Proc. Natl. Acad. Sci. USA Vol. 73, No. 11, pp. 4169-4173, November 1976
Genetics
Localization of gene functions in polyoma virus DNA (marker rescue/hr-t function/ts-a function/transformation)
JEAN FEUNTEUN, LAUREN SOMPAYRAC, MICHELE FLUCK, AND THOMAS BENJAMIN Department of Pathology, Harvard Medical School, Boston, Massachusetts 02115
Communicated by Baruj Benacerraf, September 7,1976
ABSTRACT Polyoma virus mutants of four functionally distinct groups have been mapped by the marker rescue technique using restriction enzyme fragments of wild-type viral DNA. Nontransforming host-range mutants map in the proximal part of the early region of the viral genome. The same DNA fragment that restores a normal host range also restores normal transforming ability to these mutants. ts-25D, a temperaturesensitive (ts-a class mutant, maps in the distal part of the early region. ts-3 and ts-1260 map in the proximal and distal parts of the late region, respectively.
Host-range mutants of polyoma virus have been isolated using polyoma-transformed mouse 3T3 cells as a permissive host and normal 3T3 cells as a nonpermissive host (1). These mutants form a single complementation group, and all have lost the ability to transform cells in culture*. This group has been designated hr-t. NG-18, a prototype hr-t mutant, is unable to induce tumors in newborn hamsters (ref. 2; and R. Siegler and T. Benjamin, manuscript in preparation). In normal 3T3 cells, such mutants produce T-antigen, substantial amounts of viral DNA and capsid protein(s), and a yield of progeny virus which is 2-5% that of a wild-type infection (3, *; unpublished results). A regulatory mode of action has been proposed for the hr-t viral gene. According to the model, expression of two classes of cellular genes is caused by the action of the hr-t viral gene: those necessary for productive viral infection and those involved in various aspects of the transformed phenotype (3, 4, *). Temperature-sensitive (ts) mutants fall into three complementation groups-an early group (ts-a) that is defective in the initiation of transformation, and two late groups that are unaffected in transformation (5). One. mutant, ts-3, fails to complement all other mutants, and is probably defective in a virion protein (6). Physiological properties of hr-t mutants set them clearly apart from all known temperature-sensitive mutant groups, but do not permit one to decide whether they are altered in an "early" or a "late" viral function. In productive infection they complement well with ts mutants of both early and late classes; they also complement with ts-a mutants for transformationtt. We have employed the physical mapping procedure, devised and applied to bacteriophage XX-174 by Hutchison and Egdell (7), in order to map hr-t and ts mutants of polyoma virus. This procedure has recently been applied to ts mutants of simian virus 40 (SV40) (8-10) and polyoma virus (11). Our results show that hr-t mutants map in the proximal part of the region of viral DNA that is transcribed early during productive infection and in transformed cells. Alterations within the same small segment Abbreviations: Hpa, Haemophilus parainfluenzae; Hind, Haemophilus influenzae d; Hha, Haemophilus haemolyticus; Hae, Haemophilus aegyptius; Eco, Escherichia coli; PFU, plaque-forming units; ts, temperature sensitive; SV40, simian virus 40; WT, wild type. * R. Staneloni, M. Fluck, and T. Benjamin, manuscript submitted. t M. Fluck, R. Staneloni, and T. Benjamin, manuscript submitted.
f W. Eckhart, manuscript submitted.
4169
of the viral DNA give rise to the defects in both growth and transformation in hr-t mutants. This has previously been a "silent" part of the viral DNA, not represented by any of the known ts mutants. MATERIALS AND METHODS Virus Strains. The wild-type virus was a small-plaque strain from Pasadena. hr-t mutants NG-18, NG-59, and 3A-1 have been described (1,*). Temperature-sensitive mutants ts-1260 and ts-3 were obtained from W. Eckhart; ts-1260 is a late mutant (12); ts-3 has been described (6). ts-25D was obtained from C. Basilico; it is the same as HA-25, which is an early a group mutant (13). Preparation of Viral DNAs. Baby mouse kidney cultures were infected at a multiplicity of infection of 2-20 plaqueforming units (PFU)/cell with stocks of recently plaque-purified virus. Infected cultures were incubated at 370 for wild-type and hr-t strains, or at 330 for ts strains. After 42 hr at 370 or 90 hr at 330, viral DNA was extracted by the Hirt procedure (14) and purified by equilibrium centrifugation in cesium chloride/ethidium bromide density gradients. Only supercoiled form I DNA was used. Restriction Enzymes and Preparation of DNA Fragments. HindIll was kindly supplied by D. Livingston and K. Backman, and HhaI by R. Kamen. HaeIII was purchased from New England Biolabs, Beverly, Mass. HpaII was prepared by the second procedure of Sharp et al. (15) from cells given to us by C. Thomas. EcoRI was prepared as described (16). HindIll and HhaI fragments were separated by electrophoresis in 1.4% agarose gels. HpaII and HaeIII fragments were separated on 4% and 7.5% polyacrylamide gels, respectively. The gel and running buffer was 0.04 M Tris-HCI at pH 7.8, 2 mM EDTA, and 0.02 M sodium acetate. Preparation of Singly Nicked Molecules. Barzilai has shown that, in the presence of ethidium bromide, pancreatic DNase introduces only one single-stranded break in SV40 DNA (17). We have modified his procedure by using restriction enzyme EcoRI in place of pancreatic DNase. About 5 jig of form I mutant viral DNA was incubated for 1 hr at 37° with 5 units of EcoRI (1 unit converts 1 ,ug of polyoma form I DNA to the linear form III in 1 hr at 37°) in 10 mM Tris, pH 7.5, 10 mM MgCI2, and 75 Asg/ml of ethidium bromide. The DNA was then phenol extracted and dialyzed against 15 mM NaCl, 1.5 mM sodium citrate, pH 7.0 (SSC/10), 1 mM EDTA. This procedure yields essentially 100% form II molecules with one single-strand break (L. Sompayrac and J. Feunteun, unpublished results). Heteroduplex Formation. Each 50 ,l reaction mixture contained 0.05-0.1 Asg of form II mutant viral DNA, and a 5to 10-fold molar excess of a single wild-type viral DNA fragment. In the case of HpaII and HaeIII fragments, the reaction mixture also contained a 2- to 3-fold molar excess of an unfractionated digest of the same mutant DNA. The mixture was denatured by adding 5 Al of 1 M NaOH and incubating for 10
4170
Genetics: Feunteun et al.
Proc. Natl. Acad. Sci. USA 73 (1976) Table 1. Rescue of NG-18 by wild-type HindIII fragments DNA inoculum
Virus yield
NG-18 form II alone WT 56% fragment alone WT 44% fragment alone NG-18 form II + WT 56% fragment NG-18 form II + WT 44% fragment
-Hind
II
0 0 0 140 3
Yields are given as PFU/0.1 ml of wild-type (WT) virus in DNAinfected cultures. See Fig. 1 for location of fragments.
