Over the past decade, technological progress has facilitated profiling single cell and single nuclei transcriptomes through innovative next-generation sequencing approaches that empowered the exploration of cellular phenotypes in healthy and diseased tissues with unprecedented precision. This dissertation aimed to unravel a deep understanding of human development and disease biology by studying the underlying cellular architecture and molecular mechanisms using single nuclei RNA sequencing (snRNA-seq). Specifically, two distinct systems were investigated, namely, a) maternal-fetal interface in healthy and early-onset pre-eclamptic pregnancies that comprised of human decidua and placenta and b) pancreatic neuroendocrine carcinoma and its comparison with healthy adult pancreatic cell types or states.
Pre-eclampsia (PE) stands out as one of the most severe pregnancy disorders, characterized by hypertension, proteinuria, cardio-metabolic dysfunctions, and various multi-organ complications, ultimately leading to preterm delivery, maternal mortality, and associated morbidities. Early-onset pre-eclampsia (eoPE) is particularly formidable, typically necessitating delivery before the 34th week of gestation and leading to approximately 80,000 maternal and 500,000 fetal deaths annually. Importantly, eoPE currently lacks adequate biomarkers for early screening and clinical management. Diagnosis relies solely on clinical and biomarker signs in late pregnancy, where maternal and fetal morbidity is often irreversible. This dissertation presents a comprehensive cell atlas of the human maternal-fetal interface, comprising of around 225,000 nuclei from the maternal decidua and fetal placenta obtained during the first trimester, healthy term and eoPE pregnancies, utilizing complex snRNA-seq data harmonization. A novel nuclear state termed juvenile syncytiotrophoblast (STBjuv) was identified, demonstrating previously unexplored transcriptomic diversity and a division of labor within the placental syncytiotrophoblast. Notably, the comparative analysis of decidua and placenta samples from term controls and eoPE revealed differences in composition, differential gene expression and transcriptional regulation patterns, along with changes in signaling pathways. These findings provide mechanistic insights into cell type or state-specific dysregulations occurring in eoPE. Moreover, employing advanced spatial transcriptomics techniques such as In-situ RNA sequencing (ISS) and 10X Visium sequencing facilitated the spatial delineation of cell types and states within the human placenta and established a connection between spatial and transcriptomic heterogeneity, presenting a novel discovery in this evolving field of pregnancy research.
The computational analysis identified a dysregulated syncytiotrophoblast development to serve as an initiation point for eoPE that is characterized by perturbations in transcriptional factors/co-activators, specifically the master-regulator EP300, FOXO1, SCRT2, FOXO4, FOS, and PAX5. Significantly, the enriched transcriptional regulators exhibited a noteworthy overlap in their downstream targets, imparting functional implications across various signaling pathways, including HIF1, AP1, TGFb, Wnt, PI3K-Akt signaling, and vesicle-mediated transport. Of note, a significant proportion of the perturbed syncytiotrophoblast differentiation drivers included EP300 (or, p300) regulated fusogenic targets that suggest impaired trophoblast syncytialization as a significant contributing factor in the development of eoPE.
Significantly, the discoveries from this work indicate that eoPE possibly originates in the outer syncytiotrophoblast sub-states in the fetal placenta and is marked by an augmented senescence-associated secretory phenotype (SASP) profile. The heightened senescence resulted from elevated ligand pressure, facilitated by the secretion of GDF15, INHBA, HSPG2, MIF, TGM2, ADAM9, and ADAM12 that could potentially traverse the maternal-fetal interface, and translate the disease from the fetal to the maternal side. Of note, In-situ RNA sequencing (ISS) analysis unveiled a statistically significant proximity between the senescent marker (INHBA) and markers of fetal vessels in eoPE. This association is not detected in term controls. Consequently, the findings of this thesis emphasize that a disrupted communication between syncytiotrophoblast sub-states and maternal decidua is the key to a dysregulated maternal-fetal barrier and potentially compromised maternal uterine vessel remodeling in eoPE. Remarkably, the presented data suggest a potential strategy for the prevention, intervention, and clinical management of eoPE through the pharmacological inhibition of these ligands associated with the senescence-associated secretory phenotype (SASP), including GDF15, HSPG2 and INHBA.
The subsequent chapter of this dissertation delved into the molecular intricacies guiding the development of healthy pancreatic islets— specifically, the beta cell type, and next assessed the role of these differentiation drivers in the context of pancreatic cancer. Specifically, this thesis focused on high-grade pancreatic neuroendocrine carcinoma (panNEC) with large cell morphology, a subtype presenting challenges in classification and treatment. While advancements in molecular genetics have progressively uncovered significant inter-tumor heterogeneity, there remains an unexplored realm regarding the extent of intra-tumoral heterogeneity and lineage plasticity. Like the pre-eclampsia study, a snRNA-seq approach was utilized to deconstruct the cellular landscape of panNEC that delineated both shared and unique malignant sub-states associated with specific signaling pathways and transcriptomic regulatory programs driving tumor pathophysiology and heterogeneous clinical behavior.
Notably, this work identified a shared neuroendocrine sub-state characterized by robust induction of heat shock protein encoding mRNA (HSP+), exhibiting signatures indicative of activation of the unfolded protein response, hypoxia, mTORC/PI3K-AKT signaling, and glycolytic shift. In one patient sample, a unique stromal sub-state depicted enriched YAP/TAZ-associated Hippo signaling alongside mesenchymal and basaloid programs expression, reflecting transcriptomic similarities with pancreatic ductal-adenocarcinoma (PDAC). Furthermore, one of the shared neuroendocrine sub-states was highly proliferative and was characterized by overexpressed E2F targets, including Enhancer of Zeste homolog 2 (EZH2). Notably, this sub-state demonstrated significant enrichment for PTF1A-regulated targets specific to the brain while repressing the pancreas-specific targets. This observation suggests a shift from a pancreatic lineage program in the cell of origin towards a more generic neuronal phenotype and a concomitant enrichment of signatures related to the DNA damage response, vulnerabilities in cancer stem cells, and chemotherapeutic resistance. Of note, this de-differentiated neuronal program could be exploited as the achilles heel of panNEC for devising therapeutic strategies.
Hence, the presented data unmasked considerable heterogeneity and therapeutic vulnerabilities in high-grade panNEC, emphasizing the importance of tumor profiling for personalized treatment approaches. Of note, it highlights the prospect of clinical intervention to target two shared neuroendocrine sub-states suggesting the feasibility of personalized combination therapies in clinical settings.
In a nutshell, this dissertation extensively explored two distinct but critical systems at a cellular resolution in the context of disease development and progression. The snRNA-seq approach presented an unbiased and deep understanding of disease biology that demonstrated significant translational potential. Notably, in both systems, specific cell (or, nuclear) sub-states were ascribed to the disease origin that added to our understanding on how a disease evolves on a molecular level and suggested potential avenues for therapeutic development. This transition from concentrating on a single molecular target to addressing the underlying cellular dysfunction presents novel opportunities for the clinical targeting of cellular signatures for complex diseases and for developing a new generation of therapeutics.
