Cellular life is highly complex. In order to expand our understanding of the workings of human cells, in particular in the context of health and disease, detailed knowledge about the underlying molecular systems is needed. The unifying theme of this thesis concerns the use of data derived from sequencing of RNA, both within the field of transcriptomics itself and as a guide for further studies at the level of protein expression. In paper I, we showed that publicly available RNA-seq datasets are consistent across different studies, requiring only light processing for the data to cluster according to biological, rather than technical characteristics. This suggests that RNA-seq has developed into a reliable and highly reproducible technology, and that the increasing amount of publicly available RNA-seq data constitutes a valuable resource for meta-analyses. In paper II, we explored the ability to extrapolate protein concentrations by the use of RNA expression levels. We showed that mRNA and corresponding steady-state protein concentrations correlate well by introducing a gene-specific RNA-to-protein conversion factor that is stable across various cell types and tissues. The results from this study indicate the utility of RNA-seq also within the field of proteomics. The second part of the thesis starts with a paper in which we used transcriptomics to guide subsequent protein studies of the molecular mechanisms underlying malignant transformation. In paper III, we applied a transcriptomics approach to a cell model for defined steps of malignant transformation, and identified several genes with interesting expression patterns whose corresponding proteins were further analyzed with subcellular spatial resolution. Several of these proteins were further studied in clinical tumor samples, confirming that this cell model provides a relevant system for studying cancer mechanisms. In paper IV, we continued to explore the transcriptional landscape in the same cell model under moderate hypoxic conditions. To conclude, this thesis demonstrates the usefulness of RNA-seq data, from a transcriptomics perspective and beyond; to guide in analyses of protein expression, with the ultimate goal to unravel the complexity of the human cell, from a holistic point of view. / <p>QC 20161121</p>

No description available.

While the etiology of ovarian cancer (OC) is not completely understood, evidence suggests that chronic inflammation may promote malignant transformation. However, familial history remains the strongest risk factor for developing OC and is associated with germline BRCA1 mutations, such as the 185delAG mutation. Normal human ovarian surface epithelial cells expressing the 185delAG mutant, BRAT, exhibit molecular and pathological changes that may contribute to OC oncogenesis. In the current study, I sought to determine whether BRAT could promote an inflammatory phenotype by investigating BRAT's impact on the expression of the proinflammatory cytokine, Interleukin-1β (IL-1β). Using a culture model system of normal human ovarian surface epithelial (OSE) cells with and without the BRCA1 185delAG frameshift mutation, BRAT, I investigated BRAT's role in IL-1β expression. OSE cells stably expressing the 185delAG mutation and ovarian surface epithelial cells with endogenous 185delAG were analyzed for differential target gene expression by real time PCR, western blot, ELISA, luciferase reporter and siRNA assays. Normal and malignant breast epithelial cell lines transiently expressing BRAT were also evaluated by real time PCR to determine whether BRAT-induced IL-1β expression is tissue specific. BRAT-expressing OSE cells exhibited enhanced IL-1β mRNA and protein expression. However, expression of BRAT in all breast cell lines failed to significantly alter IL-1β expression levels so that BRAT-mediated IL-1β expression promoting a chronic inflammatory phenotype conducive to malignant transformation may be limited to the ovary. Secondly, since OSE cells expressing the BRCA1 185delAG mutation have increased levels of IL-1β that may contribute to malignant transformation, in a pilot study, I sought to assess whether elevated urinary levels of IL-1β are associated with OC as well as compare urinary IL-1β levels with clinical parameters. Urinary and serum levels of IL-1β were analyzed by ELISA and biostatistical analysis from a patient cohort consisting of healthy women (N=10), women with ovarian benign disease (N=23), women with OC (N=32), women with other benign gynecological conditions (N=22), and women with other gynecological cancers (N=6). Urinary IL-1β levels were elevated in patients with ovarian benign disease and a first degree family history of ovarian and/or breast cancer. Urinary IL-1β levels were also correlated with increased body mass index. Urinary and serum IL-1β levels were increased in ovarian benign and OC patient samples supporting the theory of elevated urinary IL-1β being associated with cancer progression. Lastly, I sought to begin early molecular characterization of BRCA1 185delAG to better understand its role in ovarian transformation. I isolated 185delAG protein expressed in E. coli and utilized web tools to analyze the amino acid sequence to determine the molecular and structural characteristics. The study results showed the predicted BRCA1 185delAG protein product is an ordered, self-aggregating, alpha helical protein structurally and molecularly distinct from wild-type BRCA1. The BRCA1 185delAG amino acid sequence contained domains with resemblance to the Peptidase M20 family. Isolation of the BRCA1 185delAG protein product will allow for further protein analysis to better understand its' oncogeneic functions; as well as, elucidate the mechanism of tissue-specific BRAT-mediated IL-1β expression since increased IL-1β expression may represent an early step contributing to OC.