6
Hha
I Hha
FIG. 1. The cleavage map of polyoma DNA by HpaII is shown in the annulus with fragments 1 through 8, and the orientations of early and late RNA are indicated by arrows [from Kamen et al. (21)]. Cleavage sites by HhaI and HindIII are shown by arrows. HaeIII cleavage sites are shown in the upper left quadrant; A and B refer to the HaeIII subfragments of HpaII fragment 4. Positions of HaeIII fragments 1 and 2 have been determined by B. Griffin (personal communication).
min at room temperature, and neutralized by addition of 10
,gl of 1:1 mixture of 1 M HCO and 1 M Tris, pH 7.5. Reannealing
performed for 10 min at 68°. Ligation of Fragments. DNA ligation was carried out using a modification of the method of Hamer and Thomas (18). About 0.25 ,ug of each of the two fragments to be ligated was suspended in 20,gl of 0.05 M Tris, pH 7.8, and 0.01 M MgCl2 and placed at 40. One microliter of 1.4 mM ATP and 0.025 units of phage T4 ligase (Miles) were then added. After 24 hr at 40, another 1 ,l of ATP and 0.025 units of ligase were added, and the incubation was continued for an additional 24 hr. DNA Infection. NIH-3T3 cells were seeded at 1.5 X 105 cells in 35 mm plastic petri dishes 24 hr prior to infection. The monolayers were washed once with phosphate-buffered saline and infected by adding the heteroduplex mixture in 100 1.l of Dulbecco's modified Eagle's medium containing 750 jtg/ml of DEAE-dextran and 50 mM Tris, pH 7.5. After 30 min at room temperature, the inoculum was removed and the monolayers were washed and fed with 2.5 ml of Dulbecco's modified Eagle's medium containing 5% calf serum. After 48 hr at 37° for hr-t mutants or at 390 for ts mutants, cells and fluids were harvested, sonicated, and assayed on NIH-3T3 cells at 370 or 390 to determine yields of wild-type virus. Transformation Assay. Individual plaques of rescued virus were picked and grown into stocks. Transformation was assayed by the soft agar procedure (19), using rat (NRK) or hamster (BHK) kidney cells. was
RESULTS In marker rescue experiments partial heteroduplexes are made between complete strands of a mutant viral DNA and various restriction enzyme fragments of wild-type viral DNA. These
heteroduplexes are then used to infect cells under conditions that are nonpermissive for the mutant. Wild-type virus is produced only in cases where the wild-type DNA fragment contains the sequence altered in the mutant (7). Fig. 1 shows the cleavage patterns of polyoma DNA by the various restriction enzymes used in this study. The results of marker rescue experiments are presented in all the tables as the amount of wild-type virus (PFU/0. 1 ml) in the culture after a single cycle of growth following DNA infection. Controls in all experiments included incubation of cells with wild-type DNA
fragments alone and with mutant form II DNAs alone; these were always negative. Rescue of hr-t Mutant NG-18 with HindIII Fragments. HindIII cleaves polyoma DNA in two places, resulting in pieces of 56% and 44% length (see Fig. 1). The results shown in Table 1 demonstrate rescue of NG-18 by the 56% piece, containing the origin of replication and sequences homologous to the 5' ends of the stable early and late messages. The small amount of rescue by the 44% piece is probably due to cross-contamination of the fragments. These results place NG-18 in the half of the genome opposite to where ts mutants of both early and late classes have been mapped; all ts mutants tested were rescued by the HindIII 44% fragment (11), which includes the distal (3') portions of the early and late regions. Rescue of hr-t and ts Mutants with HpaII Fragments. Rescue with HpaII fragments has allowed hr-t mutants to be localized within the early region. Table 2 shows results with three hr-t mutants. Fragment 4, from the proximal part of the early region, is the most efficient in restoring wild-type host range to these mutants. ts-25D of the ts-a group is rescued most efficiently by fragment 2. Miller and Fried (11) mapped a total of seven ts-a class mutants, two mapping in fragment 2 above the HhaI site, and five between the two HhaI sites (see Fig. 1). We have further mapped ts-25D (different from ts-25 reported by Miller and Fried, ref. 11), and find it to reside in the upper (5') portion of HpaII fragment 2, between the HindIII and HhaI sites (results not shown). ts-3 is rescued most efficiently by fragment 3 from the proximal part of the late region. Similar results have been obtained by W. Eckhartt. ts-1260 is rescued only by fragment 1, in agreement with earlier results (11). The information in fragment 1, besides restoring wild-type growth properties to this late mutant, also specifies its hemagglutinating properties. ts-1260, a large-plaque strain, hemagglutinates well at 40 but not at 370, while the Table 2. Rescue of hr-t and ts mutants by wild-type HpaII fragments
HpaII fragments Mutant DNA
1
2
3
4
5
6
7
8
1 0 5 7 1 25
0 3 0 160 0 0
2 10 0 0 45 0
35 160 50 0 0 0
0 0 0 0 0 0
0 0 0 0 3 0
0 0 0 0 0 0
0 0 0 0 0 0
hr-t class
NG-18 NG-59 3A-1 ts-25D ts-3 ts-1260
Results are given as PFU/0.1 ml of wild-type virus in DNA-infected cultures. See Fig. 1 for location of fragments.
Genetics: Feunteun et al.
Proc. Natl. Acad. Sci. USA 73 (1976)
Table 3. Hemagglutinating properties of ts-1 260 R-1
ts-1260 Wild type ts-1260 R-1
40
370
3200 1600 200
1600 1 1
ts-1260 R-1 is a wild-type virus produced by rescue of ts-1260 with HpaII fragment 1. Titers shown are reciprocal of highest dilution giving agglutination of sheep erythrocytes at 2 x 107 cells per ml.
