Essential role for the Pak4 protein kinase in extraembryonic tissue development and vessel formation

MECHANISMS OF DEVELOPMENT

1 2 6 ( 2 0 0 9 ) 7 1 0 –7 2 0

available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/modo

Essential role for the Pak4 protein kinase in extraembryonic tissue development and vessel formation Yanmei Tiana, Liang Leia, Marta Cammaranob, Tanya Nekrasovaa, Audrey Mindena,* a

Susan Lehman Cullman Laboratory for Cancer Research, Department of Chemical Biology, Ernest Mario School of Pharmacy, Rutgers, The State University of New Jersey, 164 Frelinghuysen Road, Piscataway, NJ 08854, USA b Intercos, 200 Route 303 North, Congers, NY 10920, USA

A R T I C L E I N F O

A B S T R A C T

Article history:

Pak4 is a member of the group B family of Pak serine/threonine kinases, originally identified

Received 18 August 2008

as an effector protein for the Rho GTPase Cdc42. Pak4 knockout mice are embryonic lethal

Received in revised form

and do not survive past embryonic day 11.5. Previous work on Pak4 knockout mice has

11 May 2009

focused on studying the phenotype of the embryo. Abnormalities in the extraembryonic tis-

Accepted 13 May 2009

sue, however, are common causes of early embryonic death in knockout mice. Extraembry-

Available online 21 May 2009

onic tissue associated with the Pak4-null embryos was therefore examined. Abnormalities in both yolk sacs and placentas resulted when Pak4 was deleted. These included a lack of vas-

Keywords:

culature throughout the extraembryonic tissue, as well as an abnormally formed labyrin-

Pak4

thine layer of the placenta. Interestingly, epiblast-specific deletion of Pak4 using a

Placenta

conditional knockout system, did not rescue the embryonic lethality. In fact, it did not even

Yolk sac

rescue the extraembryonic tissue defects. Our results suggest that the extraembryonic tissue

Angiogenesis

abnormalities are secondary to defects that occur in response to epiblast abnormalities.

Development

More detailed analysis suggests that abnormalities in vasculature throughout the extraem-

Vascularization

bryonic tissue and the epiblast may contribute to the death of the Pak4-null embryos.  2009 Elsevier Ireland Ltd. All rights reserved.

1.

Introduction

Pak4 is a member of the group B family of Pak serine/threonine kinases (Abo et al., 1998; Jaffer and Chernoff, 2002). Other members of the group B family of Paks include Pak5 and Pak6. Pak5 and Pak6 have limited tissue specific expression patterns, with high expression in the brain (Li and Minden, 2003; Nekrasova et al., 2008). Pak4 in contrast, appears to be expressed ubiquitously in different tissues, and its expression begins early in development (Qu et al., 2003). The group A Paks consist of Paks 1, 2, and 3. In this case Pak2 is ubiquitously expressed, whereas Pak1 and Pak3 have more restricted tissue specific expression patterns. All of the Paks have a GTPase binding domain (GBD) through which they mediate an interaction with Rho GTPases, including Cdc42 and Rac.

Pak4 was originally identified as a protein which binds strongly to Cdc42, and less strongly to Rac. Pak4 was also shown to mediate Cdc42 induced filopodia formation (Abo et al., 1998). Like other Paks, however, Pak4 can also respond to Rho GTPase independent stimuli. In MDCK epithelial cells, for example, Pak4 is activated by Hepatocyte Growth Factor (HGF) via a signaling pathway that requires PI3 kinase. Pak4 is in turn thought to contribute to HGF induced changes in cytoskeletal organization and cell adhesion (Wells et al., 2002). Pak4 was also shown to interact with the cytoplasmic domain of the keratinocyte growth factor (KGF) receptor in a transformed kidney cell line, and may have a role in KGFR mediated cell survival pathways (Lu et al., 2003). Finally, Pak4 was shown to interact with integrin aVb5, and in the breast cancer cell line MCF7, it redistributes from the cytosol to lamellipodial structures when cells are attached to vitronectin (Zhang et al., 2002). These data

* Corresponding author. Tel.: +1 732 445 3400x253; fax: +1 732 445 0687. E-mail address: [email protected] (A. Minden). 0925-4773/$ - see front matter  2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.mod.2009.05.002

