Item EmbargoInvestigating in vivo utilization of VLP vaccines adjuvanted with single, dual, or triple pathogen-associated molecular patterns(Bilkent University, 2023-08) Ceylan, Yasemin; Gürsel, İhsanCOVID-19 was first identified in Wuhan, a Chinese city, and it has been a part of our lives since late December 2019. The WHO declared this disease, caused by SARS-CoV-2, a pandemic due to its high transmission rate, a status that remained in effect until May 2023. The escalating death toll and virus spread expedited vaccine development efforts. Various platforms for vaccine development have been utilized, including inactivated virus, mRNA, and subunit vaccines. Given the virus's propensity for mutations due to its high transmission rate, the chosen platform must be easy to produce, effective, and safe for immunocompromised individuals. One such platform is virus-like particles (VLPs). VLP vaccines consist of viral proteins without genetic material and are known for their ease of production. A VLP vaccine, adjuvanted with Alum and K-type CpG ODN against SARS-CoV-2, is tested in Phase II clinical trials. In addition to CpG ODN, this study aimed to evaluate the efficacy of other adjuvants and their combinations to explore potential clinical advantages. The additional adjuvants included poly(I:C) and a member of cyclic dinucleotides, 2,3-cGAMP, which are Toll-like receptor 3 (TLR3) and STING agonists, respectively. Nine groups of male mice (C57BL/6, 6–8 weeks old) received intraperitoneal injections three times with a mixture of VLP vaccine using different combinations of adjuvants, including CpG ODN, CDN, and poly(I:C). Blood samples were collected from the tail vein two weeks after each injection. Antibody ELISA was employed to assess humoral immunity against each combination of vaccines, with plates coated using in-house recombinant 6P-Spike and commercial recombinant RBD. ELISA results indicated that VLPs combined with CpG plus poly(I:C) and VLPs combined with all three adjuvants induced high levels of total IgG against RecS. Two weeks after the third injection, the animals' spleens were isolated, and cytokine profiles were assessed through an antigen recall assay. As a result, only one group exhibited the expected cytokine profile: VLPs combined with the triple adjuvant combination. This profile demonstrated the requisite Th1-biased cytokine response necessary for vaccine development. This study suggests that the triple adjuvant combination holds promise as a vaccine adjuvant for addressing future pandemics. Item EmbargoAssessment of immune potency of SARS-CoV-2 VLP vaccine in mice, enhanced through different adjuvants and liposome complexation(Bilkent University, 2023-08) Abraş, İrem Fatma; Gürsel, İhsanSARS-CoV-2, emerging in December 2019 in Wuhan, China, led to a swift global pandemic declaration in March 2020, prompting widespread vaccine development. Most vaccines target specific regions, mainly the spike protein. In our lab, we have employed an innovative virus-like particle (VLP) vaccine approach encompassing all four structural proteins of the virus: spike, nucleocapsid, membrane, and envelope. Our study utilizes sterically stabilized cationic liposomes (SSCL) to encapsulate VLPs expressing the Delta variant spike protein, along with various adjuvants: CpG ODN, poly(I:C), and 2’3’-cGAMP. We characterized liposomes using tunable resistive pulse sensing for size and concentration. In C57BL/6 mice, we administered primary and two booster injections on Day 0, Day 15, and Day 73, respectively, collecting blood samples at intervals (Day 14, Day 28, Day 42, Day 72, and Day 90). To assess vaccine impact on mouse humoral immunity, we conducted ELISAs for total IgG, IgG1, and IgG2c antibodies against recombinant Spike and the receptor-binding domain (RBD). IgG titers increased until Day 42, remained stable or slightly decreased on Day 72, and significantly rose on Day 90. We calculated IgG2c/IgG1 ratios, reflecting Th1 immune responses, revealing enhanced cellular immunity potential in groups with adjuvants compared to the VLP-only group. This study underscores the effectiveness of our VLP vaccine strategy in stimulating robust immune responses and opens avenues for further research and development. Item EmbargoPrognostic biomarker identification and classification of colorectal cancer patients: a dual gene-based and sample-based approach(Bilkent University, 2023-08) Naeemaee, Ronak; Güre, Ali OsmayColorectal cancer (CRC) is one of the most heterogeneous cancer types, with high mortality rates making it the one of the deadliest cancer among men and women. The heterogeneity of CRC comes from the numerous clinicopathological characteristics of these tumors, including; KRAS/BRAF mutation, Microsatellite Instability (MSI), and stage. Another essential factor recently emphasized is the tumor location (proximal or distal). Consequently, many studies have focused on finding prognostic biomarkers for CRC patients to increase the efficiency of their treatment plans. However, despite the attempts, these biomarkers fail in clinical transition as they lack robustness and consistent results in their validation studies. Moreover, understanding the mechanism behind CRC progression can significantly help the personalization of treatments. Recently, the cancer neuroscience field has been focusing on elucidating neuropeptides' role in cancer and CRC as they have been proven to be involved in cancer progression. Accordingly, the thesis was divided into two approaches. The first approach was to further examine the role of neuropeptides by finding a subset of neuropeptides for the classification of the CRC samples and following functional analysis to understand the mechanism of their involvement. Moreover, the second approach attempted the determination of robust prognostic biomarkers in a specific sample group (Proximal, Stages 2 and 3) while controlling for the inconsistencies. In the first approach, a subset of 9 neuropeptide genes was found through Principle Component Analysis (PCA) with the ability to stratify the CRC samples into high and low expression groups. Functional analyses of these groups identified an association between the up-regulation of these neuropeptides and Hedgehog's (HHG) signaling pathway, and these activities are hypothesized to be primarily specific to the stroma of the tumor. Up-regulation of these neuropeptides was also linked with other pathways involved in cancer progression, such as; EMT, angiogenesis, and TGFβ activities. The second approach utilized a