Proc. Nati. Acad. Sci. USA Vol. 75, No. 12, pp 6144-6148, December 1978
Genetics
Marker rescue and partial replication of bacteriophage T7 DNA (DNA synthesis/UV irradiation/cross-reactivation)
KATHY BAUMAN BURCK AND ROBERT C. MILLER, JR. Department of Microbiology, University of British Columbia, Vancouver, British Columbia, Canada, V6T 1W5
Communicated by A. H. Doermann, August 21,1978
ABSTRACT Experiments reported here show that some UV-irradiated wild-type 17 phage markers can be rescued efficiently by coinfection with 17 amber mutant phage in a permissive host. Other results show that the segments of a UV-irradiated genome that replicate efficiently are those that also are rescued efficiently during a marker rescue experiment. At higher doses, fewer markers are rescued efficienty and fewer segments of the genome replicate efficiently. The ruts clearly indicte that the probability of marker rescue is coreated with the ability of the DNA containing the marker to replicate. Sucrose density gradient analysis shows that UV irradiation does not produce double-strand scissions in 17 DNA at doses used here. Therefore, the partial replication and rescue of markers from the left end of thegenome is not due simply to injection of only the left end of the Ti DNA. Cross-reactivation refers to the ability of a genetically marked phage to rescue a particular locus from a coinfecting, UV-irradiated phage. Womack (1) has documented differences in the ability of various markers to be rescued from a UV-irradiated wild-type T4D (T4D+) phage genome. When UV-irradiated T4D+ phage were coinfected with various T4D mutants, the ability of a particular mutant to rescue the damaged T4D+ genome was dependent on the map position of the mutant. Recent publications have suggested that there may be a correlation between origins of DNA replication and the ability of a particular segment of the genome to be efficiently rescued (2,3). It has been postulated that reactivation occurs as a result of recombination between partial replicas of the genome (4, 5). Partial replicas of the UV-irradiated T4D+ genomes are thought to arise as a result of the inability of the replication process to pass through a UV-damaged area. Therefore, the farther a segment of the genome is from an origin of replication, the higher is the likelihood of an intervening area of UV damage, and the lower is the likelihood of being replicated. Consequently, the likelihood of rescue of a marker far from an origin would be lower, also. Several experiments are consistent with this hypothesis. The number of peaks in the distribution of marker rescue in the Womack experiment is consistent with the number of origins imagined to operate during T4 replication (1, 6, 7). One of the peaks is near gene 43, and this is an area of the map reported to contain an origin of replication (8). Rayssiguier and Vigier (3) have analyzed the recombinant clone size distribution of progeny phage produced by multiplicity reactivation of genetically marked, UV-irradiated parental phage. This analysis is consistent with the idea that partial replicas of the damaged genomes reassociate by recombining primarily at their extremities. An origin of replication of phage TI7 DNA has been mapped by electron microscopic analysis of partially replicated T7 DNA (9). This makes it possible to examine the correlation between the efficiency of rescue of a segment of UV-irradiated wild-type
T7 (T7+) DNA and its proximity to a known origin of T7+ replication. Furthermore, McDonell et al. (10) have aligned restriction fragments of the T7 genome with specific regions of the genetic map. Therefore, 32P-labeled, T7+ progeny DNA made by a UV-irradiated T7+ phage can be annealed with known restriction fragments of DNA to examine the correlation between the segments that replicate and the segments that are rescued efficiently. The results of such an analysis indicate clearly that the segments of the UV-irradiated T7 DNA that replicate are the segments that can be rescued. As the dose of UV irradiation is increased, fewer segments replicate and fewer segments are rescued. MATERIALS AND METHODS Bacteria. Escherichia colh B23 (su) does not suppress amber mutations in essential genes. E. coli BR3 (sul) (11) and E. coil 011' (su+) (12) have been described. E. coli BR3-4 is a singlecolony isolate of E. coil BR3. Bacteriophage. All bacteriophage mutants were generously provided by F. W. Studier. The various strains used here were T7+, 1am193, Iam323, Iam342a, 2am64, 3am29, 4am208, 5am28, 6am233, 8amll, 1lam37, 14amI40, 16am194, 17am290, 19amlO (12), 0.7amJs62a, T7 deletion mutant ALG-3 (11), and 0.3amCrlOb (13). Namx is the general designation for amber mutants, with N the gene number and x the specific mutant. Media and Buffers. Tris/Casamino acids/glucose medium (TCG) and H-broth have been described (14,15). Tris/NaCl/ EDTA buffer (TNE) contains 10 mM Tris-HCl, 150 mM NaCl, and 15 mM EDTA at pH 7.4. Preincubation mixture (PM) has been described (16). T7 Tris/salt is 1 M NaCl/50 mM Tris-HCl, pH 7.4. Preparation of 32P- and 3H-labeled T7 bacteriophage, alkaline sucrose gradient sedimentation, CsCl density gradient sedimentation, DNA extraction and purification, and determination of label incorporation into acid-insoluble material have been described (17). Ultraviolet Irradiation of Phage. Phage to be irradiated were diluted to 1 X 101" bacteriophage per ml in T7 Tris/salt and placed in a plastic petri dish on ice. Irradiation was for 10, 20, or 30 sec (T7+) and 30 sec (T7 ALG-3) from a distance of 30 cm by a General Electric G15T8 15-W germicidal lamp. Lethal events were quantitated by plotting survival curves of the irradiated phage for each experiment. The survival curve of UV-irradiated T7+ is biphasic. This phenomenon has been observed by several groups (18-20). The maximum phage fraction showing a lower sensitivity to UV comprises only about 3% of the population. Because the phage suspension is composed of two subpopulations, each distinguished by a unique and sharply defined exponential survival curve, the 3% subpopulation will not influence the data to be presented in a significant
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U. S. C. §1734 solely to indicate this fact.
Abbreviations: T4D+ and 17+, wild-type strains of bacteriophage T4D and T7; TCG, Tris/Casamino acids/glucose; TNE, Tris/NaCI/EDTA; PM, preincubation mixture.
6144
Genetics: Burck and Miller
way, as will be seen in the Results section. The slope of the curve representing the "resistant fraction" is about half that-of the "sensitive" fraction. Preparation of Filters Charged with Gel-Separated ¶7+ DNA Cut by Endo R Fnu C. T7 DNA was digested overnight *with restriction endonuclease Fnu C. Endo R-Fnu C has the same specificity as endo R-L pn II and endo R-Mbo I (A. Lui and M. Smith, personal communication). The fragments were electrophoresed at 70-80 V for 7-9 hr in 1.5-mm, 1% agarose gels containing 90 mM Tris, 90 mM boric acid, 25 mM EDTA at pH 8.3, and ethidium bromide at 1 ttg/ml. Only fragments A-E are retained on the gel after electrophoresis under these conditions. The procedure for transferring the fragments to nitrocellulose was modified from that of Southern (21). DNA in the gels was denatured for 30-0 min (in a buffer containing 0.2 M NaOH, 0.6 M NaCl, and thymol blue as a pH indicator), rinsed briefly in water, and neutralized by soaking 30-60 min in 1 M TrisHCI, pH 7.4/1.5 M NaCI.
