COMMENTARY
COMMENTARY
Complexity of metastasis-associated SDF-1 ligand signaling in breast cancer stem cells Nicholas A. Grahama,b and Thomas G. Graebera,b,c,d,e,1 a Department of Molecular and Medical Pharmacology, bCrump Institute for Molecular Imaging, cUniversity of California at Los Angeles Metabolomics and Proteomics Center, d Jonsson Comprehensive Cancer Center, and eCalifornia NanoSystems Institute, University of California, Los Angeles, CA 90095
Tumor metastasis is the main cause of cancer’s lethality. Much research has been dedicated to understand the signaling networks underlying tumor metastasis, with the ultimate goal of identifying signaling nodes or pathways whose inhibition will prevent metastasis. Although previous studies have identified the interaction of the chemokine ligand stromal cell-derived factor-1 (SDF-1) and the G protein-coupled receptor chemokine (C-X-C motif) receptor 4 (CXCR4) as essential for breast cancer metastasis (1), the signaling pathways downstream of SDF-1/CXCR4 have not been well established. In PNAS, Yi et al. use quantitative, liquid chromatography tandem mass spectrometry (LC-MS/MS) to greatly expand our knowledge of the protein phosphorylation networks downstream of SDF-1/CXCR4 in breast cancer stem cells (2). Yi et al. (2) begin by enriching for a subpopulation of CXCR4-expressing cells with an elevated ability to initiate tumors in vivo, that is, breast cancer stem cells, which are required for tumor recurrence and metastasis (3). Following induction with the chemokine SDF-1, they measured protein phosphorylation by LC-MS/MS. To verify that changes in phosphorylation were caused by the specific interactions of SDF-1/CXCR4, the authors compared cells with and without transient knockdown of CXCR4. Among the proteins with significantly changing phosphorylation sites (13% of the total observed), over 20% were related to cell adhesion, migration, and cytoskeleton, highlighting and further confirming the role of SDF-1/CXCR4 signaling in biophysical processes related to metastasis. Therapeutic Targets
The significance of Yi et al.’s (2) work in breast cancer stem cells derives from the hope that identification of the signaling pathways downstream of SDF-1/CXCR4 will suggest pathways or nodes that can be therapeutically targeted. To date, the most successful class of druggable targets are kinases, www.pnas.org/cgi/doi/10.1073/pnas.1405991111
and Yi et al. identify 50 kinases with significantly changing phosphorylation sites, including 44 kinases not previously associated with SDF-1/CXCR4 (Fig. 1), many of which are involved in cell migration. The data of Yi et al. also greatly expand the known functions of SDF-1/CXCR4 signaling beyond cell migration to include the cell cycle (e.g., CDK1, CDK3, and CDK7), DNA damage repair (e.g., BRCA1, Abraxas), and intercellular communication (e.g., GJA1, SLC38A2). As such, this study adds to the growing use of phosphoproteomics to identify therapeutic signaling targets in specific cancer contexts (4). Integration of the Literature and Other Data Sources
One of the fundamental challenges of phospho-proteomics is that most identified phosphorylation sites lack a known functional role. In this study (2), Yi et al. identify “core phosphosites” downstream of SDF-1 that exhibit evolutionary conservation across eukaryotic species, thereby suggesting that these phosphosites are more likely to play important functional roles in vivo. Before direct functional validation, further insight can be gained by using computational and bioinformatic methods to place phosphorylation events within the context of known biological pathways (5–7). Here, the authors combine their phosphorylation data with established kinase/phosphatase substrate interactions and known SDF-1/CXCR4 signaling pathways to construct the candidate signaling pathway map for SDF-1. This approach allowed the authors to uncover a novel PKA-MAP2K2-ERK pathway downstream of SDF-1/CXCR4, which they subsequently validated using a pharmacological inhibitor of PKA. Because of the complex nature of signaling networks, the field will need to continue developing methods that generate biological insight from quantitative phosphoproteomic datasets.
