Colin Ross

Hypersensitivity reactions to l-asparaginase pose a major limitation in pediatric acute lymphoblastic leukemia (ALL) therapy, often necessitating treatment modifications that jeopardize remission and survival. This dissertation investigates the genetic underpinnings of these reactions and explores strategies for identifying high-risk patients.A systematic review of studies examining the genetic associations of l-asparaginase-induced hypersensitivity in human case-control studies confirmed a consistent, albeit modest, effect of HLA-DRB1*07:01 (p = 3.51 x 10⁻¹⁸, OR = 1.96 , I² = 18.42%) on hypersensitivity risk. By contrast, genome-wide association studies produced heterogeneous findings. These observations emphasize the importance of robust phenotyping and ancestry-aware study designs.In a genome-wide associations study of l-asparaginase-induced hypersensitivity in a large discovery cohort (n = 926) of pediatric cancer patients, large effect variants (OR = 3.9-8.5, p 10⁻⁵) were identified within regulatory regions of genes (CYP1B1, OPLAH, SORCS2, SEC16B) involved in amino acid stress response, endoplasmic reticulum stress, and l-asparaginase clearance. These findings were then validated in an independent replication cohort (n = 180). When top-risk variants were incorporated into a genetic risk score, carriers of two or more risk alleles displayed a 25-fold increased risk of experiencing hypersensitivity (OR = 25.2 , p = 1.0 x 10⁻¹⁰), underscoring the power of polygenic models to enhance individual risk stratification.Predictive modeling revealed that incorporating multiple polygenic risk scores from diverse studies, along with clinical variables and drug usage data, improved risk prediction by 36% over single-variant models. The most comprehensive machine learning frameworks (neural networks, support vector machines, logistic regression) achieved area under the curve (AUC) scores > 0.70 in an independent replication cohort, with significant differences for the prediction of cases of hypersensitivity compared to controls (OR = 3.4 - 4.5, p  3.86 x 10⁻⁴), demonstrating discrimination between patients who developed hypersensitivity and those who did not.These findings highlight the complex genetic architecture of l-asparaginase hypersensitivity and the potential of comprehensive risk scores in refining patient stratification. By integrating genetic, clinical, and pharmacological data, this work provides a framework for pharmacogenetic testing to preserve treatment efficacy while minimizing adverse events, advancing precision medicine in pediatric ALL care.

View record

Vincristine is a cornerstone chemotherapeutic agent widely used for the treatment of various pediatric cancers. However, its clinical utility is often compromised by the occurrence of vincristine-induced peripheral neuropathy, which significantly impacts quality of life due to pain and disability, and may require dose reduction, treatment delays or discontinuation that increases the risk of relapse. While several studies have identified genetic factors that are associated with vincristine peripheral neuropathy risk, these variants do not fully explain the variability in the risk of this reaction, and the clinical relevance of many variants remains unclear. As result, there are currently no established predictors of peripheral neuropathy risk prior to the start of vincristine treatment. This dissertation aims therefore to identify, and validate, novel genomic markers of vincristine-induced neuropathy in pediatric cancer patients.A systematic review of existing genetic associations with vincristine-induced peripheral neuropathy was performed and demonstrated that variations in genes implicated in vincristine pharmacokinetics, pharmacodynamics, and neuropathy-related pathways are associated with vincristine-induced peripheral neuropathy susceptibility. Among these, the CEP72 rs924607 gene variant emerged as the most robustly supported variant influencing neuropathy risk. I subsequently conducted a genome-wide association study uncovering 13 novel genomic variants significantly associated with vincristine neuropathy risk or protection. Of these, I successfully validated two variants an independent replication cohort. Notably, patients carrying protective variants in both NRG3 and ACTN1 genes exhibited a substantial reduction in neuropathy incidence over time (HR=0.36, 95% CI: 0.124–0.602; p0.001). A follow-up pathway analysis revealed involvement of four key pathways, including nerve development, nerve myelination, neuronal transmission, and cytoskeleton/microfibril functions-related pathways. Finally, a targeted analysis of the CEP72 rs924607 variant was conducted in a subset group of patients revealing its association with significant increase in neuropathy risk following prolonged vincristine therapy (OR=3.457, 95% CI: 1.143-10.45; p=0.028).These findings present potential actionable genomic markers of vincristine-induced neuropathy and offer opportunities for tailored interventions to improve vincristine safety in children with cancer. Overall, the work discussed in this dissertation contributes to expanding the existing vincristine pharmacogenomic literature and provides needed insights into the underlying pathology of vincristine-induced peripheral neuropathy reaction in pediatric oncology patients.

