Making use of a numerical self-consistent field method and polymer brush concepts, we model the solvated corona of neurofilaments (NF) composed of projection domains (unstructured tails) of constituent proteins. Projections are modeled with amino acid resolution. We focus on the importance of the two shortest ones (alpha-internexin and NF-L) in regulating the conformations of the two longer ones (NF-M and NF-H) in an isolated NF. We take the wild-type NF with no alpha-internexin as the reference, for which the phosphorylation-induced translocation of M- and H-tails has been examined previously. We demonstrate that a subbrush of L-tails creates an electrostatic potential profile with an approximately parabolic shape. An experimentally relevant (2:1) ratio of L- to alpha-projections reduces the charge density of the L subbrush and shifts the translocation transition of the H-tails to slightly higher degrees of phosphorylation. Replacing all L-tails by alpha-projections destroys the substructure of the NF corona and this alters the NF response to the phosphorylation of long tails.