. Although Smith et al. (1990b) have shown changes in cell spreading behavior in dissociated animal cap cells exposed to media containing activin, there has been little investigation of early cellular responses to activin or bFGF in intact embryos. In the present study we report that the incidence of transfer of Lucifer yellow between ventral animal pole cells at the 32-cell stage is enhanced by ectopic expression of either activin B or Wnt-1, but not by bFGF. Furthermore, since a difference exists in the axial defect produced by Writ-1 and activin B (Sokol et al, 1991), we more closely examined the effects of ectopic expression of each polypeptide on the transfer of Lucifer yellow between 32-cell stage blastomeres. We report that there is a difference in dye transfer between animal and vegetal hemisphere cells, which is dependent on whether activin B or Wnt-1 is ectopically expressed, and on the site of dye injection. Moreover, ectopically expressed bFGF has no effect on gap junctional permeability nor on development of the dorsoventral axis. These results suggest that ectopic expression of activin B, like some Wnts, has effects on both gap junctional permeability in 32cell embryos and on the subsequent dorsoventral axis. While this is an interesting correlation, future work is necessary to test the roles of gap junctional coupling in the establishment or maintenance of a normal dorsoventral axis. Finally, these data demonstrate an interesting difference in cellular responses to ectopic expression of the two mesoderm-inducing growth factors, activin B and bFGF. METHODS
AND
MATERIALS
Handling of Frogs and Injection of RNA Adult Xenopus laevis were induced to spawn, and the eggs were fertilized in vitro. After removal of the jelly coat, but prior to first cleavage, the fertilized eggs were injected (Moon and Christian, 1989) with approximately 0.1-0.5 ng of in vitro-transcribed activin B RNA (Sokol et ah, 1991), 0.1-1.0 ng of mouse Wnt-1 RNA (McMahon and Moon, 1989), or 2-3 ng of bFGF RNA (Kimelman and Maas, 1992). The injection sites were random relative to the future dorsal-ventral axis. It has been previously demonstrated that Wnt RNA injected in this manner is expressed in much of the embryo (McMahon and
Permeability
205
Moon, 1989). Embryos were maintained in 5% Ficoll in 0.1~ modified Barth solution (Moon and Christian,
1989) at 1’7°C. Some embryos were allowed to develop to stage 24 (Nieuwkoop and Faber, 1967) to ascertain the phenotype resulting from the expression of the injected mRNA. Determination of Dorsalventral Polarity Xenopus embryos were allowed to develop to the 8-cell stage when the difference in cell pigmentation is readily apparent (ventral being darker). The dorsal side of the first cleavage furrow was marked with 1% Nile blue. Some embryos were allowed to develop to stage 10 to document the accuracy of Nile blue marking (Olson et ah, 1991). Injection of Lucifer Yellow and Scoring of Dye Transfer At the 32-cell stage, symmetrically cleaving embryos were injected with approximately 1 nl of 4% Lucifer yellow (Aldrich) into single tier 1 animal pole blastomeres (previously referred to as “hl” or “al” blastomeres) (Guthrie, 1984) or into single tier 3 vegetal pole blastomeres (previously referred to as “h3” or “a3” blastomeres) (Guthrie et ah, 1988) on the ventral side of the embryo. Ten or 20 min after dye injection, the embryos were transferred to fixative consisting of 0.25% glutaraldehyde and 3% paraformaldehyde in 0.05 M phosphate-buffered saline (PBS) (pH 7.4) for l-2 hr at 4°C. The embryos were then maintained in 0.05 M PBS at 4°C until examination. As a control, fluorescein isothiocyanate (FITC-dextran; 10,000 M,), which does not pass through gap junctions (Simpson et ah, 1977; Guthrie, 1984; Olson et ah, 1991), was injected as above into a similar group of embryos. In addition, some embryos were injected with Lucifer yellow into single dorsal tier 1 animal pole cells (previously referred to as “dl” or “el” blastomeres) (Guthrie, 1984) or into single dorsal tier 3 vegetal pole cells (previously referred to as “d3” or “e3” blastomeres) (Guthrie et ah, 1988) to determine if enhanced coupling was specific only for the ventral side in the presence of ectopically expressed activin B or Wnt-1 RNA. The results were compared to the transfer of dye between dorsal or ventral animal pole cells, or to the transfer of dye between the vegetal and animal hemispheres, in embryos not previously injected with RNA. Lucifer yellow or (FITC)-dextran was localized with a Leitz Dialux 20 epifluorescence microscope. Embryos were scored as transferring fluorescent material only when it extended beyond the injected cell and its sister cell, which are often connected by cytoplasmic bridges (Guthrie et ah, 1988).
206
DEVELOPMENTALBIOLOGY
VOLUME151,1992
TABLE 1 GAP JUNCTIONALCOMMUNICATION BETWEENNONSISTERDORSALORVENTRALBLASTOMERESIN 32-CELLXenopus EMBRYOS PREVIOUSLYINJECTEDWITH ACTIVIN B ORbFGF RNA Embryos showing transfer of Lucifer yellow” (%) Side of embryo injected with dye
Time after dye injection (min)
Ventral Ventral Dorsal
20
10 10
Activin B injected 50 (142) 56 (16) 48 (25)
injected
Uninjected injected
16 (70)
10 (134)
49 (34)
5 (20) 43 (14)
bFGF
Embryos showing transfer of FITC-dextran (%) Activin B
Uninjected
5 (21)
8 (13)
a The results reflect data pooled from eight different experiments. The number of embryos used are indicated in parentheses.
indicating that the RNA had indeed been translated in vivo. Thus, the premature expression of activin B is capaTo determine whether the ectopic expression of bFGF ble of enhancing the incidence of Lucifer yellow dye results in axial defects or changes in the permeability of transfer between ventral animal pole cells. gap junctions, bFGF RNA was injected into fertilized To determine if activin B is capable of influencing gap Xenopus eggs, followed by injection of Lucifer yellow junctional permeability on the dorsal side of the emdye into a single tier 1 ventral or dorsal animal pole cell bryo, Lucifer yellow was injected into dorsal tier 1 aniat the 32-cell stage. Embryos developed normally, with mal pole blastomeres of embryos previously injected no effect on the dorsoventral axis, as previously re- with activin B RNA. There was no statistically signifiported (Kimelman and Maas, 1992). With regard to the cant difference in the percentage of embryos showing incidence of dye transfer, in ventral animal pole cells dye transfer between dorsal nonsister cells in unindye transfer was observed in 16% of embryos injected jetted embryos versus those injected at the l-cell stage with activin B RNA (Table 1). Thus, the ectopic expreswith bFGF RNA, similar to dye transfer in uninjected control embryos (Table 1). That the bFGF RNA was in- sion of activin B RNA does not appear to influence the tact and expressed in embryos was established by em- gap junctional permeability between dorsal animal hemisphere blastomeres, which are normally relatively ploying aliquots of the same RNA used by Kimelman and Maas (1992), which led to the induction of dorsal permeable to Lucifer yellow, but activin B does enhance the incidence of transfer of Lucifer yellow between venand ventral mesoderm in isolated animal caps. To determine whether the ectopic expression of acti- tral blastomeres in 32-cell stage embryos. To test whether the above transfer of Lucifer yellow vin B, like W&-l or Xwnt-8 (Olson et al., 1991), is capable of enhancing cell coupling between ventral animal hemi- reflected changes in gap junctional permeability, experiments were repeated using (FITC)-dextran, which does sphere cells, activin B mRNA was injected into fertilized Xenopus eggs prior to first cleavage. At the 32-cell not transfer between cells by gap junctions (reviewed in stage, Lucifer yellow was injected into single tier 1 ven- Olson et ah, 1991). It was found that 5% of the embryos tral animal pole blastomeres of RNA-injected and unin- injected with activin B RNA subsequently transferred jetted embryos, and 10 min later embryos were placed in (FITC)-dextran between ventral animal pole cells (Tafixative. Embryos previously injected with activin B ble 1). This indicates that the observed enhanced RNA were found to transfer dye between nonsister ven- transfer of Lucifer yellow in embryos injected with actitral animal pole cells in 50% of the embryos (Table 1; vin B RNA, relative to uninjected controls, did not result Fig. 1B). In contrast, the transfer of Lucifer yellow be- from the persistence of cytoplasmic bridges or from the tween ventral animal pole cells was observed in only disruption of membranes between blastomeres. Given that some Wnts (Olson et ah, 1991) and activin B 10% of uninjected control embryos (Table 1; Fig. lA), (above data) affect Lucifer yellow dye transfer, the relasimilar to that found previously (Guthrie, 1984; Guthrie tive effects of ectopic expression of activin B and Writ-1 et al., 1988). This difference was found to be statistically on gap junctional permeability were then directly comsignificant by x2 analysis (P < 0.001). When dye transfer was scored 20 min after the injection of Lucifer yellow, pared for the ventral side of the embryo. Sixty-four perthe results were similar, indicating that the enhanced cent of the activin B-injected embryos displayed Lucifer dye transfer was not cell cycle dependent. Injection of yellow transfer from ventral tier 1 cells to the vegetal activin B RNA resulted in embryos with phenotypes sim- hemisphere, similar to that observed in Wnt-l-injected ilar to those previously described (Sokol et al, 1991), embryos, and significantly greater than the 17% of conRESULTS
OLSONANDMOON Gap Junctional Permeability
207
FIG. 1. Transfer of dye from a single ventral tier 1 animal pole cell injected with Lucifer yellow dye at the 32-cell stage. (A) Control embryo showing dye transfer only to a sister cell. Scale bar is 0.12 mm. (B) Embryo injected with activin B RNA prior to first cleavage, showing extensive dye transfer between ventral animal pole cells, including dye transfer to the vegetal hemisphere (direction of arrow). Scale bar is 0.09 mm. Cells injected with Lucifer yellow are denoted by a triangle in each panel, and intensity of fluorescence is decreased in animal pole cells due to pigmentation of this hemisphere.
