The B56α subunit of PP2A is necessary for mesenchymal stem cell commitment to adipocyte
2021; Springer Nature; Volume: 22; Issue: 8 Linguagem: Inglês
10.15252/embr.202051910
ISSN1469-3178
AutoresEric A. Hanse, Min Pan, Wenzhu Liu, Ying Yang, Mari B. Ishak Gabra, Thai Q. Tran, Xazmin H. Lowman, Bryan Ruiz, Qiong Wang, Mei Kong,
Tópico(s)RNA modifications and cancer
ResumoArticle7 July 2021free access Source DataTransparent process The B56α subunit of PP2A is necessary for mesenchymal stem cell commitment to adipocyte Eric A Hanse Eric A Hanse orcid.org/0000-0002-1234-0198 Department of Molecular Biology and Biochemistry, University of California, Irvine, Irvine, CA, USAThese authors contributed equally to this work. Search for more papers by this author Min Pan Min Pan Department of Computational Biology, St. Jude Medical Center, Memphis, TN, USAThese authors contributed equally to this work. Search for more papers by this author Wenzhu Liu Wenzhu Liu Department of Molecular Biology and Biochemistry, University of California, Irvine, Irvine, CA, USA Search for more papers by this author Ying Yang Ying Yang Department of Molecular Biology and Biochemistry, University of California, Irvine, Irvine, CA, USA Search for more papers by this author Mari B Ishak Gabra Mari B Ishak Gabra Department of Molecular Biology and Biochemistry, University of California, Irvine, Irvine, CA, USA Search for more papers by this author Thai Q Tran Thai Q Tran orcid.org/0000-0001-7559-0525 Department of Molecular Biology and Biochemistry, University of California, Irvine, Irvine, CA, USA Search for more papers by this author Xazmin H Lowman Xazmin H Lowman Department of Molecular Biology and Biochemistry, University of California, Irvine, Irvine, CA, USA Search for more papers by this author Bryan Ruiz Bryan Ruiz Department of Molecular Biology and Biochemistry, University of California, Irvine, Irvine, CA, USA Search for more papers by this author Qiong A Wang Qiong A Wang orcid.org/0000-0003-2224-4287 Department of Molecular Endocrinology, Diabetes and Metabolism Institute, City of Hope Medical Center, Duarte, CA, USA Search for more papers by this author Mei Kong Corresponding Author Mei Kong [email protected] orcid.org/0000-0001-8139-2349 Department of Molecular Biology and Biochemistry, University of California, Irvine, Irvine, CA, USA Search for more papers by this author Eric A Hanse Eric A Hanse orcid.org/0000-0002-1234-0198 Department of Molecular Biology and Biochemistry, University of California, Irvine, Irvine, CA, USAThese authors contributed equally to this work. Search for more papers by this author Min Pan Min Pan Department of Computational Biology, St. Jude Medical Center, Memphis, TN, USAThese authors contributed equally to this work. Search for more papers by this author Wenzhu Liu Wenzhu Liu Department of Molecular Biology and Biochemistry, University of California, Irvine, Irvine, CA, USA Search for more papers by this author Ying Yang Ying Yang Department of Molecular Biology and Biochemistry, University of California, Irvine, Irvine, CA, USA Search for more papers by this author Mari B Ishak Gabra Mari B Ishak Gabra Department of Molecular Biology and Biochemistry, University of California, Irvine, Irvine, CA, USA Search for more papers by this author Thai Q Tran Thai Q Tran orcid.org/0000-0001-7559-0525 Department of Molecular Biology and Biochemistry, University of California, Irvine, Irvine, CA, USA Search for more papers by this author Xazmin H Lowman Xazmin H Lowman Department of Molecular Biology and Biochemistry, University of California, Irvine, Irvine, CA, USA Search for more papers by this author Bryan Ruiz Bryan Ruiz Department of Molecular Biology and Biochemistry, University of California, Irvine, Irvine, CA, USA Search for more papers by this author Qiong A Wang Qiong A Wang orcid.org/0000-0003-2224-4287 Department of Molecular Endocrinology, Diabetes and Metabolism Institute, City of Hope Medical Center, Duarte, CA, USA Search for more papers by this author Mei Kong Corresponding Author Mei Kong [email protected] orcid.org/0000-0001-8139-2349 Department of Molecular Biology and Biochemistry, University of California, Irvine, Irvine, CA, USA Search for more papers by this