Análise da estrutura e da função da proteína morfogenética RodZ de Bacillus subtilis

Resumo: RodZ é um componente do sistema morfogenético das células bacterianas. É uma proteína transmembranar que localiza em bandas ao longo do eixo longitudinal da célula. Em Bacillus subtilis, RodZ consiste numa porção citoplasmática, RodZn, e em uma parte extra-citoplasmática, RodZc. RodZn contém um domínio em helixturn- helix (HTH), enquanto que RodZc pode ser dividido num domínio coiled-coil e num domínio terminal C, de função desconhecida. Um segmento transmembranar (TM) único separa RodZn de RodZc. A eliminação de rodZ causa alongamento do nucleóide e leva à produção de células polares nucleadas. Aqui, mostramos que RodZn é estruturado, estável e em hélice α. Descobrimos que as substituições Y32A e L33A na suposta hélice de reconhecimento (3) do motivo HTH, bem como as substituições Y49A e F53A, fora do motivo HTH (4), causam divisão assimétrica, mas apenas as últimas levam à deslocalização sub-celular de RodZ. Sugerimos que as hélices 3 e 4 são utilizadas para uma interacção proteína-proteína ou proteína- DNA essencial para divisão celular enquanto que 4 deve contactar um componente do citosqueleto, possivelmente MreB, uma vez que a correcta localização sub-celular de RodZ depende desta proteína. Em todos os mutantes as células polares são anucleadas, pelo que concluímos que o alongamento do nucleóide não é um prérequisito para divisão assimétrica. RodZc é largamente não estruturado mas com conteúdo de folha , sendo estabilizado pelo domínio coiled-coil. Mostramos uma relação homóloga entre RodZc e a bomba de transporte Na+/Ca2+ NCX1 e identificámos dois resíduos no domínio C, G265 e N275, essenciais para a manutenção da forma celular. Estes resíduos fazem parte de um motivo em gancho que pode actuar como um local de interacção com um ligando desconhecido. RodZn e RodZc são monoméricos em solução. Contudo, na membrana, RodZ interage consigo própria num sistema de dois híbridos (Split-Ubiquitin) em levedura, sugerindo que possa formar multímeros in vivo.-----------ABSTRACT: RodZ is a transmembrane component of the bacterial core morphogenic apparatus. RodZ localizes in bands long the longitudinal axis of the cell, and it is though to functionally link the cell wall to the actin cytoskeleton. In Bacillus subtilis, RodZ consists of a cytoplasmic moiety, RodZn, and an extracytoplasmic moiety, RodZc. RodZn contains a predicted helix-turn-helix domain, whereas RodZc is thought to contain a coiled-coil region and a terminal C domain of unknown function. A single transmembrane domain separates RodZn from RodZc. Deletion of rodZ causes elongation of the nucleoid and leads to the production of polar minicells containing DNA. Here, we have studied the structure and function of RodZn and RodZc. We show that RodZn is a stable, folded, -helical domain. We discovered that the Y32A and L33A substitutions within the presumptive recognition helix (3) of the HTH motif, as well as the Y49A and F53A substitutions outside of the HTH motif (in 4) cause asymmetric cell division. However, only the substitutions in 4 cause sub-celular delocalization of RodZ. We suggest that 3 and 4 are used for a protein-protein or protein-DNA interaction important for cell division, whereas 4 is likely to contact a cytoskeletal component, presumably MreB. The polar cells formed by all the mutants are anucleate. We conclude that nucleoid elongation is not a prerequisite for asymmetric division. RodZc appears to be a largely unstructured domain, with some -sheet content, and is stabilized by the coiled-coil region. We show a homology relationship between RodZc and the NCX1 Na+/Ca2+ transporter and we found two residues within the C domain, G265 and N275, that are important for cell shape determination. These residues are predicted to be essential determinants of a claw-like motif, which may act as a binding site for an unknown ligand. Both the isolated RodZn and RodZc proteins are monomeric in solution. However, because full-length RodZ interacts with itself in a split-ubiquitin yeast two-hybrid assay, we suggest that it may dimerize or form higher order multimers in vivo.

Molecular definition of inflammasomes

Inflammasomes are crucial actors of the innate immunity. They consist of cytoplasmic multiprotein complexes controlling the biological activities of the inflammatory cytokines IL-1ß and IL-18. Inflammasome assembly depends on protein domain interaction. In this chapter, we focus on the biochemical aspects characterizing the components of the inflammasomes and their assembly into a high moelcuar weight protelytic complex.

