To identify the proteins of the B¢¢ family in Xenopus laevis oocytes, a prophase Xenopus oocyte cDNA library was screened using human PR130 cDNA as a probe.. By cloning the cDNA from a h
Trang 1Identification and characterization of B¢¢-subunits of protein
Evidence for an independent N-terminal splice variant of PR130
and an extended human PR48 protein
Ilse Stevens1, Veerle Janssens1, Ellen Martens1, Stephen Dilworth2, Jozef Goris1and Christine Van Hoof1
1
Afdeling Biochemie, Faculteit Geneeskunde, Katholieke Universiteit Leuven, Leuven, Belgium;2Department of Metabolic
Medicine, Imperial College Faculty of Medicine, Hammersmith Hospital, London, UK
Protein phosphatase 2A is a phosphoserine/threonine
phosphatase implicated in many cellular processes The core
enzyme comprises a catalytic and a PR65/A-subunit The
substrate specificity and subcellular localization are
deter-mined by a third regulatory B-subunit (PR55/B, PR61/B¢
and PR72/130/B¢¢) To identify the proteins of the B¢¢ family
in Xenopus laevis oocytes, a prophase Xenopus oocyte cDNA
library was screened using human PR130 cDNA as a probe
Three different classes of cDNAs were isolated One class is
very similar to human PR130 and is probably the Xenopus
orthologue of PR130 (XPR130) A second class of clones
(XN73) is identical to the N-terminal part of XPR130 but
ends a few amino acids downstream of the putative splicing
site of PR130 To investigate how this occurs, the genomic
structure of the human PR130 gene was determined This
novel protein does not act as a PP2A subunit but might compete with the function of PR130 The third set of clones (XPR70) is very similar to human PR48 but has an N-ter-minal extension Further analysis of the human EST-data-base and the human PR48 gene structure, revealed that the human PR48 clone published is incomplete The Xenopus orthologue of PR48 encodes a protein of 70 kDa which like the XPR130, interacts with the A-subunit in GST pull-down assays XPR70 is ubiquitously expressed in adult tissues and oocytes whereas expression of XPR130 is very low in brain and oocytes Expression of XN73 mainly parallels XPR130 with the exception of the brain
Keywords: phosphatase; PP2A; Xenopus laevis; cell cycle
Reversible protein phosphorylation is a key mechanism that
regulates a number of different cellular events The
phos-phorylation state of a protein is dependent on the opposing
activities of kinases and phosphatases Among the
phos-phatases, protein phosphatase 2A is a major
serine/threo-nine protein phosphatase involved in many cellular events,
such as signal transduction, DNA replication, transcription,
translation, apoptosis and the cell cycle [reviewed in 1,2]
The core enzyme of PP2A is a heterodimer consisting of a
catalytic (C) and structural (PR65/A) subunit Different
mechanisms are known to regulate the phosphatase activity
The PR65/A subunit serves as a scaffolding protein that interacts with a third regulatory subunit (B-subunit) There
is evidence that PP2A is an obligate trimer in vivo [3,4], although other evidence suggests that a substantial propor-tion of PP2A can also exist in vivo as a dimer [5] The association of a Bsubunit with the AC core dimer can result
in altered substrate specificity, catalytic activity and subcel-lular localization To date, three different B-subunit families have been described in eukaryotes: PR55/B, PR61/B¢ and PR72/130/B¢¢ Each family of B-subunits harbours different genes, some giving rise to different splice variants PP2A activity is also subject to regulation by post-translational modifications, such as methylation Methylation occurs on the carboxyl group of the C-terminal Leu309 by a specific methyltransferase [6,7] and is reversible by a methylesterase [8–10] There is evidence that methylation is required for the association of certain regulatory B-subunits with the AC heterodimer [11–13]
PP2A has been shown to be required at various stages of the eukaryotic cell cycle [reviewed in 1] Xenopus laevis oocytes are often used as a model system to study specific aspects of the cell cycle Oocytes taken from the ovary of the frog are arrested in prophase of the first meiotic division Addition of progesterone results in meiotic progression and
a second arrest in metaphase of the next meiotic division At this stage, oocytes are also called eggs and are ready to be fertilized The whole process is called oocyte maturation Mitotic [14] and DNA replicating events [15] can be studied
Correspondence to J Goris, K.U Leuven, Faculteit Geneeskunde,
Gasthuisberg O/N, Afdeling Biochemie, Herestraat 49,
B-3000 Leuven, Belgium.
Fax: + 32 16 345995, Tel.: + 32 16 345794,
E-mail: jozef.goris@med.kuleuven.ac.be
Abbreviations: PP2A, protein phosphatase type 2 A; OA, okadaic
acid; CDK1, cyclin dependent kinase 1; GVBD, germinal vesicle
breakdown; PP1, protein phosphatase 1; GST, glutathione
S-transferase; IVTT, in vitro transcription-translation.
Note: The reported nucleotide sequence data of the B¢¢ subunits of
PP2A are available in the DDBJ/EMBL/GenBank under the accession
numbers: AY043260 (Xenopus PR130), AY043259 (Xenopus PR70),
AY043258 (Xenopus N73) and BK000521 (Human PR70).
