(Received for publication, April 25, 1997, and in revised form, May 23, 1997)
From the Eppley Institute for Research in Cancer,
University of Nebraska Medical Center, Omaha, Nebraska 68198-6805 and
the § Program in Molecular Medicine, University of
Massachusetts Medical Center, Worcester, Massachusetts 01605
Tyrosine kinases of the Src family are regulated via their Src homology 2 (SH2) and SH3 domains. The Nef protein of human immunodeficiency virus-1 (HIV-1) has previously been shown to bind with high affinity and specificity in vitro to the SH3 domain of Hck, a Src family member expressed primarily in myeloid cells. However, the effect of Nef on Hck activity in living cells is unknown. Here we show that Rat-2 fibroblasts co-expressing Hck and Nef rapidly developed transformed foci, whereas control cells expressing either protein alone did not. Nef formed a stable complex with Hck and stimulated its tyrosine kinase activity in vivo. Mutagenesis of the Nef proline-rich motif essential for SH3 binding completely blocked complex formation, kinase activation, and transformation, indicating that the Nef SH3-binding function is required for its effects on Hck. These results provide direct evidence that SH3 engagement is sufficient to activate a Src family kinase in vivo and suggest that Hck may be activated by Nef in HIV-infected macrophages.
The Src family of non-receptor tyrosine kinases consists of nine members (Src, Lck, Hck, Fyn, Fgr, Yes, Blk, Lyn, and Yrk) that share common structural features and regulation (reviewed in Ref. 1). The N-terminal region of each kinase contains a signal for myristoylation and in some cases palmitoylation; both lipid modifications are involved in membrane targeting. Adjacent to the N-terminal region is a Src homology (SH3)1 domain consisting of approximately 60 amino acids. SH3 domains bind to target proteins via specific proline-rich sequences that adopt a polyproline type II helical conformation (2-4). Protein-protein interactions mediated by Src kinase SH3 domains contribute to substrate recruitment, subcellular localization, and negative regulation of kinase activity (see below). C-terminal to the SH3 domain is the SH2 domain. Comprised of approximately 100 amino acids, SH2 domains bind tightly to specific tyrosine-phosphorylated sequences. As for SH3 domains, SH2 domains also contribute to negative regulation of kinase activity, subcellular localization, and interaction with signaling partners. The SH2 domain is followed by the kinase domain and a C-terminal tail with a highly conserved tyrosine residue essential to kinase regulation.
SH2 and SH3 domains cooperate in the negative regulation of Src family kinase activity (1). Phosphorylation of the conserved tyrosine residue in the C-terminal tail (Tyr-527 in Src) by the regulatory kinase Csk (5) is postulated to promote intramolecular binding to the SH2 domain, contributing to a closed and inactive kinase conformation. A large body of genetic and biochemical evidence supports this model. For example, conversion of the conserved Tyr residue in the tail to Phe or mutation of the SH2 domain results in a constitutively active kinase capable of producing a transformed phenotype in rodent fibroblasts (6, 7). Such mutations presumably interfere with intramolecular SH2-tail interaction. Many studies have shown an essential role for the SH3 domain in the negative regulation of Src family kinase activity as well. Mutants of Src lacking a functional SH3 domain are also resistant to Csk-mediated negative regulation and are often transforming (8-12). These findings suggest that interaction of Src kinase SH2 or SH3 domains with target proteins may destabilize the closed conformation in vivo, leading to kinase activation. Such a mechanism may be responsible for the activation of Src kinases following recruitment via their SH2 domains to autophosphorylated growth factor receptors (13, 14). Whether or not SH3-mediated interaction of Src kinases with target proteins is sufficient to activate the kinase domain in vivo is less clear.
