Production of highly knotted DNA by means of cosmid circularization inside phage capsids

Abstract

Background: The formation of DNA knots is common during biological transactions. Yet, functional implications of knotted DNA are not fully understood. Moreover, potential applications of DNA molecules condensed by means of knotting remain to be explored. A convenient method to produce abundant highly knotted DNA would be highly valuable for these studies.

Results: We had previously shown that circularization of the 11.2 kb linear DNA of phage P4 inside its viral capsid generates complex knots by the effect of confinement. We demonstrate here that this mechanism is not restricted to the viral genome. We constructed DNA cosmids as small as 5 kb and introduced them inside P4 capsids. Such cosmids were then recovered as a complex mixture of highly knotted DNA circles. Over 250 mug of knotted cosmid were typically obtained from 1 liter of bacterial culture.

Conclusion: With this biological system, DNA molecules of varying length and sequence can be shaped into very complex and heterogeneous knotted forms. These molecules can be produced in preparative amounts suitable for systematic studies and applications.

Figure 1

P4 phage cosmids. (a) A cosmid containing the cos sequence of phage P4 is cleaved and threaded into a viral capsid in the course of a bacterial phage infection. Circularization of the cosmid by its cohesive ends inside the phage capsid produce a highly knotted DNA circle. DNA knots are removed by topoisomerase II. (b) Scheme of circularized P4 DNA and of cosmids P4cos-8, P4cos-5 and P4cos-3, in which an EcoRI-BamH1 1189 bp fragment containing the P4-cos sequence is inserted.

Figure 2

Packaging efficiency and knotting probability of cosmid DNA in P4 phage particles. DNA was extracted from either no cosmid (a), P4cos-8 (b), P4cos-5 (c) and P4cos-3 (d). Samples were loaded in the gel before and after unknotting the DNA with topoisomerase II (lanes 1 and 2, respectivelly, in a-d). Gel electrophoresis were done in 0.6% agarose gels with TBE buffer (1.4 volt/cm × 30 hours) in a and b; and in 0.8% agarose gels with TBE buffer (1.6 volt/cm × 20 hours) in c and d. Lambda DNA digested with HindIII was loaded in lane 0 in b, c, d. Nicked-circular (C_COS) and linear (L_COS) forms of each corresponding cosmid, produced by restriction endonuclease treatment of the supercoiled plasmids, were loaded as markers in lane 3 in b, c, d. Positions of the unknotted nicked-circle (C_P4) and the linear form (L_P4) of P4 DNA are indicated.

Figure 3

Analysis of cosmid knots by two-dimensional gel electrophoresis. DNA extracted from P4 vir1 del22 viral particles, which were amplified in bacteria that harboured P4cos-5, was electrophoresed in a gel slab of 0.4% agarose equilibrated with TBE buffer. The first gel dimension (top to bottom) was at 0.8 V/cm for 36 h at room temperature. The second gel dimension (left to right) was done in the same electrophoresis buffer at 3.4 V/cm for 4 h at room temperature. (a) Ethidium bromide staining of the gel. (b) The gel blotted to a nylon membrane and probed for P4cos-5. (c) Scheme showing the gel positions of linear molecules (L), unknotted circles (C), and knotted circles (K) of P4 DNA (white) and P4cos-5 (black). Dimeric P4cos-5 molecules (2 × 4.7 Kb) would ran nearly as P4 DNA (10 kb).

Figure 4

Knot distribution of cosmid DNA. (a) P4 viral particles containing P4cos-5 were purified by equilibrium density centrifugation. Extracted DNA was then electrophoresed in a 0.8% agarose gel equilibrated with TBE. The first dimension (top to bottom) was at 0.8 V/cm for 40 h at room temperature. The second dimension (left to right) was done in the same buffer at 3.4 V/cm for 4 h at room temperature. The gel blot was probed for P4cos-5. Positions of linear molecules (L), unknotted circles (C), individual knot populations containing three to eight crossings (3–8), and the tail of more complex knots (K) are indicated. (b) The electrophoresis velocity at low voltage (first gel dimension) is projected to estimate the complexity of the knot distribution. The histogram plots relative amounts of knot populations of increasing complexity by considering a linear relation between their gel velocity and their crossing number. The position and ideal configuration of the achiral knot of four crossings (4₁) is indicated.