UV irradiation impairs in vivo encapsidation of bacteriophage T4 DNA
Journal of Virology
T4 DNA structural requirements for encapsidation in vivo were investigated, using thin-section electron microscopy to quantitate the kinetics and yields of head intermediates after synchronous DNA packaging into accumulated processed proheads. UV irradiation (254 nm) of T4-infected bacteria just before initiation of encapsidation resulted in a reduction in the rate of DNA packaged measured by electron microscopy and in the yield of viable phage progeny. In UV-irradiated infections with
... tions with excision-deficient mutants (denV-), the extent of packaging decline was proportional to the UV dose and phage yields were lower than expected based on the packaging levels observed by microscopy. Rescue analysis of progeny from such infections revealed elevated levels of nonviable virions. Pyrimidine dimers were encapsidated in denV-infections, but in excision-competent infections (denV+) dimers were not packaged. A UV-independent, 15 to 20% packaging arrest was also observed when denV endonuclease was inactive during encapsidation, indicating a denV requirement to achieve normal T4 packaging levels. Pyrimidine dimers apparently represent or induce transient blockage of DNA encapsidation or both, causing a decline in the rate. This is in contrast to other DNA structural blocks to packaging induced by mutations in T4 genes 30 and 49, which appear to arrest the process. DNA encapsidation in bacteriophage T4 development involves linkage of newly replicated concatemeric DNA to a functional cleaved prohead by phage-specified linkage/packaging proteins and subsequent translocation of the DNA into the prohead (for review, see reference 5). Many mutations have been characterized which prevent DNA packaging as a consequence of alterations in the structure of the prohead or in the linkage/packaging proteins. However, only two examples of mutations inducing changes in DNA structure which interfere with encapsidation have been demonstrated. Mutations in T4 gene 49 (endonuclease VII) result in the intracellular accumulation of expanded, partially DNA-filled head structures attached to concatemeric, very fast sedimenting DNA (6, 9, 10, 16, (18) (19) (20) (21) (22) 43) . In vitro the gene 49 nuclease cleaves very fast sedimenting DNA to form slower sedimenting linear concatemeric molecules (17, 30) and can resolve Holliday structures in DNA (27). In vivo gene 49 nuclease apparently cleaves recombination-derived branched strands in concatemeric DNA which would otherwise prevent DNA encapsidation (16, (25) (26) (27) . Mutations in T4 gene 30 (DNA ligase) and in Escherichia coli DNA ligase, like gene 49 mutations, reversibly arrest T4 DNA packaging (7, 13, 45) . The ligase function may be necessary to seal nicks in the DNA concatemer which presumably preclude encapsidation, and ligase could also function as a part of the packaging mechanism and therefore be required during the translocation process itself (6). In this study we found that after UV irradiation of replicated concatemeric T4 DNA in vivo there is an impairment of its subsequent encapsidation compared with unirradiated DNA. This packaging impairment was most evident as a decline in the rate of packaging and was shown to be due largely to the formation and repair of pyrimidine dimers in T4 DNA. We have also found a UV-independent packaging impairment associated with mutations in the T4 gene denV. These results provide support for the hypothesis that the rate * Corresponding author. of T4 DNA encapsidation in vivo is influenced by DNA structure and by processes, such as repair, which modify DNA structure. MATERIALS AND METHODS Bacteria and bacteriophage. E. coli B40 (SU1+), an ambersuppressing strain, was used in experiments involving electron microscopy and marker rescue; E. coli BE, an ambernonsuppressing strain, was used in experiments involving radioactive labeling and marker rescue; and E. coli AS19 (kindly provided by L. Gold), a rifampin-permeable strain, was used for rifampin experiments. The following T4 strains were used in this study: T4D+; cs2O(N33) ts2l(N12) amt(A3); cs20(N33) ts2l(N12) amt(A3) denV,; am23(H11) am23(B17); am55(BL292); cs2O(N33) tsdenV(431) [tsdenV(431) mutant kindly provided by M. Sekiguchi]. Hbroth was used for growth and phage dilution medium, and M9A was used for labeling experiments (8). Temperature shift experiments for electron microscopy. E. coli B40 was grown in H-broth at 37°C to a concentration of 2 x 108 cells per ml. The