Just a guess, but it would seem like when the design was originally created, these function properties were assigned to symbols through logic that was imported from Concept HDL, or another schematic capture program.
Then, instead of importing logic again by the same method (Concept HDL), you simply imported the logic thru a standard netlist file and it wrote over existing function properties in your SiP design database.
If the logic you are importing just needs to append to the existing the design, you could de-select the option: "Supercede all logical data" when importing logic. This will keep existing logic and properties intact. If your changes are more extensive and your design needs to synchronize with an existing schematic etc., then it may be safer to explore the source of the existing logic and re-import changes thru the original method.