Clonal Evolution

Molecular Clock

6 min read

Molecular Clock

Definition

The molecular clock in cancer genomics is the use of neutral (passenger) somatic mutations as a temporal record of a tumor’s evolutionary history. Because passenger mutations accumulate at a rate governed by the background mutation rate and have no effect on cell fitness, the number of passenger mutations unique to a given cell lineage serves as a measure of the time elapsed since that lineage diverged from its ancestor (Greaves & Maley, 2012). The concept transforms the cancer genome into a fossil record: each subclonal branch carries a passenger-mutation tally proportional to its evolutionary age.

Mechanism

The clock operates on a simple principle: somatic mutations arise stochastically during DNA replication and, if neutral, are neither selected for nor against. They persist in daughter cells by hitchhiking on clonal-expansions driven by driver-mutations. Over evolutionary time, the count of passengers in a lineage grows roughly linearly with cell divisions, making it a proxy for elapsed time (Greaves & Maley, 2012; Graham & Sottoriva, 2017).

flowchart TD
    A["Passenger mutations<br>accumulate during cell division"] --> B["Molecular clock:<br>passenger count proportional to<br>evolutionary time"]
    B --> C1["Relative timing of<br>clonal branching events"]
    B --> C2["Timing of driver<br>mutation acquisition"]
    B --> C3["Timing of copy-number<br>alterations"]
    B --> C4["Timing of mutational<br>process changes"]
    D1["Constant-rate assumption<br>valid in ~60% of samples"] -.->|supports| B
    D2["Spectrum shifts in ~40%<br>(Gerstung et al., 2020)"] -.->|complicates| B
    D3["Detection limit: CCF below 0.05-0.10<br>(Tarabichi et al., 2021)"] -.->|constrains| B

Figure: The molecular clock principle and its applications in cancer genomics. Passenger mutations accumulate during cell division, establishing a temporal record that enables four major classes of evolutionary timing inference. Two caveats — mutational spectrum shifts in ~40% of tumors and detection limits for rare subclones — constrain clock precision but do not invalidate the principle. Synthesized from Greaves & Maley (2012), Graham & Sottoriva (2017), Gerstung et al. (2020), and Tarabichi et al. (2021).

Types of timing inference enabled by the clock

1. Relative timing of clonal branching

By comparing the passenger mutation load on different branches of a phylogenetic-tree, the relative order and temporal separation of clonal branching events can be inferred. Mutations with cancer-cell-fraction (CCF) = 1.0 are clonal (truncal) — they occurred before the most recent common ancestor of all cancer cells and are therefore the earliest events. Mutations with CCF < 1.0 are subclonal — they occurred later, after population divergence, and the lower the CCF, the more recent the event (McGranahan & Swanton, 2017).

2. Timing of driver mutation acquisition

Gerstung et al. (2020) used the ratio of clonal to subclonal mutations to time driver acquisition across 2,658 cancers. Their key finding: many driver mutations precede clinical diagnosis by years or even decades. Early oncogenesis involves a constrained set of driver genes, while near-fourfold diversification occurs in later stages. The long latency revealed by the clock has implications for cancer-early-detection: the founding driver mutations in a tumor may be present and detectable long before the cancer becomes clinically apparent.

3. Timing of copy-number alterations

Chromosomal gains provide a particularly informative clock marker. Mutations that occurred before a chromosomal gain are present on both copies of the gained segment (clonal with respect to the gain event); mutations occurring after the gain are present on only one copy (subclonal). By counting the ratio of two-copy to one-copy mutations on gained segments, Gerstung et al. (2020) timed chromosomal gains and found they occur predominantly early in tumor evolution. This method also revealed that whole-genome-duplication is a common event that precedes extensive subclonal diversification.

4. Timing of mutational process activity

The mutational-signature composition of early (clonal) vs. late (subclonal) mutations reveals which DNA damage and repair processes were active at different evolutionary epochs. Nik-Zainal et al. (2012) demonstrated this in breast cancer: APOBEC-mutagenesis signatures (Signatures 2 and 13) appeared enriched in later subclones, while age-associated signatures (Signature 1) were prominent early. Gerstung et al. (2020) extended this finding: in ~40% of samples, the mutational spectrum changes significantly between early and late evolution, implying changing exposures, repair deficiencies, or mutagenic environments over the tumor’s lifetime.

Limitations and caveats

Constant-rate assumption

The clock assumes a stable neutral mutation rate. This is known to be violated in a substantial minority of cases: Gerstung et al. (2020) found that ~40% of samples show significant mutational spectrum shifts during evolution. When the spectrum changes, the overall rate of passenger mutation accumulation may also change, reducing the clock’s precision. For the ~60% without significant shifts, the constant-rate assumption holds as a reasonable approximation.

Detection limits create a blind zone

At standard sequencing depths (~100×), mutations with CCF below 0.05–0.10 are undetectable (Tarabichi et al., 2021). Because each doubling of the cancer cell population halves the frequency of newly arising mutations, mutations from the most recent ~7 population doublings are invisible at 100× depth, and those from the most recent ~10 doublings are invisible even at 1,000× depth (Turajlic et al., 2019). The clock cannot time the most recent evolutionary events — a blind zone that grows with tumor size.

Single-timepoint limitation

All molecular clock inferences in the current literature are cross-sectional: they reconstruct evolutionary history from a single bulk sample per patient. No source provides longitudinal multi-timepoint validation. The clock’s temporal inferences are model-dependent and rest on assumptions — constant rate, neutral passenger accumulation, correct CCF estimation — that have not been directly validated by serial sampling of the same tumor over time.

Model violations in aneuploid genomes

The copy-number-based timing method used by Gerstung et al. (2020) assumes that mutations before a chromosomal gain are present on both copies. In highly aneuploid genomes, this assumption may be violated, introducing uncertainty into CNA timing estimates.

Potential for selection to masquerade as neutrality

Graham & Sottoriva (2017) note that a specific combination of selected subclones could theoretically masquerade as a 1/f distribution under neutral-evolution. Conversely, clonal-interference and the weak average fitness advantage of drivers (~0.4% per Greaves & Maley, 2012) mean that some genuine subclonal selection may be missed by the 1/f test.

Clinical and evolutionary significance

The molecular clock transforms the cancer genome from a static catalog of mutations into a dynamic historical record. Its principal contributions are:

  • Long latency of drivers: Gerstung et al. (2020) demonstrated that many founding driver mutations precede diagnosis by years or decades, creating a window for cancer-early-detection that is far wider than previously appreciated.
  • Predicting progression: Subclonal genetic diversity, which the clock helps quantify, is “a robust biomarker for predicting progression to malignancy” (Greaves & Maley, 2012).
  • Therapeutic relevance: chromosomal-instability and whole-genome-duplication, which the clock can time relative to other events, are associated with worse clinical outcome (McGranahan & Swanton, 2017).
  • Understanding mutational process dynamics: The clock reveals which mutational processes were active at different times, providing insight into the temporal evolution of DNA damage, repair deficiency, and mutagen exposure.

Graph View

Backlinks