new methodology pipeline that aimed to ensure the selection of genes with no discrepancies among their probesets and across different technologies. Following the pipeline, 3 genes were identified, associated with favorable and non-favorable prognoses for Proximal, Stage 2, and 3 samples. However, although a very stringent methodology was used and various clinicopathological parameters such as the stage and location were considered, the prognostic associations observed were not as consistent, indicating the importance of the sample's molecular characteristics. This study also pointed out potential implications of neuropeptides in CRC progression and development. More elaborative studies are required for the clarification of the interactions of neuropeptides with the HHG signaling pathway. Furthermore, the identified prognostic biomarkers need to be validated through comprehensive validation studies in their associating subgroups of samples, as they are robust biomarkers with the potential to be used in clinics. Item Open AccessInvestigation of human IL18 structures towards understanding the dynamics and self-assembly of IL18-IL18BP heterodimers(Bilkent University, 2023-08) Yazıcı, Yılmaz Yücehan; Belkaya, SerkanInterleukin-18 (IL18) is an inflammatory cytokine involved in the regulation of both innate and adaptive immunity. IL18 activity is tightly regulated by its constitutively secreted inhibitor, IL18 binding protein (IL18BP). All known IL18BPs from human to poxviruses share an identical protein fold. Investigation of known crystal structures of human IL18 illustrated that the region between the 68th and 81st amino acids is either missing or disordered in all IL18-IL18BP complexes while adopting a 310 helical structure in the free or signaling complexes of IL18. The C74 residue in this region was reported to form a novel intermolecular disulfide bond in the human IL18-IL18BP tetrameric assembly. Yet, the impact of this small surface epitope on the stability and self-assembly of IL18 is unclear. First, we investigated the dynamics of this small surface region of IL18-IL18BP heterodimers by molecular dynamics simulations. In simulations, we used all known crystal structures of IL18-IL18BP and two additional computer models generated by AlphaFold2 and homology modeling to observe the differential folding of small surface regions. Next, we generated the proposed human and Yaba-like disease virus tetramers and analyzed them with or without intermolecular disulfide bonds. We found that the helical structure of small surfaces stabilizes the backbone of the complex through reducing its flexibility. While our computational model folded into a more stable and less flexible short IL18 epitope, similar to free IL18 or bound IL18 on the signal complex, IL18 from the complexes with human or virus IL18BPs displayed flexible regions with a less stable backbone. The core of the human crystal loosened, becoming more exposed to solvent in the human crystal structure. Particularly, a salt bridge stabilized the human IL18 structure at the helix configuration, linking the helix to the core region. We observed that the bivalent binding mode for the human tetramer, which is disturbing one side of the tetramer by breaking the disulfide bonds, did not affect the stability of the complex, and thereby the tetramer remained intact. Hence, the tetramer formation of IL18-IL18BP can be beneficial to the host as it provides an additional stability advantage. Overall, our results show that the short IL18 epitope between the amino acids 68 and 81 mediates stability and self-assembly of IL18-IL18BP heterodimers. Item EmbargoPhosphoproteomic analysis of class IA P110β isoform-specific signal transducers upon PI3K pathway activation(Bilkent University, 2023-08) Daloğlu, Beril; Çizmecioğlu, OnurPI3K pathway activation is a common event observed in various human cancers. As it regulates cell proliferation, migration and metabolism, it has been widely targeted for anticancer therapies and alteration of drug resistance. Studies focusing on PI3K Class IA isoform-specific inhibition has become critical to achieve alternative targeted treatment methods. Although previous findings show that Class IA isoforms differ in their routes of activation, the downstream targets specific to the isoforms have not been fully profiled yet. This study focuses on the phosphoproteomic analysis of p110 isoform-specific downstream molecules upon PI3K pathway activation. We have created a doxycycline-inducible system in Mouse Embryonic Fibroblasts (MEFs) to express wildtype PIK3CB gene encoding p110β or its catalytic and helical domain over-activating mutants, E1051K and N553S. By immunoblotting, the changes in downstream phosphorylation events were analyzed for MEF p110β mutants or for MEF p110β WT cells upon mitogenic PI3K stimulation. These results were collaborated with the comparative phosphoproteomic data obtained from LC-MS/MS Mass Spectrometry analysis. Our findings show that upon short-term stimulation of the PI3K pathway, proteins associated with transcriptional regulation, cytoskeletal rearrangement, cellular signalling, migration and metabolism are differentially phosphorylated. After longer stimulations on the other hand, proteins involved in cell cycle progression, phosphatase activity, nucleocytoplasmic transport and RNA metabolism become more prominent in the comparative phosphoproteome. Similar to early response proteins in β WT cells, the phosphoproteome of β E1051K cells were associated with cytoskeletal rearrangement and cellular migration. Aligning with the morphological changes observed, proteins involved in anatomical structure regulation were found. Additionally, some tumour suppressors and oncogene proteins were found among the differentially regulated phosphoproteins. On the other hand, the phosphoproteome of β N553S cells were enriched for functions of RNA metabolism, nuclear import and apoptotic regulation. Proteins involved in structural organization and microtubule assembly were also observed. 