author Author Information Eric A Hanse1, Min Pan2, Wenzhu Liu1, Ying Yang1, Mari B Ishak Gabra1, Thai Q Tran1, Xazmin H Lowman1, Bryan Ruiz1, Qiong A Wang3 and Mei Kong *,1 1Department of Molecular Biology and Biochemistry, University of California, Irvine, Irvine, CA, USA 2Department of Computational Biology, St. Jude Medical Center, Memphis, TN, USA 3Department of Molecular Endocrinology, Diabetes and Metabolism Institute, City of Hope Medical Center, Duarte, CA, USA *Corresponding author. Tel: +1 949 824 5244; E-mail: [email protected] EMBO Reports (2021)22:e51910https://doi.org/10.15252/embr.202051910 PDFDownload PDF of article text and main figures. Peer ReviewDownload a summary of the editorial decision process including editorial decision letters, reviewer comments and author responses to feedback. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions Figures & Info Abstract Adipose tissue plays a major role in maintaining organismal metabolic equilibrium. Control over the fate decision from mesenchymal stem cells (MSCs) to adipocyte differentiation involves coordinated command of phosphorylation. Protein phosphatase 2A plays an important role in Wnt pathway and adipocyte development, yet how PP2A complexes actively respond to adipocyte differentiation signals and acquire specificity in the face of the promiscuous activity of its catalytic subunit remains unknown. Here, we report the PP2A phosphatase B subunit B56α is specifically induced during adipocyte differentiation and mediates PP2A to dephosphorylate GSK3β, thereby blocking Wnt activity and driving adipocyte differentiation. Using an inducible B56α knock-out mouse, we further demonstrate that B56α is essential for gonadal adipose tissue development in vivo and required for the fate decision of adipocytes over osteoblasts. Moreover, we show B56α expression is driven by the adipocyte transcription factor PPARγ thereby establishing a novel link between PPARγ signaling and Wnt blockade. Overall, our results reveal B56α is a necessary part of the machinery dictating the transition from pre-adipocyte to mature adipocyte and provide fundamental insights into how PP2A complex specifically and actively regulates unique signaling pathway in biology. SYNOPSIS PP2A phosphatase subunit B56α is specifically induced to regulate Wnt signaling during adipocyte differentiation. B56α depletion is detrimental to adipocyte development in vitro and gonadal WAT development in vivo. B56α expression is tightly regulated during adipocyte development. B56α knock-out mice have a deficiency in gonadal white adipose development. GSK3b, Axin and b-catenin are found in a complex with B56α protein. B56α binds directly to GSK3b and correlates with its phosphorylation status. B56α protein expression is controlled by the master adipogenic transcriptional activator, PPARg. Introduction The PP2A phosphatase plays an important role in a variety of tissues and cells removing phosphorylation moieties from substrate proteins during signaling cascades. As such, contextual control of PP2A activity is critical to normal cellular function. Not surprisingly, PP2A dysregulation is implicated in several diseases (Janssens & Goris, 2001; Eichhorn et al, 2009; Sangodkar et al, 2016). The PP2A holoenzyme consists of the structural scaffold A subunit, the catalytic C subunit, and the variable B subunit (Shi, 2009). The B subunit conveys specificity and directs the catalytic C subunit to its target for de-phosphorylation (reviewed in (Virshup & Shenolikar, 2009)). These fourteen B subunits are diverse in size and domain architecture. However, they are highly conserved in eukaryotes and coded by different genes scattered throughout the genome. B subunit expression is generally regulated transcriptionally and based on tissue- and context-dependent cues (Reid et al, 2013; Seshacharyulu et al, 2013). Genetic deletion of the PP2A C or other B subunits causes a variety of embryonic defects, suggesting an important role for PP2A in cellular development (reviewed in (Gotz & Schild, 2003)). Better understanding of B subunit specificity unlocks a whole new avenue of therapeutic targets with the potential to rival the success kinase inhibitors have had in the clinic. Obesity rates are