Cloning of the Bacteriophage T 4 uvsX Gene and Purification and Characterization of the T 4 uvsX Recombination Protein *

The bacteriophage T4 UVSX gene is a nonessential gene required for normal levels of DNA repair, recombination, and replication. We demonstrate that plasmids containing the T4 DNA approximately 300-2900 base pairs upstream of T4 gene 41 express a biologically active uvsX protein. This UVSX protein imparts increased survival to UV-irradiated T4 UVSXphage and decreases the T4 uusXmutant suppression of a conditionally lethal T4 mutant in the gene 49 recombination nuclease. The uvsX protein purified from cells with a UVSX’ plasmid catalyzes ATP hydrolysis to ADP and AMP and, in the presence of the T4 gene 32 helix-destablizing protein, ATP-dependent strand exchange between homologous circular single-stranded and linear duplex DNA, These results agree with the recent characterization of uvsX protein from T4-infected cells by Yonesaki et aZ. (Yonesaki, T., Ryo, Y., Minagawa, T., and Takahashi, H. (1985) Eur. J. Biochem. 148,127-134) and by Formosa and Alberts (Formosa, T., and Alberts, B. M. (1984) Cold Spring Harbor Symp. Quant. Biol. 49, 363-370). Tn addition, we find that under some reaction conditions strand exchange is catalyzed by UVSK protein in the absence of 32 protein. The level of the uvsX protein expressed by the UVSX’ plasmids is high and independent of the orientation of the T4 DNA within the vector. This suggests that transcription promoter($ lie upstream of the UVSX gene on the cloned T4 DNA. In vitro transcription of T4 DNA restriction fragments reveals two tandem promoters whose transcripts initiate approximately 500 and 600 nucleotides upstream of the UVSX gene and extend through the gene.

The bacteriophage T4 UVSX gene is a nonessential gene required for normal levels of DNA repair, recombination, and replication. We demonstrate that plasmids containing the T4 DNA approximately 300-2900 base pairs upstream of T4 gene 41 express a biologically active uvsX protein. This UVSX protein imparts increased survival to UV-irradiated T4 UVSXphage and decreases the T4 uusXmutant suppression of a conditionally lethal T4 mutant in the gene 49 recombination nuclease.
The level of the uvsX protein expressed by the UVSX' plasmids is high and independent of the orientation of the T4 DNA within the vector. This suggests that transcription promoter($ lie upstream of the UVSX gene on the cloned T4 DNA. In vitro transcription of T4 DNA restriction fragments reveals two tandem promoters whose transcripts initiate approximately 500 and 600 nucleotides upstream of the UVSX gene and extend through the gene. The genome of bacteriophage T4 is a linear duplex of 166 kilobase pairs which is circularly permuted and terminally redundant (1,2). Genetic analyses of T4 mutants displaying altered DNA synthesis have implicated many phage genes in DNA replication and metabolism (3,4). Generally, these mutants have been classified as belonging to one of the following three categories: DNA-(no DNA synthesis), DNA delay (delay in onset of DNA synthesis), and DNA arrest (early stop of DNA synthesis). In vitro studies (5)(6)(7) have demonstrated that the products of six DNA" genes (gene 43-DNA polymerase, 44-, 62-, 45-polymerase accessory proteins, 32-helix-destabilizing protein, and 41-primae component) * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. plus one DNA delay gene (61-primase component) together catalyze efficient strand displacement synthesis of DNA on a nicked circular template and prime discontinuous synthesis on the displaced strand. These studies suggest that these seven T4 gene products are the proteins required in vivo for new DNA synthesis on the leading and lagging strands of a replication fork. As DNA synthesis proceeds in uiuo, the main form of T4 DNA observed is a complex network (8) which sediments faster than T4 DNA extracted from the phage (9). Mosig and eo-workers have proposed (reviewed in Ref. 10) that this network arises from strand invasion by the singlestranded DNA present at the unreplicated terminus of one DNA molecule into the homologous duplex region of another T4 genome. It is proposed that the invading single strand is then extended by the complex of replication enzymes. Since the T4 genome is circularly permuted, such a recombination/ replication process would result in the type of network of highly branched concatenated structures observed for replicating T4 DNA.