trols which exhibited detectable dye transfer (Table 2). The similar effects on dye transfer of activin B and Writ1 were interesting, in that the phenotypes of the embryos were somewhat different when analyzed later in development. That is, the ectopic expression of W&-l resulted in duplication of the embryonic axis or enhanced dorsal-anterior structures (McMahon and Moon, 1989; Sokol et ak, 1991). In contrast, the ectopic expression of activin B resulted in axial defects or defects in gastrulation, as previously noted (Thomsen et aC, 1990;
Sokol et al, 1991). In light of these different phenotypes, we further investigated Lucifer yellow dye transfer to test whether we might find differences between Writ-1 and activin B-injected embryos. Remarkably, when transfer of Lucifer yellow dye was compared in the opposite direction, from ventral tier 3 cells toward the animal hemisphere, a significant difference (P < 0.001) was found for embryos injected with activin B compared to those injected with W&-l RNA. Embryos displayed permeability to dye transfer in a vegetal to animal hemisphere direction in 22% of the activin B-injected embryos, similar to the 20% of uninjected controls showing dye transfer in this direction (Table 3; Figs. 2A and 2B). TABLE 2 These results are similar to the data of Guthrie et al. EFFECTSOF Wnt-1 ANDACTIVINBONGAPJUNCTIONALCOMMUNICATIONBETWEENTIER~ANIMALPOLEANDTIER~VEGETALPOLEBLAS(1988), who injected dye into h3 or a3 blastomeres (eight TOMERESIN~~-CELLX~~~W EMBRYOS embryos assayed for each blastomere), and also reported only a low incidence of dye transfer from tier 3 Embryos transferring Lucifer blastomeres toward the animal pole. In contrast, emSide of Time after yellow to vegetal pole cells” ( W) bryos transferred Lucifer yellow from a vegetal to aniembryo injected dye injection with dye (min) Activin B Wnt-1 Uninjected mal hemisphere direction in 60% of W&-l-injected embryos. Interestingly, the result for W&-l, but not actiVentral 10-20 64 (52) 65 (45) 17 (74) vin B-injected embryos is similar to the incidence of dye Dorsal 10 43 (22) transfer from dorsal tier 3 cells to the animal hemia The results are pooled from four different experiments. The num- sphere in uninjected controls (Table 3). Taken together, ber of embryos analyzed are indicated in parentheses. the data (Fig. 3) indicate that ectopic expression of both
208
DEVELOPMENTALBIOLOGYVOLUME151,1!%2
TABLE3 cells of Xenopus blastula respond to media containing EFFEIXSOFWrit-~ANDA~~I~INBONGAPJIJN~TIONALC~MM~ICA-activin by increased spreading and migration (Smith et TIONBETWEEN TIER 3 VEGETALCELLSANDANIMAL POLEBLASTOal, 1990b), supporting the idea that cell interactions and MERESIN~~-CELLX~EYBRYOS
adhesive properties may be affected by this and perhaps other secreted factors expressed in early embryos. Side of Time after Secondarily, our data support the previously noted embryo injected dye injection correlation between increased ventral cell gap juncwith dye (min) Activin B Wnt-1 Uninjected tional permeability and a propensity for enhanced formation of anterodorsal axial structures (Olson et aZ., Ventral 10-20 22 (56) 60 (59) 20 (50) 10 Dorsal 99 (22) 1991). This correlation was first documented in experiments in which exogenous application of lithium chlo“The results are pooled from four different experiments. The numride (LiCl), which is known to dorsalize Xenopus ember of embryos analyzed in each case is indicated in parentheses. bryos (Kao et ab, 1986), enhanced gap junctional permeability between ventral animal pole cells (Nagajski et ah, 1989). The notion of dorsal-ventral character being activin B and Wnt-1, but not bFGF, can affect gap junc- reflected in cell coupling is further reinforced by studies tional permeability, but that there are differences in the with ultraviolet light, administered just after fertilizaextent of Lucifer yellow dye transfer under these experi- tion. This results in a ventral-posterior phenotype mental conditions. (Scharf and Gerhart, 1983) and decreases dye transfer on the dorsal side of the embryo (Nagajski et al., 1989). It DISCUSSION is important to note, however, that greater gap juncXenopus embryos exhibit a polarity in gap junctional tional permeability on the dorsal relative to the ventral permeability at the 32-cell stage, as dorsal blastomeres side of the normal 32-cell embryos should only be considered a correlation with future dorsal fate. Significant transfer Lucifer yellow more frequently than do ventral blastomeres (Guthrie, 1984; Guthrie et al., 1988). We additional work is necessary to test the roles of gap have previously demonstrated that ectopic expression of junctions, if any, in establishing or maintaining a dorsal or ventral fate in regions of the early embryo. two members of the Writ-gene family, Wnt-1 and XwntWhile the above discussion highlights similarities be8, enhance the incidence of Lucifer yellow transfer between the ventral animal pole cells (Olson et al, 1991). tween the effects on gap junctions of ectopic expression We report here a similar result when synthetic activin B of Wnt-1 and activin B, it is important to note that submRNA is injected into fertilized eggs. That is, gap junc- tle differences exist in the transfer between cells of Lutional permeability is enhanced between ventral animal cifer yellow in embryos injected with these RNAs. Specifically, when looking at the transfer of dye between pole blastomeres in response to activin B, as monitored hemispheres, we found that the ectopic expression of by the incidence of transfer of Lucifer yellow dye. Interestingly, injection of mRNA encoding another meso- activin B resulted in enhanced cell coupling primarily in derm-inducing growth factor, bFGF, has no effect on the direction from the animal to vegetal hemispheres gap junctional permeability and has no effect on the whereas ectopic expression of Wnt-1 resulted in ensubsequent embryonic phenotype. Therefore, the most hanced gap junctional communication in either direcimportant conclusion from these data is that ectopic ex- tion between the animal and vegetal hemispheres. Thus, pression of both activin B and some Wnts,but not bFGF, Lucifer yellow transfer on the ventral side of embryos leads to rapid cell physiological responses in the early injected with Wnt-1, but not activin B, RNA is similar to embryo, as manifested by changes in gap junctional per- the dye transfer observed on the dorsal side of control meability. This conclusion may facilitate dissection of embryos. The mechanisms underlying the differences in the signalling pathways employed by these factors, as dye transfer between Wnt-1 and activin B-injected emwell as suggest future experiments on other cell physio- bryos are wholly unclear at present, but merit further investigation. It is an untested possibility that these logical responses to Wnts and growth factors. In particular, it is interesting to note that dissociated animal cap differences in gap junctional coupling contribute to the Embryos transferring Lucifer yellow to animal pole cells’ (% )
FIG. 2. Transfer of dye from a single ventral tier 3 vegetal pole cell injected with Lucifer yellow at the 32-cell stage in embryos previously injected with activin B or W&-l RNA prior to first cleavage. (A) Activin B RNA expression enhances ventral vegetal cell coupling but dye transfer in a vegetal to animal pole direction is less frequent. Scale bar is 0.17 mm. (B) Higher magnification view of a different embryo injected as in A, scale bar is 0.06 mm. (C) Writ-1 RNA expression enhances ventral vegetal cell coupling, which also includes coupling to the animal hemisphere. Scale bar is 0.17 mm. (D) Higher magnification view of a different embryo injected as in C, scale bar is 0.08 mm. In all panels, cells injected with dye are indicated with a triangle, and the direction of the animal pole is shown with an arrow.
