UC Berkeley

Viruses are fundamental drivers of cellular evolution. Myriad immune systems evolved to combat viral infections, but these systems face a fundamental challenge: distinguishing self from non-self. Immune systems must mount a potent response to infection, but immunity must be specific and target only pathogens or infected cells. Non-self immune responses run the dire risk of autoimmunity. How do organisms evolve specific immune systems? This work explores two widespread immune systems in prokaryotes, CRISPR-Cas and Hachiman, to understand how molecular solutions evolved to address the core constraint of autoimmunity. These studies reveal that DNA topology manipulation and genome integrity sensing represent connected mechanisms for specifically targeting non-self.How do CRISPR-Cas adaptive systems avoid autoimmunity? CRISPR immune systems capture DNA fragments from mobile genetic elements and integrate them into the host genome as templates for RNA-guided immunity. The Cas1-Cas2 integrase is responsible for driving adaptation and must therefore distinguish self from non-self, but the mechanism was unclear. My biochemical and structural analyses reveal that a Cas1–Cas2/exonuclease fusion (trimmer-integrase) performs intricate DNA topological manipulations to protect and remove a precursor motif (PAM) that serves as an immunological tag distinguishing self from non-self.Most natural immune systems are innate rather than adaptive, and many recently discovered systems encode helicases. While helicases sense pathogen-associated nucleic acids in eukaryotic immunity, their prokaryotic roles remain understudied. I characterized the widespread Hachiman system using in vivo assays, biochemistry, cryo-EM, and phylogenetics, demonstrating it functions as a nuclease-helicase complex (HamAB). HamB acts as a sensor helicase that detects DNA damage, triggering allosteric activation of the HamA nuclease to degrade cellular DNA. DNA-damaging agents activate Hachiman without phage presence, indicating this system monitors genome integrity as a proxy for infection. I propose a thresholding model explaining how Hachiman-like systems avoid autoimmunity and achieve broad-spectrum defense.In addition to core immunity mechanisms, I discovered regulatory layers controlling immune responses. Gene context analysis revealed coevolving regulatory gene pairs acquired by multiple immune systems that encode a specific protease that activates a partner pro-nuclease to trigger genome destruction. This analysis was extended to identify caspase-nuclease (canu) family systems. Canu confers antiviral defense through a pathway reminiscent of eukaryotic caspase activation. This study demonstrates how modularity and effector exchange drive immune evolution and sets the stage to uncover key events in caspase evolution.These findings establish genome integrity sensing as a paradigm of prokaryotic immunity with broad evolutionary implications. The Hachiman helicase shares ancestry with critical eukaryotic helicases including Ski2 and Brr2, suggesting origins of eukaryotic regulatory strategies in immunity. Modular evolution revealed through nuclease-protease pairs and canu systems demonstrates how immune components evolve as interchangeable units, accelerating adaptation and presenting a new avenue for discovery. Our ability to trigger immune activation without viral presence challenges traditional models of immune specificity and reveals these systems may monitor cellular homeostasis rather than directly detecting pathogens. This work establishes molecular signatures of non-self in the genome across diverse immune systems and informs our understanding how antiviral defenses function and evolve.