In this study, we investigated the effects of SH3 domain engagement on the activity of a Src-related kinase in living cells. Our model system consisted of Hck, a Src family member expressed primarily in macrophages and neutrophils (15-17), and the Nef protein of the human immunodeficiency virus-1 (HIV-1). Nef is essential for the high titer replication of HIV and SIV and for the induction of AIDS-like disease in monkeys (reviewed in Refs. 18 and 19). Previous reports have shown that Nef specifically binds to the SH3 domain of Hck in vitro with the highest affinity known for an SH3-mediated interaction (20, 21). These findings suggested that Nef may represent an excellent probe of Hck SH3 function in vivo. As shown below, co-expression of Nef with Hck in Rat-2 fibroblasts resulted in the appearance of a transformed phenotype that correlated with the SH3-dependent activation of Hck in vivo. These results suggest that interaction of Src family kinases with SH3 ligands may be sufficient to activate Src family kinases in vivo.
The Nef gene from HIV isolate SF2 was amplified by
polymerase chain reaction and subcloned into the retroviral vector
pSRMSVtkneo (22). Wild-type (WT), kinase-defective (K269E), and
activated (Y501F) forms of human p59 Hck were also subcloned into the
same vector. The mutant forms of Hck were generated by a standard
polymerase chain reaction-based strategy as described elsewhere (23).
The resulting constructs were used to generate high titer retroviral stocks by co-transfection of 293T cells (24) with an ecotropic packaging vector. Control retroviruses were prepared using the parent
vector. For each infection, 2.5 × 104 Rat-2 cells
were plated per well in 6-well tissue culture plates and incubated with
viral stocks in a final volume of 6 ml. To enhance the efficiency of
infection, cultures were centrifuged at 1,000 × g for
4 h at 18 °C (25) in the presence of 4 µg/ml Polybrene.
Following infection, the virus was replaced with fresh medium, and G418
(400 µg/ml) was added 48 h later. Transformed foci were counted
after 2 weeks.
To detect Hck-Nef protein complexes, cells were sonicated briefly in 1.0 ml of detergent buffer (50 mM Tris-HCl, pH 7.4, 50 mM NaCl, 1 mM EDTA, 10 mM MgCl2, and 1% Triton X-100). Clarified supernatants were incubated with 2 µg of anti-Hck antibody (Santa Cruz Biotechnology) and 15 µl of protein G-Sepharose (50% slurry; Pharmacia Biotech Inc.) for 2.5 h at 4 °C. Immunoprecipitates were washed three times with 1.0 ml of RIPA buffer (50 mM Tris-HCl, 150 mM NaCl, 1% Triton X-100, 0.1% SDS, 1 mM EDTA, and 1% sodium deoxycholate). Immunoprecipitates were heated in SDS-PAGE sample buffer, resolved by SDS-PAGE, transferred to polyvinylidene difluoride membranes, and probed for associated Nef by immunoblotting with anti-Nef monoclonal antibodies (kindly supplied by Dr. James A. Hoxie, University of Pennsylvania).
For analysis of protein-tyrosine phosphorylation or protein expression, whole cell lysates were prepared by heating cells directly in SDS-PAGE sample buffer. Lysate proteins were separated by SDS-PAGE, transferred to polyvinylidene difluoride, and immunoblotted with anti-Hck monoclonal antibodies (Transduction Laboratories), anti-phosphotyrosine antibodies (PY20; Transduction Laboratories), or anti-Nef monoclonal antibodies. Immunoreactive bands were visualized colorimetrically using alkaline phosphatase-conjugated secondary antibodies and nitro blue tetrazolium/5-bromo-4-chloro-3-indolyl phosphate as substrate.
To
determine whether interaction with Nef influences Hck activity under
physiological conditions, Rat-2 fibroblasts were co-infected with
recombinant retroviruses containing wild-type, activated, or
kinase-defective forms of Hck and either a Nef retrovirus or a control
virus containing only the bacterial neo resistance marker. After 2 weeks, transformed foci were readily observed in cells expressing wild-type Hck and Nef (Fig. 1 and Table
I). In contrast, no transformed foci were observed in
cells expressing wild-type Hck or Nef alone. The foci resulting from
the co-expression of Nef and Hck were morphologically indistinguishable
from those produced by an activated allele of Hck in which the negative
regulatory tyrosine residue in the C-terminal tail was mutated to
phenylalanine (Hck-YF mutant) (Fig. 1). Fewer foci were observed with
the co-infected cells relative to the Hck-YF cells (Table I), probably
because transformation is dependent upon successful co-infection and
expression of both Hck and Nef within the same cell. No transformed
foci were observed following co-expression of Nef with a
kinase-defective Hck mutant (Hck-KE; Ref. 23), demonstrating the
requirement for Hck tyrosine kinase activity in the initiation of
transformation by Nef. We also observed that the transforming activity
of Hck-YF was not detectably affected by co-expression with Nef (Table
I), despite the ability of these two proteins to associate in
vivo (see below).
|
To verify that the Nef and Hck proteins were expressed in the
transformants as well as the non-transformed control cells, immunoblots
were performed on whole cell lysates. As shown in Fig.