(Received 27 August 2002, revised 14 November 2002,
accepted 25 November 2002)
Trang 2easily in specific egg extracts and X laevis is also a useful
model to study embryogenesis, as all the material required
for early development are accumulated in the oocyte A role
for PP2A in the cell cycle was initially suggested through
experiments using okadaic acid (OA), an inhibitor of PP1
and PP2A Induction of meiotic oocyte maturation
occurred after injection of OA into Xenopus [16] or starfish
oocytes [17] This inhibitory activity of PP2A is shown to be
due to the presence of a trimeric form of PP2A containing
the PR55/B-subunit [18] Furthermore, immuno-depletion
of PP2A from Xenopus egg extracts resulted in strong
inhibition of chromosomal DNA replication [19]
Currently, we are investigating the role of PP2A during
the cell cycle using X laevis oocytes as model system To
achieve this goal, an inventory of the different PP2A
subunits from X laevis oocytes is required At the moment,
two PP2AC(a and b) [20], one PR65/A [21,22], one PR55/B
[22] and two PR61/B¢ subunits [23,24] have been cloned
from X laevis It has been shown that the PR65/A subunit
in oocytes is exclusively of the b-type, whereas the PR55/Bis
most similar to the human PR55a isoform [22] However,
nothing is currently known about the X laevis PR72/130/
B¢¢ family of PP2A subunits As the Bsubunits are
important components of the control of PP2A activity, it
is essential that the complement of Bsubunits expressed in
the oocyte is defined In mammalia, three different B¢¢
family members have been identified: human PR72/130 [25],
mouse PR59 [26] and human PR48 [27] PR72 was initially
identified as the third subunit of a novel trimeric form of
PP2A from rabbit skeletal muscle [28] By cloning the
cDNA from a human heart cDNA library, a second cDNA
was isolated coding for a protein of 130 kDa with the same
deduced C-terminal protein sequence as PR72 but a
different N-terminal region PR72 is specifically expressed
in skeletal muscle and heart, while PR130 is ubiquitous The
function of PR72 and PR130 is currently not known PR59
[26] and PR48 [27] were identified in a yeast two-hybrid
assay as proteins interacting with p107 (a relative of the
retinoblastoma protein) and cdc6, respectively Both
pro-teins were found to show high sequence identity with the
PR72 protein especially in their central part They have both
been found to be involved in DNA replication and the G1/S
phase transition of the cell cycle
It has been suggested that PP2A may be a tumour
suppressor gene, as inactivation of the regulatory PR65/A
subunit by gene mutation [29] or ablation [30,31] occurs
frequently in human tumours Moreover, the cellular target
for the tumour promoter okadaic acid has been shown to be
PP2A (and its relatives) In spinocerebellar ataxia, a
neurodegenerative disorder, an expanded CAG repeat has
been identified immediately upstream of the PR55b gene
that might affect PR55b expression and might be implicated
in the aetiology of the disease [32] To evaluate the role of
the PP2A as a putative tumour suppressor gene it is
important that the genomic structures of the PP2A subunits
are defined
To identify all of the proteins of the PR72/130/B¢¢ family
in oocytes, a prophase Xenopus oocyte cDNA library was
screened using a full-length human PR130 cDNA as a
probe Three different classes of cDNAs were isolated and
analysed The first is the Xenopus orthologue of PR130,
the second proved to be a novel splice variant and the third is
the Xenopus orthologue of PR48 Moreover, analysis of the structure of the latter made it possible to improve the current knowledge of the mammalian PR48 cDNA To extend these findings, the genomic structures of the human PR130 and PR48 genes (PPP2R3A and PPP2R3B, according to the human gene nomenclature) were also analysed Together the data illustrate the complexity of PP2A regulation by a still growing list of B-type regulatory subunits
Materials and methods
cDNA library screening and sequencing
A Xenopus oocyte kgt10 library [33] was screened using full-length human PR130 cDNA as the probe This specific probe was obtained by performing PCR using Pwo polymerase (Roche), the start primer (5¢-TGGGTCGAC TATGGCAGCAACTTACAGACTTGTGG-3¢) and the
GATTGAAGC-3¢) and human PR130 cDNA (accession number L12146) as template The PCR product was subsequently labelled with a random priming DNA label-ling kit (Roche) One million plaques were analysed, positives were isolated and the plaques purified as described previously [34] Isolated clones were digested with EcoRI and cloned into the pBluescript SK– vector Nucleotide sequences were determined with a Thermo Sequenase fluorescent labelled primer cycle sequencing kit and the A.L.F Express Sequenator (Amersham-Pharmacia Bio-tech) In addition to the original clones, we used subclones obtained by restriction enzyme digestion as templates for DNA sequencing
