The infection process of intracellular opportunistic microsporidia involves the forcible eversion of a coiled hollow polar filament that pierces the host cell membrane, allowing the passage of infectious sporoplasm into the host cell cytoplasm. Although the exact mechanism of spore activation leading to polar filament discharge is unknown, we have shown that spore adherence to host cells, which is mediated by sulfated glycosaminoglycans, may play a vital role. When adherence is inhibited, host cell infection decreases, indicating a direct link between adherence and infection. The goal of this study was to evaluate the effects of exogenous divalent cations on microsporidia spore adherence and infection. Data generated using an in vitro spore adherence assay show that spore adherence is augmented by manganese (Mn2+) and magnesium (Mg2+), but not by calcium (Ca2+). However, each of the three divalent cations contributed to increased host cell infection when included in the assay. Finally, we show that Mn2+ and Mg2+ may activate a constituent on the microsporidia spore, not on the host cell, leading to higher infection efficiency. This report further supports recent evidence that spore adherence to the host cell surface is an important aspect of the microsporidial infection process.

Myelin basic protein (MBP) has fascinated researchers and clinicians alike due to its major structural role in myelin and the central nervous system, and its potent auto-immunogenic properties that cause demyelination in animal models. The charge variants of MBP have been of particular interest. The C1 component, the least modified and most cationic of the variants, is the most abundant form of MBP in healthy adult myelin. The C8 component, the most modified and the least cationic variant, has been found in higher proportions in myelin of MS patients and children. Here, an investigation of the structural differences between C1 and C8 components of MBP was conducted. The spectral and hydrodynamic properties of these variants were monitored via a number of biophysical/biochemical techniques. The effect of zinc (Zn2+) on the conformational behaviour of MBP was examined. Zn2+ is an abundant metal in the brain, and had been previously shown to induce hydrodynamic compaction in MBP. Both variants have a loose tertiary arrangement with subtle differences. This arrangement is deficient in secondary structure and undergoes non-cooperative temperature-induced melting. Zn2+ stabilizes a molten globular-like state with enhanced ANS fluorescence, and promotes oligomerization.

Cette thèse, menée en collaboration avec Michelin dans le cadre de la fabrication de matériaux composites, porte sur la stabilité colloïdale du latex de caoutchouc naturel (NR). Ce latex est une dispersion colloïdale polydisperse d'un polymère naturel dans un sérum aqueux. Les particules sont stabilisées par une couche complexe de phospholipides et de protéines. Cette dispersion peut être coagulée de façon irréversible par voie physique (sous cisaillement) ou par voie physico-chimique (ajout de cations divalents ou de particules hydrophobes). Dans ce travail, nous avons étudié la coagulation du latex par les cations divalents. Pour cela nous avons utilisé des outils physico-chimiques : diagrammes de stabilité en présence de divers cations, suivi cinétique de l'agrégation des particules et mesures rhéologiques. Ensuite, nous avons sondé la surface de particules de NR, par électrophorèse, en présence de tensioactifs et aussi selon la taille des particules. Pour expliquer la coagulation, nous suggérons que l'augmentation de la force ionique écrante les interactions répulsives et que les cations forment des ponts ioniques entre les particules. Ces ponts les maintiennent au contact et l'irréversibilité du phénomène est assurée par l'adhésion entre les chaines de polymères proches de la surface. L'adsorption des tensioactifs modifie la surface des particules et la coagulation du latex. Une légère différence de charge de surface a été observée entre les particules de NR selon leur taille. Enfin, un phénomène de coagulation similaire a été obtenu avec un latex synthétique, ce qui ouvre la voie à l'exploitation industrielle de ce phénomène de coagulation. / This work, in collaboration with Michelin for the fabrication of composite materials, deals with the colloidal stability of the latex of natural rubber (NR). The NR latex is a polydisperse colloidal dispersion of a bio-polymer in an aqueous serum. The particles are stabilized by a complex layer of phospholipids and proteins. This dispersion can be coagulated by a physical way (under shearing), or by a physical-chemical way (addition of divalent cations or hydrophobic particles). In this thesis, we studied the coagulation of the NR latex by divalent cations with physical-chemical tools (stability diagrams with different cations, aggregation kinetic of particles, rheological measuremments). Then, the surface of NR particles has been characterized by electrophoresis, firstly in the presence surfactants and secondly depending on the particles size. From the results of the first part we suggest that the coagulation with divalent cations is due to a screening effect because of the increase of ionic strength and that divalent cations can bridge the particles together. This allows keeping them in contact. Adhesion between polymer chains near the surface ensures the irrversible cohesion. In a second time, we saw that the adsorption of surfactants changes the particles surface and the coagulation of NR latex with cations. A small difference of surface charge is observed between the biggest and the smallest NR particles. Finally, a similar behavior has been obtained with a synthetic latex in presence of divalent cations. Our results could be used to develop an industrial process based on this coagulation phenomenon.

