By Nathan Eisenberg
Mapping is the phenomenon where the spatial arrangement or molecular distribution of neuronal networks reflects the functional interactions of said network. One type of mapping is that of topographic mapping, in which the spatial or otherwise systematic character of the neurons’ up- and down-stream targets is borne out in the physical proximities between the neurons. A canonical example of this is in the tonotopic mapping of the auditory cortex. This system is arranged so groups of neurons in charge of processing tones of similar frequencies are adjacent to each other. Another sort of mapping is more dependent on the molecular identity of the cells in question. This means that cells that exhibit similar concentrations of particular proteins, and so share certain functional characteristics, project to or receive projections from the same places. In the olfactory system, the array of receptor neurons in the nose, each with a specific molecular profile that determines which airborne chemicals it can detect best, are interspersed so as to account for an unpredictable distribution of odorants. They cannot be said to be topographically arranged, and for good reason. However, receptor neurons with similar molecular identities, i.e. similar sensory functions, project to the same downstream processing centers in the cortex, regardless of their position relative to other receptors.
Neuronal contact with muscles frequently also involves mapping organization, for more exquisite control over muscular expenditure. Spinal motor neurons are known to organize into segregated “pools”, arranged into regional “columns” that tend to preferentially project to different anatomical centers and are identifiable from the expression rates of various transcription factors. A recent paper by Elena Demireva and colleagues attempted to determine whether this precise spatial arrangement was crucial for overall map integrity or not. They focused on genetic manipulations of proteins within the catenin-cadherin system, specifically the b and g catenin isoforms and N-cadherin. Catenins are intracellular proteins that participate in the signaling pathway of the developmental regulator Wnt and anchor cadherins to structural actin fibers within the cell. Cadherins are cellular adhesion proteins the dot the cell surface and bind to each other to stick neighboring cells onto each other. It was shown, using fluorescent microscopy in which a target protein is marked with light-emitting molecules, that b and g catenins have overlapping expression in spinal motor neurons.
The researchers created mice mutants in which either b or g catenin, or both, are knocked out. Gross morphological traits were preserved in these mutants, with the number of neurons expressing a given molecular identity the same in all conditions. The b/g double mutants showed intermixing between the different columns, exposing an underlying breakdown of compartmental integrity. However, the b and g only mutants were similar to wild type controls, in that there was negligible intermixing of columns. This suggests a certain redundancy in the function the two isoforms, as the presence of just one or the other can recover spatial partitioning, but the lack of both abolishes such function. The researchers did find, however, that at the level of neural pools, all mutants exhibited some degree of intermixing, so clearly both isoforms are needed for fine-tuning.
The concentration of N-cadherin dotting the surface of the cells was markedly decreased in b/g mutants, with increases in intracellular concentration, suggesting a disruption of N-cadherin placement. This lead to the hypothesis that the catenins act through their modulation of N-cadherin functionality. In N-cad mutants, overall morphology in terms of cell number, identity, and columnar organization was the same as in the control, but at the level of pools, the expected high degree of intermixing was present.
Most interestingly, the functional map was still intact, in either case, despite such mixing. Injection of a tracer verified that neurons with similar identities still projected to similar axonal targets, despite being far from their natural sites and interspersed with qualitatively different neurons. All in all, the functional map remained, meaning that this particular map is more dependent upon cellular identity than it is on relative position. The question remains, however, why a complex signaling system such as the catenin-cadherin is employed, presumably with expenditure of much energy and information processing, to attain precise topographic mapping if the real map is determined by transcriptional identity. The researchers speculate that spatial clustering, while not the cause of the motor map, does facilitate its coherence, as the availability of gap junctions – direct connections and exchange of materials – between adjacent neurons increases their ability to coordinate. Additionally, there is evidence that incoming axons from sensory neurons project in a tiered fashion to spinal motor pools, suggesting a role of spatial organization in establishing sensorimotor connections.
Much work remains to be performed to elucidate how extensively anatomy is oriented map-wise to physiologically relevant organs elsewhere in the body, particularly in the realm of neurons, which fulfill a conspicuously representational role. Though it remains to be seen, research into the development and various intricacies of functional maps may shed light on philosophically intractable problems such as the nature of reference and representation. All that can be said for now is that, despite its almost ungraspable complexity, the organization of our bodies is anything but arbitrary.
For additional information, you can see a graphical summary of the research or read the full-length paper. There’s also a companion paper from the same lab that talks about how the identity map is maintained.