5¢-RACE technique 5¢-RACE reactions were performed on about 1.5 lg of total RNA of prophase X laevis oocytes with 1 U of AMV-reverse transcriptase (5¢/3¢RACE kit, Roche) For the first strand cDNA synthesis, to amplify 5¢ ends of XN73, the primer 5¢-GACCACTGCTTTCTCCTCCACC-3¢ was used A nested primer (5¢-GATTTCCTGTTAAAAGA GTCAGGACGG-3¢) was used together with the oligo(dT)-anchor primer for amplification with Pwo polymerase (Roche) Subsequently, two other consecutive PCR reac-tions were performed, each time using a nested primer (5¢-ACACTCCTCACCACCAACTACC-3¢ for PCR 2 and 5¢-CTCTCCACCTATTATTTTATGAGTAACATCC-3¢ for PCR 3) and the anchor primer supplied in the 5¢-RACE kit For amplification of specific XPR130 5¢-end RACE fragments, the primer 5¢-TGGACTGACAGTTGGAG TAAGG-3¢ was used for the RT reaction, and the primer
together with the oligo(dT)-anchor primer in the first PCR reaction and the primers 5¢-TCTCCAAATCCAACAGC AATTT-3¢ and 5¢-GACCACTGCTTTCTCCTCCACC-3¢ were used for the two consecutive PCR reactions together with the anchor primer Similar 5¢-RACE reactions were performed to amplify the 5¢ ends of XPR48 The primer 5¢-TCATTGGACTTCCATAAGAACCAGG-3¢ was used for the RT reaction and the primers 5¢-GGAGACAGT GACAAGTTATCTAGTGCT-3¢ and 5¢-TCATTGAG GAGCAACTGTGTTGG-3¢ were used for PCR 1 and 2,
Trang 3respectively PCR products were cloned into the pBluescript
II SK–vector and sequenced as described
In vitro transcription and translation
Full-length cDNAs of the different B¢¢-subunits were
obtained using conventional cloning techniques with the
appropriate restriction enzymes Truncations of XPR48
cDNA were generated by PCR using Pwo polymerase
(Roche) with different start primers and the specific stop
primer The resulting cDNA fragments were ligated into the
pBluescript SK–vector GST- and His6-tagged-XN73 were
generated by cloning first the full-length cDNA into
the pGex-vector (Amersham-Pharmacia Biotech) and the
pRSET vector (invitrogen), respectively Subsequently, the
cDNAs with the GST and His6-tags were further cloned
into the pBluescript SK– vector [35S]Methionine-labelled
translation products were obtained from these templates
using the TNT-coupled reticulocyte lysate system
(Pro-mega) and the appropriate RNA polymerase (T3 and T7)
The translation products were subjected to SDS/PAGE
(12%, w/v) and visualised by autoradiography
Sequence analysis
Sequence comparisons and alignments were performed
using the BLAST tool of NCBI (http://www.ncbi.nlm.nih
gov/blast/) and the MULTALIN alignment program (http://
prodes.toulouse.inra.fr/multalin/multalin.html) Human
genome sequences were obtained by doingBLASTsearches
against the HTGS database and the human genome site
(http://www.ncbi.nlm.nih.gov/genome/guide/human/)
Monoclonal antibody isolation
cDNAs encoding full length human PP2A PR65a/A or
Ca-subunit were inserted into the pQE 30 bacterial
expres-sion vector (Qiagen) to generate a fuexpres-sion protein containing
a His6-tag at the N-terminus of the inserted sequence
PR65/A was expressed and isolated by the methods
described by the vector manufacturer, whereas the
recom-binant C-subunit was insoluble After isolation of the
inclusion bodies, the protein was solubilized in 0.05% (w/v)
SDS Young Balb/c· CBA F1 crossed mice were
immu-nised by subcutaneous injection with 100 lg of fusion
protein emulsified with an equal volume of Titremax Gold
(CytRx Corporation) The immunisation was repeated
three times at 2- to 3-monthly intervals, then the mice
rested for six months A further 100 lg of fusion protein in
NaCl/Piwas injected intraperitoneally at 6 and 3 days prior
to sacrifice All animal experiments were conducted under
the guidelines and regulations contained in the UK Animal
Scientific Procedures Act, 1986 Hybridoma lines were then
established by fusing splenocytes from the immunised
animal with the myeloma line Sp2/0-Ag14 by poly(ethylene
glycol) treatment using conventional procedures [35]
Indi-vidual wells were screened for antibody production by a
modified dot blot procedure using bacterially expressed
fusion protein [36], and single cells cloned a minimum of
three times before being grown for antibody isolation Each
monoclonal antibody was screened for specificity by
West-ern blotting against total protein from SDS derived tissue
culture cell lysates Line C5 1G7 was found to be specific for the PP2A PR65/A subunit (data not shown), in both human and Xenopus lysates, which recognises the PR65 a and b isoforms equally well The best anti-C monoclonal was found to be F2 6A10, and it recognizes PP2ACin rabbit, mouse and Xenopus oocytes and tissues
GST-pulldown assays and Western blotting cDNAs of the different B¢¢-subunits (XPR130, XN73 and XPR70) without UTR sequences were first cloned into a pGex-vector (Amersham-Pharmacia Biotech) The corres-ponding GST-fusions and GST alone were subsequently cloned between the UTR sequences of b-globin into the pGem high expression vector [37] Capped synthetic mRNA transcripts were synthesised by using T7 RNA polymerase according to the manufacturer’s instructions (mCAP RNA capping kit, Stratagene) Fifty nL of about 0.5 lgÆlL)1of this synthetic mRNA solution was microinjected into fully grown stage VI oocytes prepared as described [38] Oocytes were incubated for 24 h at 18C After incubation,
