It is widely known that the inactivation of tumor suppressor genes is critical for the initiation and progression of cancer. Coordinate inactivation of tumor suppressors and activation of oncogenes leads to the dysregulation of growth and cell death that is the hallmark of malignancy.1 Many of the genes involved in oncogenesis disrupt phosphorylation events that are critical to signaling pathways. Various kinases have been shown to drive cancer cell growth, such as the epidermal growth factor receptor (EGFR) and the BCR-ABL oncoprotein.2 A number of phosphatases have been shown to be important tumor suppressors, being frequently inactivated in a variety of cancers. For example, inactivation of PTEN by mutation or deletion has been demonstrated to be common in a number of tumors such as glioblastoma, prostate cancer and breast cancer.3
Recent work has now shown that specific members of the protein tyrosine phosphatase receptor (PTPR) family are broadly inactivated in human cancers and function as tumor suppressor genes. The PTPRs are receptor-like membrane proteins with a single transmembrane domain, variable extracellular domains, and intracellular phosphatase domains.4 The catalytic activity of the phosphatase domains function with a high degree of specificity, often dictated by side chain interactions involving regions adjacent to the phosphatase active site.5
A number of studies have suggested that PTPRs are involved in oncogenesis. PTPRJ (DEP1) is a direct antagonist of several growth promoting receptor-tyrosine kinases such as the platelet derived growth factor β (PDGFRβ). Loss of hereozygosity (LOH) has been observed at this locus in colon cancer.6,7 Furthermore, several survey studies sequencing the cancer genome have pointed to inactivation of tyrosine phosphatases as common events in colon cancer.8,9 Included among the many candidate genes in these studies was a relatively low frequency of PTPRD mutation and a somewhat higher frequency of PTPRT mutation.
In a recent report, we used a multifaceted genomic approach to shown that PTPRD is a broadly targeted multisite tumor suppressor gene inactivated by a number of mechanisms.10 This analysis examined the PTPRD locus for copy number loss, epi-genetic silencing by hypermethylation, and mutation across a number of human malignancies. Our copy number analysis was intriguing. PTPRD is located on chromosome 9p23–24, which resides on the region of 9p that is commonly lost in many cancers, including glioblastoma (GBM), lung cancer, and head and neck cancer. Array comparative genomic hybridization (aCGH) analysis of GBM tumors revealed that the PTPRD locus encompasses a second minimal commonly deleted region next to CDKN2A and that 9p loss is likely driven by not only loss of p16/ARF, but also PTPRD. In many GBMs that did not harbor PTPRD copy number loss, epigenetic silencing of the gene occurred via promoter CpG hypermethylation. Furthermore, mutations were found in GBM as well as in lung cancers and head and neck cancers, which were determined to be deleterious. In total, aberrations at the PTPRD locus exists in greater than 50% of GBMs. Our results highlight the importance of using a multifaceted approach when examining tumor suppressor genes, looking at multiple mechanisms of gene inactivation.
Next, a functional analysis of PTPRD was undertaken. Depletion of PTPRD in immortalized human astrocytes resulted in faster growth of cells both in vitro and in vivo. PTPRD was found to possess strong growth suppressive properties. These growth suppressive properties were abrogated by every tumor-derived PTPRD mutation that we examined. Next, using a candidate approach, we identified the signal transducers and activators of transcription 3 (STAT3) as a substrate of PTPRD. PTPRD dephosphorylated STAT3 in cells, in vitro, and regulated STAT3 transcriptional activity.10 STAT3 is a well known oncogene and STAT3 overactivity has been documented a large number of human cancers.11,12 The ability to regulate STAT3 was abrogated by mutations in PTPRD. Our observations, therefore, suggest that the tumor suppressive function of PTPRD is linked to its ability to control STAT3 function.