,small-plaque wild-type virus used for rescue agglutinates well at 37°. As shown in Table 3, ts-1260 R-1 (rescued by fragment 1) has acquired the wild-type property of hemagglutinating at 370. These results add to the likelihood that fragment 1 encodes part of the capsid structure. HpaII Fragment Pattern of Rescued NG-18. Fig. 2 compares the HpaII patterns of DNAs from wild-type virus, mutant NG-18, and NG-18 rescued by wild-type fragment 4. Fragments 3, 4, and 5 of NG-18 DNA are each smaller than the corresponding wild-type fragments. Rescued NG-18 virus has recovered a normal size fragment 4, while retaining the small deletions in fragments 3 and 5. The latter deletions appear not to contribute to the phenotype of hr-t mutants, and indeed have been seen frequently in various wild-type stocks. Rescue of hr-t Mutants with HaeIII Subfragments of HpaII Fragment 4. Cleavage of HpaII fragment 4 by HaeIII enzyme produces three subfragments: two of nearly equal size representing almost the entire fragment 4 (subfragments A and B) and a third small fragment of about 50 base pairs. In order to determine the positions of the two larger subfragments, use was made of NG-18 DNA with its deletion of about 150 base pairs in HpaII fragment 4. First, NG-18 and wild-type DNAs were digested with HaeIII enzyme and run on a 7.5% acrylamide gel. The HaeIII fragments which overlap HpaII fragment 4 have been ordered by Beverly Griffin (personal communication) as shown in Fig. 1. Fig. 3a shows that HaeIII fragment 2 is missing in NG-18 DNA; it has been replaced by a new smaller fragment of the expected size (not seen in figure). Therefore, the deletion must be located toward the 5' end of HpaII fragment 4, in the region overlapped by HaeIII fragment 2. Second, HpaII fragment 4 was prepared from NG-18 and wild-type DNAs and then digested with HaeIII enzyme to yield the three subfragments. Fig. 3b shows that the larger (A) subfragment is absent in NG-18 DNA, again being replaced
2 3
_3
FIG. 3. (a) Left lane, HpaII fragment 4 of wild-type DNA; middle lane, HaeIII fragments 1 through 3 (2 missing) of NG-18 DNA; right lane, HaeIII fragments 1 through 3 of wild-type DNA. Arrows and numbers refer to HaeIII fragments 1 through 3, and HpaII fragment 4. (b) From left to right: HpaII fragment 4 of NG-18 DNA, HaeIII subfragments (B, A') of HpaII fragment 4 of NG-18 DNA, HaeIII subfragments (A, B) of HpaII fragment 4 of wild-type DNA, and HpaII fragment 4 of wild-type DNA. Arrows with numbers and letters refer to HaeIII subfragments A, B, A', and HpaII fragment 4.
by a smaller piece (A'). The small (50 base pair) subfragment cannot be seen in this gel; however, other experiments show this fragment to be present in NG-18 DNA. The order of the subfragments is therefore as shown in Fig. 1, with the NG-18 deletion residing in the 5' part of HpaII fragment 4 somewhere within the A subfragment. The two larger subfragments A and B from wild-type DNA were prepared and tested for rescue of hr-t mutants. As shown in Table 4, rescue occurs with the A subfragment for mutants NG-59 and 3A-1. Neither subfragment rescues NG-18, although HpaII fragment 4 again rescues well. The important difference between NG-18 and the other two mutants may lie in the fact that the latter do not have detectable deletions in HpaII fragment 4*. Since the deletion in NG-18 has removed about 40% of the A subfragment, heteroduplexes may not be sufficiently stable to register in the marker rescue experiment. Thus, the mutation in NG-18 could reside in the nondeleted part of A, or be constituted by the deletion itself. It is also possible that the NG-18 mutation resides in the 50 base pair subfragment which was not tested. Construction of Wild-Type and Double Mutant Viruses by DNA Fragment Ligation. hr-t and ts-a mutants constitute separate early complementation groups, each representing a distinct function required for both virus growth and transformationf*. These two mutant classes map on opposite fragments produced by the HindIII enzyme (see Tables 1 and 2, and ref. 11). Ligation of appropriate HindIII halves of these mutant viral DNA molecules should permit the construction of wildtype and double mutant viruses. Table 5 and Fig. 4 show results of such experiments. Table 4. Rescue of hr-t mutants by HaeIII subfragments of HpaII fragment 4
5 6
FIG. 2. HpaII fragment patterns: left lane, wild-type DNA; middle lane, DNA of NG-18 rescued by HpaII fragment 4 of wild-type DNA; right lane, NG-18 DNA. The numbering of the fragments begins at the top, with 1, the largest, slowest fragment.
HpaII
HaelHI subfragments
Mutant DNA
fragment 4
A
B
NG-18 NG-59 3A-1
50 70 50
0 70 10
0 3 1
Results are given as PFU/0.1 ml of wild-type virus in DNA-infected cultures. See Fig. 1 for location of fragments.
4172
Genetics: Feunteun et al.
Proc. Natl. Acad. Sci. USA 73 (1976)
Table 5. Ligation of HindIII fragments from hr-t and ts-a mutant viral DNAs
320
(a)
Virus source Ligated recombinants ts-25D (56%) + NG-18 (44%) NG-18 (56%) + ts-25D (44%) Virus controls Wild type NG-18 ts-25D
Plaques measured on .~~~~~~~~~~~~~~~
(b)
390
(c)
(d)
Permissive cell
Nonpermissive cell
Permissive cell
Nonpermissive cell
280 120
140
280
280
0
0
0
70 110 200
40
40 70
30
0
100
0
0
0
HindM fragments of DNA from NG-18 and ts-25D (see Figs. 1 and 4) were ligated together and used to infect UC1-B cells at 320 as described in Materials and Methods. Plaques resulting from this infection were picked and propagated on UC1-B at 320, and assayed as indicated: (a) UC1-B at 320; (b) NIH-3T3 at 320; (c) UC1-B at 390; and (d) NIH-3T3 at 39°. Results shown are of lysates from individual plaques, and are given as PFU/0.1 ml.
Ligation of the 56% piece of ts-25D with the 44% piece of NG-18 leads to recovery of virus with wild-type growth properties, growing well under all conditions tested, including nonpermissive cells at high temperature. One plaque resulting from this ligation was propagated further and tested by HpaII digestion to confirm its origin. As shown in Fig. 4, ts-25D has a small deletion in fragment 3, while NG-18 has deletions in fragments 3, 4, and 5. All of these deletions reside within the HindIII 56% piece (see Fig. 1). The ligated wild-type recombinant shown in lane c has the expected HpaII pattern, with only the fragment 3 deletion indicating its derivation from the 56% piece of ts-25D. The construction of wild-type virus from appropriate parts of the two mutant viral DNAs confirms the map locations of the mutants deduced on the basis of marker rescue results. Ligation of the 56% piece of NG-18 with the 44% piece of ts-25D leads to recovery of virus with the properties expected of a double mutant. This virus exhibits the growth restrictions of both parental mutant viruses, and is capable of growing only on a permissive host at the permissive temperature (see Table 5). This fragment ligation procedure should prove generally useful in structural and functional studies of the polyoma genome, and will make possible physiological studies of virus
FIG. 4. HpaII fragment patterns: (a) wild-type DNA, (b) ts-25D DNA, (c) ligated recombinant of HindIII 56% fragment of ts-25D and HindIII 44% fragment of NG-18, (d) NG-18 DNA.
strains with multiple mutations. A similar procedure has already been applied to the construction of hybrid genomes of wild-type polyoma strains (20). Table 6. Transformation by rescued viruses PFU in
inoculum Virus
X 10-5
Transformants per culture
I. hr-t mutants rescued by HpaII fragment 4 Wild-type control
NG-18 NG-18 NG-18 NG-18 NG-18 NG-18 NG-59 NG-59
R-41
R-42 R-43
R-44
R-4S R-46
R-41
R-42 II. hr-t mutants rescued by HaeIII subfragment A of HpaII fragment 4 3A-1 R-A NG-59 R-A III. ts mutants rescued by HpaII fragments ts-25D R-2 (fragment 2) Wild-type control ts-3 R-3 (fragment 3) Wild-type control IV. Ligated recombinant NG-18 (44%) + ts-25D (56%) NG-18 control ts-25D control Wild-type control
20 2 3 30 3 0.4 0.5 3 0.7 1
>500 200 320 >500 200 120 120 >500 50 30
7 5
120
10 20 10 20
120 300 20 30
100 500 60 300
35 0 0 180
200
Cultures containing 5 X 104 cells in 35 mm dishes were infected as indicated, and transferred to soft agar 8-16 hr later. Transformants were counted macroscopically after 10-15 days. Wild-type controls are given to normalize transformation efficiency in different experiments.
