One of the long-standing questions puzzling antibody researchers is why somatic hypermutations are concentrated in the three short, non-consecutive complementarity determining regions (CDRs) in the antibody gene sequence. First documented by the labs of David Baltimore, Klaus Rajewsky, and Leroy Hood in two Cell papers in the early 1980s, CDR hypermutation is well accepted and taken for granted by immunologists as an axiom in textbooks, although the mechanism is unknown. The study, published online on April 24, 2023 in Cell, reports that the flexible single-stranded DNA feature is the key to CDR-hypermutation. Using a cutting-edge, high-throughput biochemical assay that can test many DNA substrates for their deamination/mutation ability, the researchers found that the CDR hypermutation is evolutionarily conserved in species that use somatic hypermutation to diversify their antibody repertoire (Figure 1). Using the powerful passenger antibody gene allele mouse model system, which allows the detection of unselected mutational events, and the CRISPR/Cas9 system, which allows rapid generation of mice with DNA sequence alterations, the researchers found that the CDR hypermutability depends on the DNA sequence context at the mesoscale level (5-50 bp) (Figure 2). Using a combination of molecular dynamics simulations and single-molecule biochemistry, the researchers demonstrated that the targeting preference of the DNA mutator enzyme, activation induced cytidine deaminase (AID), is directly regulated by the flexibility of the single-stranded DNA substrate (Figure 3). Analysis of the antibody gene sequence showed that DNA sequences encoding the CDRs have evolved highly flexible properties to facilitate hypermutation, revealing a non-coding role of these sequences, and explaining the hypermutation pattern in lymphoma. Finally, regions that are normally “cold” for hypermutation can be made “hot” by engineering flexible DNA sequences in the “cold” regions (Figure 4), opening the door to the next generation of humanized animal models for antibody discovery.