No description available.

L’un des plus importants défis de la toxicologie est de pouvoir extrapoler les résultats issus des différentes phases de l’analyse du risque sanitaire à partir de systèmes expérimentaux vers les populations humaines. Dans ce contexte, les techniques globales dites "omiques" sont de plus en plus utilisées pour caractériser les différents états des systèmes biologiques. Ainsi, la toxicogénomique permet non seulement d’étudier les mécanismes d’action des composés, d’identifier des marqueurs d’exposition, mais aussi de générer des signatures moléculaires à potentiel prédictif. En effet, des composés ayant des signatures moléculaires semblables ont également de forts risques de présenter les mêmes effets toxicologiques. L’objectif de cette thèse est d’appliquer ce concept de manière systématique, en explorant les données publiées et disponibles dans les banques dédiées à la toxicogénomique via des modèles statistiques multivariés. Ces analyses ont pour objectif de permettre le regroupement et donc la classification des composés sur la base des gènes dont ils affectent l’expression. L’appartenance de produits toxiques bien caractérisés aux classes ainsi constituées permet alors d’émettre des hypothèses quant à la toxicité des autres composés. Un jeu de données quantitatives intégrant 18 études réalisées avec la même technologie de puce à ADN et chez une seule espèce a été assemblé. De ce jeu de données, 3022 signatures toxicogénomiques correspondant à 452 composés différents ont été obtenues. Des approches de classification non supervisées afin de définir des classes de traitements induisant des altérations transcriptionnelles proches ont été mises en place. 95 et 104 classes ont été obtenues en fonction des méthodes utilisées. Finalement, une attention toute particulière a été portée sur les potentiels nouveaux perturbateurs endocriniens et xénobiotiques reprotoxiques sur-représentés dans trois classes spécifiquement. 22 composés sont en cours de test sur une lignée cellulaire humaine exprimant les enzymes de la stéroïdogenèse (NCI-H295R) pour évaluer leur potentiel effet perturbateur endocrinien. L’ensemble de ce travail a ainsi permis de démontrer la pertinence de nos approches de toxicogénomique pour la prédiction des effets toxiques de composés testés dans d’autres organes et/ou chez d’autres espèces. Il se poursuit à l’heure actuelle par la mise en place d’une base de données, TOXsIgN, permettant l’hébergement et l’accès à l’ensemble de signatures de toxicogénomique générées dans ce projet. De même, mon travail a également permis la mise en place de plusieurs outils dédiés à la toxicologie prédictive et à la visualisation de données, tels que le navigateur de génomes comme le ReproGenomics Viewer (RGV). / The core aim of my thesis project is to develop predictive toxicology approaches based on the integration of massive toxicogenomics datasets using bioinformatics and biostatistics methodologies. Specific objectives include: (1) classification of chemicals based on toxicogenomics signatures, i.e. the set of genes whose expression is known to be positively or negatively altered after exposure to these compounds; (2) the association of the resulting classes with human disorders or deleterious phenotypes based on the well-known toxicants present in those classes; (3) the prediction of novel reprotoxicants and/or endocrine disruptors based on toxicogenomics signature similarities with known chemicals affecting testis development and function. The assembled toxicogenomics dataset contains 23,657 samples covering 7092 experimental conditions (one chemical, one dose, one exposure time, one tissue) for 541 chemicals in seven distinct tissues in the rat from 18 different studies. From this dataset, 3,022 experimental conditions corresponding to 452 distinct compounds are associated to a toxicogenomics signature containing more than ten genes showing an altered expression pattern after exposure. Using unsupervised classification methods, 95 chemical clusters were defined showing close toxicogenomics signatures. The phenotype association analysis using data extracted from de Comparative Toxicogenomics Database (CTD) allowed us to identify three clusters significantly enriched in known endocrine-disrupting chemicals. Currently, 22 compounds are being tested on a human cell line expressing the enzymes of steroidogenesis (NCI-H295R) to evaluate their potential endocrine disrupting effects. These researches allowed us to demonstrate the relevance of integrating massive toxicogenomics datasets to predict adverse effects of compounds tested in different organs. It is currently being pursued through the development of a novel repository, TOXsIgN. This resource provides a flexible environment to facilitate online submission, storage and retrieval of toxicogenomics signatures by the scientific community. Similarly, the current PhD project also yielded to the implementation of several tools dedicated to predictive toxicology and data visualization including the ReproGenomics Viewer (RGV).

