Abbreviations GL, hepatic glycogen-targeting subunit of PP1 encoded by the gene PPP1R43B; GMalso termed RGL, skeletal muscle glycogen-targeting subunit of PP1 encoded by the gene PPP1R33
Trang 1that is regulated by insulin and shows differential tissue distribution in humans and rodents
Shonagh Munro1, Hugo Ceulemans2, Mathieu Bollen2, Julie Diplexcito1and Patricia T.W Cohen1
1 Medical Research Council Protein Phosphorylation Unit, University of Dundee, UK
2 Katholieke Universiteit Leuven, Faculteit Geneeskunde, Afdeling Biochemie, Belgium
Insulin-stimulated glycogen synthesis is decreased in
type 2 diabetes [1,2] One of the routes by which
insu-lin stimulates this pathway is through activation of the
rate-limiting enzyme, glycogen synthase (GS), via the
phosphatidylinositol-3-kinase⁄ protein kinase B
path-way, which leads to the inhibition of glycogen synthase
kinase 3 (GSK3) [3,4] Activation of GS results from a
net dephosphorylation of serine residues that are
phos-phorylated by GSK3 and dephosphos-phorylated by
glyco-gen-associated protein phosphatase 1 (PP1) [5–8] In
order to determine how insulin modifies GS activity, it
is therefore crucial to understand the mechanisms by
which insulin may activate glycogen-targeted PP1 The
latter mainly exists as heterodimeric complexes of the catalytic subunit, PP1c, bound to a regulatory subunit [9] In striated muscles the most abundant glycogen-binding subunit GM (124–126 kDa, encoded by the gene PPP1R3A) targets PP1c to the sarcoplasmic reti-culum as well as to glycogen particles [10–12] A much smaller protein, GL (33 kDa, encoded by the gene PPP1R3B), is the most abundant glycogen-targeting subunit of PP1 in liver, although it is only 23% identi-cal to the N-terminal region of GM[13,14] Two other glycogen-binding subunits, R5⁄ PTG (36 kDa, encoded
by PPP1R3C) with 40% identity to GL and R6 (33 kDa, encoded by PPP1R3D) with 30% identity
Keywords
diabetes; glycogen metabolism; glycogen
synthase; insulin; PP1
Correspondence
P T W Cohen, MRC Protein
Phosphorylation Unit, School of Life
Sciences, MSI ⁄ WTB Complex, University of
Dundee, Dow Street, Dundee DD1 5EH, UK
Fax: +44 1382 223778
Tel: +44 1382 344240
E-mail: p.t.w.cohen@dundee.ac.uk
(Received 23 September 2004, revised 16
December 2004, accepted 26 January 2005)
doi:10.1111/j.1742-4658.2005.04585.x
Stimulation of glycogen-targeted protein phosphatase 1 (PP1) activity by insulin contributes to the dephosphorylation and activation of hepatic cogen synthase (GS) leading to an increase in glycogen synthesis The gly-cogen-targeting subunits of PP1, GL and R5⁄ PTG, are downregulated in the livers of diabetic rodents and restored by insulin treatment We show here that the mammalian gene PPP1R3E encodes a novel glycogen-target-ing subunit of PP1 that is expressed in rodent liver The phosphatase activ-ity associated with R3E is slightly higher than that associated with R5⁄ PTG and it is downregulated in streptozotocin-induced diabetes by 60– 70% and restored by insulin treatment Surprisingly, although mRNA for R3E is most highly expressed in rat liver and heart muscle, with only low levels in skeletal muscle, R3E mRNA is most abundant in human skeletal muscle and heart tissues with barely detectable levels in human liver This species-specific difference in R3E mRNA expression has similarities to the high level of expression of GL mRNA in human but not rodent skeletal muscle The observations imply that the mechanisms by which insulin regu-lates glycogen synthesis in liver and skeletal muscle are different in rodents and humans
Abbreviations
GL, hepatic glycogen-targeting subunit of PP1 encoded by the gene PPP1R4(3B); GM(also termed RGL), skeletal muscle glycogen-targeting subunit of PP1 encoded by the gene PPP1R3(3A); GS, glycogen synthase; GSK3, glycogen synthase kinase-3; GSP, glycogen synthase phosphatase; GST, glutathione S-transferase; MBP, maltose-binding protein; NCBI, National Center for Biotechnology Information USA; PCR, PP1, protein phosphatase 1; PP1c, protein phosphatase 1 catalytic subunit; R5 (also termed PTG), regulatory subunit of PP1 encoded
by the gene PPP1R5(3C); R6, regulatory subunit of PP1 encoded by the gene PPP1R6(3D).