MECHANISMS OF DEVELOPMENT

suggest that multiple types of signaling proteins may converge upon Pak4 regulation. Another important function of Pak4 is its role in regulating cell survival. When overexpressed, Pak4 is associated with protection from apoptosis (Gnesutta and Minden, 2003; Gnesutta et al., 2001). Conversely, cells lacking Pak4 have an increased susceptibility towards apoptosis (Li and Minden, 2005). Pak4 promotes cell survival by different mechanisms, depending on the stimulus. In response to serum withdrawal, Pak4 protects cells via a kinase dependent mechanism that is associated with its ability to phosphorylate the pro-apoptotic protein Bad (Gnesutta et al., 2001). In response to stimuli that activate death domain containing receptors, however, such as TNF or Fas ligand, Pak4 functions via a kinase independent mechanism, which involves inhibition of caspase-8 recruitment and activity (Gnesutta and Minden, 2003). Pak4 also has a role in activating cell survival pathways, which lead to NFKB and ERK activation (Li and Minden, 2005). The developmental roles of the Paks are still incompletely understood. Our lab has been studying the developmental functions of the group B Paks, which include Pak4, Pak5, and Pak6. The developmental functions of these genes were studied by generating mouse knockouts. Pak5 and Pak6 knockout mice are viable and appear phenotypically normal (Li and Minden, 2003; Nekrasova et al., 2008). Pak5/Pak6 double knockout mice, however, have abnormalities in learning and locomotion (Nekrasova et al., 2008). In sharp contrast to Pak5 and Pak6, Pak4 knockout mice are embryonic lethal, and die by embryonic day 11.5. Pak4-null embryos have a number of abnormalities, including defects in the nervous system, and abnormalities in the heart (Qu et al., 2003). The exact cause of death in the knockout embryos, however, is still not known. Previous work on the Pak4 knockout mice has focused on studying the phenotype of the embryo (Qu et al., 2003). Abnormalities in the extraembryonic tissue, however, are common causes of death in mouse knockouts that die at this early embryonic stage (Copp, 1995; Rossant and Cross, 2001). In this study we therefore examined the extraembryonic tissue associated with the Pak4-null embryos. We found abnormalities in both yolk sacs and placentas when Pak4 is deleted. These included a lack of vasculature throughout the extraembryonic tissue, as well as an abnormally formed labyrinthine layer of the placenta. Interestingly, epiblast-specific deletion of Pak4 did not rescue the embryonic lethality, and it did not even rescue the extraembryonic tissue defects. Our results suggest that the extraembryonic tissue abnormalities are secondary defects that occur in response to abnormalities in the epiblast. More detailed analysis suggests that abnormalities in vasculature throughout the extraembryonic tissue and the epiblast may contribute to the death of the Pak4-null embryos.

2.

Results

2.1. Deletion of Pak4 leads to growth retardation of the embryo and abnormalities in the extraembryonic tissues Our previous study showed that Pak4-null embryos do not survive past E11.5 of gestation (Qu et al., 2003). Here we ana-

1 2 6 ( 2 0 0 9 ) 7 1 0 –7 2 0

711

lyzed E8.5–E10.5 embryos from intercrosses of Pak4+/ mice. Wild-type and Pak4-null embryos showed identical development up to E9.5. Beginning at E9.5 (20 somites), some of the Pak4/ embryos looked slightly smaller and paler than the Pak4+/+ or Pak4+/ controls, while others still looked similar to the wild-type controls. Gradually, as the wild-type embryos grew rapidly, the differences between the Pak4/ and the wild-types became more significant. At E10.5, all Pak4null embryos showed severe growth retardation compared to their littermates. They looked markedly smaller and paler than the controls (Fig. 1A), though some of them still had beating hearts. The growth retardation was associated with an overall increase in apoptosis and a decrease in cell proliferation, especially in the nervous system. This could be seen as early as E9.5 (Fig. 1B) even though Pak4 knockout embryos were still quite similar to the wild-type embryos in appearance at that time point. Extraembryonic tissue associated with the Pak4-null embryos also appeared abnormal. In particular, there was a lack of vasculature in the yolk sac, and the placenta appeared to be pale and contained less blood (see Fig. 1C). Abnormalities in the extraembryonic tissue are common causes of growth retardation and death in embryos of this stage, and this is consistent with our finding that Pak4 is expressed in both yolk sac and placenta, including fetal and maternal parts (Fig. 1D). We therefore examined the abnormalities in these tissues in more depth.

2.2.

Pak4 deletion results in yolk sac defects

Morphological analysis of whole yolk sacs at E10.5 showed that wild-type yolk sacs had well-organized vascular networks and contained large branching vitelline vessels. In contrast, Pak4/ yolk sacs were pale and looked thicker and rougher than wild-type yolk sacs (Fig. 1B). Whole-mount staining with hematoxylin and anti-Cd31, an antibody specific for platelet endothelial cell adhesion molecule (PECAM), showed that Pak4/ yolk sacs contained no large vitelline vessel but only a honeycomb-like primitive vascular plexus (Fig. 2A and B). Histological analysis of tissue sections revealed that wild-type E10.5 yolk sacs have generally small well-circumscribed capillary vessels and some large collecting vessels filled with red blood cells, whereas the vessels of mutant yolk sacs were dilated and contained fewer blood cells (Fig. 2C). These data suggest that disruption of Pak4 did not affect early vasculogenesis but prevented later angiogenesis.

2.3.

Pak4 mutants have placental defects

No significant abnormality was found in Pak4/ placentas at E9.5 when chorioallantoic fusion has occurred and vascular penetration has just initiated. At E10.5 however, Pak4/ placentas appear paler in the fetal part compared to that of control placentas (Fig. 1B). Histological analysis of mutant placenta sections revealed the major defects to be in the labyrinth layer (Fig. 3). At E10.5, the labyrinth layer in the wild-type placenta was well-vascularized with embryonic blood vessels, which contain fetal nucleated red blood cells, and maternal blood vessels, which contain enucleated red blood cells. In contrast, the labyrinth layer in the mutant placenta was rather compact and

712

MECHANISMS OF DEVELOPMENT

1 2 6 ( 2 0 0 9 ) 7 1 0 –7 2 0

Fig. 1 – Morphology of wild-type (+/+) and Pak4 homozygous mutant (/) embryos. (A) The Pak4/ embryo is developmentally retarded and smaller than its Pak4+/+ littermate (E10.5). (B) Analysis of cell proliferation (a) and apoptosis (b) in E9.5 Pak4+/+ and Pak4/ embryos. (a) Immunostaining with Ki67 antibody (brown) shows high level of proliferation throughout the Pak4+/+ embryo, and the highest level in the neural tissue indicated by arrows; whereas dramatically decreased proliferation was observed in Pak4-null embryos at this stage. Lower panel shows higher magnification of the boxed areas in upper panel. (b) Apoptosis detected by cleaved caspase-3 level (brown) was increased in Pak4/ embryo, especially in the neural tissue indicated by arrowheads, compared to Pak4+/+ embryo. Lower panel shows higher magnification of the boxed areas in upper panel. (C) Extraembryonic tissues of Pak4+/+ and Pak4/ embryos at E10.5. Note that the vitelline vessels visible in the Pak4+/+ yolk sac are missing in the Pak4/ yolk sac and that the Pak4/ placenta contains less blood and looks paler than the Pak4+/+ placenta. (D) Pak4 expression in extraembryonic tissues. FP, fetal placenta; MP, maternal placenta; YS, yolk sac; WT, wild-type mouse embryonic fibroblast (MEF) control; KO, Pak4/ MEF control.