Nanoparticles (particles less than 100 nm in at least one dimension) exhibit chemical properties that differ from their bulk counterparts. Furthermore, they exhibit increased potential for systemic toxicities due to their deposition deep within pulmonary tissue upon inhalation. Thus, standard regulatory assays alone may not always be appropriate for evaluation of their full spectrum of toxicity. Systems biology (e.g., the study of molecular processes to describe a system as a whole) has emerged as a powerful platform proposed to provide insight in potential hazard, mode of action and human disease relevance. This work makes use of systems biology to characterize carbon black nanoparticle-induced toxicities in pulmonary and extra-pulmonary tissues (i.e., liver and heart) in mice over dose and time. This includes investigations of gene expression profiles, microRNA expression profiles, tissue-specific phenotypes and plasma proteins. The data are discussed in the context of potential use in human health risk assessment. In general, the work provides an example of how toxicogenomics can be used to support human health risk assessment.

Nanoparticles (particles less than 100 nm in at least one dimension) exhibit chemical properties that differ from their bulk counterparts. Furthermore, they exhibit increased potential for systemic toxicities due to their deposition deep within pulmonary tissue upon inhalation. Thus, standard regulatory assays alone may not always be appropriate for evaluation of their full spectrum of toxicity. Systems biology (e.g., the study of molecular processes to describe a system as a whole) has emerged as a powerful platform proposed to provide insight in potential hazard, mode of action and human disease relevance. This work makes use of systems biology to characterize carbon black nanoparticle-induced toxicities in pulmonary and extra-pulmonary tissues (i.e., liver and heart) in mice over dose and time. This includes investigations of gene expression profiles, microRNA expression profiles, tissue-specific phenotypes and plasma proteins. The data are discussed in the context of potential use in human health risk assessment. In general, the work provides an example of how toxicogenomics can be used to support human health risk assessment.

Nanoparticles (particles less than 100 nm in at least one dimension) exhibit chemical properties that differ from their bulk counterparts. Furthermore, they exhibit increased potential for systemic toxicities due to their deposition deep within pulmonary tissue upon inhalation. Thus, standard regulatory assays alone may not always be appropriate for evaluation of their full spectrum of toxicity. Systems biology (e.g., the study of molecular processes to describe a system as a whole) has emerged as a powerful platform proposed to provide insight in potential hazard, mode of action and human disease relevance. This work makes use of systems biology to characterize carbon black nanoparticle-induced toxicities in pulmonary and extra-pulmonary tissues (i.e., liver and heart) in mice over dose and time. This includes investigations of gene expression profiles, microRNA expression profiles, tissue-specific phenotypes and plasma proteins. The data are discussed in the context of potential use in human health risk assessment. In general, the work provides an example of how toxicogenomics can be used to support human health risk assessment.