Functional analysis of the post-transcriptional regulator RsmA reveals a novel RNA-binding site.

The RsmA family of RNA-binding proteins are global post-transcriptional regulators that mediate extensive changes in gene expression in bacteria. They bind to, and affect the translation rate of target mRNAs, a function that is further modulated by one or more, small, untranslated competitive regulatory RNAs. To gain new insights into the nature of this protein/RNA interaction, we used X-ray crystallography to solve the structure of the Yersinia enterocolitica RsmA homologue. RsmA consists of a dimeric beta barrel from which two alpha helices are projected. From structure-based alignments of the RsmA protein family from diverse bacteria, we identified key amino acid residues likely to be involved in RNA-binding. Site-specific mutagenesis revealed that arginine at position 44, located at the N terminus of the alpha helix is essential for biological activity in vivo and RNA-binding in vitro. Mutation of this site affects swarming motility, exoenzyme and secondary metabolite production in the human pathogen Pseudomonas aeruginosa, carbon metabolism in Escherichia coli, and hydrogen cyanide production in the plant beneficial strain Pseudomonas fluorescens CHA0. R44A mutants are also unable to interact with the small untranslated RNA, RsmZ. Thus, although possessing a motif similar to the KH domain of some eukaryotic RNA-binding proteins, RsmA differs substantially and incorporates a novel class of RNA-binding site.

Smad3 deficiency in mice protects against insulin resistance and obesity induced by a high-fat diet.

OBJECTIVE-Obesity and associated pathologies are major global health problems. Transforming growth factor-beta/Smad3 signaling has been implicated in various metabolic processes, including adipogenesis, insulin expression, and pancreatic beta-cell function. However, the systemic effects of Smad3 deficiency on adiposity and insulin resistance in vivo remain elusive. This study investigated the effects of Smad3 deficiency on whole-body glucose and lipid homeostasis and its contribution to the development of obesity and type 2 diabetes.RESEARCH DESIGN AND METHODS-We compared various metabolic profiles of Smad3-knockout and wild-type mice. We also determined the mechanism by which Smad3 deficiency affects the expression of genes involved in adipogenesis and metabolism. Mice were then challenged with a high-fat diet to study the impact of Smad3 deficiency on the development of obesity and insulin resistance.RESULTS-Smad3-knockout mice exhibited diminished adiposity with improved glucose tolerance and insulin sensitivity. Chromatin immunoprecipitation assay revealed that Smad3 deficiency increased CCAAT/enhancer-binding protein beta-C/EBP homologous protein 10 interaction and exerted a differential regulation on proliferator-activated receptor beta/delta and proliferator-activated receptor gamma expression in adipocytes. Focused gene expression profiling revealed an altered expression of genes involved in adipogenesis, lipid accumulation, and fatty acid beta-oxidation, indicative of altered adipose physiology. Despite reduced physical activity with no modification in food intake, these mutant mice were resistant to obesity and insulin resistance induced by a high-fat diet.CONCLUSIONS-Smad3 is a multifaceted regulator in adipose physiology and the pathogenesis of obesity and type 2 diabetes, suggesting that Smad3 may be a potential target for the treatment of obesity and its associated disorders.

Cooperative binding of estrogen receptor to imperfect estrogen-responsive DNA elements correlates with their synergistic hormone-dependent enhancer activity.

The Xenopus vitellogenin (vit) gene B1 estrogen-inducible enhancer is formed by two closely adjacent 13 bp imperfect palindromic estrogen-responsive elements (EREs), i.e. ERE-2 and ERE-1, having one and two base substitutions respectively, when compared to the perfect palindromic consensus ERE (GGTCANNNTGACC). Gene transfer experiments indicate that these degenerated elements, on their own, have a low or no regulatory capacity at all, but in vivo act together synergistically to confer high receptor- and hormone-dependent transcription activation to the heterologous HSV thymidine kinase promoter. Thus, the DNA region upstream of the vitB1 gene comprising these two imperfect EREs separated by 7 bp, was called the vitB1 estrogen-responsive unit (vitB1 ERU). Using in vitro protein-DNA interaction techniques, we demonstrate that estrogen receptor dimers bind cooperatively to the imperfect EREs of the vitB1 ERU. Binding of a first receptor dimer to the more conserved ERE-2 increases approximately 4- to 8-fold the binding affinity of the receptor to the adjacent less conserved ERE-1. Thus, we suggest that the observed synergistic estrogen-dependent transcription activation conferred by the pair of hormone-responsive DNA elements of the vit B1 ERU is the result of cooperative binding of two estrogen receptor dimers to these two adjacent imperfect EREs.