significantly reduced in thickness and contained few embryonic blood vessels. This may have resulted from the failure of further vascular invasion after chorioallantoic fusion (Fig. 3). In situ hybridization analysis using markers for giant cells and spongiotrophoblasts show a somewhat reduced signal in these layers of the mutant placenta, suggesting that the trophoblast cells may also be underdeveloped and thinner in the absence of Pak4 (see Fig. 4).

2.4. Epiblast-specific deletion of Pak4 does not rescue embryonic lethality or extraembryonic abnormalities Abnormalities in the extraembryonic tissue in the Pak4 knockouts were quite striking. Since extraembryonic tissue is needed to support growth of the embryo, we hypothesized that the abnormalities in the extraembryonic tissue may be the actual cause of death in the Pak4-null embryos. To assess this, we used a conditional knockout system in which mice containing

floxed alleles of Pak4 were crossed with mice expressing Sox2Cre or Mox2Cre, which expresses Cre primarily in the epiblast. This cross should result in Pak4 deletion in the epiblast but not throughout the extraembryonic tissue. While the Sox2Cre drives complete recombination in the epiblast, the Mox2Cre strain drives variably mosaic recombination (Hayashi et al., 2002; Tallquist and Soriano, 2000). Both Sox2Cre and Mox2Cre systems have been previously reported to circumvent placental defects caused by certain gene knockouts, thus rescuing the early embryonic lethality (Bissonauth et al., 2006; Galabova-Kovacs et al., 2006; Wu et al., 2003). A diagram of the Pak4 floxed allele and the breeding strategy is illustrated in Fig. 5A and described in Section 4. Briefly, female mice with two floxed alleles of Pak4 (Pak4f/f) were crossed with mice containing Sox2Cre, along with one wild-type allele of Pak4 and one Pak4 deletion allele (Sox2Cre;Pak4+/D). Surprisingly, no live Sox2Cre;Pak4D/f mice resulted in a total of >200 offspring, indicating that epiblast-specific deletion of Pak4 was not able to rescue

MECHANISMS OF DEVELOPMENT

1 2 6 ( 2 0 0 9 ) 7 1 0 –7 2 0

713

Fig. 2 – Vasculature in Pak4+/+ and Pak4/ yolk sacs at E10.5. Whole-mount staining with hematoxylin (A) and anti-Cd31 antibody (B) show poor vascular development in the Pak4/ yolk sac compared with the Pak4+/+ yolk sac. (C) Yolk sac sections stained with hematoxylin/eosin. Left panel: low magnification; right panel: high magnification. Note that the vessels in the Pak4/ yolk sac are abnormally dilated and contain fewer blood cells compared to +/+ yolk sac.

embryonic death. When E10.5 embryos were examined, Sox2Cre;Pak4D/f embryos had retarded growth similar to those seen with conventional Pak4/ embryos, while embryos with other genotypes looked normal (see Fig. 5B). Strikingly, even the abnormalities in the extraembryonic tissue were seen in the conditional knockout mice. The yolk sacs appeared abnormal, similar to the conventional knockout mice, as seen in Fig. 5C. Histological analysis of placentas from Sox2Cre;Pak4D/f mice revealed that the labyrinth layer defects also were not corrected by epiblast-specific Pak4 deletion (see Fig. 6). Identical results were obtained when Mox2Cre mice were used (data not shown).

2.5. Abnormalities in vasculature throughout the Pak4null embryos The results described above suggest that although Pak4 knockout embryos have extraembryonic tissue abnormalities,

these may not be the primary cause of death in the Pak4-null embryos. Instead, they may be secondary to other abnormalities, which originate in the epiblast. As described above, a decrease in vasculature was apparent throughout the extraembryonic tissue. To determine whether vasculature is also affected in the epiblast, wild-type and Pak4/ E9.5 embryos were stained with anti-Cd31 antibody. Staining was dramatically reduced throughout the entire embryo in the knockouts (Fig. 7A and B). At this stage, wild-type embryos have started to develop branched networks of small vessels, in addition to the major networks of large blood vessels. In the knockouts, the networks of small vessels were missing. These results indicate that Pak4 is required for normal branching of vessels throughout the epiblast and extraembryonic tissue. Western blot analysis shows that Pak4 is expressed in normal vascular endothelial cells (Fig. 7C), suggesting a cell autonomous requirement of Pak4 in the vascular development.

714

MECHANISMS OF DEVELOPMENT

1 2 6 ( 2 0 0 9 ) 7 1 0 –7 2 0

Fig. 3 – Hematoxylin/eosin stained radial sections of E10.5 Pak4+/+ and Pak4/ placentas. (Upper panel) Low magnification. Note that the labyrinth layer (La) of the Pak4/ placenta is severely reduced in size and disorganized compared to the welldeveloped Pak4+/+ placenta. The dashed lines mark the boundaries between the labyrinth and the spongiotrophoblast layer. Al, allantois; Ch, chorionic plate; Sp, spongiotrophoblast; Gi, giant cell zone. (Lower panel) Higher magnification of the boxed areas in (upper panel), to show the architecture of labyrinth layer. Note in the Pak4+/+ labyrinth layer, the extensive intermingling of the fetal (arrowheads, identified by dark-colored large nucleated RBCs) and the maternal (arrows, identified by small enucleated RBCs) blood vessels. In contrast, in Pak4/ placenta, fetal blood vessels fail to penetrate into the chorionic plate to form the labyrinth layer.

