(Received for publication, August 31, 1996, and in revised form, November 4, 1996)
From the Department of Biochemistry, Vanderbilt University School of Medicine, Nashville, Tennessee 37232-0146
The intraperitoneal injection of a
vanadate/H2O2 mixture (peroxovanadate) into
mice resulted within minutes in the appearance of numerous
tyrosine-phosphorylated proteins in the liver and kidney. These effects
are presumably due to the inhibition of phosphotyrosine
phosphatase activity. Three of the tyrosine-phosphorylated proteins
have been identified as the receptors for epidermal growth factor,
insulin, and hepatocyte growth factor. The injection of peroxovanadate
also enhanced the tyrosine phosphorylation of many of the proteins
known to function downstream of these receptors, including SHC, signal
transducer and activator of transcription (Stat) 1,
, Stat 3, Stat
5, phospholipase C-
, insulin receptor substrate 1, GTPase-activating
protein,
-catenin,
-catenin, p120cas, SHP-1, and SHP-2.
The administration of peroxovanadate also induced nuclear translocation
of a number of tyrosine-phosphorylated Stat proteins. In addition, the
global effects on tyrosine phosphorylation permitted the detection of a
number of novel intracellular protein interactions, including an
association of Tyk2 with
-catenin. The in situ
administration of peroxovanadate may prove useful in the search for
novel tyrosine-phosphorylated proteins and the identification of new
interactions between previously identified tyrosine-phosphorylated
substrates.
Many of the cellular responses induced by hormones, growth factors, and cytokines are mediated by the activation of intracellular kinases. Some of these kinases are capable of autophosphorylation and of phosphorylating specific proteins on tyrosine residues. The timely appearance and disappearance of these tyrosine-phosphorylated proteins is critical for proper cell function. The tight regulation of this process is accomplished by the ubiquitous presence of tyrosine phosphatases. To detect tyrosine-phosphorylated proteins in cell extracts, specific phosphatase inhibitors, such as sodium vanadate (1), must be added. It is of interest that vanadium salts, in addition to being potent tyrosine phosphatase inhibitors, have been reported to have "insulin-like" and other biological effects both in intact animals and in cell cultures (reviewed in Refs. 2, 3, 4).
H2O2, like vanadate, mimics some of the actions of insulin in a number of cell culture systems (5, 6, 7) and, like vanadate, inhibits protein tyrosine phosphatase activity (8); in the latter instance this is presumably due to oxidation of essential sulfhydryl groups on these enzymes. The unexpected synergistic effect of vanadate/H2O2 mixtures (peroxovanadate) on intact cells as a potent insulin-mimetic agent and as an inhibitor of protein tyrosine phosphatase has been reported in a number of studies (9, 10, 11, 12).
In an attempt to amplify the tyrosine phosphorylation of potential
substrates of the insulin receptor, peroxovanadate, other peroxovanadium compounds, and insulin were injected intravenously into
anesthetized rats (13, 14, 15). Treatment with peroxovanadates resulted in:
1) the rapid tyrosine phosphorylation of multiple liver proteins,
including the 95-kDa -subunit of the insulin receptor,
IRS-1,1 and PLC-
, and 2) an inhibition
of phosphotyrosine phosphatase activity.
We have previously reported that the intraperitoneal injection of EGF
into mice leads to a rapid increase in the tyrosine phosphorylation of
a number of proteins in all organs examined (16). Using this in
situ system we were able to identify the EGF receptor, SHC and the
transcription factors Stat 1,
, Stat 3, and Stat 5 as EGF-induced
tyrosine-phosphorylated proteins in liver (17, 18, 19, 20).
The identification of cellular proteins that are phosphorylated on tyrosine in response to extracellular activating ligands has been hampered by their low abundance and the ubiquitous presence of tyrosine phosphatases (21). To circumvent these problems we combined our in situ system, which provides large quantities of material, with the potent phosphatase inhibitor peroxovanadate. In this report we demonstrate that the intraperitoneal injection of vanadate/H2O2 mixtures (peroxovanadate) into mice, in the absence of any added ligand, induces the rapid and massive tyrosine phosphorylation of multiple cellular proteins in both the liver and kidney. Not only were previously identified EGF-responsive tyrosine-phosphorylated proteins detected after peroxovanadate treatment, but we also were able to detect the enhanced tyrosine phosphorylation of receptors and downstream effectors of other hormones, growth factors, and cytokines. In addition, the extensive tyrosine phosphorylation facilitated the detection of novel intracellular interactions.