Abstract Although the critical role of the small GTPases Ras and Ral in oncogenesis has been well documented, much remains to be investigated about the molecular mechanism by which these GTPases regulate malignant transformation. The work under this thesis made two major contributions to this field. The first is the discovery that K-Ras, RalA and/or RalB are required for the maintenance of the high levels of the anti-apoptotic protein survivin in some human cancer cells, and the second is the demonstration that down regulation of K-Ras, RalA and/or RalB, but not Raf-1 or Akt1/2, stabilizes the tumor suppressor p53 and reactivates it to inhibit malignant transformation in human cancer cells with mutant K-Ras and wild type p53. Here we found that depletion of K-Ras leads to decreased survivin levels in human cancer cells that harbor mutant K-Ras but not those with wild type Ras. The mechanism by which this occurs involves ubiquitination and subsequent proteasome-mediated degradation of survivin. The presence of mutant K-Ras alone was not sufficient to predict the effects of RalA/B depletion on survivin levels. Indeed, depletion of RalA and/or RalB reduces survivin levels in human cancer cells with wild type p53 and mutant K-Ras, but not in those with mutant p53 and/or wild type K-Ras. The functional relevance of these findings to malignant transformation was further supported by the demonstration that compromising the expression of survivin by siRNA leads to reduction of mutant K-Ras-driven invasion and anchorage-independent growth. Furthermore, in this thesis, we have discovered that down regulation of K-Ras, RalA and/or RalB using siRNA leads to increased levels of functional p53 that is capable of regulating its target genes. The mechanism by which depleting K-Ras, RalA and RalB increases the levels of p53 involves an increase in the half-life of the p53 protein concurrent with an increase in the phosphorylation of serine-15 of p53, a marker of p53 stability. Finally, we demonstrated that depletion of K-Ras, RalA and/or RalB interferes with cell cycle progression, anchorage-independent growth and invasion in a p53-dependent manner. In summary, the studies suggest that mutant K-Ras contributes to the maintenance of the aberrantly-high survivin levels by regulating its stability, and that the ability of mutant K-Ras to induce malignant transformation is, at least in part, dependent of these high levels of survivin. The work of this thesis also suggests that the expression of K-Ras, RalA and/or RalB proteins is critical to maintain low levels of p53, and that down regulation of these GTPases reactivates p53 by significantly enhancing its stability, and this contributes to suppression of malignant transformation.

