LMP student seminars: 12 November

Each week during term time, MSc and PhD candidates in the Department of Laboratory Medicine and Pathobiology present their research.

Anyone is welcome. No need to register.

Location: Medical Sciences Building, rooms 4171 or 4279, see below.

As part of the core research curriculum, students taking LMP1001/2/3: Graduate Seminars in Laboratory Medicine and Pathobiology will present their projects. Please see abstracts below.

Group 2: Cancer, Development and Aging

Location: MSB 4171

Abraam Zakhary

• Title: Eosinophilic pattern of Chromophobe Renal Cell Carcinoma is More Aggressive Than the Classic pattern: A Comprehensive Morphologic and Genomic Analysis
• Supervisor: Dr. Rola Saleeb

Group 5: Infectious Diseases, Inflammation and Immunology

Location: MSB 4279

Noor Wasfi Alsmadi

• Title: Genome-Scale Metabolic Modelling to Study and Manipulate the Metabolism of Defined Gut Bacterial Communities
• Supervisor: Dr. Dana Philpott

Abstracts

Abraam Zakhary: Eosinophilic pattern of Chromophobe Renal Cell Carcinoma is More Aggressive Than the Classic pattern: A Comprehensive Morphologic and Genomic Analysis

Chromophobe renal cell carcinoma (chRCC) is generally considered an indolent malignancy; however, studies have produced conflicting results regarding the differences among its histologic patterns. The prognostic significance of eosinophilic morphology in chRCC remains controversial, in part due to inconsistent diagnostic criteria and inclusion of morphologic mimickers. In this study, we analyzed 165 chRCC cases from three institutional cohorts. Cases were classified by three expert genitourinary pathologists into eosinophilic, classic, and mixed chRCC pattern based on strict morphologic and immunohistochemical criteria. We assessed their clinicopathologic features, copy number variations (CNVs), targeted sequencing of TP53, PTEN, and RB1, and we performed pathway analyses using RNAseq and multiplex gene expression profiling. Clinicopathological and survival analysis were assessed using GraphPad Prism software. The final cohort included 24 eosinophilic chRCC, 47 classic chRCC, and 39 mixed chRCC. CNV frequencies were comparable across patterns. The eosinophilic and mixed patterns showed higher proportion of RB1 aberrations. Pathway analysis revealed significant enrichment of oncogenic signaling - including TP53, PTEN, RB1, DNA damage/repair, mTOR, and cell cycle pathways - in eosinophilic chRCC and mixed chRCC, with the strongest enrichment in eosinophilic chRCC. Clinically, eosinophilic chRCC and mixed chRCC were associated with larger tumour size, more advanced clinical stage, and significantly worse disease-free survival (DFS) and overall survival (OS) compared to classic chRCC. On multivariate analysis, histological pattern and stage remained independent predictors of DFS. As outlined in the results, eosinophilic chRCC exhibited a distinct molecular phenotype characterized by enrichment of oncogenic pathways and worse clinical outcomes compared with classic chRCC. Mixed chRCC demonstrated an intermediate but similarly aggressive profile. These findings indicate that eosinophilic morphology has potentially more aggressive biology than classic chRCC, underscoring the need for introduction of specific diagnostic criteria for eosinophilic chRCC and its differentiation in clinical practice.

Noor Alsmadi: Genome-Scale Metabolic Modelling to Study and Manipulate the Metabolism of Defined Gut Bacterial Communities

The production of metabolites - like butyrate - by the gut microbiota depends on complex cross-feeding interactions among bacterial species, making it difficult to design strategies that predict and enhance its levels for applications to increase its abundance. To address this challenge, my project proposes to develop a computational genome-scale metabolic modelling tool – called MetaGutPred – to study and manipulate the metabolism of gut microbial communities as a way to optimize the production of certain bacterial metabolites. Genome-scale metabolic models (GEMs) are mathematical representations of the biochemical reaction networks of an organism that enable the study of metabolic capabilities and cross-feeding interactions of microbial communities. Through careful integration of experimental-based constraints, I hypothesize that GEM-based community modelling can be leveraged to improve in silico predictions of gut bacterial community metabolic potential. To validate the model generated, I am using an 8-member gut bacterial community in mice called Human Defined Community- (HDC-)1. MetaGutPred takes a number of features of experimental data as input, including individual bacterial GEMs, abundance of members in the community, and used metabolite concentrations and then outputs the predicted concentration of metabolites produced by the community. To generate individual GEMs for each of the strains in the community, I first conducted whole genome sequencing and assembled the bacterial genomes. Then, functional genome annotation and the individual GEMs were created using gapseq. The GEMs created were then integrated into a community model using MetaGutPred for validation against targeted metabolomics data. Information on the abundance of the bacteria was obtained using qPCR on fecal samples from HDC1-colonized mice. Targeted metabolomics will be run on age- and diet-matched fecal samples from HDC1-colonized and GF mice. This will be used to quantify the abundance of metabolites in the fecal samples to set up uptake constraints & further validate the model against concentration of produced metabolites like butyrate. In the future, we hope to use this modelling framework as a device to test the effectiveness of gut microbiome-based therapeutics – including probiotics and prebiotics – in the context of diverse microbiotas in humans.

Contact

No need to register.

Contact lmp.grad@utoronto.ca with any questions.