Fig. 1. Expansion of the SDF-1/CXCR4 migratory signaling network in breast cancer stem cells. Using a model of human breast cancer, Yi el al. (2) isolated cells with tumor-initiating capacity (i.e., cancer stem cells). Following stimulation with SDF-1, the authors used quantitative (isotope reductive dimethylation) LC-MS to identify and quantify over 11,000 protein phosphorylation sites. This analysis revealed many previously unrecognized downstream phosphorylation events involving 60 cell mobilityrelated proteins, 50 kinases, and 8 negative-feedback regulators, of which 43, 44, and 6 proteins, respectively, had not been previously associated with SDF-1/CXCR4 signaling. To derive biological insight from this complex data, the authors used computational methods to integrate their phospho-proteomic data with known pathways from the literature, which allowed for identification of a novel PKA-MAPK2K-ERK pathway. Purple circles represent measured phosphorylation sites.
By analogy to the human genome project, early mass spectrometry-based phosphoproteomic studies (8, 9) and manually curated compilation databases (10, 11) have provided a basic blueprint of the protein sites modified by phosphorylation. Computationally derived kinase–substrate interactions and networks Author contributions: N.A.G. and T.G.G. wrote the paper. The authors declare no conflict of interest. See companion article on page E2182. 1
To whom correspondence should be addressed. E-mail: tgraeber@ mednet.ucla.edu.
PNAS | May 27, 2014 | vol. 111 | no. 21 | 7503–7504
(6) and tools for hypothesis generation based on proteomic data (12) have further expanded these resources. As with genome sequencing, technology advances are allowing phosphoproteomic studies to play more than a resource generating role. Newer mass spectrometers can achieve low attomolar sensitivity for peptides in complex mixtures (13), and these improvements in sensitivity and corresponding reductions in the required amount of sample input [10 mg of input protein used in this study (2)] have helped make studies on cellular subpopulations, such as the cancer stem cells studied here, more feasible. Phosphoproteomics and Iterative Experimentation
As with the $1,000 genome concept, additional technological advances and reductions in instrument costs will further bring phosphoproteomics to more laboratories and to everyday iterative experimentation (5, 14). The complexity of the signaling networks downstream of SDF-1/CXCR4 argues that inhibition of any single signaling pathway or node will be insufficient to prevent metastasis or to kill all cancer stem cells. To fully understand and appreciate the complexity of these signaling networks, phosphoproteomics needs to become a method for repeated, iterative assessment of model systems following genetic and pharmacological perturbations, rather than a method for simple lead generation followed by low-throughput validation (e.g., Western immunoblotting). Toward this goal, proteomics will require faster and simpler methods for sample preparation (15), as well as better and cheaper methods for absolute quantitation (16). This need is particularly true given that protein or peptide mass spectrometry can reveal the dynamics of multiple phosphorylation sites on a single protein with quantitative resolution that phospho-specific antibodies cannot match (17). Although there is a large difference in sensitivity between unbiased MS-based phosphoproteomics and antibody-based approaches, such as Western immunoblotting or fluorescent bead-based immunoassays, studies such as that of Yi et al. (2) demonstrate how much information can be missed if one relies on preexisting antibody resources. New technologies are needed and expected to help fill this gap. Feedback and Kinetics
It is a common occurrence that unbiased systems-wide signaling investigations point to positive- and negative-feedback mechanisms
7504 | www.pnas.org/cgi/doi/10.1073/pnas.1405991111
that shape and regulate network outputs, especially when hyperactivated oncogenic signaling is involved (5, 18–20). The findings by Yi et al. (2) highlight the importance of negative feedback in SDF-1 signaling by demonstrating increased phosphorylation of myosin phosphatase components PPP1Cα, PPP1Cβ, and MYPT1. In validation studies, the authors also demonstrate that negativefeedback regulates the phosphorylation status of MEK, ERK, and δ-catenin. Targeted inhibition of mutated, active kinases by clinically approved drugs, including imatinib (BCR-ABL) and vemurafenib (BRAF V600E), relieves the negative-feedback loops that attempt to counter the oncogenic signaling, which can counterintuitively lead to inhibitor-initiated pathway reactivation and, ultimately, resistance. Thus, kinetic phosphoproteomics studies will be critical to fully elucidate how feedback-loop dynamics contribute to the efficacy of any therapeutic approach (5, 9, 21). This need for kinetic studies is another motivating factor for improved sensitivity and smaller sample input requirements. Currently, once a panel of phosphorylation events of interest is defined, techniques such as targeted-proteomics (e.g., multiple reaction monitoring, selected ion monitoring) can be used to study hundreds of these events in a wider panel of samples (22). In summary, Yi et al. (2) have greatly expanded the repertoire of phosphorylation
ACKNOWLEDGMENTS. Research in T.G.G.’s laboratory is supported by the American Cancer Society (Grant RSG12-257-01-TBE), the Melanoma Research Alliance (Grant 20120279), and the National Institutes of Health National Cancer Institute (Grants P01 CA168585 and R21 CA169993). N.A.G. is supported by the University of California at Los Angeles Scholars in Oncologic Molecular Imaging Program, National Institutes of Health Grant R25T CA098010.