Dans ce travail de thèse, nous avons étudié les interactions entre un polyoside anionique (pectine) et des cations monovalent (Na+) et divalents (Ca2+, Zn2+, Ba2+, Mg2+) en régime dilué (c < c*) et concentré (c ≈ c*). Ainsi, une pectine faiblement méthylée (LMP) a été étudiée en comparaison avec l’acide polygalacturonique (PGA). L’affinité de ces polyosides pour fixer les ions calcium diminue quand la concentration en NaCl augmente. Elle est plus élevée dans le cas du Ca-polyGal en comparaison avec Ca-LMP ; ceci peut être expliqué par la faible rigidité des chaînes observée pour le polyGal. Les interactions entre les quatre cations divalents (Ca2+, Zn2+, Ba2+, Mg2+) et les deux biopolymères (PolyGal et LMP) en régime dilué ont été étudiées afin d’obtenir des informations sur la structure du réseau, le mode d’association et l’énergie d’association. Nous avons donc proposé un mécanisme d’association qui est composé de deux étapes : i) formation de monocomplexes et de réticulations ponctuelles et ii) formation de dimères. Le passage de l’étape (i) à l’étape (ii) est caractérisé par un ratio molaire critique (R*=[M2+]/[Gal]) qui dépend du nombre et de la stabilité des réticulations ponctuelles entre le polymère et le cation. Pour le Mg-polyGal, l’association est due à une condensation des ions magnésium autour des chaînes du polyGal. Les résultats de simulations ont montré que l’association de 4 cations Zn2+ avec 2 chaînes composées de 8 unités Gal est similaire à une structure de type « boîte à œuf ». Ce modèle n’est pas applicable à la structure obtenue par l’association des cations Ca2+ et Ba2+. Les associations entre les polyGals et les cations divalents à une concentration en polymère proche de la concentration de recouvrement (c*) permettent d’obtenir des gels uniquement pour les trois cations (Ca2+, Zn2+, Ba2+). Les propriétés viscoélastiques de ces gels ainsi que la cinétique de gélification ont été étudiés. Dans le cas des gels, la première étape du mécanisme d’association proposé (formation de monocomplexes et de réticulations ponctuelles) s’accompagne d’une augmentation de l’épaisseur du gel ; alors que la deuxième étape (formation de dimères) conduit à une densification du gel. Nous avons remarqué que le coefficient de diffusion du front de gel suit l’ordre suivant : Ba2+ > Ca2+ > Zn2+ > Mg2+ ; ceci peut être relié à l’affinité entre les molécules d’eau de la sphère de coordination et le cation. En effet, l’affinité du cation pour l’eau augmente selon l’ordre inverse : Ba2+ < Ca2+ < Zn2+ < Mg2+. Enfin, nous avons utilisé ces trois polyosides (PGA, LMP et ALMP - pectine faiblement méthylée amidée) en association à des ions calcium pour fabriquer des microparticules contenant la rutine afin de cibler sa libération au niveau intestinal. Nous avons ainsi relié la cinétique de libération de la rutine à la structure du réseau mis en place lors de l’étape de gélification. Les microparticules à base de pectine ALMP présentent une capacité à fixer l’eau et un taux de libération de la rutine plus élevés que les microparticules à base de LMP et PGA. Le gel Ca-ALMP est plus flexible et présente des modules viscoélastiques plus faibles que les gels Ca-PGA et Ca-LMP. Nous avons attribué ceci à la distribution aléatoire des groupements ester et/ou amide dans ALMP qui gênerait la