The Physical Mapping of Bacteriophage T5 Transfer tRNAs*

Transfer RNAs, isolated from Escherichia coli F cells infected with T5 bacteriophage, were charged with radioactive amino acids and used in RNA. DNA hybridization studies to detect and locate T5 tRNA cistrons in the T5 DNA chromosome. Hybridization of 14 SH-aminoacyl-tRNA species, including purified Heteroduplex mapping examined

Heteroduplex mapping of eight mutant T5 DNA deletions has enabled us to locate and determine the size of these deleted segments. By correlating this information with the presence and absence of specific tDNA sequences in these mutants, as determined by tRNA.DNA hybridization, we were able to define the physical limits of four tDNA-containing loci along the T5 DNA molecule. A physical map for 15 tRNA species examined indicates that the structural genes for these' tRNAs are clustered within a segment length of T5 DNA that represents approximately 11.2% of the total wild type T5 DNA. The existence of the deletion mutants indicates that T5 tRNAs are dispensable for T5 replication under the growth conditions and for the host employed.
Studies with the DNA isolated from bacteriophage T5 have shown that the DNA is a nonpermuted, linear duplex with an approximate molecular weight of 75 to 80 x lo6 (l-3); that it is terminally redundant (4); and that it has the unique feature of containing four to five single strand breaks in one of its DNA strands (5)(6)(7)(8)(9) which may be genetically determined (6,8). A recent report by Herman and Moyer (10) indicates that the single strand interruptions are repaired, in vivo, prior to DNA replication.
Another unusual feature of this phage is that infection of sensitive Escherichia coli is accomplished by the injection of T5 DNA into its host via a two-step process (11). However, as with other bacterial DNA viruses, T5 infection leads to T5 DNA-directed RNA synthesis in an ordered series of events such that distinct time-classes of T5 transcripts (pre-early, early, and late RNAs) are identifiable (12)(13)(14). Some of the physical and transcriptional features of T5 DNA have been incorporated in the T5 DNA map illustrated in Fig.  1, which was constructed from the maps of Hayward and Smith (13), Hendrickson and Bujard (14), and Hayward (3).
Work in this laboratory (15) as well as others (16, 17) initially demonstrated that RNA transcripts following T-even phage infection of E. coli include the synthesis of phage-coded transfer RNAs. Subsequent studies in several laboratories have confirmed and extended these observations (18,19). Following this discovery, Scherberg and Weiss (20) reported that T5 phage induces the synthesis of some 14 different tRNA species; and recently, this same laboratory demonstrated that two isoacceptor T5 methionine tRNA species, tRNAMet and tRNA,Met, are present in T5-infected cells (21). Unpublished results obtained by our group indicate that the synthesis of T5 tRNA falls into the time-class category of an early phage DNA transcript, similar to that reported for T4 tRNA synthesis (22). According to the physical transcription map of T5 DNA (Fig.  l), regions B, C, D, and E of the T5 chromosome code for early RNA synthesis. One would expect, therefore, that T5 tRNA genes are located in one or several of these DNA segments.
In this report, we describe the physical location of tRNA genes in T5 DNA for 15 tRNA species by use of the techniques of aminoacyl-tRNA .DNA hybridization and heteroduplex DNA mapping. The results show that almost all of the phage tRNA genes are clustered in the C segment of T5 DNA. In agreement with the transcript map shown in Fig. 1 (23).

MATERIALS AND METHODS
Growth and Purification of Phages-Wild type T5+ and all deletion mutants of T5 were grown at 37" in NCG medium (8 g nutrient broth (Difco)/lO g casamino acids (Difco)/ZO g glycerol/2.5 ml of 1 M Tris-HCl, pH 7.8, in a total volume of 1 liter) (9), with E. coli Fused as host. from other reported models (3,13,14). The heauy lines represent the heavy and light DNA strands of T5' wild type DNA with the polarities of the chains indicated by 3' and 5' designations.
The molecular weight values shown (x10-') are for the light chain segments created by the single strand interruptions as reported by Hayward (3). However, measurements for some of the light chain segments in this report differ slightly, e.g. A = 3.1 * 0.1, B = 3.7 + 0.5, and C = 6.0 l 0.1 x lo6 which in turn gives the percentage of total length for these same segments as 7.7, 9.0, and 14.7, respectively, using a single strand molecular weight of 40.6 x lo6 for T5+ DNA (see "Results"), The broken and solid lines with arrows show the region and direction of T5 pre-early, early, and late RNA transcription.
Richardson (9). After CsCl banding and dialysis, phage was stored at 4O in MGM medium ( Fig. 2) were loaded purified except for T5 tRNA, and tRNA, which were purified by hyonto filters, annealed with T5+ tRNAs individually charged with the bridization to T5+ DNA and RPC-5 chromatography (21). The input labeled amino acids shown below, and the total radioactivity fixed to counts per min shown below indicates the total radioactivity used in the filters determined as described under "Materials and Methods." the annealing mixture which varied in volume from 0.6 to 1.8 ml. The All of the amino acids listed were labeled with SH except for me-hybrid counts per minute per Kg was calculated from the total "fixed" thionine which contained " S. The tRNAs used for charging and hyradioactivity after subtraction of control counts.   Fig. 8 were obtained for each aminoacyl-tRNA.
