Function of kinase-deficient Trk receptor isoforms. TrkB and TrkC encode a number of isoforms, including those that lack the catalytic tyrosine kinase domain. Little is known about the function of these kinase deficient isoforms in Trk signaling. In vitro studies, and our own in vivo studies, have shown that truncated Trk receptors can inhibit the function of kinase-active receptor isoforms in a dominant-negative manner or by ligand sequestration. The physiological relevance of this activity is, however, still unclear. The high degree of sequence conservation of the intracellular domains of truncated receptors suggests the potential for specific interactions with cytoplasmic proteins and signaling capabilities. Indeed, it has been reported recently that BDNF induces the production of calcium waves in astroglia through the truncated TrkB T1 receptor. However, the molecular mechanism(s) linking the TrkB T1 receptor to calcium mobilization and its physiological role is still unknown. Interestingly, TrkB T1 is 50% overexpressed in the brain of the trisomy 16 (Ts16) mouse model of Down syndrome and Ts16 hippocampal neurons die prematurely in culture. Neurodegeneration is commonly associated with Down syndrome in humans and TrkB T1 is also overexpressed in Alzheimers patients. To further investigate the role of TrkB T1 in neuronal survival, we generated a mouse lacking specifically the TrkBT1 kinase-deficient receptor isoform. This mutation caused no gross phenotype and could be used to correct the levels of TrkB T1 in Ts16 mice in vivo. Importantly, hippocampal neurons from TrkB T1 -/+; Ts16 mice escaped the premature cell death of Ts16 neurons in vitro (Dorsey et al. in preparation). This is a very exciting result because it contrasts with earlier hypotheses that neurodegeneration occurs due to insufficient supply of neurotrophic factors. Rather, our studies suggest that modulation of cell death and survival can occur at the level of the Trk receptor. We are now investigating the molecular mechanism underlying the detrimental effect of elevated TrkB T1 expression. Specifically, we are addressing both the effects of TrkBT1 on the activity of the full-length TrkB receptor and on the intracellular regulation of Ca++ levels. In this regard we have further investigated the role of truncated TrkB.T1 in vivo under physiological conditions. We found a gain-of function phenotype in TrkB.T1-/- mice both in an in vivo assay of gross neuromuscular function (e.g. increased forepaw grip strength) as well as in a direct assessment of nerve evoked muscle performance. The enhancement in performance could not be explained by alterations in muscle fiber plasticity or synaptic function. Instead, we found that one mechanism underlying this phenotype is increased muscle specific contractility and calcium signaling. These novel post-synaptic findings suggest that physiological TrkB.T1 acts to delimit muscle contractility with contributions from a pathway involving Ca2+ signaling. Truncated TrkC receptors, such as TrkC TK-, have never been implicated in intracellular signaling. To identify proteins that might bind the highly conserved intracellular domain of TrkC TK-, we conducted a yeast two-hybrid screen and identified an adaptor protein (GRASP/tamalin) that interacts specifically with TrkC TK- in a ligand dependent manner. Both tamalin and TrkC TK- are expressed in the brain with overlapping anatomical and subcellular distribution. We also found that NT-3 initiation of the TrkCTK-/Tamalin complex leads to activation of Rac1 GTPase through the ADP-ribosylation factor 6 (ARF6). NT-3 binding to TrkCTK-/Tamalin induces ARF6 translocation to the membrane, which in turn causes membrane ruffling and formation of cellular protrusions. Thus, we have shown that a truncated TrkC receptor lacking kinase activity can activate a specific intracellular signaling pathway that links NT-3 to key components of neuronal development and plasticity, such as regulation of the actin cytoskeleton and membrane trafficking. Moreover, we have established NT-3 as an unsuspected upstream activator of ARF6, a regulator of endosome membrane trafficking, endocytosis and actin remodeling at the cell surface; processes that are important for cell motility. (Esteban et al., submitted) To further investigate the function of tamalin in the nervous system, we have generated mice with a targeted deletion of tamalin. These mice are viable and are currently under investigation. Our studies showing that TrkC TK- can affect cell motility raise an important question concerning the role of Trk receptors in tumors. While expression of Trk receptors has been found in many tumor types (e.g. neuroblastoma, medulloblastoma, pancreatic carcinoma, melanoma, etc.), it has been difficult to explain how expression of one receptor in some tumors is associated with a favorable outcome (e.g. TrkC expression in medulloblastoma or neuroblastoma) while in other tumors it is linked to a negative prognosis (e.g. TrkC expression in pancreatic carcinoma). Virtually no attention has been paid to which specific Trk receptor isoforms (kinase active or deficient) are expressed in these tumors and whether they confer specific biological properties to neoplastic growth. Our findings raise the intriguing possibility that expression of truncated receptors alone or in association with the tyrosine kinase isoforms may account for some of these tumor growth characteristics. Thus, we plan to determine whether there is differential expression of Trk receptor isoforms among established tumor cell lines. We also want to determine whether modulating the expression of specific receptor isoforms can influence the proliferative and metastatic potential of tumor cells. Taken together, our findings suggest that truncated Trk receptors affect signaling involved in cell survival, vesicular transport and cell motility. These are all key cell biological processes that are altered in pathological conditions. Thus, we plan to continue our dissection of truncated Trk receptor activities in vitro and further extend our analysis in vivo by using the conditional mouse models we have generated
Showing the most recent 10 out of 14 publications