2, Hck and Nef proteins were readily detected in cells
infected with the corresponding retroviruses. Taken together, these
results demonstrate that Nef and Hck cooperate to induce fibroblast
transformation in a Hck kinase-dependent manner.
Hck Forms a Stable Complex with Nef Resulting in Constitutive Activation of the Hck Tyrosine Kinase in Vivo
The biological data
described above support a model in which Nef binds to the Hck SH3
domain in vivo, leading to activation of the kinase domain
and cellular transformation. To test this model, we assessed Hck
autophosphorylation and tyrosine phosphorylation of endogenous proteins
in lysates from control and transformed cell lines. Cells were lysed
directly in SDS-PAGE sample buffer, and tyrosine phosphoproteins were
visualized by immunoblotting with anti-phosphotyrosine antibodies. As
shown in Fig. 3, transformed cells expressing both Hck
and Nef showed several prominent tyrosine-phosphorylated proteins.
These include an ~60-kDa phosphoprotein with electrophoretic mobility
identical to Hck as well as a second strong band of ~40 kDa. A nearly
identical pattern of protein-tyrosine phosphorylation was observed in
cells transformed with the activated form of Hck (Hck-YF), including
the 40-kDa transformation-related protein. Although we have not
identified the endogenous 40-kDa transformation-associated tyrosyl
phosphoprotein, immunoblotting experiments indicate that it is not the
Src substrates c-Crk, annexin I, or annexin II which are in this
molecular mass range (data not shown). None of the untransformed
control cell lines showed evidence of protein-tyrosine phosphorylation,
including cells expressing Nef alone, Hck-WT alone, or cells
co-expressing Nef and the kinase-defective Hck mutant. These results
show that co-expression with Nef leads to constitutive activation of
the Hck tyrosine kinase in living cells.
We next investigated whether Nef directly interacts with Hck in
vivo. Hck was immunoprecipitated from transformed cells and probed
for associated Nef by immunoblotting. As shown in Fig. 4, Hck-Nef complexes were readily immunoprecipitated
from the Rat-2 cells transformed by co-expression of these proteins.
Nef also formed stable complexes with the kinase-defective and
activated forms of Hck, indicating that the SH3 domains of all three
Hck proteins are accessible to Nef. Control experiments show that Nef
does not cross-react with the anti-Hck antibody. These experiments provide the first evidence for stable Hck-Nef complexes in intact cells. However, the possibility that the observed complexes form in
solution after lysis of the cells cannot be formally ruled out.
Activation of Hck by Nef and Cellular Transformation Requires the Proline-rich Region of Nef
A final series of experiments
investigated the dependence of Hck kinase activation and cellular
transformation on the SH3-binding activity of Nef. Hck was co-expressed
with a mutant of Nef in which the proline residues at positions 72 and
75 were converted to alanine (Nef-PA mutant) (20). Mutagenesis of these
prolines has been shown previously to completely block Nef-SH3 binding in vitro (20). Recent structural studies of Nef-SH3
complexes have confirmed that these Nef residues form part of the
polyproline type II helix essential for interaction with the SH3 domain
(26, 27). Co-expression of Hck with the Nef-PA mutant failed to induce transformation (Fig. 1, Table I), despite strong expression of both
proteins in the co-infected cultures (Fig. 5).
