Advertisement

The B56 Family of Protein Phosphatase 2A (PP2A) Regulatory Subunits Encodes Differentiation-induced Phosphoproteins That Target PP2A to Both Nucleus and Cytoplasm*

  • Brent McCright
    Affiliations
    Division of Molecular Biology and Genetics, Department of Oncological Sciences, the University of Utah School of Medicine, Salt Lake City, Utah 84112

    Program in Human Molecular Biology and Genetics, the University of Utah School of Medicine, Salt Lake City, Utah 84112
    Search for articles by this author
  • Ann M. Rivers
    Affiliations
    Division of Molecular Biology and Genetics, Department of Oncological Sciences, the University of Utah School of Medicine, Salt Lake City, Utah 84112

    Program in Human Molecular Biology and Genetics, the University of Utah School of Medicine, Salt Lake City, Utah 84112
    Search for articles by this author
  • Scott Audlin
    Affiliations
    Combined Program in Molecular Biology, and the University of Utah School of Medicine, Salt Lake City, Utah 84112
    Search for articles by this author
  • David M. Virshup
    Correspondence
    To whom correspondence should be addressed: Bldg. 533, Rm. 4420A, University of Utah, Salt Lake City, UT 84112. Tel.: 801-585-3408; Fax: 801-585-3501;
    Affiliations
    Division of Molecular Biology and Genetics, Department of Oncological Sciences, the University of Utah School of Medicine, Salt Lake City, Utah 84112

    Program in Human Molecular Biology and Genetics, the University of Utah School of Medicine, Salt Lake City, Utah 84112

    Department of Pediatrics, University of Utah School of Medicine, Salt Lake City, Utah 84112
    Search for articles by this author
  • Author Footnotes
    * This work was supported in part by research grants from Jason Overman Cancer Research Fund, the American Heart Association, and the American Cancer Society and by National Institutes of Health Grant R01 AI-31657 and Training Grant T32-CA09602. Grant P30 CA42014 from the National Institutes of Health subsidized the cost of oligonucleotide synthesis. The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
      Protein phosphatase 2A is a heterotrimeric protein serine/threonine phosphatase consisting of a 36-kDa catalytic C subunit, a 65-kDa structural A subunit, and a variable regulatory B subunit. The B subunits determine the substrate specificity of the enzyme. There have been three families of cellular B subunits identified to date: B55, B56 (B′), and PR72/130. We have now cloned five genes encoding human B56 isoforms. Polypeptides encoded by all but one splice variant (B56γ1) are phosphoproteins, as shown by mobility shift after treatment with alkaline phosphatase and metabolic labeling with [32P]phosphate. All labeled isoforms contain solely phosphoserine. Indirect immunofluorescence microscopy demonstrates distinct patterns of intracellular targeting by different B56 isoforms. Specifically, B56α, B56β, and B56ϵ complexed with the protein phosphatase 2A A and C subunits localize to the cytoplasm, whereas B56δ, B56γ1, and B56γ3 are concentrated in the nucleus. Two isoforms (B56β and B56δ) are highly expressed in adult brain; here we show that mRNA for these isoforms increases severalfold when neuroblastoma cell lines are induced to differentiate by retinoic acid treatment. These studies demonstrate an increasing diversity of regulatory mechanisms to control the activity of this key intracellular protein phosphatase and suggest distinct functions for isoforms targeted to different intracellular locations.

      INTRODUCTION

      Reversible protein phosphorylation is one of the major mechanisms for the regulation of cellular processes. The serine/threonine phosphorylation events catalyzed by the thousands of protein kinases encoded in the human genome are reversed by a much smaller number of protein phosphatases. Although the diverse regulation and specificity of the protein kinases (and the protein tyrosine phosphatases) are in large part determined by the sequence of their catalytic subunits, the protein serine/threonine phosphatases appear regulated largely by regulatory and targeting subunits (Hubbard and Cohen,
      • Hubbard M.J.
      • Cohen P.
      ; Mumby and Walter,
      • Mumby M.C.
      • Walter G.
      ; Mayer-Jaekel and Hemmings,
      • Mayer-Jaekel R.E.
      • Hemmings B.A.
      ; Wera and Hemmings,
      • Wera S.
      • Hemmings B.A.
      ). PP2A
      The abbreviations used are: PP2A
      protein phosphatase 2A
      HA
      hemagglutinin
      4HA
      quadruple hemagglutinin
      DMEM
      Dulbecco's modified Eagle's medium
      PAGE
      polyacrylamide gel electrophoresis
      mAb
      monoclonal antibody.
      is involved in a broad range of cellular processes, including membrane receptor desensitization, signal transduction, intermediary metabolism, transcriptional regulation and control of DNA replication, and mitosis. This diversity of PP2A function is conferred by a diversity of targeting/regulatory subunits and several levels of post-translational modification.
      PP2A exists in the cell as a heterotrimeric complex consisting of a 36-kDa catalytic C subunit, a 65-kDa structural/regulatory A subunit, and a variable regulatory B subunit. The A and C subunits are each encoded by two highly related (85 and 97% identity, respectively) and widely expressed genes. The variable B subunits identified to date are encoded by three unrelated gene families. The most prevalent isoform, denoted variously PR55, B55, or simply B, is encoded by at least three genes with several splice variants (DePaoli-Roach et al.,
      • DePaoli-Roach A.A.
      • Park I.K.
      • Cerovsky V.
      • Csortos C.
      • Durbin S.D.
      • Kuntz M.J.
      • Sitikov A.
      • Tang P.M.
      • Verin A.
      • Zolnierowicz S.
      ; Kamibayashi et al.,
      • Kamibayashi C.
      • Estes R.
      • Lickteig R.L.
      • Yang S.-I.
      • Craft C.
      • Mumby M.C.
      ). The PR72/130 regulatory subunit appears encoded by a single gene that generates two splice forms that vary in the size of their amino terminus (Hendrix et al.,
      • Hendrix P.
      • Mayer-Jaekel R.E.
      • Cron P.
      • Goris J.
      • Hofsteenge J.
      • Merlevede W.
      • Hemmings B.A.
      ). Using a two-hybrid cloning approach, we have recently described a new family of B subunits with at least three genes, which we named the B56 family (McCright and Virshup,
      • McCright B.
      • Virshup D.M.
      ). The B56 family appears to encode both the 54-kDa B′ and the 74-kDa B′/δ regulatory subunits recently cloned by several other groups (Csortos et al.,
      • Csortos C.
      • Zolnierowicz S.
      • Bakó E.
      • Durbin S.D.
      • DePaoli-Roach A.A.
      ; Tanabe et al.,
      • Tanabe O.
      • Nagase T.
      • Murakami T.
      • Nozaki H.
      • Usui H.
      • Nishito Y.
      • Hayashi H.
      • Kagamiyama H.
      • Takeda M.
      ; Tehrani et al.,
      • Tehrani M.A.
      • Mumby M.C.
      • Kamibayashi C.
      ).
      The B subunits of PP2A appear to have a number of functions. First, they contain the targeting information that directs the heterotrimer to distinct intracellular locations. Thus, Mumby and co-workers have identified an isoform of the 55-kDa B subunits, B55α, that directs PP2A to microtubules (Sontag et al.,
      • Sontag E.
      • Nunbhakdi-Craig V.
      • Bloom G.S.
      • Mumby M.C.
      ). Second, PP2A B subunits determine the substrate specificity of the enzyme. For example, a PP2A heterotrimer with a 72-kDa (PR72) regulatory subunit dephosphorylates casein kinase I sites on SV40 large T antigen, whereas PP2A with a 55-kDa regulatory B subunit dephosphorylates cyclin-dependent kinase sites (Cegielska et al.,
      • Cegielska A.
      • Shaffer S.
      • Derua R.
      • Goris J.
      • Virshup D.M.
      ). Third, tissue-specific and developmentally regulated patterns of the B subunit gene expression likely determine what substrates are dephosphorylated in specific tissues. This has been best demonstrated for the Drosophila homolog of the PR55 gene, where flies with mutant alleles demonstrate both imaginal disc and cellular (mitotic) abnormalities and have a reduced ability to dephosphorylate substrates phosphorylated by cyclin-dependent kinases (Mayer-Jaekel et al.,
      • Mayer-Jaekel R.E.
      • Ohkura H.
      • Gomes R.
      • Sunkel C.E.
      • Baumgartner S.
      • Hemmings B.A.
      • Glover D.M.
      ; Uemura et al.,
      • Uemura T.
      • Shiomi K.
      • Togashi S.
      • Takeichi M.
      ; Mayer-Jaekel et al.,
      • Mayer-Jaekel R.E.
      • Ohkura H.
      • Ferrigno P.
      • Andjelkovic N.
      • Shiomi K.
      • Uemura T.
      • Glover D.M.
      • Hemmings B.A.
      ). Lastly, the regulatory subunits of PP2A may act as receptors of second messengers. Specifically, several studies have suggested that the lipid ceramide activates PP2A, perhaps acting through the B subunit (Dobrowsky et al.,
      • Dobrowsky R.
      • Kamibayashi C.
      • Mumby M.C.
      • Hannun Y.A.
      ; Nickels and Broach,
      • Nickels J.T.
      • Broach J.R.
      ). It is of interest that at least three distinct DNA tumor viruses encode proteins that regulate PP2A activity by binding to or displacing the endogenous B subunit (Pallas et al.,
      • Pallas D.C.
      • Shahrik L.K.
      • Martin B.L.
      • Jaspers S.
      • Miller T.B.
      • Brautigan D.L.
      • Roberts T.M.
      ; Walter et al.,
      • Walter G.
      • Ruediger R.
      • Slaughter C.
      • Mumby M.
      ; Sontag et al.,
      • Sontag E.
      • Fedorov S.
      • Kamibayashi C.
      • Robbins D.
      • Cobb M.
      • Mumby M.
      ; Kleinberger and Shenk,
      • Kleinberger T.
      • Shenk T.
      ).
      In an effort to better understand the regulation of PP2A, we have further characterized the B56 family of regulatory subunits. First, two new human isoforms, B56δ and B56ϵ, have been identified, bringing the total number of human B56 genes to five. One B56 gene, which we named B56γ, has at least three splice variants (McCright and Virshup,
      • McCright B.
      • Virshup D.M.
      ; Tehrani et al.,
      • Tehrani M.A.
      • Mumby M.C.
      • Kamibayashi C.
      ). Second, we demonstrate that most of the B56 family members are phosphoproteins. Third, different isoforms have different targeting functions as illustrated by the localization of B56α, β, and ϵ to the cytoplasm and B56γ and δ to the nucleus. Fourth, retinoic acid-induced differentiation of neuroblastoma cell lines leads to a significant increase in specific B56 mRNAs as well. The B56 family is the first family of PP2A regulatory subunits shown to encode nuclear phosphoproteins; these polypeptides may play an important role in growth and differentiation.

