Evolutionary Dynamics, Diversity and Interactions of CRISPR Arrays

Abstract

CRISPR-Cas systems provide adaptive immunity against invading genetic elements in prokaryotes. Their CRISPR arrays store spacers acquired from previous infections, creating a chronological record of host–phage interactions. However, spacer deletions, array orientation, and interactions between multiple arrays and systems complicate the reconstruction of their evolutionary histories.

This thesis investigates CRISPR array evolution through computational and statistical methods. We developed SpacerPlacer, a framework for reconstructing the evolutionary histories of related CRISPR arrays along phylogenies while preserving spacer order. We further developed stochastic models and statistical methods to estimate evolutionary rates and test hypotheses about spacer deletion mechanisms. Additionally, we introduced CRISPR-evOr to predict the acquisition end of CRISPR arrays, providing information about their orientation and leader location.

Applying these methods to large-scale datasets revealed conserved evolutionary patterns across CRISPR-Cas types. Spacer deletions are frequent, often remove multiple adjacent spacers, occur less frequently at array boundaries, and exhibit patterns consistent with selective pressures. Furthermore, CRISPR-Cas systems display type-specific patterns of genomic organization, repeat similarity, and orientation. Their co-occurrence patterns suggest diverse interactions, ranging from independent or potentially antagonistic behavior to cooperation in spacer acquisition, expression, and interference.

Together, these findings establish a quantitative framework for studying CRISPR array evolution, revealing evolutionary dynamics and organizational patterns that improve our understanding of CRISPR-Cas functionality and microbial host–phage interactions.