The N-terminal DNA-binding 'zinc finger' of the oestrogen and glucocorticoid receptors determines target gene specificity.

Steroid hormone receptors activate specific gene transcription by binding as hormone-receptor complexes to short DNA enhancer-like elements termed hormone response elements (HREs). We have shown previously that a highly conserved 66 amino acid region of the oestrogen (ER) and glucocorticoid (GR) receptors, which corresponds to part of the receptor DNA binding domain (region C) is responsible for determining the specificity of target gene activation. This region contains two sub-regions (CI and CII) analogous to the 'zinc-fingers' of the transcription factor TFIIIA. We show here that CI and CII appear to be separate domains both involved in DNA binding. Furthermore, using chimaeric ERs in which either the first (N-terminal) (CI) or second (CII) 'zinc finger' region has been exchanged with that of the GR, indicates that it is the first 'zinc finger' which largely determines target gene specificity. We suggest that receptor recognition of the HRE is analogous to that of the helix-turn-helix DNA binding motif in that the receptor binds to DNA as a dimer with the first 'zinc finger' lying in the major groove recognizing one half of the palindromic HRE, and that protein-DNA interaction is stabilized through non-specific DNA binding and dimer interactions contributed by the second 'zinc finger'.

Carma1, a CARD-containing binding partner of Bcl10, induces Bcl10 phosphorylation and NF-kappaB activation.

Bcl10, a caspase recruitment domain (CARD)-containing protein identified from a breakpoint in mucosa-associated lymphoid tissue (MALT) B lymphomas, is essential for antigen-receptor-mediated nuclear factor kappaB (NF-kappaB) activation in lymphocytes. We have identified a novel CARD-containing protein and interaction partner of Bcl10, named Carma1. Carma1 is predominantly expressed in lymphocytes and represents a new member of the membrane-associated guanylate kinase family. Carma1 binds Bcl10 via its CARD motif and induces translocation of Bcl10 from the cytoplasm into perinuclear structures. Moreover, expression of Carma1 induces phosphorylation of Bcl10 and activation of the transcription factor NF-kappaB. We propose that Carma1 is a crucial component of a novel Bcl10-dependent signaling pathway in T-cells that leads to the activation of NF-kappaB.

