Cleavage arrest of early frog embryos by the G protein-activated protein kinase PAK I.

PAK I is a member of the PAK (p21-activated protein kinase) family and is activated by Cdc42 (Jakobi, R., Chen, C.-J., Tuazon, P. T., and Traugh, J. A. (1996) J. Biol. Chem. 271, 6206-6211). To examine the effects of PAK I on cleavage arrest, subfemtomole amounts of endogenously active (58 kDa) and inactive (60 kDa) PAK I and a tryptic peptide (37 kDa) containing the active catalytic domain were injected into one blastomere of 2-cell frog embryos. Active PAK I resulted in cleavage arrest in the injected blastomere at mitotic metaphase, whereas the uninjected blastomere progressed through mid- to late cleavage. Injection of other protein kinases at similar concentrations had no effect on cleavage. Endogenous PAK I was highly active in frog oocytes, and antibody to PAK I reacted specifically with protein of 58-60 kDa. PAK I protein was decreased at 60 min post-fertilization, with little or no PAK I protein or activity detectable at 80 min post-fertilization or in 2-cell embryos. At the 4-cell stage PAK I protein increased, but the protein kinase was present primarily as an inactive form. Rac2 and Cdc42, but not Rac 1, were identified in oocytes and throughout early embryo development. Thus, PAK I appears to be a potent cytostatic protein kinase involved in maintaining cells in a non-dividing state. PAK I activity is high in oocytes and appears to be regulated by degradation/synthesis and through autophosphorylation via binding of Cdc42. PAK I may act through regulation of the stress-activated protein kinase signaling pathway and/or by direct regulation of multiple metabolic pathways.

Significant progress has been made in understanding the activation of components in signaling pathways that control cell growth and differentiation in vertebrates. There is a large body of research showing that progression through the cell cycle is highly regulated by the maturation promoting factor (cyclin/cdc2 kinase) cascade and the Ras/MAP kinase (MAPK) 1 cascade (for review see Refs. [1][2][3][4][5]. Biochemical analysis of amphibian oocyte maturation has provided evidence for the involvement of a protein kinase cascade in natural meiotic metaphase arrest in preparation for fertilization. The mature oocyte is arrested in metaphase II by the action of a cytostatic factor. Mos, a serine/threonine protein kinase, has been identified as a major component of cytostatic factor. Mos is synthesized during oocyte maturation and is degraded shortly after fertilization (6 -8). Injection of c-mos cDNA into one blastomere of 2-cell frog embryos leads to cleavage arrest at metaphase; injection of c-mos antisense RNA reverses the effect (6,9,10). Concomitant with Mos-induced cleavage arrest, MAPK kinase (MEK) is activated, whereas coinjection of a MEK neutralizing antibody with Mos prevents cleavage arrest (11). Mos has been shown to be an activator of MAPK via phosphorylation of MEK (12). Both MAPK and MEK are active in mature oocytes and inactivated upon fertilization (12)(13)(14). Injection of MAPK into cleaving embryos results in Mos-like metaphase arrest (15).
Another protein kinase cascade structurally related to the MAPK pathway is that of the stress-activated protein kinase (SAPK), also known as the Jun N-terminal kinase (JNK), pathway (16 -20). Activation of this pathway results in the phosphorylation and activation of c-Jun, a transcription factor involved in regulation of different phases of growth, differentiation, and a number of other physiological responses (19 -25). At this time, the role of the JNK pathway is not well understood, but it appears to be activated under conditions of cell stress, as shown with inhibitors of protein synthesis, inflammatory cytokines, heat shock, ultraviolet irradiation, and osmotic imbalance (20,22). MEKK has been reported to function preferentially in the JNK pathway, specifically regulating a SAPK/JNK-activating kinase (SEK) activity (22,23). It is postulated that the small G proteins Rac and Cdc42 are linked to the MEKK-SEK-JNK signaling cascade through a p21activated protein kinase (PAK) (24,25).