37 kDa Akt substrate Poly(rC) Binding Protein 1 (Pcbp1) was found as a common differentially phosphorylated protein in the phosphoproteomes of mutant and WT p110β expressing cells. It is an RNA-binding protein associated with metastasis and EMT, and has a critical role in gene expression by coordinating RNA stability and processing. As our results have shown that p110β isoform has a major role in cytoskeletal reorganization, cell migration and transcriptional regulation, studying the regulation of p110β-Pcbp1 axis can pave the way to understand the mechanisms of increased tumour progression in p110β-dependent cells. Item Open AccessCharacterization of inherited IRAK-4 deficiency in a patient with acute HHV-6 encephalitis(Bilkent University, 2023-07) Demir, Zeynep Güneş Tepe; Belkaya, SerkanHuman herpesvirus-6 (HHV-6), a ubiquitous virus among humans, typically causes acute febrile illness in children, whereas the majority remain asymptomatic. HHV-6 infection can rarely cause encephalitis, with unknown pathogenesis. We hypothesized that inborn single-gene defects may underlie susceptibility to HHV-6 encephalitis in otherwise healthy children. We performed whole-exome sequencing on genomic DNA of a male child diagnosed with acute HHV-6 encephalitis and found a novel homozygous missense variation (NM_016123.4:c.G236A:p.C79Y) in Interleukin-1 receptor-associated kinase 4 (IRAK4), which is involved in the Toll-Interleukin-1 receptor signaling pathway. Sanger sequencing confirmed that both parents and the sibling were heterozygotes. The p.C79Y that affected an evolutionary conserved residue was predicted to be damaging by in silico algorithms. We found that IRAK-4 expression was severely reduced in patient’s leukocytes. There were similar levels of wild-type (WT) and mutant IRAK-4 when transiently over-expressed in HEK293 cells, however mutant IRAK-4 expression was dramatically decreased upon cycloheximide treatment, compared to the WT. This indicated that the p.C79Y might impair IRAK-4 stability. We found that patient’s leukocytes had diminished innate immune responses to various stimuli inducing different Toll-like receptors and cytosolic nucleic acid sensors, compared to the healthy controls. We also generated IRAK4 knockout HEK293 cells by CRISPR-Cas9 genome editing. Transient expression of mutant IRAK-4 had significantly reduced NFκB-dependent luciferase activity, compared to the WT in IRAK4 knockout cells treated with IL-18. Collectively, the p.C79Y impaired both the expression and function of IRAK-4, leading to diminished immune responses against bacterial and viral stimuli in patient’s leukocytes. Overall, this was the first study demonstrating that inborn errors of immunity could underlie isolated acute HHV-6 encephalitis. Our findings also widened the known genotypic and phenotypic spectrum of inherited IRAK-4 deficiency in humans. Item Open AccessProfiling the interactome of STK10 by using proximity-based biotinylation(Bilkent University, 2023-05) Carus, Yağmur Öykü; Çizmecioğlu, OnurBreast cancer accounts for a quarter of all newly diagnosed cancer cases globally in women. It is also the primary cause of death in female cancer patients. In 2020, it was responsible for 15% of all cancer-related deaths in women. The PI3K pathway is the most commonly deregulated pathway in breast cancer. As a result, numerous inhibitors for druggable targets within the pathway have been developed. Unfortunately, drug resistance is frequently observed after treatment with these inhibitors. The mechanisms behind this resistance are widely studied. AKT has been known to be the kinase responsible for transmitting the signal to downstream targets. However, new studies suggested there can be AKT-independent kinases relaying the signal to downstream targets causing PI3K inhibitor drug resistance. Previously, our group hypothesized that they could find an AKT-independent protein that confers resistance to PI3K inhibition. Their bioinformatical analyses identified STK10 (Serine-Threonine Kinase 10) as a possible druggable target for PI3K pathway inhibitor resistance in an AKT-independent manner. They performed wet lab experiments to show STK10’s impact on PI3K inhibitor resistance in breast cancer. They showed STK10 knock-down sensitizing resistant breast cancer cells to PI3K inhibitor. Although their experiments supported the hypothesis of STK10 affecting the PI3K inhibitor resistance of breast cancer cells, they could not identify the molecular mechanisms behind this interaction. In this study, by using APEX2 proximity-based biotinylation followed by mass spectrometry, we profiled STK10’s interactome. We aimed to understand how STK10 contributes to the progression of breast cancer and the development of resistance to treatment. Our bioinformatics analyses indicated possible interaction between STK10 and the candidate proteins we obtained from our mass spectrometry analysis. Immunofluorescence and Co-IP experiments supported some of the putative interactions we discovered. The proteins discovered near STK10 were primarily linked to the reorganization of the cytoskeleton, which is encouraging since STK10 is recognized for its involvement in cell migration through the phosphorylation of ERM proteins. Based on our observations, it is possible to hypothesize that STK10 is influencing the cytoskeletal reorganization in breast cancer cells. However, further experiments are needed to be done to understand the molecular mechanisms behind STK10’s possible functions in breast cancer. These experiments could lead to uncovering the role of STK10 in PI3K inhibitor resistance and help us to identify new therapeutic options to battle breast cancer. Item Open AccessDetection of phosphorylation signatures specific to cancer-related PI3-Kinase isoforms p110α and p110β(Bilkent University, 2023-01) Sulaiman, Mahnoor; Çizmecioğlu, OnurThe PI3K signaling pathway is required for many physiological activities, but it is commonly disrupted during cancer formation. The PI3K p110α and βisoforms, encoded by the PIK3CA and PIK3CB genes, are lipid kinases that phosphorylate PIP2 to PIP3 to activate the PI3K pathway. However, the distinct molecular targets of these isoforms have yet to be discovered, making targeted treatment problematic. According to cancer genomics research, the PIK3CA gene is commonly altered in cancers, but the PIK3CB gene is frequently amplified. The clinical usage of Pan-PI3K inhibitors has resulted in significant side effects, prompting the development of isoform-specific inhibitors. However, it has been shown that these drugs trigger alternate signaling systems downstream, leading in resistance to single-agent treatment. Our research intends to uncover distinctive protein-protein interactions of PI3K isoforms, as well as the consequent different phospho-proteomic signatures, which might be crucial determinants of specific cellular activities. This will be accomplished by using isogenic MEF cells that are only dependent on the p110α or p110β isoforms, isoform-specific pharmacological inhibitors