|
|
||||||||

* CNRS UA D-1284, Neurobiologie Moléculaire, Institut Pasteur, 75724 Paris Cedex, France; and
Centre de Recherche au Neurobiologie Cellulaire et Moléculaire, University of Liege, Belgium
1Correspondence: CNRS UA D-1284, Neurobiologie Moléculaire, Institut Pasteur, 2528 rue du Dr. Roux, 75724 Paris Cedex, France. E-mail: honghiem{at}pasteur.fr
| ABSTRACT |
|---|
|
|
|---|
-32P]-TTP, 43K rapsyn is
specifically phosphorylated with a 32P-half-maximal
incorporation at ~525 µM TTP. The presence of TTP in the cytosol
and of 43K rapsyn at the cytoplasmic face of the postsynaptic membrane,
together with TTP-dependent phosphorylation of 43K rapsyn without added
exokinases, suggests that TTP-dependent-43K-rapsyn phosphorylation may
occur in vivo. In addition, phosphoamino acid and
chemical stability analysis suggests that the residues phosphorylated
are predominantly histidines. Inhibition of phosphorylation by
Zn2+ suggests a possible control of 43K rapsyn
phosphorylation state by its zinc finger domain. Endogenous kinase(s)
present in rodent brain membranes can also use
[
-32P]-TTP as a phosphodonor. The use of a
phosphodonor (TTP) belonging to the thiamine family but not to the
classical (ATP, GTP) purine triphosphate family represents a novel
phosphorylation pathway possibly important for synaptic
proteins.Nghiêm, H.-O., Bettendorff, L., Changeux, J.-P.
Specific phosphorylation of Torpedo 43K rapsyn by
endogenous kinase(s) with thiamine triphosphate as the phosphate donor.
Key Words: nicotinic acetylcholine receptor (nAChR) cytoskeleton neuromuscular junction (NMJ) synapse phosphohistidine
| INTRODUCTION |
|---|
|
|
|---|
Phosphorylation is important in cell signaling (reviewed in refs
10
11
12
13
14
). 43K rapsyn, which contains several putative
phosphorylation sites (15)
, is partially phosphorylated on
serine residues in vivo and phosphorylated in
vitro by endogenous protein kinase A (PKA) (16)
.
However, this phosphorylation is not specific for 43K rapsyn and can
occur with other proteins of the postsynaptic membrane
(16)
. In view of the essential roles of phosphorylation in
cell signaling (rfs. in 1014) and of 43K rapsyn in postsynaptic
differentiation (reviewed in refs 1
, 2
, 17
), we searched
for a specific phosphorylation of this synaptic protein.
Thiamine is essential to cell life and may play a role in the central
nervous system and in synaptic transmission (18
19
20)
. The
thiamine pathway includes thiamine and its mono- (TMP), di- (TDP), and
triphosphate (TTP) derivatives. TTP, the non-cofactor form of thiamine
activates the maxi-chloride channel permeability, possibly via
phosphorylation (21)
. TTP concentrations are low in most
cells (22)
, but comparatively high in neuronal
(23)
and excitable (24
25
26)
cells. Thus, TTP
is an interesting candidate for a specific phosphorylation of 43K
rapsyn.
In this study, we demonstrate that in the presence of radiolabeled TTP, Torpedo 43K rapsyn is the predominant protein phosphorylated by endogenous kinase(s) present in nAChR-rich postsynaptic membrane preparations. Phosphorylation occurs mostly at histidine(s) and at some serine(s). Both TTP- and ATP-dependent phosphorylation of 43K rapsyn are inhibited by TTP and ATP. TTP-dependent kinase(s) thus might share some phosphorylation site(s) with PKA. The possible regulation of 43K rapsyn phosphorylation by endogenous Zn2+ and the modulation of 43K rapsyn functions via its phosphorylation state are discussed. The extension of phosphorylation to rodent brain membranes suggests a more general use of TTP as phosphate donor for synaptic proteins and a possible novel phosphorylation pathway.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Rodent brain membrane preparations
Brain membrane preparations were performed at 4°C. Mice and
rats were anesthetized then killed by decapitation. The brains were
dissected and homogenized with a Teflon glass homogenizer in 5 volumes
of ice-cold Tris-buffer pH 7.5 containing 10% sucrose (w/w), 1 mM
EDTA, 1 mM DTT, and inhibitors of proteases (aprotinin, pefabloc,
leupeptin, antipain, pepstatin A, PMSF). The homogenates were
centrifuged at 1000 g for 5 min at 4°C. The supernatants
were further centrifuged at 30,000 g for 1 h at 4°C.