The PAK enzymes have an N-terminal regulatory domain containing a highly conserved binding site for Cdc42 and Rac, with a highly homologous catalytic domain at the C terminus, as first described by Manser et al. (26). Three distinct members of the PAK family have been identified by cDNA sequence analysis. The ␣ form is ϳ60 kDa and is found primarily in brain and spleen (26). The ␤ form is found in brain and testis and has a molecular mass of ϳ60 kDa (27). The ␥ form of ϳ58 kDa appears to be universal, and the cDNA has been sequenced from human (28), rat (29), and rabbit (30). The latter, identified as PAK I, was isolated initially as an inactive holoenzyme that could be activated by limited proteolysis (31,32). Cleavage generates a peptide of 37 kDa (p37) that contains the active catalytic domain plus a small part of the regulatory domain (30). 2 Binding of GST-Cdc42 to PAK I, but not GST-Rac1 or RhoA, results in autophosphorylation of PAK I, which stimulates the protein kinase activity (30). Inactive PAK I has been purified to apparent homogeneity from rabbit reticulocytes, and two endogenously active forms of the enzyme have also been highly purified using the same procedure. 2 A number of substrates have been identified for PAK I, including histones H2B and H4 (32,33), initiation factors 3, 4B, and 4F p220 subunit (31,36), myosin light chain from smooth muscle (35), and Rous sarcoma virus protein NC (37,38).
PAK I has been shown to have enhanced activity in quiescent and serum-starved cells, as compared with actively dividing cells, 3 and is present in a number of species and tissue types (31)(32)(33)(34)(35). These observations prompted further inquiry into the role of PAK I in cytostasis using early amphibian embryos as a model system. In this study, PAK I activity and protein were analyzed in frog eggs and early embryos and compared with the levels of small G proteins at different stages of egg development. In mature oocytes, PAK I activity was high; following fertilization, PAK I activity and protein were greatly reduced prior to cleavage. To examine whether PAK I had cytostatic properties, femtomole amounts of active PAK I (p58) or the active catalytic domain (p37) were injected into one blastomere of 2-cell embryos; cleavage was arrested in the injected blastomere, whereas the other continued cleaving normally. No effect on cleavage was observed with inactive PAK I. The data suggest that PAK I is highly regulated during early embryo development by synthesis/degradation and by cycling between active and inactive forms of the enzyme and that PAK I is a potent inhibitor of cell cleavage/division.

EXPERIMENTAL PROCEDURES
Materials-Trypsin (diphenylcarbamyl chloride-treated), soybean trypsin inhibitor, and mixed histone IIAS were obtained from Sigma. Leupeptin, pepstatin, aprotinin, and histone H4 were from Boehringer Mannheim. Chemiluminescent detection reagent, protein kinase assay kits, and [␥-32 P]ATP were obtained from Amersham Corp. Horseradish peroxidase-conjugated goat anti-rat IgG was from Organan Teknika. The GST-fusion proteins CDC42Hs, Rac1, and RhoA were generously provided by Dr. Channing Der, University of North Carolina, Chapel Hill, NC. Antibodies to Rac1, Rac2, and Cdc42 were from Santa Cruz Biochemicals. Superose 12 HR 10/30 and Mono Q HR 5/5 FPLC columns were from Pharmacia Biotech Inc. Freon (1,1,2-trichlorofluoroethane) was from Aldrich. Casein kinase II from rabbit reticulocytes was purified by chromatography on DEAE-cellulose and phosphocellulose as described (39). Protein kinase C was purified from bovine brain as described previously (40). The catalytic subunit of cyclic AMP-dependent protein kinase was generously provided by Dr. William H. Fletcher, J. L. Pettis Memorial Veterans Center, Loma Linda, CA.
Injection of Frog Embryos-Eggs from Lepidobatrachus laevis were naturally fertilized or fertilized in vitro (41). The first cleavage was observed approximately 90 min following fertilization, and cleavage continued at approximately 20-min intervals through mid-cleavage. The embryos were selected at the beginning of the first furrow and injected 10 min thereafter. The protein kinase was diluted approximately 50-fold in buffer B (10 mM MOPS, pH 7.4, 1 mM dithiothreitol), and 50 nl (0.01-1 pg, 0.3-30 fmol) was injected into one blastomere. Cleavage arrest of the injected blastomere was monitored visually under a stereomicroscope for up to 2.5 h following injection, as indicated.