1 Müller A, et al. (2001) Involvement of chemokine receptors in breast cancer metastasis. Nature 410(6824):50–56. 2 Yi T, et al. (2014) Quantitative phosphoproteomic analysis reveals system-wide signaling pathways downstream of SDF-1/CXCR4 in breast cancer stem cells. Proc Natl Acad Sci USA 111:E2182–E2190. 3 Al-Hajj M, Wicha MS, Benito-Hernandez A, Morrison SJ, Clarke MF (2003) Prospective identification of tumorigenic breast cancer cells. Proc Natl Acad Sci USA 100(7):3983–3988. 4 Drake JM, et al. (2013) Metastatic castration-resistant prostate cancer reveals intrapatient similarity and interpatient heterogeneity of therapeutic kinase targets. Proc Natl Acad Sci USA 110(49): E4762–E4769. 5 Rubbi L, et al. (2011) Global phosphoproteomics reveals crosstalk between Bcr-Abl and negative feedback mechanisms controlling Src signaling. Sci Signal 4(166):ra18. 6 Linding R, et al. (2007) Systematic discovery of in vivo phosphorylation networks. Cell 129(7):1415–1426. 7 Matsuoka S, et al. (2007) ATM and ATR substrate analysis reveals extensive protein networks responsive to DNA damage. Science 316(5828):1160–1166. 8 Ballif BA, Villén J, Beausoleil SA, Schwartz D, Gygi SP (2004) Phosphoproteomic analysis of the developing mouse brain. Mol Cell Proteomics 3(11):1093–1101. 9 Olsen JV, et al. (2006) Global, in vivo, and site-specific phosphorylation dynamics in signaling networks. Cell 127(3):635–648. 10 Keshava Prasad TS, et al. (2009) Human Protein Reference Database—2009 update. Nucleic Acids Res 37(Database issue): D767–D772. 11 Hornbeck PV, et al. (2012) PhosphoSitePlus: A comprehensive resource for investigating the structure and function of experimentally determined post-translational modifications in man and mouse. Nucleic Acids Res 40(Database issue):D261–D270.