Trang 2to GL have a wide tissue distribution [15–17]
Interest-ingly, although GLis expressed at only very low levels
in rodent skeletal muscles, it is found in human
skel-etal muscles at levels comparable with those in human
liver [18] The four glycogen-targeting subunits bind
to PP1c via a short highly conserved motif (-RVXF-)
This motif is also responsible for the interaction of
many other regulatory subunits with PP1, explaining
why the binding of targeting regulatory subunits to
PP1c is mutually exclusive [19] However, certain
inhibitor proteins have also been noted to form ternary
structures with PP1c-targeting subunit complexes
[10,20–22] In addition to the PP1 and
glycogen-bind-ing motifs, the PP1 glycogen-regulatory subunits
pos-sess a motif for the interaction with substrates [23]
The glycogen-targeting subunits can modulate the
activity of PP1c towards different substrates; for
exam-ple, GLenhances PP1c activity towards GS while
sup-pressing its activity towards phosphorylase There is
evidence that PP1-GM and PP1-GL may be regulated
acutely by insulin Assay of PP1 following insulin
infu-sion of skeletal muscle and immunopelleting of
PP1-GM showed a 1.5–2-fold increase in phosphatase
activity with insulin [24] In GMnull mice, this activity
was absent and GS could not be fully activated by
insulin [24] In contrast, studies on an independently
derived GMnull mouse model found that insulin
acti-vation of GS was in the normal range, indicating that
the PP1-GM is not required for the insulin activation
of GS [25] These workers postulated the existence of a
novel insulin-activated form of glycogen-targeted PP1
[25] In the case of hepatic glycogen-targeted PP1,
insulin is thought to exert its acute activating effect on
PP1-GL mainly through modulation of cAMP levels
and decrease of phosphorylase a, which is a potent
inhibitor of hepatic glycogen synthase phosphatase
(GSP) activity [26–32] Phosphorylase a binds to 16
amino acids at the extreme C-terminus of GL, a
seq-uence that is absent from the other three
glycogen-targeting subunits [18,31] R5⁄ PTG and R6 ⁄ PPP1R3D
are not known to be acutely regulated by insulin
Insu-lin exerts a longer term regulation on hepatic GL and
R5⁄ PTG [33,34] Diabetic rats exhibit a loss of hepatic
glycogen-bound synthase phosphatase activity that can
be restored by insulin administration [35,36] The main
underlying defects are decreased expression of the two
PP1 glycogen-targeting subunits, GL and R5⁄ PTG, in
the diabetic state [33,34] Downregulation of both the
protein and mRNA levels of hepatic GL and R5⁄ PTG
are restored by insulin treatment, but the skeletal
mus-cle R5⁄ PTG level is not altered by insulin [33,34,37]
The expression of hepatic R6⁄ PPP1R3D is also
unaf-fected in diabetic animals [34]
Ceulemans et al [38], undertook a bioinformatic approach in order to trace the evolution of regulatory subunits of PP1 Searching completed genome sequences, with the sequences of known PP1 regula-tory subunits, including the conserved PP1 and glyco-gen-binding regions of the glycogen-targeting subunits, they identified nine new potential regulatory subunits
of PP1 Of these nine sequences, three were deduced
to encode for putative human glycogen-targeting sub-units, and were given the nomenclature PPP1R3E, PPP1R3F and PPP1R3G These potential human pro-teins all contained the canonical -RVXF- motif that mediates interaction with PP1, as well as putative mod-ules for targeting to glycogen and facilitating interac-tion with PP1 substrates such as GS Here we show that the phosphatase activity of one of these novel tar-geting subunits, R3E, is under long-term control by insulin in rodent liver while being virtually absent from rodent skeletal muscle; yet surprisingly, PPP1R3E mRNA is found at appreciable levels in human skeletal muscle
Results
Cloning of human PPP1R3E and PPP1R3G from human cDNA libraries
Interrogation of the NCBI databases revealed no full-length mammalian cDNAs sharing similarities to the human genomic sequences PPP1R3E, PPP1R3F and PPP1R3G (Accession nos: AL049829, NM_033215, AL035653) with chromosomes locations 14q11.2, Xp11.23 and 6p24.3-25.3, respectively In order to establish whether these sequences were functional genes or pseudogenes, attempts were made to amplify their putative cDNAs from human libraries, using primers designed from the genomic sequences For PPP1R3E, a single cDNA of 800 bp was amplified
