COMMENTARY COMMENTARY Context is everything for dependence receptors in colorectal cancer William M. Grady1 Clinical Research Division, Fred Hutchinson Cancer Research Center, Department of Medicine, University of Washington Medical School, Seattle, WA 98109

Cancer is a consequence of the accumulation phin family and is the receptor for of and epigenetic alterations in the NT-3 (Fig. 1). This fam- oncogenes and tumor-suppressor genes. Re- ily of receptors was identified following the cent advances in high-throughput sequencing discovery of the , NGF, methods and methylation array technology BDNF, NT-3, and NT-4/5, and consists has led to genome-wide analyses of gene of TrkA, TrkB, TrkC, and p75NTR (also alterations in a wide variety of cancers, in- known as NTRK1, NTRK2, NTRK3, and cluding colorectal cancer (1). These studies NGFR, respectively). These receptors are have revealed that there are hundreds to single-pass transmembrane tyrosine kinases thousands of DNA alterations in the average that were initially found to mediate neu- colorectal cancer genome, and that there are rotrophin-induced cell survival in the ner- roughly 80–100genesthatarecommonlyal- vous system. Although initially identified tered by nonsynonymous mutations and in the nervous system, the kinases are ubiq- Fig. 1. Schematic representation of the biological effects roughly 15 mutations in candidate “driver” uitously expressed and are involved in a va- induced when TrkC is bound and not bound by NT-3. In the genes per genome. (“Driver genes” are those riety of diseases, including cancer (3, 4). unbound state, TrkC can induce through - genes that induce the formation of cancer.) With regards to TrkC and colorectal can- mediated cleavage. The receptor has two cystein-rich fi domains (black ovals), a leucine-rich domain (rectan- These studies have also revealed consid- cer, one of the rst insights into the potential gle), two Ig like domains (half-circles), and cytoplasmic ty- erable heterogeneity between colorectal role of neurotrophin receptors in colorectal rosine kinase. cancers and that the majority of gene cancer came when Bardelli et al. found that mutations are likely passenger events, NTRK3 is a commonly mutated kinase in which are bystanders in cancer formation. colorectal cancer (5). Given the classic view when not bound with a ligand, dependence Thus, one of the major challenges facing of tyrosine kinases as being oncogenic, the receptors are biologically active in both the cancer biologists now is how to identify authors proposed that the NTRK3 mutations ligand-bound (on) and -unbound (off) state (11). Initially, this phenomenon was demon- the mutated genes that are functionally rel- (i.e., G608S, I695V, and L760I) were likely NTR evant in the tumorigenesis process. The tumor promoting. However, subsequently it strated for the p75 neurotrophin recep- traditional approach to deciphering which has been shown that the neurotrophin tor, but several investigators, including Pat- genes are functionally relevant assumes receptors, and TrkC in particular, can func- rick Mehlen, have provided evidence for mutations in genes will lead to constitutive tion as both tumor-suppressor genes and other receptors functioning as dependence receptors (e.g., DCC, RET, Patched, UNC5H, tumor-promoting or tumor-suppressing oncogenes (4, 6). This contradiction led – biological effects. Thus, mutations in genes Genevois et al. (2) to carry out a series of IR) (12 14). The concept of dependence receptors is not universally accepted because for the tyrosine kinases (e.g., EGFR, BRAF, careful studies to define the role of TrkC in it is plausible that the effects attributed to the and so forth) have typically been classified colorectal cancer. This group chose to study ligand-free receptor are actually a default pro- as oncogenic because they often induce TrkC in colorectal cancer because they had gramintrinsictothecell.Inthecaseofthe a constant state of inappropriate prolifera- earlier shown that TrkC can function as a neurotrophins, it is well established that these tion, whereas other genes, such as BRCA2, “dependence receptor” and because other ligands enhance cell survival in neurons, but are most often classified as tumor-suppres- dependence receptors, such as Deleted in it is less clear how the absence of ligand indu- sor genes because these mutations impair Colon Cancer (DCC)andUNC5H, can act ces cell death. Mehlen’s research group has DNA fidelity. as tumor-suppressor genes in colorectal can- produced a series of well-designed studies In this context of deciphering the func- cer (7, 8). It is the dependence-receptor as- that support the mechanism being based on tional consequences of altered genes in the pect of the studies by Genevois et al. that dependence-receptor function (8, 15, 16). colorectal cancer genome, in PNAS Gene- make their findings particularly interesting fi TrkC fi After providing support for the depen- vois et al. (2) have identi ed (also and of importance to eld of cancer biology. dence-receptor concept in the nervous called NTRK3) as a tumor-suppressor The concept of dependence receptors is a gene in colorectal cancer that is com- controversial one that was first introduced monly silenced by aberrant DNA methyl- by Rabizadeh and Bredesen to explain neu- Author contributions: W.M.G. wrote the paper. ation (“The dependence receptor TrkC is ron death induced by the absence of NGF The author declares no conflict of interest. a putative colon cancer tumor suppres- (9, 10). Unlike with classic receptor biology, See companion article on page 3017. sor”). TrkC is a member of the neurotro- in which a receptor is in the “off” position 1E-mail: [email protected].

