LSC-Retreat 4.0 on Mathematical Modelling in Biology

“DNA damage in paleogenomics: a ‘friendly’ enemy”

Paleogenomics is a relatively new discipline that in the last couple of decades has massively contributed to the understanding of, among others, the taxonomy, phylogeny, phylogeography and ecology of extinct species as well as of ancient populations of extant species. Whereas the first studies from the 1980s were based on relatively short sequences, the improvement in sequencing (Next Generation Sequencing, NGS), wet laboratory protocols and bioinformatic methods during the last decade has allowed us to obtain and analyse a huge amount of data. Although in the pre-NGS studies, the postmortem DNA damage/degradation was already identified as a potential issue in sequence quality, its role in the analyses currently performed is very important to assess the preservation and age of the samples, and to identify possible modern DNA contamination.

“Overview of DNA repair mechanisms in eukaryotes”

The preservation of genome stability is one of the fundamental tasks that a cell must accomplish to faithfully transmit its genetic information to the next generation. However, the integrity of our DNA is under constant threat from multiple sources of damage, both endogenous (oxidative reactions, spontaneous hydrolysis, chemical modifications, replicative stress, etc.) and environmental (ionising radiation, UV-light, reactive compounds, chemotherapeutic drugs, etc.). All these genotoxic agents inflict a myriad of lesions in the DNA of each cell every day. In this talk, I will summarize the repair pathways that our cells employ to counteract each specific type of DNA damage and, therefore, avoid the accumulation of mutations that cause various diseases, including cancer.

Enzyme promiscuity has often been identified as a latent ability in metabolic networks to evolve novel activities and adapt to new environments. Understanding the mechanisms of enzyme promiscuity, thus, can bring new insights into how pathways have evolved and guide metabolic engineering and directed evolution strategies. To that end, predicting enzyme promiscuity might be attempted solely based on sequence and structural analyses but, more interestingly, they can be combined with the analysis of chemical signatures of biochemical transformations into a multidimensional chemical and sequence-based tree of life. Here, we will use the reaction chemical signatures RetroRules system to model two possible mechanisms for enzyme evolution, the progressive and the regressive models, correlating the results of their simulations to the promiscuity distribution as observed through phylogenetic studies of modern metabolic networks. Based on those models, we will develop an approach to score engineered heterologous pathways in chassis organisms, as predicted by RetroPath, based on the likelihood of the efficiency of the reactions, and will discuss a more general approach based on Selenzyme to prioritize enzyme selection for a given reaction rule and biosensor design by SensBio. Finally, multiobjective optimization strategies for sampling solutions in the extended design space of production- sensing-regulation metabolic circuits will be discussed.

“In praise of imperfection: tinkering and opportunism in the evolution of metabolism”

François Jacob published an epoch-making article in Science in 1977. Contrary to what incipient molecular biology might indicate, that cells are complex systems with a mesh of highly sophisticated mechanisms perfectly adapted to their function, Jacob invites us to appreciate evolution as something far removed from perfection. Natural selection is more like a tinkerer than an engineer following a set plan and design. Admittedly, the idea of evolutionary bricolage is not original to Jacob but to Darwin himself, who had already formulated it in his study of the co- evolution of flowers and pollinating insects. Jacob’s general formulation of molecular bricolage has materialized in numerous examples and mechanisms of metabolic evolution. A particularly interesting case is that of catalytic promiscuity and its role in enzymatic adaptation and evolution. Today, the precision and specificity observed in many enzymes are confronted with an idea of disordered biology in which noise, infidelity, stochasticity or heterogeneity may also play a crucial role in evolutionary innovation. The exploration of what is chemically possible during oxygen accumulation on Earth is a good example of the role of enzyme plasticity in metabolic evolution.

“Can mechanical cues guide anastomosis?”

Angiogenesis is the generation of new blood vessels from the pre-existing vasculature. It is a complex, multi-scale process involving cell migration through a heterogeneous and dynamic medium (extracellular matrix, ECM) in response to signaling cues (known as angiogenic factors, AF). In this thesis, we address a process crucial for the formation of a functional vascular network, anastomosis, whereby two growing sprouts join to form a perfused vessel. We will explore the feasibility of a recently proposed hypothesis that mechanical cues guide anastomosis.

“Mathematical modelling of vasculature regression during cartilage condensation”

The growth of an embryo, after it reaches a size of ca. 2mm, heavily depends on a functional vascular system. Unlike other organs, the vascular system needs to be fully functional and constantly remodelling itself to support all tissues from an early stage in embryo development. A crucial question – still not totally understood – is how the 3D architecture of blood vessel networks is controlled during organogenesis. Specifically, in the case of limb development, vasculature is

“Temporal rank dynamics of transcripts in SARS-CoV-2-infected cells”

Taylor’s power law is a ubiquitous statistical law in complex systems that describes the distribution of elements by a power relationship between their mean and variance. Here, we used Taylor’s law coupled with a ranking system to analyze the longitudinal stability of transcripts in cells infected with severe acute respiratory virus 2 (SARS-CoV-2) in single-cell RNA-sequencing (scRNA-seq) data. For all three different cell types analyzed (human bronchial epithelial cells and colon and ileum organoid cells), we found that instability increases at later stages of infection. The top 25% genes with the highest accumulated rank were then grouped into stability quartiles. The quartile of the most stable genes showed the highest similarity across cell types, followed by the quartile with the least stable genes. Genes clusters were built based on their rank dynamics. This analysis suggests that the temporal dynamics of some interferon-respondent genes is cell type- specific. Further analyses are underway to dissect the goodness-of-fit of Taylor’s law to scRNA- seq data, as well as to identify gene modules with abnormal behavior and modules that drive instability in these systems.