In the field of integrative oncology, polyphenols have gained attention for their ability to modulate key signaling pathways involved in breast cancer prevention. One noteworthy product, the Polyphenol-Enriched Blueberry Preparation (PEBP), produced by the fermentation of blueberries by the bacterium Rouxiella badensis subsp. acadiensis has demonstrated various beneficial properties, including anti-inflammatory effects and the ability to control cancer stem cells. Cancer Stem Cells (CSCs) are highly tumorigenic cells involved in carcinogenesis and can cause relapses. MicroRNAs (miRNAs) can act as regulators of CSCs, by controlling stemness and invasion. We postulate that PEBP or its polyphenolic components induce specific epigenetic changes by modulating miRNA networks, reducing CSCs, and preventing breast carcinoma. Thus, the overarching aim of this thesis is to better understand the mechanisms underlying the protective effects of the polyphenol-enriched preparation in mitigating breast cancer. The objectives are: 1. To investigate the chemopreventive effects of PEBP on breast cancer stem cell development in cell models and in vivo, as well as to study the involvement of STAT3 and MAPKs signaling pathways 2. Assess the impact of the polyphenol-enriched blueberry preparation on breast cancer by regulating the expression signatures of miRNA involved with cell proliferation, survival, and CSC self-renewal pathways in in vitro experiments. 3. Characterize PEBP and investigate the effect of a subset of its components on miRNA expression. Furthermore, validate the role of those components in regulating the functional behavior of breast cancer stem cells through experiments using a 4T1 animal model. The results have shown a decrease in the formation of CSCs by delaying the development of tumors in vivo, decreasing metastasis to the lungs, and controlling the PTEN/PI3K/AKT axis, a central node in CSC signaling and homeostasis. In addition, several miRNAs associated with different clinical-pathological characteristics of breast cancer were shown to be differentially expressed in CSCs after exposure to PEBP. Notably, the expression of hypoxamir miR-210, associated with a poor prognosis in breast cancer patients, was downregulated, while tumor suppressor miR-145, which prevents metastasis through FOXO1 was over-expressed. The chemopreventive potential of a polyphenolic mixture containing PCA, gallic acid, and catechin, found in PEBP, was also shown to successfully reduce tumour growth and metastasis in our animal model and decrease the presence of stem-like tumour cells by favouring the upregulation of tumour-suppressor miR-145. These findings provide novel evidence in translational medicine, highlighting the effectiveness of a natural epigenetic modulator in chemoprevention by specifically targeting CSCs.

In recent years, research in the area of cancer stem cells has spiked tremendously. Numerous investigators have found that several types of cancers contain a subpopulation of tumor cells that display many defining characteristics of normal tissue stem cells, including multipotent differentiation potential, long-term self-renewal capacity, and expression of molecular markers of stemness. Most importantly, these cancer stem cells (CSCs) have very high tumor initiating potential, a finding that has led to the development of the cancer stem cell model for tumor progression. This model suggests that tumors are organized in a developmental hierarchy (similar to healthy tissue), with long-term tumor progression being driven by self-renewing CSCs at the top of the hierarchy. The CSC model represents a significant shift in our understanding of tumor progression, and as such, it may be possible to expand our knowledge of other aspects of tumor biology by re-examining them in the context of the CSC model. My work focuses on investigating interactions between CSCs and the tumor vasculature. Previous work has demonstrated heterogeneity in the proangiogenic potential of cells in a tumor. Considering the possibility that angiogenesis may be driven by specific subsets of tumor cells, I investigated the contribution of the CSC fraction to tumor angiogenesis. Comparing tumors with low or high CSC fractions, I have found that CSCs contribute to tumor vascular development through promotion of endothelial cell activity and recruitment of bone marrow-derived proangiogenic cells, mediated in part by vascular endothelial growth factor (VEGF) and stromal-derived factor 1 (SDF1). Since some tissue stem cells are known to reside in a vascular niche, I investigated the possibility that CSCs may also be supported by blood vessels in the tumor microenvironment, and that consequently CSCs may be targeted by disruption of tumor vasculature with antiangiogenic therapy. By testing multiple antiangiogenic therapeutic strategies, I have found that antiangiogenic therapy sensitizes CSCs to the effects of cytotoxic chemotherapy. Taken together, my work demonstrates a bi-directional relationship in which CSCs promote tumor vascular development, and tumor vasculature supports and protects CSCs. This work has implications for our understanding of CSC biology, tumor angiogenesis and antiangiogenic therapy, and provides insight into strategies for targeting the critical CSC population.