formation des dimères : les liaisons hydrogènes entre les fonctions amines et les fonctions carboxylates seraient donc responsables de la flexibilité du réseau formé. / In this thesis, we studied the interactions between an anionic polysaccharide (pectin) and monovalent cation (Na+) and divalent cations (Ca2+, Zn2+, Ba2+, Mg2+) in dilute regime (c < c*) and concentrate regime (c ≈ c *). Thus, a low methoxy pectin (LMP) was studied in comparison with a polygalacturonic acid (PGA). The affinity to bind calcium ions for these polysaccharides decreases as the NaCl concentration increases. This binding affinity was higher for Ca-polyGal than for Ca-LMP due to the low rigidity of chains observed in the polyGal. The interactions between four divalent cations (Ca2+, Zn2+, Ba2+, Mg2+) and the two biopolymers (polyGal and LMP) in the dilute regime were studied in order to obtain information about the network structure, the mode of association and the binding energy. Therefore, we propose a mechanism of the binding which consists of two steps: i) formation of monocomplexations and point-like cross-links ii) formation of dimers. The threshold molar ratio (R* = [M2+]/[Gal]), between these two steps depends on the number and the stability of the point-like cross-links between polyGal chains and the cation. Mg2+ interacts so strongly with water that is remains weakly bound to polyGal (polycondensation) by sharing water molecules from its first coordination shell with the carboxylate groups of polyGal. Molecular dynamic simulations of galacturonate chains in explicit water showed that the « egg-box » model is more adapted for zinc cations than for calcium and barium. When the concentration of the polyGal is close to the overlap concentration (c*), the addition of divalent cations allows to obtain gels for only three cations (Ca2+, Zn2+, Ba2+). The viscoelastic properties of these gels and the gelation kinetics were studied. In the case of gel formation, the first step (formation of monocomplexations and point-like cross-links) is accompanied by an increase in the gel thickness; while the second step (formation of dimers) leads to a densification of the gel. We found that the diffusion coefficient of the gel front increased according the following order: Ba2+ > Ca2+ > Zn2+ > Mg2+; this may be related to the affinity between the water molecules from the coordination sphere and the cation. Indeed, the affinity of the cation for water molecules increases in the reverse order: Ba2+ < Ca2+ < Zn2+ < Mg2+. Finally, we have used the three polysaccharides (PGA, LMP and ALMP - amidated low methoxyl pectin) in association with calcium ions to produce microparticles containing rutin to target drug release in the intestine. We have linked the rutin release kinetics to the network structure established in the gelation step. ALMP microparticles had higher ability to uptake water and thus higher drug release rate than two others microparticles (Ca-LMP and Ca-PGA). The Ca-ALMP gel was more flexible and had the lower viscoelastic modulus than Ca-PGA and Ca-LMP gels. We attributed this to the random distribution of ester and/or amide groups in ALMP, which hinders the formation of dimers: the hydrogen bonds between the amine groups and carboxylate groups are responsible for the flexibility of the network formed.