The resulting pellets corresponding to the crude brain membrane
fractions were homogenized in ice-cold homogenization buffer without
DTT and stored at -80°C.
Phosphate donors, phosphorylation, and quantification
[
-32P]-ATP
(32P-ATP) was from ICN (Irvine, Calif.).
[
-32P]-TTP (32P-TTP)
was synthesized (27)
. nAChR membranes were phosphorylated
with (78000Ci/mol) 32P-TTP or
32P-ATP in 50 mM Tris-HCl pH 7.5, 515 mM
MgCl2, 0.08% CHAPS, inhibitors of proteases at
420°C for 6090 min. Phosphorylation was stopped with sodium
dodecyl sulfate (SDS) sample buffer.
32P-phosphorylated membranes were subjected to
SDS-PAGE (polyacrylamide gel electrophoresis) designed to separate
actin, 43K rapsyn, and
-nAChR and autoradiographed (Kodak Biomax)
and/or 32P-quantified (Molecular Dynamics
phosphorImager). Coomassie blue staining was performed when necessary.
Where specified, nAChR membranes were treated with 520 mM
diethylpyrocarbonate (DEPC) (28)
in 50 mM Na phosphate
buffer pH 6.0 and 7.4 (20 min, 16°C) prior to incubation with
32P-TTP. Common kinase effectors [cAMP;
adenosine 3'-5'-cyclic monophosphate, 8-(4-chlorophenylthio)-sodium
salt (8-CPT-cAMP); anisomycin; cGMP; calmidazolium; calphostin; cdc2
peptide; genistein; bisindolylmaleimide I (GFX); H7; H89; KN62; KT5720,
ML7; protein kinase A inhibitor (PKI); staurosporine; tumor necrosis
factor-alpha (TNF-
); phorbol-12-myristate-13-acetate (TPA)] were
tested for their effects on TTP-dependent phosphorylation of 43K
rapsyn.
Chemical stability and nature of the phosphate links
For acid treatment, SDS-PAGE gels containing
32P-ATP- or 32P-TTP-treated
membranes were cut and incubated with Tris buffer or 16% TCA at 90°C
(29
, 30)
, washed, and analyzed. Equivalent amounts of 43K
rapsyn were ensured by Coomassie blue staining. For base treatment,
32P-labeled membranes were resolved by SDS-PAGE,
electroblotted on polyvinylidene difluoride membrane (PVDF). Blots were
dried at 55°C to minimize protein loss, wet in methanol, washed with
H2O, cut, and incubated in water or 1N KOH at
46°C, and analyzed.
Phosphoamino acid analysis on PVDF-electrotransferred
(31)
32P-43K rapsyn
For determination of acid-stable phosphoamino acids, 43K rapsyn
was hydrolyzed with 40 µl 5.7N HCl (1 h, 105°C). The supernatant
was evaporated and 10 µl H2O was added.
Hydrolysates were analyzed by either 1-dimensional (pH 3.5) or
2-dimensional high-voltage electrophoresis on a thin-layer cellulose
plate (first electrophoresis, pH 1.9; second electrophoresis, pH 3.5)
(32)
. For a base-stable analysis, 43K rapsyn was
hydrolyzed in 3N KOH (3 h, 105°C), neutralized with 10%
HClO4 to pH 7.5 (33)
. Supernatants
were analyzed by thin-layer chromatography (TLC) on silica gel 60A°
plates (ICN) in solvent A (t-butanol:methyl ethyl
ketone:acetone:methanol:water:concentrated NH4OH,
10: 20: 20: 5: 40: 5, v :v), which separates phosphohistidine
(P-His) from phosphoserine (P-Ser) and phospholysine (P-Lys)
(33)
. Phosphohistidine and phospholysine were synthesized
from polyhistidine and polylysine, respectively (34)
.