Fluorescent Staining of DNA-Both blastomeres of 2-cell embryos were injected with active PAK I (1 pg/blastomere) and allowed to incubate for 30 min. Uninjected embryos developed to the 16-cell stage were used as controls. The embryos were lysed in the presence of 10 l of Hoechst fluorescent dye 50 g/ml (42), and the DNA was examined under a fluorescent microscope (50 ϫ magnification), photographed, and compared with that of Newport and Kirschner (42).
Western Blotting-Individual eggs and embryos were extracted in a volume of 250 l with freon to remove the yolk proteins (43) and frozen immediately. The extracts (20 -25 g) and the DEAE-cellulose column fractions (40 -60 l) were subjected to electrophoresis on 12.5% polyacrylamide slab gels in sodium dodecyl sulfate and transferred by electroblotting to nitrocellulose membranes. The samples were probed with antibody prepared to Rac1, Rac2, and Cdc42 diluted 1/200, and with rat polyclonal anti-PAK I antibody diluted 1/1000, and analyzed using peroxidase-conjugated goat and rabbit anti-rat secondary antibody by chemiluminescent detection. Polyclonal antibodies were raised in rat to the highly purified form of inactive form of PAK I from rabbit reticulocytes. PAK I (50 g) was subjected to electrophoresis in a 12.5% polyacrylamide gel containing sodium dodecyl sulfate; the protein band was excised from the gel, mashed with a spatula, mixed with Freund's complete adjuvant (0.5 ml), and injected subcutaneously. The rats were bled from the tail vein on days 14 and 21. Subsequent boosting of the rat was carried out in a similar manner with 10 -30 g of PAK I in Freund's incomplete adjuvant. Antibody titer was quantified by Western blot analysis with purified PAK I from rabbit; the antibody was specific for PAK I and did not react with any other proteins in preparations from mammalian or frog tissues.
Chromatography of Frog Extracts-Oocytes, zygotes, and early embryos, collected at specific stages of development, were frozen in batches of 30 in liquid nitrogen and stored at Ϫ70°C. Prior to chromatography, the eggs were suspended and homogenized in 5.0 ml of buffer A (20 mM ␤-glycerophosphate, 1 mM dithiothreitol, 1 mM EDTA, 1 mM EGTA, 1 mM sodium vanadate, 1 mM sodium pyrophosphate, 0.01 mM cAMP, 0.5 mM phenylmethylsulfonyl chloride) containing 40 g/ml of leupeptin, pepstatin, and aprotinin. The homogenate was centrifuged for 10 min at 12,000 ϫ g in a Sorvall SS 34 rotor, and the supernatant was collected and centrifuged for 1 h at 47,000 rpm in a Beckman Ty65 rotor. The amount of protein was measured by Bradford analysis, and the supernatant was applied to a 0.5-ml DEAE-cellulose column (5.0 ϫ 0.8 cm) equilibrated with buffer A and, the resin was washed with 5 ml of buffer A. Protein was eluted with a 5-ml linear gradient (0 -300 mM KCl) in buffer A; fractions of 0.33 ml were collected.
Assay for PAK I-PAK I was assayed before and after limited tryptic digestion using H4 or histone IIAS as substrate. Samples from column fractions (20 l) were preincubated for 30 s in 10 mM Tris-HCl, pH 8.0, and 5 mM 2-mercaptoethanol in the presence or absence of 4 g/ml trypsin in a volume of 40 l. Proteolysis was terminated by the addition of 4 l of a 10-fold molar excess of soybean trypsin inhibitor. Protein kinase activity was assayed with H4 (2.0 g) in phosphorylation buffer containing 50 mM Tris-HCl, pH 7.4, l0 mM MgCl 2 , 30 mM 2-mercaptoethanol, 0.20 mM [␥-32 P]ATP (specific activity 1000 cpm/pmol), and PAK I in a final volume of 70 l. Leupeptin, pepstatin, and aprotinin were added at 8 g/ml to the samples without trypsin prior to incubation and to the samples containing trypsin following incubation. Incubation was for 15 min at 30°C; the reactions were terminated by the addition of 10 l of 10 mM nonlabeled ATP. Phosphorylation of H4 was analyzed by electrophoresis on 15% polyacrylamide slab gels in sodium dodecyl sulfate, followed by autoradiography (44). Radioactivity was quantified by excising the histone band and counting in a liquid scintillation counter. Under these conditions, PAK I was limiting and incorporation was linear with time.
