Replication gaps that persist into mitosis likely represent important threats to genome stability, but experimental identification of these gaps has proved challenging. We have developed a technique that allows us to explore the dynamics by which genome replication is completed before mitosis. Using this approach, we demonstrate that excessive allocation of replication resources to origins within repetitive regions, induced by SIR2 deletion, leads to persistent replication gaps and genome instability. Conversely, the weakening of replication origins in repetitive regions suppresses these gaps. Given known age-and cancer-associated changes in chromatin accessibility at repetitive sequences, we suggest that replication gaps resulting from misallocation of replication resources underlie age-and disease-associated genome instability.SIR2 | DNA replication | repetitive sequences | replication gaps | ribosomal DNA S taggered initiation of DNA replication, which is common across eukaryotes from fungi to humans, means that, at any given time, in S phase, only a fraction of replication origins is activated. In recent years, a model has emerged to explain this pattern of DNA replication (1, 2). This model assumes that the pool of initiation factors required to fire licensed origins in S phase is limited, and therefore sufficient to fire only a subset of licensed origins at any given time. Licensed origins differ in their ability to recruit factors required for firing, so origins with higher affinity or accessibility fire earlier than those with lower affinity or accessibility. After activating the initial set of origins, firing factors are released, enabling the next set of origins to fire. This results in successive waves of origin activation. Genome replication eventually finishes when areas with the least accessible origins are replicated.In healthy human cells, the least accessible genomic regions consist of repetitive DNA, which represents about half of the genome and tends to be compacted into heterochromatin. However, recent studies suggest widespread opening of hetrochromatin during carcinogenesis and aging (3-5). Such reorganization of chromatin would expose a new suite of origins within repetitive DNA that could potentially compete initiation factors away from unique portions of the genome, thereby disrupting the normal genome-wide hierarchy of replication timing. Increased origin activity within repetitive DNA could thus compromise replication elsewhere in the genome. Given the limited pool of initiation factors, we propose that an increase in density of active origins within repetitive regions could result in a decreased density of such origins in unique regions of the genome, which, when combined with the stochastic nature of origin firing, may occasionally result in replication gaps, i.e., unique regions of the genome that are unable to complete replication before mitosis. This so-called "Random Replication Gap Problem" (RRGP) is the subject of long-standing speculation, but such gaps have never been experim...