12 Naegle KM, et al. (2010) PTMScout, a Web resource for analysis of high throughput post-translational proteomics studies. Mol Cell Proteomics 9(11):2558–2570. 13 Gallien S, et al. (2012) Targeted proteomic quantification on quadrupole-orbitrap mass spectrometer. Mol Cell Proteomics 11(12): 1709–1723. 14 Francavilla C, et al. (2013) Functional proteomics defines the molecular switch underlying FGF receptor trafficking and cellular outputs. Mol Cell 51(6):707–722. 15 Kulak NA, Pichler G, Paron I, Nagaraj N, Mann M (2014) Minimal, encapsulated proteomic-sample processing applied to copy-number estimation in eukaryotic cells. Nat Methods 11(3):319–324. 16 Gerber SA, Rush J, Stemman O, Kirschner MW, Gygi SP (2003) Absolute quantification of proteins and phosphoproteins from cell lysates by tandem MS. Proc Natl Acad Sci USA 100(12):6940–6945. 17 Prabakaran S, et al. (2011) Comparative analysis of Erk phosphorylation suggests a mixed strategy for measuring phosphoform distributions. Mol Syst Biol 7:482. 18 Graham NA, et al. (2012) Glucose deprivation activates a metabolic and signaling amplification loop leading to cell death. Mol Syst Biol 8:589. 19 Asmussen J, et al. (2013) MEK-dependent negative feedback underlies BCR–ABL-mediated oncogene addiction. Cancer Discov 4(2):200–215. 20 Lito P, et al. (2012) Relief of profound feedback inhibition of mitogenic signaling by RAF inhibitors attenuates their activity in BRAFV600E melanomas. Cancer Cell 22(5):668–682. 21 Zhang Y, et al. (2005) Time-resolved mass spectrometry of tyrosine phosphorylation sites in the epidermal growth factor receptor signaling network reveals dynamic modules. Mol Cell Proteomics 4(9):1240–1250. 22 Stergachis AB, MacLean B, Lee K, Stamatoyannopoulos JA, MacCoss MJ (2011) Rapid empirical discovery of optimal peptides for targeted proteomics. Nat Methods 8(12):1041–1043.
sites known to be regulated by SDF-1/ CXCR4. This study will serve as the basis for many follow-up studies to define the functional contribution of SDF-1/CXCR4– regulated kinases and feedback loops in breast cancer metastasis. As this and other studies begin to define clinically relevant biomarkers in metastatic disease (4), it will be necessary to develop targeted proteomic assays capable of quantifying phosphorylation events in patient tissues. The signaling pathways that control cancer phenotypes including metastasis are more complex than has been previously appreciated, and technical advances in MS-based proteomics have allowed the field to begin to interrogate these dynamic kinase/phosphatase signaling networks in a quantitative and unbiased manner. To optimize therapeutic targeting and to best tap into the networks unveiled by phosphoproteomic approaches, we need to further build our analysis pipelines and iterative experimentation approaches to connect the dots between cascades of phosphorylation events and resulting disease phenotypes.
Graham and Graeber
COMMENTARY
Complexity of metastasis-associated SDF-1 ligand signaling in breast cancer stem cells Nicholas A. Grahama,b and Thomas G. Graebera,b,c,d,e,1 a Department of Molecular and Medical Pharmacology, bCrump Institute for Molecular Imaging, cUniversity of California at Los Angeles Metabolomics and Proteomics Center, d Jonsson Comprehensive Cancer Center, and eCalifornia NanoSystems Institute, University of California, Los Angeles, CA 90095
Tumor metastasis is the main cause of cancer’s lethality. Much research has been dedicated to understand the signaling networks underlying tumor metastasis, with the ultimate goal of identifying signaling nodes or pathways whose inhibition will prevent metastasis. Although previous studies have identified the interaction of the chemokine ligand stromal cell-derived factor-1 (SDF-1) and the G protein-coupled receptor chemokine (C-X-C motif) receptor 4 (CXCR4) as essential for breast cancer metastasis (1), the signaling pathways downstream of SDF-1/CXCR4 have not been well established. In PNAS, Yi et al. use quantitative, liquid chromatography tandem mass spectrometry (LC-MS/MS) to greatly expand our knowledge of the protein phosphorylation networks downstream of SDF-1/CXCR4 in breast cancer stem cells (2). Yi et al. (2) begin by enriching for a subpopulation of CXCR4-expressing cells with an elevated ability to initiate tumors in