      EXPERIMENTAL PROCEDURES

       Isolation of Full-length B56δ and B56ϵ cDNAs

      Full-length clones of B56δ and B56ϵ were isolated from a human fetal brain cDNA library in bacteriophage lambda by standard methods (Sambrook et al.,
      • Sambrook J.
      • Fritsch E.F.
      • Maniatis T.
      ). A B56δ fragment obtained in the original two-hybrid assay (McCright and Virshup,
      • McCright B.
      • Virshup D.M.
      ) corresponding to bases 580-942 of the full-length B56δ sequence was used to screen the human cDNA library for full-length clones of B56δ (GenBank™ accession number L76702). B56ϵ fragments corresponding to bases 1661-3270 and 500-1063 of the full-length B56ϵ cDNA (GenBank accession number L76703) were used to screen the cDNA library for full-length clones of B56ϵ. From 1.5 × 106 plaques, 40 B56δ clones and 24 B56ϵ clones were isolated. B56δ and B56ϵ sequencing was done on both strands by using the Sequenase 2.0 kit (U. S. Biochemical Corp.) or the University of Utah Health Sciences Core Sequencing Facility using an ABI sequencer (Perkin Elmer). Sequence analysis was performed using the Wisconsin package (Genetics Computer Group, 1994) and GenBank searches using the BLAST algorithm (Altschul et al.,
      • Altschul S.F.
      • Gish W.
      • Miller W.
      • Myers E.W.
      • Lipman D.J.
      ) on the NCBI WWW server. Gene names were assigned to the B56 isoforms by the Human Genome Nomenclature Committee as follows: B56α, PPP2R5A; B56β, PPP2R5B; B56γ, PPP2R5C; B56δ, PPP2R5D; and B56ϵ, PPP2R5E (McCright et al.,
      • McCright B.
      • Brothman A.R.
      • Virshup D.M.
      ) and are available at http://gdbwww.gdb.org/.
      Six independent library B56δ cDNA clones were sequenced across the region containing the 32-amino acid sequence present in our human B56δ clones but absent from a previously published B56δ sequence (GenBank™ accession number D78360) (Tanabe et al.,
      • Tanabe O.
      • Nagase T.
      • Murakami T.
      • Nozaki H.
      • Usui H.
      • Nishito Y.
      • Hayashi H.
      • Kagamiyama H.
      • Takeda M.
      ). All six clones were identical in this region and contained the 32-amino acid insertion. The rabbit B56δ protein sequence recently reported (Csortos et al.,
      • Csortos C.
      • Zolnierowicz S.
      • Bakó E.
      • Durbin S.D.
      • DePaoli-Roach A.A.
      ) also contains these 32 amino acids.

       Construction of Quadruple Hemagglutinin (4HA)-tagged B56 Constructs

      The protein coding sequences of B56γ1, B56γ3 (the generous gift of C. Kamibayashi), B56δ, and B56ϵ were cloned into pCEP-4/Lerner, a cytomegalovirus promoter-driven expression vector using the polymerase chain reaction primers shown in Table I as described previously (McCright and Virshup,
      • McCright B.
      • Virshup D.M.
      ). The start site ATG was omitted to prevent second site initiation. pCEP4/Lerner is based on the pCEP4 (Invitrogen) mammalian expression vector with one hemagglutinin (HA) tag at the amino terminus. HA-B56α and HA-B56β constructs previously described (McCright and Virshup,
      • McCright B.
      • Virshup D.M.
      ) and the single HA-tagged constructs described above were further modified by the addition of three more HA tags that were added in-frame at the NotI site to give the 4HA-B56 expression vectors.
      TABLE IPCR primers utilized for cloning
      B56γ forward5′-AATTATAAGCGGCCGCGTGGTGGATGCGGCCAACTCC-3′
      B56γ3 reverse5′-TAGCGGATCCCTAGCGGCCGTCCTGGGAG-3′
      T75′-GTAATACGACTCACTATA-3′
      B56δ forward5′-AATTATAAGCGGCCGCCCCTATAAACTGAAAAAGGAGAAG-3′
      B56δ reverse5′-TAGAAGATCTTCAGAGAGCCTCCTGGCTG-3′
      B56ϵ forward5′-AATTATAAGCGGCCGCTCCTCAGCACCAACAACTACTCC-3′

