Вы находитесь на странице: 1из 8

Volume 13 Number 18 1985

Nucleic Acids Research

Knotftng of DNA molecules isolated from deletion mutants of intact bacteriophage P4

John S.Wolfson*, Gail L.McHugh, David C.Hooper and Morton N.Swartz

Infectious Disease Unit, Medical Services, Massachusetts General Hospital, Boston, MA 02114, USA
Received 18 July 1985; Revised and Accepted 27 August 1985 ABSTRACT DNA molecules isolated from tailless phage particles (capsids) of bacteriophage P4 virl dellO are known to be knotted. We have found by electron microscopy that 80Z of DNA molecules isolated from intact phage particles of P4 virl dellO also contained knots. This observation indicates that the predominant form of P4 virl dellO DNA within the intact phage particle is either knotted or in a configuration that permits knotting upon isolation. In comparison to P4 virl dellO (deleted 1000 basepairs), DNA molecules isolated from intact P4 virl del2 (deleted 650 basepairs) and P4 virl (non-deleted) contained 50% and 15Z knots respectively, showing an association of decreased size of deletion of DNA with a decreased fraction of knotted genomes.

INTRODUCTION The genome of the icosahedral bacteriophage P4 is an 11.5 kilobase (kb) non-permuted duplex linear DNA molecule with 19 base single-stranded cohesive ends (1,2). Lytic growth of P4 is dependent upon bacteriophage P2 (3), which contains a 35 kb duplex linear DNA molecule (1) that encodes all of the resolved structural proteins in the P4 phage particle (4). DNA molecules isolated from tailless phage particles (capsids) of either the P4 deletion mutant P4 virl dellO (5) or non-deleted P2 (6) are unusual in that a large fraction exists as highly knotted circles. In contrast, DNA molecules isolated from intact phage particles of P4 virl dellO, though less thoroughly characterized, have a smaller fraction with knots (5); and molecules isolated from intact P2 phage particles are unknotted (6). We report here experiments using electron microscopy to quantitate in detail the fraction of knotted DNA molecules isolated from intact phage particles of P4 virl dellO. We have found that the majority of isolated P4 virl dellO DNA molecules are knotted and that the fraction of knotted molecules is substantially less for a P4 genome with a smaller or no deletion. These findings are discussed in relation to packaging of DNA in bacteriophages.

I RL Press Limited, Oxford, England.


Nucleic Acids Research


Bacterial strains and bacteriophage P4 phages were propagated on Escherichia coli strains C-117(P2) (7) and C-1792(P2 lg cc) (8). E. coli C-1055 and C-1792(P2 1g cc) were used as indicator strains (9). P4 virl dellO (10) was prepared with the following modification of the protocol of Liu et al. (5). Phages were added at a multiplicity of infection of 0.02 to cells growing in early logarithmic phase. When cell growth reached a plateau as determined by optical density (A595), EGTA (ethylene glycol-bis-(beta-aminoethyl ether) N,N'-tetracetic acid, Sigma Chemical Co.) was added to 5 mM, and incubation was continued for 1 hour. Phages were precipitated with polyethylene glycol (PEG-6000), and the pellet was extracted twice with P4 buffer (11). The phages were then concentrated by cesium chloride equilibrium centrifugation (a step which separates intact phages from both capsids and intracellular DNA) and dialyzed against P4 buffer. Purification typically generated phage titers of 1-5 x 1012 plaque-forming units per ml. Single plaques of P4 virl and P4 virl del2 (12,13) were used to infect cultures of C-1792(P2 lg cc). The virl mutation of P4 virl causes partial insensitivity to P4 immunity (3,14,15), probably by enhancing expression of the alpha and epsilon genes (16), and thus favors lytic growth. The identities of P4 virl, P4 virl del2, and P4 virl dellO were confirmed by restriction enzyme analyses. Isolation of DNA Phage DNA was isolated from phage particles by extraction three times with redistilled phenol (Bethesda Research Laboratories, Inc.) and once with diethyl ether and was then dialyzed against P4 buffer. T4 DNA ligase (Bethesda Research Laboratories, Inc.) was used according to the recommendations of the manufacturer. Electron microscopy For electron microscopy of phage particles, a phage stock of 1-5 x 1012 plaque-forming units per ml was diluted 1:20 in P4 buffer and applied in a 20 4l drop to a Parlodian-coated 200 mesh copper grid. After the drop was blotted, the grid was stained with 0.05 M uranyl acetate (Fisher Scientific) in 0.05 M HC1, and the specimen was examined immediately in an electron microscope (JEOL lOOB, 80 kV, 60 micron aperture). DNA was prepared for electron microscopy using the formamide technique of Davis, Simon, and Davidson (17), stained with 0.05 M uranyl acetate in 0.05 M HC1, and examined in the dark-field mode. Gel electrophoresis Gel electrophoresis proceeded for four hours at 4.2 V/cm in 0.8%