To assay for activation of PAK I by CDC42, aliquots (30 l) of column fractions were incubated with the G protein preloaded with GTP␥S or GDP in 70 l of phosphorylation buffer containing 1.4 g of soybean trypsin inhibitor at 30°C for 10 min, as described by Jakobi et al. (30), resulting in autophosphorylation of PAK I. An aliquot (30 l) of the autophosphorylated enzyme was removed and assayed with H4 as described above.
Activity assays for MAPK, cdc2 kinase, and casein kinase II were carried out in 30-l reaction mixtures with 5-10 l of enzyme using synthetic peptides highly selective for MAPK and specific for cdc2 kinase and casein kinase II as described by Amersham Corp.
Purification of Active and Inactive PAK I-The inactive PAK I holoenzyme (p60) was purified to apparent homogeneity from rabbit reticulocytes by chromatography on DEAE-cellulose, SP-Sepharose fast flow, protamine-agarose, and FPLC on Mono S and Mono Q. 2 The active PAK I holoenzyme (p58) was purified using similar procedures, but had slightly different chromatographic properties. To prepare large amounts of the catalytic domain (p37), inactive PAK I purified through the protamine-agarose step was subjected to limited tryptic digestion for 30 s, followed by addition of a 10-fold excess of soybean trypsin inhibitor (33). The activated enzyme was dialyzed against buffer A (20 mM Tris-HCl, pH 7.8, 3 mM dithiothreitol) containing 0.1 mM phenylmethylsulfonyl fluoride, and applied to a Mono-Q HR 5/5 column equilibrated in buffer A. The column was washed with 10 ml of buffer A, and the enzyme was eluted with a 12.5-ml gradient of 0 -0.45 M NaCl in buffer A. The peak fraction was purified to apparent homogeneity on Superose 12 HC 10/30 in buffer A. Alternatively, highly purified PAK I was cleaved, and the p37 peptide was autophosphorylated and passed over a pad of DEAE-cellulose to remove trypsin and trypsin inhibitor. Protein kinase activity was quantified as described earlier.

Analysis of the Cytostatic Activity of PAK I-The inactive
holoenzyme of PAK I was purified to apparent homogeneity from rabbit reticulocytes by ion-exchange chromatography and FPLC and migrated as a single band at 60 kDa upon polyacrylamide gel electrophoresis in sodium dodecyl sulfate (Fig. 1,  panel A). The endogenously active enzyme was purified in a similar manner and migrated at 58 kDa as shown by immunoblotting with antibody prepared to inactive PAK I. Silver staining of purified active PAK I showed a major band at 58 kDa with a single contaminating protein, which was not present in all fractions containing PAK I. A 37-kDa peptide containing the catalytic domain of PAK I was generated by limited proteolytic digestion of the inactive holoenzyme and purified to apparent homogeneity by FPLC on Mono Q and Superose 12 ( Fig. 1,  panel B). The specific activity of the endogenously active PAK I and the proteolytically activated enzyme were similar ϳ400 pmol/min/g.
The biological activity of PAK I was analyzed using embryos from L. laevis, since the eggs are easily injected and can be fertilized in vitro without destruction of the male, have very good embryo viability postinjection, and demonstrate rapid development (41). Extracts of mature eggs, zygotes, and 2-cell embryos were analyzed to determine whether an endogenous PAK I-like protein was present (Fig. 1, panel C). Antibody prepared to mammalian PAK I reacted specifically with amphibian egg protein of 58 -60 kDa, which was identical in molecular mass to PAK I in mammalian cells. The PAK I protein was present at significantly lower levels in the zygote and was greatly diminished at the 2-cell stage. When active PAK I (p58) was injected into one blastomere of a 2-cell embryo, high levels of the protein migrated at the same position as the endogenous PAK I protein, as detected with antibody to PAK I.