In recent years, research in the area of cancer stem cells has spiked tremendously. Numerous investigators have found that several types of cancers contain a subpopulation of tumor cells that display many defining characteristics of normal tissue stem cells, including multipotent differentiation potential, long-term self-renewal capacity, and expression of molecular markers of stemness. Most importantly, these cancer stem cells (CSCs) have very high tumor initiating potential, a finding that has led to the development of the cancer stem cell model for tumor progression. This model suggests that tumors are organized in a developmental hierarchy (similar to healthy tissue), with long-term tumor progression being driven by self-renewing CSCs at the top of the hierarchy. The CSC model represents a significant shift in our understanding of tumor progression, and as such, it may be possible to expand our knowledge of other aspects of tumor biology by re-examining them in the context of the CSC model. My work focuses on investigating interactions between CSCs and the tumor vasculature. Previous work has demonstrated heterogeneity in the proangiogenic potential of cells in a tumor. Considering the possibility that angiogenesis may be driven by specific subsets of tumor cells, I investigated the contribution of the CSC fraction to tumor angiogenesis. Comparing tumors with low or high CSC fractions, I have found that CSCs contribute to tumor vascular development through promotion of endothelial cell activity and recruitment of bone marrow-derived proangiogenic cells, mediated in part by vascular endothelial growth factor (VEGF) and stromal-derived factor 1 (SDF1). Since some tissue stem cells are known to reside in a vascular niche, I investigated the possibility that CSCs may also be supported by blood vessels in the tumor microenvironment, and that consequently CSCs may be targeted by disruption of tumor vasculature with antiangiogenic therapy. By testing multiple antiangiogenic therapeutic strategies, I have found that antiangiogenic therapy sensitizes CSCs to the effects of cytotoxic chemotherapy. Taken together, my work demonstrates a bi-directional relationship in which CSCs promote tumor vascular development, and tumor vasculature supports and protects CSCs. This work has implications for our understanding of CSC biology, tumor angiogenesis and antiangiogenic therapy, and provides insight into strategies for targeting the critical CSC population.

Loss of polarity is a defining characteristic of epithelial cancers. The cytoskeletal proteins, HUGL1 and HUGL2, mediate polarity in epithelial cells through diversified roles in defining membrane identity and trafficking to the basolateral membrane. Importantly, an ortholog of these molecules can inhibit tumor growth in Drosophila, although the mechanisms of their tumor suppressive functions in mammary epithelial cells are unknown. Here, we show nonredundant tumor protective roles for HUGL1 and HUGL2 in human mammary epithelial cells. Using a three dimensional culture system, we report that loss of HUGL1 or HUGL2 causes loss of apicobasal polarity, aberrant growth of multilayered epithelium, nuclear enlargement, loss of membrane identity, and cellular overgrowth. Experiments on plastic also revealed that HUGL1 or HUGL2 loss results in induction of a phenotypic EMT in breast epithelial cells and overexpression of HUGL1 in breast cancer cells reduces proliferation.In a Drosophila model of cancer driven by loss of lgl, we have discovered the consistent dysregulation of a number of miRNAs and mRNAs including the loss of let-7 and miR-9a, which are implicated in breast cancer and associated with the suppression of stem cells. Cross comparisons revealed a set of mRNAs that are both dysregulated in vivo and represent putative targets of the miRNAs changed in lgl mutants. Among these, Thrombospondin, a component of the extracellular matrix was found to be misexpressed in both flies and human cells lacking Lgl. Moreover, genetic interaction experiments showed miR-9a to be a functional effector of lgl in controlling proliferation in the wing. Taken together, the findings reported in this dissertation suggest that HUGL1 and HUGL2 function as tumor suppressors through their roles in polarity and miRNA regulation. These two proteins, functioning as modulators of cell plasticity and promoters of differentiation, are potentially able to control the transition between a differentiated epithelial cell and a cancer stem cell. This research offers new insight into the role of HUGL1 and HUGL2 in breast cancer and reveals novel targets downstream of polarity proteins for therapeutic intervention.