Nucleic Acids Research

agarose (Sigma Chemical Co.) and buffer containing 89 mM Tris-2.8 mM EDTA-89 mM boric acid in a horizontal gel apparatus (DanKar, Inc.). DNA was stained and photographed as previously described (18).
RESULTS Studies with P4 virl dellO A preparation of bacteriophage P4 virl dellO was examined in the electron microscope and found to contain 84% intact phage particles and 16% capsids. DNA molecules extracted from the same phage stock were examined by electron microscopy using 50% formamide in the spreading solution and 20% formamide in the hypophase. Of 642 molecules scored, 62% were knotted forms, 12% were highly condensed dot-like structures (thought to be knotted, poorly extended molecules), 22% were unknotted circular or linear molecules, and 4% could not be categorized. For the knotted molecules, two-thirds had no ends protruding from the knot, and one-third had two ends protruding. The two protruding ends likely represented separation of the cohesive ends during preparation for electron microscopy. To determine if both the knotted molecules and the dot-like structures were composed of a single molecular species, the DNA preparation was heated to 65C for 10 minutes, thereby separating the cohesive ends and allowing unknotting. This heat treatment resulted in the conversion of the DNA to a single population of unit-sized unknotted linear molecules as analyzed both by electron microscopy and by agarose gel electrophoresis (data not shown). To substantiate that the highly condensed dot-like structures (Figure 1, arrow) were knotted molecules, extracted DNA was prepared for electron microscopy using enhanced denaturing conditions (60% formamide in the spreading solution and 30% formamide in the hypophase) to encourage better extension of condensed DNA. Prior to preparation of DNA for electron microscopy, the cohesive ends of molecules were sealed with T4 DNA ligase to inhibit separation of the ends and unknotting; success of ligation was confirmed by the failure of heat treatment to produce linear molecules as analyzed by agarose gel electrophoresis (data not shown). When the ligated DNA was examined by electron microscopy (Figure 1), for 559 molecules scored, 76% were knotted with no ends protruding, 18% were circular or linear unknotted molecules, 4% were condensed structures, and 2% could not be categorized. No knotted molecules had linear ends protruding. Knotted molecules appeared in a continuous spectrum from condensed structures with one small loop of DNA protruding to well-displayed molecules. These


Nucleic Acids Research

Figure 1. Electron micrograph of P4 virl dellO DNA.


DNA was extracted from

purified phage particles and ligated prior to preparation for electron

The arrow points to a dot-like structure of highly condensed DNA.

results provide evidence that the condensed dot-like structures are knotted molecules. Thus highly knotted DNA molecules were the predominant form (80%) of genomes isolated from a phage stock with 84% intact particles. Studies with P4 virl and P4 virl del2 To determine whether the knotting of DNA isolated from the intact P4 phage particle might be a general characteristic of all P4 phages or represent a special property of P4 virl dellO, the parent phage P4 virl was studied. P4 virl dellO has a deletion of 1000 basepairs (19), and P4 virl has no deletion. Fifteen percent of phenol-extracted DNA molecules from a phage stock containing 90% intact P4 virl particles were knotted by electron microscopy. Thus the fraction of knotted DNA molecules isolated from the parent phage was substantially less than that of DNA molecules isolated from the deleted derivative. Similar experiments were performed using phage P4 virl del2, a P4 virl derivative which contains a 650 basepair deletion that does not overlap the deletion in P4 virl dellO (12,19). Fifty percent of the DNA molecules

Nucleic Acids Research

extracted from a phage stock containing 88% intact particles were knotted as viewed by electron microscopy. Thus phages with a deletion that was intermediate in size between P4 virl dellO and the parent P4 virl had an intermediate fraction of knotted molecules. All experiments described above were performed at least twice. Age of a phage preparation did not alter the fraction of knotted molecules. Concurrent analyses of phage DNA by agarose gel electrophoresis suggested that the diffuse smear of knotted DNA previously reported (5) increased in conjunction with the increased fraction of knotted molecules seen by electron microscopy, but these findings were not further quantitated.
DISCUSSION Our studies indicate that the predominant form of DNA isolated from intact P4 virl dellO phage particles is a complex knot. The majority of DNA molecules isolated from particles of one other virus, the cauliflower mosaic virus, are also known to be knotted (20,21,22,23). We found the fraction of knotted P4 virl dellO DNA molecules extracted from intact phage particles to be substantially higher than the 20% that was estimated previously (5). This difference likely resulted from differences either in methods of quantitation (formamide vs. aqueous preparation of DNA for electron microscopy; electron microscopy vs. agarose gel electrophoresis) or in methods of DNA isolation. The increased fraction of knotted molecules seen with the formamide method of preparation for electron microscopy might represent preferential retention of knotted forms in the protein film; this possiblity seems unlikely, however, because linear molecules generated by heat treatment of the DNA prior to spreading appeared to be present in the same high concentration as the knotted forms seen prior to heat treatment. How and why knotted viral DNA is formed is unknown. For P4 virl dellO, whether the DNA molecule within the phage particle is knotted or is packaged such that knotting occurs upon isolation is also not known. If one assumes that infectious P4 particles contain knotted DNA, unknotting might occur either during injection of DNA or, following injection, by the action of an unknotting enzyme such as DNA gyrase (24). Alternatively, phage particles containing knotted DNA may not be infectious. However, because P4 virl dellO and P4 virl del2 give larger average phage burst sizes in a shorter period of time than P4 virl, it would seem less likely that knotting is associated with non-viable phage (Richard Goldstein,