60 oocytes were homogenized into 150 lL of NaCl/Tris containing 30 lgÆmL)1 leupeptin and 0.2 mM phenyl-methylsulfonyl fluoride and a high speed centrifugation was performed (3 min; 150 000 g; Beckman air driven ultracentrifuge, rotor A-110/18) The supernatant was diluted five times in the homogenization buffer containing
1 mgÆmL)1 bovine serum albumin and glutathione–Seph-arose beads (Amersham-Pharmacia Biotech) were added After incubation overnight at 4C, beads were washed three times with homogenization buffer containing 0.1% (v/ v) Tween-20 and 0.5M NaCl The glutathione–Sepharose beads were further eluted by the addition of SDS/PAGE sample buffer and boiled for 5 min Western blots were performed and developed as described in [39] using mono-clonal antibodies against human PR65 and an enhanced chemiluminescence detection system (ECL, Amersham-Pharmacia Biotech)
Semi-quantitative RT-PCR and Southern blotting Prophase oocytes were prepared as described [38] Meta-phase oocytes were derived from proMeta-phase oocytes after addition of 3 lMprogesterone and overnight incubation of the oocytes at 18C Several Xenopus adult tissues (liver, gall bladder, spleen, heart, muscle and brain) were excised from adult female frogs and immediately freeze clamped in liquid nitrogen Total RNA from prophase oocytes and adult tissues was prepared as described [40] RNA was further purified using a mini Qiagen RNA purification kit (Westburg) cDNA was prepared from 1.5 lg of total RNA using the primer 5¢-TGGACTGACAGTTGGAGTAA GG-3¢ for XPR130, the stop primer 5¢-GGAATTCT AACTCTCGGTTAAGGGGTA-3¢ for XN73 or the stop
for XPR70, and 30 U of M-MuLV reverse transcriptase (Fermentas) Subsequently, the cDNA was purified using a GFX DNA purification kit (Amersham-Pharmacia Bio-tech) and PCR was performed in 50 lL using 1 U Pwo-DNA polymerase (Roche), PCR buffer, 2.5 mM MgCl2,
200 lM of each dNTP, 50 pmol primer The annealing
of the primers was carried out at 55C However, to be
Trang 4in the linear DNA amplification range, only 20
amplifi-cation cycles were performed with the primers 5¢-GTTCC
TGAGCAAAGTCTTCAATG-3¢ and 5¢-GGAATTCAG
TGGAAAAACTTTACAT-3¢ for XPR130, the primers
5¢-GTTCCTGAGCAAAGTCTTCAATG-3¢ and 5¢-GG
AATTCAGTGGAAAAACTTTACAT-3¢ for XN73, and
the primers 5¢-GGAATTCATGCCGACCACAACCGTT
TTAAG-3¢ and 5¢-GGAGACAGTGACAAGTTATCTA
GTGCT-3¢ for XPR70 PCR products were resolved on a
1.5% (w/v) agarose gel and transferred to a Hybond-nylon
membrane (Biorad) by capillary blotting in 2· NaCl/Cit
Pre-hybridization and hybridization was done at 60C in
6· NaCl/Cit, 2 · Denhardt’s reagent, 0.1% (w/v) SDS,
0.05% (w/v) sodium pyrophosphate and 100 lgÆmL)1
sal-mon sperm DNA with the primer 5¢-TCTCCAAATCCAA
CAGCAATTT-3¢ for XPR130, the primer 5¢-TCTCCA
AATCCAACAGCAATTT-3¢ for XN73 or the primer
5¢-TCATTGAGGAGCAACTGTGTTGG-3¢ for XPR70
Probes were labelled with T4-polynucleotide kinase
(Fermentas) in the presence of [c32P]ATP Membranes were
washed four times for 10 min at 60C with 6 · NaCl/Cit
and 0.5% (w/v) SDS, and the PCR-DNA fragments were
visualised using a phosphoimager-intensifying screen
Results
cDNA cloning of PR72/130/B¢¢ subunits fromXenopus
laevis prophase oocytes
To clone the PR72/130/B¢¢ subunits of PP2A from Xenopus
laevis, a prophase Xenopus oocytes kgt10 library was
screened using a full-length human PR130 cDNA as a
probe Out of one million plaques, 27 positive clones were
picked and analysed These clones could be divided into
three different groups (Fig 1) The first group (eight clones)
were highly homologous to human PR130 (Fig 1A)
Unfortunately, only partial clones were isolated, because
even the longest clone lacks information about 142 amino
acids at the 5¢-end when compared with human PR130, and
only one cDNA contained the stop codon and a
polyade-nylation signal The second group (four clones) could
encode a protein homologous to the N-terminal 73 kDa
part of human PR130 and were therefore designated as
XN73 The open reading frame ended with a stop codon
21 bp downstream of a putative splicing site, which would
translate a protein with an additional stretch of seven amino
acids that are not present in either human PR130 nor human
PR72 The stop codon was followed by several
polyadeny-lation signals (one AATAAA and two ATTAAA motifs)
and a poly(A) tail (Fig 1B) There was no homology found
between the 3¢-UTR of XN73 and that of XPR130 At the 5¢
end, the longest clone isolated lacked sequences encoding 17
amino acids in comparison with human PR130, so was
probably not complete The third group contained 15 clones,
all of which were homologous to human PR48 (Fig 1C)
Because the XPR130 and XN73 cDNAs isolated were
apparently only partial clones missing the 5¢ end, we used a
5¢-RACE technique to generate full-length cDNAs Using
total RNA of prophase Xenopus oocytes as the template, we
were able to isolate DNA fragments after RT-PCR and
several nested PCR attempts To ensure that we could
generate a PCR fragment long enough to isolate the specific
5¢ ends of XN73, a primer close to the 5¢ end of the clone isolated from the library (Fig 1B) was used in the RT reaction The longest RACE fragment obtained contained