Indiana University-Purdue University Indianapolis (IUPUI) / The aldehyde dehydrogenase (ALDH) superfamily comprises a group of NAD(P)+-dependent enzymes that catalyze the conversion of aldehydes to their corresponding carboxylic acids. Of the nineteen human ALDH enzymes, members of the ALDH1A subfamily consisting of ALDH1A1, ALDH1A2, and ALDH1A3 have attracted interest as markers of cancer stem cells (CSCs) in several cancer types including lung, breast, and ovarian. CSCs represent a distinct subpopulation of highly tumorigenic cells that promote metastasis, recurrence, and resistance to conventional cancer therapies. The increased expression and activity of ALDH1A in CSCs is well-documented, as is the correlation between ALDH1A and a more aggressive cancer phenotype with poorer treatment outcomes. However, the actual functional role of ALDH1A in the context of CSCs has yet to be clearly defined. Elucidating this role will lead to a greater understanding of CSC biology and evaluate ALDH1A as a potential anti-CSC therapeutic target. In this study, previously developed and characterized selective small-molecule inhibitors of ALDH1A were used in conjunction with global transcriptomic, proteomic, and metabolomic analyses to identify pathways that could potentially establish a link between ALDH1A activity and early events in CSC formation in a triple-negative breast cancer (TNBC) model. These approaches revealed that ALDH1A inhibition is associated with mitochondrial and metabolic dysfunction and perturbation of the electron transport chain. ALDH1A inhibition also resulted in an increase in markers of endoplasmic reticulum (ER) stress and activation of the unfolded protein response (UPR), specifically mediated through the Protein kinase RNA-like endoplasmic reticulum kinase (PERK) pathway. These effects appear to occur independently of both the canonical function of ALDH1A in detoxifying reactive aldehydes as well as its potential metabolic contribution through the generation of NADH. Together, these results suggest a separate role for ALDH1A in TNBC CSCs in protecting against ER stress that warrants further study. / 2024-10-03

According to the cancer stem cell hypothesis a subpopulation of cells within a tumour has the capacity to sustain its growth. These cells are termed cancer stem cells, and are most simply defined as the cells within a primary tumour that can self-renew, differentiate and regenerate a phenocopy of that cancer when transplanted in vivo. Cancer stem cells have now been prospectively identified from numerous human tumours and are actively sought in many cancer types, both clinical and experimental. The cancer stem cell hypothesis remains controversial, with evidence both supporting and challenging its existence in human tumours and in animal models of disease. Here we prospectively identify and study brain cancer stem cells in clinically representative mouse models of the medulloblastoma and glioma. Cancer stem cells from both mouse brain tumour types are prospectively enriched by fluorescent activated cell sorting freshly dissociated cells for the surface antigen CD15, display a neural precursor phenotype, exhibit the hallmark stem cell characteristics of self-renewal and multilineage differentiation, and regenerate a phenocopy of the original tumour after orthotopic transplantation. Additionally, novel mouse medulloblastoma and glioma cancer stem cell lines were established and studied in vitro as adherent cultures in the same serum-free media conditions that support the growth of normal neural stem cells. When mouse and human glioma stem cell lines were compared, many novel molecular mediators of the tumour phenotype were identified, as were chemical compounds that selectively inhibit their growth. Our results have important implications regarding the cancer stem cell hypothesis, the mechanisms that drive brain tumour stem cell growth and the therapeutic strategies that may prove effective for the treatment of glioma and medulloblastoma.