To examine the effects of PAK I on cell cleavage, active PAK I (p58) was injected into one blastomere of embryos at the 2-cell stage of development, and the embryos were allowed to develop for an additional 150 min to mid-blastula. As shown in Fig. 2, panels A and B, the injected cell was arrested, whereas the uninjected cell continued through normal division to late cleavage. Injection of the active catalytic domain (p37) also induced cleavage arrest as shown after 60 min of development to the 16 -32-cell stage (Fig. 2, panel C). No cleavage arrest was observed in embryos injected with heat-denatured PAK I (Fig.  2, panel D).
To determine the stage of arrest of the cell cycle, active enzyme was injected into both cells of a 2-cell embryo and incubation was continued for 30 min; at this point, cell cleavage was arrested in both blastomeres. Fluorescent staining of the DNA (Fig. 2, panel E) indicated the lack of a nuclear envelope and the presence of condensed chromosomes, which would be  (Table I).
Cleavage arrest was induced in 99% of the embryos injected with PAK I. When the same amount of PAK I was subjected to heat treatment (75°C for 15 min) prior to injection, only 10% of the embryos exhibited cleavage arrest. All of the embryos injected with buffer alone proceeded through normal development. The normal division time was approximately 20 min, and arrest was observed within 30 min following injection of PAK I; cleavage of the injected blastomere did not progress beyond the 2-to 4-cell stage, while the other blastomere continued dividing up to and beyond late cleavage.
Embryos injected at later stages of development (4-and 8-cell stages) exhibited cleavage arrest similar to that observed in the 2-cell embryos (data not shown). In embryos injected with lower amounts of PAK I (0.01 pg), arrest observed at the early time periods (Յ1 h) was sometimes reversed at longer incubation times, leading to apparent normal development (data not shown). At higher levels of active PAK I used in the experiments described herein (Ն0.1 pg), reversal of arrest did not occur. This result suggests that the signal needs to be of sufficient strength and duration to negate the growth response.
When the purified endogenously active PAK I was injected into 2-cell embryos, results identical to those obtained with p37 were obtained (Table II). The active enzyme effected cleavage arrest at even lower concentrations than with the active p37 peptide. Injection of inactive PAK I at levels equivalent to those of active PAK I showed no effect on development. At 100-fold higher concentrations (20 pg), the inactive holoenzyme resulted in cleavage arrest in 50% of the embryos injected, indicating a very low level of protein kinase activity. This effect, observed at 30 and 60 min, was reversed at longer times of incubation (data not shown). These data indicate that little or no activation of inactive PAK I occurred in 2-cell embryos, and active PAK I remained active for a period of time sufficient to initiate cytostasis. Thus, the components required for activation and inactivation of PAK I were not highly active and/or present in dividing cells.
Experiments were also conducted to examine the effects of injection of three other protein kinases into 2-cell embryos (Table II). Cleavage arrest was not observed following injection of equimolar amounts of casein kinase II, the catalytic subunit of the cAMP-dependent protein kinase, or protein kinase C, at concentrations identical to those used for active PAK I. Taken together, the data indicate that cleavage arrest was specific for active PAK I, and femtomole quantities produced a definitive cytostatic response.
PAK I, Rac2, and Cdc42 Proteins in Oocytes and Developing Embryos-Extracts from frog oocytes, zygotes, and developing embryos at the 2-, 4-, 8-, and 16 -32-cell stage were analyzed by immunoblotting with antibody to PAK I. The antibody was specific for PAK I and did not react with any other protein in the extract. PAK I protein (58 -60 kDa) was present in the oocyte and the early zygote but was significantly reduced at 60 min post-fertilization and almost nondetectable at 80 min postfertilization and at the 2-cell stage (Fig. 3). At the 4-cell stage, PAK I protein was again present and increased through the 16 -32-cell stage, but the level was significantly less than in the oocyte. To analyze for potential activators of PAK I, the same extracts were examined by immunoblotting with antibody prepared to Rac1, Rac2, and Cdc42. Rac2 and Cdc42 were present in the oocyte and throughout early embryogenesis. Rac1 was not detectable in the same experiments.
