Three peptides with pore-forming activity were isolated from the cytoplasmic granules of pathogenic Entamoeba histolytica by acidic extraction, gel filtration and reversed-phase high-performance liquid chromatography. Partial amino acid sequence analysis of the three active peptides revealed that the most abundant of them was amoebapore and the other two were isoforms thereof. Cloning and sequencing of genomic DNA resolved the amino acid sequence of the two newly recognized peptides. The three peptides designated amoebapores A, B and C were found to have the same molecular size but to differ markedly in their primary structure, although all six cysteine residues are conserved. Despite sequence divergence, structural implications predict for the three peptides a similar amphipathic alpha-helical conformation stabilized by disulphide bonds. All three isoforms exhibit pore-forming activity toward lipid vesicles, but they differ in their kinetics. They also are capable of perturbing the integrity of bacterial cytoplasmic membranes and thereby kill Gram-positive bacteria. The amoebapores represent a distinct family of membrane-active peptides that may function intracellularly as antimicrobial agents but may also confer cytolytic activity on the parasite.

To provide tools for functional molecular genetics of the protozoan parasite Entamoeba histolytica, we investigated the use of the prokaryotic neomycin phosphotransferase (NEO) gene as a selectable marker for the transfection of the parasite. An Escherichia coli-derived plasmid vector was constructed (pA5'A3'NEO) containing the NEO coding region flanked by untranslated 5' and 3' sequences of an Ent. histolytica actin gene. Preceding experiments had revealed that amoebae are highly sensitive to the neomycin analogue G418 and do not survive in the presence of as little as 2 jig/ml. Transfection of circular pA5'A3'NEO via electroporation resulted in Ent. histolytica trophozoites resistant to G418 up to 100 ,ug/ml. DNA and RNA analyses of resistant cells indicated that (i) the transfected DNA was not integrated into the amoeba genome but was segregated episomally, (ii) in the amoebae, the plasmid replicated autonomously, (iii) the copy number of the plasmid and the expression of NEOspecific RNA were proportional to the amount of G418 used for selection, and (iv) under continuous selection, the plasmid was propagated over an observation period of 6 months. Moreover, the plasmid could be recloned into E. coli and was found to be unrearranged. To investigate the use of pA5'A3'NEO to coexpress other genes in Ent. histolytica, a second marker, the prokaryotic chloramphenicol acetyltransferase (CAT) gene under control of an Ent. histolytica lectin gene promoter was introduced into the plasmid. Transfection of the amoebae with this construct also conferred G418 resistance and, in addition, allowed continuous expression of CAT activity in quantities corresponding to the amount of G418 used for selection. When selection was discontinued, transfected plasmids were lost as indicated by an exponential decline of CAT activity in trophozoite extracts.

To study the role of cysteine proteinases in the pathogenicity of Entamoeba histolytica, we have attempted to overexpress the three main cysteine proteinases (EhCP1, EhCP2, EhCP5) of this parasite in trophozoites of E. histolytica as well as in non‐pathogenic Entamoeba dispar by episomal transfection. Although each of the corresponding coding sequences were cloned in identical expression plasmids, we were unable to overexpress EhCP1 and EhCP5, respectively, but could substantially induce expression of EhCP2 in both amoeba species by sevenfold, leading to a threefold increase in total cysteine proteinase activity. Overexpression of EhCP2 did not influence expression of other cysteine proteinases and could be attributed to an increase of a single 35 kDa activity band in substrate gel electrophoresis. In contrast to previous findings, which indicated that amoeba cysteine proteinases are involved in erythrophagocytosis and liver abscess formation, cells overexpressing EhCP2 showed no difference in erythrophagocytosis or liver abscess formation compared with respective controls. However, overexpression of EhCP2 in both amoeba species resulted in a marked increase of in vitro monolayer destruction.

A pore‐forming peptide is implicated in the potent cytolytic activity of pathogenic Entamoeba histolytica. Using NH2‐terminal sequence information of this peptide, the corresponding cDNA was isolated. The cDNA‐deduced amino acid sequence revealed a putative signal peptide and a mature peptide of 77 amino acids including six cysteine residues. Computer‐aided secondary structure analysis predicted that the peptide would be composed of four adjacent alpha‐helices, and CD spectroscopy indicated an all alpha‐helical conformation. The tertiary structure appears to be stabilized by three disulfide bonds; the pore‐forming activity was not sensitive to heat but was lost in the presence of reducing agents. Sequence homology was found to the saposins and to surfactant‐associated protein B, both mammalian polypeptides of similar size and secondary structure but of non‐lytic function. In particular, the six cysteine residues were found to be conserved, suggesting a common motif for stabilizing a favourable tertiary structure. Compared with previously characterized toxic peptides also containing three disulfide bonds, the amoeba peptide may represent a distinct class of biologically active peptides.

Hybrid plasmids were constructed and used for successful transfection and transient expression of the chloramphenicol acetyltransferase (CAT) gene in the protozoan parasite Entwoeba histolytica. Transfection was performed by electroporation of the amebae in a potassium phosphate-based buffer under conditions of 3000 V/cm and 25 OF, resulting in a time constant of 0.4 ms. Expression of CAT activity was achieved with constructs in which the CAT coding region was flanked by untranslated upstream and downstream sequences of E. histolytca genes. Highest activity was detected after culturing transfected cells for 48 hr. Activity was found to be proportional to the amount of DNA transfected.