BYL-719 and KIN 193, and a high-resolution mass spectrometry-based method to determine the phosphorylation levels of these protein samples. The predictive biomarkers discovered in this study can be utilized to identify patients who will benefit from PI3K-targeted drugs and to better understand the resistance mechanisms that may arise in response to pathway inhibition. Item Open AccessIdentification and targeting of deregulated metabolic pathways in metastatic prostate cancer cells(Bilkent University, 2023-01) Kaysudu, Irmak; Çizmecioğlu, OnurProstate cancer is the most diagnosed cancer type and the second leading cause of death in men globally. The pathogenesis of prostate cancer mainly relies on the androgen signaling axis. Therefore, androgen deprivation therapy is the primary treatment for prostate cancer. Nevertheless, the disease progression proceeds, followed by castration resistance and androgen independence. Aberrant androgen signaling activity intertwined with the hyperactivated PI3K-Akt signaling pathway has important oncogenic consequences for castration resistance mechanisms. PTEN, a negative regulator of the PI3K/Akt pathway, is one of the most altered tumor suppressor genes in prostate cancer. PTEN loss occurs in the initial stages of prostate cancer and the frequency of its alteration increases in metastatic and castration-resistant prostate cancer. PTEN has both lipid and protein phosphatase activity, with the former antagonizing the PI3K-Akt pathway by converting membrane-associated PIP3 to PIP2. PTEN loss may cause metabolic rewiring in metastatic prostate cancer cells and the associated metabolic vulnerabilities may be tackled for the disease therapy. To understand the impact of PTEN loss in metastatic prostate cancer cells, we created a dox-inducible system in PTEN-null metastatic and castration-naïve LNCaP cells to re-express WT-PTEN and various PTEN functional mutants, and we employed targeted metastatic prostate cancer. Our multidirectional omics studies suggest that the acquisition of resistance to castration depends on the deregulation of the sphingolipid metabolism in metastatic prostate cancer cells. Furthermore, we showed that PTEN re-expression in metastatic and castration-naïve LNCaP cells attenuated sphingosine kinase levels, which might switch the sphingolipid metabolism towards increased sphingomyelin biosynthesis and ceramide phosphorylation. Moreover, we showed decreased PI3K/Akt pathway activity when we inhibited sphingosine kinase with opaganib in LNCaP cells. Our results also showed a significant upregulation in sphingolipid metabolism in castration-resistant C4-2 cells compared to castration-naïve LNCaP. We treated these cells with several sphingolipid metabolism inhibitors and discovered that castration-resistant prostate cancer cells were more sensitive to opaganib or ARN14988, but not to fingolimod, than castration-naïve prostate cancer cells. These findings suggest that sphingolipid metabolism might be a promising target for the treatment of metastatic and castration-resistant prostate cancer. Understanding changes in sphingolipid metabolism may be critical for developing rational combinatorial targeting strategies for prostate cancer in the long run. Item Open AccessIdentification of RNA-based biomarkers associated with manic episodes and lithium response in bipolar disorder(Bilkent University, 2022-12) Niaz, Noor; Güre, Ali OsmayBipolar disorder (BD) is one of the major mood disorders. A person afflicted with this neuropsychiatric disease undergoes episodes of depression and mania. BD ranks as the highest amongst all the mood disorders for having the most negative affect on a person's life. It is a lifelong illness, which requires constant monitoring and medication. From psychotic behaviours to suicide ideation, this disease is a burden on the afflicted and their families. Although, BD has been shown to have a high heritability factor, none of the studies done so far have been able to identify any biomarkers that could be causative for this disease. It is due to the fact that this disease is multifactorial, in which environment is a major contributing factor. Therefore, we believe that a more 'pattern-seeking' approach would lead to novel findings. We performed gene-set enrichment analysis (GSEA) with BD datasets that consisted of 3 different biological phenotypes; euthymic BD, manic BD and lithium treated BD samples. Since a long time, lithium has been used as 'the leading' drug to treat BD because it functions as a mood stabilizer. Therefore, we also included BD cohorts that had undergone lithium treatment in our analysis. As a result of GSEA, we were able to discover 2 novel patterns. In pattern 1, specific genes were found to up-regulated in euthymic BD and lithium treated cohorts, while the same genes were down-regulated in manic BD cohorts. In pattern 2, the opposite trend was observed, that is; another list of specific genes was down-regulated in euthymic BD and lithium treated cohorts, while they were found to be up-regulated in manic BD. The novelty of utilizing GSEA in our analysis was in the fact that we created our own custom gene-sets. The gene-sets were formed after performing differential gene expression (DEG) analysis on all the 3 types of BD datasets. The advantage of using custom gene-sets was that these were genes representing differential expression in different BD phenotypes, therefore, they were biologically relevant to BD. In the publically available curation of gene-sets on various databases, very few gene-sets represent BD. Hence, the custom gene-sets are more relevant and specific to the disease. Then we proceeded to extract core-enriched genes and performed further analysis, such as, plotting fold-change graphs and performing non-parametric tests. In light of these results, we propose putative biomarkers associated with manic episodes and lithium response in bipolar disorder. We put forth the hypothesis that these patterns can diagnose BD accurately, indicate if a patient is responding to lithium or not, and predict an oncoming manic episode. Therefore, we propose the genes adhering to the discovery patterns as putative biomarkers. Simultaneously, we performed biological analyses and literature review with the list of putative biomarkers. We found out inflammation as a potential underlying cause of pathogenicity in BD. We hypothesize that the process of inflammation is disrupted in BD patients, especially between the different mood states and that in order to treat this disease, inflammation as a