Cell, Vol. 12, 37-44,
September
1977,
Copyright
0 1977 by MIT
Further Mapping of Late Adenovirus Genes Cell-Free Translation of RNA Selected by Hybridization to Specific DNA Fragments J. B. Lewis, Cold Spring Cold Spring
porting data) the results of some of these experiments (with Barn HI and Hpa I) (Atkins et al., 1975b). The additional experiment with Hind III demonstrates that many late Ad2 mRNAs hybridize with lower affinity to a second site between positions 17.0 and 29.1, in addition to the primary site of hybridization.
C. W. Anderson* and J. F. Atkins Harbor Laboratory Harbor, New York 11724
Summary RNA isolated from the cytoplasm of human cells at late times after infection by adenovirus type 2 (Ad2) has been fractionated by hybridization to fragments of Ad2 DNA which were produced by digestion with the restriction endonucleases Hpa I, Eco RI, Barn HI and Hind Ill. Cell-free translation of these partially purified mRNAs indicates that the genes for the late Ad2 proteins lie within the following intervals on the conventional Ad2 map: 15K (4.4-17.0 map units), IX and lVaz (7.5-17.0) llla (29.1-40.9), Ill and V (29.1-57.0), pVlll (40.957.0), pVI and II (40.9-70.7), 1OOK (59.0-83.4), pVlll (70.7-83.4) and IV (85.0-100). In addition to the primary hybridization of the late Ad2 mRNAs to the regions indicated above, most late Ad2 mRNAs (except those for 15K, IX and IVa,) exhibited some hybridization to a secondary site between 17.0 and 29.1 map units. Introduction The partial purification of mRNA by hybridization to specific fragments of viral DNA, followed by cellfree translation to identify the mRNA so selected, provides a general method for mapping the position of the gene for a given protein on the viral DNA. We have used this method to identify adenovirus type 2 (Ad2)-coded polypeptides early in permissive infection and in transformed cells, and to map on the Ad2 genome the positions of the genes encoding these polypeptides (Lewis et al., 1976). We have also used this technique to form a partial map for the late Ad2 genes (Lewis et al., 1975), using the Eco RI restriction enzyme, which cleaves the right-hand portion of the Ad2 genome into five fragments, but leaves the left-hand 58.5% uncut (Mulder et al., 1974). To locate the late genes that map within the left end of Ad2, we report similar experiments using the endonucleases Barn HI, Hpa I and Hind Ill, which also cleave Ad2 into fragments that have been ordered with respect to the Ad2 genome (Mulder et al., 1974; R. Greene and C. Mulder, unpublished observations; R. J. Roberts, J. Sambrook and P. A. Sharp, unpublished observations). We have previously reported (without sup* Present address: York 11973.
Brookhaven
National
Laboratory,
Upton,
by
New
Results Hpa I was chosen to complement our previous study with Eco RI (Lewis et al., 1975), since Hpa I produces four cleavages within the Eco RI-A fragment (Mulder et al., 1974). Because the Hpa I-A, B and C fragments do not separate well on preparative agarose gels, these three fragments were redigested with Eco RI. A total of eleven fragments of Ad2 DNA were produced, ten of which were used in preparative hybridizations to fractionate cytoplasmic mRNA prepared from cells late after infection with Ad2. The mRNAs present in each fraction were identified by electrophoretic analysis of the polypeptides produced when each fraction of RNA was used to program cell-free protein synthesis. Among the polypeptides synthesized in vitro in response to unfractionated cytoplasmic RNA from Ad2-infected cells are at least twelve polypeptides (Figure lb) that we have shown to be identical to those synthesized in infected cells [using the criteria of co-migration upon polyacrylamide gel electrophoresis and, in several cases, of tryptic peptide analysis (Anderson et al., 1974)]. The mRNAs for each of these polypeptides can be selected by hybridization to whole Ad2 DNA (Figure Id). If bacteriophage h DNA is used for the hybridization (Figure lc), the products of cell-free translation are indistinguishable from those obtained when E. coli rRNA is added instead of mRNA (Figure la). In these latter cases, the only readily apparent polypeptide is globin, since its mRNA contaminates the reticulocyte initiation factors used for translation. Similarly, no late Ad2 polypeptide is made using mRNA selected by hybridization to the left 4.4% of Ad2 DNA (Hpa I-E; Figure le). The Hpa I-C fragment (Figure If) selects predominantly the mRNAs for polypeptides IVa,, 15K and IX. The mRNAs for these three proteins are not selected in appreciable quantity by any other fragment. In addition to these three, however, small amounts of almost all the late polypeptides are seen using mRNA selected by both the Hpa I-C fragment and the adjacent F fragment (Figure lg). Generally, the amounts of the other mRNAs selected by these two fragments were substantially less than the amount selected by the fragment to which the mRNA specifically bound
Cell 38
Late Adenovirus 39
Gene
Mapping
(for example, IV in Figures If and lg as opposed to Figure 1 n). Nonetheless, the apparently nonspecific binding of most late Ad2 mRNAs to these two fragments (extending from 4.4-27.9 map units) is much greater than the binding of mRNA to a nonhomologous DNA (Figure lc). Possible explanations of this phenomenon are discussed below. The large Hpa I-A fragment (Figure lh) selects the mRNAs for II, III, Illa, V, pVI and pVII. The small Hpa I-Eco RI fragment (57.0-58.4) (Figure li) and the Eco RI-B fragment (Figure lj) both select II and pVI, but not III, Illa, V or pVll mRNAs. In addition, the Eco RI-B fragment selects the 1OOK mRNA. The Eco RI-F fragment (Figure 1 k) selects 100K mRNA and pVlll mRNA. The Eco RI-D fragment (Figure le) selects smaller amounts of these same mRNAs. The most prominent mRNA selected by the D fragment, however, is the fiber (IV) mRNA. The amount of IV mRNA selected by Eco RI-D is nonetheless less than the amount selected by Hpa I-D (Figure ln)the primary site of hybridization for IV mRNA-and is also less than the amount selected by the Hpa I-F fragment (Figure lg)-a secondary site of hybridization for many late mRNA’s. The small fragment (83.4-85.0) (Figure lm) does not select any Ad2 mRNA, even though other fragments of comparably small size (Figures lg and li) do. The vast majority of the IV mRNA is selected by the Hpa I-D fragment (Figure In). The heterogeneous background of unidentified smaller polypeptides is probably the result of incomplete translation of IV mRNA. In this experiment, a significant amount of II and 1OOK mRNAs were also selected, although in much smaller amounts (compared to the amount of IV mRNA) than with whole Ad2 DNA (Figure Id). The gene order deduced from this experiment is shown in Figure 1 with respect to a map showing the fragments of Ad2 DNA used. Barn HI was chosen to produce additional cleavFigure RI
