Freeze-Thaw Cycles: How to Shatter a Molecule

In veterinary research and diagnostics, we often freeze serum or plasma to batch samples for later testing. For chemistry panels—cortisol, insulin, thyroid hormones—this freeze-store-thaw workflow is usually perfectly acceptable. The analytes are stable.

For cfDNA, and especially for advanced analyses like Fragmentomics, repeated freeze-thaw cycles are a disaster. They don't just degrade the sample—they actively alter the DNA in ways that can produce misleading results.

Understanding the physics of what happens when plasma freezes, and why DNA is particularly vulnerable, is essential for anyone working with liquid biopsy samples.

The Physics of Freezing: Ice as a Weapon

Water Expansion

When plasma freezes, the water content crystallizes into ice. Unlike most substances that contract when they solidify, water expands by approximately 9% as it freezes. This expansion is why ice floats and why frozen pipes burst.

In a plasma sample, this expansion creates mechanical stress throughout the frozen matrix.

Ice Crystal Formation

The real damage comes from how ice crystals form:

Slow Freezing (-20°C freezer):
- Ice crystals grow slowly and can become large
- Larger crystals cause more mechanical damage
- Crystals form preferentially in the water-rich regions, excluding solutes
- Creates concentrated pockets of protein and DNA surrounded by ice walls

Fast Freezing (-80°C freezer or flash-freezing):
- Ice crystals form rapidly and remain small
- Less mechanical damage
- More uniform distribution
- Preferred for cfDNA storage

The Ice Crystal Problem

These ice crystals act like microscopic blades. As they form and grow, they physically cut through molecular structures in their path. They also create shear forces as the expanding ice pushes against surrounding material.

Why DNA is Vulnerable

The Long, Fragile Molecule

DNA is a long, strand-like molecule that is particularly susceptible to mechanical damage:

- Stretched structure: DNA exists as an extended double helix
- Vulnerable backbone: The sugar-phosphate backbone can be broken by physical stress
- Size matters: Longer fragments are more susceptible than shorter ones (more surface area to encounter damage)

cfDNA Fragment Sizes

Circulating cfDNA is already fragmented, but there's a range of sizes:

- Mono-nucleosomes: ~167 base pairs (most common in healthy cfDNA)
- Di-nucleosomes: ~334 base pairs
- Tri-nucleosomes: ~500 base pairs
- Longer fragments: Present in necrotic cell death, can be thousands of base pairs

The longer fragments are preferentially damaged during freeze-thaw cycles because they have more opportunities to encounter ice crystal damage.

What Happens During Freeze-Thaw

Freezing:
- Ice crystals form and expand
- Long DNA strands get caught between growing crystals
- Physical shearing occurs at the crystal boundaries
- Some long fragments are broken into shorter pieces

Thawing:
- Ice crystals melt
- Previously trapped DNA is released
- Some fragments that were stressed but not broken during freezing may break during thaw
- Dissolved nucleases (if any were present) become active again

Refreezing:
- The whole process repeats
- Already-stressed fragments break
- More long fragments become short fragments
- Cumulative damage with each cycle

Impact on Diagnostic Tests

1. Fragmentomics: Artificial Short Fragments

Advanced cancer detection increasingly relies on fragment size analysis (fragmentomics). The principle is that tumor-derived DNA tends to be shorter than normal cfDNA.

The Test: Calculate a DNA Integrity Index (DII)—the ratio of long fragments to short fragments.

Normal Interpretation:
- High DII (more long fragments) = Normal pattern
- Low DII (more short fragments) = Suspicious for cancer

The Freeze-Thaw Problem:

If you freeze-thaw a perfectly healthy sample three times:
- Long fragments get sheared into shorter pieces
- The DII drops artificially
- The sample now "looks" like it has a cancer-associated fragment pattern

Result: False Positive for fragmentation-based cancer suspicion.

The clinician initiates an expensive workup looking for a cancer that doesn't exist, all because of improper sample handling.