Étude du mécanisme de protection des spermatozoïdes de mammifères par le lait

Le lait écrémé est utilisé depuis plus d’un demi-siècle comme diluant protecteur des spermatozoïdes de mammifères. Depuis quelques années, il existe une demande grandissante pour des diluants exempts de produits d’origine animale. Toutefois, le mécanisme par lequel le lait protège les spermatozoïdes n’est pas connu, ce qui rend difficile de lui trouver un substitut. Les protéines majeures du plasma séminal de taureau, les protéines « Binder of SPerm » (BSP), sont néfastes lors de la conservation de la semence. Les spermatozoïdes sont en contact avec une grande concentration de protéines BSP qui stimulent une extraction continuelle de cholestérol/phospholipides de leur membrane plasmique. Les lipoprotéines de faible densité (LDL) du jaune d’oeuf, un autre composé utilisé dans les diluants, empêcheraient les protéines BSP de se lier à la membrane des spermatozoïdes de taureaux et de stimuler un efflux des lipides membranaires, ce qui les protégerait durant la conservation. Notre hypothèse était que les protéines du lait protègent les spermatozoïdes durant la conservation en séquestrant les protéines BSP. Premièrement, nous avons démontré par filtration sur gel qu’il y a une interaction entre les protéines BSP bovines et les protéines du lait. Le lait écrémé a été fractionné en trois fractions : F1 (alpha-lactalbumine, bêta-lactoglobuline et caséine kappa), F2 (toutes les protéines du lait) et F3 (sels, sucres et petits peptides). Les protéines BSP1 et BSP5 ont une affinité plus grande pour F1 que BSP3, tandis que toutes les protéines BSP ont une affinité pour F2. Le titrage calorimétrique isotherme a permis de confirmer l’interaction entre les protéines BSP et les protéines du lait. L’association entre la protéine BSP1 bovine et les micelles de caséines est caractérisée par une constante d’affinité (Ka) de 3.5 × 10^5 M-1 et un paramètre stoichiométrique (n) de 4,5 BSP1 pour une caséine. L’association entre la protéine BSP1 bovine et l’alpha-lactalbumine (une protéine du sérum principale), est caractérisée par un Ka de 2.4 × 10^5 M-1 et une valeur “n” de 0,8. Ces résultats indiquent que le lait protège les spermatozoïdes bovins en séquestrant les protéines BSP grâce à une interaction protéine : protéine, tandis que le jaune d’oeuf les protège grâce à une interaction protéine : lipoprotéine. Deuxièmement, nous avons démontré par filtration sur gel que les protéines homologues aux BSP bovines retrouvées dans le plasma séminal de porc, d’étalon et de bélier ont une affinité avec les protéines du lait, ce qui suggère que le mécanisme de protection des spermatozoïdes par le lait pourrait être le même chez ces espèces. Troisièmement, nous avons caractérisé l’interaction entre BSP1 bovine et les LDL du jaune d’oeuf qui a un Ka de 3.4 ± 0.4 × 10^6 M-1 et une valeur de « n » de 104 BSP1 pour une particule de LDL, indiquant qu’il existe des différences entre le mécanisme de protection des spermatozoïdes par le lait et le jaune d’oeuf. Nous croyons que les résultats présentés dans cette thèse aideront à créer de nouveaux diluants ne contenant pas de produits d’origine animale afin de cryoconserver les spermatozoïdes des mammifères.

Caractérisation de l’interaction entre la protéine Lin28 et le précurseur du microARN let-7g

La régulation de l’expression des gènes est ce qui permet à nos cellules de s’adapter à leur environnement, de combattre les infections ou, plus généralement, de produire la quantité exacte de protéine nécessaire pour répondre à un besoin spécifique. Parmi les joueurs les plus importants dans cette régulation de l’expression des gènes on retrouve les microARN (miARN). Ces petits ARN de 22 nucléotides sont présents chez la majorité des espèces multicellulaires et sont responsables du contrôle direct de plus de 30% des gènes exprimant des protéines chez les vertébrés. La famille de miARN lethal-7 (let-7) est composée de miARN parmi les plus connus et ayant des fonctions cruciales pour la cellule. La régulation du niveau des miARN let-7 est essentielle au bon développement cellulaire. La biogenèse de ces miARN, du transcrit primaire jusqu’à leur forme mature, est régulée principalement par Lin28, une protéine pluripotente très conservée. Cette protéine est composée d’un domaine cold shock (CSD) et de deux domaines de liaison au zinc. C’est grâce à ces domaines de liaison à l’ARN que Lin28 peut lier et inhiber la maturation des miARN let-7. L’objectif de cette thèse est de caractériser l’interaction entre Lin28 et le microARN précurseur let-7g afin de mieux comprendre le rôle de cette protéine dans l’inhibition de la biogenèse du miARN. À l’aide de techniques biochimiques et biophysiques, nous avons d’abord défini les principaux déterminants de l’interaction entre Lin28 et la boucle terminale du miARN précurseur let-7g (TL-let-7g). Nous avons conclu que le domaine C-terminal de Lin28, composé d’un motif riche en lysines et arginines ainsi que de deux motifs de liaison au zinc, permet à la protéine de lier spécifiquement et avec haute affinité un renflement riche en guanine conservé chez les précurseurs de la famille let-7. Aussi, parce que la séquence et la spécificité de liaison à l’ARN de ce domaine C-terminal sont semblables à celles de la protéine NCp7 du VIH, nous avons défini ce dernier comme le domaine NCp7-like de Lin28. Par la suite, nous avons caractérisé la multimérisation de trois protéines Lin28 sur la boucle terminale de pre-let-7g. Ceci a permis de réconcilier d’apparentes contradictions retrouvées dans la littérature actuelle concernant les sites de liaison de Lin28 lors de sa liaison aux miARN précurseurs. Nous avons identifié trois sites de liaison à haute affinité sur TL-let-7g qui sont liés dans un ordre précis par trois protéines Lin28. Lors de la formation du complexe multimérique, le CSD permet une déstabilisation de l’ARN, ce qui rend accessible plusieurs sites de liaison. Le domaine NCp7-like permet plutôt un assemblage ordonné de la protéine et facilite la liaison initiale de cette dernière. Ces nouveaux résultats rendent possible la mise au point d’un nouveau modèle de l’interaction entre Lin28 et le miARN précurseur let-7g. En conclusion, les études réalisées dans cette thèse apportent une meilleure compréhension des mécanismes moléculaires impliqués dans la régulation post-transcriptionnelle d’une importante famille de miARN et permettront de guider les futures études dans le domaine de recherche en pleine effervescence qu’est celui de la biogenèse des miARN.