rising throughout the world (Collaboration NCDRF, 2019), and understanding the complex biology involved during the development and maintenance of adipose tissue is critical as we combat the accompanying complications arising from increased body fat mass (Kusminski et al, 2016). Prolonged overnutrition initiates the recruitment of nascent pre-adipocytes from local vasculature, which in turn expands to accommodate lipid storage demands. As pre-adipocytes differentiate into adipocytes, an orchestrated intracellular signaling cascade transduces extracellular cues into phenotypic changes via a currency of phosphorylation. Elucidating the kinase and phosphatase balance during this time is critical to understanding adipocyte development. We thus became interested in exploring B subunit expression and PP2A phosphatase activity during adipocyte development. Here, we report the PP2A phosphatase plays a critical role in mediating the fate differentiation of adipocytes. Specifically, we show the PP2A-B subunit B56α is necessary for adipocyte differentiation. Using a novel mouse model, we establish B56α is necessary for the development of the gonadal white adipose tissue depot in vivo. We report that B56α is specifically induced upon adipocyte differentiation and correlates with GSK3β de-phosphorylation and Wnt signaling blockade during adipocyte development. Finally, we show B56α is a PPARγ target gene, thereby establishing a mechanism for PPARγ-driven Wnt blockade. Our results uncover a critical signaling axis during adipocyte differentiation connecting PPARγ to the control of the Wnt pathway through the relationship between PP2A and GSK3β. Results The PP2A-B subunit, B56α, is required for adipocyte differentiation in vitro The PP2A holoenzyme is directed to target substrate proteins by the specificity provided through its B subunits (Virshup & Shenolikar, 2009). The B subunits are encoded by genes across the chromosomal landscape, and their expression is context- and tissue-dependent (Gotz & Schild, 2003; Seshacharyulu et al, 2013). We initiated adipocyte development in the 3T3-L1 mouse pre-adipocyte cell line and in primary mesenchymal stem cells isolated from the outer ear of mice which differentiate to adipocytes under similar culture conditions (Rim et al, 2005). We then measured a panel of known B subunit genes for mRNA expression via RT–PCR to see whether any specific B subunits played a role during adipocyte development. We looked at day 6 when insulin maintenance has been established and adipogenic factors and morphology start to emerge. Among all the B subunits tested, we found the B56α subunit encoded by the Ppp2r5a gene was the only subunit significantly induced both transcriptionally and translationally in both 3T3-L1 and primary EMSCs, suggesting it may play a role in controlling phosphorylation during this process (Fig 1A–D). We next observed expression of the B56α protein during a time course and found B56α increased from day 2 until a maximal expression at day 6, which correlates with adipocyte differentiation markers (Figs 1E and F, and EV1A and B). Moreover, knock-down of B56α using shRNA significantly decreased the ability of 3T3-L1 and EMSCs to differentiate into adipocytes measured by Oil Red O, a stain for lipid storage (Fig 1G–J). Figure 1. The PP2A-B subunit B56α is necessary for adipocyte differentiation mRNA levels of indicated PP2A-B subunits at day 0 (blue) and day 6 (red) after induction of adipocyte differentiation in 3T3-L1 mouse pre-adipocytes. Western blot of B56α protein expression in 3T3-L1 cells at day 0 and day 6 after induction. mRNA levels of indicated PP2A-B subunits at day 0 (blue) and day 6 (red) after induction of adipocyte differentiation in mouse ear mesenchymal stem cells (EMSC). Western blot of B56α protein expression in EMSC at day 0 and day 6 after induction. Western blot of B56α expression at days indicated after induction in 3T3-L1 cells. Western blot of B56α expression at days indicated after induction in EMSCs. Western blot of B56α expression in empty vector or Ppp2r5a lentiviral shRNA-transduced 3T3-L1 cells at day 6 after induction. Oil Red O staining of empty vector or Ppp2r5a shRNA-transduced 3T3-L1 cells at day 10 following adipogenesis