Single-stranded DNA production from phagemids containing GC-rich DNA fragments.

Single-stranded (ss)DNA is a preferred template for dideoxy DNA sequencing. It is most commonly obtained by cloning the studied sequences into an M13 vector (1); however, large inserts tend to be unstable. Plasmidphage chimeric vectors (phagemids) offer an alternative (7). They replicate as normal plasmids in E. coli, but upon infection with a helper phage (5,7) they produce phagemid ssDNA and package it into phage-like particles. ssDNA can be easily recovered from the culture supernatant by a standard procedure (6). Some helper phage constructs like M13K07 and its derivatives such as VCSM13 were designed to ensure that phagemid ssDNA is packaged and exported preferentially over phage ssDNA (5,7). However, helper phage ssDNA can contaminate the phagemid ssDNA to various extents. This causes ambiguities in sequencing reactions, because some primers (e.g., the commonly used T7 or T3 primers) show weak selective binding to helper phage ssDNA (data not shown). Also, yields of phagemid ssDNA vary considerably. In our work, additional problems were encountered in sequencing Streptomyces spp. DNA [GC content of 70%–74% (2)]. Double-strand sequencing of Streptomyces coelicolor A3(2) DNA gave poor results. Sequencing of a single-stranded template obtained after VCSM13 helper phage infection was more successful, but the use of recommended methods of phagemid ssDNA production resulted in a variable amount of DNA rescue, often with an abundance of a helper phage ssDNA. We examined the following factors influencing phagemid ssDNA production: (i) initial density of cell cultures, (ii) multiplicity of infection (MOI) and (iii) insert length of a recombinant. The procedure was applied to phagemids (clone A and clone B) in which pBluescript SK(+) vector (Stratagene, La Jolla, CA, USA) contained 0.3and 1.3-kb inserts, respectively, of S. coelicolor A3(2) DNA. The host strain was XL1-Blue E. coli (Stratagene) (recA1, endA1, gyrA96, thi-1, hsdR17, supE44, relA1, lac [F′proAB, lacIqZ∆M15, Tn10(tetr)]). Cultures grown 3–4 h were used as a source of exponentially growing cells. Cell concentrations were estimated from the optical density (OD)600 of cultures, assuming an OD600 of 1.0 approximately equal to 8 × 108 cells/mL (6). The quantity of ssDNA was measured by densitometry using the CS-1 Image Documentation System Version 1.20 (Cybertech, Berlin, Germany) and single-stranded M13mp18 DNA as a standard. All measurements were based on 1.5-mL cultures (a standard centrifuge tube volume for DNA minipreparations). Contrary to reports in previously discussed protocols (3,6,7), we found exponential growth of the inoculum culture to be a prerequisite. When overnight culture or glycerol stock cultures were used as inoculum, we often observed no growth. The yield of phagemid ssDNA was often very low, and large quantities of helper phage ssDNA were produced. The most surprising observation was that for phagemid ssDNA production, MOI was much less important than initial cell density at the time of infection. This can be seen in Figure 1, lanes 1, 5 and 9, where samples of the same MOI (1.67 plaque-forming units [pfu]/cell), but different initial cell densities are compared. In lane 1 (density 6 × 105 cells/mL), recombinant ssDNA, although obtained at low efficiency, was