       Expression of 4HA-tagged B Subunits

      Testing for the expression of the 4HA-B56 constructs, the phosphatase mobility shift assay, co-immunoprecipitation, phosphoamino acid analysis, and glycerol gradient sedimentation were performed using 293 cells (human embryonic kidney cells transformed with adenovirus) transiently transfected with 4HA-B56 expression vectors. Transfections were performed in 10-cm2 dishes with 2 µg of plasmid DNA using 4 µl of lipofectamine (Life Technologies Inc.) following the manufacturer's instructions. 293 cells were grown in Dulbecco's modified Eagle's medium (DMEM; Life Technologies Inc.) plus 10% supplemented calf serum (Hyclone). Localization of 4HA-B56 by immunofluorescence was done in transiently transfected CV-1 cells (green monkey kidney cells) grown in DMEM plus 10% fetal calf serum.

       Cell Fractionation

      36 h after transfection, 293 cells were lysed in hypotonic buffer (10 mM Tris, pH 7.9, 1.5 mM MgCl2, 10 mM KCl, 1 mM phenylmethylsulfonyl fluoride, 1 mM EDTA, 1 mM EGTA, 3 µg/µl leupeptin, 3 µg/µl pepstatin, and 1 mM benzamidine), and the soluble fraction was separated from the particulate fraction by a 5-min centrifugation at 1000 × g. After the addition of (null)/1;10 volume buffer B (1.4 M KCl, 30 mM MgCl2), the soluble fraction was clarified by 15 min of centrifugation at 14,000 × g. The resulting supernatant is referred to as the soluble fraction.

       Co-immunoprecipitation

      48 h after transient transfection with the 4HA-tagged B56 constructs, 293 cells were lysed, and the soluble fraction was isolated as described above. 4HA-tagged B56 complexes were immunoprecipitated using the mouse monclonal 12CA5 (a mouse monoclonal antibody that recognizes the HA epitope, (Wilson et al.,
      • Wilson I.A.
      • Niman H.L.
      • Houghton R.A.
      • Cherenson A.R.
      • Connolly M.L.
      • Lerner R.A.
      )) and protein A-agarose. Washed immunoprecipitates were solubilized in 1 × SDS-PAGE loading buffer, separated on 11% gels, and transferred to Immobilon membrane (Millipore). The membranes were then probed with rabbit anti-PP2A-A subunit and C subunit antibodies and visualized by enhanced chemiluminescence (Amersham Corp.). Whole soluble extracts were used as a positive control, and immunoprecipitates from untransfected cells were used as a negative control. The antibodies against the PP2A A and C subunits were generated by immunization of rabbits with keyhole limpet hemocyanin-coupled synthetic peptides KYFAQEALTVLSLA (corresponding to PP2A A residues 576-589) and RRGEPHVTRRTPDYFL (corresponding to PP2Ac residues 294-309), respectively, and were affinity purified on peptide-Affi-Gel-10 columns.

       Phosphatase Assays

      Approximately 20 µg of protein from soluble fractions were treated with 20 units of alkaline phosphatase (New England Biolabs) for 30 min at 37°C. Phosphatase-treated and untreated samples were then separated by 9% SDS-PAGE and transferred to nitrocellulose. Immunoblotting was done using the anti-HA monoclonal antibody (mAb) 12CA5 and visualized by enhanced chemiluminescence (Amersham Corp.).

       Phosphoamino Acid Analysis

      Transiently transfected 293 cells were grown in DMEM with 10% calf serum for 24 h and then metabolically labeled for 6 h in 5% dialyzed calf serum, 2.5 mCi/ml H332PO4, and phosphate-free DMEM. Cells were lysed on ice in 0.2% Nonidet P-40, 200 mM NaCl, 30 mM Tris, pH 7.5, 1 mM EDTA, 3 µg/ml leupeptin, 3 µg/ml pepstatin, 1 mM phenylmethylsulfonyl fluoride, 50 µM NaVO4, 5 mM NaF, 100 nM okadaic acid, and 5 mM β-glycerol phosphate. After a 5-min 14,000 × g centrifugation, the soluble extracts containing HA-tagged proteins were immunoprecipitated with the 12CA5 mAb and protein A-agarose. The immunoprecipitates were eluted from the protein A-agarose, separated by 9% SDS-PAGE, and transferred to an Immobilon membrane. 4HA-B56 32P-labeled proteins were identified by autoradiography and isolated for acid hydrolysis (Kamps,
      • Kamps M.P.
      ). Acid hydrolysis was done in 5.7 N HCl at 110°C for 1.5 h. Hydrolyzed samples were separated by two-dimensional thin layer electrophoresis, first in pH 1.9 buffer and then pH 3.5 buffer (Boyle et al.,
      • Boyle W.J.
      • van der Geer P.
      • Hunter T.
      ). Phosphoamino acid standards were added to each sample and then stained with ninhydrin to allow identification of individual 32P-labeled amino acids.

       Glycerol Gradient Sedimentation

      Soluble fractions from 4HA-B56 expressing cells were placed on top of a 10-30% glycerol gradient prepared in a SW41 ultracentrifuge tube (Beckman). After centrifugation at 40,000 rpm for 18 h at 4°C, 19 equal fractions were collected. The fractions were then separated by 9% SDS-PAGE, transferred to nitrocellulose, and immunoblotted using mAb 12CA5. Molecular weight standards were run on separate gradients under the identical conditions. Sedimentation of molecular weight standards was determined by Bradford protein assay (Bio-Rad).

       Immunofluorescence

      CV-1 cells were plated on glass coverslips and transfected using lipofectamine. After 36 h the cells were fixed in 100% ice-cold methanol for 5 min and then washed with phosphate-buffered saline. Nonspecific sites were blocked with 0.5 M NaCl, 30 mM Tris, pH 7.5, 0.05% Tween 20, 10% calf serum, and CV-1 whole cell extract from untransfected cells. Sequential incubations with primary (mAb 12CA5) and secondary (goat anti-mouse fluorescein isothiocyanate-conjugated; Boehringer Mannheim) antibodies were performed in blocking solution at room temperature for 1.5-2.0 h each. Slides were prepared according to the manufacturer's instructions using the Anti-Fade kit (Molecular Probes Inc.).

       Northern Blots

      IMR-32 and KCN cells, two human neuroblastoma cell lines, were grown in RPMI plus 10% fetal calf serum. Subconfluent cultures were treated with 1 µM all trans-retinoic acid for 7 days. Total RNA was prepared from all trans-retinoic acid-treated cells at various time points as described (Chomczynski and Sacchi,
      • Chomczynski P.
      • Sacchi N.
      ). Northern blots were performed using 20 µg of total RNA/lane electrophoresed in 1% agarose under denaturing conditions (Sambrook et al.,
      • Sambrook J.
      • Fritsch E.F.
      • Maniatis T.
      ). Northern blots were analyzed using 32P-labeled nick-translated probes made from cDNAs encoding B56β, (bases 1048-2510, GenBank™ accession number L42374) and B56δ (bases 633-1825, GenBank™ accession number L76702). Northern blots were quantitated with a PhosphorImager (Molecular Dynamics). mRNA recovery for each sample was quantitated by probing blots for glyceraldehyde-3-phosphate dehydrogenase (Bosma and Kooistra,
      • Bosma P.J.
      • Kooistra T.
      ).