The T4 genes needed for the observation of the T4 concatenated structures constitute a recombination pathway (discussed in Refs. 11 and 12) which is thought to include (among other T4 genes) uusX, uusY, uvsW, and gene 49. Proposals support the hypothesis that the uvsX and uvsY gene products contribute to the formation of the concatenated structures which require 49 nuclease for processing (12,25). Presumably, a UUSX-49-or uvsy-49-mutant survives by packaging the few T4 progeny genomes produced by replication without the recombination intermediates.
A more precise role for the uvsX gene product in recombination has been elucidated by the biochemical studies of Yonesaki et al. (26) and by Formosa and Alberts (Ref. 27,cited in Refs. 28 and 29). These workers have purified the uvsX protein from T4-infected cells and report that this protein is a single-stranded DNA-dependent ATPase that promotes ATP-dependent homologous pairing and strand exchange of DNA. These DNA-pairing reactions are analogous to those carried out by the Escherichia coli recA protein. Griffith and Formosa (29) have recently shown that, again like the recA protein, uvsX protein binds cooperatively and tightly to double-stranded or single-stranded DNA to form filaments which can be observed by electron microscopy. Taken together, these i n vitro observations suggest that the UVSX protein, through its promotion of homologous pairing, contributes directly to the formation of the concatenated structures of replicating T4 DNA observed in vivo. Genetic 2) gives one product of 34 kDa. We assign this protein as p-lactamase, the product of the amp' gene which is present in the pBR322 sequence and has been previously shown to be a major in vitro translation product of pBR322 (46). Insertion of the T4 DNA within the vector to yield the recombinant plasmids pDH428 (panel A, lanes 3 and 4) and pDH447 (panel A, lane 5 ) results in plasmids whose major translation product is a highly expressed species of 46 kDa. The 46-kDa band is again observed as the major species if just the T4 DNA from this region, a Sal1 restriction fragment (map units 24.8 to 20.75) or an EcoRI fragment (map units 24.3 to 21.15) is transcribed and translated in vitro (Fig. 3B,  lanes 2 and 3). These results support the previous in vivo results indicating that the gene for the 46-kDa protein is present on the T4 DNA.
We have inferred the direction of the 46-kDa gene based on the translation products expressed by the plasmid pDH428A3, a plasmid identical to pDH428 except that the 800 bp between the ClaI sites at map units 22.85 and 22.00 have been deleted (Fig. hi). The 46-kDa protein is not made by this plasmid in vivo (Fig. 2,lane 9) or in vitro (Fig. 3A,lane 6). However, i n vitro, the major species is a highly expressed new protein of 30 kDa. We interpret the 30-kDa band as an abnormal protein generated by the fusion of the open reading frame of the 46-kDa protein to new DNA sequences. In addition, we assume that this fusion has occurred by joining the DNA encoding the N terminus of the normal 46-kDa protein to new sequences, rather than fusing the DNA of the C terminus to another start. This assumption is reasonable since the 30-and the 46-kDa proteins are similarly expressed at a very high level and represent the major translation products of pDH428A3 and pDH428 (or pDH447), respectively. Since the plasmid pDH428Al produces normal levels of the 46-kDa protein in vivo (Fig. 2, lane 14), the gene for this protein cannot begin downstream of the BglII site at map unit 22.01. Thus, we conclude that the production of the abnormal 30-kDa protein by pDH428A3 indicates that the normal gene starts upstream of the CZaI site at map unit 22.85 and proceeds toward gene 41. This direction is consistent with the previously determined general direction of early T4 transcription in vivo (2), the direction of the immediately down- We have observed differences from the reported restriction sites (58) in the region between the EcoRI site at map unit 24.3 and the XbaI site at 22.37. We do not observe a Pstl site reported to lie at map unit 22.40 after restriction of the purified T4 SalI fragment (map units 24.8 to 20.75) or restriction of the plasmids. In addition, our calculation of distances between the EcoRI site at 24.3 and the XbaI site at 22.37, again using the purified SalI fragment and the plasmids, differs from reported values (58). We have indicated our values of these distances which were used to determine the start of the in vitro transcripts (see text). Panel B shows the plasmid pDH428 (31) which contains T4 DNA ( I ) from the EcoRI site at map unit 24.3 to the HindIII site at map unit 20.06. The T4 DNA has been ligated to a fragment of the gal expression vector pKG1810 (37) which contains E. coli sequence (0) with the IS2 transcription terminator (T) and the galK gene followed by pBR322 sequence (-) from 2068 to 4363 (61). amp' designates the ampicillin resistance gene expressing p-lactamase. stream genes 41 and 61 (31,32), and the position of the major upstream of the ClaI site at map unit 22.85 (enough sequence promoters in this region as determined in vitro (see below). t o encode a protein of 30,000 daltons). Since the 46-kDa Based on the size of the abnormal pDH428A3 product (30 protein is observed in vivo in cells containingpDH428Al (Fig.  kDa), we estimate that the farthest upstream position for the 2, lane 14) but is not observed in those containing pDH428A2 5' end of the 46-kDa gene (assuming no introns) is 820 bp (Fig. 2, lane 12), we conclude that the 3' end of the gene In Vitro Transcription Demonstrates That the T4 DNA Upstream of the 46-kDa Gene Contains Two Promoters for E. coli RNA Polymerase-In the previous translation experiments, we observed a high level of the 46-kDa protein both in vivo and in vitro. In addition, the expression of the protein is independent of the orientation of the T4 DNA within the plasmid ( Fig. 2: compare pDH4426 with pDH447 (lanes 4 and 5 ) and pDH428 with pDH421 (lanes 7 and 8)). These results suggest either that the T4 DNA contains its own transcription promoter(s) for the expression of the 46-kDa gene or (less likely) that the gene is translated at a very high level from message generated by read-through transcription from the vector. To test whether the T4 DNA we have cloned contains promoters active in vitro, we transcribed the T4 SalI fragment (map units 24.8 to 20.75) and the EcoRI fragment (map units 24.3 to 21.15) using E. coli RNA polymerase in the presence of [m-"P]UTP. In order to favor specific transcription starts from stronger promoters, we transcribed the DNA using a low molar ratio of RNA polymerase to the DNA template and we challenged the RNA polymerase-DNA binding by the addition of heparin 2 min before the addition of the triphosphates (see "Experimental Procedures"). Use of either the SalI or the EcoRI fragment yields large RNA run-off transcription products shown on the 1% agarose gel in Fig. 4A. However, because this gel is nondenaturing the sizes for the RNAs cannot be determined, although the major species from transcription of the EcoRI fragment appears to be shorter than that obtained after transcription of the SalI template. To determine the size of the major transcript(s), we cleaved the SalI fragment with the restriction endonucleases BglII, XbaI, ClaI, PstI, and NdeI. (The XbaI digest was only 50% complete.) We then transcribed these substrates in the presence of [a-"PICTP and separated the products on a denaturing 4% acrylamide, 7 M urea gel (Fig. 4B). This experiment demonstrates a doublet of transcripts (designated by dots) which are progessively shortened as the SalI fragment is cleaved with these restriction enzymes. We interpret these RNAs as run-off transcripts which initiate within the T4 DNA and proceed to the end of the template (in each case defined by the position of the restriction site). By comparing the restriction map ( Fig. lA) with the sizes of these RNAs, we estimate that these transcripts initiate from two promoters which lie 500 and 600 bp upstream of the NdeI site. Initiation at these promoters generates transcripts that extend in vitro through the region we have assigned to the gene for the 46-kDa protein.
The 46-kDa Protein Is a T4 dsDNA-binding Protein, Absent after Infection of a T4 UUSX-Phage-The region we have assigned to the 46-kDa gene lies in a part of the T4 genome which is transcribed early after infection in vivo and which contains several genes for proteins that interact with DNA (2). To test whether the 46-kDa protein might also have an affinity for DNA, we examined the dsDNA cellulose-binding proteins expressed by DH1 cells containing either the vector pKG1810R3, the plasmid pDH447, or infected with T4 41phage (Fig. 5, left panel). The 46-kDa protein expressed by pDH447 binds to the dsDNA cellulose in the presence of 0.1 M NaCl, is eluted by stepping the salt concentration to 0.5 M NaCl (lune E), and co-migrates with a similarly bound protein present in fractions from extracts of T4 41"infected cells. As expected, no such protein is expressed by cells containing the vector.