L’étude de l’adsorption et de la mobilité des ions dans les argiles est importante pour mieux appréhender de nombreux processus géochimiques et environnementaux, de même que pour prédire le comportement des radionucléides dans les conditions du stockage géologique. A cause de leurs tailles très petites (< 2μm), il n’est pas toujours évident d’étudier les argiles à l’aide des méthodes et techniques expérimentales existantes. L’une des alternatives à ce problème consiste alors à utiliser la modélisation moléculaire pour les étudier. En plus de leurs tailles fines, les argiles présentent également des structures complexes, qui peuvent survenir en raison de la multiplicité de possibilités de distributions et d’arrangements des substitutions isomorphiques dans leurs couches. Il a été clairement démontré qu’il existe une corrélation entre la distribution des substitutions dans les couches des argiles et les propriétés de ces dernières. En revanche, ceci reste à démontrer en ce qui concerne l’arrangement de ces substitutions dans les couches de l’argile. Dans ce travail, la modélisation moléculaire est utilisée pour déterminer et comparer les propriétés d’hydratation, ainsi que la structure et la mobilité des ions (Li⁺, Na⁺, K⁺, Rb⁺, Cs⁺, Mg²⁺, Ca²⁺, Sr²⁺, Ba²⁺, Ni²⁺, UO₂²⁺) et de l’eau dans l’espace interfoliaire de trois modèles de montmorillonite, différents entre eux par l’arrangement des substitutions isomorphiques dans les couches de l’argile.L’adsorption et la diffusion des cations cités plus haut et de l’eau sont également étudiées à la surface de la montmorillonite et les résultats sont comparés à ceux obtenus dans l’espace interfoliaire à 298 K et à 363 K. Les résultats obtenus dans ce travail s’accordent bien avec les observations expérimentales, et font ressortir une corrélation plus ou moins importante entre le modèle d’argile utilisé et le type de propriété calculée. Cette corrélation dépend également de la nature du cation présent dans la structure de la montmorillonite et de sa teneur en eau. / The study of adsorption and ion mobility in clay minerals is important for a better understanding of many geochemical and environmental processes, as well as to predict the behavior of radionuclides in geological storage conditions. Because of their very small size (<2μm), it is not always easy to study clays by using the existing experimental methods and techniques. One alternative to this issue is to use computational molecular modeling to carry out clay studies. In addition to their tiny size, clays minerals also have complex structures, which can appear due to various possibilities in the distribution and arrangement of isomorphic substitutions in their layers. It has been clearly demonstrated that there is a strong correlation between the distribution of substitutions in the clay layers and their properties. However, this remains to be shown regarding the arrangement of the substitutions in the layers of the clay. In this work, computational molecular modeling techniques are used to determine and compare the hydration properties, as well as the structure and mobility of ions (Li⁺, Na⁺, K⁺, Rb⁺, Cs⁺, Mg²⁺, Ca²⁺, Sr²⁺, Ba²⁺, Ni²⁺,UO₂²⁺) and water in the interlayer space of the three models of montmorillonite, that differ from each other by the arrangement of isomorphic substitutions in the clay layers.The adsorption and diffusion of the previously listed cations and water are also studied on the surface of montmorillonite clay and the results are compared to those obtained in the interlayer space both at 298 K and at 363 K. The data generated in this work agree well with experimental observations, and show a more or less significant correlation between the clay model used and the type of property calculated.