Proc. Natl. Acad. Sci. USA 75 (1978)
6145
.0
1.1 0.
o..5
1.
.0
.5 ._4
E to
0
0*
x
CDL0
cJ
o-B
.0 o5
CD
RESULTS Marker rescue with UV-irradiated T7 phage Marker rescue experiments with bacteriophage T7 similar to those of Womack (1) with bacteriophage T4 were conducted in the following manner. Permissive E. coli were coinfected with UV-irradiated T7 bacteriophage and one of a set of defined amber mutants. The infected cells were scored for the production of wild-type infective centers from the mixed infection. Fig. IA illustrates the efficiency of rescue of UV-irradiated T7+ DNA carrying 6.5 lethal hits by particular mutants. At this UV dose, rescue of markers from 0 to 30% on the T7 genome occurs with at least 8 to 11 times the frequency of marker rescue from the region of 50 to 100%. Markers in the region 50-100% of the genome are rescued at a frequency not significantly above that of background infective center production by UV-irradiated phage alone. The maximum marker rescue observed represented about 20% of the infective centers plated on the permissive (su+) host, 011'; i.e., the maximum marker rescue was close to 4 X 107 infective centers per ml. Because the rescue frequencies are very high compared to the maximum fraction of T7+ phage having a lower sensitivity to UV irradiation, the biphasic nature of the T7+ UV survival curve does not interfere with subsequent interpretations of the results (see Materials and Methods). The curves were very similar if the infective centers were diluted, incubated until lysis, and titered for yield of T7+ progeny phage. These data are in accord with the hypothesis that DNA replication initiates at the 17% region and proceeds bidirectionally until interrupted by a UV lesion on the left and/or a UV lesion on the right. Replication facilitates rescue of the partially replicated region. Nonreplicated regions are not rescued to any appreciable extent. If the radiation dose to T7+ is increased to 9.5 lethal hits per phage particle, markers along the leftmost 10% of the molecule are rescued with an efficiency at least 8 to 13 times greater than markers on the right end and 4 to 6 times greater than that of markers in the left 10-40% of the molecule (Fig. 1C). Intermediate doses-e.g., 7.5 lethal hits (Fig. 1B)-give intermediate curves. There are three possible interpretations of these data. (i) Replication is initiating from the left end of the T7 genome, and the efficiency of subsequent rescue is the same for any partially replicated piece. (ii) Initiation of replication proceeds from the 17% region but the left end has greater efficiency of rescue. (iii) There is a second internal origin around 5-10% from the left hand end. The deletion mutant ALG-3 is missing the DNA between
C 1.0-
0.5-
D
0
II
if
0.31 llc 13
0.71l1b
2
. 4
I
a
Is
.. 1
11'7 1
14
11
16
19
la
Map position I I
I I
I
I
I
1I
0
20
40
60
80
I
100
Percent length of genome
FIG. 1. Rescue of T7 amber mutant phage with UV-irradiated T7 phage. E. coli 011' was grown to 2 X 108 bacteria per ml in H-broth at 30°C. Chloramphenicol was added at 100 mg/ml to the culture to inhibit superinfection exclusion (22). At 1-min intervals aliquots of the cells were coinfected with UV-irradiated T7 phage and one of a series of T7 amber phage. The multiplicity of infection for each phage type was close to 5 phage per bacterium and was monitored by plating the surviving bacteria. (Surviving bacteria were monitored in parallel samples infected with UV-irradiated T7+ alone and with T7 mutants alone.) Ten minutes after infection an aliquot of the infected cells was diluted into T7 antiserum, incubated 5 min at 37°C, diluted further, and plated for T7+ infective centers on a nonpermissive host. All of the amber phage represented on any particular graph were examined in a single experiment. Infective center production is not significantly influenced by incubation in chloramphenicol of up to 30 min. la, Ib, and ic stand for ambers 1-193, 1-323, and 1-342a, respectively. (A) T7+ irradiated to 6.5 phage-lethal events. Infective centers plated on BR3-4. The dotted line represents the level of infective center production by UV-irradiated phage alone. (B) T7+ irradiated to 7.5 phage-lethal events. Infective centers plated on BR3-4. (C) T7+ irradiated to 9.5 phage-lethal events. Infective centers plated on BR3-4. (D) T7 ALG-3 irradiated to 6.9 phage-lethal events. Infective centers plated on B23.
15.2 and 19.2% on the T7 map; this is the region containing the presumed initiation site (9, 11). ALG-3 UV-irradiated to 6.9 phage-lethal events was coinfected with a set of amber mutants in an experiment similar to that described in Fig. IA. Rescue of markers on the left 40% of the genome is still appreciable (Fig. 1D); however, rescue of the leftmost gene 1 mutants is 3 times greater than mutants in the 10-40% region giving the curve the overall appearance of Fig. 1C. When the 17% origin region is deleted, rescue of markers from the left end is very
6146
Genetics: Burck and Miller
IAA
37 1
1 A
A
a b
A
c
1.31 AAi 1
4
Proc. Natl. Acad. Sci. USA 75 (1978)
1
15 12 8 179loll 11 1611 7 111 A A A A A A A
5 A
A
T7 genes
me... LG-3
tI
1
-0
B I 10
G I 20
I
I
I
I
D
II
E
I 30
F
A
I 40
I 50
I
60
70
I 80
Fnu C or Dpn 11 C
II 90
T7 units
100
FIG. 2. Map of the fragments generated from T7+ DNA by cutting with endo R-Dpn II, taken from McDonell et al. (10). Endo R.Fnu C has the same specificity as endo R-Dpn II (M. Smith and A. Lui, personal communication). The position of the deletion LG-3 is indicated by the broken line. The arrow points to the 17% region containing the presumed origin site. The small triangles mark the positions of the particular amber mutants used in the marker rescue experiments.
high, rescue of genes in the area of the deleted origin is depressed, and right-hand markers plate at the background level. These data imply that the absence of an origin of replication at 17% depresses marker rescue in this area because partial replication of these segments is decreased. However, it is possible that the deletion interferes directly with recombination events essential for marker rescue in this area. This is unlikely, because the marker rescue efficiencies are very similar if the co-infections instead are with amber-ALG-3 double mutants and UV-irradiated ALG-3, compared to parallel coinfections of amber and UV-irradiated ALG-3 phage (data not shown). The efficient marker rescue at the left of the UV-irradiated deletion mutant implies that replication may originate at or near the left end of UV-irradiated T7 DNA when the 17% origin is deleted.
Annealing of partially replicated T7 DNA with T7+
DNA cut by restriction enzyme Fnu C The five large restriction fragments generated by endo R-Fnu C digestion of T7+ DNA form a convenient group for the study of partially replicated T7 DNA. The fragments generated by endo R-Dpn II digestion of T7+ DNA have been mapped (Fig. 2) (10), and endo R-Dpn II has the same specificity as endo R-Fnu C. The pattern obtained by annealing 32P-labeled DNA extracted from T7+ phage to T7+ fragments on nitrocellulose filters is illustrated in Fig. 3A. Quantitation of the bands on an autoradiogram by densitometer scanning (Fig. 3A) reveals that
the fractional contribution of annealed [32P]DNA in each band is close to the fractional contribution by weight of the fragment.