Epigenetic dysfunction is a known contributor in carcinogenesis, and is emerging as a mechanism involved in toxicant-induced malignant transformation for environmental carcinogens such as arsenicals. In addition to aberrant DNA methylation of single genes, another manifestation of epigenetic dysfunction in cancer is agglomerative DNA methylation, which can participate in long-range epigenetic silencing that targets many neighboring genes and has been shown to occur in several types of clinical cancers. Using in vitro model systems of toxicant-induced malignant transformation, we found hundreds of aberrant DNA methylation events that emerge during malignant transformation, some of which occur in an agglomerative fashion. In an arsenite-transformed prostate epithelial cell line, the protocadherin (PCDH), HOXC and HOXD gene family clusters are targeted for agglomerative DNA methylation. Aberrant DNA methylation in general occurred more often within H3K27me3 stem cell domains. We found a striking association between enrichment of H3K9me3 stem cell domains and toxicant-induced agglomerative DNA methylation. Global gene expression profiling of the arsenite-transformed prostate epithelial cells showed that gene expression changes and DNA methylation changes were negatively correlated, but less than 10% of the hypermethylated genes were down-regulated. These studies confirm that a majority of the DNA hypermethylation events occur at transcriptionally repressed, H3K27me3 marked genes. In contrast to aberrant DNA methylation targeting H3K27me3 pre-marked silent genes, we found that actively expressed ZNF genes marked with H3K9me3 on their 3' ends, are preferred targets of DNA methylation linked gene silencing. H3K9me3 mediated gene silencing of ZNF genes was widespread, occurring at individual ZNF genes on multiple chromosomes and across ZNF gene family clusters. At ZNF gene promoters, H3K9me3 and DNA hypermethylation replaced H3K4me3, resulting in a widespread down-regulation of ZNF gene expression which accounted for 8% of all the down-regulated genes in the arsenical-transformed cells. In summary, these studies associate arsenical exposure with agglomerative DNA methylation of gene family clusters and widespread silencing of ZNF genes by DNA hypermethylation-linked H3K9me3 spreading, further implicating epigenetic dysfunction as a driver of arsenical-induced carcinogenesis.

Abstract Type XIII collagen is a type II oriented transmembrane protein with a short intracellular domain, a single transmembrane domain and a large, mostly collagenous extracellular domain. Tissue localization and cell culture studies have implicated that it is involved in cell adhesion. The spatio-temporal expression of type XIII collagen mRNA and protein during murine development is studied here. Type XIII collagen mRNAs were expressed at a constant rate during development, with an increase of expression towards birth. The strongest expression was detected in the central and peripheral nervous systems of the developing mouse fetus. Cultured primary neurons expressed this collagen, and recombinant type XIII collagen was found to enhance neurite outgrowth. Strong expression was also detected in the heart, with localization to cell-cell contacts and perinatal accentuation in the intercalated discs. Other sites of type XIII collagen expression included cartilage, bone, skeletal muscle, lung, intestine and skin. Clear developmental shifts in expression suggest a role in endochondral ossification of bone and the branching morphogenesis in the lung. To elucidate the function of type XIII collagen transgenic mice were generated by microinjection of a cDNA construct that directs the synthesis of truncated α1(XIII) chains with an in-frame deletion of the central collagenous COL2 domain. This construct was thought to disrupt the assembly of normal type XIII collagen trimers. Expression of shortened α1(XIII) chains by fibroblasts derived from mutant mice was demonstrated, and the lack of intracellular accumulation in immunohistochemical analysis of tissues suggested that the mutant molecules were expressed on the cell surface. Transgene expression led to developmental arrest and fetal mortality in offspring from heterozygous mating with two distinct phenotypes. The early phenotype fetuses were aborted by day 10.5 of development due to a failure in the fusion of the chorion and allantois membranes and subsequent disruption in placentation, while the late phenotype fetuses were aborted by day 13.5 of development due to cardiovascular and placental defects. Furthermore, it was shown that the heterozygous mice that were initially of normal appearance and bred normally had an increased susceptibility to develop T-cell lymphomas and angiosarcomas later in life. The results presented here increase the evidence that type XIII collagen is involved in cell adhesion, with several important tasks during development. A role of type XIII collagen in malignant transformation of certain mesenchymal cell populations is also implicated.