Enzymatic hydrolysis was conducted with 2 µg TPCK-trypsin (Promega,
Madison, Wis.) in 40 µl of trypsin buffer (10 mM
NaHCO3, 135 mM NaCl, 0.1% SDS, 1 mM
CaCl2, pH 8.5) (90 min, 37°C). Two micrograms
of TPCK-trypsin was added (2 h, 37°C), followed by 400 µg Pronase
(Boehringer Mannheim, Mannheim, Germany) (18 h, 37°C). Supernatants
were analyzed by TLC in solvent A. Phosphoamino acids and
phosphopeptides were visualized with ninhydrin.
Phosphopeptides
Phosphopeptides were generated by tryptic digestion on
PVDF-transferred 32P-43K rapsyn with TPCK-trypsin
(o.n.; 37°C in trypsin buffer). Hydrolysates were resolved in 15%
SDS-PAGE and autoradiographed for 32P-peptide
identification.
Labeling with antibodies or
-Bungarotoxin (Bgtx)
32P membranes were resolved by SDS-PAGE,
electroblotted (35)
, treated according to ref
36
, probed with specific anti-43K rapsyn
(37)
, specific antiphosphoamino acid antibodies (Sigma,
St. Louis, Mo.), or 125I-Bgtx (Amersham,
Arlington Heights, Ill.), and analyzed.
Immunoprecipitation
32P-membranes were diluted into 1 ml 50 mM Tris-HCl
pH 8.8, 0.1% SDS, 1% Nonidet P-40, 0.5% deoxycholate, protease
inhibitors, 0.15M NaCl, precleared with 50 µl protein, A-agarose
beads (Santa Cruz, Santa Cruz, Calif.), and immunoprecipitated with
anti-43K rapsyn antipeptide antibodies that specifically recognize 43K
rapsyn (37)
. Thirty microliters of protein A beads were
added (o.n., 4°C). Beads were centrifuged, washed, and analyzed.
| RESULTS |
|---|
|
|
|---|
-32P]-TTP in the presence of various
effectors, separated by SDS-PAGE, and autoradiographed (Fig. 1B
|
A sister gel of 1A was blotted, and the blotted
32P-labeled membrane corresponding to lanes 1 to
3 and to various degrees of membrane phosphorylation were further
incubated with 125I-Bgtx (a toxin specific for
-nAChR), then autoradiographed (Fig. 1C
). With
phosphorylated membranes, two radioactive bands were observed (Fig. 1C
, lanes 1, 3). In lane 2, with phosphorylation prevented,
only one radioactive band corresponding to the
125I-Bgtx-labeled band (arrowhead) and distinct
from the 32P-labeled 43 kDa band (arrow) was
observed. This demonstrates that
-nAChR is not phosphorylated by
32P-TTP.
The 32P-labeled band was recognized by anti-43K
rapsyn antibodies (immunoblot, data not shown). To ascertain that the
32P-phosphorylated protein is 43K rapsyn,
immunoprecipitation of 32P-labeled membranes was
conducted with three specific anti-43K rapsyn antipeptide antibodies
(37)
. Figure 2A
shows that the 32P-labeled protein
was specifically immunoprecipitated by anti-43K rapsyn antibodies. One
anti-43K rapsyn antibody (Fig. 2A
, lane 1) used in a
semiquantitative analysis (Fig. 2B
) showed that the
radioactivity immunoprecipitated is directly correlated to the amount
of anti-43K rapsyn used (Fig. 2B
, lanes 35). Supernatants
of immunoprecipitation showed the opposite situation (Fig. 2C
). The specificity of the immunoprecipitation was verified
with preimmuneserum and preabsorbed antibodies (Fig. 2B
,
lanes 1, 2). This demonstrates that 43K rapsyn is the TTP-dependent
phosphorylated protein.
|
The TTP-dependent-phosphorylation is driven by endogenous kinase(s)
present in the nAChR-rich membranes
The phosphorylation of 43K rapsyn that occurs at 422°C without
externally added kinases is Mg2+ (5 mM, Fig. 1B
, lanes 3, 6), pH, and time dependent (Fig. 3
). It requires TTP, is dose dependent and saturable
(KD ~510 µM TTP, Fig. 3C
; with
one membrane preparation, KD ~25 µM TTP), and
presents characteristics of an enzymatic reaction. Thus, the
TTP-dependent phosphorylation of 43K rapsyn is driven by endogenous
kinase(s) copurified with the postsynaptic membrane. The
IC50 for TTP, ATP, GTP are around 40 µM, 500
µM, and 1000 µM, respectively (Fig. 3A
). CTP inhibits
poorly. The TTP-dependent kinase or TTP-43K-kinase activity is favored
by light alkaline pH (Fig. 3B
) and partially inhibited by
DTT (3040% inhibition/10 mM; Fig. 1B
, lane 1).