ND4 Swiss Webster mice were obtained from
Harlan-Sprague-Dawley. Immobilon-P membranes were from Millipore.
Prestained molecular weight standards were from Life Technologies, Inc.
The following polyclonal antibodies were used: Tyk2, Stat 5, and Met
(Santa Cruz); GTPase-activating protein (UBI); Fer (a gift from T. W. Wang, University of Medicine and Dentistry of New Jersey); insulin receptor and IRS-1 (gifts from C. R. Kahn, Joslin Diabetes Center); EGF
and EGF receptor (this laboratory); PLC- (a gift from G. Carpenter,
Vanderbilt University); SHP-1 and SHP-2 (a gift from Z. Zhao,
Vanderbilt University); Stat 1 (a gift from J. Larner, U.S. Food and
Drug Administration, Bethesda, MD); horseradish peroxidase-conjugated
goat anti-rabbit and goat anti-mouse antibodies were obtained from
Cappel. The following monoclonal antibodies were used: RC20H, Stat
1
,
, Stat 3, Stat 5,
-catenin,
-catenin, and tensin
(Transduction Laboratories); anti-mouse IgG was obtained from
Transduction Laboratories. Sodium vanadate (orthovanadate) was from
Fisher, enhanced chemiluminescence reagent (ECL) was from Amersham
Corp., and protein A-Sepharose and all other reagents were from
Sigma.
A 5 mM solution of sodium vanadate in PBS was prepared by heating to boiling (22). Fifteen minutes prior to use, 30% H2O2 was added to the vanadate solution at room temperature to a final concentration of 50 mM. Solutions of this peroxovanadate or PBS were injected intraperitoneally into adult mice at a dose of 10 µl/g of body weight. Mice were sacrificed in a CO2 chamber at the indicated times and the tissues were removed and immediately frozen in liquid nitrogen. A 10% (wet weight/volume) tissue lysate was prepared by Dounce homogenization in Buffer A (50 mM Tris, pH 7.4, 0.15 M NaCl, 1% Triton X-100, 0.25% deoxycholate, 1 mM EDTA, 1 mM sodium vanadate, and 50 µM sodium molybdate). When EGF was used as a ligand, solutions of EGF (1 mg/ml) were injected intraperitoneally at a dose of 10 µl/g of body weight.
Mouse liver nuclei were isolated, purified by centrifugation through 2.2 M sucrose, and extracted with 0.2 M sodium chloride as described previously (18).
Western Blotting and ImmunoprecipitationPortions (10 µl)
of tissue extracts were separated by SDS-PAGE (7.5%), transferred to
Immobilon-P membranes, and probed with RC20H (1:2500). Antibody binding
was detected by ECL. Aliquots of the nuclear extracts (25 µl) were
similarly resolved by SDS-PAGE (7.5%), transferred to Immobilon-P
membranes, and probed with RC20H. Blots were stripped and reprobed
sequentially with monoclonal antibodies to Stat 1,
, Stat 3, and
Stat 5. In each instance antibody binding was detected by incubation
with horseradish peroxidase-conjugated goat anti-mouse antibody and
ECL.
For immunoprecipitation, lysates of liver from control and peroxovanadate-treated animals were centrifuged at 100,000 × g for 30 min and precleared by incubating 1 ml of extract with 50 µl of packed protein A-Sepharose for 30 min at 4 °C. When monoclonal antibodies were used for precipitation, aliquots of cleared lysate (300 µl) were incubated with 3 µl of the specific antibody (0.75 µg) for 2-4 h or overnight at 4 °C followed by the addition of 3 µg of goat anti-mouse IgG for 1 h and 50 µl of protein A-Sepharose (50% slurry) for 1 h. The resulting precipitates were washed three times in Buffer A, and the bound proteins were eluted by boiling in 100 µl of 2 × Laemmli buffer for 5 min. When polyclonal antibodies were used for immunoprecipitation, 300 µl of cleared lysate were incubated with 3 µl of specific antibody (3 µg) for 2 h followed by the addition of 50 µl of protein A-Sepharose (50% slurry) for 1 h. Aliquots were resolved by SDS-PAGE (7.5%), transferred to Immobilon-P membranes, and immunoblotted with RC20H or the specified antibody.
