Circulating-Gas Freeze Drying

WAGMAN, JACK (U.S. Army Biological Laboratories, Frederick, Md.) AND EDWARD J. WENECK. Preservation of bacteria by circulating-gas freeze drying. Appl. Microbiol. 11:244-248. 1963.-Water-washed Serratia marcescens and Escherichia coli were freeze dried in a circulating-gas system at atmospheric pressure. This convective procedure resulted in a substantially higher survival of organisms than could be obtained by the vacuum method of freeze drying. There was little or no decrease in cell viability during convective drying when the residual moisture content was 15 %o or higher. Below this level, survival declined with decreasing moisture content. A detailed comparison of the convective and vacuum methods indicated that the advantage gained by freeze drying bacteria in air accrues in the early period of sublimation, at which time cells were found to be sensitive to vacuum drying but insensitive to air drying. An explanation for this difference is proposed, based upon the kinetics of water removal in the two processes. In brief, it is suggested that the convective method permits samples to be dried more uniformly; and regional over-drying, which may be deleterious even if transient, is thus avoided in achieving the optimal level of moisture. The most widely used of the current methods for drying bacteria and other biological materials for storage is the process known as lyophilization or freeze drying. Either of these terms usually implies that drying takes place by the removal of moisture from a frozen solution or suspension in a high vacuum. Although this method, in its many variations, has been useful in preserving many kinds of nonviable materials, especially proteins and tissues, it has, with few exceptions, not yet been successfully applied in the maintenance of viability. Bacterial cells are perhaps the most notable exception. Even in this case, freeze drying has not proved to be a generally satisfactory process for many species of bacteria have been found to be highly sensitive to it (Fry, 1954; Heckly, 1961). Lyophilization may be considered a two-stage process of freezing and drying. Freezing, as applied to cells and tissues, has been studied extensively (Smith, 1954) and generally has presented no seriouis problems for the survival of microorganisms. Even with sensitive bacteria, satisfactory recoveries can be obtained by the proper control of such factors as temperature, cooling rates, and suspending media (Mazur, Rhian, and Mahlandt, 1957). On the other hand, the difficulties experienced with the drying phase of the freeze-drying procedure, as applied to sensitive microorganisms, have proven to be much more formidable and less easily resolved. Efforts to bring about improved survival of organisms have been limited largely to a search for substances, or combinations of substances, that when added to the suspending medium exercise a protective effect during drying. This approach has been used for the past 40 years by numerous investigators (Fry, 1954; Heckly, 1961), and a very impressive list of substances has already been tried. However, claims made for the protective action of certain substances by some investigators have occasionally been contradicted by others, and, even where the protective action of a substance or combination of substances for a particular organism has been substantiated, the effect is often not observed with other organisms. Moreover, this kind of solution to the problem of the survival of bacteria during freeze drying is of limited usefulness because the added substances are, in many instances, undesirable from the point of view of the subsequent use of the dried product and may be difficult, if not impossible, to remove. This state of affairs may be attributed, in part, to the general uncertainty that exists as to which aspects of the drying phase are relevant to the survival of bacteria during lyophilization. Furthermore, there appears to be a tendency to ignore certain factors, even when the available evidence has shown them to be important. For example, Fry and Greaves (1951) obtained evidence suggesting that the most pronounced losses in the viability of bacteria during vacuum freeze drying occur at the time when the most rapid removal of water is taking place. Similarly, Hutton, Hilmoe, and Roberts (1951) observed in one experiment that the survival of Brucella abortus decreased with an increased rate of drying, and in another found what appeared to be an optimal drying rate. Despite these observations, the inclination to use the highest vacuum obtainable still persists, and even Fry (1954) recommended later that a high vacuum be used to ensure rapid removal of water. Similarly, much information has been obtained that supports several, albeit conflicting, conclusions on the importance of the residual level of moisture in the survival of dried bacteria. Indeed, the efficacy of adding glucose and perhaps other substances to the suspending medium 244 on A uust 4, 2017 by gest ht://aem .sm .rg/ D ow nladed fom PRESERVATION OF BACTERIA before freeze drying may reside (Fry and Greaves, 1951) in the