Les adénomes hépatocellulaires (AHC) sont des tumeurs bénignes rares dérivant de la prolifération d'hépatocytes et se développent dans un contexte hépatique par ailleurs normal, principalement chez des femmes suite à la prise de contraceptifs oraux. D'autres études ont identifié des facteurs de risque supplémentaires comme le syndrome métabolique et des mutations germinales d’HNF1α. Les travaux réalisés dans le laboratoire ont permis le classement des AHC en cinq sous-groupes : 1) les AHC-H avec des mutations inactivatrices du facteur de transcription HNF1α, 2) AHC-I avec une activation de la voie inflammatoire, 3) B-AHC portant des mutations activatrices du gène CTNNB1, codant pour la β-caténine 4) Sh-AHC avec une activation de la voie sonic hedgehog et 5) U-AHC sans altération moléculaire spécifique trouvée à ce jour. Les AHC montrent un faible risque de transformation maligne en carcinome hépatocellulaire (CHC). Le CHC est la deuxième cause de décès liés au cancer dans le monde entier, ce qui en fait un problème majeur de santé publique. Le CHC se développe dans 90% des cas sur un foie cirrhotique et est lié à une infection virale (VHB et HCV), à une consommation d'alcool et au syndrome métabolique. L'une des modifications génétiques les plus fréquentes dans les CHC est la réactivation de la télomérase (TERT), une enzyme clé nécessaire à la maintenance des télomères. Tout d'abord, dans nos travaux, nous avons cherché à établir une corrélation génotype-phénotype des AHC activés pour la β-caténine pour mieux comprendre l'hétérogénéité des tumeurs et le risque de transformation maligne. Notre étude a permis d'identifier trois sous-groupes d’AHC mutés CTNNB1 selon le type de mutation : mutations fortement, modérément et faiblement activatrices. De façon intéressante, chaque type de mutation présentait immuno marquage spécifique de la GS et un risque de transformation maligne diffèrent. En plus, nous avons identifié la mutation du promoteur de TERT comme un événement tardif au cours de la tumorigenèse nécessaire à la transformation des AHC-B en CHC. Deuxièmement, en raison de l’importance de la réactivation de TERT dans les tumeurs du foie, nous avons testé un nouveau traitement potentiel basé sur des oligonucléotides antisens conduisant à l’inhibition de l’expression transcriptionelle de TERT. En accord avec l'érosion progressive des séquences télomériques qui se produisent à chaque division cellulaire, nous avons pu observer un effet sur la prolifération cellulaire après traitement à long terme. Nous avons montré que les effets observés étaient liés au raccourcissement des télomères. De plus, nous avons identifié AAV2 comme un nouveau virus lié au développement de rares cas de CHC. Nos résultats sont soutenus par l'expansion clonale des cellules tumorales avec insertion virale, la surexpression des gènes insérés et l'insertion observée uniquement dans des gènes déjà connus pour être impliqués dans la tumorigenèse comme TERT, TNFSF10, MLL4, CCNA2 et CCNE1. En conclusion, dans nos études, nous avons affiné la classification des AHC, avec des possibles applications dans la prise en charge clinique des patients pour identifier le meilleur traitement selon le type de mutation et le risque de transformation maligne. De plus, nous avons obtenu des résultats prometteurs concernant l'utilisation de TERT comme cible thérapeutique. Cependant, des études supplémentaires sont nécessaires pour mieux évaluer les effets du traitement par ASO contre le TERT sur la physiologie cellulaire et les avantages possibles avec des combinaisons thérapeutiques. Enfin, nos résultats remettent en question la sécurité des thérapies génique basées sur l'AAV actuellement en cours, mais des études supplémentaires sont actuellement en cours au laboratoire pour comprendre les conséquences et les mécanismes impliqués dans de l'insertion virale. / Hepatocellular adenomas (HCA) are rare benign tumors deriving from proliferation of hepatocytes occurring in normal liver background mainly of female patients with history of oral contraceptive assumption. Further studies identified additional risk factors as metabolic syndromes and germline mutations. Works performed in the laboratory allowed the classification of HCA in five subgroups: 1) H-HCA with inactivating mutation of transcription