Nop53p interacts with 5.8S rRNA co-transcriptionally, and regulates processing of pre-rRNA by the exosome

In eukaryotes, pre-rRNA processing depends on a large number of nonribosomal trans-acting factors that form intriguingly organized complexes. One of the early stages of pre-rRNA processing includes formation of the two intermediate complexes pre-40S and pre-60S, which then form the mature ribosome subunits. Each of these complexes contains specific pre-rRNAs, ribosomal proteins and processing factors. The yeast nucleolar protein Nop53p has previously been identified in the pre-60S complex and shown to affect pre-rRNA processing by directly binding to 5.8S rRNA, and to interact with Nop17p and Nip7p, which are also involved in this process. Here we show that Nop53p binds 5.8S rRNA co-transcriptionally through its N-terminal region, and that this protein portion can also partially complement growth of the conditional mutant strain Delta nop53/GAL:NOP53. Nop53p interacts with Rrp6p and activates the exosome in vitro. These results indicate that Nop53p may recruit the exosome to 7S pre-rRNA for processing. Consistent with this observation and similar to the observed in exosome mutants, depletion of Nop53p leads to accumulation of polyadenylated pre-rRNAs.

Cys mutants in functional regions of Sticholysin I clarify the participation of these residues in pore formation

Experimental evidence shows that the mechanism of pore formation by actinoporins is a multistep process, involving binding of the water-soluble monomer to the membrane and subsequent oligomerization on the membrane surface, leading to the formation of a functional pore. However, as for other eukaryotic pore-forming toxins, the molecular details of the mechanism of membrane insertion and oligomerization are not clear. In order to obtain further insight with regard to the structure-function relationship in sticholysins, we designed and produced three cysteine mutants of recombinant sticholysin I (rStI) in relevant functional regions for membrane interaction: StI E2C and StI F15C (in the N-terminal region) and StI R52C (in the membrane binding site). The conformational characterization derived from fluorescence and CD spectroscopic studies of StI E2C, StI F15C and StI R52C suggests that replacement of these residues by Cys in rStI did not noticeably change the conformation of the protein. The substitution by Cys of Arg(52) in the phosphocholine-binding site, provoked noticeable changes in rStI permeabilizing activity; however, the substitutions in the N-terminal region (Glu(2), Phe(15)) did not modify the toxin`s permeabilizing ability. The presence of a dimerized population stabilized by a disulfide bond in the StI E2C mutant showed higher pore-forming activity than when the protein is in the monomeric state, suggesting that sticholysins pre-ensembled at the N-terminal region could facilitate pore formation. (C) 2011 Elsevier Ltd. All rights reserved.

Diphtheria toxoid conformation in the context of its nanoencapsulation within liposomal particles sandwiched by chitosan

Chitosan (alpha alpha-(1-4)-amino-2-deoxy-beta beta-D-glucan) is a deacetylated form of chitin, a polysaccharide from crustacean shells. Its unique characteristics, such as positive charge, biodegradability, biocompatibility, nontoxicity, and rigid structure, make this macromolecule ideal for an oral vaccine delivery system. We prepared reverse-phase evaporation vesicles (REVs) sandwiched by chitosan (Chi) and polyvinylic alcohol (PVA). However, in this method, there are still some problems to be circumvented related to protein stabilization. During the inverted micelle phase of protein nanoencapsulation, hydrophobic interfaces are expanded, leading to interfacial adsorption, followed by protein unfolding and aggregation. Here, spectroscopic and immunological techniques were used to ascertain the effects of the Hoffmeister series ions on diphtheria toxoid (Dtxd) stability during the inverted micelle phase. A correlation was established between the salts used in aqueous solutions and the changes in Dtxd solubility and conformation. Dtxd alpha alpha-helical content was quite stable, which led us to conclude that encapsulation occurred without protein aggregation or without exposition of hydrophobic residues. Dtxd aggregation was 98% avoided by the kosmotropic, PO