1. Translation
of Late Ad2 mRNA
Fractionated
by Hybridization
ages within the region of the large Hpa I-A fragment (C. Mulder and R. Greene, unpublished observations). Late cytoplasmic RNA from infected cells was fractionated by hybridization to each of the four Barn HI fragments of Ad2 DNA and, for comparison, to whole Ad2 DNA and bacteriophage h DNA, as in Lewis et al. (1975) (see also Figure 1). If the Barn HI-B fragment is used for the hybridization, predominantly the IVa, and 15K mRNAs are selected. Although very small amounts of II and IV mRNAs are also selected, these mRNAs are predominant in the fractions selected by hybridization to the C and A fragments. The D fragment selects the Ill, llla and V mRNAs. III and V mRNAs, but not llla mRNA, are also included in the populations selected by C fragment, which also selects II and pVll mRNAs. II mRNA is also selected by A fragment, as well as the mRNAs for lOOK, pVlll and IV. Not mapped in this experiment were pVI (which is not easily distinguished among the background of unidentified heterogeneous polypeptides produced by cell-free translation in the experiment shown in Figure 2) and component IX (which has been electrophoresed off the bottom of the gel). The gene order deduced by using the Barn HI fragments is illustrated in Figure 2. A more precise mapping of the left third of the Ad2 genome was provided by redigestion of the Barn HI-B fragment with Hind III (Figure 3). The Hind Ill-G fragment selects only the 15K mRNA (Figure 3c), and the Hind Ill-C fragment selects the IVa, and IX mRNAs, and to a lesser extent the 15K mRNA (Figure 3d). The Hind III-B fragment (Figure 3e) does not select either 15K, IX or lVaz mRNA, but does select the mRNAs of every other late Ad2 protein. The amounts of these mRNAs selected are substantially less than those selected by whole Ad2 DNA (Figure 3b), but are much larger than any interaction described as nonspecific adsorption to Fragments
of Ad2 DNA Produced
by Digestion
with Hpa I and Eco
Ad2 DNA was digested with Hpa I, and the resulting fragments were separated by preparative agarose gel electrophoresis. The A, B and C fragments were prepared together and redigested with Eco RI. Ten fragments of Ad2 DNA, as well as whole Ad2 DNA and bacteriophage A DNA, were used to fractionate late Ad2 mRNA according to sequence homology. Each hybridization used 100 pg of whole Ad2 or A DNA (or an equivalent amount of each DNA fragment) and 640 fig of cytoplasmic RNA prepared 30 hr after infection of KB cells with Ad2. About 16% of the mRNA recovered from each hybridization was used to program cell-free protein synthesis. A 25 ~1 reaction mixture contained 0.125 AzeO units of ribosomal subunits (Krebs II ascites), 2 *I of pH 5 enzyme (from Krebs II ascites cells, containing about 30-35 pg of protein and about 2 pg RNA), 0.25-0.40 ~1 of rabbit reticulocyte ribosomal wash precipitating between 30-40% saturation with (NH&SO, (6-8 pg protein), and 1 ~1 of reticulocyte ribosomal wash precipitating between 40-70% saturation with (NH&SO, (about 20 fig protein). The other ingredients were 1.0 mM ATP, 0.4 mM GTP, 10 mM creatine phosphate, 20 fig of creatine kinase per ml, 30 mM HEPES (N-2hydroxyethylpiperazine-N’-2-ethanesulfonic acid) (pH 7.2), 20 mM NH&I, 87 mM KCI. 2.4 mM Mg(OAc),, 0.6 mM spermidine. 0.5 mM dithiothreitol, 30 PM each of 19 amino acids (minus methionine), and 5-15 &i of YS-methionine at >300 Cilmmole specific activity. Incubation was for 90 min at 35°C. About 10% of each reaction mixture was analyzed by electrophoresis on a 17.5% sodium dodecylsulfate polyacrylamide gel, which was then dried down for autoradiography (2-4 day exposure). Cell-free protein synthesis was carried out with the addition to the system of (a) 2 pg E. coli rRNA, (b) 6 pg of cytoplasmic RNA prepared from KB cells 30 hr after infection with Ad2, RNA selected by hybridization to (c) bacteriophage A, (d) Ad2 DNA, (e) Hpa I-E fragment of Ad2 (O-4.4 map units), (f) Hpa I-C (4.4-25.5), (g) Hpa I-F (25.5-27.9), (h) Hpa I-A (27.9-57.0), (i) Hpa,l/Eco RI (57.0-58.5), (j) Eco RI-B (68.5-70.7), (k) Eco RI-F (70.7-75.9), (1) Eco RI-D (75.9-83.4), (m) Hpa VEco RI (83.4-85.0), (n) Hpa I-D (85.0-98.5). The order of the late genes along the Ad2 DNA deduced from this experiment is illustrated below the map of the fragments of Ad2 DNA produced by digestion with Hpa I, followed by redigestion of the Hpa I-A, B and C fragments with EGO RI.
Cell 40
Figure
2. Translation
of Late Ad2 mRNA
Fractionated
by Hybridization
to the Barn
HI Fragments
of Ad2 DNA
The fragments of Ad2 DNA produced by digestion with Barn HI were used to fractionate late Ad2 mRNA (prepared 24 hr after infection) as described in the legend to Figure 1 and by Lewis et al. (1975). Cell-free protein synthesis was programmed by RNAselected by hybridization to (a) bacteriophage k DNA, (b) Ad2 DNA, (c) Barn HI-B fragment of Ad2 DNA (O-29.1), (d) Barn HI-D (29.1-40.9), (e) Barn HI-C (40.9-59.0), (f) Barn HI-A (59.0-100.0). The order of the late genes along the Ad2 DNA deduced from this experiment is presented below the Barn HI restriction map of Ad2 DNA.
Late Adenovirus 41
Gene
Figure 3. Translation with Hind III
Mapping
of Late Ad2 mRNA
Fractionated
by Hybridization
to the Fragments
Produced
by Digestion
of the Barn HI-5
Fragment
Cell-free protein synthesis was programmed by RNA selected by hybridization to (a) bacteriophage X DNA, (b) Ad2 DNA, (c) Hind Ill-G fragment of Ad2 DNA (O-7.5 map units, (d) Hind Ill-C (7.5-17.0), (e) Hind III-B fragment trimmed by Barn HI (17.0-29.1). The order of the late genes along the Ad2 DNA deduced from these experiment is presented below the Hind Ill restriction map of the Barn HI-B fragment of Ad2 DNA.