Zn2+ inhibits the TTP-dependent phosphorylation
of 43K rapsyn in a Mg2+ independent manner
[~70% inhibition/0.53 mM Zn2+/8 µM
32P-TTP and (Fig. 1B
, lane 2)].
|
TTP-43K-kinase is not PKA
Although 43K rapsyn is, within the detection sensitivity, the
only protein phosphorylated in the presence of
32P-TTP (Figs. 1
and 4
, lane 1), additional
proteins, including nAChR subunits, are phosphorylated with
32P-ATP (Fig. 4
, lane 2), in agreement with our former results (16)
. 43K
rapsyn phosphorylation is saturated with ~2550 µM
32P-TTP (Fig. 3C
) but is not saturated
with 200 µM 32P-ATP.
32P-ATP- and
32P-TTP-dependent 43K rapsyn phosphorylation are
specifically inhibited by both TTP and ATP, suggesting the presence of
common phosphorylation sites between PKA and TTP-kinase. However,
analysis conducted with PKA effectors showed that they are different.
PKI inhibited PKA (60±13% inhibition) but not TTP-kinase (6±1%
inhibition). Exogenous PKA catalytic subunit increased the
ATP-dependent phosphorylation (603±14% 32P vs.
100±23% in control) (ref 16
; this study) while
inhibiting that driven by TTP (41±6% vs. 100±2% in control).
|
TTP-43K-kinase, a novel kinase?
Putative phosphorylation sites (15)
for PKA [Ser-406
(38)
] and tyrosine kinase [Tyr-98, Tyr-189, Tyr-325
(39)
] are present on Torpedo 43K rapsyn.
Searches on Prosite (40)
and PhosphoBase (41)
showed putative sites for CaMKI, CKI, CKII, PKA, PKC, and PKG protein
kinases. Of 18 common kinase effectors, only staurosporine caused a
slight inhibition (33±3% inhibition/200 nM). Numerous activators or
inhibitors of PKA, PKC (TPA, calphostin, GFX), MAP kinase, Protein
kinase G, CaM kinase II, JNK2 kinase, cdc2 kinase, MLCK, SAP kinase,
and TyrPK (data not shown) did not drastically alter the activity of
TTP-kinase, which is likely of a novel type.
Nature of the amino acids phosphorylated with TTP
A 2-dimensional high-voltage electrophoresis of acid hydrolysates
of 32P-ATP-dependent, phosphorylated 43K rapsyn
(ATP-32P-43K rapsyn) has shown that
phosphorylation by PKA occurs predominantly on serine(s)
(16)
. Similar analysis on TTP-dependent
32P-phosphorylated 43K rapsyn
(TTP-32P-43K rapsyn) showed different results,
with a faint radioactive signal at serine and a strong one at inorganic
phosphate (Pi). No detectable radioactivity was observed at the
phosphothreonine (P-Thr) or phosphotyrosine (P-Tyr) region (data not
shown). Similar results conducted by immunoblot analysis on
TTP-dependent 32P-labeled nAChR-rich membranes with
specific antiphosphoamino acid antibodies were obtained (data not
shown). Anti-P-Tyr strongly stained several nonradioactive bands, but
not 43K rapsyn. This suggests the presence of in situ
Tyr-phosphorylated proteins and nAChR-associated protein tyrosine
kinases (43)
, and also suggests that Tyr is not
phosphorylated in 43K rapsyn [but see (44)
]. No staining
of 32P-43K rapsyn was observed with anti-P-Thr.
Anti-P-Ser faintly stained 43K rapsyn both in control [this is
consistent with the presence of in situ P-Ser in 43K rapsyn
(16)
] and in TTP-32P- membranes. A
stronger staining of 32P-43K rapsyn suggests some
phosphorylation on serine driven by TTP and is consistent with the
reciprocal inhibition of ATP- and TTP-dependent phosphorylation of 43K
rapsyn by TTP and ATP, respectively.