OLSONANDMOON
Gap Junctional Permability
209
210
DEVEXOPMENTAL BIOLOGY
ActivinB injected
VOLUME 151, 19%
WNT-1 injected
Control
Control or bFGF injected
D
D
FIG. 3. Summary of effects of ectopic expression of W&l, activin B, and bFGF on gap junctional permeability in 32-cell Xewpus embryos. Cells injected with Lucifer yellow are denoted in red and cells accumulating transferred dye are denoted in yellow. It is important to note that this summary conveys trends of dye transfer in large numbers of embryos, as Tables 1-3 demonstrate that dye transfer is not an absolute in any of the conditions tested. The arrows indicate the direction of dye transfer scored in Tables 1-3. Abbreviations: ventral, V; dorsal, D; tier 1 blastomeres, Tl; tier 3 blastomeres, T3.
differences in the axial defects produced by the ectopic spheres is indispensible for the formation of specific emexpression of activin B versus Wnt-1 or Xwnt-8 (Thom- bryonic structures. The essential nature of gap juncsen et al., 1990; McMahon and Moon, 1989; Christian et tions for normal development in Xenopus is suggested ah, 1991). That is, ventral injection of activin RNA pro- by experiments in which disruption of gap junctional duces an incomplete secondary axis, lacking anterior communication using anti-connexin antibodies injected structures (Thomsen et ab, 1990), while similar injec- into dorsal blastomeres at the g-cell stage led to the tions of Wnt-1 and Xwnt-8 RNA produce a complete sec- concomitant loss of some anterior-dorsal structures ondary embryo (Christian et ab, 1991; Sokol et al, 1991). (Warner et al, 1984). Interestingly, disruption of cell In addition, injection of Wnt-1 or Xwnt-8, but not activin coupling in Xenoipus in this manner does not inhibit all RNA, can rescue dorsal development in uv-ventralized early inductive events, at least as monitored by muscle embryos, suggesting that these members of the Wnt gene activation (Warner and Gurdon, 1987). While fufamily lead to formation of a new Spemann organizer ture experiments are necessary to ascertain the roles of (Sokol et ab, 1991). gap junctional communication in early development, It remains to be determined whether gap junctional one untested possibility is that the enhanced gap junccommunication mediates the transfer of information tional permeability in response to ectopic expression of involved in morphogenesis in Xenopus embryos, and Wnt-1 and Xwnt-8 (Olson et ak, 1991) effects tissue comwhether cell coupling between vegetal and animal hemi- petence (Sharpe et aZ.,1987; Otte et ak, 1991) to respond
OLSON AND MOON
Gap Junctional
to mesoderm-inducing signals. At present, relatively little is known of the biochemical processes underlying tissue competence to respond to inductive signals, but it is a testable notion that gap junctional permeability is involved in this process. Supporting a correlation between altered gap junctional permeability, and tissue competence to respond to extrinsic signals, is the recent finding that bFGF is capable of inducing a broader spectrum of mesodermal tissues following ectopic expression of Xwnt-8 (Christian et al., 1992), which enhances gap junctional permeability in early embryos (Olson et aZ., 1991). However, the gap junction permeability measurements were conducted with 32-cell embryos, whereas the effects of Xwnt 8 on bFGF responsiveness were conducted with blastula embryos, hence the effects of Xwnt 8 on gap junctional permeability in blastula embryos needs to be studied to test the speculative correlation between gap junctional permeability and tissue competence to respond to some extrinsic signals. It has been suggested that cellular interactions between vegetal and animal hemispheres in embryos of other species play an important role in the establishment of dorsoventral polarity. In the mollusk Patella vulgata the induction and differentiation of the mesentoblast mother cell of one of the four vegetal macromeres determines dorsoventral polarity. This process depends on animal and vegetal cell interaction between the fifth and sixth cleavages (van den Biggelaar, 1977; van den Biggelaar and Guerrier, 1979). It is of interest that gap junctional communication in Patella increases after the fifth cleavage, at the time of animal and vegetal hemisphere interaction (Dorresteijn et al, 1982). In these species, and in Xenopus, there is a pressing need for direct tests of the hypothesis that gap junctional communication influences the dorsoventral character of regions of the embryo. The current data are consistent with the hypothesis that the ectopic expression of activin B and JIM-1 leads to activation of receptor-mediated signal transduction pathways, which likely are involved in modulating gap junctional coupling as well as other processes. Although no Wnt receptors have been reported, cDNAs encoding an activin receptor have been isolated, and this receptor has the sequence characteristics of a transmembrane serine/threonine-specific protein kinase (Mathews and Vale, 1991). This is interesting considering that CAMPdependent protein kinase is known to phosphorylate gap junction proteins, resulting in enhanced communication (Hax et al., 1974; Flagg-Newton and Loewenstein, 1981; Azarnia et al., 1981). Future work is necessary to establish whether enhanced gap junctional communication in the presence of activin B or members of the Wnt gene family is related to their ability to activate signal transduction pathways which directly post-translationally
211
Permeability
modify connexins, such as by phosphorylation, or whether these secreted proteins alter gap junctional communication more indirectly by affecting processes involving cell adhesion (Riggleman et al, 1990; Peifer and Wieschaus, 1990) or connexon assembly (Musil et al., 1990). We are indebted to Jan L. Christian, Robert L. Gimlich, Arie P. Otte, and the anonymous reviewers for their critical interpretations of the data and comments on the manuscript; to Doug Melton for the activin cDNA; to David Kimelman for the bFGF cDNA, and to Robert L. Gimlich for counseling on the complexities of gap junctions. This research was supported by the NIH (ROl HD.2’7525and K04 AR1837 to R.T.M.; DE-07023 to D.J.O.). REFERENCES Asashima, M., Nakano, H., Uchiyama, H., Sugino, H., Nakamura, T., Eto, Y., Ejima, D., Nishimatsu, S., Ueno, N., and Kinoshita, K. (1991). Presence of activin (erythroid differentiation factor) in unfertilized eggs and blastulae of Xenopus laevis. Proc. Natl. Acad. Sci. USA 88,6511-6514. Azarnia, R., Dahl, G., and Loewenstein, W. R. (1981). Cell junction and cyclic AMP. III. Promotion of junctional membrane permeability and junctional membrane particles in a junction-deficient cell type. J. Me&r.
Biol. 63,133-146.
Berridge, M. J., Downes, C. P., and Hanley, M. R. (1989). Neural and developmental actions of lithium: A unifying hypothesis. Cell 59, 411-419. Busa, W. B., and Gimlich, R. L. (1989). Lithium-induced teratogenesis in frog embryos prevented by a polyphosphoinositide cycle intermediate or a diacylglycerol analog. Dev. Biol 132,315-324. Christian, J. L., Gavin, B. J., McMahon, A. P., and Moon, R. T. (1991). Xwnt-8, a Xenopus wnt-l/i&-l related gene responsive to mesoderm inducing growth factors, may play a role in ventral mesodermal patterning during embryogenesis. DeveZopment 111,1045-1055. Christian, J. L., Olson, D. J., and Moon, R. T. (1992). Xwnt-8 modifies the character of mesoderm induced by bFGF in isolated Xenopus ectoderm. EMBO J. 11,33-41. Dorresteijn, A. W. C., Bilinski, S. M., van den Biggelaar, J. A. M., and Bluemink, J. G. (1982). The presence of gap junctions during early Patella embryogenesis: An electron microscopical study. Dev. Biol. 91,397-401.
Flagg-Newton, J. L., and Loewenstein, W. R. (1981). Cell junction and cyclic AMP. II. Promotion of junctional membrane permeability and junctional membrane particles in a junction-deficient cell type. .I Membr. Biol. 63,123-132.
Gimlich, R. L. (1986). Acquisition of developmental autonomy in the equatorial region of the Xenopus embryo. Dev. Biol. 115,340-352. Guthrie, S. C. (1984). Patterns of junctional permeability in the early amphibian embryo. Nature 311,149-151. Guthrie, S., Turin, L., and Warner, A. (1988). Patterns of junctional communication during development of the early amphibian embryo. Development 103,769-783. Hax, W. M. A., van Venrooij, G. E. P. M., and Vossenberg, J. B. J. (1974). Cell communication: A cyclic AMP mediated phenomenon. J. Membr. Biol. 19,253-265.
Kageura, H. (1990). Spatial distribution of the capacity to initiate a secondary embryo in the 32-cell embryo of Xenopus laevis. Dev. Biol. 142,432-452.
Kao, K. R., Masui, Y., and Elison, R. P. (1986). Lithium-induced respecification of pattern in Xenopus laevis embryos. Nature 322, 371373.