View record

Cystic fibrosis (CF) is among the most widespread monogenic diseases globally and stands as the most common lethal genetic disorder among Canadian youths and young adults. It arises from single mutations in the cystic fibrosis transmembrane conductance regulator (CFTR) gene. CRISPR-based gene editing offers a groundbreaking opportunity for correcting monogenetic diseases like CF. Nevertheless, the effective delivery of CRISPR-tools through the lung epithelium and its protective mucosal lining poses a significant challenge. Lipid nanoparticles (LNPs) emerged as versatile non-viral gene delivery systems that could potentially overcome this challenge. However, substantial knowledge gaps persist, especially concerning diseases like CF.This thesis delved into the fundamental understanding of interactions between Cas9 mRNA or ribonucleoprotein (RNP)-loaded LNPs and mucus. Notably, LNP-mRNA demonstrated higher mucus diffusivity than LNP-RNP in healthy mucus, likely due to smaller particle sizes. Mucin sialylation significantly impeded LNP diffusivity, along with high mucin concentrations. Conversely, high ionic strength (>100 mM) and moderate acidic conditions enhanced LNP diffusivity. Importantly, increasing LNP’s PEGylation, particularly when employing a mixture of PEG species rather than a single type, significantly improved LNP diffusivity in CF-mucus, while ensuring robust cell transfection in primary normal human bronchial epithelial (NHBE) cells derived from CF patients.Subsequently, the thesis concentrated on devising strategies to enhance gene editing efficacy of Cas9/sgRNA loaded LNPs in both healthy and CF primary NHBE cells, which are notoriously challenging to genetically manipulate. Adjusting sgRNA to Cas9 ratio and incorporating endosomal escape enhancers (saponin) proved fruitful, elevating editing rates to ~15%. Importantly, through screening LNPs, LNP-H (pKa 7.1) demonstrated robust editing, achieving ~30% editing in NHBE cells. Furthermore, when LNP-H was used with highly modified sgRNA, ~50% editing in CF-NHBE cells was achieved. In physiologically relevant 3D NHBE models, LNP-H exhibited high uptake, as confirmed by fluorescence microscopy, yet achieved 7% editing in healthy models and 5% in CF models. The lower editing efficiencies in 3D models were expected because of the ciliated epithelium/mucosal barrier. Various reports indicate that normal lung function requires only 5-10% of normal CFTR, suggesting that the study's LNP-mediated strategies could be effective for CRISPR-based editing as a therapy for CF.

View record

Anthracyclines (e.g., doxorubicin) are effective chemotherapeutics used to treat a broad-spectrum of childhood and adult cancers. However, the clinical utility of anthracyclines is considerably limited by anthracycline-induced cardiotoxicity (ACT). ACT remains a significant burden in the clinic, where >50% of patients receiving anthracyclines experience some degree of ACT, initially presenting as left ventricular dysfunction, and up to 16% of patients develop heart failure. Advancements in patient genetics show substantial promise as a predictive measure of ACT. The gene variants RARG-rs2229774, UGT1A6-rs17863783, and SLC28A3-rs7853758 have the strongest evidence for association with ACT in anthracycline-treated patients. First, this thesis aimed to understand the scope of genetics, focusing on the proposed roles of genes RARG, UGT1A6, and SLC28A3 in the development and prevention of ACT. Second, I investigated the roles of UGT1A6 and RARG in the development of ACT using cellular models. The research replicated the ACT susceptibility of the UGT1A6 locus using a cellular model (HEK293) of doxorubicin-induced cytotoxicity. Further, I identified that ligand-mediated activation of RARG, using all-trans retinoic acid (ATRA) and CD1530, significantly increases the cell viability of H9c2 cardiomyoblasts. Lastly, I evaluated the efficacy of ATRA in preventing the development of cardiotoxicity using a human cardiomyocyte and a mouse model of ACT. In human cardiomyocytes, ATRA enhanced cell survival during doxorubicin exposure. The protective effect of ATRA was also observed in a mouse model (B6C3F1/J) of ACT, in which ATRA treatment improved heart function of doxorubicin-treated mice. Histological analyses also indicated that ATRA treatment reduced the pathology associated with ACT. Further, in human cardiomyocytes, ATRA induces the gene expression of retinoic acid receptors (RARG, RARB), while repressing topoisomerase-II enzyme genes (TOP2A, TOP2B), which encode for the molecular targets of anthracyclines, and repressed downstream ACT response genes. This mechanism partially explains the protective effect of ATRA and why patients presenting with dysregulation of this pathway may be susceptible to ACT. Overall, the work performed in this thesis shows that gene variants that modify ACT risk belong to critical biological pathways of ACT. Further, these molecular mechanisms also provide a unique opportunity to develop safe and effective interventions to mitigate ACT.

