ent of HT29 monolayers was cell wall -associated, likely proteinaceous and could be the same component causing the loss of HT29 bioreduction capacity. In contrast, the toxic agent of Ah was not necessarily associated with the bacterial cell, since bacteria-free culture filtrates also caused hemolysis and cytolysis. Furthermore, while Av-RAG-1 did not grow in presence or absence of HT29 it did produce an Ahlike, filterable, extracellular-cytotoxic activity when grown in LB broth. This activity was heat sensitive, whereas the Ab toxicant was not. Several studies have identified cytotoxic factors from various Ab strains. In particular, Ab produces outer membrane vesicles that contain a number of virulence-related factors. One of these factors, Ab outer membrane protein A, is suggested to play an important role in pathogenesis as evidenced by its pleiotropic effects. Purified AbOmpA was localized to the mitochondria of HEp-2 human laryngeal epithelial cells, and resulted in the release of pro-apoptotic molecules such as cytochrome c and apoptosis-inducing factor into the cytosol, caspase-3 activation, and DNA fragmentation. None of our strains demonstrated a difference in the presence of genes for ompA or another reported virulence factor gene, epsA. Our studies used the same Ab strain as in previous work of Choi and colleagues, and yet we did not observe Ab invasion of HT29 colonic epithelial cells, supporting earlier observations that other cell types were not as permissive for invasion as respiratory epithelial cells. Comparison of wild-type and AbOmpA-negative isogenic mutants of Ab by Choi et al demonstrated that AbOmpA was also involved in epithelial cell attachment, which we have observed during our studies. Since there are concerns about antibiotic resistance of some Acinetobacter strains, we also tested inhibitory effects of several antibiotics, and in particular, if the test strains carry known mutations in the constitutive genes gyrA and parC conferring fluoroquinolone resistance. The genes for gyrA and parC were sequenced, and then translated and aligned in silico. None of the tested strains had the serine to leucine transition implicated in the fluoroquinolone resistance phenotype. Functional susceptibility to fluoroquinolone was confirmed for all strains with an MIC assay. Furthermore, this assay demonstrated that the Ab strain exhibited the most resistance to the antibiotics in general, and the least effective antibiotics towards all strains were cefotaxime, ceftazidime, and trimethoprim. The MIC assay also revealed that none of the strains were resistant to the fluoroquinolone, ciprofloxacin. As a measure of immune response, we screened for bacterialinduced release of select cytokines known to be associated with our mammalian cell models. For HT29, all of PubMed ID:http://www.ncbi.nlm.nih.gov/pubmed/22189542 the Acinetobacter strains induced IL-8 levels that were consistent 
with those recorded for other non-invasive bacteria. This neutrophil chemoattractant has been used previously to differentiate between pathogens and generally harmless bacteria. The present study showed that Ab induced less IL-8 production in HT29 cells, compared to Aj, Ah, and RAG-1. Our results agree with a recent study by de Breij and colleagues who reported that human epithelial cells produced less IL-8 in response to A. 5(6)-Carboxy-X-rhodamine baumannii strains than to A. junii strains. They also suggested that A. baumannii appeared to be more virulent compared to the other strains, since infection caused by A. baumannii was a