Genetics: Feunteun et al. Transformation by Rescued Viruses. Each of 19 hr-t mutants isolated on the basis of a reduced host range has proven to be defective in transformation (1, *). The inability to trans-
form cells, arising as an unselected property in these mutants, strongly suggests that single mutations are responsible for the defects in both virus growth and transformation. It is therefore expected that hr-t mutants which have been rescued with respect to growth properties will also have regained the ability to transform. This expectation is confirmed by the data shown in Table 6. Six wild-type plaques arising from rescue of NG-18 (see Fig. 2) and two from rescue of NG-59, all by HpaII fragment 4, were grown into stocks and tested for transforming ability; all eight proved capable of transforming cells with an efficiency similar to that of a standard wild-type stock (Table 6, part I). Therefore, mutations affecting transformation in hr-t mutants also reside in fragment 4. Similar and more persuasive evidence that host range and transformation defects are due to one and the same mutation comes from the fact that mutants 3A-1 and NG-59 rescued by the HaeIII subfragment A of HpaII fragment 4 also transform like wild-type virus (Table 6, part II). Thus, if separate alterations are involved, they would have to have arisen together within this small segment (about 375 base pairs) of the viral genome.
ts-25D of the ts-a group is also defective in transformation (13). As in the case of the hr-t mutants, the fragment that restores wild-type growth properties (in this case HpaII fragment 2) also restores wild-type transforming ability. ts-3, with an apparent defect in a virion protein that may affect its transforming efficiency in certain cells (W. Eckhart, personal communication), is also restored to a normal transforming capability by the same fragment (HpaII fragment 3) that restores normal growth (Table 6, part III). Transformation is also observed by the ligated recombinant with wild-type host range formed from the nonmutated parts of ts-25D and NG-18 (Table 6, part IV; see also Table 5 and Fig.
4). DISCUSSION Four physiologically distinct mutant groups of polyoma virus have been mapped on the viral DNA by the marker rescue technique. ts mutants of the early (ts-a) and late complementation groups map in the distal (3') parts of the early and late regions, respectively. These results confirm earlier findings and demonstrate a close homology in the organization of the polyoma and SV40 viral genomes. The ts-a group of polyoma and the ts-A group of SV40 share functional similarities and occupy similar positions on the viral DNA; the same is true for the genes coding for the major capsid proteins (8-11). The mutation in polyoma mutant ts-3 is located in the proximal (5') part of the late region. By virtue of its location and certain of its biological properties, this mutant appears to be homologous to SV40 mutants of the D group (6, 9, *). hr-t mutants have been mapped within the proximal portion of the early region of the viral DNA, a region of the genome
known to be transcribed early in productive infection and in transformed cells (21). A wild-type DNA fragment representing about 7% of the viral genome is sufficient to restore both wild-type host range and wild-type transforming ability, indicating the likelihood that one mutation causes both the re-
Proc. Natl. Acad. Sci. USA 73 (1976)
4173
stricted host range and the inability to transform in these mutants. None of the temperature-sensitive mutants of SV40 resemble hr-t mutants phenotypically, and none of them map in the proximal part of the early region in SV40 DNA. Small deletions in this region of the SV40 genome have been described, but appear to have no major phenotypic manifestations (22). hr-t and ts-a mutants represent different physiological functionstt, and they map distinctly apart from one another within the early region. ts-a mutants are affected in T antigen (23, 24), while hr-t mutants are not (2). It is therefore possible that a second early protein exists, the product of the hr-t gene, which is distinct from T antigen. Alternatively, a single bifunctional protein coded by the early region could carry out both ts-a and hr-t functions. It is also possible that the hr-t function resides in a control sequence of the viral DNA if, for example, cellular functions come under cis-acting viral control after integrationt. The authors wish to acknowledge the expert technical assistance of Ingrid Lane, Anne Caesar, and Susan Tafler. Portions of the work in this laboratory have been supported by Grant no. DRG-13-F from the Damon Runyon-Walter Winchell Cancer Fund, and Contract no. NO1-43299 from the National Cancer Institute. J.F. is recipient of a long-term European Molecular Biology Organization Fellowship, L.S. is recipient of a Fellowship from the Massachusetts Division of the American Cancer Society, and M.F. is recipient of a Fellowship from the Swiss League Against Cancer and National Institutes of Health Training Grant no. T 32 CA-9031. T.B. is a Scholar of the Leukemia Society of America, Inc. 1. Benjamin, T. L. (1970) Proc. Natl. Acad. Sci. USA 67, 394399. 2. Siegler, R. & Benjamin, T. (1975) Proc. Am. Assoc. Cancer Res. 16,99. 3. Goldman, E. & Benjamin, T. L. (1975) Virology 66,372-384. 4. Schaffhausen, B. S. & Benjamin, T. L. (1976) Proc. Natl. Acad. Sci. USA 73, 1092-1096. 5. Eckhart, W. (1974) Annu. Rev. Genet. 8, 301-317. 6. Eckhart, W. & Dulbecco, R. (1974) Virology 60,359-369. 7. Hutchison, C. & Edgell, M. (1971) J. Virol. 8, 181-189. 8. Lai, C. & Nathans, D. (1974) Virology 60,466-475. 9. Lai, C. & Nathans, D. (1975) Virology 66, 70-81. 10. Mantei, N., Boyer, H. W. & Goodman, H. M. (1975) J. Virol. 16, 754-757. 11. Miller, L. K. & Fried, M. (1976) J. Virol. 18, 824-832. 12. Eckhart, W. (1969) Virology 38, 120-125. 13. Di Mayorca, G., Callender, J., Marin, G. & Giordano, R. (1969) Virology 38, 126-133. 14. Hirt, B. (1967) J. Mol. Biol. 26,365-369. 15. Sharp, P. A., Sugden, B. & Sambrook, J. (1973) Biochemistry 12, 3055-3063. 16. Yoshimori, R. N. (1971) PhD Dissertation, University of California at San Francisco. 17. Barzilai, R. (1973) J. Mol. Biol. 74, 739-742. 18. Hamer, D. & Thomas, C. (1976) Proc. Natl. Acad. Sci. USA 73, 1537-1541. 19. MacPherson, I. & Montagnier, L. (1964) Virology 23, 291294. 20. Miller, L. K. & Fried, M. (1976) Nature 259,598-601. 21. Kamen, R., Lindstrom, D. M., Shure, H. & Old, R. (1974) Cold Spring Harbor Symp. Quant. Biol. 39, 187-198. 22. Shenk, T. E., Carbon, J. & Berg, P. (1976) J. Virol. 18, 664671. 23. Oxman, M. N., Takemoto, K. K. & Eckhart, W. (1972) Virology 49, 675-682. 24. Paulin, D. & Cuzin, F. (1975) J. Virol. 15,393-397.