2. Total DNA Concentration: Debris Lysis

Even in properly prepared "cell-free" plasma, there are often residual components:

- Platelets: If double-spin wasn't performed or wasn't complete
- Microvesicles: Small membrane-bound particles
- Apoptotic bodies: Remnants of dying cells

Freezing ruptures these residual structures:

- Platelets release mitochondrial DNA
- Microvesicles release their DNA cargo
- Apoptotic bodies release trapped DNA fragments

Result: A spike in total cfDNA concentration upon thawing that wasn't present in the original fresh sample.

3. Mutation Detection: Signal Dilution

For mutation-based cancer detection (looking for specific tumor mutations):

- Freeze-thaw releases additional normal DNA from debris
- This dilutes the tumor mutation signal
- If the original tumor fraction was borderline, it may drop below detection

Result: Potential false negative for mutation detection.

The "One Freeze, One Thaw" Rule

Best practice for cfDNA is simple and absolute:

One Freeze. One Thaw. No exceptions.

The Workflow

1. Process plasma: Perform double-spin protocol immediately after collection
2. Aliquot before freezing: Divide plasma into appropriate volumes for expected tests (typically 0.5-1.0 mL per aliquot)
3. Freeze once: Place aliquots immediately in -80°C freezer (preferred) or -20°C
4. Thaw when ready: Remove only the aliquot(s) needed for the specific test
5. Never refreeze: Any thawed sample should be analyzed immediately or discarded

Why Aliquoting Is Essential

Aliquoting takes an extra 2-3 minutes during processing but saves you from impossible situations later:

Without aliquoting:
- You have one large tube of frozen plasma
- You need to run Test A
- You thaw the entire tube
- Test A fails (technical issue) and needs to be repeated
- You now must refreeze the sample (damaging it) or waste the remaining plasma

With aliquoting:
- You have four 0.5 mL aliquots of frozen plasma
- You need to run Test A
- You thaw Aliquot #1
- Test A fails and needs to be repeated
- You pull Aliquot #2 from the freezer—fresh, never-thawed
- Problem solved, no quality compromise

Recommended Aliquot Strategy

| Scenario | Recommended Aliquots | Volume Each |
|----------|---------------------|-------------|
| Single test expected | 2 aliquots | 0.5-1.0 mL |
| Multiple tests likely | 3-4 aliquots | 0.5 mL |
| Research/longitudinal study | 4-6 aliquots | 0.3-0.5 mL |

Label each aliquot clearly with patient ID, date, and aliquot number.

Storage Temperature Considerations

-80°C (Preferred)

- Smallest ice crystals
- Most stable long-term storage
- Industry standard for cfDNA biobanking
- Samples stable for years

-20°C (Acceptable)

- Larger ice crystals form
- More thermal fluctuation in frost-free freezers
- Suitable for short-term storage (weeks to months)
- Not ideal for long-term biobanking

4°C (Refrigerator)

- Not appropriate for storage beyond 24 hours
- DNA degradation by residual nucleases continues
- Only for very short-term holding before freezing

Avoid Frost-Free Freezers for Long-Term Storage

Frost-free freezers maintain low frost by cycling through warming periods. These temperature fluctuations can cause partial thaw-refreeze cycles that damage samples over time, even if they never fully thaw.

Troubleshooting: Signs of Freeze-Thaw Damage

If you suspect a sample has undergone multiple freeze-thaw cycles:

Warning Signs:
- Documentation gaps (unclear storage history)
- Sample tube shows frost or ice crystals inside
- Previous aliquots from same batch showed unexpected results
- Results don't match clinical picture

What to Do:
- Note the concern in documentation
- Consider redrawing if the result is critical
- Interpret results with appropriate caution

Summary: Protecting DNA Through Proper Handling

| Practice | Correct | Incorrect |
|----------|---------|----------|
| Freezing temperature | -80°C preferred | Room temperature thaw, back to freezer |
| Number of freeze-thaw cycles | ONE maximum | Multiple cycles |
| Aliquoting | Before first freeze | After thawing |
| Thawed sample handling | Analyze immediately | Refreeze "for later" |
| Documentation | Record all freeze/thaw events | "It's probably fine" |

The bottom line: Never put a thawed cfDNA sample back in the freezer. Each freeze-thaw cycle shatters DNA molecules and degrades your diagnostic accuracy. Aliquot before freezing, and respect the one-thaw rule.