      RESULTS

       Identification and Cloning of Two New B56 Family Members, B56δ and B56ϵ

      Human B56α, B56β, and B56γ1 were previously identified as regulatory (B) subunits of PP2A using the two-hybrid interaction method (McCright and Virshup,
      • McCright B.
      • Virshup D.M.
      ). Further analysis of additional clones obtained in the two-hybrid assay revealed two clones to be single isolates of novel B56 family members, which we have denoted B56δ and B56ϵ. Full-length cDNAs were obtained for B56δ (GenBank accession number L76702) and B56ϵ (GenBank accession number L76603) by screening a human fetal brain cDNA library using partial cDNA sequences. B56ϵ and B56δ predicted amino acid sequences and a dendrogram showing their relatedness to other B56 family members appear in Fig. 1 (a and b).
      Figure thumbnail gr1
      Fig. 1a, the B56 family consists of at least five related genes. The PP2A B56 family of B subunits is encoded by at least five different genes, and the members are 71-88% identical to each other over the 400-amino acid conserved region. Human B56γ has at least three splice variants shown as B56γ1 (McCright and Virshup,
      • McCright B.
      • Virshup D.M.
      ), B56γ2, and B56γ3 (Tehrani et al.,
      • Tehrani M.A.
      • Mumby M.C.
      • Kamibayashi C.
      ). Predicted amino acid sequences were aligned using the Genetics Computer Group PILEUP program (GCG, 1994) and displayed using SeqVu (Gardner,
      • Gardner J.
      ). Amino acid identities are shaded; similarities are boxed; and gaps introduced to optimize the alignments are indicated by a dash. The potential cAMP-dependent protein kinase phosphorylation site (B56δ residues 569-572) is shown bold type and underlined. b, the B56 family can be divided into subgroups based on amino acid similarities. A dendrogram showing relative relatedness of the B56 family members was created by the PILEUP program of the Wisconsin Package (Genetics Computer Group, Madison, WI.), release 8.1. The horizontal branch lengths are proportional to the distance between the sequences.
      The human B56δ cDNA (clone D3) is 3048 bases in length and codes for a 602-amino acid protein with a predicted molecular mass of 69,947 Da. The first in-frame ATG at nucleotide 188 is in a reasonable context for translational initiation (Kozak,
      • Kozak M.
      ). There are, however, no in-frame upstream stop codons. Northern blot analysis detects a single band of B56δ mRNA that migrates with a mobility of 3100 bases (Fig. 2). Our B56δ sequence differs by a 96-nucleotide (32 amino acid) insertion beginning at residue 85 from the B56δ sequence (D78360) recently reported for the 74-kDa PP2A regulatory subunit originally purified from human erythrocytes (Tanabe et al.,
      • Tanabe O.
      • Nagase T.
      • Murakami T.
      • Nozaki H.
      • Usui H.
      • Nishito Y.
      • Hayashi H.
      • Kagamiyama H.
      • Takeda M.
      ).
      Figure thumbnail gr2
      Fig. 2The size of B56δ and B56ϵ transcripts correspond to the cDNAs obtained. A Northern blot made using total RNA from IMR-32 cells (a human neuroblastoma cell line) was probed consecutively using fragments from B56δ and B56ϵ as described under “Experimental Procedures.”
      The full-length human cDNA for B56ϵ (clone E12) is 3270 bases and codes for a 467-amino acid protein of 54,664 Da (Fig. 1). The initiator methionine at nucleotide 568 is in an optimum context for a translational start site, and there is an in-frame stop codon at nucleotide 371. By Northern blot analysis there are two B56ϵ mRNAs expressed, one approximately the same size as the E12 clone and one, expressed at much lower levels, of approximately 6800 nucleotides. (Fig. 2). Human B56ϵ is similar to a recently reported 767-base pair partial rabbit cDNA (Csortos et al. (
      • Csortos C.
      • Zolnierowicz S.
      • Bakó E.
      • Durbin S.D.
      • DePaoli-Roach A.A.
      ); and see Table II).
      TABLE IINomenclature of the B56 family of PP2A regulatory subunits
      McCright and Virshup (
      • McCright B.
      • Virshup D.M.
      ) and this study (human)
      Genome data base names (McCright, 1996)Csortos et al. (
      • Csortos C.
      • Zolnierowicz S.
      • Bakó E.
      • Durbin S.D.
      • DePaoli-Roach A.A.
      ) (rabbit)
      Tehrani et al. (
      • Tehrani M.A.
      • Mumby M.C.
      • Kamibayashi C.
      ) (human)
      Tanabe et al. (
      • Tanabe O.
      • Nagase T.
      • Murakami T.
      • Nozaki H.
      • Usui H.
      • Nishito Y.
      • Hayashi H.
      • Kagamiyama H.
      • Takeda M.
      ) (human)
      Tissue expression of mRNASize of human polypeptideTargeting function
      amino acid
      B56αPPP2R5AHeart to skeletal muscle486Cytosolic
      B56βPPP2R5BB′αBrain497cytosolic
      B56γ1PPP2R5CB′β1-4B′α3 (mouse)Heart, skeletal muscle440nuclear
      475
      514
      γ2B′α2
      γ3B′α1
      B56δPPP2R5DB′γB′ or δ (74 kDa)Brain > heart, skeletal muscle602nuclear and cytosolic
      B56ϵPPP2R5EB′δ (partial)467cytosolic
      A comparison of the B56 family members indicates that B56δ is most closely related to B56γ3 (Fig. 1, a and b). They share a 3-amino acid insertion at B56δ residue 190 and have 52% identity and 70% similarity over the last 88 amino acids, a region not conserved between other B56 family members. This sequence similarity is consistent with the phosphorylation state and distinct subcellular localization of this branch of the B56 family (see below).
      One approach to the question of whether more than five B56 genes exist is made possible by the recent proliferation of expressed sequence tags present in the data base of expressed sequence tags (Boguski et al.,
      • Boguski M.S.
      • Lowe T.M.
      • Tolstoshev C.M.
      ). BLAST searches for homologs of the five known B56 family members as March 28, 1996, has determined that for each isoform there are between eight and 40 virtually identical expressed sequence tags; using this method, no additional B56 isoforms have been identified.