by two rounds of polymerase chain reaction (PCR) from both human testis and brain cDNA libraries The size of these products was consistent with the expected size for a putative sequence for a glycogen-targeting subunit of PP1 assuming the genomic sequence has two coding exons separated by one intron (Fig 1A) A cDNA corresponding to the genomic sequence of PPP1R3G was obtained by PCR from the human brain library, but no full-length cDNA products were obtained for PPP1R3F Subcloning and sequencing of the PCR products confirmed the identity of the testis and brain cDNAs for PPP1R3E and showed that the protein translated from this sequence was composed of
279 amino acids with a predicted molecular mass of 30.6 kDa (Fig 1B) The PPP1R3G cDNA sequence
Trang 3was verified as being identical to the genomic sequence
with one coding exon specifying a protein of 358
amino acids and a molecular mass of 38 kDa (data
not shown)
Comparison of the glycogen-targeting subunits
of PP1
Searching mouse and rat genomic sequences in the
NCBI databases identified predicted rodent cDNAs
from genes homologous to human PPP1R3E and
PPP1R3G The encoded rat and mouse R3E proteins share around 97% amino acid identity and are 89% identical to their human orthologue, indicating that this regulatory protein is very well conserved in mam-mals (Fig 2A) R3G is slightly less highly conserved; the rodent orthologues are 90% identical and they are
11 amino acids shorter than their human orthologue, sharing 67% identity (Fig 2B) A phylogenetic tree depicting the relationship between known glycogen-tar-geting subunits of PP1 and the novel glycogen-target-ing subunits is shown in Fig 2D Although all seven
Gene
A
B
mRNA
E2 E3 E4 E5
1 (ATG) 837 (STOP)
E1
E5
3176
-105 gaagcggacccagcgacttctgcgctgacgcggggcgggcgggagagaggaagagaggggagcgcggtggcgctgcgagctggccccgccggggaaggggctgcc -1
1 ATG TCC CGT GAG CGG CCC CCG GGC ACC GAC ATT CCC CGC AAC CTG AGC TTC ATC GCC GCG CTA ACG GAG CGC GCC 75
1 M S R E R P P G T D I P R N L S F I A A L T E R A 25
76 TAC TAC CGT AGC CAG CGG CCC AGC CTC GAG GAG GAG CCG GAG GAG GAG CCA GGC GAG GGC GGG ACG CGG TTC GGG 150
26 Y Y R S Q R P S L E E E P E E E P G E G G T R F G 50
151 GCC CGA TCC CGC GCT CAC GCA CCG AGT CGG GGC CGC CGG GCC CGA TCT GCA CCA GCC GGA GGC GGC GGG GCC CGG 225
51 A R S R A H A P S R G R R A R S A P A G G G G A R 75
226 GCG CCC CGC AGC CGT AGC CCA GAC ACC CGC AAG AGA GTG CGT TTC GCC GAC GCA CTG GGG TTG GAG CTG GCT GTC 300
76 A P R S R S P D T R K R V R F A D A L G L E L A V 100
301 GTG CGC CGC TTC CGT CCC GGT GAG CTG CCC CGG GTG CCC CGC CAC GTG CAG ATC CAA TTG CAG AGG GAC GCC CTC 375
101 V R R F R P G E L P R V P R H V Q I Q L Q R D A L 125
376 CGC CAC TTC GCG CCC TGC CAG CCC CGC GCC CGC GGC CTC CAG GAG GCG CGC GCC GCC CTG GAG CCG GCC AGC GAG 450
126 R H F A P C Q P R A R G L Q E A R A A L E P A S E 150
451 CCC GGC TTC GCC GCC CGC TTG CTG ACG CAG CGC ATC TGC CTG GAA CGC GCC GAG GCG GGC CCG CTG GGC GTG GCC 525
151 P G F A A R L L T Q R I C L E R A E A G P L G V A 175
526 GGG AGC GCG CGC GTG GTG GAC CTG GCC TAC GAG AAG CGC GTG AGC GTG CGC TGG AGC GCC GAC GGC TGG CGG AGC 600
176 G S A R V V D L A Y E K R V S V R W S A D G W R S 200
601 CAA CGC GAG GCG CCA GCC GCC TAC GCC GGT CCG GCC CCG CCC CCG CCG CGC GCC GAC CGC TTC GCC TTC CGC CTG 675
201 Q R E A P A A Y A G P A P P P P R A D R F A F R L 225
676 CCC GCG CCG CCG ATT GGG GGC GCC CTG CTC TTC GCC TTG CGC TAC CGT GTG ACA GGT CAC GAG TTC TGG GAC AAC 750
226 P A P P I G G A L L F A L R Y R V T G H E F W D N 250
751 AAC GGC GGC CGT GAC TAT GCT CTA CGT GGG CCC GAG CAC CCG GGC AGT GGC GGA GCT CCG GAG CCG CAG GGC TGG 825
251 N G G R D Y A L R G P E H P G S G G A P E P Q G W 275
826 ATC CAC TTT ATC TGA gacgaggcgcctgcggccgacggcggaaaacaccaaaggcacccgggggcggggcgacccgatgtggcggggaggagtag 920
Fig 1 (A) Diagram of human PPP1R3E mRNA compared with PPP1R3E gene (Accession no: ENSG00000129525) Nucleotide numbers at the start and end of each exon are given relative to the first nucleotide of the initiating methionine codon The exon ⁄ intron structure within the coding region was determined experimentally by PCR of cDNA libraries, whereas that for the untranslated region is predicted from the genomic sequence and partial cDNAs in the database (B) Human PPP1R3E cDNA and the encoded protein determined by PCR of human brain and testis cDNA libraries The PP1 binding motif is underlined, glycogen-targeting domain is double underlined, and the substrate-bind-ing sequence is underscored by a wavy line Oligonucleotides primers used for PCR are indicated by arrows.
Trang 4MOUSE R3E 1
A
B
C
D
MSPERPPRTDIPRNLSFIAALTERAYYRSQRPSLEEESEEEPGEGGTRPGARSRAHVPG
RAT R3E 1 MSHERPPRNDIPRNLSFIAALTERAYYRSQRPSLEEESEEEPGEGGTRPGARSRAHVPG
HUMAN R3E 1 MSRERPPGTDIPRNLSFIAALTERAYYRSQRPSLEEEPEEEPGEGGTRFGARSRAHA S
MOUSE R3E 60 RGRRARSAPAGGGGARTARSRSPDTRKRVRFADALGLELAVVRRFRPGEPPRVPRHVQV