an ATG start codon at the same position as in human PR130 When we used specific primers to amplify only XPR130, we obtained RACE fragments similar to XPR130 and XN73 clones, but we could not amplify a RACE fragment that contains the ATG start codon The predicted amino acid sequence of the N-terminal region of XPR130 and XN73 are different in several positions (Fig 2A) This would normally indicate that these proteins are encoded by different genes However, the genome of Xenopus laevis has been duplicated during evolution and so has tetraploid characteristics [41] Because similar variations were found in the clones isolated from the oocyte library as well as in the PCR RACE fragments, it is likely that XPR130 and XN73 are derived from the same gene and represent splice variants However,
as genome sequences of Xenopus available in the GenBank database are still rare, the correctness of this statement cannot be verified The full-length cDNA of XPR130 and XN73 were then used in an in vitro transcription–translation reaction, and the products separated by SDS/PAGE Proteins of about 130 kDa and 50 kDa, respectively, were obtained (Fig 2B) Since XN73 runs in SDS/PAGE with an apparent molecular mass of about 50 kDa instead of the calculated 72 964 Da, we verified the correctness of the expected initiation methionine by translating the fusion protein GST-XN73 and His6-XN73 As can be seen in Fig 2B, these fusion proteins run at the expected height, proving that no internal methionine was used as initiation start in the translation of the wild type mRNA of XN73 and that the XN73 protein runs with an aberrant molecular mass
in SDS/PAGE It was previously proven that also PR72, as isolated from rabbit skeletal muscle runs with an aberrant molecular mass (about 72 kDa) in SDS/PAGE, different from its calculated molecular mass (61 096 Da) In this case the deviation was in the opposite direction [25] The underlying physical cause is not known, but it is possible that both aberrations compensate in the PR130 protein, running in SDS/PAGE at its expected height
Together, the data suggest that the group one clones completed by 5¢-RACE techniques represent the Xenopus orthologue of human PR130 (51% amino acid identity) The group two clones encode a protein similar to the N-terminal region of HPR130 and almost identical to the N-terminal part of XPR130, but which differs completely at the C-terminal end, where it abruptly stops seven amino acids downstream of a putative splicing site If produced, this protein would be a new protein never described before
in other species An alignment of the amino acid sequences
of human PR130 with Xenopus PR130 and Xenopus N73 is shown in Fig 2A cDNA sequences were submitted to the EMBL GenBank database and have accession numbers AY043260 for XPR130 and AY043258 for XN73
Genomic structure of human PR130 favours the possibility that XN73 is a splice variant The results obtained from Xenopus cDNAs suggest the existence of a new potential PP2A B¢¢ subunit which we have designated as XN73 To determine whether this protein could exist in other species, we first performed a
Trang 5search with XN73 cDNA as bait This revealed no similar
EST-clones in other species We next examined the PR130
gene structure in the human genome database A blast
search of human PR130 (accession number L12146) against
the human genome NCBI-site revealed a perfect match with
a finished genomic contig (NT_005567.9), designed PPP2R3
A in the human gene nomenclature The complete gene
spans approximately 375.7 kbp and the human PR130
cDNA consists of 14 exons varying in size between 49 (exon 9) and 2849 nucleotides (exon 14) The splice sites and exon-intron boundaries are presented in Table 1 All splice donor and acceptor sites follow the canonical GT/AG rule [42] A detailedMAPVIEWanalysis placed the gene PPP2R3A at the chromosomal locus 3q22.3
Detailed analysis of the exon–intron boundaries revealed that a large part of the 5¢-UTR and the complete N-terminal
Fig 1 Screening of a prophase Xenopus oocyte kgt10 library identified three classes of clones (A) Class I: eight clones were isolated that were homologous to human PR130 All clones were partial as even the longest clone lacks about 142 amino acids at the N-terminal part compared to human PR130 Arrows shown on the longest clone represent 5¢-RACE strategy of XPR130 as described in materials and methods (B) Class II: four clones were isolated that were homologous to the N-terminal 73 kDa fragment of human PR130 The longest clone lacks 17 amino acids at its N-terminal part compared to the N-terminus of human PR130 The open reading frame ends seven amino acids after the putative splicing site The stop codon (vertical line) is followed by a polyadenylation signal and a poly(A) tail Arrows shown on the longest clone represent 5¢-RACE strategy
of XN73 as described in materials and methods (C) Class III: 15 clones were isolated that were homologous to human PR48 Three clones have sequences, upstream of the putative initiation start, which are homologous to sequences of other members of the PR72/130/B¢¢ family Arrows shown on the longest clone represent 5¢ RACE strategy of XPR70 as described in Materials and methods.
