New Page 1

Overview: Cell Signaling and Transformation

2. Requirement for a glycine in the kinase domain of Sf-Stk in Friend erythroleukemia

3. Requirement for the C-terminal tyrosine of Sf-Stk in Friend erythroleukemia

Friend virus (FV) is a murine retrovirus consisting of two viruses, spleen focus forming virus (SFFV), and MuMLV which provides helper function. SFFV is an oncogenic virus, causing erythroleukemia in infected mice through the expression of a viral glycoprotein, gp55. C57Bl/6 mice are resistant to the effects of the virus, due to diminished expression of an N-terminally truncated form of the STK receptor tyrosine kinase (Sf-Stk). These mice harbor a mutation in and internal promoter which drives expression of Sf-Stk. The viral protein, gp55, has been demonstrated to interact with the erythropoietin receptor (EpoR) and Sf-Stk, suggesting the possibility that this trimeric complex drives the Epo-independent expansion of FV infected cells. This configuration is strikingly similar to that found in the chicken v-sea retrovirus, which expresses a fusion protein composed of a viral glycoprotein and the chicken homologue of STK, c-sea, and induces the development of erythroleukemia and sarcomas in chickens.

In order to study the role of STK in Friend erythroleukemia, we have developed a strategy by which we introduce Sf-Stk into primary bone marrow from C57Bl/6 mice, infect these cells with FV, and examine the growth of Epo-independent erythroid colonies. Using this approach, we have shown that introduction of wild-type Sf-Stk into primary bone marrow from C56Bl/6 mice rescues the ability of these cells to form colonies in response to FV infection. In order to map the domains of Sf-Stk required for the expansion of FV-infected cells, we generated a series of mutations in Sf-Stk and introduced these mutant forms of Sf-Stk into the primary culture system described above. The results of these studies demonstrate that Sf-Stk kinase activity is critical for the Epo-independent growth of FV-infected bone marrow cells. Furthermore, we have identified two amino acid residues in Sf-Stk that are absolutely required for this response. The first is a glycine in the activation loop of the kinase domain. This glycine is found at a position normally occupied by aspartic acid in most receptor tyrosine kinases. Furthermore, mutation of this aspartic acid in the context of the Kit receptor has been demonstrated in patients with mastocytomas, and has been demonstrated to confer constitutive kinase activity and to alter substrate specificity of the receptor. The second amino acid residue in Sf-Stk that is critical for Epo-independent growth of FV-infected cells is the c-terminal docking site tyrosine. This site has been shown to bind a number of downstream signaling molecules upon receptor activation, including Grb2, p85, Shc, SHP2 and PLCg. This work was performed primarily by Lisa Finkelstein, a former post-doctoral fellow in the laboratory.

2. Requirement for a glycine in the kinase domain of Sf-Stk in Friend erythroleukemia

As described above, we have demonstrated that a glycine in the activation loop of the kinase domain is essential for the growth of primary bone marrow cells in response to FV. Surprisingly, while this glycine is conserved among all murine strains examined, the human homologue of STK (RON), contains the more conserved aspartic acid at this residue. Furthermore, while full-length STK is resported to be transforming in a standard NIH3T3 focus formation assay, the RON receptor is not reported to be transforming in this same assay. Interestingly, mutation of the conserved aspartic acid in RON to a valine, unleashes the transforming potential of RON in a 3T3 assay. Therefore, we speculated that the difference in the transforming potential of STK and RON was due, at least in part, to the substitution of a glycine for aspartic acid in STK. In fact, when we swap the glycine for aspartic acid in STK, and perform the converse susbstitution in RON, we find that we can swap the transforming potential of these receptors. In order to determine whether this amino acid substitution results in changes in substrate specificity, we examine a number of downstream signaling pathways induced by the wild-type and mutant forms of STK and RON. While this amino acid in the kinase domain does not alter significantly the ability of STK and RON to activate the MAPK and PI3K pathways, only receptors containing the glycine are capable of activating the b-catenin pathway. Previous studies have also demonstrated that mutation of this amino acid in the Kit receptor results in the activation of Stat3 by Kit. Therefore, we are currently investigating the ability of wild-type and mutant STK and RON to activate induce the phosphorylation of Stat3. This work is being carried out by QingPing Liu and Xin Wei.

In order to determine whether the activation of b-catenin or Stat3 by Sf-Stk is required for the expansion of Friend virus-infected erythroblasts, we have taken the approach of expressing dominant negative signaling molecules in primary bone marrow from sensitive Balb/c mice followed by infection with Friend virus and the examination of Epo-independent colony formation. Using this strategy, we have demonstrated that expression of two dominant negative forms of Stat3 blocks the Epo-independent growth of Friend virus-infected bone marrow cells. However, expression of a dominant negative Stat1 does not block this response. Surprisingly, Stat3 is reported not to be phosphorylated in the spleens of Friend virus infected mice two weeks following infection. Our data indicate that phosphorylation of Stat3 occurs at earlier time points in the spleens of infected mice, consistent with a role of Stat3 in the expansion phase of the disease. We are currently testing the requirement for b-catenin in this response by utilizing dominant negative b-catenin and TCF constructs. This work is being carried out by Shuang Ni and Shihan He in the lab.