This allows one to examine the specificity of 32P-labeled progeny DNA synthesized by parental phage receiving various doses of UV irradiation. The annealing pattern for DNA extracted from cells infected with UV-irradiated T7+ phage containing 0, 4.6, 7.0, and 9.4 phage lethal hits was examined (Fig. 3). At the lowest dose (Fig. 3C), the relative amounts of 32P-labeled progeny DNA hybridizing in bands B, D, and E increase compared to the non-UV control. These segments are from the left end of the T7 genome. Annealing in bands C and A on the right end of the genome decreases. At increasing UV doses, these effects become even more pronounced (Fig. 3D and E). By 9.4 lethal hits per phage particle (Fig. 3E) annealing of intracellular DNA is to the B, D, and E fragments and not at all to A and C. These results show that UV-damaged T7 phage replicate the left end of the DNA molecule preferentially and that the extent of replication is dependent on the density of UV damage. In other words, the 17% replication origin is apparently located
near the right end of fragment B (Fig. 2); this implies that if replication proceeds in two directions from this point to the nearest site of UV damage, then the replication of segments B, D, and E should be most resistant to UV irradiation. This is in fact the case. The results are in agreement With the rescue data presented above; segment B is replicated the most efficiently and the genes contained in segment B are rescued most efficiently. Hybridization of intracellular 32P-labeled DNA from nonirradiated T7+ phage is illustrated in Fig. 3B. The relative amounts of annealing in these bands do not correspond exactly with the molecular weights of the T7+ fragments. The amount of fragment A is slightly depressed. This result suggests that initiation at the 17% region produces slightly more of the B, D, and E segments in vwo during normal wild-type infection; that is , a few molecules at the time of isolation of T7+ progeny [32P]DNA were partially replicated. The amount of the labeled C fragment appears to he slightly higher than its fractional contribution to the T7 genome. This is consistent with progeny DNA forming head-to-tail concatemers that replicate bidirectionally (23-25). Fragment C then would be replicated by a growing fork moving to the left'i* the T7 genome and would appear in a higher relative am6toi in a partially replicated molecule. There may appear to be more replf4tion of the C fragment than one might expect when the parenr~ phage has received 4.6 lethal hits (Fig. 3C). There are several possible explanations for this apparent resistance of replication to UV irradiation. First, formation of concatemers could make partial replication of the C fragment more resistant to UV than the position of the C fragment might indicate at first glance. Concatemers might arise as a result of multiple infection by UV-irradiated T7+ phage or might arise directly from the fraction of T7+ phage having a lower sensitivity to UV irradiation (see Materials and Methods). The slope of the UV inactivation curve representing the fraction of T7+ phage having a lower sensitivity to UV irradiation might be partly due to that fraction containing multiple genomes. These genomes could be similar to the "giant" phage of T4, which contain concatemeric DNA (26). If these T7+ multiple genomes were arranged in the normal head-to-tail arrangement (23-25), then they presumably could replicate the C fragment as described above. A second possibility is that some phage-lethal events between 17% and the C fragment do not block all 32P-labeled nucleotide incorporation in the C fragment. Yet another possibility is that another origin of replication exists near or in the C fragment that may be used infrequently by a UV-irradiated T7+ genome.
Genetics: Burck and Miller
Proc. Natl. Acad. Sci. USA 75 (1978)
6147
FIG. 4. Size of UV-irradiated, parental T7 DNA. E. coli B23 was grown to 3 X 108 bacteria per ml in TCG containing 6.5 mM KH2PO4 at 300C. The culture then was infected with UV-irradiated, 3H-labeled T7+ phage receiving 7.0 lethal hits at a multiplicity of infection of 5. At 5 and 20 min after infection, aliquots were withdrawn into 21/2 vol of ice-cold TNE, sedimented, and resuspended at a concentration of 6 X 108/ml. Infected cells were lysed with sodium dodecyl sulfate, treated with Pronase, and extracted with phenol. Samples were mixed with reference [32PIDNA extracted from T7+ phage and sedimented through sucrose density gradients at neutral and alkaline pH. Samples on alkaline gradients were denatured in 0.2 M NaOH prior to centrifugation. Approximately 30 fractions were collected dropwise from the bottom of the gradient and assayed for radioactivity. (A) Fiveminute infective centers, neutral pH; (B) 20-min infective centers, neutral pH; (C) 5-min infective centers, alkaline pH; (D) 20-min infective centers, alkaline pH. 0, 3H-Labeled intracellular DNA; *, 32P-labeled reference DNA from T7+ phage particles. ED C
B A Fnm C fragment
FIG. 3. Annealing of partially replicated DNA with endo R-Fnu C restriction fragments. E. coli B23 was grown to 3 X 108 bacteria/ml in TCG containing 0.13 mM KH2PO4 at 300C and infected with T7+ at a multiplicity of infection of 5. Eight minutes after infection 500 jtCi of 32p (1 Ci = 3.7 X 1010 becquerels) was added to the 2-ml culture. The culture was chilled by the addition of 3 vol of ice-cold TNE 17 min later. The infected cells were sedimented and resuspended in 1/2 vol of cold TNE. The cells were lysed with sodium dodecyl sulfate, treated with Pronase, and extracted with phenol. Samples were sedimented to equilibrium in CsCl density gradients. Approximately 50 fractions were collected dropwise from the bottom of the tube; aliquots were precipitated with trichloroacetic acid and assayed for radioactivity. Fractions containing the DNA were pooled and dialyzed against 50 mM Tris-HCl, pH 7.4/150 mM NaCl/2 mM EDTA. This 32P-labeled material was >986 resistant to digestion with 1.0 M KOH for 18 hr at 370C. The nitrocellulose filters containing the T7 restriction fragments were preincubated 6 hr in 40 ml of PM at 650C. [32P]DNA was sonicated, heat denatured, and ice chilled, then added to the hybridization mixture. Hybridization proceeded overnight (12-16 hr) at 650C. Filters then were washed in 300 mM NaCl/30 mM Na citrate four times and exposed for autoradiography on Kodak X-Omat R XR2 film. Developed films were scanned with a Quick Scan Jr. densitometer, Helena Laboratories. (A) Annealing of 32P -labeled DNA extracted from T7+ phage. (B) Annealing of 32P-labeled replicative DNA extracted from cells infected with nonirradiated T7+ phage. (C) Annealing of 32P-labeled replicative DNA extracted from cells infected with UV-irradiated T7+ phage receiving 4.6 lethal hits. (D) Annealing of 32P-labeled replicative DNA extracted from cells infected with UV-irradiated T7+ phage receiving 7.0 lethal hits. (E) Annealing of 32P-labeled replicative DNA extracted from cells infected with UV-irradiated T7+ phage receiving 9.9 lethal hits.