culture was then transferred to 20°C and infected with the desired phage at a multiplicity of infection of 5. After 22 min, the cells were superinfected at a multiplicity of infection of 5. After 85 min at 20'C, the infected cells maintained at 4°C were centrifuged and suspended in phosphate buffer (8), exposed to UV irradiation, and then centrifuged and suspended in H-broth. Samples were removed for assay of infective centers or for fixation for electron microscopy, and the culture was transferred to 42°C. Samples for phage yield or for fixation were removed at various times after transfer to 42°C. When the inhibitor rifampin (Sigma Chemical Co.) was used, it was added (final concentration, 200 ,ug/ml) 5 min before UV exposure and shift to 42°C. Each experiment was repeated two to five times, and results presented are mean values for the replicates. Temperature shift experiments for radiolabeling. E. coli BE was grown in M9A at 37°C to a concentration of 3 x 108 cells per ml. Deoxyadenosine (250 ,ug/ml) and [methVl-3H]thy-293 on May 9, 2020 by guest http://jvi.asm.org/ Downloaded from 294 ZACHARY AND BLACK FIG. 1. Thin sections through UV-irradiated (51 J/m2) excision repair-competent (V+) and -deficient (V-) T4-infected E. coli B40 cells, from a representative temperature shift experiment. (A) V+ infection at 10 mps showing EH, PFH, and FH, indicating that packaging has occurred. Prohead I structures (PHI) formed after shift to 42°C cannot mature because of the 21 ts mutation. (B) V+ infection at 20 mps showing that EH and PFH have been converted to FH. (C) V-infection at 10 mps; predominance of EH (some are expanded) indicates reduced packaging. Arrow indicates a tailed EH. (D) Vinfection at 30 mps. Presence of several DNA containing heads indicates that some packaging has occurred; however, many EH remain. Magnification is the same in all micrographs; bar in (C) = 100 nm. mine (5 XCi!ml) were added, and the culture was grown for 30 min at 37°C. The culture was transferred to 20°C, Ltryptophan (40 ,ug/ml) was added, and the cells were infected with the desired phage at a multiplicity of infection of 5. After 20 min, [methyl-3H]thymidine (5 ,uCi/ml) was added, and after 22 min the cells were superinfected at a multiplicity of infection of 5. After 85 min at 20°C, the infected cells, maintained at 4°C, were centrifuged and suspended in phosphate buffer, exposed to UV irradiation, centrifuged and suspended in unlabeled M9A, and then transferred to 42°C. After 30 min at 42°C, the cells were lysed with chloroform and treated with DNase (100 p.g/ml; Sigma) for 30 min at 37°C, and the phage were purified by differential centrifugation followed by CsCl step gradient centrifugation. Preparation of DNA. CsCl step gradient-purified phage were extracted with an equal volume of redistilled phenol (saturated with buffer and adjusted to pH 7.4) and 0.5 volume of chloroform-isoamyl alcohol (24:1, vol/vol) until the aqueous layer cleared after low-speed centrifugation. The aqueous layer was dialyzed against endonuclease buffer (see below). DNA solutions were stored at 4°C over chloroform. Pyrimidine dimer assay. Pyrimidine dimers in DNA can be detected as sites sensitive to nicking by the endonuclease activity of phage T4 gp denV (11, 42) . H-labeled DNA (>103 cpm), purified from phages produced in temperature shift experiments, was suspended in endonuclease buffer (10 mM Tris-chloride, pH 8.0, 10 mM Na2 EDTA, pH 8.0, 100 mM NaCl) and treated with 1 pd of purified T4 gp denV (kindly provided by E. H. Radany; activity, 1.83 pmol of photoreleasable thymine per min at 37°C; see reference 23 for details) for 15 min at 37°C. Duplicate phage DNA samples were treated with buffer only. The reaction was stopped by addition of 0.2 N NaOH for 15 min at room temperature. The DNA was then layered on 5 to 20% alkaline sucrose (0.15 M NaCl, 0.2 N NaOH) gradients and sedimented for 80 min at 45,000 rpm and 4°C in an SW50.1 rotor. Gradient samples were collected from a puncture in the bottom of the tube directly into ACS (Amersham Corp.) scintillation cocktail and counted in a Packard scintillation J. VIROL. on May 9, 2020 by guest http://jvi.asm.org/ Downloaded from UV IMPAIRMENT OF DNA PACKAGING 295 I 51 102 153 L (100) 100 l O 51 102 153 U.V DOSE (J/m2) FIG. 2. Effect of UV dosage on packaging of T4 DNA in V+ and V-infections at 9 mps. Hatched bars represent PFH, solid bars represent FH, and solid dots represent head structures containing some DNA (FH plus PFH). For all microscopy data, 400 to 600 head structures were