Structural basis for continuous DNA-end protection during ligation of double-strand breaks in yeast Non-Homologous End-Joining

This study utilizes cryo-electron microscopy to reveal the structural mechanisms of DNA-PKcs-independent Non-Homologous End-Joining in yeast, demonstrating how Dnl4-mediated synaptic complexes achieve continuous DNA-end protection and strand-specific ligation on microhomology substrates while explaining the inefficiency of blunt-end joining through a non-aligned, ligation-incompatible configuration.

The Gist

The Big Picture: Fixing a Broken Book

Imagine your DNA is a massive library of books. Sometimes, a page gets ripped right in half (a "double-strand break"). If you don't fix it, the story is lost, and the cell might die.

Cells have a repair crew called NHEJ (Non-Homologous End Joining). Think of this crew as a team of emergency glue technicians. Their job is to grab the two torn edges of the book, line them up perfectly, and glue them back together.

In humans (and other vertebrates), this crew has a very heavy-duty foreman named DNA-PKcs. This foreman holds the pages steady, checks the alignment, and calls in extra tools if the tear is messy.

The Mystery: Yeast (a tiny fungus) doesn't have this heavy-duty foreman. They have a smaller, simpler crew. Scientists have always wondered: How does the yeast crew manage to fix broken DNA without the big boss holding everything together? Do they just guess, or is there a secret trick?

This paper uses a super-powerful microscope (Cryo-EM) to take 3D snapshots of the yeast repair crew in action. They discovered that the yeast crew has a clever, two-step "dance" to keep the broken ends safe and glued together.

---

The Cast of Characters

Before we get to the action, let's meet the yeast repair team:

• Ku (The Anchors): Two proteins that slide onto the broken DNA ends like a ring on a finger. They hold the line.
• Dnl4 (The Glue Gun): The actual enzyme that does the gluing. Interestingly, the yeast team brings two of these glue guns.
• Lif1 & Nej1 (The Scaffold): These are the structural beams that hold the two glue guns in place, forming a bridge over the break.

---

The Discovery: Three Different "Dances"

The researchers tested the yeast crew on three different types of "tears" in the DNA and found three different ways they handle the job.

1. The Perfect Match (Microhomology)

The Scenario: The torn edges have a tiny bit of matching pattern (like two puzzle pieces that almost fit).
The Action:
The yeast crew snaps into a "Ligation-Competent" pose.

• One glue gun (Dnl4) locks onto the first tear and gets ready to glue.
• The second glue gun hangs out nearby, waiting its turn.
• The Analogy: Imagine two people trying to tape a ripped piece of paper. One person holds the tape and applies it to the top half. The second person stands right next to them, ready to tape the bottom half immediately after. They are perfectly aligned, and the job gets done quickly.

2. The "Alternating" Dance (Two 5' Phosphates)

The Scenario: The DNA ends are ready to be glued, but both sides need work.
The Action:
This is the paper's biggest discovery. The yeast crew uses a "Hand-Off" strategy.

• Both glue guns grab the DNA at the same time, acting like a protective cage.
• However, they can't both glue at once because they would bump into each other.
• The Analogy: Think of two people trying to close a heavy door that has two latches. They can't both push the latches at the exact same time without hitting each other's hands. So, Person A pushes the top latch, then steps back slightly. Person B then pushes the bottom latch. They take turns, but they never let go of the door completely. They keep the door "protected" the whole time, just shifting their grip to get the job done.
• Why this matters: This explains how yeast fixes DNA without the heavy-duty foreman (DNA-PKcs). They use a "protective cage" made of two glue guns that take turns working, ensuring the broken ends never drift apart and get lost.

3. The "Stuck" State (Blunt Ends)

The Scenario: The DNA is torn straight across with no matching patterns (a "blunt" cut). This is the hardest type of break to fix.
The Action:
The yeast crew grabs the ends, but they get stuck in a weird position.