vivo, that is, breast cancer stem cells, which are required for tumor recurrence and metastasis (3). Following induction with the chemokine SDF-1, they measured protein phosphorylation by LC-MS/MS. To verify that changes in phosphorylation were caused by the specific interactions of SDF-1/CXCR4, the authors compared cells with and without transient knockdown of CXCR4. Among the proteins with significantly changing phosphorylation sites (13% of the total observed), over 20% were related to cell adhesion, migration, and cytoskeleton, highlighting and further confirming the role of SDF-1/CXCR4 signaling in biophysical processes related to metastasis. Therapeutic Targets
The significance of Yi et al.’s (2) work in breast cancer stem cells derives from the hope that identification of the signaling pathways downstream of SDF-1/CXCR4 will suggest pathways or nodes that can be therapeutically targeted. To date, the most successful class of druggable targets are kinases, www.pnas.org/cgi/doi/10.1073/pnas.1405991111
and Yi et al. identify 50 kinases with significantly changing phosphorylation sites, including 44 kinases not previously associated with SDF-1/CXCR4 (Fig. 1), many of which are involved in cell migration. The data of Yi et al. also greatly expand the known functions of SDF-1/CXCR4 signaling beyond cell migration to include the cell cycle (e.g., CDK1, CDK3, and CDK7), DNA damage repair (e.g., BRCA1, Abraxas), and intercellular communication (e.g., GJA1, SLC38A2). As such, this study adds to the growing use of phosphoproteomics to identify therapeutic signaling targets in specific cancer contexts (4). Integration of the Literature and Other Data Sources
One of the fundamental challenges of phospho-proteomics is that most identified phosphorylation sites lack a known functional role. In this study (2), Yi et al. identify “core phosphosites” downstream of SDF-1 that exhibit evolutionary conservation across eukaryotic species, thereby suggesting that these phosphosites are more likely to play important functional roles in vivo. Before direct functional validation, further insight can be gained by using computational and bioinformatic methods to place phosphorylation events within the context of known biological pathways (5–7). Here, the authors combine their phosphorylation data with established kinase/phosphatase substrate interactions and known SDF-1/CXCR4 signaling pathways to construct the candidate signaling pathway map for SDF-1. This approach allowed the authors to uncover a novel PKA-MAP2K2-ERK pathway downstream of SDF-1/CXCR4, which they subsequently validated using a pharmacological inhibitor of PKA. Because of the complex nature of signaling networks, the field will need to continue developing methods that generate biological insight from quantitative phosphoproteomic datasets.
Fig. 1. Expansion of the SDF-1/CXCR4 migratory signaling network in breast cancer stem cells. Using a model of human breast cancer, Yi el al. (2) isolated cells with tumor-initiating capacity (i.e., cancer stem cells). Following stimulation with SDF-1, the authors used quantitative (isotope reductive dimethylation) LC-MS to identify and quantify over 11,000 protein phosphorylation sites. This analysis revealed many previously unrecognized downstream phosphorylation events involving 60 cell mobilityrelated proteins, 50 kinases, and 8 negative-feedback regulators, of which 43, 44, and 6 proteins, respectively, had not been previously associated with SDF-1/CXCR4 signaling. To derive biological insight from this complex data, the authors used computational methods to integrate their phospho-proteomic data with known pathways from the literature, which allowed for identification of a novel PKA-MAPK2K-ERK pathway. Purple circles represent measured phosphorylation sites.
By analogy to the human genome project, early mass spectrometry-based phosphoproteomic studies (8, 9) and manually curated compilation databases (10, 11) have provided a basic blueprint of the protein sites modified by phosphorylation. Computationally derived kinase–substrate interactions and networks Author contributions: N.A.G. and T.G.G. wrote the paper. The authors declare no conflict of interest. See companion article on page E2182. 1
To whom correspondence should be addressed. E-mail: tgraeber@ mednet.ucla.edu.