Active and Inactive Forms of PAK I in Oocytes-As it is not possible to accurately measure PAK I activity in cell extracts, PAK I was analyzed following chromatography on DEAE-cellulose. Fractions from mature oocytes were assayed for active kinases on arrest of cleavage Each sample (50 nl) was injected into one blastomere of a 2-cell embryo at the concentration indicated and analyzed at 1-h postinjection. Three highly purified protein kinases, casein kinase II from rabbit reticulocytes, catalytic subunit of cAMP-dependent protein kinase from bovine heart, and protein kinase C from bovine brain, were similarly injected. These data were compiled from at least two experiments.  PAK I under kinetically valid conditions, and phosphorylation of H4 was quantified by SDS-PAGE followed by autoradiography (Fig. 4, left panel). H4 is specific for PAK I and is not phosphorylated by other major protein kinases. Two peaks of active PAK I (identified by arrows 2 and 3) were observed to elute around 0.09 M KCl (fractions 5 and 6) and 0.16 M KCl (fractions [7][8][9]. Fractions eluting at salt concentrations above 0.2 M contained protein kinase activity that phosphorylated a major frog oocyte protein of 32 kDa and other endogenous proteins to a lesser extent (data not shown). Aliquots of the DEAE-cellulose fractions were subjected to Western blot analysis using antibody prepared against purified PAK I from rabbit (Fig. 4, right panel). The antibody reacted specifically with proteins of 58 -60 kDa in the fractions containing H4 kinase activity. The antibody also reacted with a protein doublet of the same molecular mass in fraction 4 (0.05 M KCl) that had little active enzyme but eluted at the position of inactive PAK I from other species and tissues. The initial peak of active PAK I contained a doublet of 58 -60 kDa, whereas the later peak contained a single protein of 59 kDa.
To identify the inactive form of PAK I, aliquots of the DEAEfractions were subjected to limited tryptic digestion prior to assay for PAK I or preincubated with CDC42(GTP) and GDP (Fig. 4, left panel). Three peaks of activity were identified, as indicated by the arrows. The early peak (eluting around 0.05 M KCl) contained inactive PAK I, which was observed only after limited proteolytic digestion or following activation with CDC42(GTP). The degree of activation was similar (6 -7-fold) under both conditions. Peak 2 (0.09 M KCl) contained two different forms of PAK I, one requiring activation and a second active enzyme (the activity of each form is shown). Most if not all of the enzyme in peak 3 was active. The total amount of active enzyme in peak 3 was 3-fold higher than that observed in peak 2. The data were consistent with the Western blot profile in which protein migrating between 58 and 60 kDa coincided with all three peaks of PAK I activity (Fig. 4, left panel).
Active and Inactive PAK I in Zygotes and Early Embyros-To examine PAK I activity following fertilization and during early cleavage, postribosomal supernatants from zygotes and embryos at the 4-and 16 -32-cell stages were chromatographed on DEAE-cellulose and assayed for PAK I activity. Following fertilization, PAK I activity was greatly reduced; little active or inactive enzyme was detected (Fig. 4, left panel). These data coincided with the diminished level of PAK I protein at this stage (Fig. 3). By the 4-cell stage, a small amount of active enzyme was observed in peak 2, with some inactive PAK I in peaks 2 and 3 (Fig. 4, left panel). At the 16 -32-cell stage, active PAK I chromatographed primarily with peak 2, whereas three peaks of inactive PAK I were observed. Thus, two chromatographically distinct forms of active PAK I and three forms of inactive PAK I were identified; the presence of these forms could be correlated with the stage of development and were consistent with the presence of PAK I protein that migrated at 58 -60 kDa.