3.

Discussion

The placenta is required for fetal survival, and is the first organ to form during development. The placenta connects embryonic and maternal tissue, and serves as a site for exchange of material between the mother and the fetus. This includes delivery of nutrients and oxygen to the fetus, and transfer of waste products back to the mother. Since mammalian embryos cannot survive without this exchange of material, mutations that disrupt genes important for placenta development are typically essential for survival of the embryo (for reviews, see Kunath et al., 2004; Rossant and Cross, 2001). Formation of the placenta begins early in development when cells of the blastocysts diverge into two cell types. These are the cells of the inner cell mass (ICM), in the interior of the blastocysts, and the cells of the trophectoderm layer, on the outside of the blastocyst. The trophectoderm layer will ultimately invade the uterine wall, and become the placenta. Most cells of the ICM will form the epiblast, but some of them will also contribute to the placenta. At midgestation, the allantois, which is derived from the epiblast, fuses with the chorion, the outermost layer, penetrates deep into the pla-

centa, and makes branches to form fetal blood vessels in the placenta. These fetal vessels intermingle together with maternal blood vessels to form the labyrinth-structured layer. The labyrinth layer must be fully vascularized, with proper branching, in order for the embryo to receive its essential nutrient supply. It is this layer that is poorly developed in the Pak4 knockout mice, apparently due to the lack of branching vessels. The yolk sac, another important extraembryonic tissue, is a bilayer structure composed of extraembryonic mesoderm cells and visceral endoderm cells. The yolk sac is the first site of blood formation and is also critical for transporting important substances between the mother and the fetus (Downs and Bertler, 2000; Kunath et al., 2004; Rossant and Cross, 2001). The yolk sac also appears to be abnormal in the Pak4 knockouts. In the yolk sac a vascular plexus could be seen, indicating that the beginning of vessel formation has occurred, but there is almost a complete lack of vessel branching. This again appears to be due to a decrease in vessels and in vessel branching. Pak4 knockout mice were previously shown to die by embryonic day E11.5 (Qu et al., 2003), although the cause of death remained unknown. We show that there is a decrease in

MECHANISMS OF DEVELOPMENT

1 2 6 ( 2 0 0 9 ) 7 1 0 –7 2 0

715

Fig. 4 – Expression of trophoblast differentiation specific markers in the Pak4/ placenta at E10.5. (Upper panel) The giant cell marker mPL-1 is expressed in both Pak4+/+ and Pak4/ placentas, as assessed by in situ hybridization. (Lower panel) Expression of the spongiotrophoblast maker 4311 is present in both genotypes.

proliferation and an increase in apoptosis in the embryos, particularly in the nervous system. Nervous system abnormalities do not generally cause embryonic death, however, and we do not have evidence that the decreased cell proliferation causes embryonic death. It is interesting that many knockout mice that die at around the same stage as Pak4 have abnormalities in extraembryonic tissue, as well as embryonic vasculature (Cobb and Goldsmith, 2000; Rossant and Cross, 2001). In the placentas of Pak4 knockout embryos, the labyrinth layer was poorly developed, and there was a reduction in size of the spongiotrophoblast and giant cell layers. Throughout the placenta as well as the yolk sac, there was also a decreased number of embryonic vessels, suggesting that vessels were prevented from forming, or extending into, the extraembryonic tissue. These data strongly suggest a role for Pak4 in the development of the placenta, and thus help explain why it is essential for sustained growth and survival of the embryo. A number of other knockouts have abnormalities in the extraembryonic tissue. These include p38, Mek1, B-Raf, and ERK2. These differ from Pak4 knockouts, however, because they can be rescued by either epiblast-specific deletion (Bissonauth et al., 2006; Galabova-Kovacs et al., 2005), or by tetraploid aggregation (Adams et al., 2000; Bissonauth et al., 2006; Hatano et al., 2003), where the knockout is confined to the epiblast. This suggests that the placental defect is the primary cause of death in these knockouts. Surprisingly, epiblast-specific deletion of Pak4 using the Sox2Cre conditional knockout system did not rescue the early lethality of the Pak4 knockout mice. There are several possible explanations for this. One possibility is that although the lack of Pak4 leads to abnormalities in the placenta, Pak4 has other, epiblast-specific functions, that are also required for embryonic development. We have already shown that Pak4 is required for normal development of the heart, and heart defects can be another major cause of death. Interestingly, however, conditional deletion of Pak4 in the epiblast did not even rescue the abnormalities