PNAS | May 27, 2014 | vol. 111 | no. 21 | 7503–7504
(6) and tools for hypothesis generation based on proteomic data (12) have further expanded these resources. As with genome sequencing, technology advances are allowing phosphoproteomic studies to play more than a resource generating role. Newer mass spectrometers can achieve low attomolar sensitivity for peptides in complex mixtures (13), and these improvements in sensitivity and corresponding reductions in the required amount of sample input [10 mg of input protein used in this study (2)] have helped make studies on cellular subpopulations, such as the cancer stem cells studied here, more feasible. Phosphoproteomics and Iterative Experimentation
As with the $1,000 genome concept, additional technological advances and reductions in instrument costs will further bring phosphoproteomics to more laboratories and to everyday iterative experimentation (5, 14). The complexity of the signaling networks downstream of SDF-1/CXCR4 argues that inhibition of any single signaling pathway or node will be insufficient to prevent metastasis or to kill all cancer stem cells. To fully understand and appreciate the complexity of these signaling networks, phosphoproteomics needs to become a method for repeated, iterative assessment of model systems following genetic and pharmacological perturbations, rather than a method for simple lead generation followed by low-throughput validation (e.g., Western immunoblotting). Toward this goal, proteomics will require faster and simpler methods for sample preparation (15), as well as better and cheaper methods for absolute quantitation (16). This need is particularly true given that protein or peptide mass spectrometry can reveal the dynamics of multiple phosphorylation sites on a single protein with quantitative resolution that phospho-specific antibodies cannot match (17). Although there is a large difference in sensitivity between unbiased MS-based phosphoproteomics and antibody-based approaches, such as Western immunoblotting or fluorescent bead-based immunoassays, studies such as that of Yi et al. (2) demonstrate how much information can be missed if one relies on preexisting antibody resources. New technologies are needed and expected to help fill this gap. Feedback and Kinetics
It is a common occurrence that unbiased systems-wide signaling investigations point to positive- and negative-feedback mechanisms
7504 | www.pnas.org/cgi/doi/10.1073/pnas.1405991111
that shape and regulate network outputs, especially when hyperactivated oncogenic signaling is involved (5, 18–20). The findings by Yi et al. (2) highlight the importance of negative feedback in SDF-1 signaling by demonstrating increased phosphorylation of myosin phosphatase components PPP1Cα, PPP1Cβ, and MYPT1. In validation studies, the authors also demonstrate that negativefeedback regulates the phosphorylation status of MEK, ERK, and δ-catenin. Targeted inhibition of mutated, active kinases by clinically approved drugs, including imatinib (BCR-ABL) and vemurafenib (BRAF V600E), relieves the negative-feedback loops that attempt to counter the oncogenic signaling, which can counterintuitively lead to inhibitor-initiated pathway reactivation and, ultimately, resistance. Thus, kinetic phosphoproteomics studies will be critical to fully elucidate how feedback-loop dynamics contribute to the efficacy of any therapeutic approach (5, 9, 21). This need for kinetic studies is another motivating factor for improved sensitivity and smaller sample input requirements. Currently, once a panel of phosphorylation events of interest is defined, techniques such as targeted-proteomics (e.g., multiple reaction monitoring, selected ion monitoring) can be used to study hundreds of these events in a wider panel of samples (22). In summary, Yi et al. (2) have greatly expanded the repertoire of phosphorylation
ACKNOWLEDGMENTS. Research in T.G.G.’s laboratory is supported by the American Cancer Society (Grant RSG12-257-01-TBE), the Melanoma Research Alliance (Grant 20120279), and the National Institutes of Health National Cancer Institute (Grants P01 CA168585 and R21 CA169993). N.A.G. is supported by the University of California at Los Angeles Scholars in Oncologic Molecular Imaging Program, National Institutes of Health Grant R25T CA098010.
1 Müller A, et al. (2001) Involvement of chemokine receptors in breast cancer metastasis. Nature 410(6824):50–56. 2 Yi T, et al. (2014) Quantitative phosphoproteomic analysis reveals system-wide signaling pathways downstream of SDF-1/CXCR4 in breast cancer stem cells. Proc Natl Acad Sci USA 111:E2182–E2190. 3 Al-Hajj M, Wicha MS, Benito-Hernandez A, Morrison SJ, Clarke MF (2003) Prospective identification of tumorigenic breast cancer cells. Proc Natl Acad Sci USA 100(7):3983–3988. 4 Drake JM, et al. (2013) Metastatic castration-resistant prostate cancer reveals intrapatient similarity and interpatient heterogeneity of therapeutic kinase targets. Proc Natl Acad Sci USA 110(49): E4762–E4769. 5 Rubbi L, et al. (2011) Global phosphoproteomics reveals crosstalk between Bcr-Abl and negative feedback mechanisms controlling Src signaling. Sci Signal 4(166):ra18. 6 Linding R, et al. (2007) Systematic discovery of in vivo phosphorylation networks. Cell 129(7):1415–1426. 7 Matsuoka S, et al. (2007) ATM and ATR substrate analysis reveals extensive protein networks responsive to DNA damage. Science 316(5828):1160–1166. 8 Ballif BA, Villén J, Beausoleil SA, Schwartz D, Gygi SP (2004) Phosphoproteomic analysis of the developing mouse brain. Mol Cell Proteomics 3(11):1093–1101. 9 Olsen JV, et al. (2006) Global, in vivo, and site-specific phosphorylation dynamics in signaling networks. Cell 127(3):635–648. 10 Keshava Prasad TS, et al. (2009) Human Protein Reference Database—2009 update. Nucleic Acids Res 37(Database issue): D767–D772. 11 Hornbeck PV, et al. (2012) PhosphoSitePlus: A comprehensive resource for investigating the structure and function of experimentally determined post-translational modifications in man and mouse. Nucleic Acids Res 40(Database issue):D261–D270.