counted to determine each value (see text). counter. The number of endonuclease-sensitive sites (ESS) was calculated by the method of Ahmed and Setlow (1). For controls, unirradiated T4 DNA and T4 DNA exposed to UV irradiation after extraction were analyzed for ESS in the same manner. UV irradiation. Samples were irradiated with UV light at 254 nm, using a GE15T8 germicidal lamp (General Electric Co.). Dosage was measured with a IBLAK RAY UV meter (U-V Products, J-225). Electron microscopy. Samples were partially lysed by addition of osmium tetroxide (final concentration, 0.01%) for 2 min and then fixed with phosphate-buffered (8) glutaraldehyde (final concentration, 0.25%) for 55 min at 4°C. The glutaraldehyde-fixed cell suspensions were centrifuged and the pellets were overlaid with 1% buffered osmium tetroxide for 18 h at 4°C. The samples were then dehydrated, using a graded acetone series, and embedded in Epon 812 epoxy plastic. Thin sections were stained with uranyl acetate and lead citrate before examination with a Siemens 1A electron microscope. To quantitate the types of head-related structures in thin sections, each cell profile entering the electron microscope viewing field was examined and all clearly distinguishable profiles of head-related structures were scored. For each data point, 400 to 600 capsid profiles were counted. Statistical significance (99% level) was determined by chi-square analysis. Some imprecision may be inherent in the counting method, because structures containing very small amounts of DNA could have been considered empty and structures nearly filled with DNA could have been considered full head-related structures. However, in duplicate counts, i.e., the same sample counted on different occasions, the variability was <2%. Defective phage rescue. Phage yields from temperature shift experiments were measured as plaques formed on lawns of E. coli B40, an amber-suppressing host. To determine whether defective phages (i.e., complete DNA-filled virions with a genomic defect which rendered them undetectable by plaque assay) were produced after UV Uf) w r 0 H cn 0 I 0 H CY) a w 0 w 0l lJ L 100-* (31) 50-100 102 J/m2 50 -I 100-(31) 50-(80) 10 20 30 51 J/m2 % (2) 10 20 30 102 J/m2 0 (0.4) 0 pi 0 1 20 30 0 10 20 30 MINUTES AFTER START OF PACKAGING (SHIFT TO 42°) FIG. 3. Effect of UV irradiation on packaging kinetics for V+ and Vinfections. Numbers in parentheses are viable phage yields at 30 mps expressed as a percentage of the yield for an unirradiated V+ infection. Hatched bars represent PFH, solid bars represent FH, and solid dots represent head structures containing some DNA (FH and PFH). irradiation, the phage yields were also plated on lawns of amber-nonsuppressing bacteria (E. coli BE) coinfected with T4 phages having two amber mutations in gene 23 [am23(H11), am23(B17)]. These double-amber phages, although not competent to form a plaque, could provide complementary gene products and thus rescue defective phages if present in the yields from the temperature shift experiments. The phage yields were also plated on lawns of uninfected E. coli BE as a control for background plaques expected due to the leakiness reported for the t amber mutation at 37°C (15) as well as any amber revertant formation. T4D+, a control for plating efficiency, and T4 am55 phages, a control for the efficiency of rescue of an amber L) w I-I--C], 0 w w I 0 LL 0 0--100-90-80-70-60-50-40-30-20-IO-0-IṼ OL. 50, 1984 on May 9, 2020 by guest http://jvi.asm.org/ Downloaded from 296 ZACHARY AND BLACK mutation in our system, were plated in the same manner as the phage yields. Plaques were counted after 18 h of incubation at 37°C. RESULTS in this study we used a previously developed electron microscopic technique which allows assessment of the kinetics and extent of DNA packaging in vivo, independent of early steps in head assembly and of the ultimate infectivity of the phage produced (45). Growth at 20°C of cells infected with a cs2O ts2l mutant allows normal expression of T4 functions (including the ts mutant gene product function) up to the point of DNA encapsidation. Normal DNA and late protein synthesis occur, leading to intracellular accumulation of empty head structures (EH), but no DNA-containing head structures or viable phage accumulate (14, 45). After temperature shift from 20 to 42°C, the cs20 block to prohead function is released, DNA packaging is initiated synchronously, and a rapid transition of EH formed at 20°C to heads partially filled with DNA (PFH) and then to heads completely filled