Adult
mice were treated by intraperitoneal injection of PBS alone or of PBS
containing 5 mM sodium vanadate, 50 mM
H2O2, or a mixture of 5 mM sodium
vanadate plus 50 mM H2O2
(peroxovanadate) at a dosage of 10 µl/g of body weight. After 20 min,
the livers and kidneys were excised and frozen in liquid nitrogen.
Extracts were prepared and the proteins present were separated by
SDS-PAGE and analyzed by Western blotting with anti-phosphotyrosine
antibodies. In both organs the administration of peroxovanadate
resulted in the tyrosine phosphorylation of many proteins (Fig.
1). The relative extent of tyrosine phosphorylation of
the various bands detected in liver and kidney clearly differed. In
preliminary experiments the increase of tyrosine phosphorylation in
both the liver and kidney could be detected as early as 2 min
postinjection and persisted for at least 60 min (data not shown). The
administration of sodium vanadate or of H2O2
alone had little effect on the tyrosine-phosphorylated protein content
in either the liver or kidney (Fig. 1). Only slight alterations of
protein tyrosine phosphorylation were detected in the lung, spleen, and
brain following peroxovanadate treatment (data not shown). Enhanced
tyrosine phosphorylation in liver could be detected at sodium vanadate
concentrations as low as 0.5 mM, and the extent of tyrosine
phosphorylation gradually increased as the vanadate concentration was
raised to 5 mM (data not shown). Intraperitoneal injection
of mixtures of potassium chromate, sodium tungstate, or potassium
permanganate with H2O2 at similar dosages resulted in no significant alteration of tyrosine-phosphorylated protein content in liver (data not shown).
Identification of Liver Proteins Whose Tyrosine Phosphorylation Is Induced by Peroxovanadate
To understand the biochemical
interactions that led to the massive enhancement of tyrosine
phosphorylation induced by peroxovanadate in the absence of any added
activating ligand, we sought to identify the proteins affected. These
exploratory experiments were carried out by preparing detergent lysates
of livers from control and peroxovanadate-treated adult mice and
immunoprecipitating them with an antibody specific for the protein of
interest. The precipitated proteins were separated by SDS-PAGE and
analyzed by -phosphotyrosine immunoblot.
We have identified the presence of the tyrosine-phosphorylated EGF
receptor (pp170), the -subunit of the insulin receptor (pp95), and
the receptor for hepatocyte growth factor (Met, pp140) in livers of
peroxovanadate-treated animals (Fig. 2). The 52-kDa tyrosine-phosphorylated protein that is present in antibody
precipitates of the EGF receptor was identified as SHC by reprobing the
blot with antibodies specific for SHC. Immunoprecipitation using
anti-SHC antibodies resulted in a pattern of tyrosine phosphorylation
qualitatively identical to that shown for anti-EGF receptor in Fig. 2
(data not shown). Thus, the in situ response to
peroxovanadate resembles the in situ response to EGF
(17).
In view of this similarity we directly compared the pattern of
tyrosine-phosphorylated proteins in liver extracts from PBS-, EGF-, and
peroxovanadate-treated mice. As demonstrated previously (16), 20 min
following the administration of EGF an increase in the phosphotyrosine
content of a number of liver proteins was detectable (Fig.
3). It is readily apparent that liver extracts from
peroxovanadate-treated animals contain more tyrosine-phosphorylated proteins than extracts from animals induced by EGF (Fig. 3).
In view of the enhanced tyrosine phosphorylation (and possible
activation) of receptors for EGF, hepatocyte growth factor, and insulin
in livers of peroxovanadate-treated animals, it was of interest to
determine whether known downstream substrates of these receptor/kinases
(reviewed in Refs. 23, 24, 25) were also tyrosine-phosphorylated. Livers of
peroxovanadate-treated animals were found to contain markedly higher
levels of tyrosine-phosphorylated Stat 5 (pp92), Stat 1 (pp91),
PLC-
(pp150), IRS-1 (pp185), and GTPase-activating protein (pp120)
(with its associated proteins of 190 and 64 kDa) than controls (Fig.
