Amphotericin B kills unicellular leishmanias by forming aqueous pores permeable to small cations and anions.

The polyene antibiotic amphotericin B (AmB) is known to form two types of ionic channels across sterol-containing liposomes, depending on its concentration and time after mixing (Cohen, 1992). In the present study, it is shown that AmB only kills unicellular Leishmania promastigotes (LPs) when aqueous pores permeable to small cations and anions are formed. Changes of membrane potential across ergosterol-containing liposomes and LPs were followed by fluorescence changes of 3,3' dipropylthiadicarbocyanine (DiSC3(5)). In KCl-loaded liposomes suspended in an iso-osmotic sucrose solution, low AmB concentrations (</=0.1 microM) induced a polarization potential, indicating K+ leakage, but no movement of cations and anions was allowed until AmB concentrations greater than 0.1 microM were added. In agreement with these data, it was found that AmB altered the negative membrane potential held across LPs in a manner consistent with the differential cation/anion selectivity exhibited by the channels formed in liposomes. Thus, LPs suspended in an iso-osmotic sucrose solution did not exhibit any AmB-induced membrane depolarization effect brought about by efflux of anions until 0.1 microM or higher AmB concentrations were added. By contrast, LPs suspended in an iso-osmotic NaCl solution and exposed to 0.05 microM AmB exhibited a nearly total collapse of the negative membrane potential, indicating Na+ entry into the cells. The concentration dependence of the AmB-induced permeability to different salts was also measured across vesicles derived from the plasma membrane of leishmanias (LMVs), by using a rapid mixing technique. At concentrations above 0.1 microM, AmB induced the formation of aqueous pores across LMVs with a positive cooperativity, yielding Hill coefficients between 2 to 3. Measured anion selectivity across such aqueous pores followed the sequence: SCN > NO3 > Cl > I > Br > acetate (SO2-4 being impermeable). Cell killing by AmB was followed by fluorescence changes of the DNA-binding compound ethidium bromide (EB). At low concentrations (</=0.1 microM), AmB was found to be nonlethal against LPs but, above this concentration, leishmanias were rapidly killed. The rate and extent of such an effect were found to be dependent on the type of cation and anion present in the external aqueous solution. For both NH+4 and Na+ salts, the measured rank order of AmB cell killing followed the same sequence that was determined for AmB-induced salt permeation across LMVs. Further, replacement of either extracellular Na+ by choline or Cl- by SO2-4, or its partial substitution by sucrose, in iso-osmotic conditions, led to a complete inhibition of the killing effect exerted by otherwise lethal AmB concentrations. Finally, it was shown that tetraethylammonium (TEA+), an organic cation that is known to block AmB-induced salt permeation across LMVs was able to retard the time lag observed for EB incorporation across LPs, indicating that this parameter can be taken to represent the time taken for salt accumulation inside the parasites. The present results thus indicate clearly that low AmB concentrations (</=0.1 microM) were able to form across LPs, cation channels that collapsed the parasite membrane potential but are not lytic. At high concentrations (>/=0.1 microM), a salt influx via the aqueous pores formed by the antibiotic was followed by osmotic changes leading to cell lysis. This last stage is supported by electron microscopy observations of the changes of parasite morphology immediately upon addition of AmB, which indicated that the typical elongated promastigote cell forms became rounded and the flagella swells and round up. The present work is the first demonstration of the in vitro sensitivity of Leishmania promastigotes to osmotic lysis by AmB.