with DNA (FH) occurs (45). A representative micrograph from one experiment (Fig. 1A) shows examples of the head structures described above. The kinetics and levels of encapsidation can be measured as differences in the proportion of the three types of head structures (EH, PFH, and FH) seen intracellularly after shift to 42°C and are consistent with kinetic measurements of phage yields (6, 13, 45), i.e., the half-time for packaging is 5 min after temperature shift. Packaging and phage completion after shift to 42°C occur in the presence of chloramphenicol; thus, continued protein synthesis is not required for phage maturation (14) . Any head structures assembled after shift to 42°C are blocked in maturation due to ts gp2l at the prohead I stage (cf. Fig. 1A ). These core-containing, membrane-bound, morphologically distinct prohead I structures cannot package DNA and therefore are not enumerated. Effect of UV dose on DNA packaging. T4 infections involving excision repair-competent bacteriophage (cs2O ts2l amt) will be referred to as V+ infections and excision repairdeficient phage infections (cs2O ts2l amt denV,) will be termed V-infections. -E. coli B40 infected with V+ phage were grown at 20°C for 85 min, exposed to various doses of UV irradiation, and then shifted to 42°C to initiate DNA packaging. An assessment of DNA packaging was made at 9 min postshift (mps) to 42°C by examining the proportions of intracellular head structures which had accumulated in thinsectioned cells. Unirradiated V+-infected cells showed extensive packaging, with 80% of the head structures containing some DNA: 19% PFH and 61% FH (Fig. 2) . Infected cells receiving a UV dose of 51 J/m2 had roughly 25% fewer DNA-containing heads: 40% PFH and 22% FH (Fig. 2) . At a UV dosage of 102 J/m2 there was no significant change in the percentage of DNA-containing head structures, but there was a small, statistically significant decline in the proportion of FH (from 22 to 10%); at a dose of 153 J/m2 no further impairment of packaging was observed (Fig. 2) . For Vinfections receiving UV irradiation, we observed substantially less packaging than in corresponding V' infections, and the declines in Vpackaging were more nearly proportional to increases in the UV dosage used. As dosage increased from 51 to 102 to 153 J/m2, the amount of DNAcontaining heads declined from 41 to 24 to 15%, respectively, at 9 mps (Fig. 2) . Thus, at the highest dosage (153 J/m2) at 9 mps, there were one-fourth as many DNA-containing heads in the Vas in the V' infection, indicating a clear difference in packaging of UV-irradiated DNA (Fig. 2) . TABLE 1. Data from kinetic studies' Packaging Yield % Defective Phage UV dose (% V-FH) (% V+ PFU) virions 10 mps 30 mps at 30 mps (due to UV)Y V+ 0 100 100 100 0 V+ 51 54 66 31 15 V+ 102 45 70 31 NOC V-0 71 84 80 0 V-51 25 60 2 48 V-102 8 26 0.04 NO a Packaging and yield data are compared with an unirradiated excision-competent (V+) infection. b Taken from Table 2. NO, No observation. Kinetic studies. UV-induced lesions in DNA might either remain as permanent blocks to packaging or be repaired after some delay, thus allowing DNA to be encapsidated. To test these possibilities, temperature shift experiments were conducted in which the UV dosage was 0, 51, or 102 J/m2, and samples taken at 10 and 30 mps to 42°C were examined by electron microscopy. Viable phage yields were measured at 30 mps. In irradiated V+-infected cells at 10 mps, the UV impairment of packaging was evident as a reduced level of DNA-containing heads and an increased proportion of PFH, and the decline in packaging was more evident as the UV dosage increased (Fig. 1A and 3) . However, this dosedependent difference in packaging was no longer evident at 30 mps, when packaging levels for both dosages were essentially the same (Fig. 3) . In both cases packaging continued in the latter 20 min but complete recovery of packaging to the unirradiated level was not observed ( Fig. 1B and 3) . The yield of viable phage at both UV dosages was 31% of that in the unirradiated V+ infection (Table 1; Fig. 3 ). The UV-induced packaging impairment was greater in the V-infection; at 10 mps, after a dose of 51 J/m, fewer than 50% of the heads contained DNA and only 15% were FH ( Fig. 1C and 3) . At 30 mps we observed an apparent TABLE 2. Rescue of plaque-forming ability to show presence of defective phage progeny in yields from UV-irradiated (51 J/m2) infections' Rescued back-Nonviable (% Input on ground (PFU) on: Phage UV B40Rescued -Due to PF) am23 BE background!