Skip to main content
To KTH's start page To KTH's start page

Finding order in chaos

Dissecting single-cell heterogeneity in space and time

Time: Fri 2024-05-03 10.00

Location: Air&Fire, Science for Life Laboratory, Tomtebodavägen 23, Solna

Video link:

Language: English

Subject area: Biotechnology

Doctoral student: Christian Gnann , Proteinvetenskap, Lundberg

Opponent: Professor Lucas Pelkmans, University of Zurich, Department of Molecular Life Sciences

Supervisor: Professor Emma Lundberg, Proteinvetenskap; PhD Manuel D Leonetti, Chan Zuckerberg Biohub, San Francisco

Export to calendar

QC 2024-04-08


The cell is the smallest unit of life and contains DNA, RNA, proteins and a variety of other macromolecules. In recent years, technological advances in the field of single cell biology have revealed a staggering amount of phenotypic heterogeneity between cells in a population, which were previously considered homogenous. Previous work has largely been focused on studies of RNA. As proteins however are the ultimate effectors of genetic information, this thesis aims to provide a protein-centered view on cellular heterogeneity, particularly focusing on cell cycle and cellular metabolism.

Most of my work has been performed within the framework of the Human Protein Atlas project. In the context of this project, we mapped the spatial distribution of more than 13.000 human proteins with subcellular resolution and found that around a quarter of all human proteins exhibit protein expression heterogeneity.

In Paper I, we hypothesized that a majority of the observed cellular heterogeneity can be explained by differences in cell cycle progression. Therefore, we generated a map of proteomic and transcriptomic heterogeneity at subcellular resolution, which we precisely aligned to the cell cycle position of individual cells. This approach allowed us to identify hundreds of previously unknown cell cycle-related proteins. With sustained proliferative signaling representing a hallmark of cancer, novel cell-cycle proteins could serve as potential new drug targets against cancer. We further show that a large part of cell cycle dependent proteome variability is not established by transcriptomic cycling. This suggests that post-translational modifications are a major contributor to the regulation of cell cycle dependent protein level changes. Therefore, in Paper II, we carried out a deep phosphoproteome mass spectrometry profiling of the same cellular model as in Paper I and identified almost 5,000 cell cycle dependent phosphosites on over 2,000 proteins. The unprecedented scale of our phosphoproteomic data allows us to link cell cycle dependent protein expression dynamics to phosphorylation events. Furthermore, we identify a large set of proteins with stable expression levels and fluctuating phosphorylation patterns along cell cycle progression that likely alters protein function.

Despite identifying hundreds of novel cell cycle dependent proteins in paper I, we observed that the majority of heterogeneously expressed proteins display variable expression independent of cell cycle progression, among them a large number of metabolic enzymes. Thus, we sought to describe the extent of subcellular metabolic complexity in human cells and tissues in Paper III. While we confirm metabolic compartmentalization in our dataset, we show that around 50% of metabolic enzymes localize to multiple cellular compartments. By integrating public protein-protein interaction data with our subcellular location information, we identify several enzymes with novel compartment-specific functions. Additionally, we observe a strongly elevated number of heterogeneously expressed enzymes compared to the background of the human proteome that is largely independent of cell cycle progression. We show that this heterogeneity can be manifested in the lineage of a single cell and is conserved in situ. To conclude, we suggest that the extensive metabolic heterogeneity can establish functional metabolic states in a population of human cells.

Finally, in Paper IV, we assessed the heterogeneity of the mitochondrial proteome as they are metabolic powerhouses containing an elevated number of cell cycle independent variably expressed proteins. In this study, we correlated the variable expression of over 400 mitochondrial proteins to the expression of rate limiting enzymes in important mitochondrial pathways; such as the TCA cycle and ROS metabolism. We show that enzymes in the same pathways often correlate in their expression, indicating that their expression variability may contribute to the establishment of metabolic states.

Altogether, the thesis illuminates the spatiotemporal complexity of the human proteome established by protein multilocalization and expression heterogeneity as fundamental non-genetic means of functional cell regulation.