212
DEVELOPMENTALBIOLOGY
Kimelman, D., Abraham, J. A., Haapranta, T., Palisi, T. M., and Kirschner, M. (1988). The presence of fibroblast growth factor in the frog egg: Its role as a natural mesoderm inducer. Science 242,10531056. Kimelman, D., and Kirschner, M. (1987). Synergistic induction of mesoderm by FGF and TGF B and the identification of an mRNA coding for FGF in the early Xenopus embryos. Cell 51,869-8’77. Kimelman, D., and Maas, A. (1992). Induction of dorsal and ventral mesoderm by ectopically expressed Xenopus basic fibroblast growth factor. Development 114,261-269. Mathews, L. S., and Vale, W. W. (1991). Expression cloning of an activin receptor, a predicted transmembrane serine kinase. Cell 65,973982. McMahon, A. P. (1991). A superfamily of putative developmental signalling molecules related to the proto-oncogene Writ-l/i&-l. Adv. Deu. Biol. 1,31-60. McMahon, A. P., and Moon, R. T. (1989). Ectopic expression of the proto-oncogene in&l in Xenqpus embryos leads to duplication of the embryonic axis. Cell 58,1075-1084. Moon, R. T., and Christian, J. L. (1989). Microinjection and expression of synthetic mRNAs. Technique 1,76-89. Musil, L. S., Cunningham, B. A., Edelman, G. M., and Goodenough, D. A. (1990). Differential phosphorylation of the gap junction protein connexin 43 in junctional communication-competent and -deficient cell lines. J. Cell Biol. 111,2077-2088. Nagajski, D. J., Guthrie, S. C., Ford, C. C., and Warner, A. E. (1989). The correlation between patterns of dye transfer through gap junctions and future developmental fate in Xenopus; the consequences of U.V. irradiation and lithium treatment. Development 105,747-757. Nieuwkoop, P. D., and Faber, J. (1967). Normal table of Xenqpus luevis. Daudin. North Holland, Amsterdam. Olson, D. J., Christian, J. L., and Moon, R. T. (1991). Effect of writ-1 and related proteins on gap junctional communication in Xenopus embryos. Science 252,1173-1176. Otte, A. P., Kramer, I. M., and Durston, A. J. (1991). Protein kinase C and regulation of the local competence of Xenopus ectoderm. Science 251,570-573. Peifer, M., and Wieschaus, E. (1990). The segment polarity gene armadillo encodes a functionally modular protein that is the Drosophila homolog of human plakoglobin. Cell 63,1167-1178. Riggleman, B., Schedl, P., and Wieschaus, E. (1990). Spatial expression of the Drosophila segment polarity gene armadillo is posttranscriptionally regulated by wingless. Cell 63,549-560. Scharf, S. R., and Gerhart, J. C. (1983). Axis determination in eggs of Xenopus &is: A critical period before first cleavage, identified by the common effects of cold, pressure, and ultraviolet irradiation. Dev. Biol. 99, 75-87.
Sharpe, C. R., Fritz, A., Derobertis, E. M., and Gurdon, J. B. (1987). A homeobox-containing marker of posterior neural differentiation
VOLUME151,1992
shows the importance of predetermination in neural induction. Cell 50,749-758. Simpson, I., Rose, B., and Loewenstein, W. R. (1977). Size limit of molecules permeating the junctional membrane channels. S&nce 195,294-297. Slack, J. M. W., Darlington, B. G., Heath, J. K., and Godsave, S. F. (1987). Mesoderm induction in early Xenopus embryos by heparinbinding growth factors. Nature 326,197-200. Slack, J. M. W., and Isaacs, H. V. (1989). Presence of fibroblast growth factor in the early Xenoms embryo. Development 105,147-153. Smith, J. C. (1989). Mesoderm induction and mesoderm-inducing factors in early amphibian development. Development 105,665-678. Smith, J. C., Cooke, J., Green, J. B. A., Howes, G., and Symes, K. (1989). Inducing factors and the control of mesodermal pattern in Xenopus loxvis. Development Suppl 149-159. Smith, J. C., Price, B. M. J., Van Nimmen, K., and Huylebroeck, D. (1990a). Identification of a potent Xem mesoderm-inducing factor as a homologue of activin A. Nature 345,729-731. Smith, J. C., Symes, K., Hynes, R. O., and DeSimone, D. (1990b). Mesoderm induction and the control of gastrulation in Xenopus Levis: The roles of fibrodectin and integrins. Development 108,229-238. Sokol, S., Christian, J. L., Moon, R. T., and Melton, D. A. (1991). Injected Writ but not activin RNA induces a complete dorsal axis in Xenopus embryos. Cell 67,741-752. Sokol, S., and Melton, D. A. (1991). Pre-existent pattern in Xen0pu.s animal pole cells revealed by induction with activin. Nature 351, 409-411. Sokol, S., Wong, G. G., and Melton, D. A. (1990). A mouse macrophage factor induces head structures and organizes a body axis in Xeru+ pus Science 249,561-564.
Spemann, H., and Mangold, H. (1924). Uber induktion von embryoanalagen durch implantation arffremder organisatoren. Arch. Mikrosk.
Anat. Entwicklungsmech
100,599-638.
Thomsen, G., Woolf, T., Whitman, M., Sokol, S., Vaughn, J., Vale, W., and Melton, D. A. (1990). Activins are expressed early in Xenopus embryogenesis and can induce axial mesoderm and anterior structures. CeU63,485-493. van den Biggelaar, J. A. M. (1977). Development of dorsoventral polarity and mesentoblast determination in Patella vulgata J. Morphol. 154,157-186.
van den Biggelaar, J. A. M., and Guerrier, P. (1979). Dorsoventral polarity and mesentoblast determination as concomitant results of cellular interactions in the mollusc Pate& vulgata. Deu. Biol. 68, 462-471.
Warner, A. E., and Gurdon, J. B. (1987). Functional gap junctions are not required for muscle gene activation by induction in Xenspus embryos. J. Cell Biol. 104,557-564. Warner, A. E., Guthrie, S. C., and Gilula, N. B. (1984). Antibodies to gap junctional proteins selectively disrupt junctional communication in the early amphibian embryo. Nature 311,127-131.
BIOLOGY
151,204-212 (19%)
Distinct Effects of Ectopic Expression of Writ-I, Activin B, and bFGF on Gap Junctional Permeability in 32-cell Xenopus Embryos DANIELJ.OLSONANDRANDALL T. MOON Department
of Pharmacology,
University
of Washington School of Medicine, Seattle, Washington 98195
Accepted January 16, 1992
A polarity in gap junctional permeability normally exists in 32-cell stage Xenow embryos, in that dorsal cells are relatively more coupled than ventral cells, as measured by transfer of Lucifer yellow dye. The current study extends our analysis of whether gap junctional permeability at this stage can be modulated by secreted factors, and whether the polarity in gap junctional permeability correlates with the effects of ectopic expression of these secreted factors on the subsequent phenotype of the developing embryo. Following ectopic expression of activin B or Writ-1, but not bFGF,the transfer of Lucifer yellow between ventral animal pole cells is detected in a greater percentage of 32-cell stage embryos. This increased incidence of dye transfer between ventral cells correlates with axial duplications later in development. However, there are differences in the extent of Lucifer yellow transfer between animal and vegetal hemisphere blastomeres which is dependent on whether activin B or Wnt-1 RNA had previously been injected. These results suggest that enhanced gap junctional permeability between ventral cells of 32-cell Xenoms embryos correlates with subsequent defects in the dorsoventral axis, although there are at present no direct data demonstrating a role for gap junctions in establishment or maintenance of this axis. Moreover, while both activin B and bFGF are mesoderm-inducing growth factors, only activin B has effects on gap junctional permeability in 32-cell embryos following ectopic expression, demonstrating an interesting difference in physiological responses to expression of these factors. o 1992 Academic Press, Inc.