Cell 42
when compared (Figure 3a).
to bacteriophage
A DNA control
Discussion The data presented here define those fragments of Ad2 DNA, produced by digestion with the endonucleases Eco RI, Hpa I, Barn HI and Hind III, that are complementary to the mRNAs for each of twelve late Ad2-coded polypeptides. Each of these polypeptides has been identified as a polypeptide seen in virions and/or in extracts of infected cells by the criteria of co-electrophoresis in SDS-polyacrylamide gels and, in some cases, by tryptic peptide analysis (Anderson et al., 1974). Included in these twelve is a polypeptide of 15,000 dalton apparent molecular weight (15K), which is not seen in virions but is seen in infected cells (Anderson, Baum and Gesteland, 1973; referred to as being of 14,500 dalton apparent molecular weight). We did not refer to this polypeptide in our previous studies of translation of late Ad2 mRNA (Anderson et al., 1974; Lewiset al., 1975) since this region of the gel contains many polypeptides resulting from partial synthesis of larger polypeptides, so that we could not be confident that 15K was an independent gene product. The experiment illustrated in Figure 3, demonstrates that 15K is the only polypeptide resulting from translation of the late mRNA selected by the Hind Ill-G fragment (O-7 map units) and thus must be an independent polypeptide. We have re-
Table
1. Map
Limits
of the Genes
for Late Ad2Coded
ported an early Ada-coded polypeptide of 15,000 dalton molecular weight whose gene also maps within the region 4.4-17.0 map units (Lewis et al., 1976). This region could code for about 160,000 dalton molecular weight of protein. Since the only other proteins mapping within the 4.4-17.0 region are IX and IVa, (total molecular weight 67,000 daltons), there would easily be room for two different 15,000 dalton molecular weight proteins. Preliminary peptide analysis indicates, however, that the early and late 15K polypeptides are at least partially related in sequence and may be identical (M. Harter and J. B. Lewis, unpublished observations). The 15K mRNA would thus appear different from the other early mRNAs (44-50K, 72K, 15.5K, 19K and 11 K) in that it is very active in cell-free translation when RNA is isolated late after infection. In contrast, the RNA preparations used here contained no detectable amounts of active 44-50K, 15.5K, 19K or 11K mRNAs, and only small and variable amounts of 72K mRNA. To construct a map of the physical position on the Ad2 DNA of the genes for each late Ad2 polypeptide, we have tabulated in Table 1 the regions of DNA to which each mRNA hybridizes. Only the primary sites of hybridization have been used. In particular, the secondary hybridization of most late mRNAs to the region 17.0-29.1 map units, seen with both the Hpa I- (Figure 1) and Hind Ill- (Figure 3) generated DNA fragments, has been ignored. The results of experiments using different enzymes
Proteins
Map Units Ad2 DNA Fragments
Complementary
to mRNA
Hpa I, Protein
Eco RI
Eco
RI
Barn
HI
Hind
Ill
15K
O-58.5
4.4-25.5
o-29.1
IX
O-58.5
4.4-25.5
-
7.5-l
o-1 7.0
IVa,
O-58.5
4.4-25.5
o-29.1
7.5-17.0
llla
O-58.5
27.9-57.0
29.1-40.9
-
Ill, v
O-58.5
27.9-57.0
29.1-59.0
-
pVll
O-58.5
27.9-57.0
40.9-59.0
pVI, II
O-70.7
27.9-70.7
7.0
Limits 4.4-17.0 7.5-17.0 7.5-17.0 29.1-40.9 29.1-57.0 40.9-57.0
40.9-100”
-
40.9-70.7 59.0-83.4
1OOK
58.5-83.4
58.5-83.4
59.0-100
-
pVlll
70.7-83.4
70.7-83.4
59.0-100
-
70.7-83.4
IV
83.4-100
85.0-100b
59.0-100
-
85.0-100
= Because of the background of heterogeneous polypeptides in Figure 2, pVI was not easily discernible among the products Of cell-free translation with RNA fractionated by hybridization to the Barn HI fragments. Consideration of the size of pVI mRNA (Anderson et al., 1974), plus the results of the Eco RI and Hpal mapping, necessitate that the pVI mRNA be contained no more than twelve map units from position 58.5. b The Hpa I-G fragment (98.5-100) was not included in this experiment, and thus the possibility that IV mRNA extends only to 98.5 is not excluded.
Late Adenovirus 43
Gene
Mapping
for digesting Ad2 DNA have been combined to identify the left and right limits within which the gene for a particular mRNA lies, as diagrammed in Figure 4. The fact that in many cases these limits are still wide with respect to the size of the protein is, in part, a reflection of the size of the probe used for mRNA complementarity. The average size of the DNA fragments used in this study is 13.4 map units or, in terms of coding capacity, about 170,000 daltons of polypeptide, which is 3 times the size of the average Ad2 late polypeptide. It should also, be emphasized that this technique maps the mRNA for a particular polypeptide, not the structural information for that polypeptide. Since the mRNAs for some Ad2 polypeptides appear to be much larger than needed for coding that particular polypeptide (Anderson et al., 1974), this technique cannot identify the exact position where the structural information is encoded. The mapping data in Figure 4 imply a left-to-right order of the genes for the twelve late Ad2 mRNAs which we have identified. This order is 15K, (IX, IVa,), Illa, (III, V), pVII, (pVI, II), lOOK, pVIII, IV. The relative orders of IX and IVa,, III and V, and pVI and II cannot be deduced. U. Pettersson and M. B. Mathews (manuscript submitted), however, have isolated IX mRNA by fractionation of RNA according to size, and have shown that it hybridizes to the region 9.4-10.7 map units, implying that it maps to the left of IVaz. In addition, consideration of the fact that pVI mRNA is very large (Anderson et al ., 1974), and that a large mRNA maps to the left of II mRNA and overlaps II mRNA (Chow et al., 1977a), suggests that pVI maps to the left of II. In theory, a more precise map could be constructed by using smaller DNA fragments to select mRNA. Another technique that has provided much greater precision has been to visualize individual molecules of mRNA hybridized to Ad2 DNA molecules (Chow et al., 1977a) in the electron microscope by the “R loop” technique (Thomas, White and Davis, 1976). This R loop map has been correlated with the map in Figure 4 to identify with great precision the position on the Ad2 DNA of the genes for many late Ad2 mRNAs. Thus the combination of these two techniques permits positioning some late Ad2 genes to within ~0.2 map units, in contrast to the approximate positions reported here. The secondary hybridization of all late mRNAs except 15K, IX and IVa, to the region 17.0-29.1 map units was originally difficult to explain. Clearly, all these genes could not be coded there, yet the secondary hybridization appears both with the set of fragments produced by Hpa I and Eco RI (Figure 1) and with the Hind Ill-produced fragments (Figure 3), and is thus not likely to result from contamination of DNA fragments. Interpretation of these re-
15K ~IX,iXa,,
ma lIu,P)
p-m (pm,nl
o-mn
dI.ll
Irx,mo, ,15K 0
10
,
I I
20
a,p ma
30
I p-m
40
I 50
m
IOOK p-m
60
IOOK 70
,Ip eo
Figure 4. Limits for Individual Late Genes Deduced Free Translation of mRNA Selected by Hybridization Fragments of Ad2 DNA
90
100
from Cellto Specific
The limits of each late gene demonstrated by each of the experiments in Lewis et al. (1975) and in Figures l-3 are shown schematically, and the order of the late genes deduced from these limits is given.