Human exposures to polycyclic aromatic hydrocarbons (PAHs) occur as components of complex mixtures. Evaluations of health risks posed by complex mixtures containing PAHs rely on the toxicological knowledge of prioritized PAH mixture components, assuming that these PAHs share a common mode of action and that the sum of the contributions of these PAHs equals the toxic potency of the mixture (i.e., additivity). Traditional toxicity testing methods emphasizing apical endpoints have had limited success at evaluating the validity of these assumptions. Toxicogenomic tools that are able to rapidly generate toxicologically-relevant and mechanistic information have gained acceptance in the regulatory arena for individual chemicals; however, the applicability of these tools for chemical mixtures has been inadequately addressed. This thesis used toxicogenomic tools to (1) improve the understanding of mechanisms underlying the adverse, toxicological responses induced by individual PAHs and (2) evaluate the contention that transcriptional profiles and pathway information can be used to critically examine interactions between individual PAHs in PAH-containing mixtures, and the assumption of additivity. Microarrays were used to profile gene expression changes (transcriptomes) in forestomach, liver, and lung tissues (targets of PAH exposure) from mice orally exposed to three doses of eight individual PAHs, two defined PAH mixtures, and one complex PAH-containing mixture (coal tar). The results revealed that each PAH induced transcriptional changes that were significantly associated with several pathways implicated in carcinogenesis. However, despite a uniform ability to induce DNA damage (i.e., DNA adducts), mutations, and increases in enzyme activity, the pathways differ across PAHs and tissues. A novel strategy that employs single-PAH transcriptome data to models of additivity revealed that the assumption of additivity in PAH mixtures is valid at the pathway level; however, the independent action model of additivity yielded better estimates compared to concentration addition (used in human health risk assessment of PAH mixtures) or generalized concentration addition. Additionally, predicted and observed coal tar-induced transcriptional benchmark doses were comparable to those derived from previously published coal tar-induced murine lung tumour incidence data. Overall, this thesis demonstrates the utility of toxicogenomic data to expand the current understanding regarding the toxic potential of individual PAHs and PAH-containing complex mixtures.

High-throughput transcriptomic profiling has become a ubiquitous tool to assay an organism transcriptome and to characterize gene expression patterns in different cellular states or disease conditions, as well as in response to molecular and pharmacologic perturbations. Refinements to data preparation techniques have enabled integration of transcriptomic profiling into large-scale biomedical studies, generally devised to elucidate phenotypic factors contributing to transcriptional differences across a cohort of interest. Understanding these factors and the mechanisms through which they contribute to disease is a principal objective of numerous projects, such as The Cancer Genome Atlas and the Cancer Cell Line Encyclopedia. Additionally, transcriptomic profiling has been applied in toxicogenomic screening studies, which profile molecular responses of chemical perturbations in order to identify environmental toxicants and characterize their mechanisms-of-action. Further adoption of high-throughput transcriptomic profiling requires continued effort to improve and lower the costs of implementation. Accordingly, my dissertation work encompasses both the development and assessment of cost-effective RNA sequencing platforms, and of novel machine learning techniques applicable to the analyses of large-scale transcriptomic data sets. The utility of these techniques is evaluated through their application to a toxicogenomic screen in which our lab profiled exposures of adipocytes to metabolic disrupting chemicals. Such exposures have been implicated in metabolic dyshomeostasis, the predominant cause of obesity pathogenesis. Considering that an estimated 10% of the global population is obese, understanding the role these exposures play in disrupting metabolic balance has the potential to help combating this pervasive health threat. This dissertation consists of three sections. In the first section, I assess data generated by a highly-multiplexed RNA sequencing platform developed by our section, and report on its significantly better quality relative to similar platforms, and on its comparable quality to more expensive platforms. Next, I present the analysis of a toxicogenomic screen of metabolic disrupting compounds. This analysis crucially relied on novel supervised and unsupervised machine learning techniques which I specifically developed to take advantage of the experimental design we adopted for data generation. Lastly, I describe the further development, evaluation, and optimization of one of these methods, K2Taxonomer, into a computational tool for unsupervised molecular subgrouping of bulk and single-cell gene expression data, and for the comprehensive in-silico annotation of the discovered subgroups.

