The Sample Storage Debate:
A Text by Uwe Schoen, Director of Biomedhelix
Multiple overwhelming arguments exist in favour of the storage of biological samples at liquid nitrogen temperatures versus ‑80°C temperatures of electrical freezers, leading to inferior sample quality. In fact, I would like to argue that liquid nitrogen cooled bio-repositories will terminate the usage of ‑80°C electrical freezers. In the long term ‑80°C storage systems will not prevail over ‑136°C liquid nitrogen cooled bio-repositories.
As a physicist specialising in the field of biobanking, I am passionate about cold temperature storage. After many years at the German Fraunhofer Institute with engineers, biologists and medical scientists, I was involved in leading edge biobank projects in Europe. In this proposition I would like to offer a more detailed insight on the reasons why storage temperatures below ‑136°C are so crucially important for sample quality and which technologies are recommended, also for high sample numbers.
Numerous publications conclude that temperatures of -80°C are sufficient to store various types of biological samples even for long term storage. I wonder why? No veterinarian would ever store his samples at these “high” temperatures. Neither would a gynaecologist store his patient’s IVF eggs at -80°C. In the end, these precious samples have to survive the storage period, which in the majority of cases might be years or decades, and obviously they don’t.
Sample handling can also not be cited as the reason to avoid liquid nitrogen storage temperatures of below -136°C. Why should a veterinarian transport his spermatozoa with a cumbersome -80°C freezer to a staple for artificial insemination when there are liquid nitrogen dewar vessels available for that purpose in various sizes? The same reasoning is valid for laboratories.
Liquid nitrogen is available in industrial amounts in almost all settings, even in low- and middle-income countries. In rare cases, where the supply chain for liquid nitrogen fails, you can even produce your own liquid nitrogen with your own scalable plant.
Is it cheaper to keep samples at an inferior -80°C rather than to cool them down to the lower temperatures of liquid nitrogen storage? It is not, because the energy costs are higher for an electrical freezer that stores at -80°C than for a similar-sized liquid nitrogen vessel.
If you are considering the energy costs of an electrical freezer, you also have to include the costs to remove the heat produced by the two-stage vapor-compression system. Even modern vacuum panels cannot compete with the far superior thermal qualities of the vacuum insulation of a state-of the-art dewar.
The mean time between failures are another reason to use liquid nitrogen cooled vessels rather than freezers to store biological samples. Following more than 100 years of manufacturing experience, the vacuum insulation of a robust stainless steel vessel rarely fails, while electrical freezers heat up within hours after an electricity interruption or a compressor failure.
With regard to sample quality, the most important factor is to consider the process to freeze and lower the temperatures of your biological material. Once your samples are frozen, things become easier. Once you have reached your desired storage temperature, there is sufficient evidence that liquid nitrogen storage temperatures deliver the best cryopreservation outcomes.
Nevertheless, “fit for purpose” often serves as the strategic guide for the cryopreservation temperature – ignoring unanticipated future purposes. The main goal of cryopreservation of cells, tissue and organisms – once a sample is retrieved from cryostorage – is to restore as many of the biological functionalities as possible.
The proof of the empirical evidence for the functionality of biological systems, is very often complex. Although the biophysics of the freezing process is very well investigated, the understanding of many aspects remain unsatisfactory.
As mentioned before, the quality of spermatozoa or IVF eggs and many other sample types can only be well-preserved at the low temperatures that liquid nitrogen storage offers. The reasoning for that is outcome-based. The acrosomal degradative enzymatic activity and DNA integrity are only maintained at these low temperatures.
Several hypothetic parameters have been found that harm the outcomes of cryostorage processes. One parameter where the low temperature storage with liquid nitrogen is beneficial, is the exposure to excessive salt concentration to cells when water is removed as ice. (Lovelock JE, Bishop MW (1959) Prevention of freezing damage to living cells by dimethyl sulphoxide. Nature 183(4672):1394–1395). Another injury is caused by the kinetics of ice formation in the intra and extracellular media, which is mechanically destroying. The underlying chemical reactions and physical processes ultimately damage the biological systems.
From a physical point of view, there are widely acknowledged factors that are related to the glass transition temperature of -136°C. Below -136°C the rate of diffusion of a proton is >200 years to move one molecular diameter and therefore chemical reactions are improbable (Ref: Feridoun Karimi-Busheri (Editor): Biobanking in the 21st Century (2015), p39). At temperatures below -136°C, recrystallisation as a physical process is on a standstill. Electrical freezers at -80°C with an almost 60°C higher storage temperature, host chemical reactions and recrystallisation processes with significant impact on the sample integrity. (Mazur P, Leibo SP, Chu EH (1972) A two-factor hypothesis of freezing injury. Evidence from Chinese hamster tissue-culture cells. Exp Cell Res 71(2):345–355)
Often the fact “specific biological sample properties remain stable at -80°C” is confused with “ -80°C is as good or even better as liquid nitrogen storage temperatures”.