by bFGF and activin (Christian et ah, 1991), and ectopic expression of Xwnt 8 modifies the response of Xenopus Mesodermal induction in Xenopus embryos appears to blastula caps to bFGF (Christian et al., 1992). Directly involve peptide growth factors which belong to the fibro- supporting a role for a Wnt signalling pathway in early blast growth factor (FGF) family and to the transformdevelopment, early delocalized expression of some ing growth factor B (TGF-B) family (reviewed by Smith, members of the Wnt gene family leads to effects on the 1989). Maternally expressed basic FGF (Kimelman et embryonic axis (McMahon and Moon, 1989; Christian et ah, 1988; Slack and Isaacs, 1989) may induce ventrolatal., 1991; Sokol et aZ., 1991). Ectopic expression of activin era1 mesoderm (Slack et al, 1987; Kimelman and B also has effects on the embryonic axis (Thomsen et ab, Kirschner, 1987), giving rise to mesenchyme, blood, and 1990), but only Wnts are able to induce formation of a some posterior structures, although recent evidence new Spemann organizer (Sokol et aC, 1991). In contrast, suggests that bFGF can induce both dorsal and ventral microinjection of bFGF RNA into fertilized eggs has no mesoderm (Christian et ab, 1992; Kimelman and Maas, effect on the phenotype of intact embryos (Kimelman 1992). The TGF-B-related factors, activin A and activin and Maas, 1992). The specific mechanisms and pathways B, are capable of inducing dorsal mesoderm, leading to by which ectopic expression of some Wnts and activin B secondary induction of neural structures (Smith et ab, affect axis formation are unknown. 1990a; Sokol et al, 1990; Thomsen et al, 1990). Recent Since ectopic expression of some Wnts (McMahon and evidence has also been presented that a maternal acti- Moon, 1989; Christian et al, 1991; Sokol et ak, 1991) and vin may be active during early development (Asashima activins (Thomsen et al., 1990), but not bFGF (Kimelet al., 1991). man and Maas, 1992), have effects on the embryonic The Wnt gene family encodes secreted polypeptides axis, it is reasonable to postulate that there may be which are expressed in distinct regions of vertebrate measureable cell physiological changes which precede embryos, and at distinct times during embryonic devel- these overt phenotypic effects. We recently reported an in vivo assay which demonstrated changes in gap juncopment, and which may play roles in pattern formation tional permeability, monitored by Lucifer yellow dye (reviewed in McMahon, 1991). In considering the physiotransfer (Guthrie, 1984; Guthrie et al., 1988) following logical actions of mesoderm-inducing growth factors in ectopic expression of some members of the Wnt gene development, it is interesting to also note the potential involvement of Wnt signalling pathways. For example, family (Olson et al., 1991). We reported that the ectopic expression of Xwnt-8 is induced in isolated blastula caps expression of Wnt-1 and Xwnt-8 enhances the incidence INTRODUCTION
0012-1606/92 $5.00 Copyright All rights
0 1992 by Academic Press, Inc. of reproduction in any form reserved.
204
OLSONANDMOON
Gap Junctional
of transfer of Lucifer yellow via gap junctions in ventral animal pole blastomeres at the 32-cell stage of Xenopus development, and that this correlates with the bifurcation of the embryonic axis later in development. As ectopic expression of activin (Thomsen et ah, 1990), but not bFGF (Kimelman and Maas, 1992), also affects the embryonic axis, this raises the question of whether ectopic expression of either of these mesoderm-inducing growth factors also leads to measurable changes in gap junctional permeability. Although Smith et al. (1990b) have shown changes in cell spreading behavior in dissociated animal cap cells exposed to media containing activin, there has been little investigation of early cellular responses to activin or bFGF in intact embryos. In the present study we report that the incidence of transfer of Lucifer yellow between ventral animal pole cells at the 32-cell stage is enhanced by ectopic expression of either activin B or Wnt-1, but not by bFGF. Furthermore, since a difference exists in the axial defect produced by Writ-1 and activin B (Sokol et al, 1991), we more closely examined the effects of ectopic expression of each polypeptide on the transfer of Lucifer yellow between 32-cell stage blastomeres. We report that there is a difference in dye transfer between animal and vegetal hemisphere cells, which is dependent on whether activin B or Wnt-1 is ectopically expressed, and on the site of dye injection. Moreover, ectopically expressed bFGF has no effect on gap junctional permeability nor on development of the dorsoventral axis. These results suggest that ectopic expression of activin B, like some Wnts, has effects on both gap junctional permeability in 32cell embryos and on the subsequent dorsoventral axis. While this is an interesting correlation, future work is necessary to test the roles of gap junctional coupling in the establishment or maintenance of a normal dorsoventral axis. Finally, these data demonstrate an interesting difference in cellular responses to ectopic expression of the two mesoderm-inducing growth factors, activin B and bFGF. METHODS
AND
MATERIALS
Handling of Frogs and Injection of RNA Adult Xenopus laevis were induced to spawn, and the eggs were fertilized in vitro. After removal of the jelly coat, but prior to first cleavage, the fertilized eggs were injected (Moon and Christian, 1989) with approximately 0.1-0.5 ng of in vitro-transcribed activin B RNA (Sokol et ah, 1991), 0.1-1.0 ng of mouse Wnt-1 RNA (McMahon and Moon, 1989), or 2-3 ng of bFGF RNA (Kimelman and Maas, 1992). The injection sites were random relative to the future dorsal-ventral axis. It has been previously demonstrated that Wnt RNA injected in this manner is expressed in much of the embryo (McMahon and
Permeability
205
Moon, 1989). Embryos were maintained in 5% Ficoll in 0.1~ modified Barth solution (Moon and Christian,
1989) at 1’7°C. Some embryos were allowed to develop to stage 24 (Nieuwkoop and Faber, 1967) to ascertain the phenotype resulting from the expression of the injected mRNA. Determination of Dorsalventral Polarity Xenopus embryos were allowed to develop to the 8-cell stage when the difference in cell pigmentation is readily apparent (ventral being darker). The dorsal side of the first cleavage furrow was marked with 1% Nile blue. Some embryos were allowed to develop to stage 10 to document the accuracy of Nile blue marking (Olson et ah, 1991). Injection of Lucifer Yellow and Scoring of Dye Transfer At the 32-cell stage, symmetrically cleaving embryos were injected with approximately 1 nl of 4% Lucifer yellow (Aldrich) into single tier 1 animal pole blastomeres (previously referred to as “hl” or “al” blastomeres) (Guthrie, 1984) or into single tier 3 vegetal pole blastomeres (previously referred to as “h3” or “a3” blastomeres) (Guthrie et ah, 1988) on the ventral side of the embryo. Ten or 20 min after dye injection, the embryos were transferred to fixative consisting of 0.25% glutaraldehyde and 3% paraformaldehyde in 0.05 M phosphate-buffered saline (PBS) (pH 7.4) for l-2 hr at 4°C. The embryos were then maintained in 0.05 M PBS at 4°C until examination. As a control, fluorescein isothiocyanate (FITC-dextran; 10,000 M,), which does not pass through gap junctions (Simpson et ah, 1977; Guthrie, 1984; Olson et ah, 1991), was injected as above into a similar group of embryos. In addition, some embryos were injected with Lucifer yellow into single dorsal tier 1 animal pole cells (previously referred to as “dl” or “el” blastomeres) (Guthrie, 1984) or into single dorsal tier 3 vegetal pole cells (previously referred to as “d3” or “e3” blastomeres) (Guthrie et ah, 1988) to determine if enhanced coupling was specific only for the ventral side in the presence of ectopically expressed activin B or Wnt-1 RNA. The results were compared to the transfer of dye between dorsal or ventral animal pole cells, or to the transfer of dye between the vegetal and animal hemispheres, in embryos not previously injected with RNA. Lucifer yellow or (FITC)-dextran was localized with a Leitz Dialux 20 epifluorescence microscope. Embryos were scored as transferring fluorescent material only when it extended beyond the injected cell and its sister cell, which are often connected by cytoplasmic bridges (Guthrie et ah, 1988).
206
DEVELOPMENTALBIOLOGY
VOLUME151,1992
TABLE 1 GAP JUNCTIONALCOMMUNICATION BETWEENNONSISTERDORSALORVENTRALBLASTOMERESIN 32-CELLXenopus EMBRYOS PREVIOUSLYINJECTEDWITH ACTIVIN B ORbFGF RNA Embryos showing transfer of Lucifer yellow” (%) Side of embryo injected with dye
Time after dye injection (min)
Ventral Ventral Dorsal
20
10 10
Activin B injected 50 (142) 56 (16) 48 (25)
injected
Uninjected injected
16 (70)
10 (134)
49 (34)
5 (20) 43 (14)
bFGF
Embryos showing transfer of FITC-dextran (%) Activin B
Uninjected
5 (21)
8 (13)
a The results reflect data pooled from eight different experiments. The number of embryos used are indicated in parentheses.
