DEVELOPMENTAL

BIOLOGY

145,128-138

(19%)

Changes in Temporal and Spatial Patterns of Gi Protein Expression in Postimplantation Mouse Embryos JOLYNDA JONES, CATRIONA Y. LOGAN, AND RICHARD

M. SCHULTZ

We previously demonstrated the presence of GTP-binding proteins, G proteins, in the preimplantation mouse embryo (Jones and Schultz, 1990. Dm Bid. 139.250-262). These studies have been extended to the Day 6.5,7.5, and 8.5 gestation embryo by employing PT-catalyzed ADP-ribosylation and immunoblotting techniques. We report here that the amount of embryonic ai increases from Day 6.5 to Day 7.5 of gestation, and remains at about the same level at Day 8.5. In contrast, the extent of PT-catalyzed ADP-ribosylation of G,,, protein(s) decreases between Days 6.5 and 7.5-this decrease is global and not restricted to a particular germ layer of the Day 7.5 embryo-and then dramatically increases by Day 8.5 of gestation. In the Day 8.5 gestation embryo, the extent of PT-catalyzed ADP-ribosglation of G, proteins increases along the anterior-posterior axis, whereas the amount of immunoreactive q subunit decreases along this axis. By using a combination of PT-catalyzed ADP-ribosylation and immunoprecipitation with antisera specific for ql, q2, or oi3, we report that all three oi subtypes are present in the Day 8.5 gestation mouse embryo. Results of these experiments suggest that an activation of Gi proteins occurs between Days 6.5 and 7.5 of gestation in the postimplantation embryo, a time during which the embryo is gastrulating, and that a decreasing gradient of activation exists along the anterior to posterior axis in the Day 8.5 gestation embryo. Last, we report that oocgtes, eggs, and preimplantation embryos possess * 1991 Academic Press, Inc. all three subtypes of q.

ogenesis (Greene, 1989 and references therein; Hausman and Velleman, 1981; Runyan et al., 1990; Zalin, 1977). G proteins are composed of 01, p, and y subunits and comprise a family of proteins, which vary at least in the nature of the 01subunit. When activated by hormone/receptor coupling, the LYsubunit dissociates from the /3y complex and interacts with its target effector protein. The specificity of the interaction of the 01 subunit with the different effector proteins appears to be due to the different types of N subunits. For example, o(, activates adenylyl cyclase, whereas cyican inhibit adenylyl cyclase (Gilman, 1987 and references therein; Casey et al., 1988). In addition, aj can activate K+ channels or a phospholipase C (Brass et ul., 1986; Ueda et al., 1989; Mattera et ab, 1989). Three subtypes of ai-cyil, I+, and cY,-exist and have molecular weights in the range of 40,000-41,000 daltons (Gilman, 1987 and references therein; Casey et al., 1988). These Qli subunits, which are products of separate genes (Jones and Reed, 1987; Kaziro et al., 1988), are frequently identified by their ability to serve as substrates for ADP-ribosylation catalyzed by pertussis toxin (PT) in the presence of [32P]NAD; the inactive heterotrimeric form of Gi is the substrate for PT (Bokoch et ah, 1984; Katada et al., 1984). Previously, we utilized the PT-catalyzed ADP-ribosylation of Gi, subunits, as well as immunoblotting, to identify changes in the expression of G,-like proteins

INTRODIJCTION

Early postimplantation mouse development is characterized by the sequential differentiation of the inner cell mass of the blastocyst first into a bilaminar embryo, then into a trilaminar embryo, and finally into an embryo undergoing organogenesis (Rugh, 1968; Theiler, 1989). This series of events marks a significant transition from a preimplantation embryo that has undergone a series of reductional divisions in the absence of net growth (Hogan et ah, 1986; Brinster, 1967) to a postimplantation embryo that will undergo rapid changes in growth, differentiation, and morphogenesis (Rugh, 1968; Theiler, 1989). Signal transduction pathways are implicated in cell proliferation and differentiation, as well as differential gene expression, all of which are requisite for embryogenesis. In many instances, guanine nucleotide-binding regulatory proteins (G proteins) mediate specific signal transduction pathways, which result in the production of second messengers (Gilman, 198’7, and references therein). In fact, the preimplantation mouse embryo possesses G proteins (Jones and Schultz, 1990; Allworth et al., 1990), and CAMP appears to regulate specific aspects of preimplantation embryogenesis in the mouse (Manejwala et al., 1986, 1989; Poueymirou and Schultz, 1989). Moreover, signal transduction pathways are likely to exist in the early postimplantation embryo, since such pathways are implicated in regulating organ0012-1606/91 Copyright All rights

$3.00

IL 1991 by Academic Press, Inc. of reproduction in any form reserved.

128

JONES,LOGAN,ANDSHULTZ

during preimplantation mouse development. We have used these methods to extend this study to the early postimplantation embryo and report here that Gi proteins are present in the Day 6.5, 7.5, and 8.5 gestation mouse embryo. The amount of Ni increases from Day 6.5 to Day 7.5 of gestation, and remains at about the same level at Day 8.5. In contrast, the extent of PT-catalyzed ADP-ribosylation of Gi, proteins decreases between Days 6.5 and 7.5-this decrease is global and not restricted to a particular Day 7.5 embryonic germ layerand then dramatically increases by Day 8.5 of gestation. Along the anterior-posterior axis two gradients of Gi protein expression are present in the Day 8.5 gestation embryo. The extent of PT-catalyzed ADP-ribosylation of Gi, proteins increases along the anterior-posterior axis, whereas the amount of immunoreactive LYEsubunit detected by immunoblotting decreases along this axis. Last, by using a combination of PT-catalyzed ADP-ribosylation and immunoprecipitation with antisera specific for Nil, (Yi2, or Nj3, we report that all three ai subtypes are present in the Day 8.5 mouse embryo, as well as in oocytes, eggs, and preimplantation embryos. MATERIALS