found in the placenta. One possible explanation relates to Sox2 expression. Sox2Cre was used to delete Pak4 in an epiblast-specific manner, when crossed with mice containing floxed alleles of Pak4. However, while Sox2 is expressed primarily in the epiblast, there is also some expression of Sox2 in proliferating trophoblast cells of the extraembryonic tissue (Avilion et al., 2003; Kunath et al., 2004). Thus, even in the extraembryonic tissue, Pak4 may in fact be deleted to some degree, leading to placental abnormalities. However, Mox2Cre, another epiblast-specific Cre, also failed to rescue the placenta phenotype. Another possible explanation is that abnormalities in the placenta remain because some parts of the extraembryonic tissue are derived from epiblast tissue, particularly the extraembryonic mesoderm, which gives rise to parts of the extraembryonic tissue, including blood vessels, the allantois, and parts of the yolk sac, chorion, and amnion (Avilion et al., 2003; Downs and Bertler, 2000; Kunath et al., 2004; Lu et al., 2001; Rossant and Cross, 2001). Thus, even epiblast-specific deletion of Pak4 could potentially lead to abnormalities in the extraembryonic tissue. The origin of the Pak4 phenotype thus seems to be quite different from those of other knockouts that affect the placenta. While lethality in the Pak4 knockouts seems to be caused by a mesodermal abnormality, knockouts of genes such as p38, ERK2, B-raf, and Mek1 seem to have abnormalities that originate in the trophoblast. A major abnormality in the Pak4 knockout embryos was an abnormality in embryonic vasculature. The poorly developed labyrinth layer of the placenta did not show the typical network of vessels, and the yolk sac lacked branched vessels. Abnormalities in vessel development could result from either a failure of vessels to form altogether, or from a defect in angiogenesis, where vessels form but fail to invade the labyrinthine layer or to expand and form branches throughout the embryo. To address these possibilities, embryos were stained with anti-Cd31 antibody, which recognizes endothelial cells. In the knockout embryos large vessels could still

716

MECHANISMS OF DEVELOPMENT

1 2 6 ( 2 0 0 9 ) 7 1 0 –7 2 0

Fig. 5 – Epiblast-restricted ablation of Pak4 cannot rescue embryonic lethality. (A) Diagram of floxed Pak4 vector and targeting strategy for the generation of conditional Pak4 knockout mice. LoxP sites (L) are indicated by triangles. (B) Growth retardation in the Sox2Cre;Pak4D/f embryo at E10.5 compared to its normal littermate. (C) Vascular defect in the Sox2Cre;Pak4D/f yolk sac.

be detected, but smaller branching vessels were completely absent. The Pak4 knockout yolk sac did stain positively for endothelial cells, as indicated by the presence of the vascular plexus, but did not have organized vessels. This data suggests that Pak4 affects angiogenesis and branching, rather than the initial formation of vessels. Such an abnormality in vessel expansion may help explain the failure of the extraembryonic tissue to develop normally, especially the labyrinth layer, which typically has a complex network of vessels. The lack of angiogenesis is likely associated with a defect in migration of endothelial cells. It is interesting that Pak4 knockout embryos also have abnormalities in the heart (Qu et al., 2001), characterized by extremely thin myocardial walls. The heart consists of many different cell types, which include myocytes, endothelial cells, vascular smooth muscle cells, and fibroblasts that derive from the mesoderm during development (Kattman et al., 2007). Improper migration of endothelial cells in the developing heart could help explain why the hearts develop poorly in Pak4 knockout embryos. The abnormally developed heart, as well as the lack of vessels throughout the embryo and extraembryonic tissues is likely to account for the early death of the embryos. The mechanism by which Pak4 may regulate vessel formation and extraembryonic tissue architecture is unknown. Pak4

was identified as an effector protein for Cdc42, which regulates filopodia formation (Abo et al., 1998). Filopodia are important for maintenance of cell shape, and for cell migration. Since extensive cell migration is required for angiogenesis, a role for Pak4 in filopodia formation and migration could help account for its role in the formation of new blood vessels and branches. Pak4 has also been shown to be involved in regulating cell growth, and promoting cell survival, which are also important factors in the angiogenic process. Cdc42 and Rac, which can activate members of the Pak family, have been shown to have key roles in angiogenesis. In vitro, Cdc42 and Rac regulate vacuole and lumen formation, an important part of the angiogenic process (Bayless and Davis, 2002). Rac was also shown to mediate VEGF induced chemotaxis of endothelial cells, while both Cdc42 and Rac were activated by haptotaxis (Soga et al., 2001). The roles for these proteins in regulating the formation of polymerized actin structures and adhesions to the extracellular matrix may be important for controlling the types of cell migration required for angiogenesis. It would be interesting to determine whether Pak4 is an important mediator in these Rho GTPases activated signaling pathways that control angiogenesis. Although a role for Pak4 in angiogenesis has not been reported, Pak1 is thought to play an important role in the process (Kiosses et al.,

MECHANISMS OF DEVELOPMENT

1 2 6 ( 2 0 0 9 ) 7 1 0 –7 2 0

717

Fig. 6 – Epiblast-restricted deletion of Pak4 fails to redress the placental defect. The underdeveloped labyrinth layer persists in Sox2Cre;Pak4D/f placenta compared to normal placenta. The dashed lines mark the boundaries between the labyrinth and the spongiotrophoblast layer.

1999). This is proposed to be due to its role in regulating cell migration by controlling changes in the cytoskeleton and formation of adhesions. Migration of endothelial cells requires a number of steps. These include extension of forward protrusions followed by formation of new adhesions. This is followed by retraction of the back of the cell at the trailing edge. Pak1 is proposed to play an important role in the retraction of the tail (Kiosses et al., 1999). Since Pak4 is involved in the formation of filopodial extensions in many cell types (Abo et al., 1998), it would be interesting to determine whether Pak4 may have a role in the protrusion of the leading edge cells, while Pak1 affects cell retraction at the trailing edge. Pak4 is expressed in a wide array of different tissues and there may be many explanations for the failure of Pak4-null embryos to survive. Our results here suggest that one factor in the early death of the Pak4-null embryos may be the role for Pak4 in vessel formation. Our results suggest that Pak4 may play a key role in the expansion of vessels in the embryo. Although initial formation of vessels seems to occur normally in the knockouts, formation of branches throughout the embryo does not occur. Branched vessels are almost completely absent in the labyrinth layer of the placenta. The resulting lack of exchange of nutrients may be a factor in the early cause of death of Pak4 knockout embryos.