sults has been greatly facilitated recently by several different investigations from which we conclude that many late Ad2 mRNAs contain a common 5’ end sequence that is not encoded on the Ad2 DNA adjacent to the main coding region for that gene, but rather is encoded by sequences left of map position 29. Gelinas and Roberts (1977) have shown that most late Ad2 mRNAs have a common 5’ sequence of at least an undecanucleotide. Klessig (1977) and Dunn and Hassell (1977) have demonstrated that several specific late Ad2 and Ad2+NDl mRNAs each hybridize to several regions on the Ad2 genome, including one major site specific to each mRNA, and one or more secondary sites within the region 17.0-31.5 map units. Furthermore, Klessig (1977) has shown that the 5’ ends of several late Ad2 mRNAs are not protected from ribonuclease digestion by hybridization to the DNA fragment to which the mRNA predominantly hybridizes, but are protected by hybridization to several left-end fragments of Ad2 DNA, and in particular, that the common 5’ undecanucleotide is protected by hybridization to the region 14.7-17.0 map units. Chow et al. (197713) have demonstrated by electron microscopy that the 5’ ends of most late Ad2 mRNAs (including hexon, 100K and fiber mRNAs) each hybridize to sequences at 16.6, 19.6 and 26.6 map units, remote from the main coding region for any of these mRNAs. This same phenomenon has also been demonstrated for hexon mRNA by Berget, Moore and Sharp (1977). Thus short sequences from these three widely separated regions somehow become contiguous on the 5’ ends of most, if not all, late Ad2 mRNAs coded to the right of 29 map units, explaining why these mRNAs hybridize with lowered efficiency to the region 17.0-29.1, as well as to the main coding sequence for each mRNA. Experimental
Procedures
Adenovirus DNA was prepared from virus (Pettersson and Sambrook, 1973) and digested with Barn HI (Wilson and Young, 1975) or Hpa I (Pettersson et al., 1973). The fragments of DNA produced were separated by electrophoresis on agarose gels and re-
Cell 44
covered from the gel by treatment with sodium perchlorate as previously described (Lewis et al., 1975). The fragments obtained were >QO% pure as measured by electrophoresis of an aliquot on agarose gels. For some experiments, the purified Barn HI-B fragment was redigested with Hind Ill to give three fragments. To be certain that the Barn HI-B used for this purpose was free of contamination by right-end sequences (Afragment), the Barn HI-B was obtained from a double digestion with Barn HI and Eco RI, which split the Barn HI-A fragment into five fragments, facilitating separation from the B fragment. Because the Hpa I-A, B and C fragments do not separate well on preparative gels, they were eluted together and redigested with Eco RI to give seven fragments. Cytoplasmic RNA was prepared from suspension cultures of KB cells at late times (25-30 hr) after infection with Ad2, as previously described (Anderson et al., 1974). Partially purified mRNA was prepared by hybridization to fragments of DNA (Lewis et al., 1975) and then translated in a cell-free translation system as originally described by Schreier and Staehelin (1973) and modified by Anderson et al. (1974) and by Atkins et al. (1975a). The product of cell-free protein synthesis, labeled with “S-methionine, was analyzed for the presence of late Ad2 polypeptides by sodium dodecylsulfate polyacrylamide gel electrophoresis (Anderson et al., 1973.
Acknowledgments We are grateful to P. Myers and R. J. Roberts for generous gifts of Eco RI, Hpa I, Barn HI and Hind III, and to P. Baum and R. Solem for expert technical assistance. Financial support was provided by a National Cancer Institute Cancer Center grant and by an American Cancer Society research grant to J.B.L., who is a Rita Allen Foundation Scholar. Received
June
9, 1977;
revised
July 5, 1977.
References Anderson, C. W., Baum, P. R. and Gesteland, R. F. (1973). Processing of adenovirus 2-induced proteins. J. Virol. 72, 241-252. Anderson, C. W., Lewis, J. B., Atkins, J. F. and Gesteland, R. F. (1974). Cell-free synthesis of adenovirus 2 proteins programmed by fractionated mRNA: a comparison of polypeptide products and mRNA lengths. Proc. Nat. Acad. Sci. USA 77, 2756-2760. Atkins, J. F., Lewis, J. B., Anderson, C. W. and Gesteland, R. F. (1975a). Enhanced differential synthesis of proteins in a mammalian cell-free system by addition of polyamines. J. Biol. Chem. 250, 5688-5695. Atkins, J. F., Lewis, J. B., Anderson, C. W., Baum, P. R. and Gesteland, R. F. (1975b). Mapping of adenovirus-2 genes bytranslation of RNA selected by hybridization. In INSERM Symposium 47, A. L. Haenni and G. Beaud, eds. (Paris: Inserm), pp. 293-298. Berget, S. M., Moore, Sci. USA, in press.
C. and Sharp,
P. A. (1977).
Proc.
Nat. Acad.
Chow, L. T., Roberts, J. M., Lewis, J. B. and Broker, T. R. (1977a). A map of cytoplasmic RNA transcripts from lytic adenovirus type 2, determined by electron microscopy of RNA:DNA hybrids. Cell 77, 819-836. Chow, L. T., Gelinas, R. E., Broker, T. R. and Roberts, R. J. (1977b). An amazing sequence arrangement at the 5’ ends of adenovirus 2 messenger RNA. Cell 72, 1-8. Dunn, A. R. and Hassell, J. A. (1977). A novel method to map transcripts: evidence for homology between an adenovirus mRNA and discrete multiple regions of the viral genome. Cell 72, 23-36. Gelinas, R. E. and Roberts, Ft. J. (1977). One predominant 5’undecanucleotide in adenovirus 2 late messenger RNAs. Cell 77, 533-544. Klessig,
D. F. (1977).
Two
adenovirus
mRNAs
have
a common
5’
terminal leader sequence the main coding regions
encoded for these
at least IO kb upstream from messengers. Cell 72, 9-21.
Lewis, J. B., Atkins, J. F., Anderson, C. W., Baum, P. R. and Gesteland, R. F. (1975). Mapping of late adenovirus genes by cellfree translation of RNA selected by hybridization to specific DNA fragments. Proc. Nat. Acad. Sci. USA 72, 1344-1348. Lewis, J. B., Atkins, J. F., Baum. P. R.. Solem, R., Gesteland. R. F. and Anderson, C. W. (1976). Location and identification of the genes for adenovirus type 2 early polypeptides. Cell 7, 141-151. Mulder, C., Arrand, J. R., Delius, H., Keller, W., Pettersson, U., Roberts, R. J. and Sharp, P. A. (1974). Cleavage maps of DNA from adenovirus types 2 and 5 by restriction endonucleases Eco RI and Hpal. Cold Spring Harbor Symp. Quant. Biol. 39,397-400. Pettersson, U. and Sambrook, J. (1973). The amount of viral DNA in the genome of cells transformed by adenovirus type 2. J. Mol. Biol. 73, 125-130. Pettersson, U., Mulder, C., Delius, H. and Sharp, P. A. (1973). Cleavage of adenovirus type 2 DNA into six unique fragments by endonuclease R.Rl. Proc. Nat. Acad. Sci. USA 70, 200-204. Schreier, M. and Staehelin, T. (1973). Initiation of mammalian protein synthesis: the importance of ribosome and initiation factor quality for in vitro synthesis. J. Mol. Biol. 73, 329-349. Thomas, M., White, R. L. and Davis, RNA to double strand DNA: formation Sci. USA 73, 2294-2298.
R. W. (1976). of R-loops.
Wilson, G. A. and Young, F. (1975). specific nuclease (Baml) from Baci//us Mol. Biol. 97, 723-725.
Hybridization of Proc. Nat. Acad.
isolation of a sequenceamyloliquefaciens H. J.