pathway should also be targeted. We highlight TNFa as one of the main cytokines that 3 of our biomarkers; ADAMTS9, IL-1 B and STCJ are associated through various pathways, and this disruption of the inflammatory pathway may occur due to alteration in TNFa and biomarkers' levels. We hope that after further, subsequent research, the inclusion of our biomarkers in the clinics will help with the issue of BD misdiagnosis, help save precious treatment time, help with the choice of medication, and help indicate a patient's mood state. Overall, it will help the clinicians to tailor the treatment to every patient's specific profile, making the it easier to design a more personalized treatment strategy. Item Open AccessEffects of depletion of CHRNA5 and/or TP53, and transient and stable overexpression of wildtype or mutant TP53 on expression of DLK1-MEG3 locus in MCF7 cells(Bilkent University, 2022-09) Arıcı, Burçin İrem; Karakayalı, Özlen KonuThe expression of CHRNA5 has a prominent role in lung cancer and nicotine addiction. Besides, depletion of CHRNA5 has been identified as being tumour suppressive in breast cancer. Moreover, CHRNA5 depletion causes increases in CDKN1A expression, downregulates the 14q32.31 miRNAs and decreases DLK1 expression. This thesis examined the effects of CHRNA5 and/or TP53 downregulation, as well as TP53 overexpression, on the expression of the DLK1-DIO3 region. My findings demonstrated that CHRNA5 depletion decreased the expression of the protein-coding DLK1 and the non-protein coding MEG3 genes in this locus in the presence or absence of TP53. Since these two genes have a vital role in tumour suppression and tumorigenesis, this provided more evidence for the importance of CHRNA5 in the modulation of expression in this locus in breast cancer. Since there is a relation between CHRNA5 and TP53 induction, the expression of one of the main TP53 regulators, MDM2, was also studied. Accordingly, it was found that combining CHRNA5 depletion with TP53 depletion caused a significant decrease in expression in both the MDM2 and MDM2 sequestering elements, PDLIM7 and CDH18. Additionally, the effects of both wild-type and mutant TP53 expression levels on DLK1-MEG3 locus and CHRNA5 were investigated in stable MCF7 breast cancer cells that we have generated. I have found that even if CHRNA5 depletion induced p53 expression, TP53 overexpression did not have CHRNA5-inducing effects regardless of the functionality of TP53. However, wild-type TP53 overexpressing MCF7 cells behaved differently than mutant TP53 overexpressing cells in their DLK1 expression levels. Future research should clarify the effects of CHRNA5 depletion on DLK1-MEG3 region for a given TP53 mutation status. Item Open AccessDevelopment of a shiny application for comparative transcriptomics and differential gene expression analysis(Bilkent University, 2022-09) Leka, Ronaldo; Karakayalı, Özlen KonuRNA sequencing has proven to be an effective technique for divulging information about the transcriptome in molecular biology research. Compared to microarrays and early methods of cDNA sequencing, high-throughput RNA sequencing has better resolution, low background noise, a higher range to quantify gene expression, and relatively lower cost. The development of sequencing technique has led to the development of tools for analyzing the high volume of data that is generated. Statistical methods for normalizing, filtering, performing exploratory and differential analysis, and other functional analyses based on RNA sequencing count data have made RNA sequencing one of the most popular techniques in genomics. To help facilitate the use of such statistical tools, web applications developed in R using the shiny package offer an advantageous environment where researchers can use a graphical interface to give inputs and instructions to the underlying server-side libraries that analyze and generate results in tables and plots. This thesis presents a new tool that has been developed for exploratory analysis, data normalization and filtering, differential gene expression analysis (DGEA), correlation analysis, principal component analysis, and functional analysis such as over-representation analysis and gene set enrichment analysis. When compared to other available applications, this new application offers the ability to run multiple methods for DGEA and compare results between them, leading to the output of gene sets that are discovered as DEGs in multiple tests. Here I present the features of this application in detail where I aim to improve upon the applications that are available in the literature. An example dataset from our lab was also investigated by this RNA-seq tool leading to a better understanding of Mineralocorticoid Receptor (MR) signaling in breast cancer. Item Open AccessA shiny application for pancan survival analysis with paralog/miRNA pairs and in vitro validation of miRNA synergism in liver cancer(Bilkent University, 2022-09) Tombaz, Melike; Karakayalı, Özlen KonuEmerging cancer survival tools can predict risk of disease and identify prognostic biomarkers. Multivariable Cox proportional hazards models with mRNA and microRNAs (miRNAs) expression can differentiate survival outcomes. Previous studies showed that genes that belong to the same pathways/families may act independently, and in a cancer-specific manner. In this thesis, cancer-dependent hazard ratios of paralog genes and sense-antisense strands of miRNAs were tested for TCGA PANCAN. The results were presented in a R/Shiny web application that provides gene-by-survival networks. The gene-by-survival network approach also was applied to the plasma membrane-endoplasmic reticulum (PM-ER) calcium channel geneset. Among paralogs, cancer-specific prognostic signatures and functional compartmentalization were observed. Some cancers like UVM, MESO emerged as hub cancers for PM-ER signalling. Further the proposed gene-by-survival network approach has been extended for miRNA-mRNA triplets that may act in synergy in hepatocellular carcinoma (HCC). Next, the effects of synergistic miRNA pairs provided by miRCoop algorithm were tested on cell viability and target gene expression for selected triplets. The results have revealed that the combinatorial miRNA treatments show promising results as RNAi therapeutics yet future studies with different doses and triplets are needed. Item Open AccessAnalysis of the effect of Erbin on the epithelial to mesenchymal transition related genes in breast cancer(Bilkent University, 2022-09) Şahin, Filiz; Yuluğ, IşıkBreast