To gain insight into the unexpected high 32Pi
content in TTP-32P-43K rapsyn hydrolysate, ATP-
and TTP-32P-43K rapsyn were simultaneously
hydrolyzed with HCl and analyzed by 1-dimensional electrophoresis.
Similar ninhydrin-stained phosphopeptide patterns but different
autoradiograms were obtained (Fig. 5
). ATP-32P-43K rapsyn hydrolysate (lane 1) led to
high radioactivity at P-Ser (arrowhead) and low radioactivity at Pi.
TTP-32P-43K rapsyn hydrolysate (lane 3) showed
very faint radioactivity at P-Ser (arrowhead), and high radioactivity
at Pi. The radioactivity in the P-Thr or P-Tyr regions appeared mainly
as a trailing background. This confirms serine phosphorylation with ATP
and suggests that phosphorylation with TTP occurs predominantly on
residues other than serine; furthermore, TTP driven phospholinkages are
mainly acid labile.
|
A pH stability analysis was further performed on ATP- and
TTP-32P-43K rapsyn. SDS-PAGE gels containing both
phosphoproteins were treated with TCA at 90°C and
32P-quantified (Table 1)
. TTP-dependent phosphorylated 43K rapsyn is acid sensitive and the
32P-phosphate loss is a function of time in TCA
(50±4 and 16±1% 32P after 5 and 10 min vs.
100±13% for control). In contrast, ATP-32P-43K
rapsyn is less sensitive (79±5 and 49±9% 32P
after 5 and 10 min vs. 100±8% for control). A similar test conducted
at alkaline pH (Table 1)
showed a remarkable stability of the
TTP-dependent phospholinkages (72±4% 32P after
2 h in 1N KOH at 46°C vs. 100±5% for control) in contrast with
the ATP-driven phospholinkages (18±2 vs.
100±6%32P in control). Thus the phosphate links
elicited by ATP are acid stable and alkaline labile, a signature of
O-linked phosphoamino acids phosphoserine and phosphothreonine
(45)
. Serine is indeed phosphorylated with ATP
(16
; foregoing text). Conversely the phosphoryl linkages
introduced by TTP are acid labile and alkaline stable, a characteristic
of N-phosphate linkages at phosphohistidine or phospholysine
(45)
.
|
TTP causes phosphorylation predominantly on histidine residues.
To identify the N-phosphoamino acids in TTP-phosphorylated 43K
rapsyn, a TLC of 43K rapsyn hydrolysates was performed in solvent A.
Nucleoside diphosphate kinase (NDPK) (46)
which
autophosphorylates histidine (47)
was used as control
(Fig. 6A
, B
, lanes 4). All hydrolysates (Figs. 6A
, B
,
lanes 24) displayed radioactive material migrating similarly to
phosphohistidine, the highest intensity being observed with the
enzymatic hydrolysate from TTP-32P-43K rapsyn
(Fig. 6A
, lane 2). The low radioactivity at "P-His" in
both TTP-32P-43K rapsyn (Fig. 6A
, lane
3; Fig. 6B
, lanes 2, 3) and
ATP-32P-NDPK alkaline hydrolysates (Figs. 6A
, B
, lane 4), probably derived from P-His partial
destruction during hydrolysis. Added phosphohistidine (internal
standard) comigrated with the radioactive spots (Fig. 6B
,
lanes 24). These results favor phosphorylation on histidine with TTP.
|
To assess the importance of histidine(s), nAChR membranes were
pretreated with DEPC, then incubated with 32P-TTP
[DEPC modifies histidines, thus preventing their subsequent
phosphorylation (28)
]. 43K rapsyn phosphorylation was
effectively decreased in DEPC membranes (20±2% vs. 100±19%
32P in mock membranes).
Partial tryptic digestion conducted on ATP- and TTP-32P-43K rapsyn, followed by 15% acrylamide SDS-PAGE, showed one major radioactive band at ~6.515 kDa for ATP- and several radioactive bands from ~6.5 to 35 kDa for TTP-phosphorylated 43K rapsyn (data not shown). Again, this indicates different phosphorylation sites depending on the nature of the phosphodonor. ATP probably leads to phosphorylation mainly on one serine residue; with TTP, one or several histidine residues might be mainly phosphorylated.
