Chemosensory responses of Acanthamoeba castellanii: visual analysis of random movement and responses to chemical signals.

A visual assay slide chamber was used in conjunction with time-lapse videomicroscopy to analyze chemotactic behavior of axenically grown Acanthamoeba castellanii. Data were collected and analyzed as vector scatter diagrams and cell tracks. Amebas responded to a variety of bacterial products or potential bacterial products by moving actively toward the attractant. Responses to the chemotactic peptide formyl-methionyl-leucyl-phenylalanine (fMLP), lipopolysaccharide, and lipid A were statistically significant (P < or = 0.03), as was the response to fMLP benzylamide (P < or = 0.05). Significant responses to cyclic AMP, lipoteichoic acid, and N-acetyl glucosamine were also found. Chemotactic peptide antagonists, mannose, mannosylated bovine serum albumin, and N-acetyl muramic acid all yielded nonsignificant responses (P > 0.05). There was no single optimal concentration for response to any of the attractants tested, and amebas responded equally over the range of concentrations tested. Pretreatment of amebas with chemotactic peptides, bacterial products, and bacteria reduced the directional response to attractants. Amebas that had been grown in the presence of bacteria appeared more responsive to chemotactic peptides. Treatment of amebas with trypsin reduced the response of cells to chemotactic peptides, though sensitivity was restored within a couple of hours. This suggests the ameba membrane may have receptors, sensitive to these bacterial substances, which are different from the mannose receptors involved in binding bacteria to the membrane during phagocytosis. The rate of movement was relatively constant (ca. 0.40 microns/s), indicating that the locomotor response to these signals is a taxis, or possibly a klinokinesis, but not an orthokinesis. Studies of the population diffusion rate in the absence of signals indicate that the basic population motility follows the pattern of a Levy walk, rather than the more familiar Gaussian diffusion. This suggests that the usual mathematical models of ameboid dispersion may need to be modified.