PTPRT is another PTPR family member that is inactivated in cancers. PTPRT is the most frequently mutated PTPR in colon cancer (11%) and is also mutated in stomach and lung cancer.8 In contrast to PTPRD, PTPRT is primarily targeted by mutation and, to our knowledge, rarely by other means of inactivation. Interestingly, PTPRT has also been shown to dephosphorylate and regulate STAT3 function.13 Like PTPRD, PTPRT dephosphorylates STAT3 at tyrosine 705, a phosphorylation site that is essential for STAT3 transcriptional activity. The finding that two commonly inactivated phosphatases, PTPRD and PTPRT, both regulate STAT3 is particularly compelling and further strengthens the involvement of this oncogene in human cancers. The convergence of the mechanisms of action of these two tumor suppressors, as well as that of oncogenic kinases such as the Janus kinase 2 (JAK2), into the STAT3 pathway solidifies the importance of this signaling axis in oncogenesis.14
The discovery of PTPRD inactivation and its mechanism of action via STAT3 has important implications for therapy. EGFR can act through a number of signaling pathways, including activation of signaling cascades involving Ras/Raf/MAPK, PI-3K and STAT3. Although mutations in EGFR can predict response to therapy with tyrosine kinase inhibitors (as in lung cancer), not all molecular determinants of therapy have been elucidated. Since both EGFR and PTPRD pathways converge on STAT3, PTPRD inactivation may be a determinant of response to EGFR inhibition. Mechanistically, PTPRD inactivation and deregulation of STAT3 may lead to constitutive STAT3 signaling in the setting of wild-type EGFR (Fig. 1). As such, treatment of PTPRD-null cancers with EGFR inhibitors (alone or in combination with JAK inhibitors) may be effective in halting growth in these tumors.
Figure 1.
Model of STAT3 regulation by PTPRD and receptor kinases such as EGFR. Blue spheres labelled with P denote phosphorylation events. Blue and red circles denote the cytoplasmic and nuclear membranes respectively.
Future work will be needed to address several important issues. Is PTPRD and PTPRT inactivation mutually exclusive? If not, do they act in a synergistic manner? What are the other tumor suppressors and oncogenes that are altered in PTPRD-null cancers? Can combined EGFR/JAK inhibition be used to target cancers in which PTPRD is altered? The answers to these questions will provide a more complete understanding of PTPRD in cancer biology and therapy.
Abbreviations:
- PTPR
protein tyrosine phosphatase receptor
- STAT
signal transducers and activators of transcription
- CNA
copy number alteration
- GBM
glioblastoma multiforme
References
- 1.Stratton MR, et al. Nature 2009; 458:719–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Varmus H, et al. Cold Spring Harbor Symposia on Quantitative Biology 2005; 70:1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Li J, et al. Science 1997; 275:1943–7. [DOI] [PubMed] [Google Scholar]
- 4.Ostman A, et al. Nat Rev Cancer 2006; 6:307–20. [DOI] [PubMed] [Google Scholar]
- 5.Jia Z, et al. Science 1995; 268:1754–8. [DOI] [PubMed] [Google Scholar]
- 6.Ruivenkamp C, et al. Oncogene 2003; 22:3472–4. [DOI] [PubMed] [Google Scholar]
- 7.Kovalenko M, et al. J Biol Chem 2000; 275:16219–26. [DOI] [PubMed] [Google Scholar]
- 8.Wang Z, et al. Science 2004; 304:1164–6. [DOI] [PubMed] [Google Scholar]
- 9.Wood LD, et al. Science 2007; 318:1108–13. [DOI] [PubMed] [Google Scholar]
- 10.Veeriah S, et al. Proc Natl Acad Sci USA 2009; 106:9435–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Bromberg JF, et al. Cell 1999; 98:295–303. [DOI] [PubMed] [Google Scholar]
- 12.Yu H, et al. Nat Rev Cancer 2004; 4:97–105. [DOI] [PubMed] [Google Scholar]
- 13.Zhang X, et al. Proc Natl Acad Sci USA 2007; 104:4060–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Levine RL, et al. Cancer Cell 2005; 7:387–97. [DOI] [PubMed] [Google Scholar]