September
1977,
Copyright
0 1977 by MIT
Further Mapping of Late Adenovirus Genes Cell-Free Translation of RNA Selected by Hybridization to Specific DNA Fragments J. B. Lewis, Cold Spring Cold Spring
porting data) the results of some of these experiments (with Barn HI and Hpa I) (Atkins et al., 1975b). The additional experiment with Hind III demonstrates that many late Ad2 mRNAs hybridize with lower affinity to a second site between positions 17.0 and 29.1, in addition to the primary site of hybridization.
C. W. Anderson* and J. F. Atkins Harbor Laboratory Harbor, New York 11724
Summary RNA isolated from the cytoplasm of human cells at late times after infection by adenovirus type 2 (Ad2) has been fractionated by hybridization to fragments of Ad2 DNA which were produced by digestion with the restriction endonucleases Hpa I, Eco RI, Barn HI and Hind Ill. Cell-free translation of these partially purified mRNAs indicates that the genes for the late Ad2 proteins lie within the following intervals on the conventional Ad2 map: 15K (4.4-17.0 map units), IX and lVaz (7.5-17.0) llla (29.1-40.9), Ill and V (29.1-57.0), pVlll (40.957.0), pVI and II (40.9-70.7), 1OOK (59.0-83.4), pVlll (70.7-83.4) and IV (85.0-100). In addition to the primary hybridization of the late Ad2 mRNAs to the regions indicated above, most late Ad2 mRNAs (except those for 15K, IX and IVa,) exhibited some hybridization to a secondary site between 17.0 and 29.1 map units. Introduction The partial purification of mRNA by hybridization to specific fragments of viral DNA, followed by cellfree translation to identify the mRNA so selected, provides a general method for mapping the position of the gene for a given protein on the viral DNA. We have used this method to identify adenovirus type 2 (Ad2)-coded polypeptides early in permissive infection and in transformed cells, and to map on the Ad2 genome the positions of the genes encoding these polypeptides (Lewis et al., 1976). We have also used this technique to form a partial map for the late Ad2 genes (Lewis et al., 1975), using the Eco RI restriction enzyme, which cleaves the right-hand portion of the Ad2 genome into five fragments, but leaves the left-hand 58.5% uncut (Mulder et al., 1974). To locate the late genes that map within the left end of Ad2, we report similar experiments using the endonucleases Barn HI, Hpa I and Hind Ill, which also cleave Ad2 into fragments that have been ordered with respect to the Ad2 genome (Mulder et al., 1974; R. Greene and C. Mulder, unpublished observations; R. J. Roberts, J. Sambrook and P. A. Sharp, unpublished observations). We have previously reported (without sup* Present address: York 11973.
Brookhaven
National
Laboratory,
Upton,
by
New
Results Hpa I was chosen to complement our previous study with Eco RI (Lewis et al., 1975), since Hpa I produces four cleavages within the Eco RI-A fragment (Mulder et al., 1974). Because the Hpa I-A, B and C fragments do not separate well on preparative agarose gels, these three fragments were redigested with Eco RI. A total of eleven fragments of Ad2 DNA were produced, ten of which were used in preparative hybridizations to fractionate cytoplasmic mRNA prepared from cells late after infection with Ad2. The mRNAs present in each fraction were identified by electrophoretic analysis of the polypeptides produced when each fraction of RNA was used to program cell-free protein synthesis. Among the polypeptides synthesized in vitro in response to unfractionated cytoplasmic RNA from Ad2-infected cells are at least twelve polypeptides (Figure lb) that we have shown to be identical to those synthesized in infected cells [using the criteria of co-migration upon polyacrylamide gel electrophoresis and, in several cases, of tryptic peptide analysis (Anderson et al., 1974)]. The mRNAs for each of these polypeptides can be selected by hybridization to whole Ad2 DNA (Figure Id). If bacteriophage h DNA is used for the hybridization (Figure lc), the products of cell-free translation are indistinguishable from those obtained when E. coli rRNA is added instead of mRNA (Figure la). In these latter cases, the only readily apparent polypeptide is globin, since its mRNA contaminates the reticulocyte initiation factors used for translation. Similarly, no late Ad2 polypeptide is made using mRNA selected by hybridization to the left 4.4% of Ad2 DNA (Hpa I-E; Figure le). The Hpa I-C fragment (Figure If) selects predominantly the mRNAs for polypeptides IVa,, 15K and IX. The mRNAs for these three proteins are not selected in appreciable quantity by any other fragment. In addition to these three, however, small amounts of almost all the late polypeptides are seen using mRNA selected by both the Hpa I-C fragment and the adjacent F fragment (Figure lg). Generally, the amounts of the other mRNAs selected by these two fragments were substantially less than the amount selected by the fragment to which the mRNA specifically bound
Cell 38
Late Adenovirus 39
Gene
Mapping
(for example, IV in Figures If and lg as opposed to Figure 1 n). Nonetheless, the apparently nonspecific binding of most late Ad2 mRNAs to these two fragments (extending from 4.4-27.9 map units) is much greater than the binding of mRNA to a nonhomologous DNA (Figure lc). Possible explanations of this phenomenon are discussed below. The large Hpa I-A fragment (Figure lh) selects the mRNAs for II, III, Illa, V, pVI and pVII. The small Hpa I-Eco RI fragment (57.0-58.4) (Figure li) and the Eco RI-B fragment (Figure lj) both select II and pVI, but not III, Illa, V or pVll mRNAs. In addition, the Eco RI-B fragment selects the 1OOK mRNA. The Eco RI-F fragment (Figure 1 k) selects 100K mRNA and pVlll mRNA. The Eco RI-D fragment (Figure le) selects smaller amounts of these same mRNAs. The most prominent mRNA selected by the D fragment, however, is the fiber (IV) mRNA. The amount of IV mRNA selected by Eco RI-D is nonetheless less than the amount selected by Hpa I-D (Figure ln)the primary site of hybridization for IV mRNA-and is also less than the amount selected by the Hpa I-F fragment (Figure lg)-a secondary site of hybridization for many late mRNA’s. The small fragment (83.4-85.0) (Figure lm) does not select any Ad2 mRNA, even though other fragments of comparably small size (Figures lg and li) do. The vast majority of the IV mRNA is selected by the Hpa I-D fragment (Figure In). The heterogeneous background of unidentified smaller polypeptides is probably the result of incomplete translation of IV mRNA. In this experiment, a significant amount of II and 1OOK mRNAs were also selected, although in much smaller amounts (compared to the amount of IV mRNA) than with whole Ad2 DNA (Figure Id). The gene order deduced from this experiment is shown in Figure 1 with respect to a map showing the fragments of Ad2 DNA used. Barn HI was chosen to produce additional cleavFigure RI
