PileUp was always intended as a global alignment tool, which aligns sequences along their entire lengths. Because of this, PileUp is not ideally suited to finding the best local region of similarity (such as a shared motif) among all of the sequences. Global alignment algorithms tend to overlook areas of local similarity in favor of aligning the entirety of all sequences.
If you know where the areas of conserved motifs lie, use the SeqLab Editor to manually edit the resulting global alignment from PileUp, and then select subranges to realign portions of an existing alignment.
References
(1) Henikoff, S. and Henikoff, J. G. 1992. Amino acid substitution matrices from protein blocks. Proceedings of the National Academy of Sciences USA 89, 10915-10919.
(2) Schwartz, R.M. and Dayhoff, M.O. (1979). Matrices for Detecting Distant Relationships. In Atlas of Protein Sequences and Structure, (M.O. Dayhoff, ed.), 5, Suppl. 3 (pp; 353-358), National Biomedical Research Foundation, Washington, D.C., USA.