RAT R3E 60 RGRRARSAPAGGGGARTARSRSPDTRKRVRFPDALGLELAVVRRFRPGEPPRVPRHVQV
HUMAN R3E 60 RGRRARSAPAGGGGARAPRSRSPDTRKRVRFADALGLELAVVRRFRPGELPRVPRHVQI
MOUSE R3E 119 QLQRDALRHFAPCPPRARGLQEARVALEPALEPGFAARLQAQRICLERADAGPLGVAGS
RAT R3E 119 QLQRDALRHFAPCPPRTRGLQDARIALEPALEPGFAARLQAQRICLERADAGPLGVAGS
HUMAN R3E 119 QLQRDALRHFAPCQPRARGLQEARAALEPASEPGFAARLLTQRICLERAEAGPLGVAGS
MOUSE R3E 178 ARVLDLAYEKRVSVRWSADGWRSLRESPASYAGPAPSPPRADRFAFRLPAPPVGGTLLF
RAT R3E 178 ARVLDLAYEKRVSVRWSADGWRSLRESPASYAGPAPAPPRADRFAFRLPAPPVGGALLF
HUMAN R3E 178 ARVVDLAYEKRVSVRWSADGWRSQREAPAAYAGPAPPPPRADRFAFRLPAPPIGGALLF
MOUSE R3E 237 ALRYRVTGREFWDNNGGRDYALLGPEHPAGAGAAEPQGWIHFI 279
RAT R3E 237 ALRYRVTGREFWDNNGGRDYALLGPEHPGGAGAAEPQGWIHFI 279
HUMAN R3E 237 ALRYRVTGHEFWDNNGGRDYALRGPEHPGSGGAPEPQGWIHFI 279
Rat R3G 1 MEASGEQLHRSEASSSTSSEDPPPAEELSVPEVLCVESG -TSEVPI
Mouse R3G 1 MDPSGEQLHRSEASSSTSSGDPQSAEELSVPEVLCVESG -TSETPI
Human R3G 1 MEPI ARLS-LEAPGPAPFRE PPAEELPAP V CVQG GDGGGASETPS
Rat R3G 46 PDDQLQDRLLSAQKVAALPEQEELQEYRR-SRVRSFSLPADPILQAAKLL
Mouse R3G 46 PDAQLQDRPLSPQKGAALPEQEELQEYRR-SRARSFSLPADPILQAAKLL
Human R3G 50 PDAQLGDRPLSPKEEAAPQEQEELL CRRRCRARSFSLPADPILQAAKF
Rat R3G 95 QQRQQ -AGQPSSEGGEPAGDCCSKCKKRVQFADSLGLSLASVKHFS
Mouse R3G 95 QQRQQ -AGQPSSEGGAPAGDCCSKCKKRVQFADSLGLSLASVKHFS
Human R3G 100 QQQQQQAVALGGEG EDAQLGPG CCAKCKKRVQFADTLGLSLASVKHFS
Rat R3G 140 EAEEPQVPPAVLSRLHSFPLRAEDLQQLGELLAVAKVPAPLLT R QLRP
Mouse R3G 140 EAEEPQVPPAVLSRLHSFPLRAEDLQQLGGLLAVATMPDPLLV CARLRP
Human R3G 150 EAEEPQVPPAVLSRLRSFPMRAEDLEQLGGLLAA A AAPLSAP SRLRP
Rat R3G 190 LFQLPGLIAAEERLRRQRVCLERVQCSQPPRAEVTGSGRVISCPGPRAVA
Mouse R3G 190 HFQLPE RAAEERLRRQRVCLERVQCSQPPRAEVTGSGRVISCPGPRAVA
Human R3G 200 LFQLPGPSAAAERLQRQRVCLERVQCSTASGAEVKGSGRVLSCPGPRAVT
Rat R3G 240 VRYTFTEWRTFLDVPAELHPESLEPLS - RSGNSGPGAEDSEGEPGTER
Mouse R3G 240 VRYTFTEWRTFLDVPAELDPESLEPLP - QSGDSGSKAEDSEEGPGTER
Human R3G 250 VRYTFTEWRSFLDVPAELQPEPLEPQQPEAPSGA EPGSGDAKKEPGA C
Rat R3G 289 F FSLCLPPGLQPKEGEDADTWGVAIHFAVCYRCEQGEYWDNNEGANYTL
Mouse R3G 289 FHFSLCLPPGLQPKEGEDAGAWGVAIHFAVCYRCEQGEYWDNNEGANYTL
Human R3G 300 FHFSLCLPPGLQPE EEDADERGVAVHFAVCYRCAQGEYWDNNAGANYTL
Rat R3G 339 RYVCSTDPL 347
Mouse R3G 339 RYVCSTDPL 347
Human R3G 350 RYARPAD L 358
PP1 binding motif
G M / R GL /R3A 60 GT RRV S FAD
G L /R3B 58 V KKRV S FAD
R5/PTG/R3C 81 A KKRV V FAD
R6/R3D 99 Q K RV R FAD
R3E 84 T RKRV R FAD
R3F 167 AP RRV L FAD
R3G 128 C KKRV Q FAD Glycogen binding domain
G M / R GL /R3A 144 G IRVLNV S FEK L Y VR M L D
G L /R3B 146 GTVKV Q NL A FEK T K IR M TFD T
R5/PTG/R3C 171 GTVKV K NV S FEK K Q IR I TFD S
R6/R3D 191 GTVRV C NV A FEK Q A VR Y TF SG W
R3E 176 GS A RVVDL A YEK R S VR W A G
R3F 300 G VRVLN RS FEK A H VR A H G
R3G 235 GS G RVL SCPGP R V VR Y TF TE W Substrate binding domain
G M / R GL /R3A 219 W NN N T NY
G L /R3B 221 WD S R GKNY
R5/PTG/R3C 246 WDNN D Q NY
R6/R3D 267 WDNN DH RDY
R3E 248 WDNN G GRDY
R3F 407 W NN H GRNY
R3G 339 WDNN A A NY
hR3F mR3F
R6 mR3DrR3D hR3D
mR3E rR3E hR3E
mR3G rR3G hR3G hR3B
hR3C rR3C mR3C R5/PTG
rbR3A hR3A mR3A
rR3B mR3B
G L
G M /R GL
Fig 2 Amino acid alignment of human proteins with their rat and mouse homologues (A) R3E, (B) R3G Identities are shaded in black and similarities are shaded in grey The PP1-binding motif is indicated by a single underline The sequences were aligned using CLUSTALW (http:// www.clustalw.genome.ad.jp/) and shading was performed using BOXSHADE (v3.21 K.Hofmann and M.Baron) NCBI Accession nos for predic-ted cDNAs are: XM_193763 (mouse R3E), XM_344406 (rat R3E), XM_225280 (rat R3G) Mouse R3G cDNA Accession no is AK049829 (C) Amino acid alignment of the conserved regions of the glycogen-targeting subunits of PP1 Identification of the PP1-binding motif was des-cribed in Egloff et al [43], the glycogen-binding domain in [15,31,44] and the substrate binding domain in [23,45] (D) Phylogenetic relation-ship between the glycogen-targeting subunits of PP1 The unrooted tree is derived by the neighbour-joining method in CLUSTAL W from pairwise sequence distances between the conserved PP1, glycogen and substrate-binding domains (corresponding to amino acids 85–258 of R3E) of human (h), mouse (m), rat (r) and rabbit (rb) glycogen-targeting subunits The proteins aligned and their database Accession nos are R3A(PPP1R3AG ⁄ RGL) NP_002702 (h), NP_536712 (m), A40801 (rb); R3B(PPP1R3B ⁄ GL) [18] and NP_078883 (h), NP_808409 (m), NP_620267 (r); R3C(PPP1R3C ⁄ R5 ⁄ PTG) NP_005389 (h), NP_058550 (m), XP_220048 (r); R3D(PPP1R3D ⁄ R6) Y18206 ⁄ NP_006233 (h), XP_141580 (m), XP_230940 (r); R3E(PPP1R3E) this study (h, m, r); R3F(PPP1R3F) XP_372210 (h), AAH59275 (m); R3G(PPP1R3G) this study (h, m, r).