Trang 6specific part of HPR130 reside in one exon of 2435 bp (exon
2a) Comparison of the XN73 and XPR130 sequences with
the exon–intron boundaries of the human PR130 gene
revealed that the point where divergence between XN73 and
XPR130 occurs, is at an exon–intron boundary (Fig 2A,C) These results are further evidence for XN73 being a splice variant of XPR130 However, due to the large number of base pairs between exon 2a and exon 3 in the human PR130
Trang 7gene (20876 bp), no sequence with homology to the specific
XN73 region could be found using several alignment
techniques This can be explained by the short specific
sequences (264 bp) of which only 21 bp are coding
sequences in XN73
Further analysis of the genomic organization of
PPP2R3 A confirmed that PR72 and PR130 are indeed
splice variants generated from a single gene, as originally
suggested by Hendrix and coworkers [21] (see Table 1 and
Fig 2C) The human PR72 protein is identical in its
C-terminal part to human PR130, but has a totally
different N-terminus Examination of the PPP2R3 A gene
showed that this different N-terminus is encoded by one
additional exon (exon 2b) located in the intron
down-stream of exon 2a Moreover, PR130 and PR72 start with
different 5¢-UTR sequences and might be transcribed
from separate promoters
cDNA cloning of PR48 from a prophase Xenopus oocyte library
All 15 clones making up the third group of cDNAs isolated from the Xenopus oocyte kgt10 library were very homologous to PR72/130/B¢¢ subunits (Fig 1C) An alignment of the amino acid sequences revealed that they were most similar to human PR48 This homology is most striking for their C-terminal sequences, where in addition
to the N-terminus, most of the differences between the mammalian B¢¢-subunits are located Interestingly, three clones contained extra sequences upstream of the putative initiation start that are homologous to sequences from other members of the B¢¢-family (PR72 and PR59 for example) It is feasible that the isolated Xenopus clones encode a protein that contains an N-terminal extension compared to human PR48 (accession number: AF135016)
To ensure that we had obtained a full-length cDNA, the 5¢-RACE technique was performed on total RNA of prophase Xenopus oocytes Several nested PCR products were generated The longest product was 124 bp longer than the cDNA isolated from the library, but no ATG start codon was present Therefore, these sequences probably make up a 5¢-UTR region Using an in vitro transcription and translation assay of total cDNA (5¢-UTR included) a protein of approximately 70 kDa was produced (Fig 3A) To determine where the real initiation start was, we generated clones with truncations of the cDNA sequences that could only initiate translation at downstream alternative start codons (Fig 3A) The respective ATG start codons are underlined in Fig 3B Only the cDNA missing just the putative 5¢UTR gener-ated a protein of 70 kDa Therefore, we concluded that the first ATG codon encountered is the start codon used and that the Xenopus PR48 orthologue (submitted under accession number: AY043259) encodes a protein of about
70 kDa We therefore propose to rename Xenopus PR48
as XPR70, more in agreement with its molecular mass as observed in SDS/PAGE
Fig 2 Comparison of XPR130,HPR130 and XN73 (A) Alignment of
Xenopus PR130 (AY043260) and Xenopus N73 (AY043258) with
human PR130 protein (L12146) The putative splicing site is indicated
by a closed black triangle Sequences represented in grey are obtained
by the RACE technique Identical amino acids are shown in bold.
XPR130 5¢ end sequences are generated from XN73 RACE fragments,
assuming that XN73 and XPR130 are derived from the same gene (B)
In vitro transcription and translation products of total cDNA of wild
type XN73 (Lane 1), GST-tagged XN73 (lane 2) His 6 -tagged XN73
(lane 3) and XPR130 (Lane 4) The 10 kDa markers are indicated by
arrows (C) Genomic structure of the human PR72/130 gene
(PPP2R3 A) with the exon combinations of the different mRNA splice
variants: HPR130, HPR72 and XPR74 Splicing events are denoted by
dotted lines The length of the exons, but not the introns, are scaled.
The coding regions are depicted in black, whereas the untranslated
regions are represented by open boxes with exons numbered as
indi-cated.
Table 1 Exon-intron organization of the human PR72/PR130 gene Exon sequences are given in capitals Human PR130 and Human PR72 are composed by the following exons: 1, 2a, 3–14 and 2b, 3–14, respectively.
Exon Exon size (bp) 5¢-Splice donor Intron size (bp) 3¢-Splice acceptor
Trang 8Human PR48 is similar toXenopus PR70
The identification of a variant of PR48 having an extension
in its N-terminal region in X laevis (XPR70) suggests a
possible existence of the same cDNA in other species
Therefore, we performed a BLAST search with cDNA
sequences of XPR70 as bait Several human EST clones
were identified that contained extensive homology
(acces-sion number AL555389 being the longest) Figure 3Bshows
an alignment of the first 340 bp of cDNA sequences of
human PR48 (accession number: AF135016), the first
868 bp of Xenopus PR70 and this human EST clone
(Fig 3B) Translation of the human EST sequences
gener-ated a protein with sequences similar to those of the Xenopus
form, although some frame shifts occurred possibly due to
sequencing errors.Since these human EST-sequences are
identical at their 3¢-ends to the 5¢end of the reported human
PR48, these data strongly suggest that the human PR48
cDNA is incomplete and contains a long 5¢ extension similar
to the Xenopus situation
Genomic organization of human PR70
To confirm the conclusions drawn from the human EST
clones, we examined the human PR48 gene for the presence
of the additional sequences A BLAST search against the
human genome site with a combination of HPR48 cDNA