View record

Chronic primary systemic vasculitis (PSV) describes a diverse group of debilitating and potentially life-threatening diseases, characterized by inflammation of blood vessels within various organs such as the kidneys, lungs, brain, eyes, and skin. Subtypes of the small- to medium-sized vessel vasculitides are particularly challenging to classify due to many overlapping clinical symptoms. Their differentiation is important, however, as there is evidence that different subtypes benefit from different treatment approaches. The rarity of vasculitis in children has limited pediatric-specific PSV studies and the clinical approach to pediatric PSV is adapted primarily from adult studies. Not surprisingly, adult-derived classification criteria are imperfect for children and consequently, fail to classify up to two thirds of children with small- to medium-vessel PSV. The objective of this dissertation is to identify genetic markers and select, circulating biomarkers to better our understanding and clinical approach to managing the disease and to improve outcomes of children with chronic PSV. Using biological samples and clinical metadata from a worldwide cohort of children with chronic PSV, this dissertation reports (1) genetic associations specific to pediatric autoimmune vasculitis through employing a genome-wide association study; (2) a high prevalence of autoantibodies to lysosome associated membrane protein-2 in pediatric PSV that correlate to vasculitis-associated kidney dysfunction; and (3) the identification of nine patients, originally diagnosed with chronic PSV, harbouring novel and/or rare variants in adenosine deaminase 2 (ADA2) – these genetic data alongside having abrogated ADA2 enzyme activity have led to their reclassification as having a new monogenic form of vasculitis, deficiency of adenosine deaminase 2. This dissertation reports the first large-scale genotype and biomarker study of primary vasculitis focusing solely on pediatric cases. Improved classification is critical for timely therapeutic intervention, for identification of appropriately classified children for clinical trials, and for research, all of which will improve our understanding of the disease and the quality of life of children suffering from chronic PSV.

View record

Genetic diseases are a leading cause of death and disability worldwide with immense economic and societal burdens. However, currently fewer than 5% of genetic diseases have approved treatments, often with limited therapeutic benefit to patients. Genome editing provides a new therapeutic strategy to treat genetic diseases. In the last few years, the Clustered Regularly Interspaced Short Palindromic Repeats/CRISPR-associated protein 9 (CRISPR/Cas9) system has emerged as a promising gene editing approach due to its ease in design and flexibility to target different genetic diseases. One factor limiting its broader application is that it will introduce double-stranded deoxyribonucleic acid (DNA) breaks (DSBs). New advances in the CRISPR/Cas9 system are “base” and “prime” editing, which do not generate DSBs. Base editing has shown great potential to precisely repair the majority of pathogenic point mutations in the patient’s DNA. Prime editing, while currently less efficient, can introduce a broader range of possible edits. However, there remains a major challenge: delivery of genome editors to the affected tissues.To begin to address this challenge, two reporter gene mouse models were developed. Both models introduce a point mutation to knockout the reporter gene activity. Upon gene editing, the mutation is corrected back to the wildtype sequence, thereby restoring the reporter gene activity. The efficiency of gene editing can be visualized and measured with an in vivo whole animal imaging system, which permits repeated and long-term assessments of gene editing. To validate both mouse models, base editor mRNA and a mutation-specific single guide RNA (sgRNA) were encapsulated with lipid nanoparticles (LNPs) and intravenously administered into the reporter mice. This resulted in 83.5% restoration of wildtype reporter gene activity in the liver. One major application of these mouse models will be to evaluate and optimize gene editor delivery systems targeting various tissues. As an example, intramuscular administration of LNPs encapsulated with base editing components led to significant genome editing in the injected muscle, which could be readily visualized in the live imager. These novel reporter gene mouse models will provide critical tools for the in vivo evaluation and optimization of genome editor delivery and improved genome editor formulations.

View record

Adverse drug reactions (ADRs) can have life-altering consequences for patients. Several ADRs occur in the context of chemotherapy as it often involves multiple therapeutics at high doses and the non-specific destruction of actively growing cells. Cisplatin-induced hearing loss (CIHL) and anthracycline-induced cardiotoxicity (ACT) are severe chemotherapy-induced ADRs that can cause permanent hearing loss and congestive heart failure, respectively. It is therefore important to predict these ADRs before treatment is started such that treatment plans can potentially be altered to avoid long-term effects. Several studies have identified genetic predictors for CIHL and ACT; however, these variants do not account for all instances of these ADRs, and the clinical relevance of many identified variants remains unknown. The goal of this dissertation was therefore to investigate the clinical relevance of known, and identify novel, genetic predictors for CIHL and ACT.A systematic review of known genetic associations with CIHL was performed and determined that variants in ACYP2 and TPMT had strong evidence for association with CIHL. For ACT, in a cohort of adult patients treated with anthracyclines, genetic variants in RARG and SLC28A3 were more clinically relevant in a well-characterized subset. Through genome-wide analyses, a variant with a strong, but non-significant, risk effect on ACT was identified (rs34644385; odds ratio=5.09, 95% confidence interval=2.46-10.50, P=1.09x10⁻⁵) and calcium signalling was implicated in the development of ACT in adult patients. Finally, through transcriptome-wide association study analyses in pediatric patients treated with anthracyclines, an association between dysregulated GDF5 and ACT was identified (P=1.70x10⁻⁵), replicated in an independent cohort (P=3.54x10⁻³), and was functionally validated. Genes known to be essential for survival were also implicated in the development of ACT.The results of this dissertation enable development of a new CIHL predictive pharmacogenetic test that includes ACYP2 along with TPMT and recommend early pharmacogenetic testing to prevent CIHL. The additional genetic insight into ACT may also help improve the predictive ability of pharmacogenetic testing for ACT. This dissertation contributes to expanding the pharmacogenetic literature and improving patient care through the investigation and identification of genetic predictors for CIHL and ACT.

