The choice of storage method depends on several criteria,including the intrinsic stability of the protein,the duration of storage required,and how stringent the need for preventing damage is.There is tremendous variability in these criteria.However,the following general guidelines should apply in most instances.

If the goal is simply to keep a small batch of purified protein around for a few days,then often storage in unfrozen solution at 4℃ is sufficient.If greater duration of storage is required or if there is obviously protein degradation (e.g.,visible precipitates or unacceptable loss of function),then the next option is to choose either frozen storage or storage as a salted-out precipitate (UNIT 4.5; APPENDIX 3F).Both methods are relatively straightforward and,when conducted properly,should afford several weeks or months of storage stability.Prior to attempting to design a salting out strategy,preliminary freeze-thawing experiments should be conducted to determine a given protein preparation's resistance to this stress.If the protein is stable during this treatment,then long-term storage in the frozen state will probably provide adequate stability.However,if preliminary experiments document that the protein is sensitive to acute freeze-thawing stress,it will be necessary to test the appropriate cryoprotectants or to opt to explore the utility of salting out the protein.The latter is probably somewhat easier,but it does not take a large effort to optimize freeze-thawing resistance.At this point,the choice of method to use depends on prior experience with the protein,as well as the history of how a given lab has successfully dealt with protein storage stability issues.Finally,only if all other methods fail or if there is a need for very long-term storage (e.g.,18 to 24 months)at ambient temperatures should freeze-drying be explored.

STORAGE AS FREEZE-DRIED SOLIDSA properly formulated and freeze-dried protein preparation can remain stable for years,even at room temperature (Carpenter and Chang,1996); however,development of optimal solution compositions and processing conditions for a freeze-dried formulation is well beyond the scope of the typical academic lab.One main difficulty is that the type of lyophilizers available in most labs,which usually have no capacity to control sample temperature,cannot achieve the low level of residual water (e.g.,<1% to 2% by mass)that is required for long-term storage stability of freeze-dried products.To dry samples sufficiently,their temperature must be increased to greater than room temperature during the terminal stages of the freeze-drying process.

Also,without temperature control during drying the sample can collapse into a clump,which makes it even more difficult to remove water.Even if appropriate processing conditions can be developed with a given lyophilizer,it is critical to use the correct additives to prevent protein denaturation during both freezing and drying (reviewed in Carpenter and Chang,1996; Carpenter et al.,1997).The disaccharides sucrose and trehalose are very effective at protecting proteins during these stresses and during storage in the dried solid.

These sugars are nonreducing.Reducing sugars such as glucose and maltose should not be used since they degrade proteins via the Maillard reaction.The degree of protection during freezing depends on the initial bulk concentration of sugar present,whereas that during drying depends on the sugar:protein mass ratio (Carpenter and Chang,1996).As noted above,freezing protection is due to preferen-tial exclusion of the sugar from the protein's surface,which increases the free energy of unfolding.During drying,protein molecules unfold as the water hydrating their surface is removed.Sugars prevent this damage by hydrogen bonding to the dried protein in place of the lost water.

The storage stability of a dried formulation depends on a low residual water content (e.g.,<1% to 2% by mass),retention of the native protein conformation during fre-ezing and drying (which can be accomplished with protective additives)and storage of the samples below the glass transition temperature of the amorphous phase,which contains the protein and stabilizing sugars (Carpenter and Chang,1996).Measuring these physical parameters is often beyond the capabilities of most academic labs.Thus,for most labs freeze-drying should be considered only when all other methods for enhancing storage stability have been found to be inadequate.Assuming that suboptimal processing will be necessary (i.e.,there is no way to control sample temperature in the freeze-drier)and that critical physical parameters cannot be measured,is there any chance of obtaining a stable dried formulation? The answer is "maybe."

The practical approach is to prepare the protein at a relatively high concentration (which increases resistance to freezing and reduces volume to be dried)with sufficient trehalose or sucrose to inhibit unfolding during freezing and drying.If the protein preparation is resistant to freezing,usually a sugar:protein mass ratio in the range of 1 to 2 is adequate to prevent protein unfolding during drying.If protection is required during freezing,often an initial sugar concentration of 200 to 300 mM is required for optimal protection.It may also be beneficial to include a carbohydrate polymer (e.g.,dextran,Ficoll,hydroxyethyl starch),which can provide sample bulk and increase the glass transition temperature of the amorphous phase.Sometimes adding a bulking agent is needed to prevent protein from escaping from the container during drying.Also,these polymeric amorphous bulking agents make it easier to dry the samples without collapse.

An ini-tial polymer concentration of 2% to 3% (w/v)usually provides adequate bulking.The samples are usually frozen by either placing them into a mechanical freezer or by immersing the sample containers into liquid nitrogen.It is preferred to aliquot the preparation into several small test tubes (e.g.,1.5-ml polypropylene microcent-rifuge tubes),rather than trying to dry a relatively large volume in a few containers.The frozen samples should be placed into a container (e.g.,desiccator or flask),which is attached to the lyophilizer,and immediately exposed to the system's vacuum.Sublimation cooling should keep the samples frozen during the initial phase of the process when ice is removed.After ice is removed the sample temperature will incr-ease to room temperature,during which time water is partially desorbed from the remaining non-ice phases.The samples should be kept under vacuum for sufficient time to "complete" the drying process.

Drying is never actually complete.Rather the water content eventually reaches a minimal level,after which more time under vacuum at room temperature will not facilitate further drying.The water content resulting after drying at room temperature in sucrose or trehalose formulations is usually >4% to 5% by mass.The actual duration of drying needed must be determined empirica-lly,but this can be a problem if the required instrumentation is not available.To be safe,for typical samples with a volume of 0.5 ml for example,it is advisable to keep samples under vacuum for at least 1 to 3 days.Also,to minimize degradation of the protein,it is recommended that the dried samples be stored in the freezer at -20℃ or,even better,at -80℃.Again,all of this effort seems to be excessive compared to simply freezing the solutions,especially since subzero storage tempera-tures probably will be required for suboptimally freeze-dried samples.