       Most B56 Isoforms Are Phosphoproteins

      DNA sequences encoding full-length B56 polypeptides α, β, γ1, γ3, δ, and ϵ were cloned into a cytomegalovirus expression vector with four repeats of the hemagglutinin heptapeptide fused in frame at the amino terminus. Cells transiently transfected with these constructs all expressed 4HA-B56 polypeptides of the predicted sizes (Fig. 3a and data not shown) as assessed by immunoblotting with the 12CA5 mAb. Several of the isoforms produced broad or multiple bands on immunoblots, suggesting the possibility their electrophoretic mobility was altered by post-translational modification. To determine whether the alterations in mobility were due to phosphorylation, the hypotonic lysates from transfected cells were treated with calf intestinal alkaline phosphatase. After phosphatase treatment, B56α, B56β, and B56ϵ all condensed into single bands on a 9% SDS-PAGE gel, indicating that they are likely phosphoproteins (Fig. 3b).
      Figure thumbnail gr3
      Fig. 3B56α, B56β, B56γ3, B56δ, and B56ϵ are phosphoproteins. a, 4HA-B56 subunits are expressed in transiently transfected 293 cells. Soluble protein (20 µg) from hypotonic lysates of 293 cells transfected with the indicated expression vector were separated by 9% SDS-PAGE and analyzed by immunoblotting. HA-tagged polypeptides of the predicted size for each construct were detected by immunoblot using the 12CA5 mAb. b, B56 isoforms have altered electrophoretic mobility after alkaline phosphatase treatment. Hypotonic lysates (20 µg) from cells transfected with the indicated expression vector were incubated with 20 units of alkaline phosphatase (+) or mock treated (-). After treatment, samples were analyzed by SDS-PAGE and immunoblotting with 12CA5 mAb. c, in vivo 32P labeling of PP2A B56 subunits. Nearly confluent 293 cells in 10-cm2 dishes were transfected with the indicated expression vector and metabolically labeled for 6 h with 2.5 mCi/ml H332PO4 in phosphate-free medium. Cells were lysed as described (“Experimental Procedures”); expressed proteins were immunoprecipitated with 12CA5 mAb and protein A-agarose, separated by SDS-PAGE, electrophoretically transferred to polyvinylidene difluoride membranes, and analyzed by autoradiography. d, B56 subunits are phosphorylated on serine. Immunoprecipitated 32P-labeled B56 subunits on polyvinylidene difluoride were excised, acid hydrolyzed, and separated by two-dimensional electrophoresis. No phosphoamino acids were detectable in immunoprecipitates from nontransfected cells that were analyzed in parallel when equivalent exposure times were used.
      To further assess the nature of the phosphorylation, 293 cells transfected with the 4HA-B56 expression constructs were metabolically labeled with H332PO4, and the HA-tagged proteins were immunoprecipitated with 12CA5 mAb (Fig. 3c). All isoforms that showed a mobility shift upon treatment with alkaline phosphatase had incorporated 32P. Additionally, the δ and γ3 but not the γ1 isoform appeared to be phosphoproteins. Phosphoamino acid analysis of 32P-labeled B56α, β, γ3, δ, and ϵ all demonstrated the presence of phosphoserine (Fig. 3d). These results confirm that B56α, β, γ3, δ, and ϵ are phosphoproteins in vivo. Immunoprecipitates from cells transfected with the 4HA-B56γ1 vector showed no 32P-labeled proteins of the predicted size, suggesting B56γ1 is not phosphorylated in vivo. The lack of 32P incorporation in B56γ1 was not due to failure to express or immunoprecipitate the protein, because equivalent amounts of HA-B56γ1 and γ3 were present on immunoblots of the immunoprecipitates from 4HA-B56γ1-transfected cells (data not shown).

       B56 Family Members Form Heterotrimers with PP2A A and C Subunits

      B56 isoforms α-δ have previously been shown to form soluble heterotrimeric complexes with the A and C subunits of PP2A (McCright and Virshup,
      • McCright B.
      • Virshup D.M.
      ; Csortos et al.,
      • Csortos C.
      • Zolnierowicz S.
      • Bakó E.
      • Durbin S.D.
      • DePaoli-Roach A.A.
      ; Tehrani et al.,
      • Tehrani M.A.
      • Mumby M.C.
      • Kamibayashi C.
      ). To confirm that 1) B56ϵ also formed PP2A heterotrimers and 2) the presence of the 4HA tag and overexpression driven by the cytomegalovirus promoter did not interfere with heterotrimer formation, 4HA-B56 isoforms were transiently expressed in 293 cells. 4HA-containing complexes were immunoprecipitated with 12CA5 mAb and subjected to SDS-PAGE, followed by immunoblotting with rabbit anti-PP2A A and C antibodies. As shown in Fig. 4a, both A and C subunits co-immunoprecipitate with the HA-tagged B56 subunits, indicating that the presence of the 4HA epitope did not interfere with complex formation. To investigate whether these complexes were in fact heterotrimeric, hypotonic lysates were prepared 36 h after transfection and directly analyzed (without any intervening freeze-thaw cycles) by glycerol gradient sedimentation (Fig. 4b). Nearly 100% of the soluble 4HA-tagged B56 isoforms sedimented at essentially the same rate as the PP2A A and C subunits and at a size consistent with a PP2A heterotrimeric complex. These data suggest that 4HA-tagged transiently expressed B56 subunits are fully competent to bind to and do not overwhelm the existing quantity of PP2A A and C subunits. We note that the bulk of the A and C subunits sediment slightly slower than the 4HA-B56. We attribute this to the fact that this experiment is done on a population of transfected cells where we estimate the transfection efficiency to be ~25%. The bulk of the PP2A is therefore in a complex with endogenous B55, which we have previously shown sediments slightly more slowly than HA-tagged B56 (McCright and Virshup,
      • McCright B.
      • Virshup D.M.
      ). Of note, all of the immunoreactive PP2A A and C subunits sedimented at a velocity consistent with heterotrimer in both transfected and untransfected cells (data not shown), suggesting that no free AC heterodimer was present. We conclude that soluble 4HA-B56 polypeptides are predominantly present in PP2A heterotrimers.
      Figure thumbnail gr4
      Fig. 4a, PP2A catalytic and A subunits co-immunoprecipitate with 4HA-tagged B56. Hypotonic lysates from transiently transfected 293 cells were immunoprecipitated with the 12CA5 antibody and protein A-agarose. The immunoprecipitates were separated by 11% SDS-PAGE and immunoblotted with PP2A catalytic (C sub) and A subunit (A sub) antibodies. Immunoprecipitates from untransfected cells (No Plasmid) was used as a negative control. b, transiently expressed 4HA-B56 proteins are quantitatively present in heterotrimeric PP2A complexes. Hypotonic lysates from 293 cells transiently transfected with the indicated expression vector were analyzed by velocity sedimentation on 10-30% glycerol gradients in an SW41 rotor. Molecular weight markers were analyzed in a separate gradient. After centrifugation at 40,000 rpm for 18 h, 19 equal fractions were collected and individually analyzed by 9% SDS-PAGE and immunoblotting with 12CA5 mAb. PP2A catalytic and A subunits were detected by reprobing the B56α blot with rabbit polyclonal anti-C subunit and anti-A subunit antibodies. Control extracts from untransfected cells showed no 12CA5 immunoreactive proteins in the gradient (not shown). All fractions were analyzed on the gel; only odd-numbered fractions are labeled. BSA, bovine serum albumin.