Embryo

AND

METHODS

Collection

Nulliparous CF-1 female mice (6-8 weeks of age, Harlan, Sprague-Dawley), two per cage, were housed overnight with single B6D2F,/J males (Jackson Laboratory). The next morning, females were examined for vaginal plugs as evidence of mating, and this was designated Day 0.5 of gestation. On Days 6.5,7.5, and 8.5 of gestation, embryos were dissected from the decidua in bicarbonate-free minimal essential medium (Earle’s salts) containing pyruvate (100 pg/ml), gentamicin (10 pg/ml), polyvinylpyrrolidone (3 mg/ml), and 25 mM Hepes, pH 7.2 (MEM/PVP). Riechart’s membrane, the ectoplacental cone, and all extraembryonic membranes were removed from Day 6.5 (stage 9; Theiler, 1989), Day 7.5 (stage 10; Theiler, 1989), and Day 8.5 (stage 12; Theiler, 1989) embryos. All embryos were rinsed in calcium, magnesium-free phosphate-buffered saline (CMF-PBS), pH 7.2, and stored at -70°C in CMF-PBS containing 1% Lubrol-PX, leupeptin (10 /*g/ml), aprotinin (10 pg/ml), and phenylmethylsulfonyl fluoride (10 PIM) (CMF-PBS/PI). Prior to being assayed by the methods described below, the embryos were thawed in the presence of CMF-PBS/PI and homogenized on ice with ten strokes in a Dounce homogenizer. The total amount of embryonic protein was determined using the bicinchoninic acid (BCA, Pierce, Inc.) protein assay (Smith et al., 1985). Mouse oocytes, eggs, and preimplantation mouse embryos were collected as previously described (Jones and Schultz, 1990).

G Protcirls

in A-1o~t.w

E~~~hr,qos

129

Where stated, female mice were injected with 1 I.U. of pregnant mare’s serum gonadotropin (PMSG), and 48 hr later injected with 1 I.U. of human chorionic gonadotropin (hCG); this treatment regime resulted in a higher percentage of females that mated but did not induce superovulation. Females were housed with males overnight and screened for evidence of mating as described above. Embryos collected from gonadotropin-primed mice developed at the same rate as embryos from naturally mated mice (data not shown). Oocytes, eggs, and preimplantation embryos were collected as previously described (Jones and Schultz, 1990). Germ Lager Isolat%on Anterior ectoderm, posterior ectoderm and mesoderm (i.e., primitive streak region), and visceral embryonic endoderm were isolated from Day 7.5 embryos by enzymatic treatment with 0.5% trypsin and 0.25% pancreatin in CMF-PBS (Beddington, 1987). Enzymatic treatment was inhibited by placing isolated tissues in Dulbecco’s modified media (GIBCO) containing 20% fetal bovine serum (FBS), 0.1 mM sodium pyruvate, 50 pg/ml streptomycin, and 50 IU/ml penicillin. Isolated germ layers were washed extensively in CMF-PBS prior to storage at -70°C in CMF-PBS/PI. Preparation

r?fPlasma Membranes

Crude plasma membrane preparations were derived from normal rat kidney cells (NRK) (Woolkalis and Manning, 1987), platelets (PL) (Brass et uh, 1988), and rat brain (RB) (Sternweis and Robishaw, 1984) as described previously and were generously provided by Marilyn Woolkalis, Department of Pharmacology, University of Pennsylvania. Pertussis Toxin-Cutalyxed ADP-Ribosylation Embryos, NRK, PL, and RB Membranes

qf Mouse

PT-catalyzed ADP-ribosylation of mouse embryonic homogenates and control cellular plasma membrane preparations was performed as previously described (Jones et t~l., 1989). [32P]NAD was synthesized according to the method of Cassel and Pfeuffer (1978), except that an ATP regenerating system, consisting of pyruvate kinase and phosphoenolpyruvate, was included. When PT was omitted from the reaction mixture, the radiolabeled ADP-ribosylated species of M, = 38,000 and 39,000 were not detected (Fig. 1). Equal amounts of embryonic protein were radiolabeled and subjected to electrophoresis. PT-catalyzed ADP-ribosylation of Q, was performed by adding 2 pmole each of purified recombinant rat CY, and purified bovine brain B-y to the reaction mixture at the time of addition of PT. Following the incubation, the

130

DEVELOPMENTAL

1 Mrx

2

3 1'

2'

BIOLOGY

3'

10-3

45-

29-

FIG. 1. Autoradiograms of day 6.5,7.5, and 8.5 embryo homogenates incubated with or without PT and with [32P]NAD. The experiments were performed as described under Materials and Methods, and 2.5 fig of protein was used for each sample. Lanes 1-3, PT present; lanes l’-3, PT absent. Lanes 1 and I’, Day 6.5 embryos; lanes 2 and 2, Day 7.5 embryos; lanes 3 and 3, Day 85 embryos. The experiment was performed three times with similar results. Shown is a representative example. The radiolabeled species in the Day 7.5 gestation embryo (lane 2) are present but did not reproduce well.

samples were alkylated with N-ethylmaleimide as previously described (Sternweis and Robishaw, 1984). Polyacrylamide

Gel Electrophoresis

Samples were subjected to electrophoresis in 10% polyacrylamide gels containing sodium dodecyl sulfate (SDS) according to the method of Laemmli (Laemmli, 1970). [32P]radiolabeled species were detected by autoradiography at -70°C using Kodak X AR-5 X-ray film and an intensifying screen. Exposure times ranged from 1 to 16 hr. The SDS used in these studies is the 95% grade from Sigma. This SDS, which contains higher chain length homologs relative to more pure forms of SDS, results in resolving two ADP-ribosylated proteins of M, = 38,000 and 39,000. Pure forms of SDS only resolve a single species of M, = 41,000 (Jones et al., 1989; Jones and Schultz, 1990). To compare the electrophoretic mobility of ADP-ribosylated cy, and embryonic protein that was alkylated, polyacrylamide gel electrophoresis was conducted in 11% gelan and a more pure form of SDS was used. Although this resulted in only the separation of a single species for the embryo-derived material, it afforded ample resolution of (Y~from cy,.

