Primary cDNA Libraries
*CeNular Immunology Section, iLaboratory of Chemical Biology, National Institute ofArthritis and Rheumatism Branch, National Institute of Arthritis, Musculoskeletal and Skin Diseases, Diabetes, and Digestive and Kidney Diseases, Bethesda, Maryland 20892 Received August 4, 1986 A technique for the long term storage of primary cDNA libraries in a form such that relevant DNA sequences can be readily identified and retrieved is described. cDNA libraries produced using the Xgt 10 cloning vector were plated out on host bacteria in 0.7% top agarose supplemented with 30% glycerol. Nitrocellulose lifts of these libraries were made and stored. These lifts could be screened at a later time to permit identification of bacteriophage plaques containing specific cDNA inserts. The plated libraries were then transferred to a -70°C freezer. The combination of freezing and glycerol treatment allowed the bacteriophage in these primary cDNA libraries to remain viable for significantly longer than 1 year. 0 1987 Academic Press, Inc.
KEY WORDS: freezing: storage; glycerol; bacteriophage; cDNA libraries.
Lambda bacteriophage have been used as cloning vectors since 1974 (1,2). Through successful genetic engineering, these vectors have been designed with properties which make them exceptionally valuable as tools in molecular biology (3). The Xgt 10 bacteriophage was constructed such that inserts of foreign DNA could be ligated into that vector’s single EcoRI site, thereby converting the bacteriophage from having a lysogenic to a lytic effect on bacterial strains such as Escherichia coli Ceoo hfr (4). hgt 10 is a highly efficient cloning vector and much effort has been invested in determining the conditions which will produce large, full length, primary cDNA libraries in this vector (4). The usefulness of such libraries is greatest when they are first plated out and screened. Unfortunately, present methods do not allow such primary libraries to be stored and rescreened, but instead require that they be amplified and kept in liquid suspension. Because not all bacteriophage replicate at the same rate, this amplification step frequently leads to the underrepresentation of clones
with full-sized inserts, and can even lead to the loss of some clones (5). To eliminate this problem, we have developed a method which permits cDNA libraries to be plated out, screened, and then stored without further manipulation or loss of clone viability. MATERIALS
The standard techniques used in this study are to be found in the Maniatis text on molecular cloning (5). C600 hjZ bacteria were grown in 0.2% maltose-supplemented LB2 (5) at 37” to an optical density of 1.2. The bacteria were pelleted, washed once in LB, and resuspended at 40% of the original volume in LB containing 0.0 1 M MgCI, . &t 10 bacteriophage containing cDNA inserts (4) were added to 100 ~1 of Csoo hjl in 17 X loo-mm plastic centrifuge tubes. The tubes were incubated at 37°C for 15 min, mixed with 3 ml of melted 0.7% top agarose, and then poured onto prewarmed LB agar plates (England Labs, Beltsville, MD). These ’ Abbreviations used: LB, Luria broth; DMSO, dimethyl sulfoxide; SSC, sodium citrate buffer.
’ To whom correspondence should be addressed. 85
0 1987 by Academic Press. Inc. All rights of reproduction in any form reserved.
1. hgt 10 bacteriophage containing cDNA inserts were mixed with C,, hfl bacteria and plated out in conventional top agarose (A and B) or 30% glycerol-supplemented top agarose (C and D). The plates were incubated for 24 h, sealed with parafilm, and then stored at either 4°C (A and C) or -70°C (Band D). The number of viable bacteriophage in 10 randomly chosen plaques per plate was determined at several time points over a 13-month period. FIG.
plates were incubated at 37” overnight and then sealed with parafilm before being stored. Top agarose was prepared by adding 0.7% agarose (BRL, Gaithersburg, MD) to LB and exposing it to autoclave sterilization. In certain experiments, prewarmed glycerol or DMSO was added to the melted top agarose at 43°C. Nitrocellulose lifts were made from Xgt 10 libraries constructed from T-lymphocyte cDNA enriched for T cell antigen receptor @ chain message and plated out in conventional top agarose or in 30% glycerol-supplemented top agarose. The nitrocellulose was baked, prehybridized, and then hybridized overnight to a nick-translated 0.5 kb EcoRI fragment from the cDNA clone 86T5, which recognizes the @chain of the T cell antigen receptor (6). After a 24-h hybridization in
50% formamide at 42°C the nitrocellulose lifts were washed at 60°C in 0.1% SSC for 30 min, and then exposed to Kodak AR-2 film at -70°C overnight in the presence of an intensifying screen (7). RESULTS
Preliminary experiments were performed to determine whether agents commonly used as cryopreservatives had an adverse effect on the plaquing efficiency of Xgt 10 phage. We found that inclusion of 15 or 30% glycerol in top agarose had no effect on plaquing efficiency, but did reduce plaque size approximately fivefold. In contrast, 5% DMSO reduced plaquing efficiency by lo- 15% (data not shown). We therefore elected to study the effects of freezing and 30% glycerol supplementation of top agarose on virus survival.