indicating that the RNA had indeed been translated in vivo. Thus, the premature expression of activin B is capaTo determine whether the ectopic expression of bFGF ble of enhancing the incidence of Lucifer yellow dye results in axial defects or changes in the permeability of transfer between ventral animal pole cells. gap junctions, bFGF RNA was injected into fertilized To determine if activin B is capable of influencing gap Xenopus eggs, followed by injection of Lucifer yellow junctional permeability on the dorsal side of the emdye into a single tier 1 ventral or dorsal animal pole cell bryo, Lucifer yellow was injected into dorsal tier 1 aniat the 32-cell stage. Embryos developed normally, with mal pole blastomeres of embryos previously injected no effect on the dorsoventral axis, as previously re- with activin B RNA. There was no statistically signifiported (Kimelman and Maas, 1992). With regard to the cant difference in the percentage of embryos showing incidence of dye transfer, in ventral animal pole cells dye transfer between dorsal nonsister cells in unindye transfer was observed in 16% of embryos injected jetted embryos versus those injected at the l-cell stage with activin B RNA (Table 1). Thus, the ectopic expreswith bFGF RNA, similar to dye transfer in uninjected control embryos (Table 1). That the bFGF RNA was in- sion of activin B RNA does not appear to influence the tact and expressed in embryos was established by em- gap junctional permeability between dorsal animal hemisphere blastomeres, which are normally relatively ploying aliquots of the same RNA used by Kimelman and Maas (1992), which led to the induction of dorsal permeable to Lucifer yellow, but activin B does enhance the incidence of transfer of Lucifer yellow between venand ventral mesoderm in isolated animal caps. To determine whether the ectopic expression of acti- tral blastomeres in 32-cell stage embryos. To test whether the above transfer of Lucifer yellow vin B, like W&-l or Xwnt-8 (Olson et al., 1991), is capable of enhancing cell coupling between ventral animal hemi- reflected changes in gap junctional permeability, experiments were repeated using (FITC)-dextran, which does sphere cells, activin B mRNA was injected into fertilized Xenopus eggs prior to first cleavage. At the 32-cell not transfer between cells by gap junctions (reviewed in stage, Lucifer yellow was injected into single tier 1 ven- Olson et ah, 1991). It was found that 5% of the embryos tral animal pole blastomeres of RNA-injected and unin- injected with activin B RNA subsequently transferred jetted embryos, and 10 min later embryos were placed in (FITC)-dextran between ventral animal pole cells (Tafixative. Embryos previously injected with activin B ble 1). This indicates that the observed enhanced RNA were found to transfer dye between nonsister ven- transfer of Lucifer yellow in embryos injected with actitral animal pole cells in 50% of the embryos (Table 1; vin B RNA, relative to uninjected controls, did not result Fig. 1B). In contrast, the transfer of Lucifer yellow be- from the persistence of cytoplasmic bridges or from the tween ventral animal pole cells was observed in only disruption of membranes between blastomeres. Given that some Wnts (Olson et ah, 1991) and activin B 10% of uninjected control embryos (Table 1; Fig. lA), (above data) affect Lucifer yellow dye transfer, the relasimilar to that found previously (Guthrie, 1984; Guthrie tive effects of ectopic expression of activin B and Writ-1 et al., 1988). This difference was found to be statistically on gap junctional permeability were then directly comsignificant by x2 analysis (P < 0.001). When dye transfer was scored 20 min after the injection of Lucifer yellow, pared for the ventral side of the embryo. Sixty-four perthe results were similar, indicating that the enhanced cent of the activin B-injected embryos displayed Lucifer dye transfer was not cell cycle dependent. Injection of yellow transfer from ventral tier 1 cells to the vegetal activin B RNA resulted in embryos with phenotypes sim- hemisphere, similar to that observed in Wnt-l-injected ilar to those previously described (Sokol et al, 1991), embryos, and significantly greater than the 17% of conRESULTS
OLSONANDMOON Gap Junctional Permeability
207
FIG. 1. Transfer of dye from a single ventral tier 1 animal pole cell injected with Lucifer yellow dye at the 32-cell stage. (A) Control embryo showing dye transfer only to a sister cell. Scale bar is 0.12 mm. (B) Embryo injected with activin B RNA prior to first cleavage, showing extensive dye transfer between ventral animal pole cells, including dye transfer to the vegetal hemisphere (direction of arrow). Scale bar is 0.09 mm. Cells injected with Lucifer yellow are denoted by a triangle in each panel, and intensity of fluorescence is decreased in animal pole cells due to pigmentation of this hemisphere.
trols which exhibited detectable dye transfer (Table 2). The similar effects on dye transfer of activin B and Writ1 were interesting, in that the phenotypes of the embryos were somewhat different when analyzed later in development. That is, the ectopic expression of W&-l resulted in duplication of the embryonic axis or enhanced dorsal-anterior structures (McMahon and Moon, 1989; Sokol et ak, 1991). In contrast, the ectopic expression of activin B resulted in axial defects or defects in gastrulation, as previously noted (Thomsen et aC, 1990;
Sokol et al, 1991). In light of these different phenotypes, we further investigated Lucifer yellow dye transfer to test whether we might find differences between Writ-1 and activin B-injected embryos. Remarkably, when transfer of Lucifer yellow dye was compared in the opposite direction, from ventral tier 3 cells toward the animal hemisphere, a significant difference (P < 0.001) was found for embryos injected with activin B compared to those injected with W&-l RNA. Embryos displayed permeability to dye transfer in a vegetal to animal hemisphere direction in 22% of the activin B-injected embryos, similar to the 20% of uninjected controls showing dye transfer in this direction (Table 3; Figs. 2A and 2B). TABLE 2 These results are similar to the data of Guthrie et al. EFFECTSOF Wnt-1 ANDACTIVINBONGAPJUNCTIONALCOMMUNICATIONBETWEENTIER~ANIMALPOLEANDTIER~VEGETALPOLEBLAS(1988), who injected dye into h3 or a3 blastomeres (eight TOMERESIN~~-CELLX~~~W EMBRYOS embryos assayed for each blastomere), and also reported only a low incidence of dye transfer from tier 3 Embryos transferring Lucifer blastomeres toward the animal pole. In contrast, emSide of Time after yellow to vegetal pole cells” ( W) bryos transferred Lucifer yellow from a vegetal to aniembryo injected dye injection with dye (min) Activin B Wnt-1 Uninjected mal hemisphere direction in 60% of W&-l-injected embryos. Interestingly, the result for W&-l, but not actiVentral 10-20 64 (52) 65 (45) 17 (74) vin B-injected embryos is similar to the incidence of dye Dorsal 10 43 (22) transfer from dorsal tier 3 cells to the animal hemia The results are pooled from four different experiments. The num- sphere in uninjected controls (Table 3). Taken together, ber of embryos analyzed are indicated in parentheses. the data (Fig. 3) indicate that ectopic expression of both
208
DEVELOPMENTALBIOLOGYVOLUME151,1!%2
TABLE3 cells of Xenopus blastula respond to media containing EFFEIXSOFWrit-~ANDA~~I~INBONGAPJIJN~TIONALC~MM~ICA-activin by increased spreading and migration (Smith et TIONBETWEEN TIER 3 VEGETALCELLSANDANIMAL POLEBLASTOal, 1990b), supporting the idea that cell interactions and MERESIN~~-CELLX~EYBRYOS
adhesive properties may be affected by this and perhaps other secreted factors expressed in early embryos. Side of Time after Secondarily, our data support the previously noted embryo injected dye injection correlation between increased ventral cell gap juncwith dye (min) Activin B Wnt-1 Uninjected tional permeability and a propensity for enhanced formation of anterodorsal axial structures (Olson et aZ., Ventral 10-20 22 (56) 60 (59) 20 (50) 10 Dorsal 99 (22) 1991). This correlation was first documented in experiments in which exogenous application of lithium chlo“The results are pooled from four different experiments. The numride (LiCl), which is known to dorsalize Xenopus ember of embryos analyzed in each case is indicated in parentheses. bryos (Kao et ab, 1986), enhanced gap junctional permeability between ventral animal pole cells (Nagajski et ah, 1989). The notion of dorsal-ventral character being activin B and Wnt-1, but not bFGF, can affect gap junc- reflected in cell coupling is further reinforced by studies tional permeability, but that there are differences in the with ultraviolet light, administered just after fertilizaextent of Lucifer yellow dye transfer under these experi- tion. This results in a ventral-posterior phenotype mental conditions. (Scharf and Gerhart, 1983) and decreases dye transfer on the dorsal side of the embryo (Nagajski et al., 1989). It DISCUSSION is important to note, however, that greater gap juncXenopus embryos exhibit a polarity in gap junctional tional permeability on the dorsal relative to the ventral permeability at the 32-cell stage, as dorsal blastomeres side of the normal 32-cell embryos should only be considered a correlation with future dorsal fate. Significant transfer Lucifer yellow more frequently than do ventral blastomeres (Guthrie, 1984; Guthrie et al., 1988). We additional work is necessary to test the roles of gap have previously demonstrated that ectopic expression of junctions, if any, in establishing or maintaining a dorsal or ventral fate in regions of the early embryo. two members of the Writ-gene family, Wnt-1 and XwntWhile the above discussion highlights similarities be8, enhance the incidence of Lucifer yellow transfer between the ventral animal pole cells (Olson et al, 1991). tween the effects on gap junctions of ectopic expression We report here a similar result when synthetic activin B of Wnt-1 and activin B, it is important to note that submRNA is injected into fertilized eggs. That is, gap junc- tle differences exist in the transfer between cells of Lutional permeability is enhanced between ventral animal cifer yellow in embryos injected with these RNAs. Specifically, when looking at the transfer of dye between pole blastomeres in response to activin B, as monitored hemispheres, we found that the ectopic expression of by the incidence of transfer of Lucifer yellow dye. Interestingly, injection of mRNA encoding another meso- activin B resulted in enhanced cell coupling primarily in derm-inducing growth factor, bFGF, has no effect on the direction from the animal to vegetal hemispheres gap junctional permeability and has no effect on the whereas ectopic expression of Wnt-1 resulted in ensubsequent embryonic phenotype. Therefore, the most hanced gap junctional communication in either direcimportant conclusion from these data is that ectopic ex- tion between the animal and vegetal hemispheres. Thus, pression of both activin B and some Wnts,but not bFGF, Lucifer yellow transfer on the ventral side of embryos leads to rapid cell physiological responses in the early injected with Wnt-1, but not activin B, RNA is similar to embryo, as manifested by changes in gap junctional per- the dye transfer observed on the dorsal side of control meability. This conclusion may facilitate dissection of embryos. The mechanisms underlying the differences in the signalling pathways employed by these factors, as dye transfer between Wnt-1 and activin B-injected emwell as suggest future experiments on other cell physio- bryos are wholly unclear at present, but merit further investigation. It is an untested possibility that these logical responses to Wnts and growth factors. In particular, it is interesting to note that dissociated animal cap differences in gap junctional coupling contribute to the Embryos transferring Lucifer yellow to animal pole cells’ (% )
FIG. 2. Transfer of dye from a single ventral tier 3 vegetal pole cell injected with Lucifer yellow at the 32-cell stage in embryos previously injected with activin B or W&-l RNA prior to first cleavage. (A) Activin B RNA expression enhances ventral vegetal cell coupling but dye transfer in a vegetal to animal pole direction is less frequent. Scale bar is 0.17 mm. (B) Higher magnification view of a different embryo injected as in A, scale bar is 0.06 mm. (C) Writ-1 RNA expression enhances ventral vegetal cell coupling, which also includes coupling to the animal hemisphere. Scale bar is 0.17 mm. (D) Higher magnification view of a different embryo injected as in C, scale bar is 0.08 mm. In all panels, cells injected with dye are indicated with a triangle, and the direction of the animal pole is shown with an arrow.