Furthermore, Nef-PA was unable to stimulate Hck kinase activity
in vivo as assessed by tyrosine phosphorylation of the
endogenous protein, p40 (Fig. 5). In addition, Hck autophosphorylation
was not stimulated by this Nef mutant (data not shown). Consistent with
previous in vitro work, the Nef-PA mutant was unable to bind
to Hck in the co-immunoprecipitation assay (Fig. 5). Taken together,
these findings strongly support the hypothesis that Nef binds directly to the SH3 domain of Hck in vivo, leading to Hck tyrosine
kinase activation and cellular transformation.
Data presented in this report show that engagement of a Src family kinase SH3 domain by a target protein leads to kinase activation, tyrosine phosphorylation of substrate proteins, and cellular transformation. These data provide direct new evidence that SH3 domains contribute to the negative regulation of Src family kinases in vivo. While this manuscript was being prepared for submission, the three-dimensional crystal structures of the inactive, Csk-phosphorylated forms of both c-Src and Hck were reported (28, 29). Both structures revealed the anticipated intramolecular interaction between the SH2 domain and the tyrosine-phosphorylated tail. Unexpectedly, both structures also showed an intramolecular interaction of the SH3 domain with a helical region formed by the linker connecting the SH2 and kinase domains. In light of this new structural information, the most likely mechanistic explanation for our data is that engagement of the SH3 domain by Nef disturbs this intramolecular interaction, releasing the kinase from the inactive, closed state. Such a mechanism implies that SH3-mediated activation can occur in the absence of dephosphorylation of the tail and supports a central role for the SH3 domain in negative regulation. Also supporting this view is the recent observation that Nef can stimulate Hck tyrosine kinase activity in vitro despite the presence of a phosphotyrosine residue in the tail region (30). SH3 domains of Src family kinases have been reported to interact with a wide variety of substrate proteins via their SH3 domains (23, 31-33). Our data suggest that binding of a substrate protein to the SH3 domain may be sufficient to stimulate Src tyrosine kinase activity in vivo, possibly inducing substrate phosphorylation.
Work presented here is consistent with the notion that Nef, which lacks any known catalytic function, acts by influencing kinases or other signaling pathways within the host cell (reviewed in Ref. 18). Also supporting this hypothesis are previous studies of a variant of SIV that may signal by interacting with Src. However, in contrast to our data, the unusual acutely lethal phenotype of this variant (SIVpbj14; Ref. 34) is dependent upon the presence of an SH2 recognition sequence encoded by its Nef gene that is phosphorylated by Src (35). The SIVpbj14 Nef allele was able to transform NIH3T3 cells, although the role of Src activation in the transformation event was not investigated directly (35). SH2-dependent activation of Src may not be required for HIV Nef function in general, since the SIVpbj14 Nef SH2 recognition motifs that govern its acutely pathogenic properties are not found in the Nef genes of known HIV-1 strains (36). We were unable to detect tyrosine phosphorylation of Nef in the fibroblasts transformed by co-expression of Hck and Nef, suggesting that SH2 domain interaction is not involved in the activation mechanism (data not shown).
Our data show that stimulation of Hck protein-tyrosine kinase activity by HIV-1 Nef is sufficient to produce a transformed phenotype in fibroblasts. Given the strong conservation of proline-rich motifs among various Nef alleles, these data argue that SH3-dependent activation of Src family tyrosine kinases may be a general property of primate lentiviruses. Our results, together with previous reports of the remarkable affinity and specificity of HIV-1 Nef for the Hck SH3 domain (20, 21), strongly suggest that Hck may be a relevant target for Nef in macrophages, a major site of Hck expression (15-17). While the role of macrophages in lentivirus replication and pathogenicity is unclear, recent evidence suggests that macrophage infection is necessary for the pathogenic manifestations of both HIV and SIV infection (reviewed in Ref. 37). Hck is rapidly induced following macrophage activation and has been implicated in multiple signaling events including phagocytosis, Fc receptor signal transduction, integrin signaling, and tumor necrosis factor release (38-44). Further investigation should determine which, if any, of these signaling pathways are modulated by HIV infection and whether these effects are mediated through Nef-Hck interaction.
We thank Dr. Owen Witte, Howard Hughes Medical Institute, UCLA, for the retroviral vectors and Dr. James A. Hoxie, University of Pennsylvania, for the anti-Nef antibody.