Previously, the genes for two T4 proteins which interact with DNA have been genetically mapped upstream of gene 41. One of these genes encodes P-glucosyltransferase, a protein of 46,000 daltons (47) which adds monoglucosyl moieties to DNA (reviewed in Ref. 48). We find no P-glucosyltransferase activity (see "Experimental Procedures") in crude extracts of cells with plasmids expressing the 46-kDa protein or in the dsDNA cellulose-binding fraction of pDH447/DH1 cells (data not shown). The other T4 protein which interacts with DNA and whose gene has been mapped upstream of gene 41 is uvsX. As indicated in the Introduction, this protein has been shown to be analogous to the E. coli recA protein in promoting ATP-dependent homologous pairing in vitro (26)(27)(28)(29). We have identified the 46-kDa dsDNA-binding protein present in cells with pDH447 or after infection with wild type T4D as the uvsX gene product, since this protein is not found after infection with the T4 uvsXmutant, fdsA (Fig. 5 , rightpanel, lane E ) or with T4 UVSX am1 1 (not shown). Previous workers (26,28,49) have estimated the size of the uvsX protein as approximately 40 kDa, in contrast to its apparent size of 46 kDa on our gels. This discrepancy is due to the conditions of gel electrophoresis since the uvsX protein made by pDH447 and that purified from T4-infected cells (the generous gift of T. Formosa, University of California, San Francisco) migrate as 46 kDa under our conditions (data not shown) and as 40 kDa under conditions (50) similar to those used by Formosa and Alberk3s4 The reason for the anomalous electrophoretic behavior of the uvsX protein is not clear. Although for convenience we have referred to the uvsX protein as the 46-kDa band on protein gels, its actual size will have to be determined by other techniques.
The uusX Protein Expressed by the Recombinant Plasmids Complements T4 uvsX-Phage-The phenotype of T4 uvsXmutants is pleiotropic including increased sensitivity to UV light, ionizing radiation, and chemical agents, as well as arrest of DNA synthesis and decreased burst size (see Introduction).
In collaboration with J. W. Drake (National Institute of Environmental Health Sciences), we tested the biological activity of the uvsX protein expressed by our clones by asking whether UV-irradiated T4 uvsXphage have increased survival in cells containing the 46-kDa protein. T4x, (uusX-) phage were irradiated and then plated on DH1 cells without plasmid, with the vector, or with pDH447. As controls, wild type T4D or T4 deficient in two other genes associated with recombination and repair, uvsY and denV, were also irradiated and plated. Each control phage has a distinctive survival curve which is not altered by the host (Fig. 6, solid symbols). However, after 83 s of irradiation, the irradiated uvsXphage survive approximately 8-fold better (to a level nearly like that of T4D) when plated on the pDH447/DH1 cells (open triangles) than on DH1 or DH1 with the vector (open circles and squares). This experiment indicates that the uvsX protein expressed by the clone can substitute for the T4 uvsX gene product in vivo in responding to damage by UV light.
As discussed in the Introduction, T4 uvsXmutants have been shown to suppress conditionally lethal mutations in the T4 gene 49 (22)(23)(24). Thus, at the nonpermissive temperature, T4 uvsX-ts gene 49 phage will grow but T4 ts gene 49 phage will not. We asked whether providing uvsX protein from the uvsX+ plasmid pDH428 could reduce this suppression. As seen in Table I, use of the pDH4281DHl host lowers the efficiency of plating of the T4 fdsA (uvsX-)-tsCS (ts gene 49) phage at the nonpermissive temperature by greater than 20-fold. The use of cells containing the vector or the plasmid pDH428A3 (with the 800-bp deletion which eliminates the 46-kDa product) does not affect the efficiency of plating. In addition, the ts gene 49 mutation by itself displays a tighter phenotype on pDH428/DH1 cells than on cells that do not express uvsX protein. Taken together, the results of these complementation experiments indicate that the uvsX protein expressed by our clones behaves like a biologically active uvsX gene product.
Purification of the uvsX Protein from pDH4471DHl Cells-The purification of the uvsX protein (see "Experimental T. Formosa Procedures") is summarized in Table I 1 and Fig. 7. Most of the purification is achieved by chromatography of the crude extract over dsDNA cellulose, which yields a fraction containing the uvsX protein as the major protein (Fig. 7, lane C). The remaining DEAE and phosphocellulose steps are modifications of procedures developed for the purification of uvsX protein from T4-infected cells by .3 During the last chromatographic step (phosphocellulose column) uvsX protein, apparently homogeneous by gel electrophoresis, co-chromatographs with a ssDNA-dependent ATPase activity (Fig. 7) which gives both ADP and AMP as products (see Footnote e to Table 11). The purified protein is free of nucleases acting on single-stranded, double-stranded, or supercoiled DNA under the conditions used for the strand exchange reactions below (Figs. 8 and 9).
