(2dq) A Systems Approach Towards Reconciling Single-Cell Heterogeneity and Cell Phenotype in Health and Disease

Single-cell heterogeneity, a fundamental property of multicellular systems, provides a mechanism for tissues, organs, and organisms to have a wide range of responses to a dynamic environment. Such heterogeneity is quite pervasive across neurons in the brain and is an essential feature that enables robust regulation of complex physiological functions necessary for life. However, under certain conditions, such heterogeneity can be a major contributor to disease pathology. In the context of cancer, single-cell heterogeneity plays a major role in drug resistance. Somatic mutations and transcriptional variability lead to changes in the underlying gene regulatory network states underlying tumor formation and cell-state stability. Complicating matters further, these changes often vary from cell-to-cell within a tumor (i.e. intratumoral heterogeneity) and support multiple mechanisms through which subpopulations of drug-resistant tumor cells arise or previously drug-sensitive cell populations can acquire a drug-resistant phenotype during treatment.

Motivated by the complexity of single-cell heterogeneity, my research to date has focused on understanding the role heterogeneity plays within the context of physiological functions and emergent properties in health and disease. This research, guided by a systems approach towards understanding complex biology, ranges from biomolecular characterization of single neurons/cells via high-throughput qPCR and single-cell RNA-seq (scRNA-seq), bioinformatics analysis of high-throughput data, and quantitative modeling of physiological control systems. During my graduate training, I investigated the role of single-cell heterogeneity in blood pressure set point regulation. This work revealed an organizational framework in which transcriptional heterogeneity, pervasive across phenotypically homogeneous brainstem neurons, when organized with respect to the synaptic inputs actually reflected a gradient of neuronal states that enabled varied neuronal signaling outputs necessary for robust blood pressure set point regulation. Conversely, the maladaptation of these brainstem neurons resulted in the dysregulation of cardiovascular (CV) functions, which contribute to clinical phenotypes associated with heart failure. Through computational simulations, I showed that rescuing neuronal functions could theoretically improve these clinical phenotypes by compensating for physiological damages due to heart failure. Additional projects have also included reconciling the transcriptional heterogeneity among neurons within the suprachiasmatic nucleus (SCN), the central regulator for circadian rhythms. Through analysis of individual SCN neurons using high-throughput qPCR, I identified an underlying cell-interaction network, defined by autocrine/paracrine signaling among multiple neuronal cell states, which generated coordinated signaling outputs regulating circadian rhythms.

In my postdoctoral training, I am applying the skills and concepts developed from my PhD towards understanding cell-to-cell heterogeneity in the context of cancer progression, specifically in glioblastoma (GBM) – the most prominent and lethal primary brain tumor in adults. I am currently collaborating with multiple clinicians to develop regulatory network models that would inform on targetable mechanisms at the single-cell level that drive treatment-induced resistance and tumor progression in GBM stem-like cells (GSCs), a rare subpopulation of tumor cells that drive tumorigenesis and recurrence. By applying single-cell-level analysis and network inference to time-course transcriptomic responses of patient-derived GSCs (PD-GSCs), I identified distinct transcription factor (TF) networks and corresponding expression dynamics that underlie drug-induced cell-state transitions, which resulted in an increase in drug-resistance in a subpopulation of PD-GSCs. Moreover, subsequent development of ODE models and simulation analysis revealed specific TFs whose expression when knocked down improved killing efficacy of the drug pitavastatin, a drug known to induce apoptosis in GBM tumor cells. Ultimately, this approach has revealed druggable TFs and enabled rational selection of drug combinations resulting in improved efficacy of secondary drugs that were previously ineffective against the PD-GSCs.

Moving forward, my goal is to expand on my current work by building multi-scale models that will characterize regulatory network mechanisms within multiple cell types and corresponding cell-interaction networks. I will initially focus primarily on tumor cells and macrophages, which have been shown to drive the emergence of drug-resistance within the tumor cell ecosystem in response to drug treatment. I will apply a combination of experimental and computational techniques to achieve these goals. My lab will employ multiple in vitro models including monocultures and tumor cell/macrophage co-cultures of PD-GSCs and macrophages with single-cell RNA sequencing and systems biology approaches to understand the regulatory mechanisms and cell-type interactions underlying tumor cell dynamics. The use of in vitro models coupled with a systems analytical approach will enable an extensive/detailed analysis of the cellular components, corresponding cell states, and cell-state dynamics enabling the identification of regulatory mechanisms and cellular interactions underlying tumor evolution and the emergence of drug resistance. Beyond in vitro modeling and analysis, my lab will also make use of the plethora of publically available single-cell-level omic-scale expression and epigenetic data (scRNA-seq and scATAC-seq) of GBM and GBM-associated macrophage data to infer regulatory network models that will enable hypotheses generation and target predictions. I will also seek to establish clinical collaborations to apply a systems approach to analyze tumor biopsies towards identifying cell states and corresponding regulatory network states.

Through my experience in experimental and computational systems biology, cancer biology, and collaborative experience with clinicians, I believe I am well positioned to achieve these goals.

Research Funding

• I worked with my current postdoctoral advisor to create a recently awarded R01 proposal under the National Cancer Institute (NCI) (1R01CA259469-01A1; Dec 2021 – present). This proposal, based largely on my current postdoctoral work, aims to identify regulatory mechanisms underlying treatment-induced drug resistance.
• F32 Ruth L. Kirschstein Postdoctoral Individual National Research Service Award (5F32CA247445; 2019 – 2022) through the NCI – investigating the regulatory mechanisms that underlie GBM intratumoral heterogeneity.
• F31 Ruth L. Kirschstein Postdoctoral Individual National Research Service Award (1F31AA023143; 2014 – 2016) through the National Institute for Alcohol Abuse and Alcoholism (NIAAA) – identifying gene networks driving neuronal states during alcohol withdrawal.

Teaching Interests

I have been fortunate to have had multiple opportunities to develop my STEM teaching and mentoring abilities. In graduate school, I was a teaching assistant for Process Controls and Dynamics, a senior-level core chemical engineering course for the undergraduate curriculum. This involved holding office hours as well as running the lab-based component for the course. From that experience, I was motivated to apply and was awarded a Graduate Teaching Fellowship for the same course. As a teaching fellow, I co-taught the course with two other experienced professors, preparing and delivering lectures that covered feedback control and stability, assigned homework, and created a mid-term exam. My experience as a Graduate Teaching Fellow has greatly influenced my perspective on teaching scientific material. I have had the most success in conveying scientific material when engaging students in dialogue. Emphasizing the “why” and clearly articulating the gaps and the problems that motivate the need to understand a particular topic helps place ideas in a broader context, which is essential to help improve a student’s understanding and critical thinking.

Beyond the classroom, I mentored three interns and three younger graduate students. I trained students in experimental techniques including laser capture microdissection and basic statistical analyses. The culmination of this mentorship/training led to the publication of a recent article (Frontiers in Neuroscience). As a postdoc, I have had the opportunity to train a recent college graduate in coding and bioinformatics techniques in the R platform. This student has since enrolled in a competitive bioinformatics masters programs at Georgia Tech University. Through these experiences, I have found that establishing the importance of fundamental concepts is critical to instilling deeper understanding to those trying to implement such methods.

I would be comfortable teaching a variety of topics including: bioinformatics, multivariate statistics, mathematical biology, systems biology approaches, and process control and dynamics.