Nucleic Acids Research

personal communication). How might knotting relate to current understanding of the arrangement of DNA inside the head of P4? Packing of DNA in the heads of isometric phages such as lambda, P22, P2, P4, and T4 approximates that of a string tightly wound in a concentric spool with adjacent strands in a parallel orientation (25,26,27,28). In addition, for lambda, P2, and P4, a unique end of the DNA molecule is in proximity to the phage tail (29,30,31,32,33). Harrison (34) has incorporated these observations into a model for packing of DNA in an unknotted form which minimizes the chance of entanglement during unpacking. Central to his model is the anchoring of the leading end of the linear DNA molecule near the entry point into the head. The DNA molecule then enters the phage head parallel to the long axis of the particle and adds successive windings to the distal end of the growing spool. This model could account for formation of complex knots such as those isolated from tailless particles of P4 and P2, if one postulates spatial proximity of the leading and trailing ends of the DNA molecule packed in the phage head; during extraction of the DNA from the phage particle, cohesive ends could pass through the DNA spool and produce a knot upon basepairing (a possibility originally pointed out by R. Stroud (35)]. A similar sequence of events could generate our findings for intact P4 virl dellO phages. It has been noted (6) that the anchoring of an end of the DNA molecule in the P2 phage particle (30) may block joining of the cohesive ends and thus prevent the formation of knotted molecules. For the three P4 phage mutants studied, analysis by electron microscopy revealed that a decrease in the fraction of knotted DNA molecules isolated was associated with a decrease in size of deletion. Such a correlation might be expected. DNA molecules of phages lambda and P22, and presumably P2 and P4, are known to be tightly packed inside the phage head (27,28). For packaged deleted genomes of lambda, however, the distance between adjacent DNA helices is increased (28), and the motion of DNA is greater (36,37) than that of packaged wild-type genomes, indicating less constraint of deleted DNA within the phage particle. Such reduced constraint. if present in deleted P4 DNA within particles, might facilitate entanglement of strands and thus lead to increased knotting. An alternative explanation for variations in knotting is that deleted sequences either code for factors or provide for binding of factors which decrease knot formation; this explanation seems less likely, however, because sequences deleted in P4 virl dellO and P4 virl del2 do not overlap


Nucleic Acids Research

but still were associated with an increased fraction of knotted molecules. An important question raised in recent studies of lambda and other phages is whether interparticle variation in packaging of DNA molecules exists within a population of genetically homogeneous phage particles (35,36,38,39). It seems possible that the marked heterogeneity in extension of highly knotted P4 DNA molecules observed by electron microscopy (Figure 1) represents variation among DNA molecules in complexity of knotting, and that such variation is a reflection of heterogeneity among phage particles in packaging of DNA.
ACKNOWLEDGEMENTS We wish to thank Richard Goldstein and David Kurnit for helpful discussions and critical reviews of the manuscript, Roger C. Levesque for assistance with restriction enzyme analyses, and Richard Goldstein, Rosalba Lagos, Tad Goto, and James C. Wang for bacteriophages and bacterial strains. We are grateful to Romaine R. Bruns and Jerome Gross of the Lovett Memorial Group Developmental Biology Laboratory, Massachusetts General Hospital, for the use of their electron microscope facility (supported in part by Program Project Am 3564 from the National Institute of Arthritis, Diabetes, Digestive, and Kidney Diseases). Emma Tenierello provided valuable technical assistance. JSW was a Fellow of the Charles A. King Trust, Boston, Massachusetts. This work was supported in part by New Investigator Research Award 1R23 AI19452 from the National Institutes of Allergy and Infectious Diseases to JSW.

*To whom correspondence should be addressed




3. 4. 5.

7. 8. 9.