Strand Exchange Catalyzed by the uusX Protein in Vitro-We have verified that the uvsX protein from the clone catalyzes D loop formation (data not shown) and strand exchange between circular ssDNA and homologous dsDNA (Figs. 8 and 9), as originally reported for the uvsX protein from T4infected cells (26-29). Fig. 8 shows the products of strand exchange between single-stranded circular viral 4x174 DNA and double-stranded 4x174 RF linearized by cutting with PstI nuclease. (The PstI digestion gives a 4-base singlestranded extension at the 3' terminus.) At the lower concentration of ssDNA, strand exchange is catalyzed by the uvsX protein alone (Fig. 8, lane 5 ) but is strongly stimulated by the T4 gene 32 helix-destabilizing protein (lane 4 ) . At the higher ssDNA concentration, both the uvsX and 32 proteins are apparently required (Fig. 8, lane I ) , since there are no visible products with only the uvsX protein (lune 2) or 32 protein (not shown). A portion of the products in lanes 1 , 4 , and 5 comigrate with marker nicked 4X RF 11, the expected product of complete strand exchange. Since we have not further characterized these products, it is possible that products other than simple nicked circles co-migrate with the nicked circle marker. There are also products migrating behind the nicked circle marker, which may be intermediates in which the linear duplex is only partially unwound. Products running behind the nicked circle are also seen in similar reactions catalyzed by recA protein (51). In Fig. 8, there are, in addition, large products which do not enter the gel and which are more prominent at higher ratios of dsDNAssDNA (lanes 4 and 5 versus lane I). These may be complex structures resulting from incomplete strand exchange involving several molecules. Fig. 9 demonstrates that the strand exchange reaction catalyzed by the uvsX protein alone is also stimulated by increasing the concentration of dsDNA. As described under "Experimental Procedures," we labeled the 5224-bp XhoIIPstI fragment of 6x174 RF DNA at the 3' terminus of the (-)-strand by adding nucleotides to the 3' OH of the XhoI end using the large fragment of E. coli polymerase I to give blunt ends. At the lower dsDNA concentration, the uvsX protein alone does not yield visible products (lane 4 ) . However, raising the dsDNA concentration 2-fold allows a reaction by the uvsX protein alone (compare lanes 7 and 4). The addition of 32 protein (Fig. 9, lanes 3 and 6 ) increases the extent of the reaction mediated by uvsX protein at both concentrations of dsDNA. Note that products co-migrating with nicked circular DNA and presumably representing extensive branch migra- The surviving fraction represents the titer obtained after irradiation relative to that prior to UV treatment. The T4 phage were plated as described under "Experimental Procedures." The efficiency of plating 44 "C/29 "C represents the titer of the phage at the nonpermissive temperature (44 "C) relative to that at the permissive temperature (29 "C).

TABLE I Suppression of a T4 49-mutation by a UUSX-mutation i s eliminated in cells with a uvsX+ Dlnsmid
' Plaques formed at 44 "C were very small. tion following synapsis are much more prominent in the presence of 32 protein.

Cloning of the Bacteriophage T4 Gene uvsX and Preliminary Characterization of the uvsX Protein Expressed by the Clone-
The bacteriophage T4 uvsX gene is a nonessential gene whose product, along with other proteins, is required for wild type resistance to radiation and chemical treatment, normal recombination frequencies, and the formation of a network of T4 DNA thought to arise through homologous recombination This is uvsX protein in peak fractions used for phosphocellulose chromatography. Total uvsX protein recovered from DEAE column (corrected as in Footnote c) was 17 mg.
'DNA-dependent ATPase cannot be measured reliably in earlier fractions. In both fractions I11 and IV, the ratio of AMP to ADP was 0.28, and there was no ATP hydrolysis in the absence of ssDNA. among replicating T4 genomes (see Introduction). We have found that plasmids containing the T4 DNA approximately 300-2900 bp upstream of gene 41 express a 46-kDa protein whose presence both imparts increased survival to UV-irradiated T4 uvsXphage (Fig. 6) and decreases the suppression of a conditionally lethal T4 ts gene 49 mutant by a T4 uvsXmutant (Table I). These in vivo results are consistent with our having cloned the gene for a biologically active uvsX protein, and our position for the UVSX gene agrees with genetic mapping of the uvsX gene between T4 genes 41 and 42 (25,30). The T4 uvsX+ plasmids described here, together with the T4 uvsY+ and uvsW+ plasmids previously described (19, 52), should aid in our understanding the role of each of these genes in T4 DNA recombination and repair.