(accession number: AF135016) and the longest EST clone
(AL555389) containing the N-terminal extension of
HPR48, yielded an almost perfect match (99%) with
two overlapping unfinished genomic contigs: NT_026499.3
containing almost the complete cDNA (lacking some
sequences at the 3¢-end) and NT_033320.1 containing
3¢-end sequences, but lacking some 5¢ end sequences Apart
from these two contigs covering the whole HPR48 gene,
another match was found with a working draft containing
unordered pieces (AC126765), which has sequence identity
of > 97% with the two assembled contigs The putative
N-terminal extension of Human PR48 was present in this
working draft as well The further analysis of the gene was
done with the combination of contig NT_026499.3 and
NT_033320.1 The complete gene, designated PPP2R3B according to the human gene nomenclature, spans approxi-mately 47 kb of DNA in the genome and is composed of at least 13 exons varying in size between 49 (exon 8) and 561 nucleotides or more (exon 1) and 12 introns, covering the entire coding region The splice sites and exon-intron
Fig 3 Comparison of human PR70 with Xenopus PR70 (A) In vitro
transcription–translation products of total cDNA of XPR70, 5¢-UTR
included, (lane 1), of truncated cDNA starting from downstream
al-ternative start codon 1 (lane 2), 2 (lane 3), 3 (lane 4) and 4 (lane 5) The
10 kDa markers are indicated by arrows (B) Alignment of 5¢-cDNA
sequences of Xenopus PR70 (AY043259) with human PR48
(AF135016) and a human EST clone (AL555389) cDNA sequences of
Xenopus PR70, obtained by the RACE technique, are indicated in
grey Sequences of the human PR48 EST-clone, shown in black,
demonstrate that the human EST clone is probably a 5¢ extension of
human PR48 Identical sequences are shown in bold Alternative start
codons of the truncated cDNAs used in the in vitro transcription–
translation assay shown in (A) are underlined (C) Alignment of amino
acid sequences of Xenopus PR70 (AY043259) with the extended
hu-man PR48 (HPR70) derived from a combination of huhu-man PR48
(AF135016), the EST clone DNA sequences (AL555389) and genomic
sequences (NT_026499.3) Identical sequences are shown in bold The
start codon of HPR48 as proposed by Yan and coworkers [27] is
indicated by a black triangle.
Trang 9boundaries are presented in Table 2 The intron–exon
boundaries were deduced by comparing the sequences
obtained from the genomic clones and the respective cDNA
with all the intron–exon boundaries following Chambon’s
rule (GT/AG) for the splice donor and acceptor sites [42]
The 99% match of exon 1 with the human EST clone
(AL555389) strongly suggests that it is the N-terminal
extension of the published HPR48 protein [27] The putative
translation start site located in exon 1 was deduced by
comparison with XPR70 The sequence shows a 1728 bp
open reading frame corresponding to a predicted protein
product of 575 amino acids, with a predicted molecular
mass of 65.1 kDa An alignment of the amino acid
sequences of human PR48 (accession number: BK000521)
with XPR70 is shown in Fig 3C For obvious reasons, we
propose to rename full-length human PR48 as human PR70
(HPR70)
XPR70 and XPR130 are subunits of PP2A
To determine whether the predicted proteins as isolated
from the Xenopus oocyte cDNA library are indeed third
subunits of PP2A, GST-pulldown assays were performed
Glutathione S-transferase was fused to the cDNA encoding
XPR130, XN73 and XPR70 using a GST Gene Fusion
System and subsequently inserted into the Xenopus
expres-sion vector pGemHE [37] This vector is designed to
generate high expression of the encoded proteins from an
in vitro transcribed RNA injected into X laevis oocytes
High expression is obtained as a consequence of inserting
the cDNAs between the 5¢- and 3¢-UTR of b-globin, a
protein that is highly expressed in Xenopus laevis oocytes
GST-fusion proteins of XPR130, XN73 and XPR70 were
expressed in prophase X laevis oocytes using this system
After 24 h incubation, extracts were made and glutatione–
Sepharose beads added After washing, proteins were eluted
from the beads with SDS containing sample buffer,
separated by SDS/PAGE and transferred to nitrocellulose
The presence of PR65/A subunit was detected with a
specific monoclonal antibody (Fig 4) As controls, GST
was expressed alone, or extracts produced without RNA
injection The results show a clear interaction of PR65 with
GST-XPR70 or GST-XPR130 Very little specific binding
was observed when GST-XN73 was expressed Therefore, XPR130 and XPR70 are probably incorporated into a PP2A complex, so are probably genuine PP2A subunits, whereas XN73 may not be Because it is proven that some proteins interact directly with the C subunit, independently
of PR65/A, we developed the same blot as shown in Fig 4 with a monoclonal antibody to PP2ACand found in lane 1 (GST-XN73) no signal that was any higher than found in lane 4 (GST) or 5 (extract), proving that no PP2ACwas pulled down with GST-XN73, independently of PR65/A (results not shown)
Expression of B¢¢-subunits in different tissues
ofX laevis
To compare the expression of the isolated Xenopus B¢¢-subunits in several adult tissues and in oocytes, a semiquantitative RT-PCR strategy was followed rather then the traditional Northern blot strategy, because of the low expression of some B¢¢-subunits in oocytes and some other adult tissues Total RNA from several Xenopus tissues was used as template (liver, gall bladder, spleen, heart, muscle, brain, prophase and metaphase oocytes) A reverse tran-scriptase reaction was performed using primers containing the translation stop of the different isoforms to obtain specific cDNA After purification of this cDNA, semiquan-titative PCR was performed PCR products were visualized
Table 2 Exon-intron organization of human PR70 Exon sequences are given in capitals.
