The protein phosphatase 2A B56 regulatory subunit is required for ...

DEVELOPMENTAL DYNAMICS 243:778–790, 2014 DOI: 10.1002/DVDY.24111

PATTERNS & PHENOTYPES

The Protein Phosphatase 2A B56g Regulatory Subunit is Required for Heart Development a

Prajakta Varadkar,1 Daryl Despres,2 Matthew Kraman,3 Julie Lozier,1 Aditi Phadke,4 Kanneboyina Nagaraju,4 and Brent Mccright1* 1

Division of Cellular and Gene Therapies, Center for Biologics Evaluation and Research, FDA, Bethesda, Maryland Mouse Imaging Facility, NIH/NINDS, Bethesda, Maryland 3 Cambridge Research Institute, University of Cambridge, Cambridge, United Kingdom 4 The Children’s National Medical Center, Washington, DC

DEVELOPMENTAL DYNAMICS

2

Background: Protein Phosphatase 2A (PP2A) function is controlled by regulatory subunits that modulate the activity of the catalytic subunit and direct the PP2A complex to specific intracellular locations. To study PP2A’s role in signal transduction pathways that control growth and differentiation in vivo, a transgenic mouse lacking the B56g regulatory subunit of PP2A was made. Results: Lack of PP2A activity specific to the PP2A-B56g holoenzyme, resulted in the formation of an incomplete ventricular septum and a decrease in the number of ventricular cardiomyocytes. During cardiac development, B56g is expressed in the nucleus of a-actinin-positive cardiomyocytes that contain Z-bands. The pattern of B56g expression correlated with the cardiomyocyte apoptosis we observed in B56g-deficient mice during mid to late gestation. In addition to the cardiac phenotypes, mice lacking B56g have a decrease in locomotive coordination and gripping strength, indicating that B56g has a role in controlling PP2A activity required for efficient neuromuscular function. Conclusions: PP2A-B56g activity is required for efficient cardiomyocyte maturation and survival. The PP2A B56g regulatory subunit controls PP2A substrate specificity in vivo in a manner that cannot be fully C 2014 Wiley Periodicals, Inc. compensated for by other B56 subunits. Developmental Dynamics 243:778–790, 2014. V Key words: protein phosphatase 2A regulatory B subunit; PP2A; PR61; ventricular septal defect; mouse heart development; cardiomyocyte

Submitted 24 July 2013; First Decision 16 December 2013; Accepted 19 December 2013; Published online 14 January 2014

Protein phosphatase 2A (PP2A) is an abundant intracellular Ser/ Thr phosphatase, which plays a role in regulating signaling pathways that control cell cycle progression, DNA replication, protein degradation, and many other cellular functions. Regulatory subunits enable the relatively small number of Ser/Thr phosphatases to control their activity and specificity towards a large number of substrates (Shi, 2009). Hence, PP2A exists primarily as a heterotrimeric complex consisting of a 30-kDa catalytic C subunit, a 65kDa scaffold A subunit, and a variable regulatory B subunit. There are at least three multi-member families of PP2A regulatory B subunits, identified as B/PR55/PPP2R2, B’/B56/PR61/PPP2R5, and B’’/ PR72/PPP2R3 (Shi, 2009; Virshup and Shenolikar, 2009). The B56g regulatory B subunit, the focus of this study, has multiple naturally occurring splice variants and is one of the five B56 family members (McCright and Virshup, 1995; Tehrani et al., 1996; Csortos et al., 1996). The mechanisms that B subunits use to influence PP2A activity are not well understood. Structural analysis of the PP2A-B56 heterotrimeric complex has revealed that B56 binds the A subunit via

its highly conserved HEAT repeats 2–8 and that it also contacts the C subunit (Cho and Xu, 2007). The potential substrate binding site is located on the top face of the regulatory B subunit and is close to the catalytic subunit active site, thus providing a mechanism that the B subunits could use to affect substrate specificity of the PP2A complex. While the A subunit binding regions of all the B56 protein family members are very highly conserved, the N and C-terminal regions are not, which may allow these variable regions to confer unique enzymatic activity to the PP2A catalytic unit. B subunits can also regulate the PP2A catalytic subunit’s activity by targeting the heterotrimeric PP2A complex to specific intracellular compartments and substrates. B56a, b, and e have been shown to be present primarily in the cytosol, B56g primarily localizes to the nuclear region, and B56d localization has been shown to shift between the nucleus and the cytosol depending on the cell cycle (Tehrani et al., 1996; McCright et al., 1996). A number of specific B56g binding structures have been identified. For example, B56g has been shown to localize to nuclear speckles, regions associated with the accumulation of transcription and splicing factors in both neonatal and adult cardiomyocytes (Gigena et al., 2005) and B56g has been implicated in the regulation of the distribution of

*Correspondence to: Brent McCright, 29 Lincoln Drive, Bldg 29B, Room 2NN08, Bethesda, MD 20892. E-mail: [email protected]

Article is online at: http://onlinelibrary.wiley.com/doi/10.1002/dvdy. 24111/abstract C 2014 Wiley Periodicals, Inc. V