induction. Western blot of B56α expression in empty vector or Ppp2r5a shRNA-transduced EMSC at day 6 after induction. Oil Red O staining of empty vector or Ppp2r5a shRNA-transduced EMSC at day 10 following adipogenesis induction. **P < 0.01 as measured by paired Student's t-test. Shown are representative data from experiments performed at least three times. Source data are available online for this figure. Source Data for Figure 1 [embr202051910-sup-0004-SDataFig1.zip] Download figure Download PowerPoint Click here to expand this figure. Figure EV1. The PP2A-B subunit B56α is necessary for adipocyte differentiation A, B. mRNA expression of indicated adipocyte differentiation marker genes from (A) 3T3-L1 and (B) EMSCs at day 0 (black) and day 6 (white). Shown are representative samples from three separate experiments. **P < 0.01 as measured by paired Student's t-test. Source data are available online for this figure. Download figure Download PowerPoint The PP2A-B subunit, B56α, is required for adipocyte differentiation in vivo To examine whether B56α plays a significant role in the development of adipose tissue in vivo, we generated mice homozygous for a floxed exon 5 of the Ppp2r5a gene (Fig 2A). Primary EMSCs from these mice did not express B56α protein upon infection with a Cre recombinase expressing lentivirus, demonstrating knock-out (Fig 2B). We first attempted to cross these mice with adipose-specific Cre expressing mice Adipoq-CRE and observed no phenotypical change (Fig EV2A and B). We also crossed the Ppp2r5afl/fl mice with FABP4-Cre and again observed no impact on adipose development. These data suggest the fate decision influenced by Ppp2r5a in adipocytes occurs prior to the differentiation required for expression of Adipoq and Fabp4. To better understand the adipose differentiation process influenced by B56α, we bred Ppp2r5afl/fl mice to Rosa26Cre-ERT2 mice to produce a conditional whole-body knock-out. This model allows us to control Ppp2r5a expression prior to fat pad development. Most fat pads are established in utero or immediately after birth (Berry et al, 2013). Surprisingly, and in contrast to studies where Ppp2r5a has successfully been knocked out in a whole-body system (Janghorban et al, 2017), attempts to knock out Ppp2r5a during these time frames were lethal. These data suggest Ppp2r5a may play a role that is essential for development in adipose or other tissues, at least acutely. However, the gonadal white adipose tissue fat pad (gWAT) rapidly expands after day 21 as the mouse approaches sexual maturity (Berry et al, 2013). This allowed us a window to investigate the development of an isolated adipose depot. Thus, we injected tamoxifen at 21 days of age intraperitoneally 4 times over the course of a week into WT and Ppp2r5afl/fl mice and allowed them to mature for an additional 4 weeks (Fig 2C). As previously reported, tamoxifen treatment alone slowed adipose tissue development ((Ye et al, 2015), Fig 2D). We found mice without Ppp2r5a had a significantly smaller gWAT depot than control mice treated with tamoxifen (Fig 2D–F) and exhibited dense cellularity observed by H&E staining, particularly around the tissue edges (Fig 2G). We did not observe gross histological differences in body mass nor in tissues such as subcutaneous WAT, brown fat or liver (Fig EV2C–E). Although the gWAT tissue was smaller in Ppp2r5afl/fl mice treated with tamoxifen, average cell area was not significantly different among conditions (Fig EV2F). Moreover, adipose tissue depots established earlier in development such as subcutaneous WAT were not affected by loss of Ppp2r5a consistent with the idea that this signaling pathway is particularly important during adipose development (Fig EV2G). Taken together, these data demonstrate a critical role for the Ppp2r5a gene during the initiation of adipocyte differentiation and adipose tissue. Figure 2. Ppp2r5a is necessary for gonadal white adipose development in vivo Genotyping for the floxed allele of the Ppp2r5a gene taken F1 mice. Primary EMSCs isolated from WT and fl/fl mice were transduced with LV-Cre lentivirus and cultured in adipogenic induction medium for indicated days. Western blots showing loss of the B56α protein. Timeline and experimental strategy