1. Translation
of Late Ad2 mRNA
Fractionated
by Hybridization
ages within the region of the large Hpa I-A fragment (C. Mulder and R. Greene, unpublished observations). Late cytoplasmic RNA from infected cells was fractionated by hybridization to each of the four Barn HI fragments of Ad2 DNA and, for comparison, to whole Ad2 DNA and bacteriophage h DNA, as in Lewis et al. (1975) (see also Figure 1). If the Barn HI-B fragment is used for the hybridization, predominantly the IVa, and 15K mRNAs are selected. Although very small amounts of II and IV mRNAs are also selected, these mRNAs are predominant in the fractions selected by hybridization to the C and A fragments. The D fragment selects the Ill, llla and V mRNAs. III and V mRNAs, but not llla mRNA, are also included in the populations selected by C fragment, which also selects II and pVll mRNAs. II mRNA is also selected by A fragment, as well as the mRNAs for lOOK, pVlll and IV. Not mapped in this experiment were pVI (which is not easily distinguished among the background of unidentified heterogeneous polypeptides produced by cell-free translation in the experiment shown in Figure 2) and component IX (which has been electrophoresed off the bottom of the gel). The gene order deduced by using the Barn HI fragments is illustrated in Figure 2. A more precise mapping of the left third of the Ad2 genome was provided by redigestion of the Barn HI-B fragment with Hind III (Figure 3). The Hind Ill-G fragment selects only the 15K mRNA (Figure 3c), and the Hind Ill-C fragment selects the IVa, and IX mRNAs, and to a lesser extent the 15K mRNA (Figure 3d). The Hind III-B fragment (Figure 3e) does not select either 15K, IX or lVaz mRNA, but does select the mRNAs of every other late Ad2 protein. The amounts of these mRNAs selected are substantially less than those selected by whole Ad2 DNA (Figure 3b), but are much larger than any interaction described as nonspecific adsorption to Fragments
of Ad2 DNA Produced
by Digestion
with Hpa I and Eco
Ad2 DNA was digested with Hpa I, and the resulting fragments were separated by preparative agarose gel electrophoresis. The A, B and C fragments were prepared together and redigested with Eco RI. Ten fragments of Ad2 DNA, as well as whole Ad2 DNA and bacteriophage A DNA, were used to fractionate late Ad2 mRNA according to sequence homology. Each hybridization used 100 pg of whole Ad2 or A DNA (or an equivalent amount of each DNA fragment) and 640 fig of cytoplasmic RNA prepared 30 hr after infection of KB cells with Ad2. About 16% of the mRNA recovered from each hybridization was used to program cell-free protein synthesis. A 25 ~1 reaction mixture contained 0.125 AzeO units of ribosomal subunits (Krebs II ascites), 2 *I of pH 5 enzyme (from Krebs II ascites cells, containing about 30-35 pg of protein and about 2 pg RNA), 0.25-0.40 ~1 of rabbit reticulocyte ribosomal wash precipitating between 30-40% saturation with (NH&SO, (6-8 pg protein), and 1 ~1 of reticulocyte ribosomal wash precipitating between 40-70% saturation with (NH&SO, (about 20 fig protein). The other ingredients were 1.0 mM ATP, 0.4 mM GTP, 10 mM creatine phosphate, 20 fig of creatine kinase per ml, 30 mM HEPES (N-2hydroxyethylpiperazine-N’-2-ethanesulfonic acid) (pH 7.2), 20 mM NH&I, 87 mM KCI. 2.4 mM Mg(OAc),, 0.6 mM spermidine. 0.5 mM dithiothreitol, 30 PM each of 19 amino acids (minus methionine), and 5-15 &i of YS-methionine at >300 Cilmmole specific activity. Incubation was for 90 min at 35°C. About 10% of each reaction mixture was analyzed by electrophoresis on a 17.5% sodium dodecylsulfate polyacrylamide gel, which was then dried down for autoradiography (2-4 day exposure). Cell-free protein synthesis was carried out with the addition to the system of (a) 2 pg E. coli rRNA, (b) 6 pg of cytoplasmic RNA prepared from KB cells 30 hr after infection with Ad2, RNA selected by hybridization to (c) bacteriophage A, (d) Ad2 DNA, (e) Hpa I-E fragment of Ad2 (O-4.4 map units), (f) Hpa I-C (4.4-25.5), (g) Hpa I-F (25.5-27.9), (h) Hpa I-A (27.9-57.0), (i) Hpa,l/Eco RI (57.0-58.5), (j) Eco RI-B (68.5-70.7), (k) Eco RI-F (70.7-75.9), (1) Eco RI-D (75.9-83.4), (m) Hpa VEco RI (83.4-85.0), (n) Hpa I-D (85.0-98.5). The order of the late genes along the Ad2 DNA deduced from this experiment is illustrated below the map of the fragments of Ad2 DNA produced by digestion with Hpa I, followed by redigestion of the Hpa I-A, B and C fragments with EGO RI.
Cell 40
Figure
2. Translation
of Late Ad2 mRNA
Fractionated
by Hybridization
to the Barn
HI Fragments
of Ad2 DNA
The fragments of Ad2 DNA produced by digestion with Barn HI were used to fractionate late Ad2 mRNA (prepared 24 hr after infection) as described in the legend to Figure 1 and by Lewis et al. (1975). Cell-free protein synthesis was programmed by RNAselected by hybridization to (a) bacteriophage k DNA, (b) Ad2 DNA, (c) Barn HI-B fragment of Ad2 DNA (O-29.1), (d) Barn HI-D (29.1-40.9), (e) Barn HI-C (40.9-59.0), (f) Barn HI-A (59.0-100.0). The order of the late genes along the Ad2 DNA deduced from this experiment is presented below the Barn HI restriction map of Ad2 DNA.
Late Adenovirus 41
Gene
Figure 3. Translation with Hind III
Mapping
of Late Ad2 mRNA
Fractionated
by Hybridization
to the Fragments
Produced
by Digestion
of the Barn HI-5
Fragment
Cell-free protein synthesis was programmed by RNA selected by hybridization to (a) bacteriophage X DNA, (b) Ad2 DNA, (c) Hind Ill-G fragment of Ad2 DNA (O-7.5 map units, (d) Hind Ill-C (7.5-17.0), (e) Hind III-B fragment trimmed by Barn HI (17.0-29.1). The order of the late genes along the Ad2 DNA deduced from these experiment is presented below the Hind Ill restriction map of the Barn HI-B fragment of Ad2 DNA.