Weltweit versuchen Wissenschaftler, künstliche Viren für den Gentransfer zu konstruieren, die nicht reproduktionsfähig sind. Diese sollen die Vorteile der natürlichen Viren besitzen (effizienter Transport von genetischem Material), jedoch keine Antigene auf ihrer Oberfläche tragen, die Immunreaktionen auslösen. <br><br> Ziel dieses Projektes ist es, einen künstlichen Viruspartikel herzustellen, dessen Basis eine Polyelektrolytenhohlkugel bildet, die mit einer Lipiddoppelschicht bedeckt ist. Um intakte Doppelschichten zu erzeugen, muss die Wechselwirkung zwischen Lipid und Polyelektrolyt (z.B. DNA) verstanden und optimiert werden. Dazu ist es notwendig, die strukturelle Grundlage der Interaktion aufzuklären. Positiv geladene Lipide gehen zwar starke Wechselwirkungen mit der negativ geladenen DNA ein, sie wirken jedoch toxisch auf biologische Zellen. In der vorliegenden Arbeit wurde daher die durch zweiwertige Kationen vermittelte Kopplung von genomischer oder Plasmid-DNA an zwitterionische oder negativ geladene Phospholipide an zwei Modellsystemen untersucht. <br><br> 1. Modellsystem: Lipidmonoschicht an der Wasser/Luft-Grenzfläche <br> Methoden:<br> Filmwaagentechnik in Kombination mit IR-Spektroskopie (IRRAS), Röntgenreflexion (XR), Röntgendiffraktion (GIXD), Brewsterwinkel-Mikroskopie (BAM), Röntgenfluoreszenz (XRF) und Oberflächenpotentialmessungen <br> Resultate:<br> A) Die Anwesenheit der zweiwertigen Kationen Ba2+, Mg2+, Ca2+ oder Mn2+ in der Subphase hat keinen nachweisbaren Einfluss auf die Struktur der zwitterionischen DMPE- (1,2-Dimyristoyl-phosphatidyl-ethanolamin) Monoschicht. <br> B) In der Subphase gelöste DNA adsorbiert nur in Gegenwart dieser Kationen an der DMPE-Monoschicht. <br> C) Sowohl die Adsorption genomischer Kalbsthymus-DNA als auch der Plasmid-DNA pGL3 bewirkt eine Reduktion des Neigungswinkels der Alkylketten, die auf einen veränderten Platzbedarf der Kopfgruppe zurückzuführen ist. Durch die Umorientierung der Kopfgruppe wird die elektrostatische Wechselwirkung zwischen den positiv geladenen Stickstoffatomen der Lipidkopfgruppen und den negativ geladenen DNA-Phosphaten erhöht.<br> D) Die adsorbierte DNA weist eine geordnete Struktur auf, wenn sie durch Barium-, Magnesium-, Calcium- oder Manganionen komplexiert ist. Der Abstand zwischen parallelen DNA-Strängen hängt dabei von der Größe der DNA-Fragmente sowie von der Art des Kations ab. Die größten Abstände ergeben sich mit Bariumionen, gefolgt von Magnesium- und Calciumionen. Die kleinsten DNA-Abstände werden durch Komplexierung mit Manganionen erhalten. Diese Ionenreihenfolge stellt sich sowohl für genomische DNA als auch für Plasmid-DNA ein. <br> E) Die DNA-Abstände werden durch die Kompression des Lipidfilms nicht beeinflusst. Zwischen der Lipidmonoschicht und der adsorbierten DNA besteht demnach nur eine schwache Wechselwirkung. Offensichtlich befindet sich die durch zweiwertige Kationen komplexierte DNA als weitgehend eigenständige Schicht unter dem Lipidfilm. <br> <br><br> 2. Modellsystem: Lipiddoppelschicht an der fest/flüssig-Grenzfläche<br> Methoden:<br> Neutronenreflexion (NR) und Quarzmikrowaage (QCM-D)<br> Resultate:<br> A) Das zwitterionische Phospholipid DMPC (1,2-Dimyristoyl-phosphatidylcholin) bildet keine Lipiddoppelschicht auf planaren Polyelektrolytmultischichten aus, deren letzte Lage das positiv geladene PAH (Polyallylamin) ist. <br> B) Hingegen bildet DMPC auf dem negativ geladenen PSS (Polystyrolsulfonat) eine Doppelschicht aus, die jedoch Defekte aufweist. <br> C) Eine Adsorption von genomischer Kalbsthymus-DNA auf dieser Lipidschicht findet nur in Gegenwart von Calciumionen statt. Andere zweiwertige Kationen wurden nicht untersucht.<br> D) Das negativ geladene Phospholipid DLPA (1,2-Dilauryl-phosphatidsäure) bildet auf dem positiv geladenen PAH eine Lipiddoppelschicht aus, die Defekte aufweist. <br> E) DNA adsorbiert ebenfalls erst in Anwesenheit von Calciumionen in der Lösung an die DLPA-Schicht.