VOLUME

145, 1991

length); in contrast to the large gels, the mini-gels do not resolve well the two species that are detected by PT-catalyzed ADP-ribosylation. Following electrophoresis, proteins were transferred to Immobilon-P (Millipore) or nitrocellulose (Schleicher and Schuell) as previously described (Jones and Schultz, 1990). Silver staining of mini-gels (Wray et al., 1981) or Coommassie staining of standard gels revealed that essentially complete transfer of proteins in the M, = 25,000-100,000 range occurred. Antisera specific for olilm3,cyil, ai2, ai3, o(,, or cy,, as well as an antiserum that recognizes part of the GTP-binding site common to all o( subunits were used; these antisera were generated against the peptides described in Table 1 and were generous gifts from Dave Manning, Department of Pharmacology, University of Pennsylvania. Binding of the primary antisera (1:lOO dilution used in all experiments) to specific LYsubunits on Immobilon-P or nitrocellulose was visualized using either a Vectastain peroxidase kit (Vector) and diaminobenzidine (Sigma), or 1251-protein A (Amersham, sp. act. 100 &i/ml). When 1251-protein A was utilized the Immobilon-P or nitrocellulose membranes were incubated with the appropriate primary antisera and washed in TrisHCl-buffered saline (TBS), pH 7.5, as described previously (Jones and Schultz, 1990). Following the wash, the membranes were incubated with 1251-protein A (1 pCi/ml) for 30 min. The membranes were then washed in TBS containing 0.05% Tween-20 and air dried. [1251]protein A labeled proteins were detected by autoradiography at -70°C using Kodak X AR-5 X-ray film. Immunoprecipitution Equal amounts of protein from Day 6.5, 7.5 and 8.5 gestation embryos were subjected to PT-catalyzed ADP-ribosylation in the presence of [32P]NAD. The radiolabeled proteins were then immunoprecipitated according to the method of Carlson et al. (1989) except that 1:lO dilutions of the antisera were used. In order to demonstrate antisera specificity, the specific synthetic peptides described in Table 1 for oil, ai2, or Lyi3(200 pg/ml) were initially incubated for 8 hr at 4°C with antisera directed against ail, ai2, or ai3, respectively, prior to immunoprecipitation. The synthetic peptides were a gift from Dave Manning, Department of Pharmacology, University of Pennsylvania. Densitometry

Immunoblotting Equal amounts of protein from Day 6.5, 7.5, and 8.5 gestation embryos were subjected to electrophoresis in either 10% polyacrylamide-SDS mini-gels (5 cm

Densitometry was performed on autoradiograms of samples subjected to either PT-catalyzed ADP-ribosylation or immunoblotting. For each case, the exposure time for the autoradiograms was such that the increase

G Proff,itcs

JONES,LOGAN,ANDSHULTZ

SPECIFICITY Antiserum

0~ PEPTIDE

it/ Mo1r.w

TABLE 1 ANTISERA TO (Y SIJBUNIT

131

Ev~l~qos

SUBTYPES

Recognizes

Peptide

Reference

KNNLKDCGLF CDLDRIAQPNYI CDLERIAQSDYI RADDARAAEGFDIC CEYGDKERKADSK CTGPAESKGEITPELI, CGAGESGKTIVKQMK ” This antisera recognizes Ilniwrsity of Pennsglvania).

LYEfive times

greater

Carlson vf trl.. 1989 Williams f’f ul.. 1990 Carlson c,f (I/. 1 1089 . LVilliams fat t/l.. 1990 Personal communication” Carlson c? trl.. 1989 Carlson vf trl., 1989 than

(Y, polypeptides

in densitometer units was linear with respect the amount of PT-catalyzed ADP-ribosylation immunoreactive (yi (data not shown).

to either of (Y~or

RESULTS

Since we have previously demonstrated developmental changes in the pattern of Gi protein expression during preimplantation development, we extended these studies by examining Gi protein expression during postimplantation development. As development proceeds from Day 6.5 to Day 7.5 (primitive streak stage; stage 10 Theiler, 1989) a marked decrease in the extent of PT-catalyzed ADP-ribosylation was observed (Fig. 1); densitometry revealed an average decrease of about 5fold (Fig. 2). A dramatic increase in the extent of PT-ca-

loooo l-----l

in LA 6.5

7.5

a.5

(personal

communication,

David

Manning,

Department

of Phxrmacolo~~.

talyzed ADP-ribosylation was observed between Day 7.5 and Day 8.5 of development, at which time organogenesis is occurring (Fig. 1); densitometry revealed an average increase of about 22-fold relative to Day 6.5 (Fig. 2). The ADP-ribosylated species detected in the postimplantation embryos had an M, = 38,000 and 39,000, which was very similar, if not identical to that present in preimplantation embryos (Jones and Schultz, 1990). Results of two mixing experiments that used equal amounts of Day 7.5 and Day 8.5 embryonic protein revealed that the extent of PT-catalyzed ADP-ribosglation was additive (data not shown). This result minimized the likelihood that an endogenous inhibitor was present in the Day 7.5 embryo homogenate and responsible for the decrease in the extent of PT-catalyzed ADPribosylation observed in the Day 7.5 embryo. Although there is no overt neuronal differentiation in the Day 8.5 embryo (Buse and Krisch, 1987), the increase in the extent of PT-catalyzed ADP-ribosylation could be due to an increase in the amount of G,, which is also a substrate for PT and is abundant in neuronal tissue (Sternweis and Robishaw, 1984). This possibility was most unlikely, since using an electrophoretic system that readily resolves (Y, from (yi (Sternweis and Robishaw, 1984), the ADP-ribosylated species present in the Day 8.5 embryo had an electrophoretic mobility that differed from that of recombinant CY,(Fig. 3).