cancer is a heterogeneous disease and has complex mechanisms, which brings the need to come up with different approaches for its various types. Erbin is a member of the LAP family and directly interacts with ErbB2 (HER2), which is a crucial component for breast cancer sub-typing and treatment. Research on the relationship between Erbin and breast cancer showed different results since Erbin was seen to be able to act as both a tumor suppressor and a tumor promoter. There are multiple findings that Erbin has a role in the epithelial to mesenchymal transition in different cancer types. The role of Erbin in EMT is said to be multi-faceted, it affects multiple pathways that act on EMT progress. However, like Erbin’s contradictory role in breast cancer, there are different findings that support the conflicting claims on whether Erbin is an EMT promoter or not. These findings in the literature lead us to analyze how Erbin affects EMT in breast cancer. To investigate if there is a role of Erbin in EMT regulation, we have overexpressed Erbin in MCF7 and silenced Erbin in MDA-MB-231 with MDA-66 cell lines. With this experimental setup, we aimed to see how Erbin plays a role in EMT in both epithelial and mesenchymal cells as well as Luminal A, TNBC, and ER positive breast cancer types. Our results showed that in TGFB induced EMT, Erbin levels are decreased in epithelial cells. Erbin overexpressed MCF7 cells show significant changes in the EMT markers compared to the control group. Experiments conducted with Erbin silenced MDA-MB-231 and MDA-66 cells also demonstrated significant changes in EMT marker levels. Our results suggest that Erbin has a role in EMT in breast cancer, and this role is not limited to one pathway and not streamlined, but possibly a big orchestration of different effects. Item Open AccessA-To-I RNA editing events, potential biomarkers for prognosis and chemosensitivity in gastric cancer(Bilkent University, 2022-09) Çela, Isli; Güre, Ali OsmayGastric cancer (GC) is one of the leading causes of cancer mortality, and it frequently presents in advanced stages with a poor prognosis and response to treatment. Although extensive research has identified many potential biomarkers in GC, the heterogeneity of the disease is an impediment to validation, so only a small number find limited application in clinics. RNA editing is an epigenetic modification that results in nucleotide changes in the RNA sequence. Adenosine to Inosine (A-to-I) substitutions are the most common editing events in humans, and they are mediated by Adenosine deaminases acting on RNA (ADAR) enzymes. Inosine (I) mimics Guanosine (G) and creates pairs with Cytidine (C), resulting in changes in RNA structure and stability, amino acid substitutions, alternative splicing, or gene expression regulation via miRNA target site modifications.RNA editing dysregulations have been found in breast, lung, kidney, brain, and gastric cancers, but the utility of specific editing events as biomarkers is largely unexplored. In this study we investigate the potential of A-to-I editing events as chemosensitivity and prognostic biomarkers in GC. Across multiple datasets, our analysis shows that RNA editing events at 305 unique positions correlate with drug sensitivity measures of 17 approved chemotherapeutics in GC cell lines.The most significant editing event-drug sensitivity correlations indicate that higher editing levels are associated with higher chemosensitivity. Interestingly, the expression levels of genes with identified editing events have a weaker or no correlation with drug sensitivity, implying that editing events are biomarkers independent of transcript levels. We show that, while ADAR enzymes mediate editing events, ADAR expression levels are not interchangeable with editing frequencies as chemosensitivity biomarkers in GC. We discovered a non-synonymous editing event in the C11orf80 coding sequence, resulting in an amino acid substitution (S.p133G). Also, we identified an editing event in the 3'UTR of SOGA1 that correlates with increased SOGA1 expression. The presence of this editing site in a putative target site of miR-9-5p suggests that gene expression might be regulated by miRNA target site modifications. In the TCGA and Singapore cohorts, the prognostic role of editing events in GC was investigated. Overall, higher levels of editing are associated with better survival in GC patients. In both cohorts, we found an editing event in the CLPX gene at the position 65442098 to be an independent good prognostic factor. We chose editing events that would best categorize our patients into "High" and "Low" edited groups using the Log Rank Multiple Cut-off (LRMC) plot distribution. In each dataset, we propose two editing events, one good and one bad prognostic factor that independently correlate with survival in GC patients. In the Singapore Cohort, high editing levels in ZNF587 are associated with a good prognosis, while those in DCAF16 are associated with a poor prognosis.High editing levels in CTSB correlate with better overall survival (OS) in the TCGA cohort, while those in NUP43 correlate with worse OS. Because transcript levels do not correlate with survival, the prognostic effects of these editing events are unaffected by gene expression levels.We believe that editing levels at specific positions can be used as prognostic biomarkers in a significant way, providing a more cost-effective and applicable alternative to prognostic editing signature models. Our findings suggest that editing events could be used as independent biomarkers for chemosensitivity and prognosis in gastric cancer, however more investigation is required to elucidate the mechanisms underlying the observed relationships. Item Open AccessTranscript variants of CELF2 gene as unique prognostic indicators in breast cancer(Bilkent University, 2022-08) Azizolli, Shila; Güre, Ali OsmayBreast cancer (BC) is the most common malignant tumor in women around the world. Aside from finding a cure for this disease, it is also critical to identify prognostic biomarkers that can help clinicians intervene with the appropriate treatment and prevent BC progression. Current biomarker identification methods rely primarily on multi-gene prognostic signature models. However, due to the tumors' high heterogeneity, the accuracy of these multi-gene signatures is questionable. As a result, our main objective was to conduct a comprehensive analysis to identify a reliable prognostic biomarker in BC. Previously, a group of eight carnitine metabolites and SAH were