Introduction


PP2A-B56g IS REQUIRED FOR HEART DEVELOPMENT 779

chromosomes during mitosis and meiosis by binding to the centromeric cohesin protector protein, Shugoshin (Tang et al. 2006). B56 family members are expressed during embryonic development and may influence the Wnt/b-catenin signaling pathway, which plays a major role in controlling organogenesis (Martens et al., 2004; Seeling et al., 1999). Exogenous expression of B56 subunits in Xenopus embryos prevented the Xwnt-8 induced formation of secondary dorsal body axis in Xenopus embryos by enhancing degradation of b-catenin (Li et al. 2001). Likewise, in mice that overexpress B56g in the lung, the amount of b-catenin is also decreased (Everett et al. 2002). PP2A-B56 activity has also been implicated as a regulator of insulin signaling and lipid metabolism via the regulation of Akt phosphorylation in both C. elegans (Padmanabhan et al., 2009) and Drosophila (Vereshchagina et al., 2008). Misregulation of PP2A activity has also been shown to be the cause of Opitz BBB/G Syndrome, a congenital disorder characterized by cleft lip/palate, heart defects (ventricular and atrial septal defects), dysphagia, and mental retardation (Liu et al., 2011). In vivo manipulation of signaling pathways in transgenic mice is a valuable tool for identifying biological activity and function. It has been previously shown that the knockout of the PP2A catalytic subunit results in early embryonic lethality, making it a limited tool for studying in vivo PP2A activity (Gotz et al., 1998). So far the results from only one mouse B56 knockout study, B56d, have been reported. In the absence of B56d, the microtubule-associated protein tau is hyper-phosphorylated due to an increase in glycogen synthase kinase-3b activity and the mice have impaired sensorimotor, but normal cognitive functions (Louis et al., 2011). In this study, we created a B56g knockout strain of mice in order to uncover PP2A activities that are regulated by B56g during mammalian development. We have chosen B56g to study because our interest in heart development coincided with the previously reported expression of B56 in the fetal mouse (Martens et al., 2004) and human heart (McCright and Virshup, 1995). We found that B56g expressed in the nucleus of fetal cardiomyocytes is required for the efficient survival of cardiomyocytes and the formation of a ventricular septum. Our study is the first report of a requirement for regulation of PP2A activity via a B56 subunit during heart development.

mRNA in B56c/ tissues, whereas B56g–bgeo fusion mRNA was detected in multiple B56c/ tissues (Fig. 1B). Additional RT-PCR primer combinations that tested for transcription starting after the gene trap, or alternative splicing around the gene trap, did not detect any mRNA that could result in potentially functional B56g protein expression (data not shown). B56g protein corresponding to the g2 and g3 isoforms was detected by Western blot analysis on the extracts from the wild type and heterozygous mouse embryonic fibroblasts (MEFs), but no B56g protein was detected in B56c/ MEFs (Fig. 1C). The antibody used for Western blots recognizes B56g peptide sequence near the carboxyl terminus. Therefore, it would be able to detect B56g peptides that were products of alternative mRNA splicing events that bypassed the gene trap. Since no residual native B56g protein or mRNA can be detected in B56c/ mice, it is likely that this strain represents a complete loss of function or “null” mutation. No change was observed in the protein expression of the catalytic subunit of PP2A (PP2Ac), showing that the absence of the B56g subunit did not substantially affect the amount of PP2Ac present in the extracts (Fig. 1C).

Results Generation of Mice That Express No B56g

B56c2/2 Mice Have Ventricular Septum and Myocardial Abnormalities

A mouse strain that expresses no functional B56g, B56cXP0444, was created using the embryonic stem cell line XP0444 from the Sanger Institute Gene Trap Resource (SIGTR). In the XP0444 cell line, a strong splice acceptor sequence was inserted into the second intron of B56c, and thus would be expected to produce a fusion protein consisting of the first 88 amino acids of B56g joined to bgeo (Fig. 1A). Since this protein does not contain the PP2A A subunit binding domain, any potential fusion protein would be unable to associate with the PP2A complex and would be expected to have no effect on PP2A activity (Li and Virshup, 2002). Throughout the rest of this article, wild type, heterozygous, and homozygous B56cXP0444 mice will be referred to as B56cþ/þ, B56cþ/, or B56c/ respectively. RT-PCR analysis was used to detect native B56g and gene trap mRNA in tissue samples. Note the B56g encoding region within the PCR primers used for this analysis is common to all 4 known B56g splice variants. RT-PCR analysis detected no wild type B56g

B56c/ mice appeared paler at birth than their littermates suggesting problems with their circulation, so non-invasive and histological techniques were used to analyze heart structure and function. Doppler color imaging of E16 mice detected mixing of right and left ventricular blood flow in B56c/ mice due to the lack of ventricular septum formation (Fig. 3B). Histological analysis confirmed ventricular septal defects in E16 B56c/ hearts (Fig. 3C and D). Although all B56c/ fetuses show a decrease in ventricular heart tissue, only about half (11 out 21) had a clearly observable ventricular septum defect. The septums of the B56c/ adults usually appeared thinner than their wild type littermates, but no incomplete ventricular septums were detected in adult hearts (Fig. 3E and F). This is likely due to the combination of the most strongly affected mice dying shortly after birth, and the eventual closure of the ventricular septum in surviving B56c/ mice.

B56c2/2 Have a Neonatal Growth Deficiency But Develop Into Obese Adult Mice No dead B56c/ embryos were found and no deviation from expected genotypic distribution was detected in litters analyzed during fetal development. However, intercrosses of heterozygous offspring resulted in a non-Mendelian ratio of genotypes at weaning, 33% wild type, 52% heterozygous, and 15% B56c/ mice (n ¼ 138 in total) indicating about a 40% loss of B56c/ mice during neonatal development. Neonatal B56c/ mice were less active and on average were 36% smaller in mass than their heterozygous and wild type litter mates (Fig. 2A and B,Table 1). The majority of B56c/ mice that died were lost within a day or two of birth. After weaning, B56c/ mice developed into obese adults and by age 6 months were on an average 31% heavier than their littermates (Fig. 2C and D, Table 1). The obesity found in B56c/ mice is due to a large increase in adipose tissue (Fig. 2E). Heterozygous mice do not have an intermediate growth or weight phenotype. Instead, B56cþ/ mice appear to be phenotypically wild type in their appearance, development, growth, and behavior.