1

2

Age (days) FIG. 2. Relative changes in the extent of PT-catalyzed AL)P-ribosylation in day 6.5, 7.5, and X.5 embryo homogenates. The autoradiograms were subjected to densitometry and the data from three separate experiments have been pooled and are expressed as the mean k SEM relative to the extent of PT-catalyzed ADP-ribosylation in Day 6.5 embryo homogenates. Note that the ordinate is a log scale and that the intensity of the signal obtained for the Day 8.5 embryo material is beyond the linear range of the densitometer; thus the increase that occurs by Day 8.5 is an underestimate. The differences between any two groups is significant (I’ < 0.02, Student’s t test).

FIG. 3. Region of an autoradiogram of Day 8.5 embrgonic material and recombinant q, subjected to PT-catalyzed ADP-ribosglation. The experiment was performed as described under Materials and Methods: 1.5 PB of Daq’ 8.5 emhrvo protein and 2 pmole of recombinant o, were used. Lane 1, Day 8.5 embryo; lane 2, recombinant (Y,. The exposure times were adjusted to yield relatively similar signal levels. In this gel clectrophorctic system, q migrates slower than q, (Sternweis and Robishaw, 19841.

132

DEVELOPMENTAL

BIOLOGY

VOLUME

145, 1991

Day

FIG. 4. Immunoblot of Day 6.5,7.5, and 8.5 gestation embryos using antiserum to Gi and the [‘=I]protein A detection method using serum 8730. The experiment was performed as described under Materials and Methods and 12.5 wcg of protein was used for each sample. Lane 1, NRK, 2 pg; lane 2, Day 6.5 embryos; lane 3, Day 7.5 embryos; lane 4, Day 8.5 embryos. Preimmune serum did not recognize the immunoreactive q species (data not shown).

Since equal quantities of embryonic protein were subjected to PT-catalyzed ADP-ribosylation, the changes in the extent of labeling of 01~could be attributed to either alterations in the state of dissociation of Gi-the heterotrimeric, undissociated form of Gi is the substrate for PT (Bokoch et ul., 1984; Katada et al., 1984)-or changes in the amount of Gi, or both. Immunoblotting experiments were performed to discriminate between these possibilities. Immunoblotting experiments were conducted with a polyclonal antibody 8730 (Table 1) that recognizes (Yil and cyiBequally, cyi3at 50% the level of ail and Ni2, and cy, at 510% the level of ai1 and c+ (Carlson et ah, 1989). The amount of immunoreactive ai was greater in the Day 7.5 and Day 8.5 embryos when compared to the Day 6.5 embryo (Figs. 4 and 5). The electrophoretic mobility of this ai subunit (M,. = 38,000) was the same as that of cyi present in NRK membranes (Marilyn Woolkalis, personal communication; Jones et ul., 1989; Jones and Schultz, 1990). On Day 7.5, primitive streak formation begins and results in the formation of a trilaminar embryo, which consists of embryonic ectoderm, embryonic mesoderm, and visceral endoderm. Thus, the decrease in the extent of PT-catalyzed ADP-ribosylation that was observed in the Day 7.5 embryo could have been due to a global decrease or a decrease in a specific germ layer. In order to ascertain whether regional variations in the extent of PT-catalyzed ADP-ribosylation of Gi,, occur in specific Day 7.5 embryonic germ layers, the germ layers of the Day 7.5 embryo were enzymatically isolated (Beddingtion, 1987), and equal amounts of protein were subjected to PT-catalyzed ADP-ribosylation. Since isolation of em-

6.5

Day

7.5

Day 6.5

FIG. 5. Relative changes in the amount of immunoreactive q in Day 6.5, 7.5, and 8.5 embryo homogenates. The immunoblots were subjected to densitometry and the data from two experiments in which the [r”I] detection method was used and one experiment in which the peroxidase detection method was used have been pooled and are expressed as the mean k SEM relative to the amount in the Day 6.5 embryo. When the peroxidase detection method was used, a positive negative was scanned. The increases observed at either Day 7.5 or Day 8.5 relative to Day 6.5 are significant (E < 0.01, Student’s t test) and the differences between Day 7.5 and Day 8.5 are significant (P < 0.02, Student’s t test ).

bryonic mesoderm from the posterior embryonic ectoderm was difficult, these two tissues were combined. The extent of PT-catalyzed ADP-ribosylation was similar in the embryonic anterior ectoderm, visceral embryonic endoderm, and embryonic mesoderm plus posterior ectoderm (Figs. 6 and 7). Therefore, the decrease in ADP-ribosylation of ai in Day 7.5 embryos when compared to that in the Day 6.5 embryo was not due to a decrease in the extent of PT-catalyzed ADP-ribosylation of Gi in one embryonic germ layer versus another. Also consistent with this interpretation was the observation that immunoblots of protein isolated from Day 7.5 germ layers demonstrated that although the differences in the amount of immunoreactive material in the three germ layers were significant, the three germ layers, nevertheless, contained roughly similar amounts of cyi (Fig. 8); densitometric analysis of these experiments is shown in Fig. 7.

