ABSTRACT
<p><b>OBJECTIVE</b>To predict the exported proteins of the novel bacterium Phenylobacterium zucineum HLK1(T) using genome-wide computational identification by searching the export signals including N-terminal signal peptides and alpha-transmembrane helices.</p><p><b>METHODS</b>The computational identification of signal sequences was based on a consensus between multiple predictive tools, including SignalP V3.0, LipoP V1.0, Phobius and TMHMM 2.0. Type IV signal peptides and proteins exported via TAT machinery were searched manually based on the conservative motifs. All the predicted proteins were classified according to the Cluster of Orthologous Group (COG) standard.</p><p><b>RESULT</b>In the total 3861 proteins encoded by P. zucineum HLK1(T) 1 378 (35.7%) were predicted to be exported proteins, most of which (totally 735, 19.0% of the proteome and 53.3% of all the exported proteins) were uncleavable transmembrane helices. In addition, 499 type I signal peptides (12.9%, 36.2%), 101 lipoproteins (2.6%, 7.3%) were also identified. Four Type IV signal peptides and 12 TAT proteins were detected as well. According to the COG classification standard, most of these exported proteins were P proteins related to inorganic ion transport and metabolism and S proteins whose functions were unknown.</p><p><b>CONCLUSION</b>The genome of HLK1(T) coded various types of exported proteins which may play an important role in the interaction between P. zucineum and the host cell, and facilitate the strain to invade into the cell.</p>