for gWAT development assay. Representative gWAT depots from WT and Ppp2r5a fl/fl mice. Western blot from gWAT tissue indicating loss of B56α expression. gWAT mass at 8 weeks. **P < 0.01 as measured by unpaired Student's t-test. Hematoxylin and eosin staining for gWAT from indicated mice at 10× and 20× magnification. Scale bar equal to 20 microns. Source data are available online for this figure. Source Data for Figure 2 [embr202051910-sup-0005-SDataFig2.zip] Download figure Download PowerPoint Click here to expand this figure. Figure EV2. Ppp2r5a is necessary for gonadal white adipose development in vivo Gonadal white adipose tissue percentage of total body weight for tissue-specific Adipoq-Cre 2r5afl/fl mice statistics measured by unpaired Student's t-test. H&E staining for gonadal white adipose tissue of a representative mouse from the study in (A). Scale bar equal to 20 microns. H&E staining for liver, subcutaneous white adipose tissue (SCWAT), and brown adipose tissue (BAT). Animal weight at 8 weeks. MRI measurements of percent body fat mass and percent lean. Histology of gonadal WAT was analyzed using Adiposoft through ImageJ. A minimum of six high-powered fields were analyzed from three separate mice per condition to measure adipocyte cell area. Shown are the individual measurements along with mean and standard deviation. One-way ANOVA using Dunnett's multiple comparison test was used to determine statistical significance (*P < 0.01). Percent total body weight represented by the subcutaneous fat pad. Each dot represents an individual mouse. The mean is represented by the bar. To determine statistical significance, one-way ANOVA was performed using Dunnett's multiple comparisons. Source data are available online for this figure. Download figure Download PowerPoint B56α binds to Axin1 and GSK3β during adipocyte development The canonical function for PP2A as a phosphatase led us to ask what potential substrate proteins interact with B56α during adipocyte development. To address this, we transfected a FLAG-tagged B56α construct into differentiating 3T3-L1 cells at day 6 when B56α protein was observed to be maximally expressed (Fig 1). On day 6, we collected lysates, immunoprecipitated FLAG containing protein complexes, and identified associated peptides via mass spectrometry (Fig 3A). Along with the PP2A-A and -C subunits, we found prominent members of the β-catenin destruction complex. Specifically, proteins, Axin1, GSK3β, and Dvl1 were found to associate with FLAG-B56α at day 6 (Fig 3B). We confirmed interactions with Axin1 and GSK3β via immunoprecipitation in both 3T3-L1 and EMSCs (Fig 3C and D). We were not able to confirm Dvl1 in these complexes via immunoprecipitation, however that does not preclude its involvement in this process. Likely, Dvl1's role in the complex is associative and less likely to be direct. Interestingly, we found the interaction of B56α with the β-catenin destruction complex dramatically increased upon differentiation (Fig 3C and D). Next, we looked closer at the biochemistry involved with B56α and individual members of the β-catenin destruction complex identified in the proteomics screen. Using recombinant proteins to investigate binding in vitro, we found B56α preferably binds to GSK3β (Fig 3E). Next, we further defined the biochemical basis that mediates the interaction between B56α and GSK3β. Specifically, it has been reported that B56α subunit binds to a consensus short linear sequence on interacting proteins, termed the LxxI/VxE motif (Hertz et al, 2016). Interestingly, we found that GSK3β contains a conserved LxxI/VxE motif (Fig 3F). This consensus sequence is dependent upon aspartate at position 6. Thus, we mutated E137 of GSK3β to glutamate (E137D) to minimally disturb structural integrity and found the E137D mutant showed significantly decreased binding to B56α, whereas binding to Axin was diminished but still functional (Fig 3G), suggesting LxxI/VxE motif on GSK3β is critical for the interaction between GSK3β and B56α. Figure 3. B56α protein–protein interactions during adipocyte development 3T3-L1 cells were transduced with FLAG-Ppp2r5a retrovirus and induced toward adipogenesis. At day 6, after adipogenesis induction, protein was collected and