123

45

6 &”

7”

FIG. 6. Region of an autoradiogram of Day 7.5 embryonic germ layers subjected to PT-catalyzed ADP-ribosylation. The experiment was performed as described under Materials and Methods. Lanes 1 and 4, 2 ).~cgprotein of Day 7.5 embryonic visceral endoderm; lanes 2 and 5, 2 ~g protein of Day 7.5 embryonic anterior ectoderm, lanes 3 and 6, 2 (~g protein of Day 7.5 embryonic mesoderm and posterior ectoderm. The experiment was performed three times with similar results. Results of two experiments are shown.

133

,JONES,LOGAN,ANDSHULTZ

12345678

I--

II1 AE

Endo

MPE

FIG. 7. Relative changes in the extent of PT-catalyzed ADP-ribosylation or amount of immunoreactive CY,in germ layers of Day 7.5 embryos Densitometry was performed and the data are expressed as the mean i SEM relative to the value obtained for the endoderm. The results of three PT-catalyzed ADP-ribosylation and four immunoblotting experiments were pooled. Solid bars, PT-catalyzed ADP-ribosylation; open bars. immunoreactive q. For the PT-catalyzed ADP-ribosylation experiments, the decreases in the anterior ectoderm (AE) and mesoderm and posterior ectoderm (MPE) are significant (P < 0.05, Student’s f test) when compared to endodoerm (ENDO). The differences between AE and MPE are not significantly different. For the amount of immunoreactive (yi, the difference between any two groups is significant (P < 0.001, Student’s t test).

Spafiul Chunges in the Pattern in the Day ti.5 Embryo

of G Protein

Expression

To ascertain if a gradient of Gi protein expression existed along the anterior-posterior axis of the Day 8.5

1 2

3 1’ 2’ 3’

MI x 1C3 84-

27-

FIG. 8. Immunoblot of Day 7.5 embryonic germ layers using antiserum to Gi. The experiment was performed as described under Materials and Methods using the [lwI] detection method and serum 8730, and 5 pg of protein was used for each sample. Lanes 1-3, antiserum 8730; lanes l’-3’, normal rabbit serum. Lanes 1 and l’, Day 7.5 embryonic visceral endoderm; lanes 2 and 2’, Day 7.5 embryonic anterior ectoderm; lanes 3 and 3, Day 7.5 embryonic mesoderm and posterior ectoderm. The chevron points to the region where q migrates in the mini-gel. The experiment was performed three times with similar results. Shown is a representative example. The signal along the outer edges of the lanes is due to nonspecific binding to the prestained molecular weight standards.

FIG. 9, Region of an autoradiogram of Day 85 anterior, middle, and posterior embryonic regions subjected to PT-catalyzed ADP-rihosylation. The experiment was performed as described under Materials and Methods. Lane 1, 1 gg protein NRK plasma membranes plus PT; lane 8,l pg protein NRK plasma membranes minus PT; lanes 3 and 6,2 p’g of anterior region Day 8.5 embryonic protein plus PT; lanes 4 and 7, 2 pg of middle region Day 8.5 embryonic protein plus PT; lanes 5 and 8,2 fig of posterior region Day X.5 embryonic protein plus PT. The experiment was performed three times with similar results. Results of two experiments are shown.

gestation embryo, the Day 8.5 embryo was dissected into three portions denoted as anterior, middle, and posterior and each portion was subjected to either PT-catalyzed ADP-ribosylation or immunoblotting. The anterior Day 8.5 segment consisted of that portion of the embryonic neural folds anterior to the optic sulcus (Theiler, 1989). The middle portion consisted of the embryonic material anterior to the first visible dorsal somite up to the optic sulcus, including the developing heart and presumptive anterior gastrointestinal tract. The posterior portion of the Day 8.5 gestation embryo consisted of all embryonic structures from the first dorsal somite to the most posterior portion of the embryo, excluding the allantois. When equal amounts of embryonic protein obtained from these three regions were subjected to PT-catalyzed ADP-ribosylation, an increase in the extent of ADP-ribosylation of ai was observed along the anterior to posterior axis (Figs. 9 and 10). In contrast, results of immunoblotting experiments using the 8730 antiserum and equal amounts of embryonic protein obtained from these three regions revealed a reciprocal gradient, i.e., a decrease in the amount of immunoreactive material along the anterior to posterior axis (Fig. 11); densitometric analysis of these experiments is shown in Fig. 10. Results of two mixing experiments that used equal amounts of anterior and posterior embryonic protein revealed that the extent of PT-catalyzed ADP-ribosylation was additive (data not shown). This result minimized the likelihood that an endogenous inhibitor was present in the Day 8.5 gestation embryo and responsible for differences in the extent of PT-catalyzed ADP-ribosylation along the anterior to posterior axis.

