How funny this is, the perimeter of a desmosome with this very precise and quite abrupt “beginning”. It seems that adjacent cells (with a concentric area of pretty well adhered plasmalemmae (two adjacent cells involved)) make the abrupt 45 degree bend just to (coincident with?)  the widened intercellular space at the core of the desmosome to its typical width of about 25nm (up from the space between two cells at its closest… more around 17nm).

What is most interesting is what signals the 45 degree spread (bend) on the plasmalemma just before the desmosomal disk, and why is it so electron lucent? or free or proteins?.  This is a “free” zone before the electron dense desmosomal complex proteins that make their junctional adherent  weld of about 150nm. the width of the annulus. The area of the annulus would vary depending upon where in the spot desmosome the section happened to fall, and a cut in the annulus completely would just look like a wider two-cell junction, showing no desmosome at all.  Whether the width of the annulus around the desmosome is consistent from one species, or even one cell type to another is likely not known.  I will google to see.