Size of intracellular UV-irradiated parental 17 DNA The size of the DNA recovered from E. coli infected by UVdamaged T7+ phage has been examined by sucrose gradient sedimentation (Fig. 4). Five minutes after mixing phage and bacteria, 3H-labeled, UV-irradiated T7+ DNA cosediments with 32P-labeled DNA extracted from T7+ phage at both neutral (Fig. 4A) and alkaline (Fig. 4C) pH. Similar results are obtained for the nonirradiated control. Therefore, the results reported in Fig. 1 and 3 cannot be explained by preferential injection of the left end of UV-damaged T7 DNA. After 20 min, UV-irradiated T7+ parental DNA remains size 1 in the double-stranded state (Fig. 4B) but single-strand breaks have been introduced (Fig. 4D). It has been reported that T7 injects the left end of the genome first (27). If UV irradiation led to breakage of the T7 DNA, one might argue that only the left end of the molecule could be rescued because only the left end could be injected and replicate. This figure shows that UV-irradiated parental phage DNA recovered from infected cells is not significantly different from unirradiated DNA extracted from phage particles. We have no evidence that UV at the doses used here breaks DNA. On the contrary, the evidence indicates that UV-irradiated parental DNA is not broken.
DISCUSSION The results presented in this paper support the basic idea of partial replica hypotheses (4, 5) as an explanation for the pattern of marker rescue frequencies during bacteriophage cross-
6148
Genetics: Burck and Miller
reactivation; that is, UV-irradiated phage DNA molecules replicate from an origin to a point of UV damage, are blocked from further replication, and are rescued most efficiently for markers in the segments of the genome that have replicated. Markers in the UV-irradiated T7+ genome are rescued in a specific pattern with regard to the genetic map (Fig. 1). Markers in the left end are rescued most efficiently, and markers in the right end are rescued less efficiently depending on how far to the right they are on the map. A mutation deleting the region containing a major origin of replication in the left end decreases marker rescue in that region. There are no major expansions or contractions of the genetic map (11-13) as determined by a comparison of physical and genetic characterizations of the T7 genome. Therefore, it is reasonable to assume that the marker rescue pattern reflects differences in the ability of specific segments of the genome to replicate. Different segments of the UV-irradiated T7+ DNA do replicate with different efficiencies. Segments in the left end of the molecule replicate most efficiently, whereas segments toward the right end replicate least efficiently (Fig. 3). The differences in the abilities of various segments to replicate are accentuated by increasing the dose of UV irradiation. Therefore, partial replicas of UV-irradiated T7+ DNA do exist. The segments of the genome that replicate most efficiently are the segments that are rescued most efficiently. A numerical correlation between the efficiency with which a marker replicates and is rescued can be established in the following manner. To determine the relative efficiencies of the replication of the fragments, one can divide the fractional contribution of a UV-irradiated fragment by the fractional contribution of the same fragment in the nonirradiated control; the fractional contribution is defined as the area under the peak on the densitometer tracing of the particular hybridization band divided by the summed area of all the peaks. For example, one divides the area of fragment A in Fig. 3A by the total area under the curve of Fig. 3A. This is divided in turn by the ratio of the area of fragment A in Fig. 3D to the total area under the curve of Fig. 3D. A similar calculation is made for fragment B. Thus by dividing the calculation for fragment A replication by fragment B replication, one can determine the relative efficiency of replication. At a dose of 6.5 phage-lethal events the relative efficiency of replication of fragment A compared to B is 0.1, or 10% as efficient. If one compares the efficiency of rescue of a marker in the middle of fragment A (14am140) to a marker in the middle of fragment B (Iam193) at a similar dose of UV irradiation, the ratio is about 0.05. Similarly if one calculates the relative efficiency of replication of fragment E compared to fragment B, then E replicates about 84% as efficiently as B; a marker in E (5am28) is rescued with an efficiency of 0.5 (50%) compared to a marker in B (lam193). Therefore, the rescue of a marker is correlated with the efficiency of replication of the segment of the DNA containing that marker. It is not possible to say from these results whether replication of a segment is required absolutely before a marker in the segment can be rescued at all. These results show that the efficiency of repli-
Proc. Natl. Acad. Sci. USA 75 (1978)
cation is correlated with the efficiency of rescue, but it is difficult to determine whether some rescue occurs in the complete absence of replication; it is difficult because of the low background level of T7+ infective centers produced by UV-irradiated phage alone and because UV irradiation does not inhibit replication completely. Therefore, it is not possible at this time to say whether a special feature of replicating DNA is required for marker rescue. We thank D. M. Taylor and H. W. Smith for their excellent technical assistance. We are very grateful to A. Lui and M. Smith for supplying us with restriction enzyme Fnu C and determining its specificity. We thank C. Hutchinson for his aid in setting up the system for transferring DNA from agarose gels to nitrocellulose filters. This research was supported by grants from the National Research Council of Canada. 1. Womack, F. C. (1965) Virology 26,758-761. 2. Rayssiguier, C. & Vigier, P. R. R. (1972) Molec. Gen. Genet. 115, 140-145. 3. Rayssiguier, C. & Vigier, P. R. R. (1977) Virology 78, 442452. 4. Barricelli, N. A. & Doermann, A. H. (1961) Virology 13, 460476. 5. Doermann, A. H., Chase, M. & Stahl, F. W. (1955) J. Cell. Comp. Physiol. 45, Suppl., 41-74. 6. Delius, H., Howe, C. & Kozinski, A. W. (1971) Proc. Natl. Acad. Sci. USA 68,3049-3053. 7. Howe, C. C., Buckley, P. J., Carlson, K. & Kozinski, A. W. (1973) J. Virol. 12, 130-148. 8. Mosig, G. (1970) J. Mol. Biol. 53,503-514. 9. Dressler,, D., Wolfson, J. & Magazin, M. (1972) Proc. Natl. Acad. Sci. USA 69, 998-1002. 10. McDonell, M. W., Simon, M. N. & Studier, F. W. (1977) J. Mol. Biol. 110, 119-146. 11. Studier, F. W. (1973) J. Mol. Biol. 79,227-236. 12. Studier, F. W. (1969) Virology 39,562-574. 13. Studier, F. W. (1975) J. Mol. Biol. 94,283-295. 14. Kozinski, A. W. & Szybalski, W. (1959) Virology 9,260-274. 15. Steinberg, C. M. & Edgar, R. S. (1962) Genetics 47, 187-208. 16. Denhardt, D. T. (1966) Biochem. Biophys. Res. Commun. 23, 641-646. 17. Miller, R. C., Jr., Lee, M., Scraba, D. G. & Paetkau, V. (1976) J. Mol. Biol. 101, 223-234. 18. Dulbecco, R. J. (1959) J. Bacteriol. 39,329-347. 19. Ellison, S., Feiner, R. & Hill, R. F. (1960) Virology 11, 294296. 20. Kuemmerle, N. B. & Masker, W. (1977) J. Virol. 23,509-516. 21. Southern, E. M. (1975) J. Mol. Biol. 98,503-517. 22. Benbassat, J., Burck, K. B. & Miller, R. C., Jr. (1978) Virology 87, 164-171. 23. Schlegel, R. A. & Thomas, C. A., Jr. (1972) J. Mol. Biol. 68, 319-345. 24. Watson, J. D. (1972) Nature (London) New Biol. 239, 197201. 25. Langman, L., Paetkau, V., Scraba, D., Miller, R. C., Jr., Roeder, G. S. & Sadowski, P. D. (1978) Can. J. Biochem. 56,508-516. 26. Doermann, A. H., Eiserling, F. A. & Boehner, L. (1973) J. Virology 12, 374-385. 27. Pao, C. C. & Speyer, J. F. (1973) J. Virol. 11, 1024-1026.