Two arguments are often used to justify the storage of samples in electrical -80°C freezers instead of the vapor phase of liquid nitrogen.
The first is the lack of availability of liquid nitrogen as a cooling medium. This argument is weak in almost all environments, as liquid nitrogen is supplied by large gas producers for industrial purposes and processes worldwide. It is easy to provide liquid nitrogen to the storage tank by using supply vessels and preferably vacuum insulated transfer lines, planned in a systematic manner.
The second argument is the concern that the sample container is not specified for the lower temperatures. At the Fraunhofer Institute for Biomedical Technology (IBMT) in Germany, we performed multiple sample transfers from -80°C to LIN storage, and there were never issues around that. Having said that, in rare cases sample containers are not suitable for low temperature storage in the gas phase of liquid nitrogen and a validation / qualification should be performed.
I have worked in the cryostorage automation field since 2003. Logically it follows that large sample numbers require improved sample handling and identification methods. Conventional sample handling equipment specified for room temperatures or mild negative temperatures to -20°C, such as robots, could not be exposed at the extreme low temperatures. The main reasons for discontinuing moving parts are significant ice formation on the cold surfaces inside the storage, thermal length expansion and failing lubrication of moving parts. We successfully developed modules and components for fully automated systems like a bionic robot arm or a low temperature gripper for a joint arm robot, which resisted the low temperatures of liquid nitrogen storage. More importantly, we developed devices and processes to keep parts that are sensitive to low temperatures out of the extreme cold. The engineering requirements were not much more challenging for temperatures lower than -136°C than for -80°C. In this manner we qualified the systems for the lower temperatures.
These patented products and processes lead to a new line – the ASKION C-line semi-automated and automatic cryostorage solutions for liquid nitrogen temperatures. Current -80°C electrical freezer systems like the Hamilton BiOS automated biobank don’t address the main requirements of biobanking. The storage temperatures are a minimum 56°C higher than the desired -136°C and sample integrity is damaged. Normally sample handling happens at temperatures of -20°C harming sample quality further.
A limitation is scalability. Usually, after a period of time, a collection expands, but because of the lack of variety of different sizes for the storage room and the requirement to plan for the final volume, scaling potential is limited.
The system’s failsafe security cannot be guaranteed. In case of a fatal dysfunction, it would be impossible to relocate millions of samples, despite two diverse backup cooling facilities that are connected. Sample manipulation processes are fragile because of the unavoidable ice formation on surfaces. The current design of -80°C automated storage systems usually leaves the moving parts inside the cooled space, which increases the energy consumption and makes de-icing difficult or impossible.
The successful designs for semi- or fully automated sample storage follow these basic principles that are guided by laws of physics:
- the sample storage room is a vacuum insulated dewar vessel;
- the withdrawal systems are on top of the storage tanks and allow cooling of the handling area.
Obeying the basic design rules account for an uncritical use of automats and robots for sample handling and the integrity of samples in their liquid nitrogen cooled bio-repositories at temperatures of –136°C and colder. The companies ASKION and Liconic offer these types of automated biobanks based on liquid nitrogen cooling.
From thermal considerations, thermodynamic properties of vacuum panels or other insulating materials like foam are not remotely comparable. It is impossible to achieve the same quantity of insulation by just increasing the thickness of an insulation material. For these dimensions internal heat transport phenomena become imminent. Vacuum insulation almost completely cuts off the physical heat transport through radiation and convection. The trade-off is a limitation to cylindrical storage dewars with torispherical (not flat) heads, which should have minimum insulation interruptions through opening or even heat conducting bridges like tubes or rails facilitating sample transport.
There are not many arguments left for -80°C as a storage temperature for biological samples. The implementation and running costs for the electrical systems are higher than liquid nitrogen storage systems. Failures are more likely with serious consequences for the sample integrity.
Liquid nitrogen may be an additional medium to deal with, but very easy handling allows even small installations. If large sample numbers require a higher degree of automation, corresponding liquid nitrogen cooled systems have been available for some years now. The manufacturers ASKION and Liconic offer scalable commercial solutions with references for high sample volume biorepositories like the German Helmholtz cohort (German National Cohort NAKO). In parallel, new sample identification and logistic systems based on barcode or RFID support Laboratory Information Systems are available and improve the several processes that become more critical with high sample numbers and are applicable to automated systems or large classic manual biorepositories.
In closing, after having overseen sample storage processes worldwide in biobanks over many years, I have seen sufficient evidence that liquid nitrogen storage is a far superior method over -80°C electrical freezers. Also, in view of global environmental awareness, in the long term liquid nitrogen-related technologies are resource saving. In case of existing -80°C units, the option should be to replace the systems after failure if a complete switch to liquid nitrogen storage is not feasible. In my view it is shortsighted to invest in -80°C freezers when liquid nitrogen sample storage will further strengthen its position in the cryo-storage environment.