3. Requirement for the C-terminal tyrosine of Sf-Stk in Friend erythroleukemia

As described above, we have demonstrated a requirement for the C-terminal docking site tyrosine in Sf-Stk for the Epo-independent growth of Friend virus infected bone marrow cells. By altering the sequence surrounding this tyrosine from YVNV to YVHV, which binds all of the signaling molecules tested except Grb2 which binds to a consensus YxNx sequence, we proposed that Grb2 binding to this tyrosine is essential for the response of bone marrow progenitor cells to FV. In order to further characterize the role of Grb2 in the induction of Friend erythroleukemia, we utilized mice heterozygous for a targeted deletion in the gene encoding Grb2. These mice express reduced levels of Grb2 in the bone marrow and are less sensitive to FV-induced erythroleukemia in vivo as measured by spleen weight two weeks following infection. Futhermore, bone marrow from these mice support reduced numbers of Epo-independent colonies in vitro in response to Friend virus infection, and this response can be rescued by introducing wild-type Grb2 into these cells. To map the domains of Grb2 required for this response, we mutated the two SH3 domains of Grb2 either singly or together and used these mutant forms of Grb2 to rescue the response of Grb2 +/- bone marrow to Friend virus. Utilizing this approach, we have identified the C-terminal SH3 domain as a critical regulator of the response of primary bone marrow to Friend virus. This work was performed by Lisa Finkelstein and Hami Teal.

In order to determine whether the Grb2 is required to activate the Ras/MAPK pathway by STK through the recruitment of SOS, we mutated the two docking site tyrosines in STK and examined the ability of wild-type and mutant forms of STK to induce the phosphorylation of Erk and the activation of AP1 in 293T cells. Results from these studies indicated that while the kinase activity of STK was required for Erk activation, the docking site tyrosines were dispensable for this response. Interestingly, we found that the STK and RON RTKs also exhibited a significant difference in the ability to activate Erk in this system. While STK is a strong inducer of the Erk MAPK signaling pathway, overexpression of RON does not induce this response, and activation of RON with MSP induces only a weak response when compared with STK. We also demonstrated that this difference is not due to the glycine in the kinase domain of STK described above. Another difference we have observed between STK and RON is in the juxtamembrane domain. While STK harbors a short JM domain, the JM domain of RON is significantly longer and contains two tyrosines, one of which has been demonstrated to bind to Cbl. Interestingly, when we delete this extended JM domain in RON, this receptor is now a potent inducer of Erk/AP1 activity. In addition, a chimeric receptor containing the JM domain of STK and the kinase domain and c-terminal tail of RON is also a strong inducer of this pathway. Therefore, we conclude that sequences in the juxtamembrane domain of STK, rather than the Grb2 binding site, are critical for STK-induced Erk activity. This work was performed by Xin Wei.

While both SH3 domains of Grb2 bind SOS leading to activation of the Ras/MAPK pathway, the C-terminal SH3 domain binds a family of large adaptor proteins, Grb2-associated binding proteins (GABs). These adaptors contain a PH domain at the N-terminal end of the protein required for recruitment to phosphoinositides in the membrane. We found that overexpression of a PH domain, which would act as a dominant negative for the activation of PH domain-containing proteins, in primary bone marrow cells blocked the response of erythroid progenitor cells to Friend virus. The Gab family of docking proteins contains three members, Gab1, Gab2 and Gab3. All three family members are expressed in Friend virus infected spleens, however only Gab2 appears to localize to the plasma membrane when co-transfected into 293T cells with Sf-Stk and gp55. To determine which family members can support the Epo-independent growth of Friend virus infected cells, fusion proteins containing the SH2 domain of Grb2 and the C-terminal signaling domain of Gab1, 2 and 3 were expressed in primary bone marrow followed by infection with Friend virus. Only the Grb2/Gab2 fusion protein successfully supported the Epo-independent growth of Friend virus infected cells. In order to determine whether the role of the C-terminal tyrosine in Sf-Stk was to recruit Gab2, we constructed fusion proteins of Sf-Stk, without the C-terminal docking site tyrosines, and Gab1, 2 or 3. Sf-Stk/Gab2 fusion proteins successfully rescued the ability of bone marrow from resistant C56Bl/6 mice to grow in response to infection with Friend virus. Taken together, these data suggest that a Grb2/Gab2 complex is required for the ability of Sf-Stk to drive expansion of primary erythroblasts in response to Friend virus. Recent studies have identified a critical role for Grb2/Gab2 in the transformation of cells by BCR/Abl, suggesting that this might be a common downstream signaling pathway in the leukemic transformation of cells by tyrosine kinases. This work was performed by Hami Teal and Jie Xu.

[Home] [Macrophage Biology] [Cell Signaling and Transformation]