Active and inactive PAK I were quantified at different stages of development (Fig. 5). PAK I activity was maximal in the oocyte, with 66% of the enzyme in an active state. The level of total enzyme activity decreased dramatically upon fertilization. The zygote contained only minor amounts of active and inactive PAK I as compared with the mature oocyte. An increased level of inactive PAK I was observed at the 4-and 16 -32-cell stages. 4-cell embryos contained 31% and 16 -32-cell embryos contained 41% of the total PAK I activity observed in mature oocytes, with approximately 75% of the protein kinase present as inactive enzyme.

Analysis of Other Protein Kinases in Oocytes and Developing
Embryos-The activities of other protein kinases in mature oocytes were also examined to compare the regulation of PAK I with that of other protein kinases. Levels of the protein kinase activities were analyzed with peptides specific for each enzyme and compared with the activities in eggs (Table III). MAPK activity decreased to almost undetectable levels following fertilization (zygote) and remained inactive in the 4-cell and 16 -32-cell embryos. Cdc2 activity was similarly undetected in zygotes, although low levels were present in the 4-cell (8%) and 16 -32-cell embryos (25%). In contrast to MAPK and cdc2, casein kinase II activity was present at about the same level at all stages. DISCUSSION PAK I belongs to the PAK family of protein kinases that are activated by the small G proteins Rac and Cdc42 but not by Ras and RhoA (26 -30) In these studies, PAK I has been shown to be a potent inhibitor of cleavage in early amphibian embryos. Injection of subfemtomole amounts of active PAK I into one blastomere of a 2-cell embryo arrested cleavage in the injected cell at metaphase, whereas the other blastomere continued to divide. The levels of PAK I needed to induce arrest were significantly lower than those used with MAPK and similar to those of Mos (15). Neither the inactive form of PAK I nor heat-treated PAK I were inhibitory, suggesting that arrest is due to phosphorylation of proteins that alter the cell cycle. Thus, PAK I appears to play an important role in maintaining cells in a non-dividing state. This state could be rapidly released or reversed by changes in PAK I activity and/or by degradation.
PAK I protein is highly conserved between species, since antibody prepared to rabbit PAK I reacts specifically with a protein of the same molecular weight in mature frog oocytes. High levels of PAK I activity and protein were identified in mature oocytes, consistent with the cytostatic activity of active PAK I observed upon injection into dividing embryos. Approximately 1 h following fertilization, total PAK I activity dropped to a low level and remained low through the 2-cell stage; the loss in PAK I activity appeared to be due to protein turnover, as shown by immunoblots of extracts from zygotes and 2-cell embryos. At the 4-cell stage, PAK I protein was again observed, but at lower levels than the oocyte; however, the inactive form was predominant in 4-and 16 -32-cell embryos. The identification of PAK I as a Cdc42-activated protein kinase (30), the identification of Cdc42 and Rac2 in the oocyte and throughout early development, and the activation of oocyte PAK I by CDC42 suggest that PAK I activity is regulated by Cdc42 and possibly Rac2 in vivo as well as in vitro. The identification of relatively consistent levels of Cdc42 and Rac2 in oocytes and early embryos is of interest. Rac2 has been observed previously primarily in cells of myeloid origin, and levels are highest in differentiated cells (45).
The complex chromatographic elution profile for PAK I dur-ing the early embryonic development, consisting of three individual peaks of PAK I, may be due at least in part to differential phosphorylation and to association with other proteins. PAK I is autophosphorylated following activation by proteolysis 2 or association with CDC42(GTP␥S) (30). The molecular mass of PAK I varies from 58 to 60 kDa, depending on the state of the enzyme. Thus, regulation of PAK I in germ cells appears to be a complex mechanism involving degradation/synthesis, G proteins, and phosphorylation. Multiple forms of PAK I, similar to those described herein, have also been identified in 3T3-L1 cells and rabbit reticulocytes. Three forms of PAK I (58 -60 kDa) have been highly purified from rabbit reticulocytes. 2 The three peaks of activity of PAK I activity observed upon chromatography of frog egg extracts on DEAE-cellulose is consistent with these three forms. With 3T3-L1 cells, the elution patterns of PAK I on DEAE-cellulose are dependent on the growth state of the cells. Chromatography profiles of PAK I from quiescent 3T3-L1 cells are similar to those observed with mature oocytes, and PAK I profiles from dividing cells are similar to those observed in the 16 -32-cell stage. 3 MAPK and cdc2 kinase activities were high in mature eggs and greatly diminished after fertilization. In the 4-cell and 16 -32-cell embryos, MAPK was not detectable, as shown previously in Xenopus (9,11,15,46). Low levels of cdc2 kinase  a PAK I activity was taken from Fig 5. The activity of the inactive form of PAK I was measured following limited proteolysis. activity were observed in the 4-cell and 16 -32-cell embyros. Casein kinase II remained at approximately the same level in all of the embryonic stages examined, suggesting that casein kinase II is not involved in regulation of cleavage arrest. This result is consistent with the lack of effect observed upon injection of casein kinase II into one blastomere of 2-cell frog embryos.