Trang 5human subunits and their rodent orthologues possess known or putative PP1, glycogen and substrate-bind-ing domains (Fig 2C), no two subunits share more than 40% amino acid identity Despite this, each gly-cogen-targeting subunit is particularly well conserved between rodents and humans, suggesting that each subunit may serve an important, nonredundant func-tion in mammals
Tissue distribution of PPP1R3E and PPP1R3G mRNA
The human PPP1R3E cDNA probe hybridized to two mRNA species on a human multiple tissue northern blot with sizes 7.2 and 5.9 kb (Fig 3A) These tran-scripts were predominantly present in skeletal muscle and heart, although the smaller transcript was also present in pancreas and placenta and was detectable at very low levels in liver and kidney The sizes and the tissue distribution of these transcripts are not consis-tent with those encoding any of the other characterized glycogen-targeting subunits In addition, the extremely low level of sequence similarity at the nucleotide level implies that cross-hybridization with the mRNAs for the other subunits is unlikely Hybridization of a nor-thern blot with the human PPP1R3G cDNA probe revealed a single PPP1R3G mRNA transcript of
9 kb that was present exclusively in brain (Fig 3C) Unfortunately, attempts to amplify mouse or rat PPP1R3E cDNA from tissue specific libraries were unsuccessful However, the rat PPP1R3E exons showed
a high level of conservation (86% identity) with the coding region of human PPP1R3E cDNA This cou-pled with the lack of sequence similarity to the coding regions of other glycogen-targeting subunits, allowed the human cDNA probe to be used to establish the tis-sue distribution of rat PPP1R3E mRNA on a northern blot (Fig 3B) Following a series of stringent washes and autoradiography, the probe hybridized predomi-nantly to 6.0 kb and 5.0 kb mRNA species in heart
Heart Brain Placenta Lung Liv Sk
2.0 kb
α-actin
7.2 kb 5.9 kb
9.5 kb
7.5 kb
4.4 kb
2.4 kb
1.35 kb
Human PPP1R3E mRNA blot
Heart Brain Lung Liv Sk
Spleen Testis
β-actin α-actin
2.0 kb
1.8 kb
9.5 kb
7.5 kb
4.4 kb
2.4 kb
1.35 kb
6.0 kb 5.0 kb 4.5 kb
B
C
A
Rat PPP1R3E mRNA blot
Heart Brain Placenta Lung Liver Skeletal muscle Kidney Pancreas
2.0 kb
9.5 kb
7.5 kb
4.4 kb
2.4 kb
1.35 kb
0.24 kb
~ 9.0 kb
Human PPP1R3G mRNA blot
Fig 3 Tissue distribution of (A) human PPP1R3E mRNA (B) rat PPP1R3E mRNA (C) human PPP1R3G mRNA Blots contained
2 lg of poly(A) +
RNA from different tissues The upper panels of (A) and (B) were hybridized with a probe corresponding to the entire coding region (837 bp) of human PPP1R3E and the upper panel of (C) was hybridized with a probe corresponding to the entire coding region (1074 bp) of human PPP1R3G Following auto-radiography, the membranes were stripped in 0.5% (w ⁄ v) SDS at
100 C for 5 min and subsequently re-probed with a b-actin in order
to assess whether equal amounts of the samples were loaded In heart and skeletal muscle the b-actin probe cross-hybridizes with a-actin.
Trang 6and to a 4.5 kb RNA mRNA in liver Surprisingly, the
probe hybridized only weakly to the 6.0 and 5.0 kb
transcripts in skeletal muscle The 5.0 kb transcript was
also present in brain, spleen, lung, liver, kidney and
testis, albeit at very low levels
R3E protein is present in the rat liver glycogen
fraction and phosphatase activity associated with
R3E is higher than that associated with R5/PTG
Anti-R3E(8–23) sera were raised against amino acids
8–23 in the N-terminus of human R3E, as this is the
region that shares no similarity with other
glycogen-targeting subunits These antibodies and
anti-GST-R3E(1–98) sera recognized as little as 0.2 ng of
bacteri-ally expressed GST-R3E(full length, human) (Fig 4A)
Anti-GST-R3E(1–98) was virtually specific for R3E as
it did not recognize 100 ng of GM, GL, R6 or
R5⁄ PTG (Fig 4B) The peptide antibody was
extre-mely specific as it did not cross-react with 100 ng of
GM, GL, R5⁄ PTG or R6 (data not shown)
The presence of a fairly well-conserved glycogen-binding motif in R3E suggested that it may interact with glycogen To test this hypothesis, a rat liver lysate, microsomal fraction and a glycogen fraction were prepared and the proteins in these fractions separated by SDS⁄ PAGE, transferred to nitrocellulose and immunoblotted A single R3E band was detected
in the glycogen fraction, which was consistent with the predicted size of R3E ( 31 kDa) (Fig 5A) This band was sometimes detectable at low levels in rat liver lysates (Figs 4B and 5A)
In order to establish whether R3E could bind (and therefore be regulated allosterically by phosphory-lase a, 4 lg of GST-R3E was transferred to nitrocellu-lose membrane and tested for its ability to bind to [32P]phosphorylase a [31] The32P-labelled phosphory-lase a was found to bind to GST-GL, but not to GST-R3E or GST-R5⁄ PTG (data not shown)
A specific and sensitive phosphatase immunoadsorp-tion assay has been developed [34,39], which allows characterization of the activities of the different
glyco-A
B
C
Fig 4 Specificity and characterization of
PPP1R3E antibodies (A) Recognition of
0.1–10 ng of bacterially expressed GST-R3E
by anti-R3E(8–23) and anti-GST-R3E(1–98)
sera Both antibodies were used at a
concentration of 0.2 lgÆmL)1 (B) Specificity
of anti-R3E(1–98) sera for R3E The
immunoblot of several glycogen-targeting
subunits was probed with 0.2 lgÆmL)1
affinity purified antibody Lane 1, rat liver
glycogen pellet; lane 2, rat liver lysate; lane
3, 1 ng GST-R3E(full-length); lane 4, 2 ng
GST-R3E, lane 5, 5 ng GST-R3E; lane 6,
100 ng GST-GM(1–243); lane 7, 100 ng
GST-GL; lane 8, 100 ng GST-R5 ⁄ PTG; lane 9,
100 ng GST-R6 In the lower panel, the blot
was stripped and reprobed with ant-GST
sera to show the loading of the samples.
