by Julian Warhurst, julian.warhurst@brooks.com; John Fink, john.fink@brooks.com; Tiffany Holmes, tiffany.holmes@brooks.com; Matthew Albert, matthew.albert@brooks.com; Bruce Zandi, bruce.zandi@brooks.com

It seems to be universally accepted that storing at -190°C in a liquid nitrogen (LN2 )vapor-phase freezer is the best technique for biosample preservation. Brooks Automation (CSA CSM) decided to explore this idea further, looking specifically at what happens to the innocent samples that are pulled from the freezer, the ones neither used nor touched, but going along for the warming ride. These samples may take this ride countless times throughout their storage lives, and the goal was to ensure they would not be damaged by excessive temperature excursions in the process. This brief summarizes an oral presentation on findings given at the May 2015 International Society of Cellular Therapy Conference.

Our study focused on developing a protocol to define the frequency and duration of rack extraction from a -190°C vapor-phase LN2 freezer and to ensure the samples would meet acceptable quality levels when returned. The temperature threshold chosen was the glass transition phase of water (Tg -135°C). We limited the study to 2 mL FluidX vials stored in 10×10 cryoboxes in 14-shelf freezer racks. The racks were pulled from the LN2 freezer using a mechanical device. This procedure eliminated the human variation that occurs when racks are extracted from a freezer manually, plus it enabled uniform timing. The boxes or vials were never touched or removed from the racks.

In each test, we first stabilized the samples carefully at -190°C. Sample temperatures were recorded digitally every two seconds throughout the warming cycle. Fine-wire thermocouples were placed in the centers of the vials. Three vials were placed in each box—two in the corners and one in the center. From past studies, it was known that the corners were the most temperature-affected positions in the rack. They are exposed on two sides and there are no surrounding vials to provide insulation. We created a “worst-case” cryobox with only three vials and without a cryobox underneath or above it. It was in the top position of the rack, closest to the top of the freezer. A more typical cryobox of 100 filled vials was positioned in the center of the rack with full cryoboxes both above and below it.

The Results: In the first test, the rack was pulled out of the freezer for 30 seconds then immediately returned (Figure 1). The graph shows the temperature profile of one of the corner position vials in the top cryobox. The dashed vertical line on the graph shows the point at which the rack was returned to the freezer.

It is important to note that the vial continued to warm for 2.5 minutes after it went back into the -190°C environment. This is because­ when the rack with the cryobox was pulled out it warmed up considerably more than the vials inside.

Once returned to the freezer, the vial continued to be warmed by the heat from the surrounding rack and cryobox. The vial did not begin to cool until 2.5 minutes after insertion. The vial warmed a total of 7°C when outside of the freezer and an additional 13°C after being returned to the freezer.

We extracted the same samples for four different time durations in the second test. This experiment simulated different SOPs that labs might have. For the 2.25 minute and 3-minute extraction times (Figure 2), the sample temperature in the top box, corner tube position reached or exceeded -135°C (Tg). Also, notice the sample continued to stay above its initial temperature for some time after insertion. When the rack was extracted for three minutes, the peak temperature occurred two minutes after insertion at -122°C, above Tg.

The third test evaluated the corner position vial in the center rack position with a full cryobox at the same four extraction durations. As expected, the peak temperatures were quite a bit lower (Figure 3). With a three minute exposure, the vial temperature only reached -152°C after insertion. In this experiment, the rack could be extracted for a total of six minutes before the sample temperature exceeded Tg.

This third test represents the ideal scenario though, since the cryobox and vials have neighbors and are well insulated. Recall the vial in the top cryobox didn’t fare anywhere near as well. When designing a protocol, you must design around the worst-case scenario.

One interesting and important finding was the long recovery time back to -190°C of the vial in the fully populated cryobox after it was inserted back into the LN2 freezer. You must take this slow re-cooling effect into account and consider how frequently the rack is being pulled from the freezer and its starting temperature when developing an SOP.

In the fourth test, a comparison was conducted of the corner tube versus the center tube in the two different cryobox densities. The corner and center positions of the sparsely populated box correlate quite closely (Figure 4).

That’s because there are no neighboring vials to act as an insulating wall. Contrast that to the results in the bottom two lines of the graph—the center and corner vial positions in the fully populated cryobox.

The corner vial warmed up noticeably quicker than the center vial. Note the center vial had a long delay to its peak temperature that occurred at about 25 minutes after the rack had been returned to the freezer. And it continued to stay well above its original temperature for an extended period.

The next experiment assessed the time it would take for the corner vial in the fully populated cryobox to return to its starting temperature after being extracted for three minutes and then immediately returned to the freezer. It took three hours for the vial to finally return to its starting temperature (Figure 5).

Consider the scenario if the rack was pulled again one hour after being returned to the freezer: the sample would have been 5°C warmer than when initially extracted. If a rack may be pulled multiple times during the day, a protocol must consider and compensate for this warming delta penalty. While the rack may not go over Tg on the first pull, Tg might be exceeded on the third or fourth extraction during a day.

One unexpected finding was the impact of standard heating and cooling (HVAC) vents on sample warming. When testing the corner vial in the fully populated cryobox, a 50 percent improvement in the rate of sample warming was achieved when we turned off the room HVAC vent (Figure 6).

To further the experiment, we extracted the rack into a foam insulation sleeve and additional improvement was realized. It was remarkable that such a seemingly minor and overlooked factor such as HVAC circulation could make such a substantial difference in the sample warming rate. It also points out that people must account for and correct for these factors in their lab protocols.

Conclusion: The study revealed that basic placement and handling techniques, cryobox location in a rack and the number and location of vials in a cryobox can make significant differences in sample warm-up rates and peak sample temperatures.

Innocent samples can warm above Tg from a 135-second rack extraction and then take hours to return to the -190°C starting temperature. Failure to plan and compensate for these factors in your protocols may result in failure to protect innocent samples from crossing Tg.

The following are some best-practice recommendations:
• Minimize the time samples are out of the freezer.
• Avoid empty vial positions in a cryobox. Surrounding vials provide significant thermal mass to reduce warming during exposure. It’s better to have vials filled with water than to have no surrounding vials in the cryobox.
• Avoid or minimize any direct drafts or HVAC exposure in the lab. Be aware that many lab safety protocols call for high ventilation rates.
• If possible, shroud the extracted rack to block any convective activity.
• Accurately track the amount of time the racks spend in and out of the freezer and estimate the final temperature of innocent samples. People may be retrieving the same rack several times during the day, and it is important to make sure the samples do not cross Tg.