MUTYH cancer-associated variants within the interdomain connector differentially impact glycosylase activity and cellular DNA repair

This study demonstrates that cancer-associated MUTYH variants within the interdomain connector exhibit differential functional impacts, with some impairing lesion recognition and others disrupting downstream repair steps, highlighting the necessity of combining in vitro and cellular assays for accurate variant classification.

The Gist

The Big Picture: The Body's "Spellchecker"

Imagine your DNA is a massive library of instruction manuals for building and running your body. Every now and then, a typo happens. Specifically, a letter "G" gets oxidized (damaged) and looks like an "O" (8-oxoGuanine). When the cell tries to copy this page, it mistakenly pairs the damaged "O" with an "A" instead of a "C."

If this typo isn't fixed, it leads to a permanent mutation (a G to T transversion), which can eventually cause cancer.

Enter MUTYH, the body's specialized spellchecker. Its job is to find that wrong "A" paired with the damaged "O" and cut it out so the correct "C" can be put in its place.

The Problem: The "Connector" Mystery

The MUTYH spellchecker is a complex machine with two main parts:

1. The Cutter: The front end that snips out the wrong letter.
2. The Sensor: The back end that finds the damaged "O" on the page.

Connecting these two parts is a long, flexible arm called the Interdomain Connector (IDC). Think of this arm as a swivel chair or a bridge that lets the two parts work together. It also has a special "zinc linchpin" (a metal screw) that holds the bridge steady.

Scientists have found over 1,000 different typos (mutations) in the gene that builds this spellchecker. Many of these typos happen right on that "bridge" (the IDC). The big question was: Do these typos break the spellchecker, or is it just a cosmetic scratch?

The Experiment: Testing the Spellcheckers

The researchers took a group of these "bridge" mutations and tested them in two different ways:

1. The Lab Bench Test (In Vitro)

They built the spellcheckers in a test tube and watched them work on a single piece of DNA.

• The Result: Surprisingly, most of the "bridge" mutations worked perfectly fine. They found the damage, cut the wrong letter, and passed the job along just like the original, healthy spellchecker.
• The Analogy: It's like taking a broken car to a mechanic. The mechanic starts the engine, and it purrs perfectly. The mechanic says, "This car is fine!"

2. The City Traffic Test (Cellular Assay)

The researchers then put these spellcheckers inside living human cells (HEK293 cells). This is a much more chaotic environment, like a busy city street with traffic, pedestrians, and other cars.

• The Result: This is where the truth came out. Several of the "bridge" mutations that looked perfect in the test tube failed miserably in the living cells. They couldn't finish the repair job.
• The Analogy: When that same "perfect" car is driven in heavy city traffic, it stalls. The engine is fine, but the transmission or the suspension (the parts that handle the complex environment) is broken.

What Did They Discover?

The study revealed that the "bridge" (IDC) isn't just a passive connector; it's a communication hub.

• The "Handoff" Problem: In the cell, after the spellchecker cuts out the wrong letter, it has to hand the damaged spot off to a different worker (a protein called APE1) to finish the repair. The "bridge" is the handshake zone.
  • Some mutations (like V329M and C332R) broke the handshake. The spellchecker did its job but couldn't pass the baton to the next worker. The repair stalled, and the mutation remained.
• The "Metal Screw" Problem: One mutation (C332R) was supposed to break the zinc screw holding the bridge together. Surprisingly, the screw still held in the test tube, but the bridge wobbled in the cell, causing the machine to fail under pressure.
• The "Truncated" Mystery: One mutation (Q338X) was supposed to chop off the back half of the spellchecker. Theoretically, it should be useless. But the cell showed it still did some work (about 23% of normal). It's like a car with no trunk that can still drive to the grocery store, just not very well.

Why Does This Matter?

This research teaches us a vital lesson about cancer genetics: You can't judge a book by its cover (or a protein by a test tube).

• The Trap: If doctors only looked at the "Lab Bench" results, they would have said these mutations were harmless.
• The Reality: The "City Traffic" (cellular) test showed they are dangerous.

The study highlights that some mutations are subtle. They don't break the engine; they just make the car bad at driving in traffic. This explains why some people with these mutations get cancer later in life—their spellcheckers work fine when things are calm, but they fail when the body is under stress (like inflammation or aging).

The Takeaway

This paper is a detective story that solved a mystery: Why do some MUTYH mutations cause cancer even though the enzyme looks normal?

The answer is that the "bridge" is essential for the spellchecker to talk to its neighbors in the complex environment of a living cell. By using a new, more realistic test (the cellular assay), the researchers found that several mutations previously thought to be "safe" are actually dangerous because they disrupt the team effort required to fix DNA. This helps doctors better predict who is at risk for cancer and why.

Technical Summary

1. Problem Statement

• Context: Mutations in the MUTYH gene cause MUTYH-associated polyposis (MAP), a hereditary condition significantly increasing the risk of colorectal cancer. MUTYH is a base excision repair (BER) glycosylase responsible for removing adenine (A) mispaired with 8-oxo-7,8-dihydroguanine (OG), preventing G:C to T:A transversion mutations.
• The Gap: Over 1,000 germline and somatic MUTYH variants have been identified. While "founder" mutations (e.g., Y179C, G396D) are well-characterized, the functional impact of the majority of variants, particularly rare ones, remains unknown.
• Specific Focus: A significant number of these Variants of Uncertain Significance (VUS) are located in the Interdomain Connector (IDC). The IDC links the N-terminal catalytic domain and the C-terminal OG-recognition domain. In humans, the IDC contains a unique Zinc Linchpin Motif (CysX5CysX2Cys) and serves as a hub for protein-protein interactions (e.g., with APE1, HUS1, SIRT6).
• Challenge: Predicting the pathogenicity of IDC variants is difficult because they are often distal to the active site, and their disordered nature makes structural prediction challenging. Furthermore, standard in vitro assays often fail to detect subtle defects in these variants.