Genetics
Marker rescue and partial replication of bacteriophage T7 DNA (DNA synthesis/UV irradiation/cross-reactivation)
KATHY BAUMAN BURCK AND ROBERT C. MILLER, JR. Department of Microbiology, University of British Columbia, Vancouver, British Columbia, Canada, V6T 1W5
Communicated by A. H. Doermann, August 21,1978
ABSTRACT Experiments reported here show that some UV-irradiated wild-type 17 phage markers can be rescued efficiently by coinfection with 17 amber mutant phage in a permissive host. Other results show that the segments of a UV-irradiated genome that replicate efficiently are those that also are rescued efficiently during a marker rescue experiment. At higher doses, fewer markers are rescued efficienty and fewer segments of the genome replicate efficiently. The ruts clearly indicte that the probability of marker rescue is coreated with the ability of the DNA containing the marker to replicate. Sucrose density gradient analysis shows that UV irradiation does not produce double-strand scissions in 17 DNA at doses used here. Therefore, the partial replication and rescue of markers from the left end of thegenome is not due simply to injection of only the left end of the Ti DNA. Cross-reactivation refers to the ability of a genetically marked phage to rescue a particular locus from a coinfecting, UV-irradiated phage. Womack (1) has documented differences in the ability of various markers to be rescued from a UV-irradiated wild-type T4D (T4D+) phage genome. When UV-irradiated T4D+ phage were coinfected with various T4D mutants, the ability of a particular mutant to rescue the damaged T4D+ genome was dependent on the map position of the mutant. Recent publications have suggested that there may be a correlation between origins of DNA replication and the ability of a particular segment of the genome to be efficiently rescued (2,3). It has been postulated that reactivation occurs as a result of recombination between partial replicas of the genome (4, 5). Partial replicas of the UV-irradiated T4D+ genomes are thought to arise as a result of the inability of the replication process to pass through a UV-damaged area. Therefore, the farther a segment of the genome is from an origin of replication, the higher is the likelihood of an intervening area of UV damage, and the lower is the likelihood of being replicated. Consequently, the likelihood of rescue of a marker far from an origin would be lower, also. Several experiments are consistent with this hypothesis. The number of peaks in the distribution of marker rescue in the Womack experiment is consistent with the number of origins imagined to operate during T4 replication (1, 6, 7). One of the peaks is near gene 43, and this is an area of the map reported to contain an origin of replication (8). Rayssiguier and Vigier (3) have analyzed the recombinant clone size distribution of progeny phage produced by multiplicity reactivation of genetically marked, UV-irradiated parental phage. This analysis is consistent with the idea that partial replicas of the damaged genomes reassociate by recombining primarily at their extremities. An origin of replication of phage TI7 DNA has been mapped by electron microscopic analysis of partially replicated T7 DNA (9). This makes it possible to examine the correlation between the efficiency of rescue of a segment of UV-irradiated wild-type
T7 (T7+) DNA and its proximity to a known origin of T7+ replication. Furthermore, McDonell et al. (10) have aligned restriction fragments of the T7 genome with specific regions of the genetic map. Therefore, 32P-labeled, T7+ progeny DNA made by a UV-irradiated T7+ phage can be annealed with known restriction fragments of DNA to examine the correlation between the segments that replicate and the segments that are rescued efficiently. The results of such an analysis indicate clearly that the segments of the UV-irradiated T7 DNA that replicate are the segments that can be rescued. As the dose of UV irradiation is increased, fewer segments replicate and fewer segments are rescued. MATERIALS AND METHODS Bacteria. Escherichia colh B23 (su) does not suppress amber mutations in essential genes. E. coli BR3 (sul) (11) and E. coil 011' (su+) (12) have been described. E. coli BR3-4 is a singlecolony isolate of E. coil BR3. Bacteriophage. All bacteriophage mutants were generously provided by F. W. Studier. The various strains used here were T7+, 1am193, Iam323, Iam342a, 2am64, 3am29, 4am208, 5am28, 6am233, 8amll, 1lam37, 14amI40, 16am194, 17am290, 19amlO (12), 0.7amJs62a, T7 deletion mutant ALG-3 (11), and 0.3amCrlOb (13). Namx is the general designation for amber mutants, with N the gene number and x the specific mutant. Media and Buffers. Tris/Casamino acids/glucose medium (TCG) and H-broth have been described (14,15). Tris/NaCl/ EDTA buffer (TNE) contains 10 mM Tris-HCl, 150 mM NaCl, and 15 mM EDTA at pH 7.4. Preincubation mixture (PM) has been described (16). T7 Tris/salt is 1 M NaCl/50 mM Tris-HCl, pH 7.4. Preparation of 32P- and 3H-labeled T7 bacteriophage, alkaline sucrose gradient sedimentation, CsCl density gradient sedimentation, DNA extraction and purification, and determination of label incorporation into acid-insoluble material have been described (17). Ultraviolet Irradiation of Phage. Phage to be irradiated were diluted to 1 X 101" bacteriophage per ml in T7 Tris/salt and placed in a plastic petri dish on ice. Irradiation was for 10, 20, or 30 sec (T7+) and 30 sec (T7 ALG-3) from a distance of 30 cm by a General Electric G15T8 15-W germicidal lamp. Lethal events were quantitated by plotting survival curves of the irradiated phage for each experiment. The survival curve of UV-irradiated T7+ is biphasic. This phenomenon has been observed by several groups (18-20). The maximum phage fraction showing a lower sensitivity to UV comprises only about 3% of the population. Because the phage suspension is composed of two subpopulations, each distinguished by a unique and sharply defined exponential survival curve, the 3% subpopulation will not influence the data to be presented in a significant
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U. S. C. §1734 solely to indicate this fact.
Abbreviations: T4D+ and 17+, wild-type strains of bacteriophage T4D and T7; TCG, Tris/Casamino acids/glucose; TNE, Tris/NaCI/EDTA; PM, preincubation mixture.
6144
Genetics: Burck and Miller
way, as will be seen in the Results section. The slope of the curve representing the "resistant fraction" is about half that-of the "sensitive" fraction. Preparation of Filters Charged with Gel-Separated ¶7+ DNA Cut by Endo R Fnu C. T7 DNA was digested overnight *with restriction endonuclease Fnu C. Endo R-Fnu C has the same specificity as endo R-L pn II and endo R-Mbo I (A. Lui and M. Smith, personal communication). The fragments were electrophoresed at 70-80 V for 7-9 hr in 1.5-mm, 1% agarose gels containing 90 mM Tris, 90 mM boric acid, 25 mM EDTA at pH 8.3, and ethidium bromide at 1 ttg/ml. Only fragments A-E are retained on the gel after electrophoresis under these conditions. The procedure for transferring the fragments to nitrocellulose was modified from that of Southern (21). DNA in the gels was denatured for 30-0 min (in a buffer containing 0.2 M NaOH, 0.6 M NaCl, and thymol blue as a pH indicator), rinsed briefly in water, and neutralized by soaking 30-60 min in 1 M TrisHCI, pH 7.4/1.5 M NaCI.
Proc. Natl. Acad. Sci. USA 75 (1978)
6145
.0
1.1 0.
o..5
1.