<br> F) Durch die Zugabe von EDTA (Ethylendiamintetraessigsäure) werden die Calciumionen dem DLPA/DNA-Komplex entzogen, wodurch dieser dissoziiert. Demnach ist die calciuminduzierte Bildung dieser Komplexe reversibel. / All over the world scientists are trying to engineer artificial viruses, which do not replicate, for gene delivery. These artificial viruses should have the advantages of natural viruses such as efficient transport of genetic material, but they should not carry antigens, which cause immune reactions, on their top portion.<br><br> The aim of this project is to develop an artificial virus particle that is based on a polyelectrolyte hollow capsule which is covered by a lipid bilayer. To create intact bilayers, it is crucial to understand and optimize the interaction between lipids and polyelectrolytes (e. g. DNA). Therefore the structural basis of that interaction must be elucidated. Positively charged lipids interact strongly with the negatively charged DNA but they cause toxic reactions in biological cells. Hence the present work used two model systems to study the coupling of genomic or plasmid DNA to zwitterionic or negatively charged phospholipids induced by divalent cations. <br><br> 1. Model system: Lipid monolayer at the air/water-interface <br> Methods: <br> Langmuir filmbalance in combination with IR-spectroscopy (IRRAS), X-ray reflectometry (XR), X-ray diffraction (GIXD), Brewster angle microscopy (BAM), X-ray fluorescence (XRF), and surface potential measurements<br> Results:<br> A) The presence of the divalent cations Ba2+, Mg2+, Ca2+ or Mn2+ in the subphase has no traceable influence on the structure of a zwitterionic DMPE (1,2-dimyristoyl-phosphatidyl-ethanolamine) monolayer.<br> B) DNA which is dissolved in the subphase adsorbs to the DMPE-monolayer only if divalent cations are present.<br> C) The adsorption of genomic calf thymus DNA as well as of the plasmid DNA pGL3 causes a reduction of the tilt angle of the lipid alkyl chains. The tilt reduction can be ascribed to a change in the space required by the lipid head group. This change in head group orientation increases the electrostatic interaction between the positively charged nitrogen atoms in the lipid head and the negatively charged DNA phosphates.<br> D) The adsorbed DNA exhibits an ordered structure if it is complexed by barium, magnesium, calcium or manganese ions. The spacing between parallel DNA strands depends on the size of the DNA fragments as well as on the kind of cation. The largest DNA-spacings are observed with barium ions, followed by magnesium and calcium ions. DNA-complexation with manganese ions causes the smallest spacings. This order of ions is observed for both genomic and plasmid DNA.<br> E) Compression of the monolayer does not influence the DNA spacings. Thus the interaction between the lipid monolayer and adsorbed DNA is only weak. The DNA must exist as a more or less separate layer under the lipid film.<br> <br><br> 2. Model system: Lipid bilayer at the solid/fluid-interface<br> Methods: <br> Neutron reflectometry (NR), and Quartz crystal microbalance (QCM-D)<br> Results:<br> A) The zwitterionic phospholipid DMPC (1,2-dimyristoyl phosphatidylcholine) does not form lipid bilayers on top of planar polyelectrolyte multilayers covered with the positively charged PAH (polyallylamine).<br> B) In contrast, DMPC forms a lipid bilayer with defects on top of the negatively charged PSS (polystyrolsulfonate) terminated polyelectrolyte cushion.<br> C) Genomic calf thymus DNA adsorbs only to the DMPC layer in presence of calcium ions. Different ions were not examined.<br> D) The negatively charged phospholipid DLPA (1,2-dilauryl-phosphatidic acid) also forms a lipid bilayer with defects on top of the PAH-terminated cushion.<br> E) The DNA adsorbs also to the DLPA layer only in the presence of calcium ions in the solution.<br> F) By addition of EDTA (ethylenediaminetretraacetic acid) the calcium cations are removed from the DLPA/DNA-complex and the complex dissociates. Thus the calcium induced formation of that complex is reversible.<br><br>