4.

Materials and methods

4.1.

Mice

Conventional Pak4 knockout mice (Pak4+/) (Qu et al., 2003), Pak4 floxed mice (Pak4+/f and Pak4f/f) (see below) and

Pak4 deletion mice (Pak4+/D) (see below) were backcrossed to C57BL/6J for at least five generations and maintained on a C57BL/6J background. EIIa-cre mice, Sox2Cre mice and Mox2Cre mice were obtained from Jackson Laboratory. For epiblast-restricted deletion, Pak4+/D mice were crossed to Sox2Cre mice (Hayashi et al., 2002) or Mox2Cre mice (Tallquist and Soriano, 2000) to generate the breeding male Sox2Cre;Pak4+/D or Mox2Cre;Pak4+/D. Such breeding males were then bred to Pak4f/f female mice. In timed mating, noon of the vaginal plug detection date was termed day 0.5 of gestation (E0.5).

4.2.

Generation of Pak4 floxed mice and Pak4 deletion mice

A targeting construct spanning all eight exons of Pak4 was used to generate Pak4 floxed mice. Briefly, the construct includes a LoxP site in intron 1 and a loxP-neo-loxP cassette in intron 4. After linearizing with NotI, the construct was electroporated into ES cells of 129 background. The ES clones were then screened by PCR and Southern blotting. One 5 0 and one 3 0 probe, both of which are outside of targeting construct, were used in Southern blotting. The correct targeting clones were then injected into C57BL/6J blastocysts to generate chimera founder lines. Germline transmission was observed from one chimera. The heterozygotes were then backcrossed to C57BL/6J. The establishment of the correct targeting line was again confirmed by PCR and Southern blotting. In order to excise the neo cassette, the correct targeting line described above was crossed to the deleter line EIIa-cre (Xu et al., 2001). PCR genotyping of the tail DNA was used to screen the F1 generation. EIIa-cre is expressed in oocytes.

718

MECHANISMS OF DEVELOPMENT

1 2 6 ( 2 0 0 9 ) 7 1 0 –7 2 0

Fig. 7 – Defective vasculature in Pak4-null embryo. (A and B) Whole-mount immunostaining of E9.5 Pak4+/+ and Pak4/ embryos with anti-Cd31 antibody. (A) Note that the Pak4/ embryo shows dramatically weaker staining than the Pak4+/+ embryo (upper panel). Middle and lower panels show the details in the head and the intersomitic regions (boxed areas in upper panel), respectively. Pak4+/+ embryo shows well-branched networks of small vessels (arrows) in both regions. These are not seen in Pak4/ embryo. (B) Sagittal sections of Cd31 whole-mount immunostained embryos showing the vasculature inside the embryos. Compared to Pak4+/+ embryos, strikingly reduced staining throughout the Pak4/ embryos were observed. Arrows indicate the staining of vessels. (C) The expression of Pak4 in normal human vascular endothelial cells were detected by Western blot analysis. WT, wild-type MEF control; KO, Pak4/ MEF control; A, aorta; PA, pulmonary artery; UA, umbilical artery; UV, umbilical vein.

Therefore, depending on which two of the three LoxP sites are excised, a mosaic of progenies would be obtained in the F1 generation, which includes Pak4 floxed (recombination between LoxP2 and LoxP3) and Pak4 deletion (recombination between LoxP1 and LoxP3) (see Fig. 5A). The candidate F1 mice were then backcrossed to C57BL/6J. Another round of PCR genotyping was done to confirm the establishment of Pak4+/f line and Pak4+/D line. Pak4f/f line was obtained from cross between Pak4+/f mice.

4.3.

Genotyping

Genomic DNA was isolated from tails, yolk sacs or embryos and genotyping was performed by PCR. The following primers were used for genotyping Pak4 alleles: primer 1, 5 0 CAGCAGCCATTCAGTCTT-3 0 ; primer 2, 5 0 -GGTGGATGTGGAATGTGT-3 0 ; primer 3, 5 0 -TGGTCACTGGGAGAGGATG-3 0 ; primer 4, 5 0 -ATTGTCACCCACACCAGGTACCATAG-3 0 ; primer 5, 5 0 -AAG CACACAAAGCCGCACACATCC-3 0 ; primer 6, 5 0 -ATATGTCAAA TAGGCACCTCAGC-3 0 . Primers 1, 2, and 3 are used for conventional Pak4 knockout mice and amplify a 200-bp fragment of Pak4 allele and a 307-bp fragment of Pak4+ allele. Primers 4 and 5 amplify a 469-bp fragment of Pak4+ allele and a 503bp fragment of Pak4f allele. Primers 5 and 6 amplify a 500-

bp fragment of Pak4D allele. Primers 5 0 -GGTTTCCCGCA and 5 0 -GCTAAGTGCCTTCTCTACACC-3 0 GAACCTGAAG-3 0 were used to detect a 455-bp fragment of Cre allele.

4.4.