Genetics
Localization of gene functions in polyoma virus DNA (marker rescue/hr-t function/ts-a function/transformation)
JEAN FEUNTEUN, LAUREN SOMPAYRAC, MICHELE FLUCK, AND THOMAS BENJAMIN Department of Pathology, Harvard Medical School, Boston, Massachusetts 02115
Communicated by Baruj Benacerraf, September 7,1976
ABSTRACT Polyoma virus mutants of four functionally distinct groups have been mapped by the marker rescue technique using restriction enzyme fragments of wild-type viral DNA. Nontransforming host-range mutants map in the proximal part of the early region of the viral genome. The same DNA fragment that restores a normal host range also restores normal transforming ability to these mutants. ts-25D, a temperaturesensitive (ts-a class mutant, maps in the distal part of the early region. ts-3 and ts-1260 map in the proximal and distal parts of the late region, respectively.
Host-range mutants of polyoma virus have been isolated using polyoma-transformed mouse 3T3 cells as a permissive host and normal 3T3 cells as a nonpermissive host (1). These mutants form a single complementation group, and all have lost the ability to transform cells in culture*. This group has been designated hr-t. NG-18, a prototype hr-t mutant, is unable to induce tumors in newborn hamsters (ref. 2; and R. Siegler and T. Benjamin, manuscript in preparation). In normal 3T3 cells, such mutants produce T-antigen, substantial amounts of viral DNA and capsid protein(s), and a yield of progeny virus which is 2-5% that of a wild-type infection (3, *; unpublished results). A regulatory mode of action has been proposed for the hr-t viral gene. According to the model, expression of two classes of cellular genes is caused by the action of the hr-t viral gene: those necessary for productive viral infection and those involved in various aspects of the transformed phenotype (3, 4, *). Temperature-sensitive (ts) mutants fall into three complementation groups-an early group (ts-a) that is defective in the initiation of transformation, and two late groups that are unaffected in transformation (5). One. mutant, ts-3, fails to complement all other mutants, and is probably defective in a virion protein (6). Physiological properties of hr-t mutants set them clearly apart from all known temperature-sensitive mutant groups, but do not permit one to decide whether they are altered in an "early" or a "late" viral function. In productive infection they complement well with ts mutants of both early and late classes; they also complement with ts-a mutants for transformationtt. We have employed the physical mapping procedure, devised and applied to bacteriophage XX-174 by Hutchison and Egdell (7), in order to map hr-t and ts mutants of polyoma virus. This procedure has recently been applied to ts mutants of simian virus 40 (SV40) (8-10) and polyoma virus (11). Our results show that hr-t mutants map in the proximal part of the region of viral DNA that is transcribed early during productive infection and in transformed cells. Alterations within the same small segment Abbreviations: Hpa, Haemophilus parainfluenzae; Hind, Haemophilus influenzae d; Hha, Haemophilus haemolyticus; Hae, Haemophilus aegyptius; Eco, Escherichia coli; PFU, plaque-forming units; ts, temperature sensitive; SV40, simian virus 40; WT, wild type. * R. Staneloni, M. Fluck, and T. Benjamin, manuscript submitted. t M. Fluck, R. Staneloni, and T. Benjamin, manuscript submitted.
f W. Eckhart, manuscript submitted.
4169
of the viral DNA give rise to the defects in both growth and transformation in hr-t mutants. This has previously been a "silent" part of the viral DNA, not represented by any of the known ts mutants. MATERIALS AND METHODS Virus Strains. The wild-type virus was a small-plaque strain from Pasadena. hr-t mutants NG-18, NG-59, and 3A-1 have been described (1,*). Temperature-sensitive mutants ts-1260 and ts-3 were obtained from W. Eckhart; ts-1260 is a late mutant (12); ts-3 has been described (6). ts-25D was obtained from C. Basilico; it is the same as HA-25, which is an early a group mutant (13). Preparation of Viral DNAs. Baby mouse kidney cultures were infected at a multiplicity of infection of 2-20 plaqueforming units (PFU)/cell with stocks of recently plaque-purified virus. Infected cultures were incubated at 370 for wild-type and hr-t strains, or at 330 for ts strains. After 42 hr at 370 or 90 hr at 330, viral DNA was extracted by the Hirt procedure (14) and purified by equilibrium centrifugation in cesium chloride/ethidium bromide density gradients. Only supercoiled form I DNA was used. Restriction Enzymes and Preparation of DNA Fragments. HindIll was kindly supplied by D. Livingston and K. Backman, and HhaI by R. Kamen. HaeIII was purchased from New England Biolabs, Beverly, Mass. HpaII was prepared by the second procedure of Sharp et al. (15) from cells given to us by C. Thomas. EcoRI was prepared as described (16). HindIll and HhaI fragments were separated by electrophoresis in 1.4% agarose gels. HpaII and HaeIII fragments were separated on 4% and 7.5% polyacrylamide gels, respectively. The gel and running buffer was 0.04 M Tris-HCI at pH 7.8, 2 mM EDTA, and 0.02 M sodium acetate. Preparation of Singly Nicked Molecules. Barzilai has shown that, in the presence of ethidium bromide, pancreatic DNase introduces only one single-stranded break in SV40 DNA (17). We have modified his procedure by using restriction enzyme EcoRI in place of pancreatic DNase. About 5 jig of form I mutant viral DNA was incubated for 1 hr at 37° with 5 units of EcoRI (1 unit converts 1 ,ug of polyoma form I DNA to the linear form III in 1 hr at 37°) in 10 mM Tris, pH 7.5, 10 mM MgCI2, and 75 Asg/ml of ethidium bromide. The DNA was then phenol extracted and dialyzed against 15 mM NaCl, 1.5 mM sodium citrate, pH 7.0 (SSC/10), 1 mM EDTA. This procedure yields essentially 100% form II molecules with one single-strand break (L. Sompayrac and J. Feunteun, unpublished results). Heteroduplex Formation. Each 50 ,l reaction mixture contained 0.05-0.1 Asg of form II mutant viral DNA, and a 5to 10-fold molar excess of a single wild-type viral DNA fragment. In the case of HpaII and HaeIII fragments, the reaction mixture also contained a 2- to 3-fold molar excess of an unfractionated digest of the same mutant DNA. The mixture was denatured by adding 5 Al of 1 M NaOH and incubating for 10
4170
Genetics: Feunteun et al.