linked to a poor prognosis in BC (Dr. Waqas Akbar, Unpublished Data). We discovered the genes associated with these metabolites using correlation analysis, and then we identified CELF2 as a good prognostic biomarker in BC. We validated CELF2's prognostic role in RNA-Seq and Microarray datasets in-silico. We demonstrate that the CELF2 1554569_a_at probeset is more consistent in its association with a favorable prognosis direction than the 202157_s_at probeset. When compared to the other probeset, CELF2 – 202157_s_at is expressed at higher levels and in a broader range of tissues. We were unable to find a clear significant association between CELF2 expression and prognosis during the in-vitro immunohistochemistry validation experiments. We hypothesized that this could be because our polyclonal anti-CELF2 antibody also recognized the less consistent 202157_s_at CELF2 probeset. We discovered that there are probeset-specific CELF2 transcript variants that are associated with different prognosis while testing this hypothesis. We created a Risk Score model by combining the expression levels of good and less-favorable CELF2 prognostic transcripts to improve prognosis prediction accuracy. The model successfully stratified the patients and predicted a higher overall survival in the Low-Risk group versus the High-Risk group. Overall, our findings suggest that each unique transcript variant of a gene can be associated with different prognosis directions. Therefore, we propose that studying prognostic associations at a gene transcript level could be a rich resource for the development of more robust biomarkers and therapeutics in cancer in the future. Item Open AccessInvestigation of Pre-clinical and Phase II clinical studies of VLP-58-1023- AL-K3-PII vaccine for Alpha variant(Bilkent University, 2022-08) Saraydar, Berfu; Gürsel, İhsanIn the late December of 2019, SARS-CoV-2, a new coronavirus, was discovered in Wuhan, China and described as the causative agent of Coronavirus Disease 2019 (COVID-19). The disease has spread rapidly across the world due to its high transmissibility and has been declared a pandemic by the World Health Organization (WHO). The development of an effective vaccine has become the most significant issue to constrain the pandemic. Several COVID-19 vaccines have been authorized for human use and others are in clinical trials. Although SARS-CoV-2 encodes four structural proteins, which are Spike (S), Nucleocapsid (N), Membrane (M) and Envelope (E), most of the current vaccines used only Spike as antigen in order to generate antibodies for preventing the virus entry and replication. However, concerns have raised about Spike-based vaccines with the emerging of variants as they can moderately escape from neutralizing antibodies. For these purposes, we developed Virus-like particle (VLP) vaccine which displays hexaproline prefusion-stabilized spike (S-6p), N, M, E proteins, and adjuvanted with Alum and K3-CpG ODN. Rather than using wild type, we preferred to use the sequence of Alpha variant because of its high mortality risk and selection advantages. At the beginning of the study, we designed three different vaccine formulations and based on the results of humoral immune response in mice we determined the optimal formulation and dosage for human use. Our pre-clinical studies revealed that the best vaccine combination was high dose antigen and low dose adjuvants. Next, we wondered whether a 3rd dose has an impact on long-lasting immunity or enhancing immunogenicity in mice so that its applicability to humans could be determined. It was found that 3rd dose injection increased the antibody levels much higher than 2nd dose administration and prevented humoral immunity from decreasing after a certain amount of time. Further, both humoral and cellular immunity were studied with serum and PBMC samples from 117 volunteers who participated in the Phase II clinical trial. All IgG ELISA experiments indicated that VLP-58-1023-AL-K3-PII vaccine induced great amount of humoral immune responses against S,N proteins and WT, Alpha, Delta RBDs. In terms of T cell responses, it is known that Alum-induced robust Th2 response can be redirected to the Th1 axis with the use of CpG ODN. So, we investigated whether Th1 or Th2 type of cell response was dominant after vaccination. All cytokine levels specific to SARS-CoV-2 peptides demonstrated that the vaccine elicited Th1-biased responses. Taken together, this study revealed that VLP-58-1023-AL-K3-PII vaccine for Alpha variant successfully elicited both humoral and cellular immune responses, its effectiveness against other variants was indicated and the efficiency of vaccine could be increased with the administration of 3rd dose, in terms of ensuring long-lasting immunity. Item Open AccessInvestigation of cellular and humoral immune responses induced by SARS-CoV-2 VLP vaccine in Pre-clinical and clinical studies(Bilkent University, 2022-08) Bildik, Kadriye Tuğçe; Gürsel, İhsanCOVID-19 pandemic, caused by SARS-CoV-2, emerged in China in late 2019, and as of March 2020 has become a serious concern for the whole world due to severe progression of the disease especially in the elderly and high transmission rate of the virus. The ravages of pandemic on health sector, economy and social life accelerated the vaccine race as vaccination is the most critical measure to hamper the pace of the pandemic and the most important step for acquiring herd immunity. Various vaccine studies are registered by WHO most of which target Spike protein or a certain region of Spike protein, such as receptor binding domain (RBD) for inducing immune reactions. Alternatively, the use of virus-like particle (VLP) technology for vaccine development broadens the magnitude of immune responses as the four main structural proteins of the virus, Spike, Nucleocapsid, Membrane and Envelope, are incorporated. Therefore, we developed a VLP vaccine against SARS-CoV-2 and adjuvanted it using Alum and K3 CpG ODN. Herein, we investigated the humoral and cellular immune responses induced by SARS-CoV-2 VLP vaccine in pre-clinical and clinical studies. The results from pre-clinical experiments conducted on vaccinated BALB/c mice indicated that VLP vaccine triggered effective antibody response against Spike, Nucleocapsid, Wild Type RBD, Alpha RBD and Delta RBD proteins as shown by ELISA. Additionally, it was proven that the diminishing antibody responses can be boosted and a prolonged immune response was achieved by a third dose of VLP vaccine. For the clinical studies, the humoral