All hybridiza-I icn values greater than 10% of wild type DNA were considered positive (+) while those less than 10% of wild type were scored as negative ( - Since the exact size and location of the individual mutant DNA deletions were not precisely known, it was necessary to obtain this information before we could attempt to construct a physical tRNA gene map for T5+ DNA. It was first necessary to measure the intact genome of T5. Such measurements were made by electron microscopy, as described under "Materials and Methods." Native T5+ DNA was measured as 23.86 + 0.32 x @X174 RF DNA in 50% formamide. If a value of 1.7 x lo6 daltons is used as the length of @X174 DNA (29), the length of native T5+ DNA is 81.1 + 1.1 x lo6 daltons, and the single strand molecular weight is 40.6 l 0.5 x 10'. Measurements were also made against an external standard (a replica grating), yielding values of 1.63 f 0.04 pm for @X174 RF DNA and 38.9 f 0.5 pm for T5+ DNA. Thus, for T5+ DNA in 50% formamide, the mass to length ratio is 2.06. Heteroduplexes were then measured relative to the length of T5+ DNA. In Fig. 9 is shown a schematic representation of the heteroduplex forms which one might expect to obtain by cross-hybridization of the DNA strands from T5+ wild type (Form 1) and a Group A deletion mutant (Form 2). It should be possible to obtain two different DNA heteroduplex structures by the renaturation of opposing heavy (H) and light (L) DNA strands (Forms 3 and 4). Form  Mapping of T5 tRNAs 543 structure in the T5+ H-strand since the equivalent complementary DNA segment is missing in the mutant L-strand. A reciprocal loop structure should also form from the annealing of wild type L-strand segments and mutant H-strand if the deletion does not include a single strand interruption as in the case of the Group B T5 mutants. Another type of heteroduplex structure, Form 4, can also be predicted if hybridization occurs between the H-strand of a Group A mutant DNA and the L-strand segments of wild type DNA. In this case, a doublebranched single strand structure should appear since the deletion includes the C-D single strand interruption which preempts the formation of a DNA loop. For either of the heteroduplex forms, the size and position of the DNA deletion can be ascertained by measurement of the lengths of the DNA loop or the branched structures as well as the distance of their branch points to either end of the heteroduplex chain.
In Fig. 10  sents the st(0) deletion segment. As described under "Materials and Methods," &X174 bacteriophage DNA and mouse mitochondrial DNA served as standards for the measurement of single-and double-stranded heteroduplex regions, respectively.
The electron micrographs in Fig. 12 show two heteroduplex structures which formed after annealing of denatured T5+ DNA and T5 st(20) DNA. The loop structure on the right corresponds to Form 3 (Fig. 9), whereas the branched structure on the left with its two single-stranded segments corresponds to Form 4. The branched segments labeled C and D respresent that portion of T5 DNA (to the left and right of the C-D interruption) which is missing or deleted in st(20) DNA. The small circular structures are single-stranded $X174 DNA. Table III summarizes our measurements of the different heteroduplex forms obtained between wild type T5+ DNA and the various DNAs isolated from the T5 mutant phages. For each of the mutant DNAs examined, the data offer length measurements from the left end and right end of the deletion loop to the left end of the T5 chromosome. The difference between these two measurements is the size of the deletion indicated in the last column. The position of the C-D single strand interruption was ascertained from similar measurements of the double-branched structure seen in the heteroduplex between T5+ and St(O) DNAs. In addition, we have measured the A, B, and C segments of T5+ DNA (not shown in Table III)  As shown in Fig. 13, a physical map of the C segment containing the approximate locations of T5 tRNA genes was constructed from the hybridization data of Table I and the size and location of the DNA deletions shown in Table III The conditions for heteroduplex DNA formation with T5+ and T5 st(20) DNAs were as described under "Materials and Methods." The loop structure shown on the right corresponds to Form 3 (Fig. 9). The branched structure shown on the left with branches C and D correspond to Form 4 (Fig. 9). The small circular structures in the lower right section of the photograph are single-stranded #X174 DNA molecules.   give the kilobase length for the three spaces separating adjacent tRNA regions (I-11, II-III, and III-IV) where no tRNA genes have been positioned as well as the kilobase distances from the B-C nick to the left end of region I and the C-D nick to the left end of region IV. From these measurements, the total C segment length is estimated to contain 18.02 kilobases which is equivalent to a molecular size of 5.97 x 10' daltons, as a single-stranded DNA chain. In support of this suggestion, our results show that the genes which code for T5 tRNAs map almost exclusively in the C segment of T5 DNA, and RNA transcription in this region of the T5 chromosome has been reported to occur only from the heavy strand (13, 14). Although the hybridization data of Table I do not completely eliminate the possibility that the light DNA segments contain some information for tRNA synthesis, it seems unlikely that the same tDNA sequences would be present in both heavy and light chains for any single tRNA species.