.0
.5 ._4
E to
0
0*
x
CDL0
cJ
o-B
.0 o5
CD
RESULTS Marker rescue with UV-irradiated T7 phage Marker rescue experiments with bacteriophage T7 similar to those of Womack (1) with bacteriophage T4 were conducted in the following manner. Permissive E. coli were coinfected with UV-irradiated T7 bacteriophage and one of a set of defined amber mutants. The infected cells were scored for the production of wild-type infective centers from the mixed infection. Fig. IA illustrates the efficiency of rescue of UV-irradiated T7+ DNA carrying 6.5 lethal hits by particular mutants. At this UV dose, rescue of markers from 0 to 30% on the T7 genome occurs with at least 8 to 11 times the frequency of marker rescue from the region of 50 to 100%. Markers in the region 50-100% of the genome are rescued at a frequency not significantly above that of background infective center production by UV-irradiated phage alone. The maximum marker rescue observed represented about 20% of the infective centers plated on the permissive (su+) host, 011'; i.e., the maximum marker rescue was close to 4 X 107 infective centers per ml. Because the rescue frequencies are very high compared to the maximum fraction of T7+ phage having a lower sensitivity to UV irradiation, the biphasic nature of the T7+ UV survival curve does not interfere with subsequent interpretations of the results (see Materials and Methods). The curves were very similar if the infective centers were diluted, incubated until lysis, and titered for yield of T7+ progeny phage. These data are in accord with the hypothesis that DNA replication initiates at the 17% region and proceeds bidirectionally until interrupted by a UV lesion on the left and/or a UV lesion on the right. Replication facilitates rescue of the partially replicated region. Nonreplicated regions are not rescued to any appreciable extent. If the radiation dose to T7+ is increased to 9.5 lethal hits per phage particle, markers along the leftmost 10% of the molecule are rescued with an efficiency at least 8 to 13 times greater than markers on the right end and 4 to 6 times greater than that of markers in the left 10-40% of the molecule (Fig. 1C). Intermediate doses-e.g., 7.5 lethal hits (Fig. 1B)-give intermediate curves. There are three possible interpretations of these data. (i) Replication is initiating from the left end of the T7 genome, and the efficiency of subsequent rescue is the same for any partially replicated piece. (ii) Initiation of replication proceeds from the 17% region but the left end has greater efficiency of rescue. (iii) There is a second internal origin around 5-10% from the left hand end. The deletion mutant ALG-3 is missing the DNA between
C 1.0-
0.5-
D
0
II
if
0.31 llc 13
0.71l1b
2
. 4
I
a
Is
.. 1
11'7 1
14
11
16
19
la
Map position I I
I I
I
I
I
1I
0
20
40
60
80
I
100
Percent length of genome
FIG. 1. Rescue of T7 amber mutant phage with UV-irradiated T7 phage. E. coli 011' was grown to 2 X 108 bacteria per ml in H-broth at 30°C. Chloramphenicol was added at 100 mg/ml to the culture to inhibit superinfection exclusion (22). At 1-min intervals aliquots of the cells were coinfected with UV-irradiated T7 phage and one of a series of T7 amber phage. The multiplicity of infection for each phage type was close to 5 phage per bacterium and was monitored by plating the surviving bacteria. (Surviving bacteria were monitored in parallel samples infected with UV-irradiated T7+ alone and with T7 mutants alone.) Ten minutes after infection an aliquot of the infected cells was diluted into T7 antiserum, incubated 5 min at 37°C, diluted further, and plated for T7+ infective centers on a nonpermissive host. All of the amber phage represented on any particular graph were examined in a single experiment. Infective center production is not significantly influenced by incubation in chloramphenicol of up to 30 min. la, Ib, and ic stand for ambers 1-193, 1-323, and 1-342a, respectively. (A) T7+ irradiated to 6.5 phage-lethal events. Infective centers plated on BR3-4. The dotted line represents the level of infective center production by UV-irradiated phage alone. (B) T7+ irradiated to 7.5 phage-lethal events. Infective centers plated on BR3-4. (C) T7+ irradiated to 9.5 phage-lethal events. Infective centers plated on BR3-4. (D) T7 ALG-3 irradiated to 6.9 phage-lethal events. Infective centers plated on B23.
15.2 and 19.2% on the T7 map; this is the region containing the presumed initiation site (9, 11). ALG-3 UV-irradiated to 6.9 phage-lethal events was coinfected with a set of amber mutants in an experiment similar to that described in Fig. IA. Rescue of markers on the left 40% of the genome is still appreciable (Fig. 1D); however, rescue of the leftmost gene 1 mutants is 3 times greater than mutants in the 10-40% region giving the curve the overall appearance of Fig. 1C. When the 17% origin region is deleted, rescue of markers from the left end is very
6146
Genetics: Burck and Miller
IAA
37 1
1 A
A
a b
A
c
1.31 AAi 1
4
Proc. Natl. Acad. Sci. USA 75 (1978)
1
15 12 8 179loll 11 1611 7 111 A A A A A A A
5 A
A
T7 genes
me... LG-3
tI
1
-0
B I 10
G I 20
I
I
I
I
D
II
E
I 30
F
A
I 40
I 50
I
60
70
I 80
Fnu C or Dpn 11 C
II 90
T7 units
100
FIG. 2. Map of the fragments generated from T7+ DNA by cutting with endo R-Dpn II, taken from McDonell et al. (10). Endo R.Fnu C has the same specificity as endo R-Dpn II (M. Smith and A. Lui, personal communication). The position of the deletion LG-3 is indicated by the broken line. The arrow points to the 17% region containing the presumed origin site. The small triangles mark the positions of the particular amber mutants used in the marker rescue experiments.
high, rescue of genes in the area of the deleted origin is depressed, and right-hand markers plate at the background level. These data imply that the absence of an origin of replication at 17% depresses marker rescue in this area because partial replication of these segments is decreased. However, it is possible that the deletion interferes directly with recombination events essential for marker rescue in this area. This is unlikely, because the marker rescue efficiencies are very similar if the co-infections instead are with amber-ALG-3 double mutants and UV-irradiated ALG-3, compared to parallel coinfections of amber and UV-irradiated ALG-3 phage (data not shown). The efficient marker rescue at the left of the UV-irradiated deletion mutant implies that replication may originate at or near the left end of UV-irradiated T7 DNA when the 17% origin is deleted.
Annealing of partially replicated T7 DNA with T7+
DNA cut by restriction enzyme Fnu C The five large restriction fragments generated by endo R-Fnu C digestion of T7+ DNA form a convenient group for the study of partially replicated T7 DNA. The fragments generated by endo R-Dpn II digestion of T7+ DNA have been mapped (Fig. 2) (10), and endo R-Dpn II has the same specificity as endo R-Fnu C. The pattern obtained by annealing 32P-labeled DNA extracted from T7+ phage to T7+ fragments on nitrocellulose filters is illustrated in Fig. 3A. Quantitation of the bands on an autoradiogram by densitometer scanning (Fig. 3A) reveals that
the fractional contribution of annealed [32P]DNA in each band is close to the fractional contribution by weight of the fragment.