www.pnas.org/cgi/doi/10.1073/pnas.1300758110 PNAS | February 19, 2013 | vol. 110 | no. 8 | 2697–2698 Downloaded by guest on September 25, 2021 system, for the last decade Mehlen’steamhas and have provided support for the depen- receptors can bind the unprocessed form been assessing the role of this class of recep- dence receptor concept in cancer biology. of neurotrophins, called proneurotrophins, tors in cancer. In studies from over a decade Their results also extend a context-depen- and can induce cell death (22). It is possible ago, they resolved the mystery behind the dence model of mutations in cancers by that some of the ligand-independent effects role of DCC in colorectal cancer. Although reinforcing the idea that tumor-suppres- of TrkC are actually secondary to p75NTR DCC was one of the first putative tumor- sor genes can be conditional tumor-sup- activity. In addition, the expression of TrkC suppressor genes found in colon cancer, pressor genes, the effects of which vary andNT-3inthenormalcolonandinco- mouse-model studies and other functional depending on not only the state of other lon adenomas remains to be defined. This studies failed to reveal tumor-suppressor ac- mutant genes in the cell, but also on the information would provide a richer un- tivity (17). Mehlen’s group revealed that the presence of ligands in the case of depen- derstanding of the consequences of TrkC reason for this discrepancy was because DCC dence receptors. deregulation in colorectal cancer formation is a dependence receptor and that it had Although the studies by Genevois et al. (2) and would address how NT-3 and TrkC in- opposing effects on intestinal epithelial cells, providestrongevidencethatTrkCisinacti- teract to create a selective advantage for depending on whether its ligand, netrin-1, vated in colon cancer and that it functions as TrkC inactivation in cancer cells. The stud- was present (18). Genevois et al. and other a tumor-suppressor gene, their findings also ies by Genevois et al. (2) have provided investigators have further demonstrated raise a number of questions. Given that TrkC compelling evidence to justify further study a role for other dependence receptors in is a member of a family of receptors, it would of the neurotrophins in colorectal cancer. the pathogenesis of colorectal cancer, includ- help to know the expression levels of the These additional studies will further define ing RET and UNC5H (6, 19, 20). other neurotrophins and their receptors in the potential for neurotrophins and their In PNAS, Genevois et al. (2) have now dem- colorectal cancer. The expression of p75NTR receptors to be the targets of anticancer onstrated that TrkC has tumor-suppressor and sortilin is particularly germane, as these therapies. activity in the colon and, perhaps more im- portantly, they have provided evidence that TrkC is a conditional tumor suppressor that 1 Leary RJ, et al. (2008) Integrated analysis of homozygous 12 Ellerby LM, et al. (1999) Kennedy’s disease: Caspase cleavage of mediates its effects through its dependence- deletions, focal amplifications, and sequence alterations in breast the is a crucial event in cytotoxicity. J Neurochem and colorectal cancers. Proc Natl Acad Sci USA 105(42): 72(1):185–195. – receptor activities (21). The authors found 16224 16229. 