TTP is not a phosphodonor for NDPK
NDPK is a highly conserved enzyme that plays a key role in growth
and metastasis control (47)
. As the enzyme
autophosphorylates histidine and presents a broad specificity,
phosphorylation was assayed with TTP. NDPK was strongly phosphorylated
with 32P-ATP but not with
32P-TTP (data not shown).
TTP, a phosphate donor in the central nervous system (CNS)
Mouse and rat brain membranes incubated with
32P-TTP or 32P-ATP were
phosphorylated. However, as with Torpedo postsynaptic
membranes (Fig. 7A
, B
, lanes 1), ATP phosphorylated many proteins in mouse
brain membranes (Fig. 7A
, lane 2) whereas TTP phosphorylated
very few. Two major 32P-labeled bands were
observed at ~4346 kDa (Fig. 7B
, lane 2). Phosphorylation
was partially inhibited by ATP (Fig. 7A
, lane 3) or TTP
(Fig. 7B
, lane 3). Thus, in vitro, TTP is also a
phosphodonor for proteins in the CNS.
|
| DISCUSSION |
|---|
|
|
|---|
The TTP-dependent kinase(s), (a) novel kinase(s)?
Specific phosphorylation of 43K rapsyn with TTP as phosphodonor
occurs at 430°C, temperatures compatible with that of the sea water
surrounding the Torpedo. 43K rapsyn is localized at the
postsynaptic membrane inner face (5
, 6)
, hence
topologically accessible to the high cytosolic TTP content (~430
nmol/g wet tissue; 24, 26). Thus, conditions necessary for a successful
endogenous phosphorylation of 43K rapsyn are met, supporting the notion
that phosphorylation of 43K rapsyn with TTP as phosphoryl donor occurs
in vivo in Torpedo electrocytes.
The phosphorylation is Mg2+ and TTP dependent, with characteristics of an enzymatic reaction driven by endogenous kinase(s) present in the nAChR-rich postsynaptic membrane and specific for TTP, although with some affinity for ATP. They were named TTP-dependent-43K rapsyn kinase(s) or TTP-kinase(s).
These TTP-kinase(s) seem to be of a novel type, different from PKA, PKC, or common kinases. Their affinity is not drastically affected by inhibitors of PKA, activators (TPA) or inhibitors (calphostin, GFX) of PKC, or effectors of other common kinases. The question of their classification in a new eukaryotic protein kinase family or as members of known kinase families will be solved with their identification, purification, characterization, and sequencing. However, the possibility of a TTP-dependent autophosphorylation of 43K rapsyn must also be considered.
Histidine phosphorylation of 43K rapsyn
Phosphoamino acid and antibody analysis suggests that besides some
minor phosphorylation on serine residues, histidines are predominantly
phosphorylated. The inhibition of TTP-dependent 43K phosphorylation by
both ATP and TTP suggests that TTP-kinase might share some common
phosphorylation sites with PKA. However, since most of the detectable
phosphoryl groups introduced by TTP are on histidine(s) and those by
ATP on serine (16)
, the inhibition through shared serine
site(s) should be only partial. The strong inhibition of
phosphorylation by high concentrations of heterologous triphosphate ATP
may reside on the ability of TTP-kinase to recognize and link either
ATP or TTP, although with different affinities. In addition, a
modification of the histidine(s) microenvironment brought about by
phosphorylation of serine(s) might occur and result in a decrease of
histidine phosphorylation by TTP-kinase(s) (Ser-406, a strong PKA
consensus site, is close to His-384 and His-387).
Zn2+ modulates the activity of many proteins and
may play a role in synaptic transmission (48)
; we have
shown that TTP-kinase activity is prevented by 500 µM
Zn2+. At its carboxyl terminus, ahead of Ser-406,
43K rapsyn displays two zinc finger motifs that could be important for
nAChR clustering (42
, 49
, 50)
. In addition, two conserved
histidines, His-384 and His-387, are present in the zinc finger motifs.
In vitro, 43K rapsyn binds Zn2+
probably through the two histidines (42)
, which
consequently might become less available for an eventual
phosphorylation. Binding of Zn2+ might also
elicit conformational changes inducing a decrease of 43K rapsyn
accessibility for histidine phosphorylation by TTP-kinase. If
Zn2+ binds to 43K rapsyn in vivo, the
zinc finger domain might play a role in regulation of the protein
phosphorylation state. An intrinsic sensitivity of TTP-kinase to
Zn2+ should account only partially at these
Zn2+ concentrations.