       B56 Family Members Form PP2A Complexes That Are Located in Different Intracellular Regions

      PP2A has been found in membrane, cytosolic, and nuclear locations (Sontag et al. (
      • Sontag E.
      • Nunbhakdi-Craig V.
      • Bloom G.S.
      • Mumby M.C.
      ); Turowski et al. (
      • Turowski P.
      • Fernandez A.
      • Favre B.
      • Lamb N.J.
      • Hemmings B.A.
      ); Wera and Hemmings (
      • Wera S.
      • Hemmings B.A.
      ) and references therein). It is likely that such localization is determined in large part by the nature of the associated B subunit. The yeast homolog of the B56 family, SCS1/RST1, has been reported to be cytosolic, (Shu and Hallberg,
      • Shu Y.
      • Hallberg R.L.
      ), whereas the FLAG epitope-tagged γ2 and γ3 isoforms of Tehrani et al. (
      • Tehrani M.A.
      • Mumby M.C.
      • Kamibayashi C.
      ) (B′α2 and B′α1 by their nomenclature; see Table II) were localized to the nucleus of transfected cells. B56δ (B′) was purified from human erythrocytes, suggesting a non-nuclear localization. To assess the intracellular location of multiple B56 isoforms under uniform conditions, African green monkey kidney (CV-1) cells were transiently transfected with 4HA-B56 isoforms and subject to indirect immunofluorescence microscopy. CV-1 cells were utilized instead of 293 cells because they were more adherent to glass coverslips.
      4HA-B56α, β, and ϵ were all predominantly cytoplasmic and appeared excluded from the nucleus (Fig. 5a). Additionally, 4HA-B56α and 4HA-B56ϵ appeared concentrated in a perinuclear spot of unknown function. Overexpression of these isoforms had no gross effect on the morphology of transfected cells when compared with nontransfected neighboring cells. No staining was present in nontransfected cells, either in cells adjacent to transfected cells as demonstrated by phase contrast micrographs (Fig. 5) and in cells either mock-transfected or stained without primary antibody (data not shown). The γ1, γ3, and δ isoforms had distinctly different localization. Isoforms B56γ1 and γ3 were concentrated in the nuclei of transfected cells, with a diffuse pattern (Fig. 5b). The location of 4HA-B56δ appeared to vary with the cell cycle (Fig. 5c). It was present in both cytoplasm and nucleus in interphase cells but appeared concentrated in the nucleus in cells undergoing mitosis and cells that had recently divided.
      Figure thumbnail gr5
      Fig. 5Differential localization of distinct B56 isoforms. a, B56α, B56β, and B56ϵ are localized in the cytoplasm. CV-1 cells were transfected with 4HA-B56α (A and B), B56ϵ (C and D), or B56β expression vectors (E and F). After growth for 36 h in DMEM plus 10% fetal calf serum, cells were fixed in 100% methanol. Detection of 4HA-B56 protein was achieved using 12CA5 mAb and fluorescein isothiocyanate-conjugated goat anti-mouse secondary antibody (A, C, and E). Coverslips were mounted with an anti-fade kit (Molecular Probes Inc.), and micrographs were made at 400× magnification. B, D, and F are phase contrast micrographs of A, C, and E. Bar, 20 µm. b, B56γ1 and B56γ3 are concentrated in the nucleus. CV-1 cells were transfected with 4HA-B56γ1 (A and B) or B56γ3 expression vectors (C, D, E, and F). B, D, and F are phase contrast micrographs of A, C, and E. c, B56δ is concentrated in the nucleus during mitosis. CV-1 cells were transfected with 4HA-B56δ. A, B, C, and D shown nonmitotic cells with B56δ present in both the nucleus and the cytoplasm. E, F, G, and H show mitotic cells with B56δ present in the nucleus. I and J depict postmitotic cells with B56δ still present in the nucleus. B, D, F, H, and J are phase contrast micrographs of A, C, E, G, and I.
      Several considerations suggest these results with epitope-tagged, overexpressed proteins reflect the actual distribution of endogenous B56. First, although PP2A is known to be present in both nucleus and cytoplasm, no other PP2A B subunits have been identified that target PP2A to the nucleus. Second, velocity sedimentation results presented above (Fig. 4) indicate that the soluble HA-tagged B56 subunits are not present in excess of endogenous PP2A A and C subunits and are thus unlikely to have overwhelmed the cellular transport machinery. Third, the nuclear localization of the two tested B56γ splice variants confirm the findings of Tehrani et al. (
      • Tehrani M.A.
      • Mumby M.C.
      • Kamibayashi C.
      ) and are consistent with the high degree of sequence relatedness of the γ and δ isoforms illustrated in Fig. 1 (a and b). As demonstrated in Fig. 1, the nuclear forms δ and γ are most related to each other, whereas the cytosolic forms α, β, and ϵ fall into a separate grouping. These results support the targeting model of phosphatase localization in which phosphatases are targeted to specific cellular locations and substrates by their association with regulatory subunits. In this case, different isoforms of the B56 family appear able to target PP2A to distinct intracellular compartments.

       Retinoic Acid-induced Up-regulation of B56 Isoform Expression

      Because mRNA encoding B56 is expressed at low levels in several tissue culture cell lines relative to its expression in terminally differentiated cell types such as brain (data not shown), we asked whether induced differentiation of neuronal cells would lead to changes in B56 gene expression. IMR-32 cells are derived from a human neuroblastoma and differentiate into neuron-like cells when treated with 1 µM all-trans retinoic acid for several days (Fig. 6a). Subconfluent IMR-32 cells were treated with 1 µM all trans-retinoic acid; total RNA was then harvested at various time points. Northern blot analysis was performed to determine the effect of retinoic acid on B56 expression, normalized to glyceraldehyde-3-phosphate dehydrogenase mRNA levels. The isoforms highly expressed in brain, B56β and B56δ, have reproducible increases in mRNA expression of 3-5-fold after several days of retinoic acid treatment (Fig. 6, b and c), whereas the isoforms less highly expressed in brain, B56α, B56γ, and B56ϵ, showed a 1.4-fold or less increase in mRNA levels (data not shown). To determine whether these results were cell line-specific, a second cell line, KCN, also derived from a human neuroblastoma, was similarly analyzed. Again, B56β mRNA levels increased 4-fold 7 days after starting treatment with retinoic acid (data not shown).
      Figure thumbnail gr6
      Fig. 6Retinoic acid induced expression of neuronal isoforms of B56. a, IMR-32 cells differentiate with 1 µM retinoic acid treatment. Subconfluent IMR-32 cells were treated with 1 µM retinoic acid (Sigma). At days 0 and 5 micrographs were made at 320×. Bar, 50 µm. Cells treated with vehicle alone did not differentiate nor show increase B56 expression (data not shown). b and c, B56β and B56δ expression increases as IMR-32 cells differentiate. After treatment with 1 µM retinoic acid, total RNA was prepared from cells at various time points and analyzed by Northern blotting as described under “Experimental Procedures.” In C, expression levels were normalized to that of glyceraldehyde-3-phosphate dehydrogenase.

      DISCUSSION

      The ability of protein phosphatase 2A to dephosphorylate specific intracellular proteins depends on its activity and appropriate intracellular localization, which in turn is largely determined by the nature of its regulatory subunit. In this report we show that the newly described human B56 gene family encodes retinoic acid-inducible phosphoproteins that can differentially target PP2A to the cytoplasm and nucleus. Thus, regulation of PP2A activity may occur by post-translational modification of B56, by the targeting of PP2A heterotrimers to specific intracellular locations by association with B56, and by tissue and developmental specific expression of B56 genes.
      We also report here the cloning of an additional human B56 family member, B56ϵ, and the cloning of a more complete form of human B56δ than previously identified. This brings the total number of B56 genes to five, with human B56γ additionally having three splice variants. As a result of the rapid proliferation of B56 isoforms, the nomenclature has become quite confusing. A summary of the mammalian genes cloned to date and their names is presented in Table II.