780 VARADKAR ET AL.

Fig. 1. Generation of B56c/ mice. A: Diagram of the 50 region of the B56c gene showing the insertion of the gene trap cassette with a splice acceptor (SA), b-galactosidase-neomycin resistance (b-geo) construct, and polyadenylation sequence (pA). B: Semi-quantitative RT-PCR was used to detect native (WT, g1 to g4) and B56g XP0444 specific (MT, g1 to 302) mRNA in wild type (þ/þ), heterozygous (þ/) and homozygous B56c/ (/) tissues. S, skeletal muscle; B, brain; H, heart; K, kidney; P, pancreas. C: Western blot analysis of E13 mouse embryonic fibroblast (MEF’s) extracts (n¼3) was used to detect native PP2A B56g3, B56g2, and PP2A catalytic subunit (PP2Ac) protein expression in wild type (þ/þ), heterozygous (þ/), and gene trap homozyogous (/) samples. No expression of native PP2A B56g3 or B56g2 was detected in the homozygous (/) samples.

Endogenous B56g Is Expressed in a-Actinin-Positive Cardiomyocytes To determine the cell types that express B56g from E10–E16, colocalization experiments were performed using an affinity purified B56g-specific antibody, and antibodies for cardiovascular cell markers. High levels of B56g expression were found in a-actinin-positive cardiomyocytes throughout the heart including in the ventricular septum (Fig. 4A–C). Little B56g expression was observed in the a-smooth muscle actin (aSMA) positive cells present in the periphery of the developing ventricles at E13 or E16 (Fig. 4E and F), but there was co-localization with aSMA at E10 (Fig. 4D). Note that aSMA is present in blood vessel smooth muscle tissue and is also transiently expressed in immature embryonic cardiomyocytes (Ya et al., 1997; Clement et al., 2007), whereas a-actinin is localized to the sarcomeric Z-bands of mature cardiomyocytes. Therefore, during cardiac development, B56g appears to be primarily expressed in more mature a-actinin positive cardiomyocytes. B56g is not expressed in the endocardial or endothelial cells of the heart (Fig. 4G–I).

B56g Is Localized to the Nucleus of E16 Cardiomyocytes At E10, endogenous B56g expression was seen to be primarily localized in the perinuclear and cytoplasmic regions of the cell

(Fig. 5A–C) whereas at E16 the majority of B56g appears to be localized in the nucleus, as shown by the co-localization of endogenous B56g with Hoechst stain (Fig. 5D–F). The nuclear expression of B56g was confirmed by the immunohistochemical analysis of cardiomyocytes, cultured and isolated from the E16 wild type hearts (Fig. 5G–I). The intracellular localization of B56g appears to shift to being predominately nuclear in the later stages of cardiomyocyte development.

Loss of B56g Expression Reduces the Number of Cardiomyocytes With Sarcomeric Z-Bands Because there appeared to be a decrease in heart tissue in B56c/ mice, markers corresponding to different stages of cardiomyocyte development were used to determine if specific cell populations are affected by the lack of PP2A-B56g specific activity. Immunohistochemical analysis showed a marked reduction in the number of cardiomyocytes expressing sarcomeric a-actinin in E16 B56c/ hearts, relative to wild type (Fig. 6A–H). Decreases in the number of a-actinin positive Z-bands was observed in both the ventricular wall and septum of B56c/ hearts. Using a similar approach, no change was observed in the number of aSMA positive cells indicating that B56g is not required for cardiomyocyte specification and the early stages of cardiomyocyte development (Fig. 6I and J).



Fig. 2. B56c/ neonatal mice are small but become obese adults. A,B: Age-matched, wild type (þ/þ) and small-sized B56c/ (/) mice at 5 days old (P5) and 21 days old (P21). C,D: Wild type (þ/þ) and obese B56c/ (/) mice at 3 months (3 m) and 6 months (6 m). E: Comparison of wild type (þ/þ) and obese B56g/ (/) mice at 6 months showing excess of adipose tissue (arrows) in B56c/ mice.

Loss of B56g Expression Causes a-Actinin-Positive Ventricular Cardiomyocytes to Undergo Apoptosis In order to understand the mechanism behind the decreased cell number in B56c/ hearts, cellular apoptosis during heart development was analyzed using the TUNEL assay and cleaved Caspase-3 staining. TUNEL assay revealed that at day 10 of gestation (E10), B56c/ mice had similar low levels of apoptotic cells present in the heart (Fig. 7A and B). In contrast, B56c/ mice had elevated apoptosis in localized regions of the ventricu-

lar trabeculae and septum at E13 and E16 relative to the wild type hearts (Fig. 7C–F). The percentage of TUNEL-positive cells in E16 B56c/ hearts was found to be 24.6% 6 0.82 as compared to 4.7% 6 0.54 in B56cþ/þ hearts. Values represent means 6 SEM (n ¼ 6, P < 0.001). The TUNEL results were further confirmed using Caspase-3 antibody staining, which also detected increased cell death in E16 B56c/ hearts (Fig. 7G and H). To identify the cell types that were undergoing apoptosis, we performed co-localization studies with cardiac markers. Colocalization studies with the cardiac markers showed that the

782 VARADKAR ET AL.

TABLE 1. Comparison of the Body Weights of Wild Type and B56c-/- Mice Age

Strain

Day 1

WT [n¼6] KO [n¼6] WT [n¼6] KO [n¼6] WT [n¼6] KO [n¼6] WT [n¼6] KO [n¼6] WT [n¼6] KO [n¼6]

Day 20 2 months 3 months 5 months

Body weight (gms) 1.68 1.07 7.78 6.23 18.33 21.95 20.25 25.93 23.03 30.21

6 6 6 6 6 6 6 6 6 6

0.09 0.15* 0.35 0.25* 0.82 1.67y 1.32 2.67y 0.61 3.05y

% Effect on size KO vs. WT 36 % # 20 % # 19 % " 28 % " 31 % "


WT, wild type mice; KO, B56c-/- mice. *P < 0.001. y P < 0.01.