immunoprecipitation was performed. Shown is a silver stain of the precipitates. List of highly enriched peptide IDs co-precipitating with FLAG-B56α as identified via mass spectrometry. Arrows indicate members of the PP2A phosphatase and β-catenin destruction complex. Immunoprecipitation and Western blot of FLAG-B56α transfected 3T3-L1 cells at day 6 after induction. Immunoprecipitation and Western blot of EMSCs. Representatives of at least three independent experiments are shown. GST pulldown and resulting Western blot (left) and silver stain (right). Schematic of GSK3β consensus sequence and proposed mechanism of PP2A/B/C de-phosphorylation. Western blot of 293T co-expression and immunoprecipitation with GSK3β point mutation E147D. Source data are available online for this figure. Source Data for Figure 3 [embr202051910-sup-0006-SDataFig3.zip] Download figure Download PowerPoint The phosphorylation of GSK3β is dependent upon the expression of B56α GSK3β phosphorylates β-catenin initiating its destruction and repressing its transcriptional activation. Downstream of Wnt ligand engagement however, one way Wnt activity is regulated is through GSK3β inhibition. GSK3β is phosphorylated at serine 9 and inactivated allowing β-catenin to accumulate and drive transcription (van Noort et al, 2002). In the context of adipocyte development, sustained Wnt signaling or β-catenin activation pushes pre-adipocytes toward osteoblast differentiation and away from adipocytes (Kang et al, 2007; Zeve et al, 2012). Since the phosphorylation status of GSK3β could affect relay of Wnt signaling, we monitored GSK3β phosphorylation at serine 9 throughout adipocyte development in vitro. Phosphorylation at Ser9 decreases starting at day 2 and was maintained at low levels until day 6. We found this pattern of phosphorylation is inversely correlated with B56α expression (Fig 4A). The phosphatase relationship with B56α is specific to Ser9 as phosphorylation at Tyr216, a separate activation mark (Hughes et al, 1993), did not correlate with B56α protein levels. We further tested the relationship between B56α expression and ph-GSK3β Ser9 by modulating expression of the Ppp2r5a gene. In EMSCs treated with shRNA targeting Ppp2r5a mRNA, we found GSK3β phosphorylation at Ser9 was sustained throughout the differentiation timeline (Fig 4B). Conversely, we decreased GSK3β Ser9 phosphorylation in cells transfected with a plasmid directing overexpression of the Ppp2r5a gene in 3T3-L1 cells (Fig 4C). Next, we utilized the Ppp2r5a knock-out system to test the relationship between ph-GSK3β Ser9 and B56α in isolated EMSCs from Ppp2r5afl/fl mice and WT mice. Following transfection with lentiviral-Cre, we induced adipocyte differentiation and found GSK3β phosphorylation was sustained throughout the differentiation window in Ppp2r5afl/fl compared to wild type (Fig 4D). Further, the loss of B56α completely blocked these cells from differentiating into cells capable of storing lipid (Fig 4E). These results demonstrate B56α expression has a profound effect on the phosphorylation status of GSK3β during adipocyte development. Figure 4. The phosphorylation of GSK3Β is dependent upon the expression of B56α Western blots from 3T3-L1 cells and EMSCs over time after adipocyte induction. Control or Ppp2r5a lentiviral shRNA-transduced EMSCs were used for adipocyte induction. Cell lysates were collected for Western blots at day 0 and day 6 after induction. Western blot using cell lysates from B56α-FLAG overexpression retrovirus infected 3T3-L1 cells. Western blots from EMSCs isolated from wild-type and Ppp2r5a fl/fl mice. Cells were transduced with LV-Cre lentivirus and cultured in adipogenic induction medium for indicated days. Oil Red O staining of EMSCs isolated from wild-type and Ppp2r5a fl/fl mice. Cells were transduced with LV-Cre lentivirus and cultured in adipogenic induction medium for 10 days. Source data are available online for this figure. Source Data for Figure 4 [embr202051910-sup-0007-SDataFig4.zip] Download figure Download PowerPoint Loss of B56α leads to accumulation of β-catenin and expression of osteoblast markers The phosphorylation of GSK3β inactivates the kinase which should, in turn, result in increased accumulation of β-catenin