Subtypes of q Present in the Day 8.5 Embryo In order to assessthe complement of G proteins present in the early postimplantation embryo, Day 8.5 embryonic protein was screened with a series of antisera

134

DEVELOPMENTAL

BIOLOGY

VOLUME

145,

123456789

1991 10

11

A B

A

M

P

10. Relative changes in the extent of PT-catalyzed ADP-ribosylation and amount of immunoreactive q in different regions of the Day 8.5 embryo. Densitometrp was performed and the data are expressed as the mean i- SEM relative to the value obtained for the anterior region. The results of three PT-catalyzed ADP-ribosylation and four immunoblotting experiments were pooled. Solid bars, PTcatalyzed ADP-ribosylation; open bars, immunoreactive q. For the PT-catalyzed ADP-ribosylation experiments, the increases in the middle (M) and posterior (P) regions are significant (1’ < 0.02, Student’s t test) when compared to the anterior (A) region; the differences between the M and P regions are also significant (P < 0.0.5, Student’s t test). Similarly, for the amount of immunoreactive (yi, the decreases in M and P are significant (P < 0.001, Student’s f test) when compared to A, and the difference between M and P are significant (P < 0.025, Student’s t test). FIG.

directed against the specific peptides listed in Table 1. Day 8.5 gestation embryos, which were from gonadotropin-primed mice, were used for two reasons. First, sufficient quantities of Day 8.5 embryonic protein could be obtained easily for the immunoblotting and immuno-

1 2

3

4

5 1' 2' 3' 4' 5'

Mrx lo3 84-

FIG. 11. Immunoblot of anterior, middle, and posterior Day 8.5 embryonic regions using antiserum to Gi. The experiment was performed as described under Materials and Methods using the peroxidase detection method and serum 8’730, and 6 pg of protein was used for each sample. Lanes l-5, antiserum 8730; lanes l’-5’, normal rabbit serum. Lanes 1 and l’, 4 pg protein NRK plasma membranes; lanes 2 and 2’,2 pg protein NRK plasma membranes; lanes 3 and 3, anterior region Day 8.5 embryonic protein; lanes 4 and 4’, middle region Day 8.5 embryonic protein; lanes 5 and 5, posterior region Day 8.5 embryonic protein. The chevron points to the region where q migrates in the minigel. The experiment was performed three times with similar results. Shown is a representative example.

FIG. 12. Region of an autoradiogram of NRK plasma membranes and Day 8.5 gestation embryo homogenates subjected to PT-catalyzed ADP-ribosylation and immunoprecipitation with antisera listed in Table 1. The experiment was performed as described under Materials and Methods. (A) NRK plasma membranes (5 eg protein each lane); (B) Day 8.5 gestation embryo homogenates (5 pg protein each lane). Lane 1, normal rabbit serum; lane 2, 8730 antiserum; lane 3,3646 antiserum; lane 4, 3646 antiserum initially incubated with 200 Kg/ml ql specific synthetic peptide; lane 5, 1521 antiserum; lane 6, 1521 antiserum initially incubated with 200 @g/ml qZ specific synthetic peptide; lane 7,1518 antiserum; lane 8,1518 antiserum initially incubated with 200 pg/ml q3 specific synthetic peptide; lane 9, 2918 antiserum: lane lo,1398 antiserum; lane 11.2919 antiserum. The experiment was performed two times with similar results. Note that the two bands present in lane 3 did not reproduce well in the photograph. In addition, the two bands present in lanes 2 and 10 in each panel are not obvious, since the autoradiogram is overexposed for these lanes.

precipitation procedures. Second, Day 8.5 of gestation marks the beginning of organogenesis (Rugh, 1968; Theiler, 1989) and G proteins modulate a series of second messenger pathways known to regulate organ formation. For example, CAMP is implicated in palate (Greene, 1989 and references therein) and muscle (Hausman and Velleman, 1981; Zalin, 1977) development, while protein kinase C and G proteins are implicated in cardiac development (Runyan et al., 1990). Results of immunoblotting experiments that used 100 pg of protein from Day 8.5 gestation embryos and the antisera listed in Table 1 revealed a consistent signal for ai (1521) and GTP-binding site common (1398) antisera when either the Vectastain or [lz51] protein A method of detection was used (data not shown). Only the Vectastain method of detection demonstrated a weak signal for cr, using the 2919 antiserum (data not shown). Since efforts to increase the sensitivity of detection of (yil, (yi3, and 01,in the immunoblots were not successful, immunoprecipitation of the PT-catalyzed ADP-ribosylated species in the Day 8.5 gestation embryo was used. Antisera 8730 and 1398 (Table 1) immunoprecipitated species of M,. = 38,000 and 39,000 from Day 8.5 embryo homogenates, as well as from membrane preparations derived from platelets, NRK cells, or rat brain, which served as positive controls (Figs. 12 and 13). Antiserum 3646, which is specific for Lyi1,also immunoprecipitated both of these radiolabeled species, except that the species of AJ, = 39,000 was predominant in the Day 8.5 embryonic protein samples as compared to the M, = 38,000 species (Fig. 12). A similar situation was observed for rat brain, whereas an equal intensity of

JONES, LOGAN,

AND SHULTZ

12345678

A

**-

B FIG. 13. Region of an autoradiogram of rat brain plasma membranes and platelet plasma membranes subjected to PT-catalyzed ADP-ribosylation and immunoprecipitation with antisera listed in Tahle 1. The experiment was performed as described under Materials and Methods. (A) Rat brain plasma memhranes (5 kg protein each lane); (B) platelet plasma membranes (5 PK protein each lane). Lane 1, normal rabbit serum; lane 2, 8730 antiserum; lane 3, 3646 antiserum; lane 4,1521 antiserum; lane .i, 1518 antiserum; lane 6,2918 antiserum; lane 7, 1398 antiserum; lane 8, 2919 antiserum. The experiment was performed two times with similar results.