12 Naegle KM, et al. (2010) PTMScout, a Web resource for analysis of high throughput post-translational proteomics studies. Mol Cell Proteomics 9(11):2558–2570. 13 Gallien S, et al. (2012) Targeted proteomic quantification on quadrupole-orbitrap mass spectrometer. Mol Cell Proteomics 11(12): 1709–1723. 14 Francavilla C, et al. (2013) Functional proteomics defines the molecular switch underlying FGF receptor trafficking and cellular outputs. Mol Cell 51(6):707–722. 15 Kulak NA, Pichler G, Paron I, Nagaraj N, Mann M (2014) Minimal, encapsulated proteomic-sample processing applied to copy-number estimation in eukaryotic cells. Nat Methods 11(3):319–324. 16 Gerber SA, Rush J, Stemman O, Kirschner MW, Gygi SP (2003) Absolute quantification of proteins and phosphoproteins from cell lysates by tandem MS. Proc Natl Acad Sci USA 100(12):6940–6945. 17 Prabakaran S, et al. (2011) Comparative analysis of Erk phosphorylation suggests a mixed strategy for measuring phosphoform distributions. Mol Syst Biol 7:482. 18 Graham NA, et al. (2012) Glucose deprivation activates a metabolic and signaling amplification loop leading to cell death. Mol Syst Biol 8:589. 19 Asmussen J, et al. (2013) MEK-dependent negative feedback underlies BCR–ABL-mediated oncogene addiction. Cancer Discov 4(2):200–215. 20 Lito P, et al. (2012) Relief of profound feedback inhibition of mitogenic signaling by RAF inhibitors attenuates their activity in BRAFV600E melanomas. Cancer Cell 22(5):668–682. 21 Zhang Y, et al. (2005) Time-resolved mass spectrometry of tyrosine phosphorylation sites in the epidermal growth factor receptor signaling network reveals dynamic modules. Mol Cell Proteomics 4(9):1240–1250. 22 Stergachis AB, MacLean B, Lee K, Stamatoyannopoulos JA, MacCoss MJ (2011) Rapid empirical discovery of optimal peptides for targeted proteomics. Nat Methods 8(12):1041–1043.
sites known to be regulated by SDF-1/ CXCR4. This study will serve as the basis for many follow-up studies to define the functional contribution of SDF-1/CXCR4– regulated kinases and feedback loops in breast cancer metastasis. As this and other studies begin to define clinically relevant biomarkers in metastatic disease (4), it will be necessary to develop targeted proteomic assays capable of quantifying phosphorylation events in patient tissues. The signaling pathways that control cancer phenotypes including metastasis are more complex than has been previously appreciated, and technical advances in MS-based proteomics have allowed the field to begin to interrogate these dynamic kinase/phosphatase signaling networks in a quantitative and unbiased manner. To optimize therapeutic targeting and to best tap into the networks unveiled by phosphoproteomic approaches, we need to further build our analysis pipelines and iterative experimentation approaches to connect the dots between cascades of phosphorylation events and resulting disease phenotypes.
Graham and Graeber