TUNEL-positive cells expressed a-actinin at E13 and E16 (Fig. 7I and K) but did not co-localize with aSMA-positive cells (Fig. 7J and L). Together, our results show that the presence of nuclear PP2A-B56g is required to prevent the cell death of a portion of the a-actinin positive ventricular cardiomyocytes during later stages of fetal heart development.

Loss of B56g Expression Does Not Affect Cell Proliferation or Amount of b-Catenin Bromodeoxyuridine (BrdU) incorporation was used to determine the effect that losing B56g had on fetal cardiac cell proliferation. BrdU analysis revealed that there was no apparent difference in the number of proliferating cells in wild type (Fig. 8A and C) as compared to B56c/ (Fig. 8B and D) embryonic hearts at E10 and E16. Colocalization studies show that some of the proliferating E16 cells co-express B56g protein indicating that B56g is not expressed exclusively in terminally differentiated cells (Fig. 8E and F). Because B56 and PP2A has been shown to interact with the Wnt signaling pathway by regulating the stability of b-catenin protein (Seeling et al. 1999), the amount of b-catenin present in E16 cardiomyocytes was assayed by Western blot. We found no difference in the amount of b-catenin present in B56c/ extracts, relative to B56cþ/ or B56cþ/þ extracts, suggesting that the mechanism associated with the cell death was independent of the Wnt signaling pathway (Fig. 8G).

B56c2/2 Mice Have Muscle Strength and Coordination Defects B56c/ mice were observed to have a wobbly gait that suggested a neuromuscular defect. To qualitatively assess muscle function, a “Hanging wire test” was used to measure grip strength and coordination. B56c/ mice released their grip on the wire lid almost immediately, with an average hang time less than one second, whereas the wild type littermates suspended themselves for considerably longer times, with an average of 17 sec (n ¼ 9, P < 0.0001). This deficiency lasts throughout the life of the mouse. To further investigate the loss of coordination, the mice were evaluated by rotarod testing. There was approximately a 10-fold decrease in the amount of time that B56c/ mice are

able to maintain their balance on the rotarod device as compared to their wild type littermates (Fig. 9).

Discussion Most of the currently available information on how B subunits regulate PP2A catalytic activity has been derived from cell culture experiments. However, studies of intact organisms are needed to associate biological functions with specific regulatory B subunits and to determine to what extent are the B subunits functionally redundant. In this study, we have examined the phenotypes present in a mouse strain containing a null mutation of B56g to identify functional requirements for this PP2A regulatory subunit. We found that mice lacking the B56g regulatory subunit of PP2A have a defect in the formation of the ventricular septum of the heart. This lack of a complete ventricular septum is caused by a reduction in myocardial tissue due to cell death in aactinin-positive cardiomyocytes that normally express B56g. PP2A-B56g activity does not appear to be required for cardiomyocyte specification or progenitor proliferation because these attributes appear unaffected by its absence. Instead, the timing of the apoptosis suggests that B56g may play a role in controlling the transition from highly proliferating, to terminally differentiated cardiomyocytes. When PP2A containing B56g is not present, this transition does not occur appropriately, causing a portion of the cardiomyocytes to die. Depending on the context, PP2A has been shown to have the potential to regulate cell growth and proliferation by either promoting or preventing apoptosis (Janssens and Rebollo, 2012). Our findings suggest that B56g normally regulates PP2A activity towards nuclear localized substrates required for promoting the efficient survival of differentiated cardiomyocytes late in fetal development. The phenotypes observed in the B56c/ hearts have similarities to mice that expressed high levels of a dominant negative mutant of the A subunit (AD5) that can still bind the C subunit but cannot bind B subunits (Brewis et al., 2000). The AD5 mice also had thin ventricular walls similar to what we observed in the B56c/ fetal hearts. The AD5 phenotypes and our data are consistent with B56g incorporating into the PP2A heterotrimer to regulate PP2A activity towards specific phosphoproteins whose phosphorylation status is critical in maintaining healthy cardiomyocytes.



Fig. 3. B56c/ mice exhibit ventricular septum and myocardial abnormalities. A,B: Ultrasound 2D imaging showing the frontal plane of an E16 fetus. Color Doppler flow mapping shows mixing of the blood flow through the ventricular septum of a B56c/ heart (B, arrow), compared to the normal flow in the wild type heart (A). C,D: Histological sections from E16 wild type (C), and E16 B56c/ hearts showing a ventricular septum defect (D, VSD, arrow). Scale bar ¼ 0.2 mm. E,F: Histology sections from 6-month-old wild type (E) and B56c/ (F) hearts showing thinning of the ventricular septum (arrows). Scale bar ¼ 1 mm.