O meduloblastoma é o tumor maligno do sistema nervoso central mais frequente na infância e adolescência. A expressão de genes tipicamente expressos em células-tronco está correlacionada com pior prognóstico em pacientes com meduloblastoma e a expressão de POU5F1 se mostrou capaz de distinguir pacientes com desfecho clínico desfavorável e pior sobrevida. Apesar do seu valor prognóstico, não há evidências diretas da contribuição de OCT4 para a aquisição de fenótipos mais agressivos em meduloblastoma. Nesse contexto, o presente trabalho investigou o papel da isoforma OCT4A em características pró-tumorigênicas de meduloblastoma in vitro e in vivo, e também avaliou as alterações moleculares que podem ser responsáveis pela aquisição de fenótipo mais agressivo em células de meduloblastoma humano. Para tanto, foi realizada a superexpressão de OCT4A mediada por retrovírus em três linhagens celulares de meduloblastoma (Daoy, D283Med e USP-13-Med). As células de meduloblastoma com superexpressão de OCT4A exibiram maior proliferação e alterações no ciclo celular. Foram observados também aumentos na atividade clonogênica, geração de esferas tumorais e desenvolvimento tumoral em modelo subcutâneo, sendo esses efeitos dependentes dos níveis de OCT4A. A avaliação da mobilidade celular in vitro demonstrou diminuição na adesão celular e aumento da invasão celular de esferoide 3D. Em modelo ortotópico de meduloblastoma, as células com superexpressão de OCT4A geraram tumores mais desenvolvidos, com fenótipos mais agressivos, infiltrativos e metastáticos. A superexpressão de OCT4A foi associada a maior instabilidade genômica, entretanto, as aberrações em números de cópias variaram em frequência e tipo de alteração dependendo da linhagem celular, e sendo pouco associada com os genes diferencialmente expressos. De forma interessante, uma relevante expressão diferencial de RNAs não-codificadores de proteínas foi observada em células de meduloblastoma com superexpressão de OCT4A, incluindo os recém descobertos e pouco caracterizados RNAs não codificadores longos, além de múltiplos RNAs pequenos nucleolares. Assim, os resultados aqui apresentados fundamentam a relevância de fatores envolvidos em pluripotência para o agravamento de traços associados com desfecho clínico desfavorável em meduloblastoma e destacam o valor prognóstico e terapêutico de OCT4A neste tumor pediátrico do sistema nervoso central / Medulloblastoma is the most common malignant brain tumor in infants. The expression of typical pluripotency genes is correlated with poor prognosis in medulloblastoma and POU5F1 expression was shown capable of discriminating patients with poor survival outcome. Despite this prognostic value, direct evidences of OCT4 contribution to more aggressive traits in medulloblastoma are missing. In this context, we investigated the role of OCT4A isoform on pro-tumorigenic features of medulloblastoma in vitro and in vivo and evaluated molecular alterations that could be responsible for acquisition of a more aggressive phenotype in medulloblastoma cells. Retroviral-mediated overexpression of OCT4A were performed in three medulloblastoma cell lines (Daoy, D283Med and USP-13-Med). Medulloblastoma cells overexpressing OCT4A displayed enhanced cell proliferation and cell cycle alterations. Increased clonogenic activity, tumorsphere generation capability and subcutaneous tumor development were also observed, and these effects were OCT4A expression level-dependent. Evaluation of cell mobility in vitro showed loss of cell adhesion and greater 3D-spheroid invasion. In an orthotopic model of medulloblastoma, OCT4A overexpressing cells generated more developed, aggressive, infiltrative and metastatic tumors. OCT4A overexpression was associated with chromosomal instability but copy number aberrations varied in frequency and type according to the cell line, with little association with differently expressed genes. Interestingly, marked differential expression of non-coding RNAs, including newly discovered, still poorly characterized, long non-coding RNAs and multiple small nucleolar RNAs were observed in medulloblastoma cells with OCT4A overexpression. Altogether, our findings support the relevance of pluripotency-related factors in the aggravation of medulloblastoma traits classically associated with poor clinical outcome, and underscore the prognostic and therapeutic value of OCT4A in this challenging type of pediatric brain cancer