Histology and immunohistochemistry

Placentas and yolk sacs for histology were fixed in 4% paraformaldehyde, embedded in paraffin, cut into 5–10 lm sections and stained with hematoxylin and eosin according to a standard protocol. For immunohistochemistry, embryos were frozen in OTC medium and cut into 10 lm sections. The cryosections were fixed in 4% paraformaldehyde and immunostained with Ki67 (1:50, DakoCytomation) or cleaved caspase-3 (1:400, Cell Signaling) antibodies using a standard avidin–biotin peroxidase complex method (Liu et al., 2008). Sections were lightly counter-stained with eosin to show the structure of embryos. The procedure for whole-mount immunohistochemistry with anti-Cd31 antibody was as described (Schlaeger et al., 1995). Briefly, embryos and yolk sacs were fixed in 4% paraformaldehyde, dehydrated in the series of methanol and bleached in 5% hydrogen peroxide in methanol. The bleached samples were rehydrated, blocked in 3% non-fat

MECHANISMS OF DEVELOPMENT

milk in PBS with 0.1% Triton X-100 and then incubated with anti-Cd31 antibody (BD Pharmingen). Horseradish peroxidase-conjugated secondary antibody (Sigma) was used and detected using DAB substrate (Vector Laboratories). To show the staining inside the embryos, some of the whole-mount stained embryos were frozen in OTC medium and cut into 20 lm sections.

4.5.

In situ hybridization

The riboprobes for 4311 and mPL-1 were kindly provided by Janet Rossant (Guillemot et al., 1994; Lescisin et al., 1988) and generated by using a DIG RNA labeling kit (Roche). In situ hybridization was performed as described previously (Li and Minden, 2003). Briefly, paraffin sections of placenta were dewaxed and treated with proteinase K. The hybridization was performed at 42 C in hybridization buffer (50% formamide, 10% dextran sulfate, 2· SSPE [1· SSPE is 0.15 M NaCl, 10 mM NaH2PO4, and 1 mM EDTA; pH 7.7], 2.5· Denhardt solution, 100 lg of herring sperm DNA/ml, 100 lg of yeast RNA/ml, 5 mM dithiothreitol, 40 U of RNase inhibitor/ml). The digoxigenin was detected according to the protocol from the DIG detection kit (Roche).

4.6.

Primary endothelial cell samples and Pak4 antibody

The lysates for endothelial cells from normal human aorta, pulmonary artery, umbilical artery, and umbilical vein were purchased from ScienCell Research Laboratories. Western blot analysis was performed using polyclonal Pak4 antibody from Cell Signaling.

Acknowledgments We thank The Iowa University Targeting Facility for assistance in generating the Pak4 floxed mice. We thank Janet Rossant for generous gift of riboprobes. This grant was supported by ACS grant (RSG-04-178-01-DDC) and NIH grant (2 R01 CA076342-06) to A.M.

R E F E R E N C E S

Abo, A., Qu, J., Cammarano, M.S., Dan, C., Fritsch, A., Baud, V., Belisle, B., Minden, A., 1998. PAK4, a novel effector for Cdc42Hs, is implicated in the reorganization of the actin cytoskeleton and in the formation of filopodia. EMBO J. 17, 6527–6540. Adams, R.H., Porras, A., Alonso, G., Jones, M., Vintersten, K., Panelli, S., Valladares, A., Perez, L., Klein, R., Nebreda, A.R., 2000. Essential role of p38alpha MAP kinase in placental but not embryonic cardiovascular development. Mol. Cell 6, 109– 116. Avilion, A.A., Nicolis, S.K., Pevny, L.H., Perez, L., Vivian, L., LovellBadge, R., 2003. Multipotent cell lineages in early mouse development depend on SOX2 function. Genes Dev. 17, 126–140. Bayless, K.J., Davis, G.E., 2002. The Cdc42 and Rac1 GTPases are required for capillary lumen formation in three-dimensional extracellular matrices. J. Cell Sci. 115, 1123–1136.