       Potential Regulation of PP2A by Phosphorylation of B56 Regulatory Subunits

      Most of the B56 isoforms are phosphoproteins, with phosphorylation on serine, suggesting that additional levels of regulation of PP2A function may result from phosphorylation of the regulatory subunit. Previous studies have demonstrated alterations in PP2A activity upon tyrosine or serine/threonine phosphorylation of the catalytic subunit, but no post-translational modifications of the structural and regulatory subunits have previously been identified. B56γ3 and the related B56δ are serine phosphorylated, whereas the splice variant B56γ1, 67 residues shorter than B56γ3 and B56δ, is not. The 67-amino acid region present in the phosphorylated forms B56δ and B56γ3 but absent in the nonphosphorylated form B56γ1 contains only three serines, one of which is in a cAMP-dependent protein kinase consensus phosphorylation site, RRKSEL. Of note, the phosphorylation state of the B56 polypeptides does not appear to regulate their binding to the AC complex, because velocity sedimentation analysis demonstrated that B56 polypeptides of distinct electrophoretic mobilities (and thus distinct phosphorylation states) all sedimented as though they were present in PP2A heterotrimers.

       Targeting Function of the B Subunits

      Intracellular targeting of the serine/threonine phosphatases by tightly bound noncatalytic subunits appears to be a widespread mechanism for localization of activity. Thus, protein phosphatase 1 is bound to glycogen and muscle via the G subunit (Hubbard and Cohen,
      • Hubbard M.J.
      • Cohen P.
      ), and calcineurin (PP2B) is localized at postsynaptic densities by binding to AKAP79 (Coghlan et al.,
      • Coghlan V.M.
      • Perrino B.A.
      • Howard M.
      • Langeberg L.K.
      • Hicks J.B.
      • Gallatin W.M.
      • Scott J.D.
      ). It has long been appreciated that PP2A is present in both cytosolic and nuclear forms, but the subunit composition of these nuclear PP2A enzymes has remained elusive. The recent report by Tehrani et al. indicating that B56γ3 and B56γ2 encode nuclear proteins was the first indication of a PP2A regulatory subunit that is localized to the nucleus (Tehrani et al.,
      • Tehrani M.A.
      • Mumby M.C.
      • Kamibayashi C.
      ). Interestingly, the yeast homolog of B56, SCS1/RTS1, has been shown to be cytosolic (Shu and Hallberg,
      • Shu Y.
      • Hallberg R.L.
      ).
      In this report we show that B56α, B56β, and B56ϵ target PP2A holoenzyme to the cytosol, whereas B56δ, B56γ3, and B56γ1 target PP2A to the nucleus. Thus, the localization of PP2A to nucleus or cytoplasm is determined by the nature of the specific B56 isoform bound to the PP2A holoenzyme. It is interesting to note that although B56δ shares with B56γ3 a potential bipartite nuclear localization signal (KR TVETEAVQML KDIKK, starting at residue 547 of B56δ), B56γ2 (Tehrani et al.,
      • Tehrani M.A.
      • Mumby M.C.
      • Kamibayashi C.
      ) and B56γ1 (this report) both lack this sequence yet appear to be efficiently targeted to the nucleus.
      Analysis of the differential phosphorylation, alternative intracellular localization, and the sequence alignments suggests that B56 is composed of several modules. The conserved central core of the polypeptide is likely to be involved in binding to the A and C subunits of PP2A. In fact, in preliminary experiments we have detected interaction between the PP2A A subunit and a B56α polypeptide lacking the first 59 and the last 103 residues (data not shown). The amino- and carboxyl-terminal domains of the B56 family members are much more variable, and our data suggest that at a minimum, these variable regions determine the intracellular localization and phosphorylation state of the enzyme. Whether these variable regions can confer differential substrate recognition as well remains to be determined.

       Possible Functions of B56-containing PP2A in the Nucleus

      PP2A has been implicated in a variety of nuclear processes. Nuclear forms of PP2A have been implicated in the regulation of transcription by CREB and AP-1, in the dephosphorylation of p53, and in controlling the activity of the retinoblastoma protein (Alberts et al., 1993a
      • Alberts A.S.
      • Deng T.
      • Lin A.
      • Meinkoth J.L.
      • Schünthal A.
      • Mumby M.C.
      • Karin M.
      • Feramisco J.R.
      , 1993b
      • Alberts A.S.
      • Thorburn A.M.
      • Shenolikar S.
      • Mumby M.C.
      • Feramisco J.R.
      ; Wadzinski et al.,
      • Wadzinski B.E.
      • Wheat W.H.
      • Jaspers S.
      • Peruski Jr., L.F.
      • Lickteig R.L.
      • Johnson G.L.
      • Klemm D.J.
      ). One additional in vivo activity of the B56-containing nuclear forms of PP2A may be to regulate the initiation of SV40 DNA replication. Heterotrimeric forms of PP2A remove inhibitory phosphoryl groups from large T antigen and thereby activate T antigen to unwind the SV40 origin of replication (Virshup et al.,
      • Virshup D.M.
      • Russo A.A.R.
      • Kelly T.J.
      ,
      • Virshup D.M.
      • Cegielska A.
      • Russo A.
      • Kelly T.J.
      • Shaffer S.
      ; Cegielska et al.,
      • Cegielska A.
      • Shaffer S.
      • Derua R.
      • Goris J.
      • Virshup D.M.
      ). These inhibitory phosphoryl groups are rapidly turned over in the nucleus of infected cells (Scheidtmann,
      • Scheidtmann K.H.
      ). Both the PR72- and B56-containing PP2A heterotrimers can dephosphorylate these inhibitory sites in vitro (Cegielska et al.,
      • Cegielska A.
      • Shaffer S.
      • Derua R.
      • Goris J.
      • Virshup D.M.
      ) and data not shown); however, B56-containing heterotrimers are the only form to date actually shown to target PP2A to the nucleus. Thus, B56-containing heterotrimers may play a role in viral DNA replication.

       Developmental Regulation of B56 Isoforms

      The net phosphorylation state of key intracellular proteins can be regulated during development by alterations in intracellular kinase and phosphatase activity. Developmentally regulated expression of PP2A C and A genes has been shown in HL-60 cells induced to terminally differentiate with retinoic acid (Nishikawa et al.,
      • Nishikawa M.
      • Omay S.B.
      • Toyoda H.
      • Tawara I.
      • Shima H.
      • Nagao M.
      • Hemmings B.A.
      • Mumby M.C.
      • Deguchi K.
      ); in these cells, the levels of B55 mRNA remained constant. The possibility that B56 genes might change during differentiation was suggested by the observation that expression was low in several tissue culture cell lines and high in terminally differentiated tissues such as skeletal muscle and brain (McCright and Virshup (
      • McCright B.
      • Virshup D.M.
      ); and data not shown). The observed increased expression of B56β and δ during neuronal differentiation is likely to result in an increase in the intracellular concentration of B56-containing PP2A complexes and therefore an increase in phosphatase activity toward as of yet unknown substrates. Increasing B56 levels may also displace B55 from PP2A holoenzymes, much as viral tumor antigens displace endogenous B subunits during viral infection. In contrast to the effect of the viral proteins, however, which stimulate cell growth by altering PP2A activity, it seems likely that the B56 isoforms highly expressed in terminally differentiated tissues play distinctly different roles in cellular function.

      Acknowledgments

      We thank Jessica Habashi and Andrew Thorburn for assistance, members of the lab for helpful discussions, and Craig Kamibayashi for sharing the B56 cDNA clones.