View record

Nature presents a solution to drug resistant infections in the form of cationic host defense peptides (CHDPs), such as LL-37, that exhibit wide-ranging antimicrobial, anti-biofilm, and immunomodulatory properties. Innate defense regulators (IDRs), such as IDR-1002 and IDR-1018, are synthetic derivatives of CHDPs with enhanced effects on the host immune system but diminished direct antimicrobial activity. Many barriers exist in the development of CHDPs and IDRs. These include toxicity arising from the cationic and amphiphilic nature of the peptides, which gives them a propensity to lyse cell membranes and precipitate. Formulations have been reported, such as polymer conjugates, that improve their biocompatibility. However, a fundamental lack of knowledge exists around the pharmacokinetics of CHDPs and IDRs, whether in their native or formulated state. Understanding how these peptides are absorbed, distributed, metabolized and excreted is important to fully understand their mechanism, safety and efficacy profile, and to rationally drive their formulation and clinical translation. In this thesis, a detailed pharmacokinetic study was undertaken for LL-37, IDR-1002 and IDR-1018. The peptides were radiolabeled with gallium-67 and injected into healthy mice at doses between 1-40 mg/kg via several routes of administration— intravenous (IV), subcutaneous (SC), intratracheal (IT), intraperitoneal (IP). The peptides were then quantitatively tracked up to 48 h post-injection using nuclear techniques. At low doses, all peptides were well-tolerated across all routes of administration, rapidly absorbed from extracellular sites and rapidly cleared from the blood compartment. Higher doses of IDRs caused acute respiratory failure and death when given IV or IT. Pronounced lung distribution and disruption to the air-exchange surfaces was observed with high doses of IV IDRs. Delayed systemic absorption was noted upon dose escalation in the SC and IP cohorts, and visual signs of precipitation were observed at the SC injection site for IDR-1002. To overcome some of the toxicity and pharmacokinetic limitations observed, IDR-1002 was also reformulated in this thesis onto hyperbranched polyglycerol (HPG) via a redox-sensitive linkage. The conjugates had enhanced colloidal stability and maintained the biological profile of the IDR. Conjugates with a 5:1 peptide-to-polymer ratio, but not higher, circulated well when administered IV compared to native IDR-1002.

View record

There are currently no effective treatments for 95% of the over 7,000 known genetic diseases [1, 2]. The use of CRISPR/Cas9 has recently emerged as a powerful approach to directly repair disease-causing mutations . The goal of this project was to optimize a novel CRISPR/Cas9 ‘base editing’ approach to treat genetic diseases by specifically repairing the disease-causing mutation in situ in the patients’ DNA. This technology uses a fusion of a partially deactivated Cas9 (nCas9) protein coupled with a nucleotide modifying enzyme to achieve 15-30% editing, and in some cases up to 75% editing in vitro . Two-thirds of all genetic diseases are caused by point mutations and base editors have now been developed that can target more than 68% of all disease-causing point mutations . Despite the growing potential of base editing, limitations in the ability to quantify gene editing quickly and precisely have resulted in a lack of robust high throughput studies to compare and optimize base editors. We hypothesized that developing base editor reporter model systems would enable us to efficiently test and optimize the safety and effectiveness of base editing. To this end, we successfully developed both in vitro and in vivo reporter models and utilized these models to test the safety, delivery, and effectiveness of base editing. This series of novel model systems address a critical gap in the field of CRISPR/Cas9 gene editing. Using these, we were able to significantly increase the efficiency of base editing in vitro. We were also able to compare the safety of different base editors using a new application of my in vitro reporter model. We also demonstrated the therapeutic feasibility of our base editing approach in vitro to repair three prevalent mutations that cause a debilitating genetic disease: Lipoprotein Lipase Deficiency. Lastly, using the in vivo reporter model, we were able to demonstrate sustained correction and efficient delivery of base editors using lipid nanoparticles. Overall, these experiments have demonstrated critical steps towards the implementation of new treatment options for patients with a wide variety of genetic diseases.

View record