Several heat-stable (st) mutants of T5 phage carry a deletion in the D and C regions of T5+ DNA (Figs. 4 and 7). We had previously reported that the St(O) T5 deletion mutant lacked the tRNA genes for arginine, tyrosine, and phenylalanine, and that the length and location of the St(O) deleted DNA segment could be determined accurately by electron microscopic examination of heteroduplex molecules formed between wild type and mutant DNAs (23). In subsequent studies by Scheible and Rhoades (31), heteroduplex mapping of heat-stable mutants of T5 bacteriophage has also been described. By correlating the physical locations and the sizes of the mutant deletions with the presence or absence of detectable tRNA genes, we have defined the positions and the maximal sizes of four tDNA-containing loci in T5+ DNA (Figs. 13 and 14).
The hybridization technique for locating tDNA sequences by use of labeled aminoacyl-tRNAs depends on several critical factors, which include the presence of functional tRNA molecules capable of accepting amino acids; the efficiency of enzyme charging; the specific activity of the individual radioactive amino acids; the stability of aminoacyl-tRNAs under the annealing conditions employed; the stability of the charged tRNA .DNA complex following hybridization, which includes treatment with Tl RNase; as well as the presence of a sufficient number of tRNA gene copies to permit detection. Variations in the levels of hybridization observed for some of the T5 mutant DNAs could be related to one or several of these factors. A negative hybridization result does not necessarily exclude the possible existence of the corresponding tRNA gene in the DNA genome. However, positive results do indicate the presence of tRNA complementary sequences in T5 DNA which, in turn, imply the presence of tRNA cistrons and transcription into active tRNA molecules. A molecule of T5 DNA should not need to possess an entire tRNA sequence in order to retain detectable radioactivity from hybridization with a labeled aminoacyl-tRNA. Partial DNA sequences of a tRNA molecule could give rise to a branched tRNA.DNA hybrid structure which, if stable, would be sufficient for its detection by the method used here. If the 3'.aminoacyl end of the tRNA molecule were part of the branched unannealed portion of the hybrid, and if a guanine residue were present in this branch, the hybrid complex would go undetected since our procedure includes Tl RNase treatment following hybridization.
If the guanine residue were close to the point of branching of the partial hybrid structure, it might not be readily accessible to the action of Tl RNase, thus causing incomplete digestion. It is possible therefore, since various deletion mutant DNAs are being used, that the number of imperfect hybrid structures formed would vary and that expected levels of hybridization relative to wild type DNA would not be observed. It is not clear whether this problem is related to some of the lower levels of hybridization detected for certain mutant DNAs. On the other hand, if T5 wild type DNA contains multiple tRNA cistrons for a single tRNA species (isoacceptor or repeated tRNA genes), and if the mutant DNA contains a smaller number of gene copies for this same tRNA species, one would expect to find lower levels of aminoacyl-tRNA hybridization to the mutant DNA. It is, therefore, important that when low hybridization values relative to wild type DNA, are encountered to distinguish (if possible) between intrinsic variations due to the procedure itself and meaningful values which reflect the size, location, and number of tRNA cistrons in the mutant DNAs.
The hybridization curves shown in Fig. 8 typify the results obtained for the 15 T5 aminoacyl-tRNA species. Those hybridization values for the mutant DNAs which differ slightly from the results obtained with T5+ DNA probably can be attributed to intrinsic variations in the assay procedure. However, the very low values found for phenylalanyl-tRNA hybridization to st(14) DNA (25% of that of wild type DNA), and the approximate half-values found for isoleucyl-tRNA hybridization to five of the mutant DNAs, cannot readily be ascribed to some intrinsic error, but are more probably due to the deletion of tDNA sequences (either partial or complete tDNA cistrons). This laboratory has reported on the presence of two isoacceptor T5 methionine tRNAs, tRNAfMet and tRNAmMet (21). In addition, we have recently been able to demonstrate that there are two T5 isoacceptor tRNAs for isoleucine, one of which is missing in the T5 mutant b2.3 These findings suggest that there are two different genes for tRNA"'" at two separate loci in the C segment of T5 DNA. This tentative conclusion would explain the 50% lower hybridization seen in Fig. 80 for the b2 mutant, and most probably for the b3, st(8), b4, and St(O) mutants as well. Experiments to confirm this conclusion are currently in progress.
The T5 tRNA cistron map (Fig. 13) derived from the hybridization and deletion measurements in this study shows four groups of tRNAs located close to one another at four separate regions in the C segment of T5 DNA. Region IV, which contains only the cistron(s) for tRNAA'g, encompasses the C-D single strand interruption; hence, the tDNAArg sequences could be positioned on either side of the C-D nick within the 3020-nucleotide length that comprises this region. Region III,