The art and science of freezing bacteria
10 December 2025
When I started working with bacterial cultures at Nottingham University, frozen stock preparation seemed straightforward enough. We mixed bacterial culture with glycerol, popped it in the freezer, and called it a day. It seemed simple, so I never thought much about it, especially as we always could retrieve frozen strais without difficulty. However, as it turns out, there is a surprising amount of nuance hiding in those innocuous little vials.
The Nottingham method
Our setup at Nottingham was delightfully lo-fi: 7 ml glass urine sample vials filled with a 2:1 mixture of bacterial culture to 80% glycerol. This gave us a final glycerol concentration of just over 25%, and we stored everything at -20°C. Very quick, and very convenient. But the problem with 25% glycerol is that it freezes solid at -20°C. Anyone who's tried to scrape frozen culture out of a vial at 9 AM on a Monday knows this is suboptimal. So, after talking to some colleagues and collaborators, I tweaked the protocol: equal parts culture and 80% glycerol, bringing the final concentration to 40%. At this concentration, the stocks stayed liquid at -20°C, or at least slushy enough to pipette directly. No waiting, no thawing, just grab your stock from a metal ice block, streak a plate, and you're done. The tweak seemed beneficial, because it reduced handling time in the relative warmth of the lab.
What I didn't appreciate at the time was how well this system worked, not just for convenience, but for long-term viability. Those large glass vials kept cultures alive for years. I had stocks that remained viable for my entire time at Nottingham, and beyond.
The Brunel reality check
Then I moved to Brunel University and everything changed. The glass urine sample vials were discontinued, so we switched to standard 1.8 ml cryo vials. Same protocol, different tubes. What could go wrong?
Everything, as it turns out.
First, we noticed that bacteria in the non-frozen 40% glycerol stocks would slowly settle to the bottom of the tubes. Taking a small aliquot from the top would not result in any growth on a plate, simply because all the cells had settled at the bottom of the tube. Not a disaster: just vortex thoroughly before use. But annoying.
But then we started to see a lack of growth after vortexing the culture, clearly highlighting a drop in viability. Stocks that should have lasted years were dying within months. DH5α strains carrying plasmids sometimes couldn't be revived after just 12 months. Next were strains that seem more susceptible due to some of the alleles or gene deletions they are carrying. Some of these were dying noticeably faster than wild-type strains.
What was going on? I still don't have a complete answer, but I suspect the larger volume in those glass vials mattered more than I realized. More volume means more cells, yes, but perhaps there's also something about the physical dynamics of freezing in different container geometries, or the glass material itself providing better insulation. It's one of those empirical observations that worked beautifully in practice but remains frustratingly unclear in theory.
The solution was pragmatic rather than elegant: we now maintain stocks at -20°C for daily use and keep backup copies at -80°C for long-term storage. This two-tier system works well, but it only works now that we have reliable -80°C storage facilities - for years -80°C storage was notoriously unreliable in the Division, and multiple sets of samples from colleagues got lost when -80°C freezers broke down.
What does the literature actually say?
Here's the slightly embarrassing part: despite years of hands-on experience, I'd never actually looked up what the scientific literature had to say about optimal bacterial storage conditions. All my knowledge came from lab folklore, troubleshooting sessions and conversations with colleagues and collaborators. Unsurprisingly, a lot is known about stocks.
The Glycerol Standard
The literature confirms that 15-20% glycerol is the sweet spot for most applications. The typical protocol is simple: mix equal volumes of bacterial culture with 50% glycerol, giving a final concentration of 25% - exactly the concentration we started with at Nottingham. This concentration provides excellent cryoprotection while still allowing the cultures to freeze, which is actually desirable at proper storage temperatures.
My 40% glycerol innovation? Turns out it's outside the typical range, though not unheard of. There are studies that mention frozen stocks made with 40% glycerol, so I wasn't completely off the reservation. The advantage of staying liquid at -20°C is real, but it comes with trade-offs in long-term viability.
Temperature matters (a lot)
This is where the literature got uncomfortably specific:
-20°C storage: Good for about one year maximum
-80°C storage: Reliable for between 3 and 10 years
Liquid nitrogen (-196°C): The gold standard, maintaining viability for up to 30 years
So my long-term reliance on -20°C storage in small cryo tubes at Brunel was essentially playing Russian roulette with my stock collection. The fact that it worked at Nottingham (possibly due to those magical glass vials) was the exception, not the rule.