Tryptic digestions suggest that ATP probably leads to the
phosphorylation of one main serine (possibly Ser 406, a strongly
conserved PKA consensus site) whereas TTP may drive phosphorylation on
one or several histidines. 43K rapsyn possesses 13 histidines that are
potential candidates. Ten of these residues are conserved among chick
(51)
, human (52)
, mouse (53)
,
Torpedo (54)
, Xenopus
(55)
. Some of the conserved histidines also have their
neighboring sequence conserved: e.g., His-154; His-239; His-256;
His-384 and His-387 of the tandem zinc fingers. Highly homologous,
although not totally conserved, neighboring sequences of His-53,
His-329, His-348, and His-353 are located in regions possibly important
for 43K rapsyn functions. His-53 is present in a domain involved in 43K
rapsyn self-association (56)
. Mutations of His-384 and
His-387 reduce 43K rapsyn ability to form clusters (42)
.
His-348 and His-353 are located between these two important regions of
43K rapsyn. The neighboring sequence RYAH of His-154 is conserved
in K. aerogenes (57)
, N.
meningitidis (58)
, and E. coli
(59)
and has been identified as a phosphorylation site
essential for polyphosphate kinase activity in prokaryotes
(59)
. The phosphate in phosphohistidine is of a high
energy state and is often further transferred to an acceptor residue
(on the same or another molecule), an important step in the
two-component signaling mechanisms in cell regulation (60
, 61)
. It will be of interest to identify the histidine(s)
phosphorylated by TTP and determine by mutational analysis whether a
similar role of histidine phosphorylation can be related to 43K rapsyn
phosphorylating and clustering functions in the postsynaptic domain.
TTP-dependent phosphorylation of 43K rapsyn, TTP-kinase(s), and
nAChR clustering.
43K rapsyn is present as cytosolic and membrane-attached pools in
a ratio depending on tissue maturation (37)
. The question
of a relationship between 43K rapsyn phosphorylation and its cellular
compartmentalization is raised.
nAChR phosphorylation has been reported in several instances
(62
63
64
65)
. 43K rapsyn regulates tyrosine phosphorylation of
several postsynaptic membrane proteins, including the nAChR ß-subunit
(44)
. nAChR tyrosine phosphorylation regulates the rapid
rate of receptor desensitization and may play a role in nerve-induced
nAChR clustering (65
66
67)
. Two protein tyrosine kinases
associated with the nAChR have been cloned in Torpedo
electrocyte (43)
. The TTP-kinase(s) that drive specific
phosphorylation(s) of 43K rapsyn predominantly on histidine(s) are also
present in nAChR-rich postsynaptic membrane. Their purification (see
above) will allow further analysis of their possible involvement in the
cascade responsible for nAChR phosphorylation and clustering.
Of 18 common protein kinase effectors tested, only staurosporine, a
potent but nonspecific protein kinase inhibitor (68)
,
causes some inhibition. Staurosporine also inhibits agrin-induced nAChR
phosphorylation and aggregation (69)
. This raises the
question of an eventual connection between these events and 43K rapsyn
phosphorylation via TTP. Agrin plays an important role in NMJ
differentiation (70
71
72)
. Cotransfected 43K rapsyn causes
clustering of dystroglycan, the agrin binding component of the
dystrophin glycoprotein complex (73)
. It also induces
clustering and activation of MuSK, a synapse-associated muscle specific
kinase (74
75)
and component of the agrin-MuSK-MASC
signaling complex responsible for nAChR clustering and postsynaptic
differentiation. It will be interesting to study the influence of
TTP-dependent phosphorylation on the involvement of 43K rapsyn in the
agrin-dystroglycan-MuSK-MASC cascade.
Phosphorylation of 43K rapsyn through TTP also suggests the possibility
of an interplay between 43K rapsyn and the thiamine pathway in
excitable cells. Increased nervous activity leads to dephosphorylation
from TTP and TDP to TMP and thiamine (18
, 76)
, and
deafferentation of the cerebellum decreases turnover of thiamine
phosphate derivatives (77)
.
