Restriction of infectious bacteriophage fd DNA's and an assay for in vitro host-controlled restriction and modification.

Arber2 has shown that bacteriophage fd, which contains single-stranded circular DNA (Marvin and Schaller12) is restricted by E. coli B and E. coli (P1) +. This report presents detailed data on the restriction of both the infectious, singlestranded fd-phage DNA and the infectious, double-stranded fd-replicative form (RF)18 DNA. It also describes an infectivity assay for in vitro host-controlled restriction and modification. This assay has been used to find restriction enzymes in fractionated extracts of E. coli (Linn and Arber11). Materials and Methods.-Bacterial strains: E. coli F+meth-gal-rK -mK-18 E. coli F+ metKleuKrB+mB +threoB+, E. coli F+metgal(P1) +, and E. coli B galmetrB mB+ were kindly provided by Prof. Werner Arber (for derivation of strains, see Arber2). Phage strains: Bacteriophage fd was donated by Prof. Werner Arber; a phage mutant which is only partially restricted by E. coli B (Smith, Arber, and Kuehnlein'7) was the gift of Urs Kuehnlein. Purification and characterization of phage and phage DNA's: The 6X174 phage was purified to the first high-speed centrifugation step described by Rueckert, Zillig, and Huber."1 The fd phage was harvested from the medium of Fraser and Jerrel7 and purified by polyethylene glycol precipitation and CsCl density-gradient centrifugation. The spectrum of the purified phage coincided with the spectrum reported by HoffmannBerling, Marvin and Duerwald.8 Phage DNA was extracted as described by Benzinger, Delius, Jaenisch, and Hofschneider.3 Following exhaustive dialysis against 0.01 M Tris buffer, pH 8, containing 1 mM EDTA, the DNA was further purified by CsCl density gradient centrifugation. The fd RF was isolated according to slight modifications of the procedure of Jansz, Pouwels, and Schiphorst.'0 The filtered RF was sedimented on linear 5-20% sucrose gradients and fractions were collected. Electron micrographs of these fractions (kindly prepared by Dr. H. Delius) showed that more than 80% of the DNA was circular. The fast-sedimenting RF fractions were mostly twisted circles; the slower-sedimenting fractions were mostly open circles.'4 Contamination by single-stranded phage DNA was ruled out as described below. Infectious DNA assays: Lysozyme-EDTA spheroplasts were prepared according to the following modifications of the procedure of Ray, Bscheider, and Hofschneider:13 the harvested cells were resuspended in storage medium (plus 10 ,g/ml of required amino acids) at a titer of 1 X 1010 to 2 X 1010 cells/ml. They were then aerated vigorously for 30 min at 370 (Bscheider, personal communication). The bacteria were treated with lysozyme-EDTA in the presence of Povite albumin for 1 min as described by Ray et al."3 Following addition of MgSO4, the spheroplasts were incubated without aeration for another 30 min at 370 (Bscheider, personal communication). After the spheroplasts had been chilled in an ice bath, they were ready for testing with infectious fd DNA's. The titer of infectious DNA with these preparations remained constant for at least 5 days and did not vary more than threefold for 30 consecutive different preparations. Infectivity assays were performed according to the method of Benzinger et al.3 with the following modifications: DNA was adsorbed to spheroplasts in 15-mm tubes for only 8 min without agitation and 0.5% Povite albumin was added to the sucrose soft agar. To obtain large distinct plaques, all assays were plated onto EMB bottom agar' containing 1%