B56c/ mice clearly accumulate more white adipose tissue than their littermates (Fig. 2E), but the mechanism behind this 100% penetrant phenotype has not been determined. One possibility is that B56g is normally responsible for directing PP2A activity towards a substrate that helps control metabolism and fat storage. Interestingly, heterotrimeric PP2A containing B56

family members have been identified as modulators of metabolism via the insulin/IGF pathway in both C. elegans (Padmanabhan et al., 2009) and Drosophila (Vereshchagina et al., 2008; Hahn et al., 2010). The Drosophila B56-1 homologue, which is most similar to B56g, has been shown to direct PP2A to dephosphorylate S6K in Drosophila. When B56-1 is knocked


784 VARADKAR ET AL.

Fig. 4. B56g is expressed in a-actinin positive cells throughout heart development. A–C: B56g co-localizes with a-actinin-positive cells. B56g (green) and a-actinin (red) co-localize (yellow, arrows) in ventricular cardiomyocytes at E10, E13, and E16. D–F: B56g was found to co-localize (yellow, arrows) with a-smooth muscle actin–positive cells (SMA) at E10 but not at E13 or E16. G–I: B56g is not expressed in PECAM positive endothelial cells. Representative immunofluorescence images from the embryonic hearts at E10, E13, and E16. Data are representative of three mice from each group. Figure 4 was imaged using a 40 objective.

out, the increase in S6K phosphorylation results in a reduction of tryglycerides. In the other studies, the B56 homologue that more closely resembles B56b, was shown to specifically direct PP2A activity to dephosphorylate Akt-1 in both Drosophila and C. elegans, also resulting in a reduction in triglyceride formation. Control of fat synthesis and utilization is complex and other signaling pathways could influence its accumulation. For instance, PP2A containing B56a has been shown to inhibit lipolysis (Kinney et al., 2010). Conversely, the increase of fat in B56c/ mice could be a secondary effect, caused by a decrease in physical activity that may be related to their heart and coordination phenotypes and/or an increase in food intake.

PP2A inhibitors such as okadaic acid are known tumor inducers and the small t antigen of SV40 transforms cells in culture by replacing B subunits suggesting that certain PP2A B subunits might have tumor suppressor activity (Ruediger et al., 1994; Janssens et al., 2005). In addition, several prior in vitro studies have specifically identified B56g as a potential tumor suppressor (Chen et al., 2004; Li et al., 2007; Shouse et al., 2011). So far we have not seen any increase in tumor formation in B56c/ mice under normal colony maintenance conditions, including mice that were 18 months old. However, the B56c/ mice could be more susceptible to carcinogenic reagents or may genetically interact in an additive manner with mouse strains containing oncogenic transgenes.



Fig. 5. B56g expression varies its cellular localization during heart development. A, D: Immunohistochemistry using anti-B56g (green) detects B56g protein in the embryonic hearts of wild type mice at E10 and E16. B, C: Higher magnification of the boxed region in A shows perinuclear and cytoplasmic B56g (B, C, arrows) not co-localizing with nuclear Hoechst 33342 stain (C). E, F: Higher magnification of the boxed region in D shows the nuclear localization of B56g (E, F, arrows) with Hoechst 33342 (F) at E16. G–I: Immunohistochemistry shows native B56g (green) expression in cardiomyocytes cultured from E16 wild type hearts co-localizes in the nucleus with DAPI. A and D were imaged using a 10 objective, G and H were imaged using a 40 objective, B, C, E, F, and I were imaged using a 60 objective.

B56g is expressed in both heart and skeletal muscle but we could not detect any structural developmental defects in the muscles of the B56c/ mice (data not shown). Therefore, at this time we do not have a cause and effect mechanism to explain the observed muscular function defects. However, it is interesting to note that like the B56c/ mice, B56d knockout mice have also been reported to have motor control and coordination deficiencies (Louis et al., 2011). It is therefore possible that these two subunits have some overlapping function, possibly cerebellar in origin, necessary for coordination of complex movements. In all of our analyses, the heterozygous mice appeared to be wild type in appearance. This finding is most consistent with the presence of a surplus capacity of PP2A-B56g activity being present in normal tissues and a tolerance for some alteration in PP2A-B subunit stoichiometry. It is also plausible that the function of B56g is being compensated by other PP2A-B56 complexes in B56gþ/ mice because there are multiple B56 family members

present in cardiomyocytes and most other tissues. However, the B56c/ phenotypes support the hypothesis that B56 regulatory subunits preferentially regulate PP2A activity towards specific substrates in vivo and that their function cannot be fully compensated for by other B subunits present in the cell. The B56cXP0444 mice and other mouse strains containing PP2A transgenes and knockouts can provide valuable tools to further identify and define PP2A mechanisms used for controlling cellular behavior both in vivo and in vitro.

Experimental Procedures Generation of B56g

XP0444

Mice

An embryonic stem cell line, XP0444, containing a gene trap disruption of the PP2A regulatory subunit B56c was obtained from the Sanger Institute Gene Trap Resource (SIGTR). Germ line

786 VARADKAR ET AL.

transmission was achieved by injecting the XP0444 cell line into C57BL/6 blastocyts and breeding the chimeric offspring onto a C57BL6 background. PCR, followed by genomic DNA sequencing,

confirmed the location of the gene trap near the start of the second intron of B56c, at base 36,819 relative to the predicted start of transcription. Genotyping DNA primers B56g1 (CAGAAGCTACGCCAGTGTTGTG) and B56g2 (GATACTCAGTACAGCTCCTTACAG) were used for identifying wild type mice, and B56g1 and B56g3 (TTGGGACCACCTCATCAGAAGC) were used for identifying B56cXP0444 mice (Fig. 1B). Experimental data was collected from mice outcrossed onto the C57BL/6 background for more than 5 generations. All the experiments conducted in these studies were performed in accordance with protocols approved by the Institutional Animal Care and Use Committee at our FDA research facility.

