DT is a standard nucleic acid base name for T in DNA, as of PDB format v3.x (if I remember it correctly). So the error message from HADDOCK means that it is strict (picky) with its input of PDB format. As a test, you may download the DNA structure (e.g., 355d) from the RCSB PDB, and submit it to HADDOCK to see what happens.
I've re-created a new version of the PDB filedna-new2.pdb (attached) for you to try: it has T instead of DT etc.
Best regards,
Xiang-Jun
As a side note, such HADDOCK-related questions are better directed to its creators/maintainers. They know exactly what PDB format they support, and may provide you some working examples.