Proc. Natl. Acad. Sci. USA 73 (1976) Table 1. Rescue of NG-18 by wild-type HindIII fragments DNA inoculum
Virus yield
NG-18 form II alone WT 56% fragment alone WT 44% fragment alone NG-18 form II + WT 56% fragment NG-18 form II + WT 44% fragment
-Hind
II
0 0 0 140 3
Yields are given as PFU/0.1 ml of wild-type (WT) virus in DNAinfected cultures. See Fig. 1 for location of fragments.
6
Hha
I Hha
FIG. 1. The cleavage map of polyoma DNA by HpaII is shown in the annulus with fragments 1 through 8, and the orientations of early and late RNA are indicated by arrows [from Kamen et al. (21)]. Cleavage sites by HhaI and HindIII are shown by arrows. HaeIII cleavage sites are shown in the upper left quadrant; A and B refer to the HaeIII subfragments of HpaII fragment 4. Positions of HaeIII fragments 1 and 2 have been determined by B. Griffin (personal communication).
min at room temperature, and neutralized by addition of 10
,gl of 1:1 mixture of 1 M HCO and 1 M Tris, pH 7.5. Reannealing
performed for 10 min at 68°. Ligation of Fragments. DNA ligation was carried out using a modification of the method of Hamer and Thomas (18). About 0.25 ,ug of each of the two fragments to be ligated was suspended in 20,gl of 0.05 M Tris, pH 7.8, and 0.01 M MgCl2 and placed at 40. One microliter of 1.4 mM ATP and 0.025 units of phage T4 ligase (Miles) were then added. After 24 hr at 40, another 1 ,l of ATP and 0.025 units of ligase were added, and the incubation was continued for an additional 24 hr. DNA Infection. NIH-3T3 cells were seeded at 1.5 X 105 cells in 35 mm plastic petri dishes 24 hr prior to infection. The monolayers were washed once with phosphate-buffered saline and infected by adding the heteroduplex mixture in 100 1.l of Dulbecco's modified Eagle's medium containing 750 jtg/ml of DEAE-dextran and 50 mM Tris, pH 7.5. After 30 min at room temperature, the inoculum was removed and the monolayers were washed and fed with 2.5 ml of Dulbecco's modified Eagle's medium containing 5% calf serum. After 48 hr at 37° for hr-t mutants or at 390 for ts mutants, cells and fluids were harvested, sonicated, and assayed on NIH-3T3 cells at 370 or 390 to determine yields of wild-type virus. Transformation Assay. Individual plaques of rescued virus were picked and grown into stocks. Transformation was assayed by the soft agar procedure (19), using rat (NRK) or hamster (BHK) kidney cells. was
RESULTS In marker rescue experiments partial heteroduplexes are made between complete strands of a mutant viral DNA and various restriction enzyme fragments of wild-type viral DNA. These
heteroduplexes are then used to infect cells under conditions that are nonpermissive for the mutant. Wild-type virus is produced only in cases where the wild-type DNA fragment contains the sequence altered in the mutant (7). Fig. 1 shows the cleavage patterns of polyoma DNA by the various restriction enzymes used in this study. The results of marker rescue experiments are presented in all the tables as the amount of wild-type virus (PFU/0. 1 ml) in the culture after a single cycle of growth following DNA infection. Controls in all experiments included incubation of cells with wild-type DNA
fragments alone and with mutant form II DNAs alone; these were always negative. Rescue of hr-t Mutant NG-18 with HindIII Fragments. HindIII cleaves polyoma DNA in two places, resulting in pieces of 56% and 44% length (see Fig. 1). The results shown in Table 1 demonstrate rescue of NG-18 by the 56% piece, containing the origin of replication and sequences homologous to the 5' ends of the stable early and late messages. The small amount of rescue by the 44% piece is probably due to cross-contamination of the fragments. These results place NG-18 in the half of the genome opposite to where ts mutants of both early and late classes have been mapped; all ts mutants tested were rescued by the HindIII 44% fragment (11), which includes the distal (3') portions of the early and late regions. Rescue of hr-t and ts Mutants with HpaII Fragments. Rescue with HpaII fragments has allowed hr-t mutants to be localized within the early region. Table 2 shows results with three hr-t mutants. Fragment 4, from the proximal part of the early region, is the most efficient in restoring wild-type host range to these mutants. ts-25D of the ts-a group is rescued most efficiently by fragment 2. Miller and Fried (11) mapped a total of seven ts-a class mutants, two mapping in fragment 2 above the HhaI site, and five between the two HhaI sites (see Fig. 1). We have further mapped ts-25D (different from ts-25 reported by Miller and Fried, ref. 11), and find it to reside in the upper (5') portion of HpaII fragment 2, between the HindIII and HhaI sites (results not shown). ts-3 is rescued most efficiently by fragment 3 from the proximal part of the late region. Similar results have been obtained by W. Eckhartt. ts-1260 is rescued only by fragment 1, in agreement with earlier results (11). The information in fragment 1, besides restoring wild-type growth properties to this late mutant, also specifies its hemagglutinating properties. ts-1260, a large-plaque strain, hemagglutinates well at 40 but not at 370, while the Table 2. Rescue of hr-t and ts mutants by wild-type HpaII fragments
HpaII fragments Mutant DNA
1
2
3
4
5
6
7
8
1 0 5 7 1 25
0 3 0 160 0 0
2 10 0 0 45 0
35 160 50 0 0 0
0 0 0 0 0 0
0 0 0 0 3 0
0 0 0 0 0 0
0 0 0 0 0 0
hr-t class
NG-18 NG-59 3A-1 ts-25D ts-3 ts-1260
Results are given as PFU/0.1 ml of wild-type virus in DNA-infected cultures. See Fig. 1 for location of fragments.
Genetics: Feunteun et al.
Proc. Natl. Acad. Sci. USA 73 (1976)
Table 3. Hemagglutinating properties of ts-1 260 R-1
ts-1260 Wild type ts-1260 R-1
40
370
3200 1600 200
1600 1 1
ts-1260 R-1 is a wild-type virus produced by rescue of ts-1260 with HpaII fragment 1. Titers shown are reciprocal of highest dilution giving agglutination of sheep erythrocytes at 2 x 107 cells per ml.