factor HNF1α, 2) I-HCA with activation of the inflammatory pathway, 3) B-HCA carrying activating mutations of CTNNB1 gene, coding for β-catenin 4) Sh-HCA with activation of the sonic hedgehog pathway and 5) U-HCA with no specific alteration found to date. HCA show an average low risk of malignant transformation in hepatocellular carcinoma (HCC). HCC is the second cause of cancer-related deaths worldwide making it a major problem of public health. HCC development occurs in 90% of the cases on cirrhotic background and is linked to virus infection (HBV and HCV), alcohol intake and metabolic syndromes. One of the most frequent genetic alterations in HCC is the reactivation of TERT, a key enzyme necessary for telomeres maintenance. First, in our works we aimed to perform a correlation between genotype and phenotype of β-catenin activated HCA to better understand tumor heterogeneity and risk of malignant transformation. Our study allowed the identification of three subgroups of CTNNB1 mutated HCA according to the mutation type: highly, moderately and weakly activating mutations. Interestingly each type of mutation showed specific IHC staining for GS and risk of malignant transformation. Additionally we identified TERT promoter mutation as a late event in tumorigenesis necessary for HCA transformation in HCC with a significant correlation with CTNNB1 alterations. Second, due to the relevance of TERT reactivation in tumors, we tested a new potential treatment based on antisense oligonucleotides leading to TERT transcript ablation and lack of protein production. According to the slow erosion of the telomere sequences that occurs at each cell division, we were able to observe some effect on cell proliferation with long-term treatment. We showed that the observed effects were linked to telomere shortening. Further, we identified AAV2 as a new virus linked to tumor formation. Our results are supported by the clonal expansion of cells with viral insertion, overexpression of the inserted genes and insertion observed only in genes already known to be involved in tumorigenesis like TERT, TNFSF10, MLL4, CCNA2 and CCNE1. Taken together, in our studies we refined HCA classification, with possible direct applications in the clinical care of the patients in identifying the best treatment according to mutation type and risk of malignant transformation. Additionally we obtained promising results regarding using TERT as a therapeutic target. Still, additional studies are needed to better assess the effects of ASO treatment against TERT on cell physiology and possible advantages in using combination therapies. Finally, our findings challenge the safety of AAV based targeted therapies currently ongoing but additional studies to understand the effective consequences and mechanisms of viral insertion are object of current work in the laboratory.

The energy metabolism of most tumor cells relies on aerobic glycolysis (Warburg effect) characterized by an increased glycolytic flux that is accompanied by the increased formation of the cytotoxic metabolite methylglyoxal (MGO). Consequently, the rate of detoxification of this reactive glycolytic byproduct needs to be increased in order to prevent deleterious effects to the cells. This is brought about by an increased expression of glyoxalase 1 (GLO1) that is the rate-limiting enzyme of the MGO-detoxifying glyoxalase system. Here, we overexpressed GLO1 in HEK 293 cells and silenced it in MCF-7 cells using shRNA. Tumor-related properties of wild type and transformed cells were compared and key glycolytic enzyme activities assessed. Furthermore, the cells were subjected to hypoxic conditions to analyze the impact on cell proliferation and enzyme activities. Our results demonstrate that knockdown of GLO1 in the cancer cells significantly reduced tumor-associated properties such as migration and proliferation, whereas no functional alterations where found by overexpression of GLO1 in HEK 293 cells. In contrast, hypoxia caused inhibition of cell growth of all cells except of those overexpressing GLO1. Altogether, we conclude that GLO1 on one hand is crucial to maintaining tumor characteristics of malignant cells, and, on the other hand, supports malignant transformation of cells in a hypoxic environment when overexpressed.

