A PDSE does not have directory block structure like PDS. There really isn't a concept of defining more directory entries in a PDSE since they are automatically kept interspersed with the data component.
Also PDSEs work differently than PDSs and that they should not run out of space when properly allocated (with secondary extents and with enough space on the volume). This allows for dynamic extent allocation.
For these reasons the CA PDSMAN's ALTERDIR (for adding directory blocks) and CA PDSMAN's ADDSPACE (for adding extents of cylinders or tracks) only apply to a PDS datasets.