13 Bordeaux MC, et al. (2000) The RET proto-oncogene induces 2 that the majority of colorectal cancers silence Genevois A-L, et al. (2013) Dependence receptor TrkC is a putative apoptosis: A novel mechanism for Hirschsprung disease. EMBO J colon cancer tumor suppressor. Proc Natl Acad Sci USA NTRK3 19(15):4056–4063. through aberrant DNA methylation 110:3017–3022. 14 Thibert C, et al. (2003) Inhibition of neuroepithelial patched- and that reconstitution of TrkC in colorectal 3 Hondermarck H (2012) Neurotrophins and their receptors in breast induced apoptosis by sonic hedgehog. Science 301(5634):843–846. cancer. Cytokine Rev 23(6):357–365. cancer cell lines suppressed hallmark behav- 15 Mehlen P, et al. (1998) The DCC gene product induces apoptosis 4 Ivanov SV, Panaccione A, Brown B, et al. (2012) TrkC signaling is iors of cancer (2). Having established that activated in adenoid cystic carcinoma and requires NT-3 to stimulate by a mechanism requiring receptor proteolysis. Nature 395(6704): NTRK3 acts as a tumor-suppressor gene in invasive behavior. Oncogene, in press. 801–804. 5 Bardelli A, et al. (2003) Mutational analysis of the tyrosine kinome 16 Mehlen P, Bredesen DE (2004) The dependence receptor the colon, they then showed that the addi- in colorectal cancers. Science 300(5621):949. hypothesis. Apoptosis 9(1):37–49. tion of the ligand for TrkC, NT-3, rescues 6 Bouzas-Rodriguez J, et al. (2010) Neurotrophin-3 production 17 Arakawa H (2004) Netrin-1 and its receptors in tumorigenesis. the cells from the tumor-suppressor effects of promotes human neuroblastoma cell survival by inhibiting Nat Rev Cancer 4(12):978–987. – 18 TrkC. The authors also showed that a natu- TrkC-induced apoptosis. J Clin Invest 120(3):850 858. Mazelin L, et al. (2004) Netrin-1 controls colorectal tumorigenesis 7 Mehlen P, Llambi F (2005) Role of netrin-1 and netrin-1 by regulating apoptosis. Nature 431(7004):80–84. rally occurring NTRK3 , E543D, dependence receptors in colorectal cancers. Br J Cancer 93(1):1–6. 19 Luo Y, Tsuchiya KD, Il Park D, et al. (2012) RET is a potential 8 lacks the proapoptotic activity of TrkC. Tauszig-Delamasure S, et al. (2007) The TrkC receptor induces tumor suppressor gene in colorectal cancer. Oncogene, 10.1038/ apoptosis when the dependence receptor notion meets the onc.2012.225. These last sets of experiments demonstrated neurotrophin paradigm. Proc Natl Acad Sci USA 104(33): 20 Shin SK, et al. (2007) Epigenetic and genetic alterations in ’ 13361–13366. the conditional nature of TrkC stumor- Netrin-1 receptors UNC5C and DCC in human colon cancer. 9 Rabizadeh S, et al. (1993) Induction of apoptosis by the – suppressor activity and its role as a depen- low-affinity NGF receptor. Science 261(5119):345–348. Gastroenterology 133(6):1849 1857. 21 dence receptor in colorectal cancer. 10 Rabizadeh S, Bredesen DE (2003) Ten years on: Mediation of cell Mehlen P (2005) The dependence receptor notion: Another way fi fi death by the common neurotrophin receptor p75(NTR). Cytokine to see death. Cell Death Differ 12(8):1003. The signi cance of these ndings is two- 22 Growth Factor Rev 14(3-4):225–239. Yang L, et al. (2005) Targeted disruption of Smad4 in mouse fold. Genevois et al. (2) have found a unique 11 Mehlen P, Fearon ER (2004) Role of the dependence receptor DCC epidermis results in failure of hair follicle cycling and formation of skin tumor-suppressor gene for colorectal cancer in colorectal cancer pathogenesis. JClinOncol22(16):3420–3428. tumors. Cancer Res 65(19):8671–8678.

2698 | www.pnas.org/cgi/doi/10.1073/pnas.1300758110 Grady Downloaded by guest on September 25, 2021