OLSONANDMOON
Gap Junctional Permability
209
210
DEVEXOPMENTAL BIOLOGY
ActivinB injected
VOLUME 151, 19%
WNT-1 injected
Control
Control or bFGF injected
D
D
FIG. 3. Summary of effects of ectopic expression of W&l, activin B, and bFGF on gap junctional permeability in 32-cell Xewpus embryos. Cells injected with Lucifer yellow are denoted in red and cells accumulating transferred dye are denoted in yellow. It is important to note that this summary conveys trends of dye transfer in large numbers of embryos, as Tables 1-3 demonstrate that dye transfer is not an absolute in any of the conditions tested. The arrows indicate the direction of dye transfer scored in Tables 1-3. Abbreviations: ventral, V; dorsal, D; tier 1 blastomeres, Tl; tier 3 blastomeres, T3.
differences in the axial defects produced by the ectopic spheres is indispensible for the formation of specific emexpression of activin B versus Wnt-1 or Xwnt-8 (Thom- bryonic structures. The essential nature of gap juncsen et al., 1990; McMahon and Moon, 1989; Christian et tions for normal development in Xenopus is suggested ah, 1991). That is, ventral injection of activin RNA pro- by experiments in which disruption of gap junctional duces an incomplete secondary axis, lacking anterior communication using anti-connexin antibodies injected structures (Thomsen et ab, 1990), while similar injec- into dorsal blastomeres at the g-cell stage led to the tions of Wnt-1 and Xwnt-8 RNA produce a complete sec- concomitant loss of some anterior-dorsal structures ondary embryo (Christian et ab, 1991; Sokol et al, 1991). (Warner et al, 1984). Interestingly, disruption of cell In addition, injection of Wnt-1 or Xwnt-8, but not activin coupling in Xenoipus in this manner does not inhibit all RNA, can rescue dorsal development in uv-ventralized early inductive events, at least as monitored by muscle embryos, suggesting that these members of the Wnt gene activation (Warner and Gurdon, 1987). While fufamily lead to formation of a new Spemann organizer ture experiments are necessary to ascertain the roles of (Sokol et ab, 1991). gap junctional communication in early development, It remains to be determined whether gap junctional one untested possibility is that the enhanced gap junccommunication mediates the transfer of information tional permeability in response to ectopic expression of involved in morphogenesis in Xenopus embryos, and Wnt-1 and Xwnt-8 (Olson et ak, 1991) effects tissue comwhether cell coupling between vegetal and animal hemi- petence (Sharpe et aZ.,1987; Otte et ak, 1991) to respond
OLSON AND MOON
Gap Junctional
to mesoderm-inducing signals. At present, relatively little is known of the biochemical processes underlying tissue competence to respond to inductive signals, but it is a testable notion that gap junctional permeability is involved in this process. Supporting a correlation between altered gap junctional permeability, and tissue competence to respond to extrinsic signals, is the recent finding that bFGF is capable of inducing a broader spectrum of mesodermal tissues following ectopic expression of Xwnt-8 (Christian et al., 1992), which enhances gap junctional permeability in early embryos (Olson et aZ., 1991). However, the gap junction permeability measurements were conducted with 32-cell embryos, whereas the effects of Xwnt 8 on bFGF responsiveness were conducted with blastula embryos, hence the effects of Xwnt 8 on gap junctional permeability in blastula embryos needs to be studied to test the speculative correlation between gap junctional permeability and tissue competence to respond to some extrinsic signals. It has been suggested that cellular interactions between vegetal and animal hemispheres in embryos of other species play an important role in the establishment of dorsoventral polarity. In the mollusk Patella vulgata the induction and differentiation of the mesentoblast mother cell of one of the four vegetal macromeres determines dorsoventral polarity. This process depends on animal and vegetal cell interaction between the fifth and sixth cleavages (van den Biggelaar, 1977; van den Biggelaar and Guerrier, 1979). It is of interest that gap junctional communication in Patella increases after the fifth cleavage, at the time of animal and vegetal hemisphere interaction (Dorresteijn et al, 1982). In these species, and in Xenopus, there is a pressing need for direct tests of the hypothesis that gap junctional communication influences the dorsoventral character of regions of the embryo. The current data are consistent with the hypothesis that the ectopic expression of activin B and JIM-1 leads to activation of receptor-mediated signal transduction pathways, which likely are involved in modulating gap junctional coupling as well as other processes. Although no Wnt receptors have been reported, cDNAs encoding an activin receptor have been isolated, and this receptor has the sequence characteristics of a transmembrane serine/threonine-specific protein kinase (Mathews and Vale, 1991). This is interesting considering that CAMPdependent protein kinase is known to phosphorylate gap junction proteins, resulting in enhanced communication (Hax et al., 1974; Flagg-Newton and Loewenstein, 1981; Azarnia et al., 1981). Future work is necessary to establish whether enhanced gap junctional communication in the presence of activin B or members of the Wnt gene family is related to their ability to activate signal transduction pathways which directly post-translationally
211
Permeability
modify connexins, such as by phosphorylation, or whether these secreted proteins alter gap junctional communication more indirectly by affecting processes involving cell adhesion (Riggleman et al, 1990; Peifer and Wieschaus, 1990) or connexon assembly (Musil et al., 1990). We are indebted to Jan L. Christian, Robert L. Gimlich, Arie P. Otte, and the anonymous reviewers for their critical interpretations of the data and comments on the manuscript; to Doug Melton for the activin cDNA; to David Kimelman for the bFGF cDNA, and to Robert L. Gimlich for counseling on the complexities of gap junctions. This research was supported by the NIH (ROl HD.2’7525and K04 AR1837 to R.T.M.; DE-07023 to D.J.O.). REFERENCES Asashima, M., Nakano, H., Uchiyama, H., Sugino, H., Nakamura, T., Eto, Y., Ejima, D., Nishimatsu, S., Ueno, N., and Kinoshita, K. (1991). Presence of activin (erythroid differentiation factor) in unfertilized eggs and blastulae of Xenopus laevis. Proc. Natl. Acad. Sci. USA 88,6511-6514. Azarnia, R., Dahl, G., and Loewenstein, W. R. (1981). Cell junction and cyclic AMP. III. Promotion of junctional membrane permeability and junctional membrane particles in a junction-deficient cell type. J. Me&r.