This allows one to examine the specificity of 32P-labeled progeny DNA synthesized by parental phage receiving various doses of UV irradiation. The annealing pattern for DNA extracted from cells infected with UV-irradiated T7+ phage containing 0, 4.6, 7.0, and 9.4 phage lethal hits was examined (Fig. 3). At the lowest dose (Fig. 3C), the relative amounts of 32P-labeled progeny DNA hybridizing in bands B, D, and E increase compared to the non-UV control. These segments are from the left end of the T7 genome. Annealing in bands C and A on the right end of the genome decreases. At increasing UV doses, these effects become even more pronounced (Fig. 3D and E). By 9.4 lethal hits per phage particle (Fig. 3E) annealing of intracellular DNA is to the B, D, and E fragments and not at all to A and C. These results show that UV-damaged T7 phage replicate the left end of the DNA molecule preferentially and that the extent of replication is dependent on the density of UV damage. In other words, the 17% replication origin is apparently located
near the right end of fragment B (Fig. 2); this implies that if replication proceeds in two directions from this point to the nearest site of UV damage, then the replication of segments B, D, and E should be most resistant to UV irradiation. This is in fact the case. The results are in agreement With the rescue data presented above; segment B is replicated the most efficiently and the genes contained in segment B are rescued most efficiently. Hybridization of intracellular 32P-labeled DNA from nonirradiated T7+ phage is illustrated in Fig. 3B. The relative amounts of annealing in these bands do not correspond exactly with the molecular weights of the T7+ fragments. The amount of fragment A is slightly depressed. This result suggests that initiation at the 17% region produces slightly more of the B, D, and E segments in vwo during normal wild-type infection; that is , a few molecules at the time of isolation of T7+ progeny [32P]DNA were partially replicated. The amount of the labeled C fragment appears to he slightly higher than its fractional contribution to the T7 genome. This is consistent with progeny DNA forming head-to-tail concatemers that replicate bidirectionally (23-25). Fragment C then would be replicated by a growing fork moving to the left'i* the T7 genome and would appear in a higher relative am6toi in a partially replicated molecule. There may appear to be more replf4tion of the C fragment than one might expect when the parenr~ phage has received 4.6 lethal hits (Fig. 3C). There are several possible explanations for this apparent resistance of replication to UV irradiation. First, formation of concatemers could make partial replication of the C fragment more resistant to UV than the position of the C fragment might indicate at first glance. Concatemers might arise as a result of multiple infection by UV-irradiated T7+ phage or might arise directly from the fraction of T7+ phage having a lower sensitivity to UV irradiation (see Materials and Methods). The slope of the UV inactivation curve representing the fraction of T7+ phage having a lower sensitivity to UV irradiation might be partly due to that fraction containing multiple genomes. These genomes could be similar to the "giant" phage of T4, which contain concatemeric DNA (26). If these T7+ multiple genomes were arranged in the normal head-to-tail arrangement (23-25), then they presumably could replicate the C fragment as described above. A second possibility is that some phage-lethal events between 17% and the C fragment do not block all 32P-labeled nucleotide incorporation in the C fragment. Yet another possibility is that another origin of replication exists near or in the C fragment that may be used infrequently by a UV-irradiated T7+ genome.
Genetics: Burck and Miller
Proc. Natl. Acad. Sci. USA 75 (1978)
6147
FIG. 4. Size of UV-irradiated, parental T7 DNA. E. coli B23 was grown to 3 X 108 bacteria per ml in TCG containing 6.5 mM KH2PO4 at 300C. The culture then was infected with UV-irradiated, 3H-labeled T7+ phage receiving 7.0 lethal hits at a multiplicity of infection of 5. At 5 and 20 min after infection, aliquots were withdrawn into 21/2 vol of ice-cold TNE, sedimented, and resuspended at a concentration of 6 X 108/ml. Infected cells were lysed with sodium dodecyl sulfate, treated with Pronase, and extracted with phenol. Samples were mixed with reference [32PIDNA extracted from T7+ phage and sedimented through sucrose density gradients at neutral and alkaline pH. Samples on alkaline gradients were denatured in 0.2 M NaOH prior to centrifugation. Approximately 30 fractions were collected dropwise from the bottom of the gradient and assayed for radioactivity. (A) Fiveminute infective centers, neutral pH; (B) 20-min infective centers, neutral pH; (C) 5-min infective centers, alkaline pH; (D) 20-min infective centers, alkaline pH. 0, 3H-Labeled intracellular DNA; *, 32P-labeled reference DNA from T7+ phage particles. ED C
B A Fnm C fragment
FIG. 3. Annealing of partially replicated DNA with endo R-Fnu C restriction fragments. E. coli B23 was grown to 3 X 108 bacteria/ml in TCG containing 0.13 mM KH2PO4 at 300C and infected with T7+ at a multiplicity of infection of 5. Eight minutes after infection 500 jtCi of 32p (1 Ci = 3.7 X 1010 becquerels) was added to the 2-ml culture. The culture was chilled by the addition of 3 vol of ice-cold TNE 17 min later. The infected cells were sedimented and resuspended in 1/2 vol of cold TNE. The cells were lysed with sodium dodecyl sulfate, treated with Pronase, and extracted with phenol. Samples were sedimented to equilibrium in CsCl density gradients. Approximately 50 fractions were collected dropwise from the bottom of the tube; aliquots were precipitated with trichloroacetic acid and assayed for radioactivity. Fractions containing the DNA were pooled and dialyzed against 50 mM Tris-HCl, pH 7.4/150 mM NaCl/2 mM EDTA. This 32P-labeled material was >986 resistant to digestion with 1.0 M KOH for 18 hr at 370C. The nitrocellulose filters containing the T7 restriction fragments were preincubated 6 hr in 40 ml of PM at 650C. [32P]DNA was sonicated, heat denatured, and ice chilled, then added to the hybridization mixture. Hybridization proceeded overnight (12-16 hr) at 650C. Filters then were washed in 300 mM NaCl/30 mM Na citrate four times and exposed for autoradiography on Kodak X-Omat R XR2 film. Developed films were scanned with a Quick Scan Jr. densitometer, Helena Laboratories. (A) Annealing of 32P -labeled DNA extracted from T7+ phage. (B) Annealing of 32P-labeled replicative DNA extracted from cells infected with nonirradiated T7+ phage. (C) Annealing of 32P-labeled replicative DNA extracted from cells infected with UV-irradiated T7+ phage receiving 4.6 lethal hits. (D) Annealing of 32P-labeled replicative DNA extracted from cells infected with UV-irradiated T7+ phage receiving 7.0 lethal hits. (E) Annealing of 32P-labeled replicative DNA extracted from cells infected with UV-irradiated T7+ phage receiving 9.9 lethal hits.