,small-plaque wild-type virus used for rescue agglutinates well at 37°. As shown in Table 3, ts-1260 R-1 (rescued by fragment 1) has acquired the wild-type property of hemagglutinating at 370. These results add to the likelihood that fragment 1 encodes part of the capsid structure. HpaII Fragment Pattern of Rescued NG-18. Fig. 2 compares the HpaII patterns of DNAs from wild-type virus, mutant NG-18, and NG-18 rescued by wild-type fragment 4. Fragments 3, 4, and 5 of NG-18 DNA are each smaller than the corresponding wild-type fragments. Rescued NG-18 virus has recovered a normal size fragment 4, while retaining the small deletions in fragments 3 and 5. The latter deletions appear not to contribute to the phenotype of hr-t mutants, and indeed have been seen frequently in various wild-type stocks. Rescue of hr-t Mutants with HaeIII Subfragments of HpaII Fragment 4. Cleavage of HpaII fragment 4 by HaeIII enzyme produces three subfragments: two of nearly equal size representing almost the entire fragment 4 (subfragments A and B) and a third small fragment of about 50 base pairs. In order to determine the positions of the two larger subfragments, use was made of NG-18 DNA with its deletion of about 150 base pairs in HpaII fragment 4. First, NG-18 and wild-type DNAs were digested with HaeIII enzyme and run on a 7.5% acrylamide gel. The HaeIII fragments which overlap HpaII fragment 4 have been ordered by Beverly Griffin (personal communication) as shown in Fig. 1. Fig. 3a shows that HaeIII fragment 2 is missing in NG-18 DNA; it has been replaced by a new smaller fragment of the expected size (not seen in figure). Therefore, the deletion must be located toward the 5' end of HpaII fragment 4, in the region overlapped by HaeIII fragment 2. Second, HpaII fragment 4 was prepared from NG-18 and wild-type DNAs and then digested with HaeIII enzyme to yield the three subfragments. Fig. 3b shows that the larger (A) subfragment is absent in NG-18 DNA, again being replaced
2 3
_3
FIG. 3. (a) Left lane, HpaII fragment 4 of wild-type DNA; middle lane, HaeIII fragments 1 through 3 (2 missing) of NG-18 DNA; right lane, HaeIII fragments 1 through 3 of wild-type DNA. Arrows and numbers refer to HaeIII fragments 1 through 3, and HpaII fragment 4. (b) From left to right: HpaII fragment 4 of NG-18 DNA, HaeIII subfragments (B, A') of HpaII fragment 4 of NG-18 DNA, HaeIII subfragments (A, B) of HpaII fragment 4 of wild-type DNA, and HpaII fragment 4 of wild-type DNA. Arrows with numbers and letters refer to HaeIII subfragments A, B, A', and HpaII fragment 4.
by a smaller piece (A'). The small (50 base pair) subfragment cannot be seen in this gel; however, other experiments show this fragment to be present in NG-18 DNA. The order of the subfragments is therefore as shown in Fig. 1, with the NG-18 deletion residing in the 5' part of HpaII fragment 4 somewhere within the A subfragment. The two larger subfragments A and B from wild-type DNA were prepared and tested for rescue of hr-t mutants. As shown in Table 4, rescue occurs with the A subfragment for mutants NG-59 and 3A-1. Neither subfragment rescues NG-18, although HpaII fragment 4 again rescues well. The important difference between NG-18 and the other two mutants may lie in the fact that the latter do not have detectable deletions in HpaII fragment 4*. Since the deletion in NG-18 has removed about 40% of the A subfragment, heteroduplexes may not be sufficiently stable to register in the marker rescue experiment. Thus, the mutation in NG-18 could reside in the nondeleted part of A, or be constituted by the deletion itself. It is also possible that the NG-18 mutation resides in the 50 base pair subfragment which was not tested. Construction of Wild-Type and Double Mutant Viruses by DNA Fragment Ligation. hr-t and ts-a mutants constitute separate early complementation groups, each representing a distinct function required for both virus growth and transformationf*. These two mutant classes map on opposite fragments produced by the HindIII enzyme (see Tables 1 and 2, and ref. 11). Ligation of appropriate HindIII halves of these mutant viral DNA molecules should permit the construction of wildtype and double mutant viruses. Table 5 and Fig. 4 show results of such experiments. Table 4. Rescue of hr-t mutants by HaeIII subfragments of HpaII fragment 4
5 6
FIG. 2. HpaII fragment patterns: left lane, wild-type DNA; middle lane, DNA of NG-18 rescued by HpaII fragment 4 of wild-type DNA; right lane, NG-18 DNA. The numbering of the fragments begins at the top, with 1, the largest, slowest fragment.
HpaII
HaelHI subfragments
Mutant DNA
fragment 4
A
B
NG-18 NG-59 3A-1
50 70 50
0 70 10
0 3 1
Results are given as PFU/0.1 ml of wild-type virus in DNA-infected cultures. See Fig. 1 for location of fragments.
4172
Genetics: Feunteun et al.
Proc. Natl. Acad. Sci. USA 73 (1976)
Table 5. Ligation of HindIII fragments from hr-t and ts-a mutant viral DNAs
320
(a)
Virus source Ligated recombinants ts-25D (56%) + NG-18 (44%) NG-18 (56%) + ts-25D (44%) Virus controls Wild type NG-18 ts-25D
Plaques measured on .~~~~~~~~~~~~~~~
(b)
390
(c)
(d)
Permissive cell
Nonpermissive cell
Permissive cell
Nonpermissive cell
280 120
140
280
280
0
0
0
70 110 200
40
40 70
30
0
100
0
0
0
HindM fragments of DNA from NG-18 and ts-25D (see Figs. 1 and 4) were ligated together and used to infect UC1-B cells at 320 as described in Materials and Methods. Plaques resulting from this infection were picked and propagated on UC1-B at 320, and assayed as indicated: (a) UC1-B at 320; (b) NIH-3T3 at 320; (c) UC1-B at 390; and (d) NIH-3T3 at 39°. Results shown are of lysates from individual plaques, and are given as PFU/0.1 ml.
Ligation of the 56% piece of ts-25D with the 44% piece of NG-18 leads to recovery of virus with wild-type growth properties, growing well under all conditions tested, including nonpermissive cells at high temperature. One plaque resulting from this ligation was propagated further and tested by HpaII digestion to confirm its origin. As shown in Fig. 4, ts-25D has a small deletion in fragment 3, while NG-18 has deletions in fragments 3, 4, and 5. All of these deletions reside within the HindIII 56% piece (see Fig. 1). The ligated wild-type recombinant shown in lane c has the expected HpaII pattern, with only the fragment 3 deletion indicating its derivation from the 56% piece of ts-25D. The construction of wild-type virus from appropriate parts of the two mutant viral DNAs confirms the map locations of the mutants deduced on the basis of marker rescue results. Ligation of the 56% piece of NG-18 with the 44% piece of ts-25D leads to recovery of virus with the properties expected of a double mutant. This virus exhibits the growth restrictions of both parental mutant viruses, and is capable of growing only on a permissive host at the permissive temperature (see Table 5). This fragment ligation procedure should prove generally useful in structural and functional studies of the polyoma genome, and will make possible physiological studies of virus
FIG. 4. HpaII fragment patterns: (a) wild-type DNA, (b) ts-25D DNA, (c) ligated recombinant of HindIII 56% fragment of ts-25D and HindIII 44% fragment of NG-18, (d) NG-18 DNA.