The uvsX protein that we have purified from plasmidcontaining cells is a ssDNA-dependent ATPase which promotes homologous strand exchange in vitro, as has been previously shown for uvsX protein from T4-infected cells (26-29). Our finding that the plasmid-encoded uvsX protein hydrolyzes ATP to both ADP and AMP (Table 11) is in agreement with the more extensive analysis of the products formed by the uvsX protein from infected cells (27-29) and eliminates the possibility that one of these products was produced by a contaminating phage protein. uvsX protein from cells with the plasmid mediates ATP-dependent strand exchange between the (-)-strand of a linear duplex and its single-stranded circular (+)-strand complement in a reaction which is stimulated by, but not absolutely dependent upon, the T4 gene 32 helix-destabilizing protein (Figs. 8 and 9). Strand exchange in the absence of 32 protein is highly dependent on the concentration of both ssDNA and dsDNA (Figs. 8 and 9), which may explain previous reports of a 32 protein requirement for the strand exchange reaction (cited in Ref. 29). However, we cannot rule out the possibility of a fundamental difference between the uvsX protein from the clone and that from T4-infected cells. Our results, indicating a stimulation but not a requirement for 32 protein in uvsX protein-mediated strand exchange, resemble those previously observed for the E. coli ssDNA binding (SSB) protein and the E. coli recombination protein recA ((51); for a recent review, see Ref. 53). More detailed investigations of the T4 recombination proteins will be required in order to understand precisely how homologous recombination is accomplished by purified T 4 proteins in vitro as well as during the life cycle of T4.
Expression of the uvsX Protein-Our plasmids which contain the T4 uvsX gene cloned in a multicopy plasmid vector express a high level of the protein both in vivo (Fig. 2) and in vitro (Fig. 3). We have presented evidence that at least part of this expression is due to transcription promoter(s) on the T4 DNA upstream of the uvsX gene. First, expression of the protein is independent of the orientation of the T4 DNA within the plasmid (Fig. 2), suggesting that in the plasmid the uvsX gene is not expressed simply by read-through transcription from the vector DNA. Second, our in vitro transcription of T4 DNA fragments from this region identifies 2 tandem promoters lying upstream of the uvsX gene which produce transcripts that extend through the uvsX gene (Fig. 4). We have observed transcripts from these promoters under conditions (see "Experimental Procedures" and "Results") which should favor specific initiation from stronger promoters over nonspecific starts or initiation a t weaker promoters (54). A previous study by Gram et al. (55) to identify strong T4 promoters used in vitro by E. coli RNA polymerase failed to reveal the ones we have observed upstream of the uvsX gene. This discrepancy probably reflects the difference in salt concentrations used in their study (200 mM KCl) and our study) (75 mM) since we find that the intensity of the RNA bands observed after transcription of the T4 Sal1 fragment (map units 23.8 to 20.75) is diminished severalfold using the higher salt c~ncentration.~ However, even at the higher salt concentration, the initiation of transcription is very specific, yielding product which comigrates with the major transcription band observed in Fig. 4A, lane 1. I n vitro transcription/translation of the plasmid with the largest T 4 DNA insert, pDH428, produces several minor species (Fig. 3)  exposures of this autoradiogram reveal that two of the fainter of these bands co-migrate with either p-lactarnase, the amp' gene product observed with the vector alone, or with T4 gene 41 protein, which we have previously shown is expressed by pDH428 in uiuo (31). Most of the remainder of the bands migrate faster than the 46-kDa UVSX product and are dependent on the T4 DNA for expression. These minor bands could arise from premature transcription or translation stops of the highly expressed uvsX gene or could be unidentified T4 proteins. In fact, T4 gene 40, whose product, a 14-kDa protein, is required for proper phage head assembly has been mapped, like the uvsX gene, between genes 41 and 42 and is transcribed in the same direction as gene 41 (56). Whether one of the minor protein bands we have observed after translation of the uvsX plasmids is this gene product is not yet known.