Size of intracellular UV-irradiated parental 17 DNA The size of the DNA recovered from E. coli infected by UVdamaged T7+ phage has been examined by sucrose gradient sedimentation (Fig. 4). Five minutes after mixing phage and bacteria, 3H-labeled, UV-irradiated T7+ DNA cosediments with 32P-labeled DNA extracted from T7+ phage at both neutral (Fig. 4A) and alkaline (Fig. 4C) pH. Similar results are obtained for the nonirradiated control. Therefore, the results reported in Fig. 1 and 3 cannot be explained by preferential injection of the left end of UV-damaged T7 DNA. After 20 min, UV-irradiated T7+ parental DNA remains size 1 in the double-stranded state (Fig. 4B) but single-strand breaks have been introduced (Fig. 4D). It has been reported that T7 injects the left end of the genome first (27). If UV irradiation led to breakage of the T7 DNA, one might argue that only the left end of the molecule could be rescued because only the left end could be injected and replicate. This figure shows that UV-irradiated parental phage DNA recovered from infected cells is not significantly different from unirradiated DNA extracted from phage particles. We have no evidence that UV at the doses used here breaks DNA. On the contrary, the evidence indicates that UV-irradiated parental DNA is not broken.
DISCUSSION The results presented in this paper support the basic idea of partial replica hypotheses (4, 5) as an explanation for the pattern of marker rescue frequencies during bacteriophage cross-
6148
Genetics: Burck and Miller
reactivation; that is, UV-irradiated phage DNA molecules replicate from an origin to a point of UV damage, are blocked from further replication, and are rescued most efficiently for markers in the segments of the genome that have replicated. Markers in the UV-irradiated T7+ genome are rescued in a specific pattern with regard to the genetic map (Fig. 1). Markers in the left end are rescued most efficiently, and markers in the right end are rescued less efficiently depending on how far to the right they are on the map. A mutation deleting the region containing a major origin of replication in the left end decreases marker rescue in that region. There are no major expansions or contractions of the genetic map (11-13) as determined by a comparison of physical and genetic characterizations of the T7 genome. Therefore, it is reasonable to assume that the marker rescue pattern reflects differences in the ability of specific segments of the genome to replicate. Different segments of the UV-irradiated T7+ DNA do replicate with different efficiencies. Segments in the left end of the molecule replicate most efficiently, whereas segments toward the right end replicate least efficiently (Fig. 3). The differences in the abilities of various segments to replicate are accentuated by increasing the dose of UV irradiation. Therefore, partial replicas of UV-irradiated T7+ DNA do exist. The segments of the genome that replicate most efficiently are the segments that are rescued most efficiently. A numerical correlation between the efficiency with which a marker replicates and is rescued can be established in the following manner. To determine the relative efficiencies of the replication of the fragments, one can divide the fractional contribution of a UV-irradiated fragment by the fractional contribution of the same fragment in the nonirradiated control; the fractional contribution is defined as the area under the peak on the densitometer tracing of the particular hybridization band divided by the summed area of all the peaks. For example, one divides the area of fragment A in Fig. 3A by the total area under the curve of Fig. 3A. This is divided in turn by the ratio of the area of fragment A in Fig. 3D to the total area under the curve of Fig. 3D. A similar calculation is made for fragment B. Thus by dividing the calculation for fragment A replication by fragment B replication, one can determine the relative efficiency of replication. At a dose of 6.5 phage-lethal events the relative efficiency of replication of fragment A compared to B is 0.1, or 10% as efficient. If one compares the efficiency of rescue of a marker in the middle of fragment A (14am140) to a marker in the middle of fragment B (Iam193) at a similar dose of UV irradiation, the ratio is about 0.05. Similarly if one calculates the relative efficiency of replication of fragment E compared to fragment B, then E replicates about 84% as efficiently as B; a marker in E (5am28) is rescued with an efficiency of 0.5 (50%) compared to a marker in B (lam193). Therefore, the rescue of a marker is correlated with the efficiency of replication of the segment of the DNA containing that marker. It is not possible to say from these results whether replication of a segment is required absolutely before a marker in the segment can be rescued at all. These results show that the efficiency of repli-
Proc. Natl. Acad. Sci. USA 75 (1978)
cation is correlated with the efficiency of rescue, but it is difficult to determine whether some rescue occurs in the complete absence of replication; it is difficult because of the low background level of T7+ infective centers produced by UV-irradiated phage alone and because UV irradiation does not inhibit replication completely. Therefore, it is not possible at this time to say whether a special feature of replicating DNA is required for marker rescue. We thank D. M. Taylor and H. W. Smith for their excellent technical assistance. We are very grateful to A. Lui and M. Smith for supplying us with restriction enzyme Fnu C and determining its specificity. We thank C. Hutchinson for his aid in setting up the system for transferring DNA from agarose gels to nitrocellulose filters. This research was supported by grants from the National Research Council of Canada. 1. Womack, F. C. (1965) Virology 26,758-761. 2. Rayssiguier, C. & Vigier, P. R. R. (1972) Molec. Gen. Genet. 115, 140-145. 3. Rayssiguier, C. & Vigier, P. R. R. (1977) Virology 78, 442452. 4. Barricelli, N. A. & Doermann, A. H. (1961) Virology 13, 460476. 5. Doermann, A. H., Chase, M. & Stahl, F. W. (1955) J. Cell. Comp. Physiol. 45, Suppl., 41-74. 6. Delius, H., Howe, C. & Kozinski, A. W. (1971) Proc. Natl. Acad. Sci. USA 68,3049-3053. 7. Howe, C. C., Buckley, P. J., Carlson, K. & Kozinski, A. W. (1973) J. Virol. 12, 130-148. 8. Mosig, G. (1970) J. Mol. Biol. 53,503-514. 9. Dressler,, D., Wolfson, J. & Magazin, M. (1972) Proc. Natl. Acad. Sci. USA 69, 998-1002. 10. McDonell, M. W., Simon, M. N. & Studier, F. W. (1977) J. Mol. Biol. 110, 119-146. 11. Studier, F. W. (1973) J. Mol. Biol. 79,227-236. 12. Studier, F. W. (1969) Virology 39,562-574. 13. Studier, F. W. (1975) J. Mol. Biol. 94,283-295. 14. Kozinski, A. W. & Szybalski, W. (1959) Virology 9,260-274. 15. Steinberg, C. M. & Edgar, R. S. (1962) Genetics 47, 187-208. 16. Denhardt, D. T. (1966) Biochem. Biophys. Res. Commun. 23, 641-646. 17. Miller, R. C., Jr., Lee, M., Scraba, D. G. & Paetkau, V. (1976) J. Mol. Biol. 101, 223-234. 18. Dulbecco, R. J. (1959) J. Bacteriol. 39,329-347. 19. Ellison, S., Feiner, R. & Hill, R. F. (1960) Virology 11, 294296. 20. Kuemmerle, N. B. & Masker, W. (1977) J. Virol. 23,509-516. 21. Southern, E. M. (1975) J. Mol. Biol. 98,503-517. 22. Benbassat, J., Burck, K. B. & Miller, R. C., Jr. (1978) Virology 87, 164-171. 23. Schlegel, R. A. & Thomas, C. A., Jr. (1972) J. Mol. Biol. 68, 319-345. 24. Watson, J. D. (1972) Nature (London) New Biol. 239, 197201. 25. Langman, L., Paetkau, V., Scraba, D., Miller, R. C., Jr., Roeder, G. S. & Sadowski, P. D. (1978) Can. J. Biochem. 56,508-516. 26. Doermann, A. H., Eiserling, F. A. & Boehner, L. (1973) J. Virology 12, 374-385. 27. Pao, C. C. & Speyer, J. F. (1973) J. Virol. 11, 1024-1026.