Les môles hydatiformes sont une prolifération placentaire prétumorale pouvant évolueren tumeur alors traitée par chimiothérapie. Afin de réduire la mortalité et d’optimiser laprise en charge thérapeutique, l’objectif de cette thèse est d’identifier les gènespermettant de prédire la transformation en tumeur post môlaire et la chimiorésistance.Concernant la prédiction de la transformation, l’analyse de l’expression de gènescandidatssur tissu molaire décrit la relocalisation apicale de la Syncytine-1 en cas detransformation maligne, sans modification de transcription de ses récepteurs ni de deuxautres enveloppes rétrovirales placentaires. L’analyse sans à priori du transcriptome par3 méthodes différentes n’a pas permis d’identifier de gène différentiellement expriméselon la transformation. Cela suggère que la variabilité interindividuelle et les diverscritères utilisés pour le diagnostic de tumeur nuisent à l’identification de biomarqueursrobustes.Concernant la prédiction de la chimiorésistance, une approche transcriptomique largespectre sur tissu tumoral de choriocarcinome identifie une réduction de transcriptiond’HLA-G en cas de monochimiorésistance, confirmée au niveau protéique par immunohistochimie. L’analyse en réseaux de l’ensemble des gènes différentiellementexprimés suggère que la monochimiorésistance est associée à une altération de ladifférenciation des lymphocytes T alors que la polychimiorésistance est associée à unealtération de la prolifération des cellules sanguines.In fine, l’objectivation de l’expression trophoblastique du point de contrôle PD-L1 aconduit à évaluer l’efficacité d’un anti PD-L1 chez les patientes chimiorésistantes. Lesrésultats encourageants de cet essai et la possibilité de stratifier les patientes à l’aidedes marqueurs HLA-G et Syncytine-1 incitent à évaluer la place de l’anti PD-L1 associéà une monochimiothérapie en première ligne de traitement des tumeurstrophoblastiques. / Hydatidiform moles are a pretumoral placental proliferation which can turn into a tumorrequiring chemotherapy. In order to reduce mortality and propose an optimal therapeuticmanagement, the aim of this thesis is to identify genes which are predictive of postmolartumor transformation and chemoresistance.Concerning the prediction of transformation, the expression analysis of candidate-geneson molar tissue shows a relocalization of Syncytin-1 at the syncytiotrophoblast apicalborder in moles followed by malignant transformation, without modification oftranscription of its receptors and two other retroviral placental envelopes. A wholetranscriptomeapproach using 3 different microarrays-based methods did not identify anydifferentially expressed gene according to the post molar evolution. This may reflect thatinter-individual variability and the different criteria used for tumor diagnosis impede theidentification of robust biomarkers.Concerning the prediction of chemoresistance, a broad-spectrum transcriptomicapproach on choriocarcinoma tumor tissue identifies a down regulation of HLA-G in case of monochemoresistance, confirmed at the protein level by immunohistochemistry.Pathway analysis of the differentially expressed genes suggests thatmonochemoresistance is associated with impaired T-cell differentiation, whereaspolychemoresistance is associated with impaired proliferation of blood cells.Ultimately, the evidence of trophoblastic ubiquitous expression of the PD-L1 immunecheckpoint led us to the evaluation of the efficacy of PD-L1 blockade in chemoresistantpatients. The encouraging results of this trial and the possibility of stratifying patientswith HLA-G and Syncytin-1 markers encourages the assessment of PD-L1 blockadecombined with monochemotherapy as a first line treatment for trophoblastic tumors.