1 2 6 ( 2 0 0 9 ) 7 1 0 –7 2 0

719

Bissonauth, V., Roy, S., Gravel, M., Guillemette, S., Charron, J., 2006. Requirement for Map2k1 (Mek1) in extra-embryonic ectoderm during placentogenesis. Development 133, 3429– 3440. Cobb, M.H., Goldsmith, E.J., 2000. Dimerization in MAP-kinase signaling. Trends Biochem. Sci. 25, 7–9. Copp, A.J., 1995. Death before birth: clues from gene knockouts and mutations. Trends Genet. 11, 87–93. Downs, K.M., Bertler, C., 2000. Growth of the pre-fusion murine allantois. Anat. Embryol. 202, 323–331. Galabova-Kovacs, G., Kolbus, A., Matzen, D., Meissl, K., Piazzolla, D., Rubiolo, C., Steinitz, K., Baccarini, M., 2006. ERK and beyond: insights from B-Raf and Raf-1 conditional knockouts. Cell Cycle 5, 1514–1518. Galabova-Kovacs, G., Matzen, D., Piazzolla, D., Meissl, K., Plyuschch, T., Chen, A.P., Silva, A., Baccarini, M., 2005. Essential role of B-Raf in ERK during extraembryonic development. Proc. Natl. Acad. Sci. USA 103, 1324–1330. Gnesutta, N., Minden, A., 2003. Death receptor induced activation of initiator caspase-8 is antagonized by the serine/threonine kinase PAK4. Mol. Cell. Biol. 23, 7838–7848. Gnesutta, N., Qu, J., Minden, A., 2001. The serine/threonine kinase PAK4 prevents caspase activation and protects cells from apoptosis. J. Biol. Chem. 276, 14414–14419. Guillemot, F., Nagy, A., Auerbach, A., Rossant, J., Joyner, A.L., 1994. Essential role of Mash-2 in extraembryonic development. Nature 371, 333–336. Hatano, N., Mori, Y., Oh-hora, M., Kosugi, A., Fujikawa, T., Nakai, N., Niwa, H., Miyazaki, J., Hamaoka, T., Ogata, M., 2003. Essential role for ERK2 mitogen-activated protein kinase in placental development. Genes Cells 8, 847–856. Hayashi, S., Lewis, P., Pevny, L., McMahon, A.P., 2002. Efficient gene modulation in mouse epiblast using a Sox2Cre transgenic mouse strain. Mech. Dev. 119 (Suppl. 1), S97–S101. Jaffer, Z.M., Chernoff, J., 2002. P21-activated kinases: three more join the Pak. Int. J. Biochem. Cell Biol. 34, 713–717. Kattman, S.J., Adler, E.D., Keller, G.M., 2007. Specification of multipotential cardiovascular progenitor cells during embryonic stem cell differentiation and embryonic development. Trends Cardiovasc. Med. 17, 240–246. Kiosses, W.B., Daniels, R.H., Otey, C., Bokoch, G.M., Schwartz, M.A., 1999. A role for p21-activated kinase in endothelial cell migration. J. Cell Biol. 147, 831–844. Kunath, T., Strumpf, D., Rossant, J., 2004. Early trophoblast determination and stem cell maintenance in the mouse – a review. Placenta 25 (Suppl. A), S32–S38. Lescisin, K.R., Varmuza, S., Rossant, J., 1988. Isolation and characterization of a novel trophoblast-specific cDNA in the mouse. Genes Dev. 2, 1639–1646. Li, X., Minden, A., 2003. Targeted disruption of the gene for the PAK5 kinase in mice. Mol. Cell. Biol. 23, 7134–7142. Li, X., Minden, A., 2005. PAK4 functions in tumor necrosis factor (TNF) alpha-induced survival pathways by facilitating TRADD binding to the TNF receptor. J. Biol. Chem. 280, 41192–41200. Liu, Y., Xiao, H., Tian, Y., Nekrasova, T., Hao, X., Lee, H.J., Suh, N., Yang, C.S., Minden, A., 2008. The Pak4 protein kinase plays a key role in cell survival and tumorigenesis in athymic mice. Mol. Cancer Res. 6, 1215–1224. Lu, C.C., Brennan, J., Robertson, E.J., 2001. From fertilization to gastrulation: axis formation in the mouse embryo. Curr. Opin. Genet. Dev. 11, 384–392. Lu, Y., Pan, Z.Z., Devaux, Y., Ray, P., 2003. P21-activated protein kinase 4 (PAK4) interacts with the keratinocyte growth factor receptor and participates in keratinocyte growth factormediated inhibition of oxidant-induced cell death. J. Biol. Chem. 278, 10374–10380. Nekrasova, T., Jobes, M.L., Tingh, J.H., Wagner, G.C., Minden, A., 2008. Targeted disruption of the Pak5 and Pak6 genes in mice

720

MECHANISMS OF DEVELOPMENT

leads to deficits in learning and locomotion. Dev. Biol. 322, 95– 108. Qu, J., Cammarano, M.S., Shi, Q., Ha, K.C., de Lanerolle, P., Minden, A., 2001. Activated PAK4 regulates cell adhesion and anchorage-independent growth. Mol. Cell. Biol. 21, 3523–3533. Qu, J., Li, X., Novitch, B.G., Zheng, Y., Kohn, M., Xie, J.M., Kozinn, S., Bronson, R., Beg, A.A., Minden, A., 2003. PAK4 kinase is essential for embryonic viability and for proper neuronal development. Mol. Cell. Biol. 23, 7122–7133. Rossant, J., Cross, J.C., 2001. Placental development: lessons from mouse mutants. Nat. Rev. Genet. 2, 538–548. Schlaeger, T.M., Qin, Y., Fujiwara, Y., Magram, J., Sato, T.N., 1995. Vascular endothelial cell lineage-specific promoter in transgenic mice. Development 121, 1089–1098. Soga, N., Namba, N., McAllister, S., Cornelius, L., Teitebaum, S.L., Dowdy, S.F., Kawamura, J., Hruska, K.A., 2001. Rho family GTPases regulate VEGF-stimulated endothelial cell motility. Exp. Cell Res. 269, 73–87.

1 2 6 ( 2 0 0 9 ) 7 1 0 –7 2 0

Tallquist, M.D., Soriano, P., 2000. Epiblast-restricted Cre expression in MORE mice: a tool to distinguish embryonic vs. extra-embryonic gene function. Genesis 26, 113–115. Wells, C.M., Abo, A., Ridley, A.J., 2002. PAK4 is activated via PI3K in HGF-stimulated epithelial cells. J. Cell Sci. 115, 3947–3956. Wu, L., de Bruin, A., Saavedra, H.I., Starovic, M., Trimboli, A., Yang, Y., Opavska, J., Wilson, P., Thompson, J.C., Ostrowski, M.C., Rosol, T.J., Woollett, L.A., Weinstein, M., Cross, J.C., Robinson, M.L., Leone, G., 2003. Extra-embryonic function of Rb is essential for embryonic development and viability. Nature 421, 942–947. Xu, X., Li, Cuiling, Garrett-Beal, Lisa, Larson, Denise, WynshawBoris, Anthony, Deng, Chu-Xia, 2001. Direct removal in the mouse of a floxed neo gene from a three-loxp conditional knockout allele by two novel approaches. Genesis 30, 1–6. Zhang, H., Li, Z., Viklund, E.K., Stromblad, S., 2002. P21-activated kinase 4 interacts with integrin alpha v beta 5 and regulates alpha v beta 5-mediated cell migration. J. Cell Biol. 158, 1287– 1297.