Biol. 63,133-146.
Berridge, M. J., Downes, C. P., and Hanley, M. R. (1989). Neural and developmental actions of lithium: A unifying hypothesis. Cell 59, 411-419. Busa, W. B., and Gimlich, R. L. (1989). Lithium-induced teratogenesis in frog embryos prevented by a polyphosphoinositide cycle intermediate or a diacylglycerol analog. Dev. Biol 132,315-324. Christian, J. L., Gavin, B. J., McMahon, A. P., and Moon, R. T. (1991). Xwnt-8, a Xenopus wnt-l/i&-l related gene responsive to mesoderm inducing growth factors, may play a role in ventral mesodermal patterning during embryogenesis. DeveZopment 111,1045-1055. Christian, J. L., Olson, D. J., and Moon, R. T. (1992). Xwnt-8 modifies the character of mesoderm induced by bFGF in isolated Xenopus ectoderm. EMBO J. 11,33-41. Dorresteijn, A. W. C., Bilinski, S. M., van den Biggelaar, J. A. M., and Bluemink, J. G. (1982). The presence of gap junctions during early Patella embryogenesis: An electron microscopical study. Dev. Biol. 91,397-401.
Flagg-Newton, J. L., and Loewenstein, W. R. (1981). Cell junction and cyclic AMP. II. Promotion of junctional membrane permeability and junctional membrane particles in a junction-deficient cell type. .I Membr. Biol. 63,123-132.
Gimlich, R. L. (1986). Acquisition of developmental autonomy in the equatorial region of the Xenopus embryo. Dev. Biol. 115,340-352. Guthrie, S. C. (1984). Patterns of junctional permeability in the early amphibian embryo. Nature 311,149-151. Guthrie, S., Turin, L., and Warner, A. (1988). Patterns of junctional communication during development of the early amphibian embryo. Development 103,769-783. Hax, W. M. A., van Venrooij, G. E. P. M., and Vossenberg, J. B. J. (1974). Cell communication: A cyclic AMP mediated phenomenon. J. Membr. Biol. 19,253-265.
Kageura, H. (1990). Spatial distribution of the capacity to initiate a secondary embryo in the 32-cell embryo of Xenopus laevis. Dev. Biol. 142,432-452.
Kao, K. R., Masui, Y., and Elison, R. P. (1986). Lithium-induced respecification of pattern in Xenopus laevis embryos. Nature 322, 371373.
212
DEVELOPMENTALBIOLOGY
Kimelman, D., Abraham, J. A., Haapranta, T., Palisi, T. M., and Kirschner, M. (1988). The presence of fibroblast growth factor in the frog egg: Its role as a natural mesoderm inducer. Science 242,10531056. Kimelman, D., and Kirschner, M. (1987). Synergistic induction of mesoderm by FGF and TGF B and the identification of an mRNA coding for FGF in the early Xenopus embryos. Cell 51,869-8’77. Kimelman, D., and Maas, A. (1992). Induction of dorsal and ventral mesoderm by ectopically expressed Xenopus basic fibroblast growth factor. Development 114,261-269. Mathews, L. S., and Vale, W. W. (1991). Expression cloning of an activin receptor, a predicted transmembrane serine kinase. Cell 65,973982. McMahon, A. P. (1991). A superfamily of putative developmental signalling molecules related to the proto-oncogene Writ-l/i&-l. Adv. Deu. Biol. 1,31-60. McMahon, A. P., and Moon, R. T. (1989). Ectopic expression of the proto-oncogene in&l in Xenqpus embryos leads to duplication of the embryonic axis. Cell 58,1075-1084. Moon, R. T., and Christian, J. L. (1989). Microinjection and expression of synthetic mRNAs. Technique 1,76-89. Musil, L. S., Cunningham, B. A., Edelman, G. M., and Goodenough, D. A. (1990). Differential phosphorylation of the gap junction protein connexin 43 in junctional communication-competent and -deficient cell lines. J. Cell Biol. 111,2077-2088. Nagajski, D. J., Guthrie, S. C., Ford, C. C., and Warner, A. E. (1989). The correlation between patterns of dye transfer through gap junctions and future developmental fate in Xenopus; the consequences of U.V. irradiation and lithium treatment. Development 105,747-757. Nieuwkoop, P. D., and Faber, J. (1967). Normal table of Xenqpus luevis. Daudin. North Holland, Amsterdam. Olson, D. J., Christian, J. L., and Moon, R. T. (1991). Effect of writ-1 and related proteins on gap junctional communication in Xenopus embryos. Science 252,1173-1176. Otte, A. P., Kramer, I. M., and Durston, A. J. (1991). Protein kinase C and regulation of the local competence of Xenopus ectoderm. Science 251,570-573. Peifer, M., and Wieschaus, E. (1990). The segment polarity gene armadillo encodes a functionally modular protein that is the Drosophila homolog of human plakoglobin. Cell 63,1167-1178. Riggleman, B., Schedl, P., and Wieschaus, E. (1990). Spatial expression of the Drosophila segment polarity gene armadillo is posttranscriptionally regulated by wingless. Cell 63,549-560. Scharf, S. R., and Gerhart, J. C. (1983). Axis determination in eggs of Xenopus &is: A critical period before first cleavage, identified by the common effects of cold, pressure, and ultraviolet irradiation. Dev. Biol. 99, 75-87.
Sharpe, C. R., Fritz, A., Derobertis, E. M., and Gurdon, J. B. (1987). A homeobox-containing marker of posterior neural differentiation
VOLUME151,1992
shows the importance of predetermination in neural induction. Cell 50,749-758. Simpson, I., Rose, B., and Loewenstein, W. R. (1977). Size limit of molecules permeating the junctional membrane channels. S&nce 195,294-297. Slack, J. M. W., Darlington, B. G., Heath, J. K., and Godsave, S. F. (1987). Mesoderm induction in early Xenopus embryos by heparinbinding growth factors. Nature 326,197-200. Slack, J. M. W., and Isaacs, H. V. (1989). Presence of fibroblast growth factor in the early Xenoms embryo. Development 105,147-153. Smith, J. C. (1989). Mesoderm induction and mesoderm-inducing factors in early amphibian development. Development 105,665-678. Smith, J. C., Cooke, J., Green, J. B. A., Howes, G., and Symes, K. (1989). Inducing factors and the control of mesodermal pattern in Xenopus loxvis. Development Suppl 149-159. Smith, J. C., Price, B. M. J., Van Nimmen, K., and Huylebroeck, D. (1990a). Identification of a potent Xem mesoderm-inducing factor as a homologue of activin A. Nature 345,729-731. Smith, J. C., Symes, K., Hynes, R. O., and DeSimone, D. (1990b). Mesoderm induction and the control of gastrulation in Xenopus Levis: The roles of fibrodectin and integrins. Development 108,229-238. Sokol, S., Christian, J. L., Moon, R. T., and Melton, D. A. (1991). Injected Writ but not activin RNA induces a complete dorsal axis in Xenopus embryos. Cell 67,741-752. Sokol, S., and Melton, D. A. (1991). Pre-existent pattern in Xen0pu.s animal pole cells revealed by induction with activin. Nature 351, 409-411. Sokol, S., Wong, G. G., and Melton, D. A. (1990). A mouse macrophage factor induces head structures and organizes a body axis in Xeru+ pus Science 249,561-564.
Spemann, H., and Mangold, H. (1924). Uber induktion von embryoanalagen durch implantation arffremder organisatoren. Arch. Mikrosk.
Anat. Entwicklungsmech
100,599-638.
Thomsen, G., Woolf, T., Whitman, M., Sokol, S., Vaughn, J., Vale, W., and Melton, D. A. (1990). Activins are expressed early in Xenopus embryogenesis and can induce axial mesoderm and anterior structures. CeU63,485-493. van den Biggelaar, J. A. M. (1977). Development of dorsoventral polarity and mesentoblast determination in Patella vulgata J. Morphol. 154,157-186.
van den Biggelaar, J. A. M., and Guerrier, P. (1979). Dorsoventral polarity and mesentoblast determination as concomitant results of cellular interactions in the mollusc Pate& vulgata. Deu. Biol. 68, 462-471.
Warner, A. E., and Gurdon, J. B. (1987). Functional gap junctions are not required for muscle gene activation by induction in Xenspus embryos. J. Cell Biol. 104,557-564. Warner, A. E., Guthrie, S. C., and Gilula, N. B. (1984). Antibodies to gap junctional proteins selectively disrupt junctional communication in the early amphibian embryo. Nature 311,127-131.