Die Zahl der Kolonkarzinome in den westlichen Industrieländern steigt in den letzten Jahren stetig an. Zu den Verbindungen, die mit der Zubereitung der Nahrung entstehen, mit ihr aufgenommen werden und die Kolonkanzerogenese möglicherweise begünstigen, gehört das heterozyklische aromatische PhIP, das bei der Erhitzung proteinreicher Nahrungsmittel entsteht. Neben zahlreichen Fütterungsversuchen an Nagern existieren auch Zellkulturmodelle zur Untersuchung der molekularen Mechanismen der PhIP-induzierten Kolonkanzerogenese. Die chemische Transformation von Zellen sollte durch wiederholte Exposition gegenüber dem hydroxylierten Metaboliten des Kanzerogens (N2-OH-PhIP) erzielt werden. Es wurden IEC-18-Zellen der Ratte und HCEC-Zellen des Menschen zur Untersuchung verwendet. Die Behandlung der IEC-18-Zellen führt nach 25 Behandlungszyklen mit Konzentrationen von 5 bis 20 µM nicht zur Transformation der Zellen. Die Anwesenheit von N2-OH-PhIP führt zu einer zehnfach erhöhten Induktion der GST-Aktivität, insbesondere der Untereinheiten GST-A1, -A3, -Pi und -T2, die für die effiziente Detoxifizierung des N-Acetoxy-Metaboliten vom N2-OH-PhIP verantwortlich sind. Bereits nach drei Behandlungen mit 1,5 µM N2-OH-PhIP konnte eine maligne Transformation der HCEC-Zellen erzielt werden. Die Zellen zeigten die charakteristischen Zeichen der Transformation: veränderte Wachstumseigenschaften wie klonales dreidimensionales Zellwachstum („pilling up“), Hemmung der Zell-Zell-Kontaktinhibierung, verkürzte Populationsverdopplungszeiten und tumorigene und metastasierende Eigenschaften. Außerdem exprimierten die N2-OH-PhIP-exponierten humanen Kolonzellen mit steigender Anzahl der Behandlungen größere Mengen des trunkierten APC-Proteins. Die bekannten PhIP-spezifischen Mutationen im APC-Gen resultieren in der Expression eines trunkierten Proteinproduktes und werden als frühe Ereignisse in der Kolonkanzerogenese betrachtet. Die zusammenfassende Betrachtung aller Ergebnisse zeigt, dass die IEC-18-Zelllinie zur chemischen Transformation durch N2-OH-PhIP ungeeignet ist. Dagegen wurde erstmalig eine vollständige chemische Transformation von Humandickdarmepithelzellen in vitro durch Exposition der humanen Kolonepithelzelllinie HCEC gegenüber dem Kolonkarzinogen N2-OH-PhIP erzielt. / In the last few years a strong increase in the incidence of colorectal cancer has been observed. As to the specific components in processed food responsible for the induction of colon cancerogenesis / it has been suggested that heterocyclic aromatic amines (HAA), e.g. the most abundant HAA 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP), which is formed in protein rich food, when it is cooked at high temperatures or over an open flame, might be involved in this process. Whereas a number of in vivo-models to study PhIP-mediated colon carcinogenesis are known, only a limited number of cell culture systems to study the HAA-mediated transformation of intestinal epithelial cells do in fact exist. In the present study IEC-18 cells (rat intestinal epithelial cells) and HCEC cells (human colon epithelial cells) were incubated with N2-OH-PhIP, the N-hydroxylated metabolite of PhIP. The IEC-18 cells could not be transformed despite 25 treatment cycles with 5 to 20 µM N2-OH-PhIP. This might be due to the fact that GST activity as well as the expression of the GST -A1, -A3, -Pi and -T2 units, which are responsible for the detoxication of the N-acetoxy derivative of PhIP were strongly induced by N2-OH-PhIP. In contrast, HCEC cells were malignantly transformed when exposed three times to 1.5 µM N2-OH-PhIP. The chemically-treated cells showed a reduced population doubling time, they lost cell-cell contact inhibition and started pilling up. Furthermore, if HCEC cells were injected subcutaneously into SCID mice tumors developed at the site of injection in all animals tested. The transformed HCEC cells also express high amounts of truncated APC protein, which in vivo appears at an early stage of colon cancerogenesis. Taken together, it has been shown that IEC-18 cells are not suitable for chemical transformation studies with the HAA metabolite N2-OH-PhIP. For the first time it has been shown that the HAA metabolite N2-OH-PhIP is indeed able to malignantly transform human colon epithelial cells in vitro.