and activation of canonical Wnt signaling (Salic et al, 2000). In the context of pre-adipocyte differentiation, sustained Wnt signaling could push development toward osteoblasts. Therefore, we took a closer look at Wnt signaling in the absence of Ppp2r5a. We first looked at the expression of both Wnt target genes and osteoblast differentiation genes in the absence of Ppp2r5a in EMSCs and found Wnt target genes Axin2 (Yan et al, 2001) and C-myc (He et al, 1998) were significantly increased in the absence of Ppp2r5a at day 5 (Fig 5A). The osteoblast differentiation marker and Wnt target gene Runx2 also increased significantly after differentiation (Gaur et al, 2005) indicating a potential push toward osteoblast differentiation in these cells. Moreover, we found increased expression of Wnt10b, an adipogenic inhibitor protein that has also been reported to increase upon GSK3β inhibition (Bennett et al, 2002). Our in vivo data suggest the gWAT is the primary organ affected by the loss of Ppp2r5a. To address this biology directly from the tissue, we harvested the stromal vascular fraction taken from adolescent fat pads of WT and Ppp2r5afl/fl mice and induced adipocyte differentiation. We found these cells behaved similarly to EMSCs (Fig EV3A), suggesting the biological pathways found in 3T3-L1 and EMSCs are functional in gWAT. We next asked whether loss of B56α could produce sustained Wnt signaling and inhibit adipocyte differentiation markers in vivo. In RNA isolated from gonadal WAT, we found significant increases in the Wnt and osteoblast markers Runx2, Pref1, and Wnt10b, and a coordinate decrease in the adipose differentiation markers Fabp4 and Plin1 (Fig 5B). We next used EMSCs derived from WT or Ppp2r5afl/fl mice to look for activation of the Wnt pathway using the TCF/LEF reporter TOPFlash. We found that the luciferase reporter activity of Ppp2r5afl/fl EMSCs was significantly higher than in the wild type in the absence of adipogenic stimuli. These data suggest increased activation of Wnt signaling in the absence of Ppp2r5a (Fig 5C). In mice knocked out for Ppp2r5a, we observed a dramatic increase in cell density, particularly near the edges of the tissue (Fig 2G). We used immunohistochemistry to stain for the presence of β-catenin in the gWAT collected from Ppp2r5a knock-out mice and compared the staining to wild-type samples. We found evidence of increased β-catenin staining in gWAT collected from these mice, particularly in areas of dense cellularity, which tended to be focused at the edge of the tissue (Fig 5D). Mice that constitutively express β-catenin in progenitor cells produce highly cellular, osteoblast-like tissue in the adipose compartments, including the storage of calcium (Zeve et al, 2012). Using Alizarin Red staining, we investigated for evidence of calcium accumulation in gWAT collected from Ppp2r5a knock-out mice. We found significant evidence of increased calcium in the adipose tissue, particularly near the dense cellular edges where β-catenin and active β-catenin staining was observed (Figs 5E and EV3B). Taken together, these data suggest loss of Ppp2r5a leads to increased β-catenin expression and an osteoblast phenotype in adipocytes and adipose tissue. Figure 5. Loss of B56α leads to accumulation of β-catenin and expression of osteoblast markers mRNA expression of indicated genes from EMSCs treated with LV-Cre. **P < 0.01 as measured by one way ANOVA. Shown are representatives of three independent experiments. mRNA expression of indicated genes from RNA isolated from the gWAT adipose tissue of mice at 8 weeks of age, 4 weeks after tamoxifen. **P < 0.01 as measured by unpaired Student's t-test. n = 6 mice each group. TOPFlash reporter assay in unstimulated EMSCs *P < 0.01 as measured by unpaired Student's t-test. Immunohistochemistry using an antibody targeting β-catenin in gWAT. Histology staining of gWAT using Alizarin Red to indicate calcium. Scale bar equal to 20 microns. Source data are available online for this figure. Source Data for Figure 5 [embr202051910-sup-0008-SDataFig5.zip] Download figure Download PowerPoint Click here to expand this figure. Figure EV3. Wnt signaling is increased in Ppp2r5a knock-out adipocytes mRNA expression of indicated
Referência(s)