Response surface models to predict broiler performance and applications for economic analysis

A study was carried out to elaborate response surface models using broiler performance data recovered from literature in order to predict performance and elaborate economic analyses. Nineteen studies published between 1995 and 2005 were retrieved using the systematic literature review method. Weight gain and feed conversion data were collected from eight studies that fulfilled the pre-established inclusion criteria, and a response surface model was adjusted using crude protein, environmental temperature, and age as independent variables. The models produced for weight gain (r² = 0.93) and feed conversion (r² = 0.85) were accurate, precise, and not biased. Protein levels, environmental temperature and age showed linear and quadratic effects on weight gain and feed conversion. There was no interaction between protein level and environmental temperature. Age and crude protein showed interaction for weight gain and feed conversion, whereas interaction between age and temperature was detected only for weight gain. It was possible to perform economic analyses to determine maximum profit as a function of the variables that were included in the model. It was concluded that the response surface models are effective to predict the performance of broiler chickens and allow the elaboration of economic analyses to optimize profit.

Biometria da carcaça e peso de vísceras e de órgãos internos de bovinos mestiços alimentados com diferentes volumosos e fontes protéicas

Quarenta bovinos machos, com peso médio de 371 kg, foram distribuídos em um delineamento em blocos casualizado com esquema fatorial 3x3, dos quais quatro foram abatidos no início do experimento como referência, para avaliar a biometria das carcaças e o peso de vísceras e órgãos internos quando alimentados com diferentes fontes protéicas (uréia, Amiferm e farelo de soja) e volumosos (cana-de-açúcar e a silagem de milho e pastagem de capim Brachiaria brizantha). Não houve interação significativa volumosos x fontes protéicas e não houve diferença significativa entre as fontes protéicas para os parâmetros estudados. Para os volumosos, os pesos (em porcentagem do peso do corpo vazio - % PCV) do conteúdo estomacal (CRROA) e do conteúdo gastrintestinal (CGI) foram significativamente maiores para os animais alimentados com cana-de-açúcar (7,52 e 8,59% PCV, respectivamente) e não diferiram entre os alimentados com silagem de milho e capim, com médias de 5,10% PCV, para CRROA, e 5,89% PCV para CGI. Os animais mantidos em pastagem não apresentaram compacidade das carcaças diferente dos animais alimentados com os outros volumosos, enquanto os alimentados com cana-de-açúcar mostraram ser menos compactos que os alimentados com silagem de milho. As fontes protéicas também influenciaram a compacidade, que foi menor nos animais que receberam Amiferm e iguais para os que receberam uréia e farelo de soja. Conclui-se que os animais alimentados com cana-de-açúcar apresentam maior enchimento dos compartimentos digestivos e carcaças menos compactas.

Crystallographic and spectroscopic characterization of a molecular hinge: Conformational changes in bothropstoxin I, a dimeric Lys49 phospholipase A2 homologue

Bothropstoxin I(BthTX-I) from the venom of Bothrops jararacussu is a myotoxic phospholipase A2 (PLA2) homologue which, although catalytically inactive due to an Asp49-->Lys substitution, disrupts the integrity of lipid membranes by a Ca2+-independent mechanism, the crystal structures of two dimeric farms of BthLTX-I which diffract X-rays eo resolutions of 3.1 and 2.1 Angstrom have been determined, the monomers in both structures are related by an almost perfect twofold axis of rotation and the dimer interfaces are defined by contacts between the N-terminal alpha-helical regions and the tips of the beta-wings of partner monomers. Significant differences in the relative orientation of the monomers in the two crystal forms results in open and closed dimer conformations, Spectroscopic Investigations of BthTX-I in solution have correlated these conformational differences with changes in the intrinsic fluorescence emission of the single tryptophan residues located at the dimer interface, the possible relevance of this structural transition in the Ca2+-independent membrane damaging activity is discussed. (C) 1998 Wiley-Liss, Inc.