RT PCR Analyses


Total RNA was isolated using the Trizol reagent (Life Technologies, Inc., Carlsbad, CA) according to the manufacturer’s protocol. B56g mRNA was reverse transcribed using B56g5 (ATCTTGTGTTCCTCCTTCAGTGG, 704 to 682 B56g mRNA) and PCR amplified using primers B56g1 (100–120 B56g mRNA) and B56g4 (GTAACTCCGCTATGCCATTGTG, 643–622 B56g mRNA). B56g-bgeo fusion mRNA was reverse transcribed using Oligo 300 TAATGGGATAGGTCACG (Bases 1,935–1,919 of gene trap vector pGT01xf). Oligo 302 AGTATCGGCCTCAGGAAGATCG (Bases 1,862–1,841 of gene trap vector pGT01xf) and B56g1 (100–120 B56g mRNA) were used to generate the B56cXP0444 PCR product (Fig. 1B). Other primer sets used to test for alternative B56 mRNA included RT primer B56g11 ggcgttgtatatcaagccgtg Bases 1,299–1,279 (exon 11) followed by PCR with B56g-9 ACTCATGATGTACTCGTTGTTCC Bases 1,190–1,169 (exon 10) and B56g-1; or B56g-12 GTTAACATGTTCCGAACCTTGC Bases 421–442 (exon 3) and B56g-9.

Rotarod Analyses Mice were trained on the rotarod (Ugo Basile) for 2 days before collecting data. Each trial consisted of placing the mouse on the rod at 10 rpm for 60 sec (stabilizing period) followed by acceleration from 10 to 40 rpm over a period of 25 sec. The latency to fall (sec) was recorded, and all six scores per mouse were averaged and were recorded as latency to fall (in seconds) for each mouse. Each trial was done twice a day (2-hr interval between sessions) for 3 consecutive days. Statistical analysis was done using Student’s two-tailed paired t-test.

Echocardiography Color flow and spectral Doppler images were obtained for E16 embryos in utero using an Acuson Sequoia C256

Fig. 6.

Fig. 6. Loss of B56g expression reduces the number of cardiomyocytes with sarcomeric Z-bands. Immunohistochemistry detects aactinin-positive cells in the left ventricular wall of E16 wild type (A) and B56c/ (B) hearts. A higher magnification of the boxed region in A (C, arrows) shows an increased number of Z-bands in a-actinin positive cardiomyocytes in the hearts of wild type mice relative to the higher magnification of the boxed region in B (D, arrows) of B56c2/2 mice. Likewise, there is a decrease in the number of a-actinin positive cells in the septum of B56c2/2 hearts (H) relative to wild type hearts (G, arrows) which are the higher magnifications of the boxed region in F and E respectively. No difference was observed in the abundance of asmooth muscle actin positive cells (SMA) in wildtype (I) and B56c/ (J) E16 embryonic hearts. Data are representative of three mice from each group. A, B, E, F, I, and J were imaged using a 20 objective, C, D, G, and H were imaged using a 40 objective.



Fig. 7. Loss of B56g causes a-actinin-positive ventricular cardiomyocytes to undergo apoptosis. A, B: TUNEL staining shows no difference in apoptosis in the ventricular regions of hearts from wild type (A) and B56c/ (B) mice at E10. C–F: Large numbers of TUNEL-positive cells were seen in B56c/ hearts at E13 (D, boxed region with arrows) and E16 (F, boxed region with arrow) compared to the wild type hearts at E13 (C) and E16 (E), respectively. G, H: Cleaved Caspase-3 staining detects more apoptosis in the ventricular region of B56c/ hearts (H) than wild type hearts (G) at E16. I, K: TUNEL-positive cardiomyocytes (green) co-localize (yellow, arrows) with a-actinin (red) in E13 and E16 B56c/ embryonic hearts respectively. J, L: TUNEL-positive cells (green) do not co-localize with a-smooth muscle actin–positive (red, SMA) cells. Data are representative of three mice from each group. A–F were imaged using a 10 objective, G–L were imaged using a 40 objective.

ultrasound system with a 15-MHz L8 linear phased array transducer.

Histology and Immunohistochemistry Paraffin-embedded histological sections and cryo-preserved frozen sections were prepared from fetal hearts fixed in 4% paraformaldehyde/PBS as previously described (Varadkar et al. 2008). After blocking was carried out with 10% donkey serum/ TTBS, primary antibodies to cleaved Caspase-3 (Cell Signaling, Danvers, MA; 1:1,000), a-smooth muscle actin (1:40 dilution, Sigma, St. Louis, MO), CD31 (1:100 dilution, BD Pharmingen, San Jose, CA), a-actinin (1:250 dilution, Sigma), and B56g (1 mg/ml dilution) were used. To create a B56g specific antibody, a peptide corresponding to amino acids 471–486 of murine B56g3 and amino acids 432–447 of B56g2 (QAQKELKKDRPLVRRK) was injected into two rabbits. The antiserum was peptide affinity purified prior to use. Secondary

antibodies, fluorescein isothiocyanate (FITC) conjugated donkey anti-rabbit (1:200 dilution, Jackson ImmunoResearch, West Grove, PA), tetrarhodamine isothiocyanate (TRITC) conjugated donkey anti-rat (1:200 dilution, Jackson ImmunoResearch), and TRITC conjugated donkey anti-mouse (1:200 dilution, Jackson ImmunoResearch) were used. Hoechst 33342 (Life Technologies) was used to visualize the nuclei. Images were captured using a Zeiss (Thornwood, NY) 710 confocal microscope.

Isolation of Embryonic Cardiomyocytes E16.5 ventricles were minced and digested with collagenase and then pre-plated for 2 hr to enrich for cardiomyocytes. The cells were then cultured in DMEM/F12 medium containing 10% FBS on fibronectin coated Lab-Tek II chamber slides. Immunohistochemistry was performed using B56g specific antibody as described in the Histology and Immunohistochemistry section.