O câncer é uma das principais causas de mortalidade em todo o mundo. O envelhecimento da população mundial, especialmente nos países em desenvolvimento, indica que esse índice tende a aumentar a cada ano. Como o câncer é uma doença complexa, uma melhor compreensão dos mecanismos moleculares envolvidos na carcinogênese e na resposta aos tratamentos atuais são essenciais para mudar esse cenário. Gemcitabina e rebecamicina são moléculas que apresentam atividade antitumoral, sendo que a primeira é usada como agente único no tratamento de adenocarcinoma pancreático, ou em conjunto com outras drogas em vários tipos de câncer, enquanto que alguns análogos da segunda já foram utilizados em testes clínicos de fase II em câncer de mama e pulmão. A gemcitabina é um análogo de desoxicitidina que interrompe a síntese de DNA, além de apresentar efeito inibitório sobre várias proteínas, como ribonucleotídeo redutase. A rebecamicina e vários de seus análogos são indolocarbazóis que incluem moléculas intercalantes de DNA, além de inibidores da topoisomerase I e de algumas proteína quinases, como PKC e Chk1. Alguns indolocarbazóis apresentam também importante atividade antimicrobiana. Os mecanismos de resistência a essas drogas são pouco conhecidos e alguns já foram sugeridos; porém, ainda não há consenso sobre as vias celulares envolvidas. A proposta desse trabalho foi identificar genes associados às respostas a gemcitabina e análogos de rebecamicina, usando Saccharomyces cerevisiae como modelo celular. Avaliamos a atividade biológica de nove análogos de rebecamicina sintéticos, destacando o composto mais promissor para estudos mais aprofundados. Por meio de análise direta de fenótipo em escala genômica, identificamos genes cuja deleção resulta em hipersensibilidade ao análogo de rebecamicina, dentre eles um sensor de estresse relacionado à via Pkc1/Mpk1 quinase (SLG1), uma lisina deacetilase (RPD3) e uma lisina demetilase (JHD2). Esses três genes estão relacionados a uma maior atividade de proteínas de resistência multidroga, sugerindo que diversas vias regulatórias são importantes na resposta a indolocarbazóis. Também verificamos, por análise quantitativa de fenótipo (ensaio Bar-Seq), que a deleção de genes com funções atualmente pouco relacionadas à gemcitabina conferiam maior sensibilidade a droga, como uma arginina metiltransferase (HMT1) e uma subunidade do complexo Cop9 sinalossomo (CSI1), o qual está relacionado a regulação da atividade de diversas proteínas, incluindo ribonucleotídeo redutase. Em conjunto, nossos resultados demonstram que vias ainda não relacionadas a resposta a esses compostos podem ser de grande contribuição para a compreensão dos mecanismos de resistência aos mesmos, auxiliando no desenvolvimento de novos tratamentos e fármacos. / Cancer is one of the main causes of death throughout the world. Population aging, especially in developing countries, points to an increase in cancer incidence over the coming years. As cancer is a complex disease, a better understanding of the mechanisms involved in carcinogenesis and in the current treatments is essential to change this scenario. Gemcitabine and rebeccamycin are molecules that present antitumoral activity. The first one is currently used as a single agent in pancreatic cancer treatment or in combination with other drugs in several types of cancer, while analogs of the second have been tested in phase II clinical trials in breast and lung cancers. Gemcitabine is a deoxycytidine analog that inhibits DNA synthesis, and also presents inhibitory effects over several proteins, such as ribonucleotide reductase. Rebeccamycin and its analogs are indolocarbazoles which include DNA intercalant molecules, as well as topoisomerase I and protein kinase inhibitors, such as PKC and Chk1. The mechanisms related to the resistance to these drugs are poorly understood and some of them have been suggested; even though, there is no consensus about the cellular pathways involved. The aim of this work was to identify genes associated to the response to gemcitabine and rebeccamycin analogs using Saccharomyces cerevisiae as a model cell. We evaluated the biological activity of nine synthetic rebeccamycin analogs, selecting the most promising compound for deeper studies. In a genome-wide qualitative phenotypic analysis, we have identified genes whose disruption resulted in hypersensitivity to the rebeccamycin analog. Among them, we found a stress-sensor involved in Pkc1/Mpk1 kinase pathway (SLG1), a lysine deacetylase (RPD3) and a lysine demethylase (JHD2). These three genes are related to an increased multidrug resistance activity, suggesting that several regulatory pathways play a role in the response to indolocarbazoles. We also found, by means of a quantitative phenotypic analysis (Bar-Seq assay), increased sensitivity to gemcitabine in null-mutants of genes whose functions are currently poorly related to the drug, like an arginine methyltransferase (HMT1) and a subunit of Cop9 signalosome complex (CSI1), which is related to activity regulation of several enzymes, including ribonucleotide reductase. Together, our results show that pathways which were previously unrelated to the response to these compounds can be useful for a better understanding of their resistance mechanisms, providing insights into new drug development and therapy.