4). The identities of the other tyrosine-phosphorylated
proteins detected in these immunoprecipitates are not known.
Proteins that mediate cell adhesion, such as the cadherin-associated
proteins -catenin (pp92),
-catenin (pp82), and pp120cas
and cytoskeletal proteins such as talin (pp120), also have been shown
to be tyrosine-phosphorylated following the activation of EGF or
hepatocyte growth factor receptors in cell cultures (26, 27, 28, 29, 30). Enhanced
tyrosine phosphorylation of
-catenin,
-catenin, and several
isoforms of p120cas are demonstrable in the liver of mice
following intraperitoneal injection of EGF (data not shown). All of
these putative signal transducing proteins are tyrosine-phosphorylated
in the livers of peroxovanadate-treated animals (Fig.
5). The bands noted in immunoprecipitates of
pp120cas are presumably isoforms (25). The identities of the
higher molecular weight bands associated with
-catenin in Fig. 5 are not known.
Among the possible cellular targets responsible for initiating the
enhanced tyrosine phosphorylation observed following treatment with
peroxovanadate are the intracellular tyrosine phosphatases. Tyrosine
phosphatases that contain Src homology 2 domains (SHP-1 and SHP-2) have
been implicated as regulators of signal transduction and tyrosine
phosphorylation of SHP-1, and SHP-2 has been demonstrated in growth
factor- and cytokine-stimulated cells (reviewed in Refs. 31 and 32).
Livers of peroxovanadate-treated animals contain both SHP-1 (pp68) and
SHP-2 (pp72) in a tyrosine-phosphorylated form (Fig. 6).
Both phosphatases appear to be associated with major
tyrosine-phosphorylated proteins in the 120-140-kDa range. The
association of SHP-1 and SHP-2 with receptors and other
tyrosine-phosphorylated proteins in various cell culture systems has
been reported (32). The tyrosine-phosphorylated bands labeled
a and b (Fig. 6) were identified as SHP-1 and
SHP-2, respectively, by Western blotting with antibodies specific for
these proteins and were present in both control and
peroxovanadate-treated animals.
Finally, we have examined liver extracts for the presence of
tyrosine-phosphorylated cytoplasmic tyrosine kinases such as Jak1,
Jak2, Tyk2, and Fer. Small amounts of all four were detected as
tyrosine-phosphorylated proteins in peroxovanadate-treated animals
(data not shown for Jak1 and Jak2). Perhaps the most interesting observations (Fig. 6) were that immunoprecipitates of Tyk2 (pp135, band c) also contained a tyrosine-phosphorylated protein
identified as -catenin (pp92, band d), and
immunoprecipitates of Fer (pp95, band f) contained
tyrosine-phosphorylated pp120cas (band e). All of
these proteins were identified by reprobing the blots with appropriate
antibodies. The association of
-catenin with Tyk2 and
p120cas with Fer also were seen in extracts of livers from
control animals in the absence of detectable tyrosine phosphorylation
of these molecules (data not shown). The association of Fer with
pp120cas in A431 cells stimulated by EGF or platelet-derived
growth factor, as well as the association of Fer with unphosphorylated
p120cas, were reported previously (33).
Since the
administration of either EGF or growth hormone to animals induces the
tyrosine phosphorylation and nuclear translocation of Stat proteins in
livers (18, 34), we compared the patterns of protein tyrosine
phosphorylation in salt extracts of liver nuclei from control,
peroxovanadate-treated, and EGF-treated animals. Both EGF and
peroxovanadate induced the appearance of major tyrosine-phosphorylated bands of Mr ~86,000-92,000 (Fig.
7A). Additional bands were detected in nuclei
from peroxovanadate-treated animals.
