Abstract:
Axonal branching mechanisms play a critical role in establishing neuronal circuitry that is the basis for a functional nervous system. A cyclic guanosine monophosphate (cGMP)-dependent signaling pathway mediates axon bifurcation in murine somatosensory neurons. Known parts of this cascade are the extracellular ligand C-type natriuretic peptide (CNP), which binds its transmembrane recep-tor guanylyl cyclase B (GC-B) thus activating the production of the second messenger cGMP, which in turn activates its effector, the cGMP-dependent kinase I (cGKI). Other components that are es-sential for the process of axon bifurcation remain elusive. This work aimed to uncover novel puta-tive cGKI substrates and GC-B interactors to gain better mechanistic understanding of the axon bifurcation process. The identification of such components could also provide a basis for new phar-macological approaches given the importance of cGMP-dependent signaling pathways in health and disease.
A mass-spectrometric (MS) phosphoproteome analysis was performed on the murine melanoma cell line B16F10 as well as on embryonic dorsal root ganglion (DRG) tissue after stimulation of cGKI activity with 8Br-cGMP and CNP, respectively. Both analyses lead to the discovery of many signifi-cantly regulated phosphorylation sites on putative direct or indirect cGKI substrates. Seven candi-dates were detected in both screening experiments: apoptotic chromatin condensation inducer in the nucleus (Acin1), glutamine-dependent carbamoyl-phosphate synthase-aspartate carbamoyltransfer-ase-dihydroorotase fusion protein (Cad), cytoplasmic linker protein-associating protein 1 (Clasp1), collapsin response mediator protein 1 (Crmp1), microtubule-associated protein 1B (Map1b), mam-malian enabled protein (Mena), and E3 ubiquitin-protein ligase NEDD4-like (Nedd4l).
The transcription of these cGKI substrate candidates was shown in both B16F10 cells and embryonic DRG tissue by RT-PCR. For Acin1, Clasp1, Crmp1, Map1b, and Mena, an expression analysis was performed to investigate the subcellular location in DRG cells as well as the tissue distribution pat-tern of these candidates. Acin1 showed only a minor presence outside of nuclei and could be found ubiquitously in the analyzed transversal embryo sections. The other candidates were present mostly in the cytosolic fraction, coinciding with cGKI; and to a smaller extent in the membrane fraction, coinciding with GC-B. Immunostaining of transversal embryo sections revealed the location of these four candidates in somata and axons of DRG neurons and some additional areas of the spinal cord, including the dorsal root entry zone (DREZ).
Aiming to verify the direct phosphorylation of Clasp1, Map1b, and Mena at the identified sites by cGKI, an analogue sensitive kinase (ASK) assay was performed on 3xFLAG-tagged constructs of the wildtype (WT) proteins as well as mutants with an alanine (A) substitution at the putatively modi-fied serine (S) residue. An ASK-variant of cGKIα could modify WT Mena and to a significantly less-er extent MenaS637A. Consequently, Mena is a direct phosphorylation substrate of cGKI and S637 is among the modified sites. While results for Map1b remained inconclusive, phosphorylation of Clasp1 by cGKIα-ASK could be shown as well, although no preferential modification of Clasp1S646 was apparent.
cGKI can interact via an N-terminal leucine zipper (LZ) domain with certain substrates that likewise contain such a domain. This can lead to a more stable substrate binding than other transient interac-tions with phosphorylation targets. Axon tracing of a mouse model expressing a mutant variant of cGKIα with an altered LZ domain affecting substrate binding but not kinase activity led to the dis-covery of an impairment in sensory axon bifurcation. In mouse embryos of this genotype, merely 23% of all investigated DRG axons entering the DREZ exhibited bifurcation. The remaining 77% of DRG axons instead completed a turn to grow along the rostro-caudal axis of the spinal cord. While a deletion of cGKI would lead to a near absolute absence of bifurcation, the substantial reduction of growth cone splits caused by disruption of the LZ domain indicates the integral nature of interac-tions via this domain.
Tissue of a GC-B reporter mouse line with an N-terminal hemagglutinin tag was used for pull-down assays to analyze GC-B interactors that co-immunoprecipitated with this enzyme. An MS analysis was performed on heart tissue of adult mice as well as DRG tissue of embryos, both leading to the discovery of putative GC-B interactors in the cytoskeletal compartment. While in heart tissue GC-B consistently interacted with a microtubule-associated protein, in DRG tissue significant interactions with four cytoskeleton-associated proteins as well as two tubulin monomer isoforms were detected. This suggests a function of GC-B as a possible anchoring point for cytoskeletal components and might be a starting point for uncovering a protein complex directing the cGMP-dependent signaling pathway in growth cones of somatosensory neurons to specific subcellular locations, thus mediating axon bifurcation.
Taken together, both phosphoproteomic screens for GKI substrates and interactome studies for GC-B have led to the identification of new components of the cGMP-dependent pathway, including several cytoskeleton-associated elements in DRG neurons. Some of the identified cGKI phosphoryla-tion sites have yet to be verified in future research, and their involvement in the process of axon bifurcation needs to be studied. While none of the candidate cGKI substrates investigated here con-tain an LZ domain, such substrates seem to play a critical role in vivo and might be the subject of future research efforts. Furthermore, they or other proteins might provide a scaffold to link GC-B and cGKI. Considering the known role of cGMP-dependent pathways in many (patho-)physiological contexts, the discovery of novel putative cGKI substrates is of more general interest in the field of cGMP research even beyond the process of somatosensory axon bifurcation.
cGMP