Exon Exon size (bp) 5¢-Splice donor Intron size (bp) 3¢-Splice acceptor
Fig 4 XPR70 and XPR130 are bona fide PP2A subunits GST-fusion proteins were expressed in prophase Xenopus laevis oocytes Proteins from high speed supernatant were bound on glutathione–Sepharose beads and after washing, the beads were subjected to Western blotting and developed with PR65 antibodies as described in Materials and methods.
Trang 10on Southern blots hybridized with internal primers derived
from the different subunits The expression pattern of the
different isoforms is shown in Fig 5 As can be seen,
XPR130 is ubiquitously expressed in all adult tissues
examined, with the exception of the brain, but is only
expressed to a very limited extent in prophase and
metaphase oocytes XN73 is expressed in all adult tissues,
but is hardly detectable in prophase and metaphase oocytes
XPR70 is ubiquitously expressed in adult tissues and
oocytes None of the RNAs in oocytes showed a difference
after meiotic maturation
Discussion
By screening a Xenopus oocyte kgt10 library with the
full-length human PR130 cDNA we have isolated two
B¢¢-subunits of PP2A, XPR130 and XPR70 This is the
first cloning of any B¢¢ subunits from Xenopus cDNAs
encoding PR72, PR59 and the newly described ton2 protein
[43] were not found in the oocyte library It is not really
surprising that PR72 is not found in oocytes, since in rabbit
tissues its presence could only be demonstrated in skeletal
muscle and heart [25] We have observed the presence of a
splice variant of XPR130 in Xenopus skeletal muscle
(unpublished results), very similar to human PR72
How-ever, it came as a surprise that we could not find PR59 in the
oocyte library, since this sequence was originally identified
from a mouse embryo library [26] So far, PR59 has not
been found in any other species Also the ton2 protein,
originally found in Arabidopsis, which shows high
homo-logy to the PR72/130/B¢¢ subunits of PP2A and thought to
have a role in the control of cytoskeletal structure [43], is not
found in the oocyte library The function of its mammalian
counterpart remains to be determined
Most sequences of the isolated XPR70 clones are
identical to human PR48 (68% identity), even in the
extreme C-terminal part where most differences between the
different B¢¢ subunits are located However, this Xenopus
protein has a long N-terminal extension, and so encodes a
protein of about 70 kDa instead of 48 kDa Screening of
the human EST database with the cDNA sequences comprising this N-terminal extension revealed a human EST clone with sequence similarity to this cDNA We have shown that this EST-clone is an extension of the known human PR48 cDNA, and moreover, that sequences of this clone are present as exons at the correct position within the human PR48 gene Consequently, we suggest that this gene
is named PPP2R3Baccording to the human nomenclature However, we cannot exclude that the ATG codon of HPR48 is also an authentic translation start site and that both initiation codons are used in different cellular condi-tions This methionine is not conserved in XPR70 (Fig 3C), though, so this hypothesis is unlikely Therefore, human PR48 probably encodes a protein of about 70 kDa Since HPR72 also encodes a protein of about 70 kDa and PR70 and PR72 have quite similar amino acid sequences, they might fulfil overlapping functions
Besides XPR130 and XPR70, a third cDNA was cloned that is nearly identical to the N-terminal region of XPR130 but with an open reading frame that ends seven amino acids downstream of the putative splicing site of gene PPP2R3A followed by a 3¢-UTR containing a polyadenylation signal and a poly(A) tail This indicates that the cDNA is real and not the result of some splicing errors or cloning artefacts Moreover, four similar clones were isolated and analysis of the gene structure of HPR130 would favour the hypothesis that XN73 is a splice variant of XPR130 However, aBLAST
search with the 3¢ end of XN73 revealed some distinct EST-clones (three different EST EST-clones in total) derived from
X laevistissue that have homologous sequences in a small part (about 55 bp) within the 3¢-UTR Other parts of these EST clones have no homology with XN73 As genome sequences of X laevis are still rare, the accuracy of the different clones cannot yet be verified Therefore, it is possible that either our clones or the EST clones in the database might not be 100% correct Moreover, there were
no EST clones found from other species having homology with the specific parts of XN73 With this caveat, we conclude that this cDNA probably encodes a new protein, the function of which is not yet known It cannot be considered as a real third-subunit of PP2A since no in vivo interaction with the PR65/A-subunit could be demonstra-ted Actually, this did not come as a surprise, as XN73 lacks the C-terminal part of PR130 which has the highest similarity with other subunits of the B¢¢-family and has been demonstrated to contain two putative A-subunit binding domains responsible for PR65/A-subunit binding [44,45] Because the C-terminal part is believed to be responsible for the interaction with the A-subunit, it is likely that the long N-terminal extension of PR130 will act in the specific function of PR130 Interaction partners, such as PP2A substrates or proteins that target PR130 to specific subcellular localizations will probably use these amino acid sequences for binding Therefore, one might speculate that XN73 has an inhibitory role in the function of PR130, as it contains the same N-terminal sequences that may compete for interaction with the same proteins as PR130 The expression patterns of XPR130 and XN73 are very similar They are ubiquitously expressed in all adult tissues with the brain as a notable exception for XPR130 The expression of the proteins is estimated to be 10 times higher in adult tissues than in oocytes, suggesting that the proteins might
Fig 5 Expression of PR72/130/B¢¢ subunits in different adult tissues of
Xenopus laevis Specific cDNA parts of XPR130, XN73 and XPR70
were amplified using semiquantitative RT-PCR and visualized by
hybridization with specific oligonucleotide probes as described in
Materials and methods.