Made available in DSpace on 2014-08-20T14:36:55Z (GMT). No. of bitstreams: 1 Dissertacao_ Marla_Farias.pdf: 256980 bytes, checksum: 6e2beba1f0290a61687c9e4ec64963b6 (MD5) Previous issue date: 2007-07-29 / The iron toxicity may cause considerable reduction in rice productivity, depending on the intensity of symptoms and on the affected farming area. It is necessary to identify the factors that determine the occurrence of this nutritional disorder, as well as, the possible involvement of other nutrients, such as phosphorus and potassium, in its intensity. The objectives of this work were: (1) to determine the effect of phosphorus and potassium adubation on the iron toxicity; (2) to evaluate the development of the symptoms in the aerial part of the plants (BR-IRGA 409 cultivar) and (3) to test whether the symptoms occur when the relative concentration of Fe2+, in relation to the sum of the concentrations of divalent cations (Fe2+, Mn2+, Ca2+, and Mg2+) in the solution is very high. Hence, in green house, a factorial experiment 5x2 (five soils and two adubation levels with phosphorus and potassium), completely randomized design, with two repetitions, was conducted. The adubation levels were zero and 100 mg dm-3 of P and K. The evaluated indicators were visual symptoms of the aerial part of the plants, production of dry matter, content of macro and micronutrients in the tissue of Ca2+, Mg2+, Fe2+ and Mn2+ in the soil solution. The statistical analysis were carried using Analysis of Variance (Duncan test) at 5% probability. The phosphorus and potassium adubation promoted accented increase in the production of dry matter of the plants in all soils, however there was no effect of this adubation neither in the contents of the divalent cations in the tissue, nor in the concentrations of those in the solution. The model of the relation between the molar fraction in the tissue and the molar fraction in the soil solution depended on the cation: Ca, Mg and Mn which showed tendency for saturation. The absorption of Fe remained constant up to a molar fraction of 0,4 (above that it increased exponentially). In molar fraction of Fe up to 0,6 (maximum value obtained in the experiment) the plants did not show symptoms of iron toxicity. / A toxidez de ferro pode causar reduções consideráveis na produtividade do arroz, dependendo da intensidade dos sintomas e da área afetada na lavoura. É necessário identificar os fatores que determinam a ocorrência desta desordem nutricional, bem como, o possível envolvimento de outros nutrientes, como o fósforo e o potássio, na sua intensidade. Os objetivos deste trabalho foram: (1) determinar o efeito da adubação com fósforo e potássio na toxidez de ferro, (2) avaliar o desenvolvimento dos sintomas na parte aérea das plantas (cultivar BR-IRGA 409) e (3) testar a hipótese de que os sintomas ocorrem quando a concentração relativa do Fe2+, em relação à soma das concentrações dos cátions divalentes (Fe2+, Mn2+, Ca2+ e Mg2+) na solução for muito alta. Para tal, foi conduzido, em casa de vegetação, um experimento fatorial 5 x 2 (cinco solos e dois níveis de adubação com fósforo e potássio), em delineamento completamente casualizado, com duas repetições. Os níveis da adubação foram zero e 100 mg dm-3 de P e K. Os indicadores avaliados foram sintomas visuais na parte aérea das plantas, produção de matéria seca da parte aérea, teores de macro e micronutrientes no tecido e concentrações de Ca2+, Mg2+, Fe2+ e Mn2+ na solução do solo. Os dados obtidos foram submetidos à Análise de Variância (teste de Duncan) a 5% de probabilidade. A adubação com fósforo e potássio promoveu acentuado aumento na produção de matéria seca das plantas em todos os solos, porém não houve efeito desta adubação nos teores dos cátions divalentes no tecido, nem nas concentrações dos mesmos na solução do solo. O modelo da relação entre a fração molar no tecido e a fração molar na solução do solo depende do cátion: o Ca, o Mg e o Mn mostraram tendência à saturação. A absorção do Fe ficou constante até uma fração molar de 0,4 (acima desta aumentou exponencialmente); em fração molar de Fe de até 0,6 (máxima obtida no experimento) as plantas não apresentaram sintomas de toxidez de ferro.