Inman, R.B., Schnos, M., Simon, L.D., Six, E.W. and Walker, D.H. (1971) Virology 44, 67-72. Wang, J.C., Martin, K.V. and Calendar, R. (1973) Biochemistry 12, 2119-2123. Six, E.W. and Klug, C.A.C. (1973) Virology 51, 327-344. Six, E.W. (1975) Virology 67, 249-263. Liu, L.F., Davis, J.L. and Calendar, R. (1981) Nucleic Acids Res. 9, 3979-3989. Liu, L.F., Perkocha, L., Calendar, R. and Wang, J.C. (1981) Proc. Natl. Acad. Sci. USA 78, 5498-5502. Bertani, L.E. (1964) Biochim. Biophys. Acta 87, 631-640. Lagos, R. and Goldstein, R. (1984) J. Bacteriol. 158, 208-214. Wiman, M., Bertani, G., Kelly, B. and Sasaki, I. (1970) Molec. Gen. Genet. 107, 1-31.


Nucleic Acids Research


12. 13.


16. 17. 18.

21. 22.

23. 24. 25.


28. 29.
30. 31. 32. 33.

34. 35. 36. 37.

38. 39.

Kahn, M., Ow, D., Sauer, B., Rabinowitz, A. and Calendar, R. (1980) Molec. Gen. Genet. 177, 399-412. Shore, D., Deho, G., Tsipis, J. and Goldstein, R. (1978) Proc. Natl. Acad. Sci. USA 75, 400-404. Kahn, M. and Hopkins, A. (1978) Virology 85, 359-363. Souza, L., Geisselsoder, J., Hopkins, A. and Calendar, R. (1978) Virology 85, 335-342. Calendar, R., Ljundquist, E., Deho, G., Usher, D., Goldstein, R., Youdarian, P., Sironi, G. and Six, E. (1981) Virology 113, 20-38. Goldstein R., Sedivy, J. and Ljundquist, E. (1982) Proc. Natl. Acad. Sci. USA 79, 515-519. Lin, C. (184) Nucleic Acids Res. 12, 8667-8684. Davis, R.W., Simon, M. and Davidson, N. (1971) In Methods in Enzymology, eds., Grossman, L. and Moldave, L., Vol. 21, pp. 413-428, Academic Press, New York. Hooper, D.C., Wolfson, J.S., McHugh, G.L., Winters, M.B. and Swartz, M.N. (1982) Antimicrob. Agents Chemother. 22, 662-671. Raimondi, A., Donghi, R., Montaguti, A., Pessina, A. and Deho, G. (1985) J. Virol. 54, 233-235. Menissier, J., de Murcia, G., Lebeurier, G. and Hirth, L. (1983) EMBO J. 2, 1067-1071. Civerolo, E.L. and Lawson, R.H. (1978) Phytopathology 68, 101-109. Volovitch, M., Drugeon, G. and Yot, P. (1978) Nucleics Acid Res. 5, 2913-2925. Richins, R.D. and Shepherd, R.J. (1983) Virology 124, 208-214. Gellert, M. (1981) Ann. Rev. Biochem. 50, 879-910. Richards, K.E., Williams, R.C. and Calendar, R. (1973) J. Mol. Biol. 78, 255-259. Earnshaw, W.C., King, J., Harrison, S.C. and Eiserling, F.A. (1978) Cell 14, 559-568. Earnshaw, W.C., Casjens, S.R. and Harrison, S.C. (1976) J. Mol. Biol. 104, 387-410. Earnshaw, W.C. and Harrison, S.C. (1977) Nature (London) 268, 598-602. Padmanabhan, R., Wu, R. and Bode, V.C. (1972) J. Mol. Biol. 69, 201-207. Chattoraj, D.K. and Inman, R.B. (1974) J. Mol. Biol. 87, 11-22. Saigo, K. and Uchida, H. (1974) Virology 61, 524-536. Thomas, J.0. (1974) J. Mol. Biol. 87, 1-9. Thomas, J.0., Sternberg, N. and Weisberg, R. (1978) J. Mol. Biol. 123, 149-161. Harrison, S.C. (1983) J. Mol. Biol. 171, 577-580. Haas, R., Murphy, R.F. and Cantor, C.R. (1982) J. Mol. Biol. 159, 71-92. Widom, J. and Baldwin, R.L. (1983) J. Mol. Biol. 171, 419-437. Ashikawa, I., Furuno, T., Kinosita, K. Ikegami, A., Takahashi, H. and Akutsu, H. (1984) J. Biol. Chem. 259, 8338-8344. Hall, S.B. and Schellman, J.A. (1982) Biopolymers 21, 1991-2010. Hall, S.B. and Schellman, J.A. (1982) Biopolymers 21, 2011-2031.