O câncer é uma das principais causas de mortalidade em todo o mundo. O envelhecimento da população mundial, especialmente nos países em desenvolvimento, indica que esse índice tende a aumentar a cada ano. Como o câncer é uma doença complexa, uma melhor compreensão dos mecanismos moleculares envolvidos na carcinogênese e na resposta aos tratamentos atuais são essenciais para mudar esse cenário. Gemcitabina e rebecamicina são moléculas que apresentam atividade antitumoral, sendo que a primeira é usada como agente único no tratamento de adenocarcinoma pancreático, ou em conjunto com outras drogas em vários tipos de câncer, enquanto que alguns análogos da segunda já foram utilizados em testes clínicos de fase II em câncer de mama e pulmão. A gemcitabina é um análogo de desoxicitidina que interrompe a síntese de DNA, além de apresentar efeito inibitório sobre várias proteínas, como ribonucleotídeo redutase. A rebecamicina e vários de seus análogos são indolocarbazóis que incluem moléculas intercalantes de DNA, além de inibidores da topoisomerase I e de algumas proteína quinases, como PKC e Chk1. Alguns indolocarbazóis apresentam também importante atividade antimicrobiana. Os mecanismos de resistência a essas drogas são pouco conhecidos e alguns já foram sugeridos; porém, ainda não há consenso sobre as vias celulares envolvidas. A proposta desse trabalho foi identificar genes associados às respostas a gemcitabina e análogos de rebecamicina, usando Saccharomyces cerevisiae como modelo celular. Avaliamos a atividade biológica de nove análogos de rebecamicina sintéticos, destacando o composto mais promissor para estudos mais aprofundados. Por meio de análise direta de fenótipo em escala genômica, identificamos genes cuja deleção resulta em hipersensibilidade ao análogo de rebecamicina, dentre eles um sensor de estresse relacionado à via Pkc1/Mpk1 quinase (SLG1), uma lisina deacetilase (RPD3) e uma lisina demetilase (JHD2). Esses três genes estão relacionados a uma maior atividade de proteínas de resistência multidroga, sugerindo que diversas vias regulatórias são importantes na resposta a indolocarbazóis. Também verificamos, por análise quantitativa de fenótipo (ensaio Bar-Seq), que a deleção de genes com funções atualmente pouco relacionadas à gemcitabina conferiam maior sensibilidade a droga, como uma arginina metiltransferase (HMT1) e uma subunidade do complexo Cop9 sinalossomo (CSI1), o qual está relacionado a regulação da atividade de diversas proteínas, incluindo ribonucleotídeo redutase. Em conjunto, nossos resultados demonstram que vias ainda não relacionadas a resposta a esses compostos podem ser de grande contribuição para a compreensão dos mecanismos de resistência aos mesmos, auxiliando no desenvolvimento de novos tratamentos e fármacos. / Cancer is one of the main causes of death throughout the world. Population aging, especially in developing countries, points to an increase in cancer incidence over the coming years. As cancer is a complex disease, a better understanding of the mechanisms involved in carcinogenesis and in the current treatments is essential to change this scenario. Gemcitabine and rebeccamycin are molecules that present antitumoral activity. The first one is currently used as a single agent in pancreatic cancer treatment or in combination with other drugs in several types of cancer, while analogs of the second have been tested in phase II clinical trials in breast and lung cancers. Gemcitabine is a deoxycytidine analog that inhibits DNA synthesis, and also presents inhibitory effects over several proteins, such as ribonucleotide