Cell 42
when compared (Figure 3a).
to bacteriophage
A DNA control
Discussion The data presented here define those fragments of Ad2 DNA, produced by digestion with the endonucleases Eco RI, Hpa I, Barn HI and Hind III, that are complementary to the mRNAs for each of twelve late Ad2-coded polypeptides. Each of these polypeptides has been identified as a polypeptide seen in virions and/or in extracts of infected cells by the criteria of co-electrophoresis in SDS-polyacrylamide gels and, in some cases, by tryptic peptide analysis (Anderson et al., 1974). Included in these twelve is a polypeptide of 15,000 dalton apparent molecular weight (15K), which is not seen in virions but is seen in infected cells (Anderson, Baum and Gesteland, 1973; referred to as being of 14,500 dalton apparent molecular weight). We did not refer to this polypeptide in our previous studies of translation of late Ad2 mRNA (Anderson et al., 1974; Lewiset al., 1975) since this region of the gel contains many polypeptides resulting from partial synthesis of larger polypeptides, so that we could not be confident that 15K was an independent gene product. The experiment illustrated in Figure 3, demonstrates that 15K is the only polypeptide resulting from translation of the late mRNA selected by the Hind Ill-G fragment (O-7 map units) and thus must be an independent polypeptide. We have re-
Table
1. Map
Limits
of the Genes
for Late Ad2Coded
ported an early Ada-coded polypeptide of 15,000 dalton molecular weight whose gene also maps within the region 4.4-17.0 map units (Lewis et al., 1976). This region could code for about 160,000 dalton molecular weight of protein. Since the only other proteins mapping within the 4.4-17.0 region are IX and IVa, (total molecular weight 67,000 daltons), there would easily be room for two different 15,000 dalton molecular weight proteins. Preliminary peptide analysis indicates, however, that the early and late 15K polypeptides are at least partially related in sequence and may be identical (M. Harter and J. B. Lewis, unpublished observations). The 15K mRNA would thus appear different from the other early mRNAs (44-50K, 72K, 15.5K, 19K and 11 K) in that it is very active in cell-free translation when RNA is isolated late after infection. In contrast, the RNA preparations used here contained no detectable amounts of active 44-50K, 15.5K, 19K or 11K mRNAs, and only small and variable amounts of 72K mRNA. To construct a map of the physical position on the Ad2 DNA of the genes for each late Ad2 polypeptide, we have tabulated in Table 1 the regions of DNA to which each mRNA hybridizes. Only the primary sites of hybridization have been used. In particular, the secondary hybridization of most late mRNAs to the region 17.0-29.1 map units, seen with both the Hpa I- (Figure 1) and Hind Ill- (Figure 3) generated DNA fragments, has been ignored. The results of experiments using different enzymes
Proteins
Map Units Ad2 DNA Fragments
Complementary
to mRNA
Hpa I, Protein
Eco RI
Eco
RI
Barn
HI
Hind
Ill
15K
O-58.5
4.4-25.5
o-29.1
IX
O-58.5
4.4-25.5
-
7.5-l
o-1 7.0
IVa,
O-58.5
4.4-25.5
o-29.1
7.5-17.0
llla
O-58.5
27.9-57.0
29.1-40.9
-
Ill, v
O-58.5
27.9-57.0
29.1-59.0
-
pVll
O-58.5
27.9-57.0
40.9-59.0
pVI, II
O-70.7
27.9-70.7
7.0
Limits 4.4-17.0 7.5-17.0 7.5-17.0 29.1-40.9 29.1-57.0 40.9-57.0
40.9-100”
-
40.9-70.7 59.0-83.4
1OOK
58.5-83.4
58.5-83.4
59.0-100
-
pVlll
70.7-83.4
70.7-83.4
59.0-100
-
70.7-83.4
IV
83.4-100
85.0-100b
59.0-100
-
85.0-100
= Because of the background of heterogeneous polypeptides in Figure 2, pVI was not easily discernible among the products Of cell-free translation with RNA fractionated by hybridization to the Barn HI fragments. Consideration of the size of pVI mRNA (Anderson et al., 1974), plus the results of the Eco RI and Hpal mapping, necessitate that the pVI mRNA be contained no more than twelve map units from position 58.5. b The Hpa I-G fragment (98.5-100) was not included in this experiment, and thus the possibility that IV mRNA extends only to 98.5 is not excluded.
Late Adenovirus 43
Gene
Mapping
for digesting Ad2 DNA have been combined to identify the left and right limits within which the gene for a particular mRNA lies, as diagrammed in Figure 4. The fact that in many cases these limits are still wide with respect to the size of the protein is, in part, a reflection of the size of the probe used for mRNA complementarity. The average size of the DNA fragments used in this study is 13.4 map units or, in terms of coding capacity, about 170,000 daltons of polypeptide, which is 3 times the size of the average Ad2 late polypeptide. It should also, be emphasized that this technique maps the mRNA for a particular polypeptide, not the structural information for that polypeptide. Since the mRNAs for some Ad2 polypeptides appear to be much larger than needed for coding that particular polypeptide (Anderson et al., 1974), this technique cannot identify the exact position where the structural information is encoded. The mapping data in Figure 4 imply a left-to-right order of the genes for the twelve late Ad2 mRNAs which we have identified. This order is 15K, (IX, IVa,), Illa, (III, V), pVII, (pVI, II), lOOK, pVIII, IV. The relative orders of IX and IVa,, III and V, and pVI and II cannot be deduced. U. Pettersson and M. B. Mathews (manuscript submitted), however, have isolated IX mRNA by fractionation of RNA according to size, and have shown that it hybridizes to the region 9.4-10.7 map units, implying that it maps to the left of IVaz. In addition, consideration of the fact that pVI mRNA is very large (Anderson et al ., 1974), and that a large mRNA maps to the left of II mRNA and overlaps II mRNA (Chow et al., 1977a), suggests that pVI maps to the left of II. In theory, a more precise map could be constructed by using smaller DNA fragments to select mRNA. Another technique that has provided much greater precision has been to visualize individual molecules of mRNA hybridized to Ad2 DNA molecules (Chow et al., 1977a) in the electron microscope by the “R loop” technique (Thomas, White and Davis, 1976). This R loop map has been correlated with the map in Figure 4 to identify with great precision the position on the Ad2 DNA of the genes for many late Ad2 mRNAs. Thus the combination of these two techniques permits positioning some late Ad2 genes to within ~0.2 map units, in contrast to the approximate positions reported here. The secondary hybridization of all late mRNAs except 15K, IX and IVa, to the region 17.0-29.1 map units was originally difficult to explain. Clearly, all these genes could not be coded there, yet the secondary hybridization appears both with the set of fragments produced by Hpa I and Eco RI (Figure 1) and with the Hind Ill-produced fragments (Figure 3), and is thus not likely to result from contamination of DNA fragments. Interpretation of these re-
15K ~IX,iXa,,
ma lIu,P)
p-m (pm,nl
o-mn
dI.ll
Irx,mo, ,15K 0
10
,
I I
20
a,p ma
30
I p-m
40
I 50
m
IOOK p-m
60
IOOK 70
,Ip eo
Figure 4. Limits for Individual Late Genes Deduced Free Translation of mRNA Selected by Hybridization Fragments of Ad2 DNA
90
100
from Cellto Specific
The limits of each late gene demonstrated by each of the experiments in Lewis et al. (1975) and in Figures l-3 are shown schematically, and the order of the late genes deduced from these limits is given.