788 VARADKAR ET AL.

Fig. 8. Loss of B56g expression does not affect the cell proliferation or b-catenin levels. A–D: Bromodeoxyuridine (BrdU, green) labeling in conjunction with DAPI (blue) was used to label proliferating cells at E10 and E16 in wild type and B56c/ embryonic hearts. E: Merged image of BrdU-positive cells (green) DAPI (blue) with B56g (red) at E16 in wild type embryonic hearts. F: Higher magnification of boxed region in E showing merged image of BrdU-positive proliferating cells (green) and B56g expression (red) in wild type mice at E16. Some of the proliferating cells express B56g (yellow, arrows). G: Western blot analysis of E16 mouse embryonic heart extracts was done to detect b-catenin in wild type (þ/þ), heterozygous (þ/), homozyogous (/) samples. A–D were imaged using a 10 objective, E and F were imaged using a 40 objective.

TUNEL (TdT-Mediated dUTP Nick-End Labeling) Assay Embryonic heart tissues were fixed and paraffin embedded. Tissues were sectioned at 7 mm, then dewaxed and rehydrated

through graded alcohols. The sections were permeabilized in 10 mM sodium citrate buffer (pH 6) by applying microwave irradiation for 1 min. Blocking was carried out with 5% mouse serum/ TTBS. The slides were incubated with the reaction mixture


References


Fig. 9. B56c/ mice have muscle coordination defects. Rotarod tests in age and gender-matched wild-type (þ/þ), heterozygous (þ/), and B56c/ (/) mice. Values represent seconds prior to fall, means 6 SEM (n ¼ 6). *P < 0.001, statistical analysis was done using a Student’s two-tailed paired t-test.

containing TdT (Roche, Indianapolis, IN) and fluorescein conjugated dUTP (Roche) for 1 hr at 37 C in a dark humidified chamber. The tissues were washed and then incubated with primary antibodies to a-smooth muscle actin and a-actinin as mentioned above. The percentage of TUNEL-positive cells was determined by counting the number of apoptotic cells and total cells in 10 sections from both B56g/ and B56gþ/þ heart sections. Slides from six individual B56g/ and B56gþ/þ mice were used for these analyses.

Bromodeoxyuridine Labeling Assay E10.5 and E16.5 pregnant mice were given 3 mg of bromodeoxyuridine (BrdU) by intraperitoneal injection 1–2 hr before the embryos were removed for processing. Tissues were fixed in 4% paraformaldehyde overnight at 4 C. Hearts were sectioned at 7 mm, then dewaxed and rehydrated through graded alcohols. Tissues were permeabilized with boiling 10 mM sodium citrate buffer (pH 6). Blocking was carried out with 5% mouse serum/ TTBS and proliferating cells were detected using an Alexa Fluor 488-conjugated mouse anti-BrdU (1:40 dilution, Sigma).

Western Blots To assay the expression of B56g protein, extracts from mouse embryonic fibroblasts (MEFs) from the gestation day–13 embryos were used. For evaluation of the amount of b-catenin present, extracts were isolated from gestation day–16 hearts. Equal amounts of protein were separated by SDS-PAGE (4–12% gels; Invitrogen, Carlsbad, CA) and then transferred onto polyvinylidene difluoride membranes (Millipore, Billerica, MA). The membranes were blocked with 10% Blocking solution (Odyssey-LiCOR) and sequentially incubated with antibodies against B56g (1 mg/ml), PP2A catalytic subunit (Millipore, 1:500), b-catenin (Cell Signaling, 1:1,000), and then with anti-rabbit and anti-mouse secondary antibody IRDye 800 CW (Odyssey-Li-COR), respectively. Western blots were imaged and analyzed by Odyssey LiCOR infrared System.

Acknowledgments The preclinical phenotyping facility at CNMC is supported by a grant from US Department of Defense grant no. W81XWH-09-10599 to Dr. Nagaraju.