Prostate cancer is a leading cause of cancer-related death in men, and while many men diagnosed with the disease will have an indolent clinical course, 20-25% of men will experience disease recurrence which is invariably lethal. There is an urgent need for prognostic biomarkers that will predict disease recurrence and risk-stratify patients upon diagnosis, allowing for personalised therapies. This thesis attempts to identify new prognostic biomarkers for prostate cancer and investigates their patterns of protein expression in human primary prostate tumour tissue. Cancer stem cells are cancer cells thought to be uniquely capable of self-renewal and tumorigenicity, and may have a role in tumour recurrence. Using a literature searching approach, potential biomarkers related to stem cells, cancer stem cells or recurrence in prostate cancer were identified, and ALDH7A1, BMI1, SDC1, MUC1-C, Nestin and ZSCAN4 were chosen for investigation. An in silico approach was also used for biomarker identification, with RS1 and SLC31A1 selected as their mRNA was found to be upregulated in recurrent tumours. The expression patterns of all 7 potential biomarkers were examined by immunohistochemistry on prostate tumour tissue and benign tissue from prostate biopsies and prostatectomies. BMI1, ALDH7A1, MUC1-C and Nestin showed no relationship to recurrence or other clinical features. RS1 protein levels increased in patients with recurrence within 5 years, negatively correlated with AR expression, and a meta-analysis showed that the RS1 gene was amplified in up to 32% of castration-resistant prostate tumours. ZSCAN4 was heterogeneously expressed in a subset of 26% of prostate tumours with unclear characteristics and was not expressed in benign tissue, but was not associated with recurrence. Finally, SDC1 expression was lost in tumour epithelium, but a population of unidentified SDC1-expressing cells were found in the stroma of a third of tumours, and an increased burden of these cells was associated with primary Gleason pattern 5 tumours. These cells do not overlap with common epithelial, mesenchymal or stromal lineages, but may be migratory. In summary, the data presented in this thesis identifies 3 potential new biomarkers for prostate cancer, and provides the basis for future characterisation of their wider roles in the disease.

Resistance to endocrine treatments in oestrogen receptor positive (ER+) breast cancer (BC) significantly contribute to patient morbidity and mortality. ER+ BC constitute 60% of all breast cancers although there is considerable clinico-pathological diversity within this group. Breast cancer initiating cells (BCICs) are implicated in tumour relapse and metastasis and are postulated to drive resistance to standard anti-cancer therapies. However little is known about the sensitivity of BCICs to endocrine therapies. We assessed the effect of tamoxifen treatment and acquired tamoxifen resistance on BCIC frequency in vitro and in vivo using breast cancer cell lines and, importantly, patient derived samples of early and metastatic ER+ breast cancer. In ER+ breast cancer, BCICs may be prospectively enriched in vitro by selecting cells by CD44+/CD24lo/ESA+ phenotype or by mammosphere initiating capacity (MIC). However the gold standard assay to determine BCIC frequency is limiting dilution transplantation in vivo. In the past it has been historically difficult to generate xenograft models of ER+ breast cancer using patient samples. In this thesis, using a novel experimental technique, patient-derived xenografts (PDX) of early and metastatic ER+ BC were generated with almost 