2. Methodology

The study employed a multi-tiered approach combining in vitro biochemical assays with a newly optimized in vivo cellular repair assay.

• Variant Selection: The authors selected several cancer-associated variants (CAVs) within the human MUTYH IDC: V329M, E331K, C332R, Q338H, Q338R, and Q338X. Corresponding mutations were engineered into the Mus musculus Mutyh protein (e.g., V329M $\rightarrow$ I297M) for purification and in vitro testing.
• In Vitro Biochemical Assays:
  • Metal Analysis: Inductively Coupled Plasma-Mass Spectrometry (ICP-MS) was used to quantify [4Fe-4S] cluster and Zn$^{2+}$ cofactor retention.
  • Glycosylase Activity: Single-turnover (STO) and multiple-turnover (MTO) assays measured the rate of adenine excision ($k_2$) and the active fraction of the enzyme.
  • Binding Affinity: Fluorescence polarization (FP) assays determined the apparent dissociation constant ($K_{1/2}$) for OG:A-containing DNA substrates.
  • APE1 Stimulation: Assays measured the rate of AP-site product release ($k_3$) in the absence and presence of AP endonuclease 1 (APE1) to test for defects in downstream repair handoff.
• Cellular Repair Assay:
  • A novel OG:A-specific repair reporter assay was developed in human HEK293FT cells (MUTYH knockout background).
  • Reporter Design: A plasmid containing an OG:A mismatch interrupts the GFP coding sequence. Successful repair restores the G:C pair, allowing GFP expression. dsRed serves as a transfection control.
  • Readout: Flow cytometry quantified the percentage of cells expressing both dsRed and GFP, normalized to wild-type (WT) levels.
  • Validation: The assay was benchmarked using known pathogenic variants (Y179C, G396D) and a catalytically inactive mutant (D236N).
  • Expression Analysis: Quantitative Western blotting correlated protein expression levels with repair efficiency across different clonal cell lines.

3. Key Results

A. In Vitro Biochemical Findings

• Cofactor Retention: Most IDC variants retained [4Fe-4S] clusters and Zn$^{2+}$ cofactors at levels comparable to WT, despite mutations in the Zn-coordinating motif (e.g., C332R).
• Glycosylase Activity & Binding:
  • Under single-turnover conditions, the catalytic rate ($k_2$) for all IDC variants was indistinguishable from WT.
  • Lesion binding affinity ($K_{1/2}$) was also similar to WT.
  • Contrast: This differs from founder mutations like Y179C, which showed severe defects in both activity and binding in vitro.
• APE1 Stimulation: Most variants showed APE1-stimulated product release similar to WT, though some (e.g., Q306H) showed slightly reduced enhancement. Notably, APE1 did not inhibit the IDC variants (unlike the Y150C variant), suggesting substrate engagement is maintained.

B. Cellular Repair Findings

• Divergent Phenotypes: Despite resembling WT in vitro, several IDC variants exhibited significantly reduced OG:A repair in mammalian cells:
  • Severe Defects: V329M (87% reduction), C332R (36% reduction), and Q338X (77% reduction) showed statistically significant repair deficits compared to WT (93%).
  • Mild/No Defects: E331K, Q338H, and Q338R showed repair levels comparable to WT.
• Expression Correlation:
  • For variants with moderate defects (e.g., Y179C, C332R), there was a positive correlation between protein expression levels and repair capacity. Higher expression could partially compensate for the functional defect.
  • For catalytically dead variants (D236N), repair remained low regardless of expression levels.
• Q338X Anomaly: The Q338X nonsense mutation (truncating the C-terminal domain) retained 23% repair activity, significantly higher than the MUTYH knockout background (5%), suggesting the truncated protein retains partial function or that read-through occurs.

4. Key Contributions

1. Discovery of "Silent" In Vitro Defects: The study demonstrates that mutations in the IDC can be functionally benign in isolated biochemical assays (glycosylase activity, binding) but pathogenic in a cellular context. This highlights the limitation of relying solely on in vitro kinetics for variant classification.
2. Mechanistic Insight: The data suggests that IDC variants likely disrupt steps downstream of base excision, such as the recruitment of APE1, interaction with the 9-1-1 complex (HUS1), or the handoff of the repair intermediate, rather than the initial lesion recognition or catalysis.
3. Optimized Cellular Assay: The authors validated a robust, high-throughput cellular reporter system capable of distinguishing between mild, moderate, and severe repair defects, and correlating them with protein expression levels.
4. Re-evaluation of VUS: The study provides functional evidence to reclassify specific VUS (e.g., V329M, C332R) as likely pathogenic due to their significant impact on cellular repair, while suggesting others (E331K, Q338H/R) may be benign or have context-dependent effects (e.g., under high oxidative stress).

5. Significance

• Clinical Impact: The findings underscore the necessity of using cellular functional assays alongside genetic data to accurately classify MUTYH variants. Relying only on in vitro data could lead to the misclassification of pathogenic variants as benign.
• Biological Understanding: The results reinforce the critical role of the IDC not just as a structural linker, but as a dynamic hub for protein-protein interactions essential for the full BER pathway.
• Future Directions: The study suggests that the phenotypic severity of these variants may depend on cellular conditions (e.g., ROS levels) and protein expression levels, offering a nuanced view of cancer risk that goes beyond simple "pathogenic/benign" binary classifications. This approach provides a framework for assessing the functional impact of rare variants in other DNA repair genes.