The cryoprotectant landscape
Glycerol dominates the field for good reason: it's effective, cheap, and works with virtually all bacterial species. But there are alternatives.
DMSO (5-10%) is effective but has drawbacks beyond its toxicity for in vivo work. As I can attest from experience, DMSO-preserved cultures freeze rock-solid, making it nearly impossible to extract small volumes for fresh cultures without completely thawing the stock, defeating much of the convenience factor.
Trehalose, sucrose, and skim milk also appear in the literature as cryoprotectants, each with specific applications. For instance, 12% sucrose works particularly well for mycoplasma strains, while 10% skim milk is effective for storing cultures at -20°C. These alternatives might be worth exploring for problematic strains, though they lack glycerol's universal applicability.
Intriguingly, some researchers combine protectants: glycerol with sucrose and inulin, for example, dramatically improved the viability of strict anaerobic gut microbes. This suggests that for difficult-to-preserve strains, cocktail approaches might be worth the added complexity.
The Devil in the Details
Beyond the big three factors (glycerol concentration, temperature, and container type), several other variables influence stock viability:
Growth phase at harvest matters. Cells from stationary phase cultures are generally more resistant to freeze-thaw damage than exponential phase cells. This aligns with the broader principle that stressed, hardy cells survive freezing better than their pampered, rapidly dividing cousins.
Freezing speed is critical. The literature emphasizes immediate freezing after adding cryoprotectant, often with a "snap freezing" step in liquid nitrogen before transferring to -80°C storage. Some protocols call for controlled cooling rates of 1°C per minute.
It is highlighted that repeated freeze-thaw cycles are death to bacterial stocks. However, in our hands we did not observe this. When we had stocks with a low viability, switching to the backup stock, which was often entirely untouched, showed pretty much the same low viability. In theory, freeze-thaw cycles must matter, but in practise we did not see any difference between stocks in use and backup stocks that were never touched. Don't ask me why.
The Viability Timeline
Perhaps the most sobering finding: after 12-18 months at typical storage temperatures, viability of bacterial species stored without cryoprotectants drops to less than 20% of the initial cell count. With proper cryoprotectants, this jumps to 80-90%.
This explains my DH5α disasters. A year at -20°C, even with 40% glycerol, was pushing the limits. Add in the potential issues with smaller tube volumes and specific genetic backgrounds, and the stock failures we experienced were entirely predictable, and perfectly in lign with what was stated in the literature.
Lessons Learned
Looking back, the Nottingham system worked despite itself. The 40% glycerol provided excellent short-term convenience, but the long-term success was likely due to factors I didn't control for: larger volumes, glass containers, and possibly just fresher stocks that never had time to deteriorate before being used up or replaced.
The Brunel experience is rather educational. Yes, the immediate problem was solved by adding -80°C backup storage. But the deeper lesson is about the difference between "works in my hands" and "works because I understand why it works."
Practical Recommendations
If you're setting up a bacterial stock system (or troubleshooting an existing one), here's what I'd suggest based on both the literature and hard-won experience:
- Use 15-20% final glycerol concentration as your default. Don't be clever like me unless you have a specific reason.
- Store at -80°C for any stock you care about long-term. -20°C is fine for working stocks you'll use within weeks, but don't trust it for archival storage. Many better Universities have excellent -80°C storage facilities with alarms and backup cover. If you do not mind freezing your fingers off: keep all stocks at -80°C. You can keep a two-tier system, with daily-use stocks at -20°C and backup copies at -80°C (or liquid nitrogen if you're fancy), but be aware that the -20°C stocks have a very limited expiry date.
- Harvest cells in stationary phase when possible. They're hardier.
- Freeze immediately after adding cryoprotectant. If you have liquid nitrogen available, snap-freeze before transferring to final storage temperature. (This looks good on paper, but is utterly unpractical on a daily basis in the lab.)
- Don't assume your system is fine just because it's worked so far. Check viability periodically, especially for critical strains.
- Read the literature before committing to a protocol. I know, I know, it seems obvious in retrospect. But when you're busy running experiments, it's easy to stick with "what we've always done" without questioning whether it's actually optimal.
So before you commit to any bacterial storage protocol, do what I failed to do years ago when I started my own lab: dig into the literature, understand the variables and design your system based on evidence rather than convenience. Your future self will thank you.