reductase. Rebeccamycin and its analogs are indolocarbazoles which include DNA intercalant molecules, as well as topoisomerase I and protein kinase inhibitors, such as PKC and Chk1. The mechanisms related to the resistance to these drugs are poorly understood and some of them have been suggested; even though, there is no consensus about the cellular pathways involved. The aim of this work was to identify genes associated to the response to gemcitabine and rebeccamycin analogs using Saccharomyces cerevisiae as a model cell. We evaluated the biological activity of nine synthetic rebeccamycin analogs, selecting the most promising compound for deeper studies. In a genome-wide qualitative phenotypic analysis, we have identified genes whose disruption resulted in hypersensitivity to the rebeccamycin analog. Among them, we found a stress-sensor involved in Pkc1/Mpk1 kinase pathway (SLG1), a lysine deacetylase (RPD3) and a lysine demethylase (JHD2). These three genes are related to an increased multidrug resistance activity, suggesting that several regulatory pathways play a role in the response to indolocarbazoles. We also found, by means of a quantitative phenotypic analysis (Bar-Seq assay), increased sensitivity to gemcitabine in null-mutants of genes whose functions are currently poorly related to the drug, like an arginine methyltransferase (HMT1) and a subunit of Cop9 signalosome complex (CSI1), which is related to activity regulation of several enzymes, including ribonucleotide reductase. Together, our results show that pathways which were previously unrelated to the response to these compounds can be useful for a better understanding of their resistance mechanisms, providing insights into new drug development and therapy.

CoordenaÃÃo de AperfeiÃoamento de Pessoal de NÃvel Superior / Amburana cearensis (cumaru) Ã uma leguminosa subutilizada da Caatinga. O alto teor proteico das sementes faz delas uma possÃvel nova fonte de proteÃnas alimentares. Foram determinadas as melhores condiÃÃes de extraÃÃo proteica para posterior produÃÃo do isolado proteico de cumaru (IPAc). As anÃlises tambÃm foram conduzidas com a farinha delipidada (FDAc) e com um isolado proteico de soja comercial (IPS). IPAc apresentou teor proteico > 90% e composiÃÃo de aminoÃcidos comparÃvel a IPS, com notÃvel deficiÃncia em metionina. Foram determinadas algumas propriedades funcionais para predizer sua adequaÃÃo como ingrediente na indÃstria alimentÃcia. Os resultados para FDAc, IPAc e IPS foram, respectivamente: solubilidade 20-79; 9-99 e 3-76%; capacidade de retenÃÃo de Ãgua 1,7; 3,9 e 7,5 g/g; capacidade de retenÃÃo de Ãleo 1,5; 6,3 e 1,6 g/g; capacidade emulsificante / estabilidade 52 / 46%; 55 / 51% e 53 / 21%; capacidade espumante / estabilidade 47 / 32%; 65 / 53% e 33 / 8 e menor concentraÃÃo geleificante 18, 16 e 12%. Para investigar se IPAc e IPS apresentam indÃcios de toxicidade, cÃlulas de adenocarcinoma humano foram expostas por 24 h aos isolados. Para comparaÃÃo, tambÃm foi realizada exposiÃÃo com os isolados de soja (IPGm), feijÃo comum (IPPv), feijÃo de porco (IPCe) e mamona (IPRc), a fraÃÃo proteica de mamona (FPRc), as lectinas PHA-E e ConA e o controle quÃmico (Crtl_Quim). O RNA das cÃlulas foi hibridizado em microarranjos contendo o genoma humano completo. Os perfis de expressÃo gÃnica apÃs exposiÃÃo Ãs proteÃnas foram comparados aos de PBS e Crtl_Quim e agrupados hierarquicamente. Crtl_Quim, IPAc e IPGm se mostraram inadequados para a continuaÃÃo das anÃlises por provocar um efeito superestimado. A anÃlise de vias e processos biolÃgicos afetados apontou que genes envolvidos com