both M, = 38,000 and 39,000 species was present in platelet plasma membranes, and 1M, = 38,000 species was more predominant in NRK plasma membranes (Figs. 12 and 13). These changes in intensity of the two immunoprecipitated species could reflect differences in either their relative amounts present in the homogenate or the efficiency of radiolabeling of the two species. In addition, it may have been possible that these differences were due to other (Yisubtypes that may have been recognized by this antiserum. Within the limits of detection, antiserum 1521, which is specific for Ni2, only precipitated the M,. = 38,000 species, and likewise antiserum 1518, which is specific for Q, only precipitated the M, = 39,000 species in the Day 8.5 embryo and the other cell types tested (Figs. 12 and 13). The immunoprecipitation was specific, since an initial incubation of the antisera directed against Nil (3646), (yi2(1521), or ai (1518) with the appropriate peptide used to generate each of these sera resulted in a dramatic reduction in the amount of immunoprecipitated material (Fig. 12). Little radiolabeled immunoprecipitate was observed when antiserum 2918, which is specific for LY,,was used and compared to normal rabbit serum (Figs. 12). Although this antiserum did not efficiently immunoprecipitate radiolabeled proteins from rat brain membrane preparations (Fig. 13), which are known to contain G, (Marilyn Woolkalis, personal communication; Jones and Reed, 1987; Sternweis and Robishaw, 1984), our previous results which demonstrated that the electrophoretic mobility of the ADP-ribosylated species present in the embryo differed from that of CY,indicated that CY,was either not present or present at very low levels in the embryo. Since G,, does not contain the requisite cysteine residue for PT-catalyzed ADP-ribosylation (Fong ef al., 1988), as expected, antiserum 2919, which is specific for cy,, did not immunoprecipitate radiolabeled proteins in any instance (Figs. 12 and 13).

G P,Y)l(Ji,,S

in Mouse

135

Etrtbr.,/ps

Subtypes of ai Presed in Oocytes, Eggs, U& Preimphltution Embryos The amount of (yi is relatively constant from the GVintact oocyte to the two-cell embryo stage (Jones and Schultz, 1990; Allworth et ub, 1990). Relative to the twocell embryo, a reduced amount of cyi is present in the eight-cell and morula stages and this amount increases by the blastocyst stage (Jones and Schultz, 1990; Allworth ef al., 1990). In order to ascertain the different types of cyipresent and whether developmental changes in the types of ai expression occurred during this developmental period, a similar immunoprecipitation appreach to that described for the Day 8.5 embryo was undertaken. Similar to the Day 8.5 embryo, oocytes, ovulated eggs, and preimplantation embryos all possessedCQ, Q, and (xi31and cyi2was apparently the most abundant species (Fig. 14). In addition, the relative amounts of each of these subtypes of Ni remained essentially constant. DISClJSSION

In this report we extend the results of our previous study that documented developmental changes in the amount of ai and extent of PT-catalyzed ADP-ribosylation of cyiin the mouse preimplantation embryo (Jones and Schultz, 1990) by performing similar experiments with the early postimplantation embryo. We observe that the amount of (yi increases almost 2-fold from Day 6.5 to Day 7.5 and then decreases in the Day 8.5 embryo

1

oocyte

2

3

I”-...

4

5

-

8-cell morula blastocyst FIG. 14. Region of autoradiograms of oocytes, eggs, and preimplantation embryos subjected to PT-catalyzed ADP-ribosylation and immunoprecipitation with antisera listed in Table 1. Lane 1, normal rabbit serum; lane 2,873O antiserum; lane 3,3546 antiserum; lane 4,152l antiserum; lane 5, 1518 antiserum.

136

DEVELOPMENTALBIOLOGY

to a value about 50% greater than that observed in the Day 6.5 embryo. In contrast, the extent of PT-catalyzed ADP-ribosylation decreases about 5-fold between Day 6.5 and Day 7.5, and then increases about 22-fold from Day 7.5 to Day 8.5. The decrease in the extent of ADP-ribosylation that occurs between Day 6.5 and Day 7.5 appears not to be restricted to a specific germ layer. In the Day 8.5 embryo, which contains all three ai subtypes, the amount of oli decreases along the anterior to posterior axis, whereas the extent of PT-catalyzed ADP-ribosylation increases along this axis. In addition, we document that oocytes, eggs, and preimplantation embryos contain all three subtypes of ai and that the relative amounts of each species remains relatively constant during this period of development; thus, the changes in expression of cuithat occur during preimplantation development are coordinate. The results of the immunoprecipitation experiments provide the likely explanation for resolving two ADPribosylated species under our electrophoretic conditions that employ 95% pure SDS (Jones et al., 1989). Both ai1 and oli3have an electrophoretic mobility less than that of oli2.The apparent equal intensity of the signals is also accounted for by the observation that the intensity of the signal from ai is greater than that of either ai1 and ai3. The presence of a doublet for ai1 may represent some proteolysis or post-translational modification, e.g., phosphorylation (Carlson et al., 1989; Gundersen and Devreotes, 1990). The decrease in the extent of PT-catalyzed ADP-ribosylation of cyibetween Day 6.5 and Day 7.5 of gestation occurs at a time when the amount of cyiincreases. Since the heterotrimeric form of Gi is the substrate for PT (Bokoch et al., 1984; Katada et ab, 1984), this decrease in the extent of PT-catalyzed ADP-ribosylation may reflect an increase in the amount of dissociated Gi proteins that are present in the Day 7.5 embryo and are capable of interacting with their appropriate effector molecules. The decrease in the extent of PT-catalyzed ADP-ribosylation is not restricted to a specific germ layer(s) in the Day 7.5 embryo. Consistent with this global decrease is that similar amounts of cyi, as detected by immunoblotting, appear to be present in each of the germ layer dissections. It is possible, however, that the enzymatic method employed to isolate the germs layers may affect the ability of the Gi proteins, which are membranebound, to serve as substrates for PT-catalyzed ADP-ribosylation, and thus artifactually mask regional changes in the extent of PT-catalyzed ADP-ribosylation. The potential global activation of Gi occurs between a time when the bilaminar embryo becomes trilaminar, i.e., the formation of mesoderm as a result of massive