• Both glue guns lock onto the DNA, but because the ends don't match, the glue guns are forced to stand far apart (about 30 Angstroms, which is huge in molecular terms).
• The Analogy: Imagine trying to tape two pieces of wood together, but the wood is so thick that the two people holding the tape can't reach each other. They are holding the wood, but the pieces are too far apart to actually stick.
• The Result: The yeast crew is "protecting" the ends (keeping them from wandering off), but they can't glue them yet. They have to wait for a different tool (a polymerase or nuclease) to come in, shave off a little bit of wood, or add a little bit of new wood to create a "matching pattern" (microhomology). Once that happens, the crew can switch to the "Alternating Dance" and finish the job.
• Why it's slow: This explains why yeast is bad at fixing blunt ends. They get stuck in this "waiting room" mode, unable to glue until the DNA is modified.

---

The "Aha!" Moment: Why Yeast is Different from Humans

In humans, the heavy-duty foreman (DNA-PKcs) holds the DNA ends in a specific shape that allows the glue guns to work efficiently, even on messy breaks.

In yeast, they don't have that foreman. Instead, they evolved a self-contained safety net.

• They use two glue guns working together as a single unit.
• They act as a "protective shield," keeping the broken ends safe from the environment while they figure out how to align them.
• They take turns working (alternating engagement) so they don't crash into each other.

The Takeaway

This paper solves a mystery that has puzzled scientists for years. It shows that even without the "big boss" (DNA-PKcs), nature has found a brilliant, mechanical solution: The yeast repair crew acts like a synchronized dance team. They lock onto the broken DNA, protect it with a dual-grip, and take turns gluing the strands back together.

However, if the break is too messy (blunt ends), their dance floor is too crowded, and they get stuck waiting for the DNA to be "prepped" before they can finish the job. This structural insight helps us understand why some DNA breaks are easy to fix and others are a nightmare, and it reveals the elegant, primitive engineering of life's repair systems.

Technical Summary

1. Problem Statement

Non-Homologous End Joining (NHEJ) is a primary pathway for repairing DNA double-strand breaks (DSBs). In vertebrates, this process relies heavily on the DNA-dependent protein kinase catalytic subunit (DNA-PKcs) to align DNA ends, recruit processing enzymes, and facilitate ligation by DNA Ligase IV (Lig4). However, many eukaryotes, including the budding yeast Saccharomyces cerevisiae, lack DNA-PKcs and the accessory factor PAXX. Despite this, yeast NHEJ (yNHEJ) functions efficiently for cohesive ends but is notably inefficient and slow when repairing blunt-ended DSBs. The structural and mechanistic basis for how a DNA-PKcs-independent system maintains DNA end protection, aligns ends, and performs sequential ligation without the regulatory scaffold provided by DNA-PKcs remained unclear.

2. Methodology

The authors employed a multi-disciplinary approach combining structural biology, biochemistry, and genetics:

• Reconstitution: They purified core yNHEJ components from S. cerevisiae: the Ku heterodimer (Yku70/Yku80), the DNA ligase IV complex (Dnl4-Lif1), and the XLF homologue (Nej1).
• Substrate Design: They designed synthetic DNA substrates to trap specific intermediates:
  • Cohesive ends: 4-bp 3' overhangs with either a single 5'-phosphate (to isolate the first ligation event) or two 5'-phosphates.
  • Blunt ends: Substrates lacking microhomology with two 5'-phosphates.
  • Modifications: Use of dideoxy nucleotides (ddC/ddG) to create non-ligatable nicks, preventing full ligation and allowing the capture of pre- or post-ligation states.
• Complex Stabilization: The fragile synaptic complexes were stabilized using GraFix (gradient fixation with glutaraldehyde) prior to vitrification.
• Cryo-Electron Microscopy (Cryo-EM): High-resolution structures were determined for the reconstituted complexes on various DNA substrates, achieving resolutions between 3.3 Å and 5.8 Å.
• Validation: In vitro ligation assays and in vivo yeast survival assays (using HO endonuclease-induced DSBs) were performed to validate structural findings, including the testing of specific point mutations in protein-protein interfaces.