We are presently examining the in vivo expression of the The vector DNA is schematically shown as the non-T4 DNA in Fig. 1  from pVH627 was ligated to the e R I /~d I I I vector fragment of pKGl8lO TO Construct pDH447 and pDH4426 (Fig. 1Al. the 3400 bp g R I fragment produce blunt end DNA at the remaining G d I I I and =HI termini) and ligated (described above). The products were treated with T4 DNA polymerase (to again. pDH447 and pDH4426, two products of this procedure, represent insertions of the 3400 bp X R I fragment in opposite orientations. To Construct the control plasmid pKGl8lOR3, the KRI/&dIII vector presence of the &HI OligOdeOxyribonuCleOtide linker pd(GGAATTCC).
fragment of pKGl8lO was treated with T4 DNA polymerase and ligated in the In Vitro Transcription/Translation Reactions -Plasmid DNA, isolated as bromide-cesium chloride gradients 1391. After removal of the ethidium bro-described (311, was purified by twice centrifuging to equilibrium in ethidium mide by extraction with CsC1-saturated isopropanol, the DNA Solution was dialyzed against TE buffer ( 1 0 mM Tris-HCI, pH 7.9, 1 mM EDTA), and the DNA precipitated by the addition of salt and ethanol. The DNA pellet was washed with 70% ethanol, dried, and redissolved in TE buffer to give a final concentration of 0.2 to 1.5 mg/ml. The DNA Was stored at -2O'C.
(amount given in legend to Fig. 3). 6 111 of E. coli I~CB', recc-5-30 extract " In vitro t r a n s c r i p t i o n / t r a n s l a t i o n reactions ( In Vitro Transcription Reactions -The T4 e 1 restriction fragment, representing map Units 24.8 to 20.75 (Fig. 1A) and isolated from T4 dC-DNA, and the T4 e n 1 fragment representing map units 24.3 to 21.15 (Fig. 1A) and isolated from the plasmid pVH627 (see above) were purified a5 described (31).
or the experiments shown in Fig. 4~. the purified SI fragment was incubated with each of the restriction enzymes indicated. After digestion. the reaction mixtures were extracted with phenol and the DNA precipitated by the addition of Salt and ethanol, washed with 70% ethanol, dried, and redissolved in TE buffer. DNA fragments were transcribed essentially as described (40) in reaction mixtures ( 1 5 ul) containing 20 mM Tris-HC1. pH 7.9, 6% glycerol, 75 mM KCl, 5 mM MgC12, 0.1 mM EDTA, 0.5 mM dithiothreitol, 50 ug/ml bovine serum albumin, 0.01 pmol DNA fragment(s1, 0.02 pmol E-a RNA polymerase phates IrNTPs). The reaction mixture containing all components except hepa-(Enzo Eiolabs), 100 ug/ml heparin, and all four 5'-ribanucleaside triphosrin and the rNTPS was first incubated at 37-c for IO minutes. Heparin was then added and the resulting mixture kept at 37'c for 2 minutes. In vitro transcription was begun by adding a solution of either 10 VM (,-3Zp)~~p (  T4 fdsA I&), T4 tSC9 (ts gene 49), or T4 fdsA-tsC9 phage were plated as described (24) on DH1 cells or DH1 cells containing the indicated plasmids except that phage were not preabsorbed before plating. Plates were incubated overnight at 29°C (permissive for T4 tsC9 growth) or 44'C (nonpermissive for T4 e).
The efficiency of plating I-epresents the titer at the nonpermissive temperature relative to that at the permissive temperature.
with a uysx+ plasmid, T4 phage were irradiated and plated essentially as To test far increased survival of irradiated T4 phage in a host described (24). T4 phage at 2 to 3 x 10' phage/ml in M9S medium (241 Were fixst irradiated for the indicated times using a 15-watt low pressure Hg bulb diluted in D broth (24)  Purification of T4 uysx Protein -a DHl/pDH447 was grown in a 50-liter fermentor to late log phase at 37OC in L broth plus 25 pg/ml ampicillin, The cell paste (129 g) was stored at -7OOC.
Extract -20.5 g of cell paste were thawed at 4°C in 6 0 ml of the sonication buffer described above supplemented with 1 mM phenylmethylsulfonyl fluoride IPMSF). The cells were sonicated in SO-second pulses, keeping the ___ temperature below 8OC with a salt ice bath, until the optical density at h concentration of 17 pg/ml, and the extract incubated for 60 minutes at 15°C 540 had decreased from 1.6 to 0.3. Pancreatic DNase was added to a final and centrifuged for 2 hours at 35,000 rpm in a Beckman 45 Ti rotor. The supernatant solution (50 ml) was dialyzed over a period of 20 hours against 4 changes of 1 liter of DC buffer with 5 0 mM NaCl and 1 mM PMSF.