a biossÃntese de colesterol, estresse do RE e resposta imune foram superexpressos e genes envolvidos com o ciclo celular e replicaÃÃo do DNA foram subexpressos apÃs exposiÃÃo a IPPv, IPCe, PHA-E e ConA. As amostras IPRc e FPRc regularam para cima genes envolvidos com a apoptose e resposta imune e para baixo genes envolvidos com o ciclo celular e sinalizaÃÃo do cAMP. AnÃlise do mapa conectivo mostrou alta correlaÃÃo de ConA, PHA-E e IPCe com drogas bloqueadoras de canais de Ca2+ e de K+; enquanto IPRc e FPRc apresentam funÃÃo biolÃgica semelhante a pelo menos seis drogas inibitÃrias da sÃntese proteica. Conclui-se que IPAc tem grande potencialidade como novo alimento proteico com grande aplicabilidade na indÃstria de alimentos, devendo, no entanto, sofrer modificaÃÃes no mÃtodo de obtenÃÃo visando a melhora na solubilidade e a eliminaÃÃo do sal residual. / Amburana cearensis (cumaru) is an underutilized legume from Caatinga. Because of their high protein content its seeds are a possible novel source of food protein. Were determined the best conditions for protein extraction and preparation of protein isolate (IPAc).Analysis were carried out also with the defatted flour (FDAc) and with a marketed soybean protein isolate (IPS). IPAc showed protein content of 90% and amino acid composition compatible with the IPSâs composition, with remarkable low methionine content. Some functional properties were measured in order to predict the adequacy of the isolates as an ingredient for food industry. The results for FDAc, IPAc and IPS were, respectively: solubility 20-79; 9-99 e 3-76%; water holding capacity 1.7; 3.9 e 7.5 g/g; oil holding capacity 1.5; 6.3 e 1.6 g/g; emulsion formation / stability 52 / 46%; 55 / 51% e 53 / 21%; foamability / stability 47 / 32%; 65 / 53% e 33 / 8; least gelling concentration (LGC) 18, 16 and 12%. In order to investigate whether IPAc e IPS exhibit signs of toxicity, two cell lines of human adenocarcinoma (MCF-7 e Caco-2) were exposed to these isolates as well as to others prepared from soyabean (IPGm), white bean (IPPv), Jack bean (IPCe) and castor bean (IPRc), to the protein fraction from castor bean (FPRc), to the lectins PHA-E and ConA (50 Âg/mL) and to the chemical control (Crtl_Quim). Subcitotoxic doses were determined by viability test to later exposure assay for 24 h. The MCF-7 cellâs RNA were extracted and hybridized to a microarray containing the complete human genome. The expression profiles after the exposition to the proteins were compared to that ones generated after PBS and Crtl_Quim and hierarchicaly clustered. Crtl_Quim, IPAc and IPGm were unsuitable for further analysis by their overestimated effect. Analysis of biological pathways and process showed that genes involved with the cholesterol biosynthesis, ER stress and immune response were upregulated and genes involved with cell cycle and DNA replication were downregulated after exposure of cells to IPPv, IPCe, PHA-E and ConA. The samples IPRc and FPRc upregulated genes involved with apoptosis and immune response and downregulated genes involved with cell cycle and cAMP signaling. Connectivity Map analysis showed high correlation of ConA, PHA-E and IPCe with drugs that are Ca2+ and K+ channel blockers, while IPRc and FPRc showed biological function similar to at least six drugs inhibitors of protein synthesis. Itâs possible to conclude that the IPAc has a high potentiality as a novel protein food with wide applicability in the food industry and should, however, be modified in the method of obtaining aiming the improvement in its solubility and eliminating the residual salt.