Active PAK I injected into early embryos at the 2-cell stage remains active for a period sufficient to arrest cleavage. Proteases activated upon fertilization that could degrade PAK I would be shut off relatively quickly to preserve the integrity of the egg and would not be expected to be present at the 2-cell stage. Inactive PAK I injected at the 2-cell stage is not activated either by G proteins or proteolysis. This is consistent with the fact that PAK I is primarily in an inactive form in the 4-cell and 16 -32-cell stages.
It thus appears that G protein activity could be limiting in the early embryos. The G proteins express an intrinsic GTPase activity and cycle between active enzyme bound to GTP and inactive GDP-bound forms (47). Three different classes of proteins modulate the cycling process; these include the GTPaseactivating proteins that enhance GTP hydrolysis resulting in down-regulation, guanine nucleotide exchange factors that exchange GDP for GTP, and guanine nucleotide dissociation stimulators (48). Any or all of these proteins could be regulated during early development resulting in changes in Cdc42 activity. Taken together, these data suggest that PAK I regulation in the oocyte and early embryonic development is through G protein stimulation of autophosphorylation (30) and by synthesis/degradation. Ras, activated in response to growth-promoting compounds, stimulates the MAPK cascade. Under these same conditions, the JNK pathway does not become activated. In contrast, the JNK pathway is activated under conditions of stress, including toxic inhibition of protein synthesis, inflammatory cytokines, heat shock, ultraviolet irradiation, and osmotic imbalance (20,22). This pathway is mediated by the G proteins Rac1 and Cdc42, and guanine nucleotide exchange factor specific for these proteins can stimulate JNK activity. Data by Coso et al. (24) and Minden et al. (25) suggest that Rac1 and Cdc42 can independently mediate this stress-regulated pathway and postulate that a PAK or PAK-related protein kinase may be involved in the regulation. Our observations with frog eggs support that postulate with the following evidence. First, PAK I has been shown to be specifically activated by Cdc42. Second, Cdc42 is present in oocytes and during early embryo development. Third, PAK I has been shown to have cytostatic properties upon injection into early frog embyros. Fourth, the high levels of activity in mature oocytes and quiescent cells are consistent with this thesis. Fifth, PAK I appears to be a universal enzyme, present in all higher animal species and tissues. Thus, it appears that PAK I may be the proposed PAK enzyme involved in dampening or shutting down cell metabolism in response to physiological stress, as well as maintaining cells (and organisms) in a non-dividing state.
The biological role of PAK I in cytostasis is under investigation at the molecular level; the current hypothesis for PAK regulation of mammalian cell growth is via phosphorylation of c-Jun through regulation of the JNK pathway (24,25). Studies are currently underway to examine the possible role of PAK I in regulation of the JNK and Ras/MAPK pathway. PAK I has also been shown to phosphorylate a number of substrates involved in macromolecular synthesis, including key translational initiation factors (31,36), H2B and H4 (32,33), and myosin light chain (35), suggesting multiple pathways may be coordinately controlled by Cdc42 via activation of PAK I. Thus, it is likely that phosphorylation of a number of key regulatory enzymes/factors/ proteins in the oocyte are required to maintain cytostasis.