VOLUME 145,lW

cell migration in the gastrulating embryo. It will be of interest to ascertain if the decrease in the extent of PTcatalyzed ADP-ribosylation occurs prior to, concomitant with, or subsequent to the initiation of gastrulation. In this regard it will also be interesting to define more precisely when the increase occurs in the extent of PT-catalyzed ADP-ribosylation between Day 7.5 and Day 8.5; this increase occurs during a period of time in which the amount of ai remains essentially constant and is consistent with a functional inactivation of Gi. Again, results of such studies may indicate roles for Gi during this period of time in which organogenesis commences. It should be noted, however, that the changes in extent of PT-catalyzed ADP-ribosylation could be due to changes in the amount of 07 present in the developing embryo, since the heterotrimeric form of Gi is the substrate for PT. In fact, results of a recent study indicate that a preferential decrease in the amount of Pr relative to N occurs during mouse oocyte maturation; ,+y decreases about 69% whereas 01decreases about 40% (Allworth et al., 1990). The maturation-associated decrease in the extent of PT-catalyzed ADP-ribosylation of cyicorrelates with these decreases in Gi protein subunits (Jones and Schultz, 1990; Allworth et al., 1990). If such changes in the ratio of o( to /3r are the molecular basis for the observed changes in the extent of PT-catalyzed ADP-ribosylation of 01~that we report here, then the dramatic loo-fold increase in the extent of PT-catalyzed ADP-ribosylation that occurs between Day 7.5 and Day 8.5 may reflect a pronounced accumulation of 07. It should be noted that this accumulation of /3r would occur during a period of time in which the results of immunoblotting experiments indicate that the amount of (Y~ stays relatively constant. Whether such changes in p subunit expression occur during this period of development could be examined by immunoblotting experiments using antibodies to the /3 subunit. The Day 8.5 embryo has several subtypes of Cyipresent -that are detected by antibodies raised to pep-ai1-3 tides specific for the different types of N subunits. The most prevalent form detected by either immunoprecipitation following ADP-ribosylation or immunoblotting is Lyiz.It is not known, however, whether this reflects the greater abundance of this type of LYprotein, or if the antiserum that detects this subtype is better in both the immunoprecipitation and immunoblotting experiments. Regardless, it is clear that the Day 8.5 embryo contains aile3, and possibly N,. It is unlikely that (Y, is present in readily detectable amounts in the Day 8.5 embryo, since the electrophoretic mobility of the ADP-ribosylated species present in the embryo differs from that of recombinant cy,. Results of immunocytochemical experiments using specific antisera or in situ hybridization experiments using specific riboprobes may reveal spatial pat-

137

JONES, LOGAN, AND SHULTZ

terns of expression that suggest possible functions for these different 01subtypes. Two opposing gradients along the anterior to posterior axis are present in the Day 8.5 gestation embryo; the amount of ai decreases along this axis, whereas the extent of PT-catalyzed ADP-ribosylation of q increases. Thus, when compared to the posterior region, the anterior region apparently has greater amounts of G, protein, and, furthermore, more of this protein is in a potentially activated form. It is interesting to note that this anterior to posterior gradient in the amount and level of potential activity of the Gi follows the anterior to posterior gradient of embryonic development (Rugh, 1968; Theiler, 1989). The anterior structures of the embryo develop prior to the posterior portions of the embryo as is evident in both somite and limb formation (Rugh, 1968; Theiler, 1989). This same anterior to posterior gradient is evident for the protein product of the oncogene id-1 (Wilkinson et al., 1987), terminal differentiation proteins such as muscle myosins (Sassoon et trl., 1989), and for transcripts of homeobox genes Hox 1.1 (McMahon ef ul., 1988), 1.3,1.4,1.5,3.1, and 6.1 (Gaunt ef (II., 1988). The pattern of Gi protein expression and activation reported here may be another manifestation of this developmental gradient. This research was supported by a grant from the NIH (HD 22681) to R.M.S. J.J. thanks the Manning lab for their )qnerous suppI>- of reagents and constant advice on their use. We are very grateful to Drs. Maurine Linder and Patrick Cases forgenerouslg providing the recombinant (Y, and [j-y, respectivclg. We also thank Greg Kopf, Dave Manning. and Marilyn Roolkalis for critically reading the manuscript and making useful sug,rgestions, as well as the reviewers for their constructivc, comments. REFERENCES ALLWORTH, A. E., HILDERBRANDT, J. D., and ZIOMEK, C. A. (1990). Differential regulation of G protein suhunit expression in mouse ooqtes, eggs, and earlg embryos. DC,,,.Biol. 142, 129-137. BEDDINGTON, R. (1987). Isolation, culture and manipulation of postimplantation mouse embryos. 11, “Mammalian development” (M. Monk, Ed.), pp. 43-69. IRL Press, Oxford. BOKOCH, G. M., KATADA, T., NORTHUP, J. K., UI, M., and GILMAN, A. G. (1984). Purification and properties of the inhihitorg guanine nucleotide-hindinK regulatory component of adenylate cq’clasc. tJ, Bid. (‘hc’m. 2.59, 3560-3567. BRASS, I,. F., LAPOSATA, M., BANGA, H. S., and RITTENHOUSE, S. E. (1986). Regulation of phosphoinositide hgdrolysis pathway in thromhin-stimulated platelets by a pertussis toxin-sensitive guanine nucleotide-binding protein. Evaluation of its contribution to platelet activation and comparisons with the adenylate cyclase inhibitory protein, Gr. J. Bird C//o,/. 261, 16X38-16847. BRASS, L. F., WOOLKALIS, M. J., and MANNING, D. R. (19X8). Interactions in platelets between G proteins and the agonists that stimulate phospholipase (: and inhibit adenylgl cgclase. ,I. Bid. Chon. 263, 534X-5‘355. . .