Extension of TTP-dependent phosphorylation to other eukaryotic
systems. TTP, a phosphodonor for mammalian synaptic proteins?
Occurrence of the TTP-dependent phosphorylation of 43K rapsyn at
the vertebrate NMJ remains to be defined as well as its potential role
in proteinprotein interactions, nAChR aggregation, and stabilization
at the NMJ.
Although TTP is not a phosphodonor for NDPK histidine [despite NDPKs
broad specificity (47)
], TTP can cause phosphorylation of
proteins present in rodent central nervous membranes. TTP thus
represents a valuable tool for defining a possibly novel
phosphorylation pathway specific for synaptic proteins.
43K rapsyn causes clustering of cotransfected
GABAA receptors (78)
and is present
in chick ciliary ganglion neurons (51)
. Analysis of a
possible involvement of TTP as a phosphodonor in the phosphorylation of
brain receptors, chick ciliary ganglion 43K rapsyn and putative brain
43K rapsyn homologues should permit a better understanding of the
molecular processes underlying synaptic functions.
The novel and specific TTP-dependent phosphorylation of 43K rapsyn highlights the possible importance of TTP-dependent phosphorylation in the modulation of synaptic organization. It also opens up a new phosphorylation pathway for synaptic proteins that differs from the more classical purine triphosphate pathway.
Histidine phosphorylation in eukaryotes
In eukaryotes, phosphorylation has been estimated to occur
predominantly on serine residues (~90%), ~9.9% on threonine
residues and only ~0.1% on tyrosine residues despite its key role in
cell modulation (32). In prokaryotes phosphorylation on histidine
(~6%) has been documented and often related to regulation processes
(61)
. Fewer cases are reported in eukaryotes
(79)
. Our findings add a new example of histidine
phosphorylation on a synaptic protein in eukaryotic cells, thus
broadening the importance of histidine in eukaryotic phosphorylation.
| CONCLUSION |
|---|
|
|
|---|
| ACKNOWLEDGMENTS |
|---|
| FOOTNOTES |
|---|
| REFERENCES |
|---|
|
|
|---|
1 polypeptide from Torpedo marmorata electric organ. Proc. Natl. Acad. Sci. USA 80,6403-6407
1) polypeptide in Torpedo marmorata electrocyte: support for a role in acetylcholine receptor stabilization. Development 113,1059-1067[Abstract]
subunit. J. Biol. Chem. 262,10506-10510This article has been cited by other articles:
![]() |
M. Gangolf, P. Wins, M. Thiry, B. El Moualij, and L. Bettendorff Thiamine Triphosphate Synthesis in Rat Brain Occurs in Mitochondria and Is Coupled to the Respiratory Chain J. Biol. Chem., January 1, 2010; 285(1): 583 - 594. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Song, L. Bettendorff, M. Tonelli, and J. L. Markley Structural Basis for the Catalytic Mechanism of Mammalian 25-kDa Thiamine Triphosphatase J. Biol. Chem., April 18, 2008; 283(16): 10939 - 10948. [Abstract] [Full Text] [PDF] |
||||
![]() |
J.-Y. Liu, D. E. Timm, and T. D. Hurley Pyrithiamine as a Substrate for Thiamine Pyrophosphokinase J. Biol. Chem., March 10, 2006; 281(10): 6601 - 6607. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. E. Schweingruber The Melaminophenyl Arsenicals Melarsoprol and Melarsen Oxide Interfere with Thiamine Metabolism in the Fission Yeast Schizosaccharomyces pombe Antimicrob. Agents Chemother., September 1, 2004; 48(9): 3268 - 3271. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. Lakaye, B. Wirtzfeld, P. Wins, T. Grisar, and L. Bettendorff Thiamine Triphosphate, a New Signal Required for Optimal Growth of Escherichia coli during Amino Acid Starvation J. Biol. Chem., April 23, 2004; 279(17): 17142 - 17147. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. Lakaye, A. F. Makarchikov, A. F. Antunes, W. Zorzi, B. Coumans, E. De Pauw, P. Wins, T. Grisar, and L. Bettendorff Molecular Characterization of a Specific Thiamine Triphosphatase Widely Expressed in Mammalian Tissues J. Biol. Chem., April 12, 2002; 277(16): 13771 - 13777. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |