Molecular Dosimetry of DNA Adducts in Mice Exposed to Ethylene Oxide

This study establishes the first reliable quantitative dose-response relationships for mutagenic O6-HE-dG and non-mutagenic N7-HE-G DNA adducts across multiple tissues in mice exposed to a wide range of ethylene oxide concentrations, revealing distinct low-dose linearity and high-dose disproportionality that inform cancer risk assessment and genotoxic hazard characterization.

The Gist

The Big Picture: A Chemical "Tag" Game

Imagine Ethylene Oxide (EtO) as a tiny, hyper-active sticker that flies through the air. It's used in factories to make plastics and to sterilize medical equipment. Because it's so reactive, if you breathe it in, it doesn't just sit there; it tries to stick to everything it touches, including your DNA (the instruction manual inside your cells).

Scientists have long known that EtO is a carcinogen (cancer-causing), but they've been arguing about how dangerous it is at low levels.

• The Debate: One group of regulators (the EPA) thinks the danger starts immediately and is very steep, even at tiny amounts. Another group (TCEQ) thinks the danger is more like a gentle, steady slope that only gets steep when you are exposed to huge amounts.
• The Goal: This study wanted to settle the argument by looking at the actual "stickers" (DNA damage) left behind in mice, rather than just guessing based on statistics.

The Experiment: The Mouse "Sticker" Hunt

The researchers took mice and exposed them to different levels of this chemical gas for 28 days. They tested a huge range:

• Tiny amounts: Like the air in a normal city (0.05 to 1 part per million).
• Huge amounts: Like a factory accident (50 to 200 parts per million).

Afterward, they took samples from the mice's lungs, liver, bone marrow, and even their mammary glands (breast tissue) to see what kind of "stickers" were left behind. They looked for two specific types of damage:

1. The "N7-HE-G" Sticker: The Abundant, Harmless Graffiti

Think of this as graffiti on the side of a house.

• What it is: It forms very easily and in huge numbers.
• Is it dangerous? Not really. It's like a sticker on a wall; it doesn't change the structure of the house or how the house functions. The body can easily wash it away or ignore it.
• What the study found: Even at the tiniest levels of exposure, this "graffiti" appeared. It increased in a straight, predictable line as the gas got stronger. This tells us the chemical reaches the cells easily, but it doesn't necessarily mean it causes cancer.

2. The "O6-HE-dG" Sticker: The Dangerous "Glitch"

Think of this as a glitch in the computer code.

• What it is: It forms much less often, but it's very dangerous. If this sticker stays on the DNA, it confuses the cell's machinery. When the cell tries to copy its DNA, it reads the wrong instruction, leading to a mutation (a typo in the genetic code). This is the "spark" that can start cancer.
• Is it dangerous? Yes, very.
• What the study found: This is the big discovery. At low levels (what normal people breathe), this dangerous glitch was completely invisible. The body's repair crew (like a team of IT specialists) fixed it instantly or it never formed in the first place.
• The Threshold: The dangerous glitch only appeared when the mice were exposed to very high levels of the gas (50 ppm and up). Even then, the damage didn't shoot up in a scary, vertical cliff; it rose in a curve that got steeper only when the system was overwhelmed.

The "Blood Test" Connection

The researchers also checked the mice's blood. They found that a marker in the blood (called HE-V) acted like a perfect thermometer for how much "graffiti" (N7-HE-G) was on the DNA.

• Why this matters: You can't easily take a biopsy of a person's lung or breast tissue. But you can take a blood test. This study proved that if you check the blood, you can accurately guess what's happening to the DNA in the organs.

The "Mammary Gland" Surprise

One interesting finding was about the mammary gland (breast tissue). Even though it's far away from the nose (where the gas enters), it got almost as much "damage" as the lungs did at high doses.

• The Analogy: Imagine a house where the front door (lungs) gets the most rain, but the back bedroom (breast tissue) gets almost as wet because the whole house is soaked. This supports the idea that EtO can cause breast cancer, not just lung cancer, but only when the exposure is high enough to soak the whole house.

The Verdict: What Does This Mean for Humans?

This study provides a "molecular reality check" for cancer risk models.

1. The "Steep Cliff" Theory is Wrong: The data does not support the idea that there is a massive, immediate spike in cancer risk at very low, everyday exposure levels. If the dangerous "glitch" (O6-HE-dG) isn't even forming at low levels, the risk of cancer shouldn't be sky-high either.
2. The "Gentle Slope" Theory is Better: The data supports a model where the risk increases slowly and steadily as exposure increases, only becoming significant when exposure gets very high.
3. The Bottom Line: The body has repair mechanisms that handle low-level exposure well. The danger really kicks in when you overwhelm those defenses with high doses.

In simple terms: Breathing a tiny bit of this chemical is like getting a few raindrops on your umbrella; your body handles it easily. Breathing a lot of it is like standing under a firehose; eventually, the umbrella breaks, and you get soaked. The study shows that the "firehose" effect only happens at high levels, not at the tiny "raindrop" levels we encounter in daily life.

Technical Summary

1. Problem Statement

Ethylene oxide (EtO) is a widely used industrial chemical and a known human carcinogen. Its mutagenic mode of action (MOA) is attributed to DNA alkylation, forming two primary adducts:

• N7-(2-hydroxyethyl)guanine (N7-HE-G): The most abundant adduct (~95%), generally considered non-mutagenic but chemically unstable.
• O6-(2-hydroxyethyl)-2'-deoxyguanosine (O6-HE-dG): A minor adduct but highly promutagenic, capable of causing G:C→A:T transition mutations.

The Knowledge Gap: While EtO is classified as a carcinogen, the precise dose-response relationships of these DNA adducts at low inhalation exposures (relevant to the general population, <3 ppm) remain unknown. Existing cancer risk assessments by agencies like the EPA and TCEQ rely on statistical models with vastly different assumptions (e.g., a steep linear slope at low doses vs. a shallow single-slope linear model). Without empirical molecular data at low doses, it is difficult to determine which statistical model is biologically plausible. Previous studies lacked the sensitivity to detect O6-HE-dG at low concentrations and did not comprehensively map dose-responses across multiple tissues (including tumor targets like the mammary gland and bone marrow) at environmentally relevant levels.

2. Methodology

The study employed a rigorous experimental design to quantify DNA adducts across a wide exposure range using highly sensitive mass spectrometry.

• Experimental Subjects: Male and female B6C3F1 mice (the strain used in the NTP EtO cancer bioassay), aged 9–12 weeks.
• Exposure Protocol:
  • Route: Whole-body inhalation.
  • Duration: 6 hours/day for 28 consecutive days (selected to reach steady-state adduct levels).
  • Doses: 8 groups: 0 (air control), 0.05, 0.1, 0.5, 1, 50, 100, and 200 ppm.
  • Phasing: Due to technical issues with low-dose generation in Phase 1, a Phase 2 was conducted for low doses (0.05–0.1 ppm) and a high-dose control (50 ppm) to ensure accuracy. Data from both phases were integrated after statistical validation.
• Tissue Collection: Immediately post-exposure, DNA was extracted from lung, liver, bone marrow, and mammary gland. Blood was also collected for hemoglobin adduct analysis (HE-V).
• Analytical Techniques:
  • Adduct Release: N7-HE-G was released via neutral thermal hydrolysis; O6-HE-dG was released via enzymatic digestion (DNase I, alkaline phosphatase, phosphodiesterase).
  • Purification: Samples underwent filtration (3 kDa MWCO) and HPLC fractionation to remove matrix interferences.
  • Quantification:
    • N7-HE-G: Measured using nanoLC-ESI-MS/MS (Q Exactive HF) for low doses and LC-MS/MS (TSQ Quantis) for high doses.
    • O6-HE-dG: Measured using nanoLC-ESI-MS/MS (Q Exactive HF).
    • Sensitivity: The method utilized stable isotope-labeled internal standards (N7-HE-G-d4 and O6-HE-dG-d4). The limit of detection (LOD) for O6-HE-dG was 3 amol (equivalent to <1 adduct per cell), representing a significant improvement over previous methods.

3. Key Contributions

• First Low-Dose Quantification: This is the first study to reliably quantify N7-HE-G adducts at exposures as low as 0.05 ppm (below the occupational limit of 1 ppm).
• Unprecedented Sensitivity: The study achieved the lowest reported LOD for O6-HE-dG (3 amol), allowing for the definitive determination that this promutagenic adduct is not detectable at low environmental exposures (<50 ppm).
• Multi-Tissue Dose-Response: Provided comprehensive dose-response data for both adducts across four distinct tissues, including the mammary gland (a tumor target), revealing systemic distribution patterns.
• Biomarker Correlation: Established a strong linear correlation between blood hemoglobin adducts (HE-V) and tissue DNA adducts (N7-HE-G), validating HE-V as a surrogate for internal tissue dose.

4. Key Results

• N7-HE-G (Non-mutagenic Adduct):
  • Low Dose (≤1 ppm): Exhibited a linear dose-response relationship in all tissues (lung, liver, bone marrow, mammary gland). Significant increases were detected even at 0.05 ppm.
  • High Dose (≥50 ppm): Showed a sharp, dose-disproportionate (non-linear) increase, particularly between 100 and 200 ppm.
  • Tissue Distribution: Lung had the highest levels, followed by mammary gland, liver, and bone marrow. Mammary gland levels were notably high, comparable to lung at 200 ppm.
• O6-HE-dG (Promutagenic Adduct):
  • Detection Threshold: Not detected in any tissue at exposures <50 ppm, despite the method's ability to detect <1 adduct per cell.
  • High Dose (≥50 ppm): Became quantifiable starting at 50 ppm. The dose-response pattern mirrored N7-HE-G, showing a steep, non-linear increase at high doses.
  • Ratio: The ratio of N7-HE-G to O6-HE-dG was approximately 200–800:1, varying by tissue (highest in liver, likely due to efficient MGMT repair).
• Correlation: Blood HE-V levels showed a strong positive linear correlation ($p < 0.001$) with tissue N7-HE-G levels across all exposure groups and tissues.

5. Significance and Implications

• Biological Plausibility for Risk Assessment: The data challenges the EPA's current cancer risk model, which assumes a very steep linear slope at low doses. Instead, the molecular data supports a shallow, single-slope linear dose-response at low exposures (≤1 ppm) for the mutagenic event (O6-HE-dG formation), as no promutagenic adducts were formed below 50 ppm.
• Mode of Action (MOA) Insights: The findings suggest that at low environmental doses, EtO exposure results in DNA adduct formation (N7-HE-G) but likely undergoes efficient repair or does not reach the threshold for promutagenic O6-HE-dG accumulation. The "upward bend" in dose-response occurs only at high doses (≥50 ppm), likely due to the saturation of detoxification pathways (glutathione depletion) and repair mechanisms.
• Regulatory Impact: The results provide empirical evidence to support the TCEQ's risk assessment model (shallow linear slope) over the EPA's model (steep initial slope), potentially reducing estimated cancer risks for the general population by orders of magnitude.
• Biomonitoring Utility: The strong correlation between HE-V and tissue DNA damage confirms that blood-based biomonitoring can effectively estimate internal tissue doses, facilitating retrospective exposure assessment in epidemiological studies where tissue samples are unavailable.

In conclusion, this study provides the first high-resolution molecular dosimetry of EtO, demonstrating that the formation of the critical mutagenic lesion (O6-HE-dG) is restricted to high-dose exposures, thereby informing a more biologically plausible, conservative, and less steep dose-response model for human cancer risk assessment.