sults has been greatly facilitated recently by several different investigations from which we conclude that many late Ad2 mRNAs contain a common 5’ end sequence that is not encoded on the Ad2 DNA adjacent to the main coding region for that gene, but rather is encoded by sequences left of map position 29. Gelinas and Roberts (1977) have shown that most late Ad2 mRNAs have a common 5’ sequence of at least an undecanucleotide. Klessig (1977) and Dunn and Hassell (1977) have demonstrated that several specific late Ad2 and Ad2+NDl mRNAs each hybridize to several regions on the Ad2 genome, including one major site specific to each mRNA, and one or more secondary sites within the region 17.0-31.5 map units. Furthermore, Klessig (1977) has shown that the 5’ ends of several late Ad2 mRNAs are not protected from ribonuclease digestion by hybridization to the DNA fragment to which the mRNA predominantly hybridizes, but are protected by hybridization to several left-end fragments of Ad2 DNA, and in particular, that the common 5’ undecanucleotide is protected by hybridization to the region 14.7-17.0 map units. Chow et al. (197713) have demonstrated by electron microscopy that the 5’ ends of most late Ad2 mRNAs (including hexon, 100K and fiber mRNAs) each hybridize to sequences at 16.6, 19.6 and 26.6 map units, remote from the main coding region for any of these mRNAs. This same phenomenon has also been demonstrated for hexon mRNA by Berget, Moore and Sharp (1977). Thus short sequences from these three widely separated regions somehow become contiguous on the 5’ ends of most, if not all, late Ad2 mRNAs coded to the right of 29 map units, explaining why these mRNAs hybridize with lowered efficiency to the region 17.0-29.1, as well as to the main coding sequence for each mRNA. Experimental
Procedures
Adenovirus DNA was prepared from virus (Pettersson and Sambrook, 1973) and digested with Barn HI (Wilson and Young, 1975) or Hpa I (Pettersson et al., 1973). The fragments of DNA produced were separated by electrophoresis on agarose gels and re-
Cell 44
covered from the gel by treatment with sodium perchlorate as previously described (Lewis et al., 1975). The fragments obtained were >QO% pure as measured by electrophoresis of an aliquot on agarose gels. For some experiments, the purified Barn HI-B fragment was redigested with Hind Ill to give three fragments. To be certain that the Barn HI-B used for this purpose was free of contamination by right-end sequences (Afragment), the Barn HI-B was obtained from a double digestion with Barn HI and Eco RI, which split the Barn HI-A fragment into five fragments, facilitating separation from the B fragment. Because the Hpa I-A, B and C fragments do not separate well on preparative gels, they were eluted together and redigested with Eco RI to give seven fragments. Cytoplasmic RNA was prepared from suspension cultures of KB cells at late times (25-30 hr) after infection with Ad2, as previously described (Anderson et al., 1974). Partially purified mRNA was prepared by hybridization to fragments of DNA (Lewis et al., 1975) and then translated in a cell-free translation system as originally described by Schreier and Staehelin (1973) and modified by Anderson et al. (1974) and by Atkins et al. (1975a). The product of cell-free protein synthesis, labeled with “S-methionine, was analyzed for the presence of late Ad2 polypeptides by sodium dodecylsulfate polyacrylamide gel electrophoresis (Anderson et al., 1973.
Acknowledgments We are grateful to P. Myers and R. J. Roberts for generous gifts of Eco RI, Hpa I, Barn HI and Hind III, and to P. Baum and R. Solem for expert technical assistance. Financial support was provided by a National Cancer Institute Cancer Center grant and by an American Cancer Society research grant to J.B.L., who is a Rita Allen Foundation Scholar. Received
June
9, 1977;
revised
July 5, 1977.
References Anderson, C. W., Baum, P. R. and Gesteland, R. F. (1973). Processing of adenovirus 2-induced proteins. J. Virol. 72, 241-252. Anderson, C. W., Lewis, J. B., Atkins, J. F. and Gesteland, R. F. (1974). Cell-free synthesis of adenovirus 2 proteins programmed by fractionated mRNA: a comparison of polypeptide products and mRNA lengths. Proc. Nat. Acad. Sci. USA 77, 2756-2760. Atkins, J. F., Lewis, J. B., Anderson, C. W. and Gesteland, R. F. (1975a). Enhanced differential synthesis of proteins in a mammalian cell-free system by addition of polyamines. J. Biol. Chem. 250, 5688-5695. Atkins, J. F., Lewis, J. B., Anderson, C. W., Baum, P. R. and Gesteland, R. F. (1975b). Mapping of adenovirus-2 genes bytranslation of RNA selected by hybridization. In INSERM Symposium 47, A. L. Haenni and G. Beaud, eds. (Paris: Inserm), pp. 293-298. Berget, S. M., Moore, Sci. USA, in press.
C. and Sharp,
P. A. (1977).
Proc.
Nat. Acad.
Chow, L. T., Roberts, J. M., Lewis, J. B. and Broker, T. R. (1977a). A map of cytoplasmic RNA transcripts from lytic adenovirus type 2, determined by electron microscopy of RNA:DNA hybrids. Cell 77, 819-836. Chow, L. T., Gelinas, R. E., Broker, T. R. and Roberts, R. J. (1977b). An amazing sequence arrangement at the 5’ ends of adenovirus 2 messenger RNA. Cell 72, 1-8. Dunn, A. R. and Hassell, J. A. (1977). A novel method to map transcripts: evidence for homology between an adenovirus mRNA and discrete multiple regions of the viral genome. Cell 72, 23-36. Gelinas, R. E. and Roberts, Ft. J. (1977). One predominant 5’undecanucleotide in adenovirus 2 late messenger RNAs. Cell 77, 533-544. Klessig,
D. F. (1977).
Two
adenovirus
mRNAs
have
a common
5’
terminal leader sequence the main coding regions
encoded for these
at least IO kb upstream from messengers. Cell 72, 9-21.
Lewis, J. B., Atkins, J. F., Anderson, C. W., Baum, P. R. and Gesteland, R. F. (1975). Mapping of late adenovirus genes by cellfree translation of RNA selected by hybridization to specific DNA fragments. Proc. Nat. Acad. Sci. USA 72, 1344-1348. Lewis, J. B., Atkins, J. F., Baum. P. R.. Solem, R., Gesteland. R. F. and Anderson, C. W. (1976). Location and identification of the genes for adenovirus type 2 early polypeptides. Cell 7, 141-151. Mulder, C., Arrand, J. R., Delius, H., Keller, W., Pettersson, U., Roberts, R. J. and Sharp, P. A. (1974). Cleavage maps of DNA from adenovirus types 2 and 5 by restriction endonucleases Eco RI and Hpal. Cold Spring Harbor Symp. Quant. Biol. 39,397-400. Pettersson, U. and Sambrook, J. (1973). The amount of viral DNA in the genome of cells transformed by adenovirus type 2. J. Mol. Biol. 73, 125-130. Pettersson, U., Mulder, C., Delius, H. and Sharp, P. A. (1973). Cleavage of adenovirus type 2 DNA into six unique fragments by endonuclease R.Rl. Proc. Nat. Acad. Sci. USA 70, 200-204. Schreier, M. and Staehelin, T. (1973). Initiation of mammalian protein synthesis: the importance of ribosome and initiation factor quality for in vitro synthesis. J. Mol. Biol. 73, 329-349. Thomas, M., White, R. L. and Davis, RNA to double strand DNA: formation Sci. USA 73, 2294-2298.
R. W. (1976). of R-loops.
Wilson, G. A. and Young, F. (1975). specific nuclease (Baml) from Baci//us Mol. Biol. 97, 723-725.
Hybridization of Proc. Nat. Acad.
isolation of a sequenceamyloliquefaciens H. J.