OF cDNA LIBRARIES
Figure 1 shows that X phage plated out in top agarose and stored at 4°C underwent a marked decline in viability over time. In contrast, there were greater than lo5 viable bacteriophage per plaque in glycerol-supplemented plates stored at -70°C. This protective effect of glycerol plus freezing was also seen in plates stored at 4°C for several weeks and then transferred to -70°C conditions (data not shown). Although freezing by itself conferred some protection on plain top agarose plates, in the absence of glycerol no bacteriophage survived for 7 months. In addition, the formation of large ice crystals in conventional top agarose plates stored at -70°C perturbed the surface architecture of these plates to such a degree that identification of individual plaques was difficult. This was not true in glycerol-supplemented plates, where ice crystal formation was inhibited. Three methods were found to be effective for harvesting viable bacteriophage from frozen plates. (i) Part of a plaque could be pried off the frozen plate with a sterile 22 gauge needle. (ii) An entire plaque could be harvested by allowing the plate to warm to - 10°C and then “drilling” the plaque with a sterile Pasteur pipet. The plate was never allowed to thaw completely; if thawed, a sol-gel transformation occurred and liquid was exuded at the surface, which smeared the plaques. (iii) A sterile piece of nitrocellulose could be pressed onto the frozen plate and then transferred to a growing lawn of Cboo hJ1 on another plate. If this was done carefully, partial replicas of the original plate could be made and plaques picked from it. This technique worked best if the original plate was at - 10°C and contained plaques greater than 1 mm in diameter. DISCUSSION
The ability to store virgin cDNA libraries is a useful addition to the techniques cur-
FIG. 2. A Xgt 10 library was produced in which approximately 2% of all bacteriophage contained inserts derived from the P chain of the T cell antigen receptor. An aliquot of this library was mixed with &, I&‘E. coli and plated out in conventional top agarose (A) or agarose supplemented with 30% glycerol (B). After 24 h of incubation, nitrocellulose lifts were made from these plates and probed with nick-translated 0 chain DNA. The strong positive signals seen in autoradiographs from the conventional (C) and glycerol-supplemented (D) plates correlated precisely with those plaques bearing p chain inserts.
rently available in molecular biology. To screen X phage cDNA libraries, bacteriophage must be plated out in semisolid agar or agarose. Under conventional conditions, the longevity of bacteriophage on such plates rarely exceeds several weeks. To maintain the library for a longer period, it must be amplified and stored as a liquid suspension. Even after amplification, a library will undergo a gradual loss in titer and heterogeneity as poorly surviving bacteriophage die. These problems may be avoided using the
techniques described in this report, wherein a primary library, once plated, can be stored without further manipulation or clone loss. Greater than 10’ viable bacteriophage per plaque could be harvested from plates stored for 1 year at -70°C when the X phage were plated in top agarose supplemented with 30% glycerol. Extrapolating from the number of viable bacteriophage present after 3, 7, 10, and 13 months of storage under these conditions suggests that greater than lo3 viable bacteriophage will remain in these plaques after 3 years of storage. The use of glycerol supplementation presents no problems in terms of conventional plate screening or clone selection from cDNA libraries. This technique should therefore be suitable for the long term storage of primary libraries produced in other X vectors. Since bacteriophage plated out in 30% glycerol and stored at 4°C can later be transferred to -70°C and still maintain significant viability, initial screening (and production of duplicate nitrocellulose lifts for future use) can be easily accomplished. One drawback to the use of glycerol-supplemented plates is the modest initial decrease in the titer of bacteriophage per plaque. We attribute this to the viscosity of glycerol, which inhibits phage diffusion and results in the production of smaller plaques. However, the density of bacteriophage per unit area of plaque is maintained, as shown by the strength of the signal which results when nitrocellulose lifts from the plates are hybridized with nick-translated probes (Fig. 2). There may even be an advantage to smaller plaque size, in that a greater number of plaques can be screened per plate with good resolution. The technique in this report permits virgin cDNA libraries to be repeatedly probed and analyzed over a prolonged period of time. In
our laboratory, we conventionally plate 2 X 1O4 colonies in l OO-cm* petri plates using 30% glycerol-supplemented top agarose with excellent retrieval of viable bacteriophage after prolonged storage. The amplification of these libraries at a later time is possible, so that there are no apparent drawbacks to adopting this methodology as a general laboratory tool. ACKNOWLEDGMENTS We thank Dr. Alfred D. Steinberg for scientific suggestions and encouragement, and Joyce Anderson for secretarial assistance. Dr. Dennis Klinman is supported in part by a fellowship grant from the Arthritis Foundation.
REFERENCES I. Ramback, A., and Tiollais, P. (1974) Proc. N&l. Acad. Sci. USA 71,3927-3930. 2. Murray, N. E., and Murry, K. (1974) Nature (London) 251,48 l-486. 3. Williams, B. G., and Blattner, F. R. (1980) in Genetic Engineering (Setlow, J. K., and Hollander, A., Eds.), Vol. II, p. 201, Plenum, New York. 4. Huynh, R. V., Young, R. A., and Davis, R. W. (1984) in DNA Cloning Techniques: A Practical Approach (Glover, D., Ed.), pp. 274-308, IRL Press, Oxford. 5. Maniatis, T., Fritsch, E. F., and Sambrook, J. (1982) in Molecular Cloning, pp. 17-74, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. 6. Hedrick, S. M., Cohen, D. I., Neilsen, E. A. and Davis, M. M., (1984), Nature. 308,149-153. 7. Klinman, D. M., Mushinski, J. F., Honda, M., and Steinberg, A. D., (1986) J. Exp. Med. 163, 1292-1307. 8. Meryman, H. T. (1956) Science 124, 515-521. 9. Panijel, J., Huppert, J., and Barbu, E., (1957), Ann. Inst. Pasteur 93, 183-193. 10. Lark, K. G., and Adams, M. H. ( 1957) Cold Spring Harbor Symp. Quant. Biol. 18, 17 I-183. II. Scherer, W. F. (1960) Exp. Cell Res. 19, 175-176. 12. Porterheld, J. S., and Ashwood-Smith, M. J. (1962) Nature