The C-terminal four residues of influenza A virus non-structural protein 1 (NS1) comprise the binding site for PDZ (the domain of PSD95, Dig and ZO-1) and is named PL domain (PDZ ligand domain). Previous study showed that NS1 proteins from different subtypes/strains of influenza A virus varied in their amino acids sequence of PL domain, which may affect the interaction between NS1 and cellular proteins and is closely associated with the pathogenicity of influenza A virus. To further explore the role of PL domain in the biological properties of NS1 protein, four wild type NS1-expressing plasmids harboring the NS1 encoding sequence from different influenza A virus subtypes (H1N1, H3N2, H5N1 and H9N2) were constructed firstly, based on pEGFP-c1 vector. The mutant NS1 (H3N2)-expressing plasmids were subsequently generated by either deletion or replacement of PL domain from other influenza A virus subtypes. Comparative analysis of the localization of these NS1 proteins in HeLa cells showed that wild type NS1 protein from H3N2 virus mainly localized in the nucleoli, whereas wild type NS1 proteins from H1N1, H5N1 and H9N2 viruses or mutant NS1 proteins from H3N2 virus mainly localized in the nuclei (but not nucleoli). In MDCK cells, none of the above NS1 proteins located in the nucleoli. The results indicated that PL domain can significantly influence the nuclear localizing pattern of NS1 protein in HeLa cells, which may be responsible for the variation of NS1 protein in its biological function; the distributed pattern of NS1 protein is closely associated with the origin of host cells.