85% efficiency in NOD/SCID IL2gammaR-/- (NSG) mice. PDX expressed ER and were able to undergo serial in vivo passage, matching the phenotype of the tumour from which they were derived. In this work, two patterns of response to tamoxifen treatment were observed in ER+ cell lines, patient derived breast cancer samples and xenografts during BCIC assays in vitro and in vivo; Limited Sensitivity (LS) or Resistance (R). In the LS group there was no change or a significant diminution in BCIC frequency in the presence of tamoxifen. In the R group, a significant increase in BCIC frequency was observed in the presence of tamoxifen. Furthermore BCIC activity was shown be enhanced by the acquisition of tamoxifen resistance using cell line models. Cellular populations enriched for BCICs in ER+ cell lines were shown to express low levels of ER compared to non-BCICs. Finally Notch (gamma-secretase inhibitor) and EGFR (gefitinib) pathway inhibitors were tested alone or in combination with tamoxifen against a panel of established and novel cell lines and ER+ patient-derived breast cancer samples for anti-BCIC activity. Tamoxifen treatment can increase BCIC frequency in vitro assays of cell lines and patient-derived samples and in vivo using patient-derived xenografts of ER+ breast cancer. However phenotypic diversity of BCIC may be present within the ER+ BC population. A pharmaceutical strategy to effectively treat BCICs alongside standard endocrine therapy is necessary for the effective future treatment of ER+ breast cancer.

Prostate cancer (PCa) is heterogeneous containing phenotypically diverse cells. It is unclear whether these phenotypically different PCa cells are functionally distinct and possess divergent tumorigenic potential. Androgen signaling plays important roles in differentiation and survival of malignant PCa cells, and prostate specific antigen (PSA) as one of the androgen signaling target genes is used as a biomarker of AR signaling to assess tumor progression and evaluate therapeutic efficiency in clinic. Here we present evidence for discordant AR and PSA expression resulting in AR⁺/PSA⁺, AR⁺/PSA⁻, AR⁻/PSA⁻, and AR⁻/PSA⁺ PCa cells in human tumors. We also show that prostate tumor PSA mRNA levels inversely correlate with poor clinical outcomes and patient survival. By employing a lentiviral reporter system, we have fractionated bulk PCa cells into PSA⁺ and PSA⁻[superscript '/lo'] cell populations, with the former being AR⁺/PSA⁺ and the latter containing both AR⁺/PSA⁻ and AR⁻/PSA⁻ cells. The PSA⁺ and PSA⁻[superscript '/lo'] PCa cells demonstrate distinct molecular, cellular, and tumor-propagating properties. PSA⁻[superscript '/lo'] PCa cells are quiescent and refractory to stresses including androgen deprivation, exhibit high clonogenic potential, and possess long-term tumor-propagating capacity. They preferentially express stem cell genes and can undergo asymmetric cell division to generate PSA⁺ cells. Of great clinical interest, PSA⁻[superscript '/lo'] PCa cells can initiate robust tumor development and resist androgen ablation in castrated hosts, and they harbor highly tumorigenic castration resistant PCa cells. In contrast, PSA⁺ PCa cells possess more limited tumor-propagating capacity, undergo symmetric division, and are sensitive to castration. Systemic androgen levels dynamically regulate the relative abundance of PSA⁺/PSA⁻[superscript '/lo'] PCa cells in the tumors, which in turn impact the kinetics of tumor growth. Further studies reveal that the PSA⁻[superscript '/lo'] PCa cell population harbors several overlapping but nonidentical tumorigenic subsets including ALDH⁺, CD44⁺, and [alpha]2[beta]1⁺ cells and ALDH⁺CD44⁺[alpha]2[beta]1⁺ can further enrich castration resistant PCa cells. These observations together suggest that heterogeneous PCa cells are organized as a tumorigenic hierarchy. Our results have important implications in understanding how different subpopulations of PCa cells manifest differential responses to current androgen deprivation therapy (ADT). / text