Single-cell RNA sequencing for subtype discovery in Plasmodium falciparum and mammalian cells

Abstract

Since the dawn of massively parallel sequencing technologies in mid-2000s their utility in profiling the expression of genes in a genome-wide fashion has matured and progressed from cell populations to individual cells. In particular, single-cell RNA sequencing (scRNA-seq) has impacted numerous domains in life sciences and hold immense promise in biology and medicine. Indeed, it has become realistic to chart the complete set of cell types and states in multicellular organisms, and projects have started to map out cell types in humans (i.e. the Human Cell Atlas project) and model organsims. In this thesis, I present the application of scRNA-seq to infectious disease and cancer as well as a computational assessment of the general possibilities and limitations of scRNA-seq for enumerating cell types and states de novo.

In Paper I, we describe the ability of scRNA-seq to profile transcriptomes from individual malaria-causing P. falciparum parasites. We reveal heterogeneity even among synchronized cultures of parasites during their red blood cell life cycle. Moreover, we identify a subset of sexually differentiated P. falciparum with a distinct gene signature, likely important for parasite transmission that may be exploited for the design of transmission-blocking drugs and/or vaccines. In Paper II, I present a computational strategy to identify the magnitude of biological gene expression differences needed for accurate inference of cell identities using scRNA-seq. Interestingly, rather large differences are needed for proper cell state discrimination, irrespective of scRNA-seq protocol, implying that large number of cell states may escape detection. In Paper III, we used scRNA-seq and bulk RNA-seq to characterize the molecular programs during the later stages of lung metastasis. We demonstrate that a transition from epithelial to mesenchymal cell characteristics occurs in cancer cells during metastasis, and that the mesenchymal properties are maintained during metastasis growth extending over a week. In Paper IV we performed transcriptome analyses on stem and progenitor populations in myelodysplastic syndrome (MDS) patients. We provide evidence that the MDS stem cells and the progenitors have distinct transcriptome.

Altogether, this thesis expands the applications of scRNA-seq towards parasite biology and cancer metastasis and we provide valuable insights into the abilities of current scRNA-seq technologies in mapping cell states in an unbiased fashion.