regulatory effect that these materials have upon the final moisture content, i.e., in preventing samples from becoming too dry. Despite its recognized importance, there has been no systematic attempt to control the residual moisture during freeze drying, particularly for sensitive bacteria. This is perhaps due to the fact that the vacuum technique is singularly unsuitable for the control of this factor. Moreover, there appears to be no convenient way in vacuum freeze drying to avoid over-drying one portion of a sample in the process of bringing the sublimation in another portion to completion. Considerations such as these resulted in the present study, in which the vacuum method of freeze drying has been replaced by a procedure based upon sublimation at atmospheric pressure. In this procedure, the resistance to the flow of water vapor from the frozen sample to the condenser is reduced by the circulation of air, or other gases, alternately past the sample and condenser surfaces. This technique, which may be designated "convective freeze drying" to differentiate it from the vacuum method, has recently been used by Meryman (1959) in the preparation of dried specimens of mammalian tissue. Besides retaining the advantages of drying from the frozen state, it permits easier control of the moisture-removal process. MATERIALS AND METHODS Pr eparation of frozen samples of bacteria. The organisms used included Escherichia coli B, grown for 18 hr on nutrient broth (Difco) at 34 C, and Serratia marcescens 8 U1K, grown for 17 hr at 31 C in Tryptose broth (Difco). In preparation for drying experiments, harvested cells were washed several times by alternate suspension and centrifugation in distilled water. Water-washed samples of S. marcescens contained at least 5 X 1012 viable organisms per g (dry weight), and for E. coli, there was a minimum of 2 X 1012 viable cells per g. After washing, cells were resuspended in sufficient distilled water to form a fairly thick suspension with a dry weight of about 14 %. For some of the experiments, frozen rod-shaped samples were prepared by filling cellulose nitrate tubes (16 in. diameter, 1% in. in length) with suspension. This was followed by immersion in a Dry Ice-acetone mixture and separation of the solidified suspension from the tubes. In other experiments, frozen bacterial pellets (18 to 14 in. diameter) were used. These were formed by bubbling the cell suspension through Freon 113 cooled in a Dry Ice-acetone bath. Without the benefit of added substances, water-washed samples of S. marcescens and E. coli, such as those described here, have been routinely found to undergo viability losses in excess of 90 % during freeze drying in vacuum. However, no stabilizing materials of any kind were added to the samples in these experiments, to preclude their interference in determining the efficacy of convective freeze drying as compared with the vacuum technique. Convective freeze-drying technique. Figure 1 illustrates the drying apparatus used in these experiments. It is essentially a closed system with a blower that circulates air at about atmospheric pressure. Sublimed water vapor is thus continuously removed from the sample -and condensed in the cold trap. An alcohol-water mixture at -8 C was circulated through the jacket of the sample compartment, and the cold trap was immersed in a Dry Ice-acetone bath. The circulating-gas temperature in the sample compartment varied from about -10 to 5 C. Early experiments were carried out with frozen rodshaped samples in an apparatus consisting of sample and condenser compartments made of glass. With this arrangement about 5 g of dried bacteria could be obtained with a moisture content of about 10 % after 30 hr. Subsequently, pelleted samples were dried in a stainless-steel fabricated apparatus in which this level of moisture was attainable in about 16 hr. The air-flow rate in this system was about 1 liter per sec. A humidity-sensing element was used to determine the approximate moisture content of the sample as drying progressed. Pelleted samples were vacuum freeze dried, in a number of experiments, for comparison with the convective technique, and in some instances to supplement air drying. Vacuum drying was carried out in a glass manifold freeze dryer at about 100 ,u of pressure with sample compartments immersed in an alcohol-water bath at -8 C and vapor condenser in a mixture of Dry Ice and acetone. Measurement of residual moisture and survival. The moisture content of freeze-dried samples was determined by heating at 100 C in a vacuum oven, usually overnight, until virtually constant weight was achieved. This procedure yielded values for residual moisture that were significantly higher than those obtained by heating at 50 C, a method that has often been used in previous studies. Allowance for the difference, which was found to be about 2.2 % of dry weight in measurements with several samples of S. marcescens, should be made in any comparisons of the data presented here with others in the literature. The survival of cells was determined by comparing the number of viable organisms per dry gram in a sample