[throwawaymaths]

it is possible to have mixed chirality in syntheic proteins. (see michael weiss work investigating insulin receptor binding to insulin)

its basically impossible (but not totally impossible) for a living creature to be able to generate any protein with mixed chiralities.

this is because a ribosome with a chamber that can support both chiralities is likely to be less efficient at protein extension. but also you need so much more trna if you want to support arbitrary d-amino acids, etc.