We have previously identified phosphorylated Stat 1,
, Stat 3, and
Stat 5 in liver nuclei as components of the 86-92-kDa tyrosine-phosphorylated bands detected after administration of EGF
(20). Western blots of nuclear extracts from control and peroxovanadate-treated animals revealed the presence of Stat 1
,
, Stat 3, and Stat 5 proteins in nuclei from peroxovanadate-treated animals and not in nuclei from control animals (Fig. 7B).
Thus, peroxovanadate mimics some of the nuclear effects induced by the administration of either EGF or growth hormone.
The activation of transmembrane and cytoplasmic tyrosine kinases by growth factors, hormones, and cytokines initiates many intracellular signaling events. Increased attention has recently been focused on the importance of protein tyrosine phosphatases in the regulation of cell function (reviewed in Refs. 31, 35, and 36).
Our experiments describing the in situ effects of peroxovanadate suggest that the control of protein tyrosine phosphatase activity may be of greater significance in cell signaling than heretofore realized. The intraperitoneal injection of a vanadate/H2O2 mixture (peroxovanadate) into mice resulted within minutes in the appearance of many tyrosine-phosphorylated proteins in both the liver and kidney (Figs. 1 and 3). The differing patterns of phosphotyrosine-containing proteins in each organ may reflect different types and amounts of tyrosine kinases or substrates present in each tissue. The liver and kidney are the organs predominantly affected by peroxovanadate, suggesting that they are able to rapidly concentrate the biologically active component from the peritoneal cavity.
It was surprising that, in the absence of any added ligand, we were
able to detect extensive tyrosine phosphorylation of not only receptors
for EGF, insulin, and hepatocyte growth factor (Fig. 2) but also of
many of the downstream substrates of these and other kinases. The
cellular signaling components in liver whose tyrosine phosphorylation
was enhanced by the peroxovanadate treatment include SHC, Stat 5, Stat
1,
, PLC-
, IRS-1, GTPase-activating protein (with its
associated proteins of 190 and 64 kDa), and the cadherin-associated
proteins
-catenin,
-catenin, and p120cas (Figs. 4 and
5).
The Src homology 2-containing phosphotyrosine phosphatases (SHP-1 and SHP-2) are presumably among those tyrosine phosphatases inhibited by peroxovanadate. These two enzymes have been reported to be targets of a number of receptor and receptor-associated tyrosine kinases (37, 38, 39, 40, 41). The tyrosine phosphorylation of both SHP-1 and SHP-2 is enhanced in the livers of peroxovanadate-treated animals, and both are associated with other unidentified tyrosine-phosphorylated proteins (Fig. 6).
Finally, our data confirm the association of Fer with p120cas
(33) and indicate, for the first time, that Tyk2 is associated with
-catenin in liver (Fig. 5).
The mimicry of receptor activation by peroxovanadate with regard to
enhanced tyrosine phosphorylation of cytoplasmic proteins extends to
mimicry of some of the nuclear changes induced by ligand-mediated receptor activation. Peroxovanadate, like EGF, induces the tyrosine phosphorylation and nuclear translocation in liver of a number of Stat
proteins (Fig. 7, A and B). It is of interest
that peroxovanadate has been shown to induce T-cell activation and
transcription of c-fos (42), the activation of
interferon--inducible and prolactin-inducible transcription factors
(Stats), the activation of Jak kinases, and finally the expression of
interferon-
-responsive endogenous genes (43, 44, 45, 46).
The lack of ligand specificity in all of the reports suggests that peroxovanadate is globally affecting many tyrosine phosphatases and/or kinases. Although it is probable that the effects of peroxovanadate on protein tyrosine phosphorylation is due to its ability to inhibit intracellular protein tyrosine phosphatases, vanadate derivatives are known to bind and inhibit a variety of other enzymes (47). Furthermore, in view of the rapid equilibria among the many complex forms of vanadate in aqueous solutions (48), it is not possible to specifically define the structure of the relevant biologically active species. Despite these uncertainties, the potent biological effects of peroxovanadate may help to elucidate the biochemical mechanisms involved in the coordination of kinases and phosphatases in cell signaling and to uncover novel intracellular interactions. Our data also emphasize the importance of phosphatase activity on the steady-state levels of phosphotyrosine in cellular proteins and the extent to which the basal activity of cellular kinases is always "on" and is sufficient to activate many signaling pathways.