Trang 7gen-targeted forms of PP1 Essentially, using specific
antibodies to a glycogen-targeting subunit of choice, it
is possible to pellet the bound PP1c activity in an
immune complex However, the interaction of
regula-tory subunits with PP1c may modify substrate
specific-ity, decreasing the activity of PP1c against some
substrates while increasing it against others The
immune pellet is therefore assayed for protein
phos-phatase activity in the absence and presence of a
pep-tide that dissociates the interaction between PP1c and
glycogen-targeting subunits [40] Inclusion of the
disso-ciating peptide relieves the modification of phosphatase
activity imposed by the glycogen-targeting subunit and
provides a means to calculate the actual amount of
PP1c bound to each subunit After immunoadsorption
of R3E with anti-R3E(8–23) serum, the spontaneous phosphorylase phosphatase activity associated with PPP1R3E (measured in the absence of the dissociating peptide) was 0.006 ± 0.0008 mUÆmg)1 Addition of the dissociating peptide to the assay increased the activity by approximately fourfold to 0.024 ± 0.005 mUÆmg)1 This provided evidence that R3E does indeed interact with PP1c and suggests that the interaction of R3E with PP1 inhibits its activity substantially with phosphorylase a as a substrate (Fig 5B) In contrast, the activity of PP1-R3E using
GS as substrate was similar in the presence and absence of dissociating peptide, demonstrating that R3E exhibited little or no inhibition of PP1c activity towards this substrate (Fig 6B) The glycogen synthase phosphatase⁄ phosphorylase phosphatase (GSP ⁄ PhP) activity ratio for R3E-PP1c of 3.7 is substantially higher than that calculated for GL (1.9), R5 (0.9) and R6 ( 2) [34] Comparison of the level of phosphory-lase phosphatase activity associated with PPP1R3E with that associated with GL and R5 in rat liver, shows that the activity associated with PPP1R3E is
30% of that bound to GL, and is slightly higher than that associated with R5⁄ PTG (Fig 5C)
Effect of induced diabetes and insulin treatment
on the expression and activity of PPP1R3E in vivo Previous studies [33,34] have shown that streptozoto-cin-induced diabetes in rats causes 75 and 60% decrea-ses in the hepatic protein phosphatase activity associa-ted with GL and R5⁄ PTG, respectively This response
is accompanied by a corresponding decrease in the hepatic levels of GL and R5⁄ PTG proteins All of these effects were restored by the intravenous adminis-tration of insulin The finding that R3E appears to be most highly expressed in rodent liver prompted investi-gation into whether this subunit may be regulated
in vivo in liver by streptozotocin-induced diabetes and changes in insulin levels
Figure 6(A,B) illustrates the results of assays of anti-R3E(8–21)–protein G–Sepharose immunopellets from liver lysates of control, diabetic and insulin-treated diabetic rats The phosphorylase phosphatase and GSP activities associated with R3E are decreased by
65–70% in the diabetic rat liver Furthermore, the phosphatase activities associated with R3E could be restored to that of control levels following intravenous administration of insulin for 96 h The same percent-age decrease in phosphorylase phosphatase and GSP activities in diabetic livers and restoration by insulin treatment was observed in the presence of the
dissoci-Phosphorylase phosphatase acti
Phosphorylase phosphatase acti
R3E 28
39
G L R5 R3E +RVXF
-RVXF
peptide peptide
0
0.01
0.02
0.03
A
0.025 0.050
0.1
0.075
0
Fig 5 Detection of PPP1R3E and its associated phosphatase
activ-ity in liver (A) Rat liver lysate (20 lg protein), microsomal fraction
(20 lg protein) and glycogen fraction (2 lg protein) were subjected
to electrophoresis on 10% SDS ⁄ polyacrylamide gels After transfer
to nitrocellulose, the blot was probed with 0.5 lgÆmL)1
anti-GST-R3E(1–98) (B) Phosphorylase phosphatase activity associated with
R3E in rat liver lysates (assayed in the presence of 4 n M okadaic
acid) The R3E complex was immunoadsorped from 100 lg of rat
liver lysate The immune pellets were then assayed for
sponta-neous phosphorylase phosphatase activity (in the absence of
disso-ciating peptide) and total phosphorylase phosphatase activity
(assayed in the presence of the PP1c-dissociating RVXF containing
peptide) Phosphatase activity is expressed in mUÆmg)1total
pro-tein in the rat liver lysate The phosphatase activity in control IgG
protein G-Sepharose immune pellets (0.001 mUÆmg)1) was
subtrac-ted Error bars indicate the SEM for assay of three liver lysates,
each assay being performed in triplicate (C) Comparison of
the total phosphorylase phosphatase activity associated with GL,
R5 and R3E (measured in the presence of the PP1-dissociating
peptide).
Trang 8ating peptide (Fig 6) The activity in IgG control
immune pellets was < 5% of the phosphatase activity
associated with R3E Analysis of the RNA in the livers
of control and streptozotocin diabetic rats showed that
the R3E mRNA levels varied in parallel with the
phos-phatase activities of PP1-R3E (Fig 6C) The data
demonstrate that, like GLand R5⁄ PTG, R3E is
down-regulated in type 1 diabetic animals
Discussion
The novel gene PPP1R3E encoding a putative
glyco-gen-targeting subunit of PP1 is shown here to express
R3E protein in rodent liver R3E shows <33% amino
acid identity to any of the other glycogen-targeting
subunits, but is highly conserved from rodents to
humans (> 86% identity), suggesting that it may serve
an important nonredundant function The R3E protein
was found to be present in the hepatic glycogen
frac-tion and to bind to PP1 The phosphorylase
phospha-tase activity associated with R3E in rat liver was
slightly higher than that bound to R5⁄ PTG and
30% of that bound to the most abundant hepatic
glycogen-targeting subunit GL However, the GSP⁄ PhP
activity ratio associated with R3E is 3.7 compared with
1.9 for GL and 0.9 for R5⁄ PTG indicating that
PP1c-R3E has the potential to contribute 60% of the GSP
activity of PP1c-GLin rat liver The data also indicate
that PP1c-R3E, like PP1c-GL, would be expected to
function mainly as a GSP, whereas PP1c-R5⁄ PTG is
more likely to function predominantly as
phosphory-lase phosphatase
Although analysis of the mRNA encoding rat R3E revealed that the main tissues of expression are liver and heart, with only very low levels being present in skeletal muscle, analysis of the human tissues indicated that PPP1R3E mRNA is most highly expressed in skeletal muscle and heart Very low levels of PPP1R3E mRNA were detected in most other human tissues examined, including liver The difference in tissue dis-tribution between humans and rats reflects in part that
Fig 6 Effect of streptozotocin-induced diabetes on R3E-associated
phosphorylase and GSP activities and R3E mRNA in rat liver.R3E
immune pellets assayed for phosphorylase phosphatase activity (A)
and GSP (B) activities assayed in the absence and presence of
the PP1c RVXF-containing dissociating peptide The activities are
expressed as mUÆmg)1of total protein in the rat liver lysate Error
bars indicate the SEM Control rats (n ¼ 3), diabetic rats (n ¼ 5),
diabetic rats +96 h insulin treatment (n ¼ 4).The differences in
spontaneous phosphorylase phosphatase activities (P < 0.01 for
control and diabetic livers, P < 0.001 for diabetic and insulin treated
livers), and the total phosphorylase phosphatase activities in the
presence of the PP1c dissociating peptide (P < 0.02 for control and
diabetic livers, P < 0.001 for diabetic and insulin-treated livers) are
statistically significant The differences in spontaneous GSP
activ-ities (P < 0.05) and total GSP activactiv-ities in the presence of the
PP1c-dissociating peptide (P < 0.05) are also statistically significant.
(C) Analysis R3E mRNA levels in the livers of control and
strepto-zotocin-induced diabetic rats The R3E and control b-actin DNA
bands obtained by multiplex RT–PCR using rat R3E-specific and
b-actin-specific primers are stained with ethidium bromide and
visu-alized under UV light.
0 0.0001 0.0002 0.0003 0.0004
Glycogen synthase phosphatase acti
+RVXF -RVXF
0 0.01 0.02
0.03
A
C
B
Phosphorylase phosphatase acti
control diabetic
diabetic +96 h i
+RVXF -RVXF
β-actin R3E
Trang 9seen for the GLglycogen-targeting subunit of PP1 [18].
GL, which is highly expressed in rodent liver but only
present at very low levels in rodent skeletal muscle, is
found at appreciable levels in human skeletal muscle
(as well as in human liver) The finding that two
glyco-gen-targeting subunits are highly expressed in human
skeletal muscle while being present at only very low
levels in rodent skeletal muscle may underlie a
funda-mental difference in the regulation and function of
gly-cogen-bound PP1 in skeletal muscle in humans and
rodents
The observation that R3E appeared to be
predomin-antly expressed in insulin-sensitive tissues, led to
inves-tigation of whether this protein is regulated by insulin
in vivo Although no evidence was found for acute
regulation via phosphorylase a as seen for PP1-GL,
PP1-R3E associated phosphorylase phosphatase and
GSP activities were substantially decreased in the livers
of diabetic rats and these activities were restored by
insulin treatment The similar decreases in activity
observed for PP1-GL and PP1-R5⁄ PTG in the livers
of diabetic animals was found to correspond to a
decrease in protein and mRNA levels for their
glyco-gen-targeting subunits [34] Because R3E protein was
barely detectable in liver lysates (Figs 4B and 5A) by
either of two different antibodies, it was not possible
to directly confirm a decrease in R3E protein in the
livers of streptozotocin diabetic rats by
immunoblot-ting However, examination of R3E mRNA levels
demonstrated a decrease to below detectable levels in
the livers of diabetic rats It therefore appears that
hepatic R3E, like GL and hepatic R5⁄ PTG, is
regula-ted at the transcriptional level by insulin and that R3E
mRNA and consequently protein levels are decreased
in streptozotocin diabetic animals
The novel PPP1R3G appears to be expressed at low
levels exclusively in brain as judged from mRNA
blot-ting and detection in brain cDNA libraries This
situ-ation is unusual, in that other PP1 glycogen-targeting
subunits are either expressed at low levels ubiquitously
or are present at significant levels in insulin-sensitive
tissues such as liver and skeletal muscle However,
gly-cogen is a major energy reserve in brain astrocytes and
glycogen mobilization is tightly coupled to neuronal
activity [41]
Conservation of the amino acid sequence of R3G
from human to rodents suggests that, like R3E, it may
perform a distinct and critical function The generation
of mice lacking the gene encoding the major striated
muscle glycogen-targeting subunit of PP1, GM, has
pro-vided evidence to suggest that there is insufficient
com-pensatory response from other subunits because mice
lacking the GM subunit have only 10% muscle
glyco-gen compared with their wild-type littermates [24,25] The homozygous deletion of PTG⁄ R5 ⁄ PPP1R3C has recently been reported to be embryonic lethal [42] Mice heterozygous for this deletion have decreased glycogen stores and GS activity in muscle, liver and adipose tis-sue Glucose intolerance, hyperinsulinaemia and insulin resistance were also observed to develop with increasing age These results indicate that PTG performs a critical role that cannot be undertaken by the other glycogen-targeting subunits The development of mice lacking particular subunits may, therefore, uncover whether there is any functional redundancy among the other glycogen-targeting subunits of PP1
The high levels of GL and PPP1R3E mRNA in human compared with rodent skeletal muscle indicates that rodents may not be appropriate models from which to gain an understanding of the hormonal regu-lation of human skeletal muscle GSP In addition, this species-specific difference in the expression of PP1 reg-ulatory subunits is likely to be relevant to the study of the mechanism of action of insulin on human skeletal muscle and liver glycogen synthesis and the pathophy-siology of human type 2 diabetes
Materials and methods
Amplification of PPP1R3E and PPP1R3G from human cDNA libraries
Full-length coding sequences of PPP1R3E and PPP1R3G were amplified from human brain and testis Matchmaker cDNA libraries (Clontech, Palo Alto, CA, USA) by two rounds of PCR using the Advantage GC-cDNA polymerase and instructions (Clontech) PPP1R3E was amplified by an initial PCR with the forward primer 1 (nucleotides )105 to )84) 5¢-GAAGCGGACCCACGGACTTCTG-3¢ and the re-verse primer 2 (complementary to nucleotides 957–937 5¢-GA CTCCCTTGGACCGCTCCCG-3¢), followed by a second round of PCR with the forward primer 3 (nucleotides 1–21 5¢-ATGTCCGCTGAGCGGCCCCCG-3¢) and the reverse primer 4 (complementary to nucleotides 837–815 5¢-GATA AAGTGGATCCAGCCCCATAGGGGCGCGG-3¢) and the reverse primer 8 (complementary to nucleotides 1074–1058 5¢-GAGCGCGTCCGCAGGGCACGC-3¢) PCR products were resolved on 1% (w⁄ v) agarose gels, gel-purified, cloned into pCR2.1 TOPO vector (Invitrogen, Carlsbad,
CA, USA) and sequenced in both directions using M13 for-ward and reverse primers DNA sequencing was performed
in conjunction with the Sequencing Service managed by
Dr Nick Helps (School of Life Sciences, University of Dundee; http://www.dnaseq.co.uk) using an Applied Biosystems 373 A DNA sequencer or Big-Dye Ver 3.1 chemistry on an Applied Biosystems model 3730 automated capillary DNA sequencer
Trang 10RNA analyses
Northern blots (Clontech) contained 2 lg poly(A)+
RNA from different tissues of fed rats post mortem and human
tis-sues collected from fed individuals no more than 3 h after
death Blots were hybridized a32P-labelled cDNA probes
according to the manufacturer’s instructions with the final
wash in 10 mmolÆL)1 NaCl, 1.5 mmolÆL)1 sodium citrate,
0.1% SDS, at 55C Levels of R3E and control b-actin mRNA
transcripts were assessed in total rat liver RNA by multiplex
RT-PCR (Promega, Madison, WI, USA) as described
previ-ously [34] The rat R3E specific forward and reverse primers
were 5¢-ATGTCCCGTGAGCGGCCCCCG-3¢ and 5¢-GAT
AAAGTGGATCCAGCCCTGCG-3¢, respectively
Treatment of animals
Diabetes was induced with either intravenous or
intraperi-toneal injection of streptozotocin into male Wistar rats
and insulin was subsequently administered intravenously
into some of the rats for 96 h [34] Blood glucose levels
were elevated ‡ fourfold in diabetic animals prior to
insu-lin treatment The rats were killed by suffocation in CO2
and tissues were excised, freeze-clamped, and stored at
)80 C All procedures were performed in accordance with
the guidelines of the ethical committees of the University
of Dundee or the Katholieke Universiteit Leuven
Immunological techniques
Homogenization of tissues was performed as detailed in
Munro et al [18] Homogenates were centrifuged at 16 000 g
for 10 min, and the supernatants were snap-frozen in liquid
nitrogen and stored at )80 C Preparation of subcellular
fractions was performed as detailed in Browne et al [34]
Proteins were separated by 10% SDS⁄ PAGE, transferred to
nitrocellulose, and probed with affinity purified antibodies
Peptides were synthesized by G Bloomberg (University of
Bristol, UK); antibodies were raised in sheep by Diagnostics
Scotland (Penicuik, Midlothian, UK) and affinity purified in
conjunction with the Division of Signal Transduction
Ther-apy, University of Dundee coordinated by H McLauchlan
and J Hastie Antibodies to human PP1b peptide (amino
acids 316–327) and human PPP1R3E(8–23) were affinity
purified against their respective peptides Antibodies to
human GST-PPP1R3E(1–98) were affinity purified against
MBP-PPP1R3E Immunoblotting followed by detection of
immunoreactive bands by enhanced chemiluminescence was
performed as described in Munro et al [18]
Protein phosphatase assays
PP1 activities were determined by release of [32P]phosphate
from phosphorylase a (10 mmolÆL)1, phosphorylated by
phosphorylase kinase) and GS (1 mmolÆL)1, phosphoryl-ated by GSK3) in the presence of 4 nm okadaic acid for
10 min at 30C For immunoadsorption of PP1-GL, PPP1R5 and PPP1R3E with GL, R5 and anti-R3E sera, respectively, lysates were prepared in the pres-ence of 100 nm okadaic acid Immune pellets from 100 lg
of liver lysate were washed five times in the presence of
4 nmolÆL)1okadaic acid, and PP1 activities in the immune pellets were assayed as described above either before (‘spontaneous’ activity) or after (‘total’ activity) preincuba-tion with 0.1 mgÆmL)1 ‘dissociating’ peptide (GKRTNLR KTGSERIAHGMRVKFNPLALLLDSC) that causes the release of free PP1c from the glycogen-targeting subunit [34,39] One unit of activity is the amount of enzyme that catalyses the release of 1 mmol of [32P]phosphate per minute Statistical significance was assessed using the Student’s t-test
Acknowledgements
The work was supported by the UK Medical Research Council, UK and Diabetes UK SM was initially the recipient of a Cooperative Awards in Science and Engineering postgraduate studentship from the Bio-technology and Biological Research Sciences Council,
UK and Novo Nordisk, Bagsvaerd, Denmark Subse-quently, SM was supported on a postdoctoral research assistantship by Diabetes UK HC is a postdoctoral fellow of the Fund for Scientific Research-Flanders
References
1 BeckNielsen H (1998) Mechanisms of insulin resistance
in non-oxidative glucose metabolism: the role of glyco-gen synthase J Basic Clin Physiol Pharmacol 9, 255–279
2 Shulman GI, Rothman DL, Jue T, Stein P & Defronzo RAS (1990) Quantitation of muscle glycogen synthesis
in normal subjects and subjects with non-insulin-depen-dent diabetes by C-13 nuclear magnetic-resonance spec-troscopy New Eng J Med 322, 223–228
3 Cohen P (1999) The Croonian Lecture 1998 Identifica-tion of a protein kinase cascade of major importance in insulin signal transduction Phil Trans R Soc Lond B
354, 485–495
4 Shepherd PR, Nave BT & Siddle K (1995) Insulin sti-mulation of glycogen synthesis and glycogen synthase activity is blocked by wortmannin and rapamycin in 3T3-L1 adipocytes: evidence for the involvement of phosphoinositide 3-kinase and p70 ribosomal protein-S6 kinase Biochem J 305, 25–28
5 Ingebritsen TS, Foulkes JG & Cohen P (1983) The pro-tein phosphatases involved in cellular regulation 2 Gly-cogen metabolism Eur J Biochem 132, 263–274