and cellular immune responses of Phase II clinical trial volunteers were analyzed by employing ELISA and CBA methods respectively. The results obtained from ELISA demonstrated that the Phase II volunteers developed effective antibody titers against Spike, Nucleocapsid, Wild Type RBD, Alpha RBD and Delta RBD proteins. Analysis of the cellular immune responses by CBA showed that VLP vaccine induced a Th1-skewed immune reaction with coexistence of Th2-, Th17- and Treg related responses owing to its formulation with K3 CpG ODN and Alum. Taken together, our results indicated that SARS-CoV-2 VLP accine triggered effective cellular and humoral immune reactions against Spike and Nucleocapsid along with RBD variants which proved the cross-protective response achieved. Elicitation of multi-functional immune response and adoptability of the VLP technology to newly emerging variants makes VLP vaccine a promising candidate for the future booster injections. Item Open AccessInvestigation of Pre-clinical and clinical results against SARS-COV-2 wild-type and Alpha variants combination of VLP-58-1023-AL-K3 vaccine(Bilkent University, 2022-08) Bartan, Aslı Gülce; Gürsel, İhsanThe SARS-CoV-2 virus, which first appeared in Wuhan, China and had a pandemic effect at the end of 2019, has urgently created need of suitable vaccine candidates’ development with its quick spread. Scientists in different laboratories from all over the world, including Turkey, have started the studies of vaccine development with different methods based on the basic proteins of the SARS-CoV-2 virus. Most of these vaccines are based on the Spike protein, which allows the virus to enter the host and survive. Although the Spike protein, which has high antigenic properties and contains the RBD region to bind ACE2 receptor in the host, is effective in the formation of neutralizing antibodies in the person, the virus essentially has 4 structural proteins, not only spike. Our vaccine strategy containing virus-like particles (VLP) was based on the self-assembly property of Spike, Membrane, Envelope and Nucleocapsid proteins, and these protein structures were transformed into plasmids with his-tag labels, and the products were collected in mammalian cells in vitro by the transfection method. VLPs were purified by affinity chromatography and formulated with 2 different adjuvants which are Alum and CpG. Since the virus, which has changed rapidly in the last 2 years, is open to new mutations over spreading, it is very important that our vaccine should be editable and adaptable to new virus variants. Due to the alpha variant effect in the world, which was first seen in the UK and has become widespread all over the world recent years, WT VLP formed with a more stable Spike protein containing a thermostable sub-proline (6p; HexaPro) mutation was used in the first dose of our vaccine, but UK VLP containing alpha variant mutations was used in the second dose. Alum and CpG adjuvants were used together for both injection formulations. Our vaccine has successfully completed Phase 2 studies. As a result of these studies, approximately 115 vaccinated volunteers were followed up for 90 days and immunological analyzes were performed with samples taken at certain day intervals (Day 21, Day 35, Day49 and Day90). As a result of these analyzes, moderate and high levels of neutralizing and non-neutralizing antibody responses were observed in many patients, and a humoral immune response was induced until day 90. In addition, it was observed that the cellular responses of the volunteers are progressed in tendency with the Th1 cell response. Item Open AccessBRCA:Cohort-vs-TCGA: a webtool for comprehensive exploration and comparison of mutational landscape of breast cancer cohorts with TCGA-BRCA(Bilkent University, 2022-08) Ahadli, Farid; Karakayalı, Özlen KonuComparisons of mutational landscapes between independent breast cancer (BRCA) cohorts in a comprehensive and statistical manner using online tools are likely to help advance our understanding of the diversity of somatic interactions and mutational signatures. In this thesis I have developed a webtool called BRCA:Cohort-vs-TCGA, which makes the mutational landscape comparison between a breast cancer cohort and TCGA-BRCA cohort accessible to the general public. BRCA:Cohort-vs-TCGA app contains modules to conduct mutational signature identification and comparison, differentially mutated gene and pathway identification and comparison, driver gene identification, and somatic interaction identification and comparison. Among the applications generated for similar purposes in literature, BRCA:Cohort-vs-TCGA provides advantages because of its comparative, statistical, and signature related features. When BRCA:Cohort-vs-TCGA was applied to compare the Gustave Roussy Institute (IGR) metastatic breast cancer cohort with the TCGA-BRCA cohort, a unique to IGR mutational signature related to SBS17b, a reactive oxygen species associated signature, which was missed by the original analysis of the authors, was identified. Interestingly, contribution of SBS17b was significantly higher in TNBC samples when compared to HER+ and HR+ samples. Differentially mutated gene analysis identified four genes, namely, ESR1, FSIP2, ARMCX4 and MALRD4, of which the latter two were unique to re-analysis by BRCA:Cohort-vs-TCGA. Differentially mutated pathway analysis on the other hand pointed to the differential mutation of DNA damage related pathways, NRF2 pathway and WNT pathway. The influence of the hypermutator samples in Fisher's exact test based analyses is well documented, as a result of many spurious mutations these samples contain. Re-analysis of the IGR cohort after removal of the hypermutator samples showed ARMCX4 was no longer differentially mutated anymore. Re-analysis of the differentially mutated pathways resulted in identification of three pathways, namely Mismatch Repair - Core, LIN37 independent P53 Targets, and U5, all of which had an odds ratio of less than one. These results, however, could be biased by the presence of hypermutators in the TCGA-BRCA cohort. In the second case study, I have demonstrated how the extracted somatic interactions for a user selected gene, i.e., CHRNA5 and TP53, can be used to prioritize and filter genes. This was done by merging the somatic interactions with the differential expression profiles of MCF7 cells treated with siRNA against CHRNA5 and TP53. Accordingly, BRCA:Cohort-vs-TCGA can help annotate/enrich other high throughput data with somatic interactions of custom genes.