3. Key Contributions and Results

A. Architecture of the Ligation-Competent Complex

• Overall Structure: The yeast synaptic complex resembles the human short-range (SR) complex, forming a "W-shaped" architecture. It consists of two Ku heterodimers flanking the break, bridged by a central Lif1-Nej1 scaffold.
• Ku Conformation: Unlike the human complex where XLF binding opens the Ku80 vWA domain, yeast Nej1 lacks the canonical Ku-binding motif (KBM). Instead, the authors identified a conserved Yku80 internal loop (residues 505–525) that occupies the KBM pocket, stabilizing the Ku80 vWA domain in an open conformation. Mutational analysis confirmed this loop is essential for NHEJ efficiency in vivo.
• Rigid Scaffold: The Lif1-Nej1 scaffold in yeast is straighter and more rigid than in humans, imposing a distinct DNA trajectory (135° bend vs. ~146° in humans) and a larger angle (85°) between the scaffold and DNA.

B. Mechanism of Sequential Ligation (Alternating Engagement)

Using substrates with two 5'-phosphates, the authors resolved two predominant, coexisting "DNA-aligned protective" states:

• State 1a & 1b: In these states, both Dnl4 catalytic cores (DBD-NTD modules) are engaged with the DNA ends simultaneously but in reciprocal geometries.
  • One Dnl4 molecule is positioned for catalysis (pre- or post-ligation), while the second is engaged but displaced or in a "waiting" mode.
  • The transition between these states involves a coordinated rotation and translation of the DNA duplex (~39° rotation, ~3.5 Å translation) to present the second nick to the active site.
• Mechanism: This supports an alternating engagement model. The two Dnl4 molecules alternately engage the DNA to seal the two strands sequentially. Crucially, the second catalytic core remains bound to the DNA even when not catalyzing, ensuring continuous protection of the DNA ends. This compensates for the lack of DNA-PKcs, which in vertebrates provides this protective tethering.

C. Structural Basis for Blunt-End Inefficiency

• Non-Aligned Protective State: On blunt-ended substrates, the complex adopts a distinct "non-aligned" configuration. Both Dnl4 catalytic modules bind the 5'-phosphates symmetrically but dimerize via a specific yeast-specific insertion (Insert2).
• Steric Hindrance: This dimerization forces the two DNA ends to remain separated by ~30 Å, a distance incompatible with ligation.
• Implication: Productive ligation of blunt ends requires a massive rearrangement where one Dnl4 module must disengage to allow the DNA ends to align. This high-energy barrier explains the slow kinetics and low efficiency of blunt-end repair in yeast, suggesting that end-processing enzymes (nucleases/polymerases) are required to create microhomologies to transition into the aligned state.

D. Evolutionary Insights

• The study highlights specific insertions in Dnl4 (Insert1, Insert2, Insert3) that differ between yeast and vertebrates.
• The yeast-specific Insert2 mediates the dimerization of the catalytic cores in the blunt-end state, a feature absent in vertebrate Lig4.
• The authors propose that the primitive yeast system relies on a stable, dual-engaged protective state to ensure repair completion, whereas vertebrates may delegate the second ligation step to alternative pathways (e.g., Ligase III/BER) or rely on DNA-PKcs for dynamic regulation.

4. Significance

• Mechanistic Clarity: This work provides the first high-resolution structural view of a DNA-PKcs-independent NHEJ machinery, revealing how it achieves end protection and sequential ligation without the kinase scaffold.
• Protection vs. Catalysis: It establishes a model where the "protective" state (keeping ends bound) and the "ligation-competent" state are distinct but interconvertible, with the dual-engagement of Dnl4 acting as a safeguard against end loss.
• Explaining Phenotypes: The structures provide a direct structural explanation for the well-known biological observation that yeast NHEJ is inefficient at repairing blunt ends, linking the inefficiency to a specific, stable, non-productive synaptic conformation.
• Therapeutic Potential: The identification of a stable dimeric intermediate in yeast (and potentially transient in humans) offers a potential new target for drug design; trapping this dimer could block NHEJ, sensitizing cells to DNA damage.

In summary, the paper elucidates how the yeast NHEJ machinery utilizes a rigid scaffold and an alternating, dual-engagement mechanism of the ligase to protect DNA ends and perform repair in the absence of DNA-PKcs, while simultaneously explaining the structural constraints that limit its ability to repair blunt-ended breaks.