      REFERENCES

        • Alberts A.S.
        • Deng T.
        • Lin A.
        • Meinkoth J.L.
        • Schünthal A.
        • Mumby M.C.
        • Karin M.
        • Feramisco J.R.
        Mol. Cell. Biol. 1993; 13 (a): 2104-2112
        • Alberts A.S.
        • Thorburn A.M.
        • Shenolikar S.
        • Mumby M.C.
        • Feramisco J.R.
        Proc. Natl. Acad. Sci. U. S. A. 1993; 90 (b): 388-392
        • Altschul S.F.
        • Gish W.
        • Miller W.
        • Myers E.W.
        • Lipman D.J.
        J. Mol. Biol. 1990; 215: 403-410
        • Boguski M.S.
        • Lowe T.M.
        • Tolstoshev C.M.
        Nat. Genet. 1993; 4: 332-333
        • Bosma P.J.
        • Kooistra T.
        J. Biol. Chem. 1991; 266: 17845-17849
        • Boyle W.J.
        • van der Geer P.
        • Hunter T.
        Methods Enzymol. 1991; 201: 110-148
        • Cegielska A.
        • Shaffer S.
        • Derua R.
        • Goris J.
        • Virshup D.M.
        Mol. Cell. Biol. 1994; 14: 4616-4623
        • Chomczynski P.
        • Sacchi N.
        Anal. Biochem. 1987; 162: 156-159
        • Coghlan V.M.
        • Perrino B.A.
        • Howard M.
        • Langeberg L.K.
        • Hicks J.B.
        • Gallatin W.M.
        • Scott J.D.
        Science. 1995; 267: 108-111
        • Csortos C.
        • Zolnierowicz S.
        • Bakó E.
        • Durbin S.D.
        • DePaoli-Roach A.A.
        J. Biol. Chem. 1996; 271: 2578-2588
        • DePaoli-Roach A.A.
        • Park I.K.
        • Cerovsky V.
        • Csortos C.
        • Durbin S.D.
        • Kuntz M.J.
        • Sitikov A.
        • Tang P.M.
        • Verin A.
        • Zolnierowicz S.
        Adv. Enzyme Regul. 1994; 34: 199-224
        • Dobrowsky R.
        • Kamibayashi C.
        • Mumby M.C.
        • Hannun Y.A.
        J. Biol. Chem. 1993; 268: 15523-15530
        • Gardner J.
        SeqVu 1.0. Garvan Institute of Medical Research, Sydney, Australia1995
        • Hendrix P.
        • Mayer-Jaekel R.E.
        • Cron P.
        • Goris J.
        • Hofsteenge J.
        • Merlevede W.
        • Hemmings B.A.
        J. Biol. Chem. 1993; 268: 15267-15276
        • Hubbard M.J.
        • Cohen P.
        Trends Biol. Sci. 1993; 18: 172-177
        • Kamibayashi C.
        • Estes R.
        • Lickteig R.L.
        • Yang S.-I.
        • Craft C.
        • Mumby M.C.
        J. Biol. Chem. 1994; 269: 20139-20148
        • Kamps M.P.
        Methods Enzymol. 1991; 201: 21-27
        • Kleinberger T.
        • Shenk T.
        J. Virol. 1994; 67: 7556-7560
        • Kozak M.
        Annu. Rev. Cell Biol. 1992; 8: 197-225
        • Mayer-Jaekel R.E.
        • Hemmings B.A.
        Trends Cell Biol. 1994; 4: 287-291
        • Mayer-Jaekel R.E.
        • Ohkura H.
        • Gomes R.
        • Sunkel C.E.
        • Baumgartner S.
        • Hemmings B.A.
        • Glover D.M.
        Cell. 1993; 72: 621-633
        • Mayer-Jaekel R.E.
        • Ohkura H.
        • Ferrigno P.
        • Andjelkovic N.
        • Shiomi K.
        • Uemura T.
        • Glover D.M.
        • Hemmings B.A.
        J. Cell Sci. 1994; 107: 2609-2616
        • McCright B.
        • Virshup D.M.
        J. Biol. Chem. 1995; 270: 26123-26128
        • McCright B.
        • Brothman A.R.
        • Virshup D.M.
        Genomics. 1996; (in press)
        • Mumby M.C.
        • Walter G.
        Physiol. Rev. 1993; 73: 673-699
        • Nickels J.T.
        • Broach J.R.
        Genes & Dev. 1996; 10: 382-394
        • Nishikawa M.
        • Omay S.B.
        • Toyoda H.
        • Tawara I.
        • Shima H.
        • Nagao M.
        • Hemmings B.A.
        • Mumby M.C.
        • Deguchi K.
        Cancer Res. 1994; 54: 4879-4884
        • Pallas D.C.
        • Shahrik L.K.
        • Martin B.L.
        • Jaspers S.
        • Miller T.B.
        • Brautigan D.L.
        • Roberts T.M.
        Cell. 1990; 60: 167-176
        • Sambrook J.
        • Fritsch E.F.
        • Maniatis T.
        Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY1989
        • Scheidtmann K.H.
        Virology. 1986; 150: 85-95
        • Shu Y.
        • Hallberg R.L.
        Mol. Cell. Biol. 1995; 15: 5618-5626
        • Sontag E.
        • Fedorov S.
        • Kamibayashi C.
        • Robbins D.
        • Cobb M.
        • Mumby M.
        Cell. 1993; 75: 887-897
        • Sontag E.
        • Nunbhakdi-Craig V.
        • Bloom G.S.
        • Mumby M.C.
        J. Cell Biol. 1995; 128: 1131-1144
        • Tanabe O.
        • Nagase T.
        • Murakami T.
        • Nozaki H.
        • Usui H.
        • Nishito Y.
        • Hayashi H.
        • Kagamiyama H.
        • Takeda M.
        FEBS Lett. 1996; 379: 107-111
        • Tehrani M.A.
        • Mumby M.C.
        • Kamibayashi C.
        J. Biol. Chem. 1996; 271: 5164-5170
        • Turowski P.
        • Fernandez A.
        • Favre B.
        • Lamb N.J.
        • Hemmings B.A.
        J. Cell Biol. 1995; 129: 397-410
        • Uemura T.
        • Shiomi K.
        • Togashi S.
        • Takeichi M.
        Genes & Dev. 1993; 7: 429-440
        • Virshup D.M.
        • Russo A.A.R.
        • Kelly T.J.
        Mol. Cell. Biol. 1992; 12: 4883-4895
        • Virshup D.M.
        • Cegielska A.
        • Russo A.
        • Kelly T.J.
        • Shaffer S.
        Adv. Protein Phosphatases. 1993; 7: 271-293
        • Wadzinski B.E.
        • Wheat W.H.
        • Jaspers S.
        • Peruski Jr., L.F.
        • Lickteig R.L.
        • Johnson G.L.
        • Klemm D.J.
        Mol. Cell. Biol. 1993; 13: 2822-2834
        • Walter G.
        • Ruediger R.
        • Slaughter C.
        • Mumby M.
        Proc. Natl. Acad. Sci. U. S. A. 1990; 87: 2521-2525
        • Wera S.
        • Hemmings B.A.
        Biochem. J. 1995; 311: 17-29
        • Wilson I.A.
        • Niman H.L.
        • Houghton R.A.
        • Cherenson A.R.
        • Connolly M.L.
        • Lerner R.A.
        Cell. 1984; 37: 767-778