Brewis N, Ohst K, Fields K, Rapacciuolo A, Chou D, Bloor C, Dillmann W, Rockman H, Walter G. 2000. Dilated cardiomyopathy in transgenic mice expressing a mutant A subunit of protein phosphatase 2A. Am J Physiol Heart Circ Physiol 279:H1307– H1318. Chen W, Possemato R, Campbell KT, Plattner CA, Pallas DC, Hahn WC. 2004. Identification of specific PP2A complexes involved in human cell transformation 2. Cancer Cell 5:127–136. Cho US, Xu W. 2007. Crystal structure of a protein phosphatase 2A heterotrimeric holoenzyme. Nature 445:53–57. Clement S, Stouffs M, Bettiol E, Kampf S, Krause KH, Chaponnier C, Jaconi M. 2007. Expression and function of alpha-smooth muscle actin during embryonic-stem-cell-derived cardiomyocyte differentiation. J Cell Sci 120:229–238. Csortos C, Zolnierowicz S, Bako E, Durbin SD, DePaoli-Roach AA. 1996. High complexity in the expression of the B’ subunit of protein phosphatase 2A0. Evidence for the existence of at least seven novel isoforms. J Biol Chem 271:2578–2588. Everett AD, Kamibayashi C, Brautigan DL. 2002. Transgenic expression of protein phosphatase 2A regulatory subunit B56gamma disrupts distal lung differentiation. Am J Physiol Lung Cell Mol Physiol 282:L1266–L1271. Gigena MS, Ito A, Nojima H, Rogers TB. 2005. A B56 regulatory subunit of protein phosphatase 2A localizes to nuclear speckles in cardiomyocytes. Am J Physiol Heart Circ Physiol 289:H285– H294. Gotz J, Probst A, Ehler E, Hemmings B, Kues W. 1998. Delayed embryonic lethality in mice lacking protein phosphatase 2A catalytic subunit Calpha. Proc Natl Acad Sci USA 95:12370–12375. Hahn K, Miranda M, Francis VA, Vendrell J, Zorzano A, Teleman AA. 2010. PP2A regulatory subunit PP2A-B’ counteracts S6K phosphorylation. Cell Metab 11:438–444. Janssens V, Rebollo A. 2012. The role and therapeutic potential of Ser/Thr phosphatase PP2A in apoptotic signalling networks in human cancer cells. Curr Mol Med 12:268–287. Janssens V, Goris J, Van Hoof C. 2005. PP2A: the expected tumor suppressor. Curr Opin Genet Dev 15:34–41. Kinney BP, Qiao L, Levaugh JM, Shao J. 2010. B56alpha/protein phosphatase 2A inhibits adipose lipolysis in high-fat diet-induced obese mice. Endocrinology 151:3624–3632. Li HH, Cai X, Shouse GP, Piluso LG, Liu X. 2007. A specific PP2A regulatory subunit, B56gamma, mediates DNA damage-induced dephosphorylation of p53 at Thr55. EMBO J 26:402–411. Li X, Virshup DM. 2002. Two conserved domains in regulatory B subunits mediate binding to the A subunit of protein phosphatase 2A. Eur J Biochem 269:546–552. Li X, Yost HJ, Virshup DM, Seeling JM. 2001. Protein phosphatase 2A and its B56 regulatory subunit inhibit Wnt signaling in Xenopus. EMBO J 20:4122–4131. Liu E, Knutzen CA, Krauss S, Schweiger S, Chiang GG. 2011. Control of mTORC1 signaling by the Opitz syndrome protein MID1. Proc Natl Acad Sci USA 108:8680–8685. Louis JV, Martens E, Borghgraef P, Lambrecht C, Sents W, Longin S, Zwaenepoel K, Pijnenborg R, Landrieu I, Lippens G, Ledermann B, Gotz J, Van LF, Goris J, Janssens V. 2011. Mice lacking phosphatase PP2A subunit PR61/B’delta (Ppp2r5d) develop spatially restricted tauopathy by deregulation of CDK5 and GSK3beta. Proc Natl Acad Sci USA 108:6957–6962. Martens E, Stevens I, Janssens V, Vermeesch J, Gotz J, Goris J, Van HC. 2004. Genomic organisation, chromosomal localisation tissue distribution and developmental regulation of the PR61/B’ regulatory subunits of protein phosphatase 2A in mice. J Mol Biol 336:971–986. McCright B, Virshup DM. 1995. Identification of a new family of protein phosphatase 2A regulatory subunits. J Biol Chem 270: 26123–26128. McCright B, Rivers AM, Audlin S, Virshup DM. 1996. The B56 family of protein phosphatase 2A (PP2A) regulatory subunits encodes differentiation-induced phosphoproteins that target PP2A to both nucleus and cytoplasm. J Biol Chem 271:22081–22089. Padmanabhan S, Mukhopadhyay A, Narasimhan SD, Tesz G, Czech MP, Tissenbaum HA. 2009. A PP2A regulatory subunit

790 VARADKAR ET AL.


regulates C. elegans insulin/IGF-1 signaling by modulating AKT-1 phosphorylation. Cell 136:939–951. Ruediger R, Hentz M, Fait J, Mumby M, Walter G. 1994. Molecular model of the A subunit of protein phosphatase 2A: interaction with other subunits and tumor antigens 1. J Virol 68:123–129. Seeling JM, Miller JR, Gil R, Moon RT, White R, Virshup DM. 1999. Regulation of beta-catenin signaling by the B56 subunit of protein phosphatase 2A 4. Science 283:2089–2091. Shi Y. 2009. Serine/threonine phosphatases: mechanism through structure. Cell 139:468–484. Shouse GP, Nobumori Y, Panowicz MJ, Liu X. 2011. ATM-mediated phosphorylation activates the tumor-suppressive function of B56gamma-PP2A. Oncogene 30:3755–3765. Tang Z, Shu H, Qi W, Mahmood NA, Mumby MC, Yu H. 2006. PP2A is required for centromeric localization of Sgo1 and proper chromosome segregation 1. Dev Cell 10:575–585.

Tehrani MA, Mumby MC, Kamibayashi C. 1996. Identification of a novel protein phosphatase 2A regulatory subunit highly expressed in muscle. J Biol Chem 271:5164–5170. Varadkar P, Kraman M, Despres D, Ma G, Lozier J, McCright B. 2008. Notch2 is required for the proliferation of cardiac neural crest-derived smooth muscle cells. Dev Dyn 237:1144–1152. Vereshchagina N, Ramel MC, Bitoun E, Wilson C. 2008. The protein phosphatase PP2A-B’ subunit Widerborst is a negative regulator of cytoplasmic activated Akt and lipid metabolism in Drosophila. J Cell Sci 121:3383–3392. Virshup DM, Shenolikar S. 2009. From promiscuity to precision: protein phosphatases get a makeover. Mol Cell 33:537–545. Ya J, Markman MW, Wagenaar GT, Blommaart PJ, Moorman AF, Lamers WH. 1997. Expression of the smooth-muscle proteins alpha-smooth-muscle actin and calponin, and of the intermediate filament protein desmin are parameters of cardiomyocyte maturation in the prenatal rat heart. Anat Rec 249:495–505.