Agar Plug/Serial Dilution Approach for Rapid PCR Screening of Phage Libraries

The “superpool”/pool format has widely been used for PCR screening of BAC, PAC, or YAC genomic libraries. While these vectors have become the tools of choice for genome analysis, such libraries are not available for all species. Furthermore, for specific applications such as the construction of knockout vectors, small insert libraries are useful, as they enable one to obtain clones containing the gene of interest and its closest sequence directly, with no need for tedious subcloning steps. With genomic libraries built in λ bacteriophages, the number of clones necessary to cover the genome (i.e., three or four genome equivalents) is generally too high for isolating single clones in microplates. Therefore, the screening is often performed by classic colony hybridization and requires the plating of a large number of phage plaques (at least 4–6 × 105 for a phage library with 15 kb average insert size). When homologous primers can be designed, PCR of “superpools” and pools has been proposed as a useful alternative to hybridization (1), but these protocols involve a liquid growth step for the phages, and this procedure sometimes results in bias and under-representation of some phages. The possibility of serial dilutions from plaques was mentioned with the use of nitrocellulose membranes that are first lifted from the agar plates and then split into a series of smaller sections (4). Here a significant improvement on these techniques that does not require radioactivity or membranes is proposed. This technique makes use of a series of dilutions and exploits the possibility of performing PCR directly on agarose slices derived from an LB-agar plate. Embryonic stem cells derived from 129 SV mice are the principal targets for knockout experiments. Therefore, obtaining DNA from a mouse 129 genomic library for the genes under scrutiny is an essential step toward their targeted inactivation. The current technique has been successfully applied to identify mouse phages encompassing two genes involved in sex determination, PISRT1 and FOXL2 (2,3). The phage library was constructed in λGEM-12 (Promega, Madison, WI, USA), with MboI-digested 129sv genomic fragments. The phage particles were used to infect E. coli strain KH802. One million bacteriophages plaques from the initial phage library were plated on 50 plates at about 20 000 plaques/8.5-cm Ø plate (Figure 1). To each plate 2 mL 10 mM MgSO4 were added, and the plates were shaken gently for 10 min. Each supernatant was poured in 50 × 2 mL Eppendorf® tubes. One hundred microliters of each solution were transferred to one well of a 96-well microplate, to which 20 μL Proteinase K (ICN Biomedicals, Costa Mesa, CA, USA) were added to a final concentration of 1 mg/mL. The plate was placed in a MJ thermal cycler (MJ Research, Waltham, MA, USA) and incubated for 1 h at 65°C and 10 min at 100°C. Five microliters were used in 20-μL reactions with the specific primers in standard conditions. Positive agar plates were identified, and the agar was divided into 50 blocks with a sterile scalpel. Each block, representing approximately 400 different phage plaques, was put in a 2-mL Eppendorf tube containing 10 mM MgSO4. After a 30-min incubation at 37°C with gentle shaking, 100 μL were placed in PCR tubes, and the Proteinase K/PCR procedure was resumed. Ten microliters from the positive plug supernatants were serially diluted in 100 μL MgSO4 in six successive 10-fold dilutions. Each of these dilutions was mixed with an equal volume of KH802 bacteria and incubated for 20 min at 37°C, mixed with 5 mL 0.5% LB-agarose, plated on an LB-agar plate, and incubated overnight at 37°C. The next morning, 2 mL 10 mM MgSO4 were added, and the supernatants were collected from the plates and tested by PCR. The agar from the positive plate with the lower phage concentration out of the six tested dilutions was cut again in 50 plugs representing 10–20 different plaques, and the procedure was resumed. After three successive steps, for the two PCR systems tested, the lower positive dilution contained less than 100 clones. Each clone was picked with a sterile toothpick into 10 μL water and incubated 10 min at 100°C. The PCR mixture was then added to a final volume of 20 μL, and the samples were subjected to cycling according to the standard protocol, allowing the identification of the positive clones. An example of this last screening test is given in Figure 1 with primers designed for the mouse homologue of PISRT1. Because of the weak level of interspecific sequence conservation, a specific amplification product of only 81 bp could be defined. In such a case, PCR screening is the alternative of choice compared with classical hybridization. At the last step, 28 individual phage clones were picked, of which Benchmarks