strains with multiple mutations. A similar procedure has already been applied to the construction of hybrid genomes of wild-type polyoma strains (20). Table 6. Transformation by rescued viruses PFU in
inoculum Virus
X 10-5
Transformants per culture
I. hr-t mutants rescued by HpaII fragment 4 Wild-type control
NG-18 NG-18 NG-18 NG-18 NG-18 NG-18 NG-59 NG-59
R-41
R-42 R-43
R-44
R-4S R-46
R-41
R-42 II. hr-t mutants rescued by HaeIII subfragment A of HpaII fragment 4 3A-1 R-A NG-59 R-A III. ts mutants rescued by HpaII fragments ts-25D R-2 (fragment 2) Wild-type control ts-3 R-3 (fragment 3) Wild-type control IV. Ligated recombinant NG-18 (44%) + ts-25D (56%) NG-18 control ts-25D control Wild-type control
20 2 3 30 3 0.4 0.5 3 0.7 1
>500 200 320 >500 200 120 120 >500 50 30
7 5
120
10 20 10 20
120 300 20 30
100 500 60 300
35 0 0 180
200
Cultures containing 5 X 104 cells in 35 mm dishes were infected as indicated, and transferred to soft agar 8-16 hr later. Transformants were counted macroscopically after 10-15 days. Wild-type controls are given to normalize transformation efficiency in different experiments.
Genetics: Feunteun et al. Transformation by Rescued Viruses. Each of 19 hr-t mutants isolated on the basis of a reduced host range has proven to be defective in transformation (1, *). The inability to trans-
form cells, arising as an unselected property in these mutants, strongly suggests that single mutations are responsible for the defects in both virus growth and transformation. It is therefore expected that hr-t mutants which have been rescued with respect to growth properties will also have regained the ability to transform. This expectation is confirmed by the data shown in Table 6. Six wild-type plaques arising from rescue of NG-18 (see Fig. 2) and two from rescue of NG-59, all by HpaII fragment 4, were grown into stocks and tested for transforming ability; all eight proved capable of transforming cells with an efficiency similar to that of a standard wild-type stock (Table 6, part I). Therefore, mutations affecting transformation in hr-t mutants also reside in fragment 4. Similar and more persuasive evidence that host range and transformation defects are due to one and the same mutation comes from the fact that mutants 3A-1 and NG-59 rescued by the HaeIII subfragment A of HpaII fragment 4 also transform like wild-type virus (Table 6, part II). Thus, if separate alterations are involved, they would have to have arisen together within this small segment (about 375 base pairs) of the viral genome.
ts-25D of the ts-a group is also defective in transformation (13). As in the case of the hr-t mutants, the fragment that restores wild-type growth properties (in this case HpaII fragment 2) also restores wild-type transforming ability. ts-3, with an apparent defect in a virion protein that may affect its transforming efficiency in certain cells (W. Eckhart, personal communication), is also restored to a normal transforming capability by the same fragment (HpaII fragment 3) that restores normal growth (Table 6, part III). Transformation is also observed by the ligated recombinant with wild-type host range formed from the nonmutated parts of ts-25D and NG-18 (Table 6, part IV; see also Table 5 and Fig.
4). DISCUSSION Four physiologically distinct mutant groups of polyoma virus have been mapped on the viral DNA by the marker rescue technique. ts mutants of the early (ts-a) and late complementation groups map in the distal (3') parts of the early and late regions, respectively. These results confirm earlier findings and demonstrate a close homology in the organization of the polyoma and SV40 viral genomes. The ts-a group of polyoma and the ts-A group of SV40 share functional similarities and occupy similar positions on the viral DNA; the same is true for the genes coding for the major capsid proteins (8-11). The mutation in polyoma mutant ts-3 is located in the proximal (5') part of the late region. By virtue of its location and certain of its biological properties, this mutant appears to be homologous to SV40 mutants of the D group (6, 9, *). hr-t mutants have been mapped within the proximal portion of the early region of the viral DNA, a region of the genome
known to be transcribed early in productive infection and in transformed cells (21). A wild-type DNA fragment representing about 7% of the viral genome is sufficient to restore both wild-type host range and wild-type transforming ability, indicating the likelihood that one mutation causes both the re-
Proc. Natl. Acad. Sci. USA 73 (1976)
4173
stricted host range and the inability to transform in these mutants. None of the temperature-sensitive mutants of SV40 resemble hr-t mutants phenotypically, and none of them map in the proximal part of the early region in SV40 DNA. Small deletions in this region of the SV40 genome have been described, but appear to have no major phenotypic manifestations (22). hr-t and ts-a mutants represent different physiological functionstt, and they map distinctly apart from one another within the early region. ts-a mutants are affected in T antigen (23, 24), while hr-t mutants are not (2). It is therefore possible that a second early protein exists, the product of the hr-t gene, which is distinct from T antigen. Alternatively, a single bifunctional protein coded by the early region could carry out both ts-a and hr-t functions. It is also possible that the hr-t function resides in a control sequence of the viral DNA if, for example, cellular functions come under cis-acting viral control after integrationt. The authors wish to acknowledge the expert technical assistance of Ingrid Lane, Anne Caesar, and Susan Tafler. Portions of the work in this laboratory have been supported by Grant no. DRG-13-F from the Damon Runyon-Walter Winchell Cancer Fund, and Contract no. NO1-43299 from the National Cancer Institute. J.F. is recipient of a long-term European Molecular Biology Organization Fellowship, L.S. is recipient of a Fellowship from the Massachusetts Division of the American Cancer Society, and M.F. is recipient of a Fellowship from the Swiss League Against Cancer and National Institutes of Health Training Grant no. T 32 CA-9031. T.B. is a Scholar of the Leukemia Society of America, Inc. 1. Benjamin, T. L. (1970) Proc. Natl. Acad. Sci. USA 67, 394399. 2. Siegler, R. & Benjamin, T. (1975) Proc. Am. Assoc. Cancer Res. 16,99. 3. Goldman, E. & Benjamin, T. L. (1975) Virology 66,372-384. 4. Schaffhausen, B. S. & Benjamin, T. L. (1976) Proc. Natl. Acad. Sci. USA 73, 1092-1096. 5. Eckhart, W. (1974) Annu. Rev. Genet. 8, 301-317. 6. Eckhart, W. & Dulbecco, R. (1974) Virology 60,359-369. 7. Hutchison, C. & Edgell, M. (1971) J. Virol. 8, 181-189. 8. Lai, C. & Nathans, D. (1974) Virology 60,466-475. 9. Lai, C. & Nathans, D. (1975) Virology 66, 70-81. 10. Mantei, N., Boyer, H. W. & Goodman, H. M. (1975) J. Virol. 16, 754-757. 11. Miller, L. K. & Fried, M. (1976) J. Virol. 18, 824-832. 12. Eckhart, W. (1969) Virology 38, 120-125. 13. Di Mayorca, G., Callender, J., Marin, G. & Giordano, R. (1969) Virology 38, 126-133. 14. Hirt, B. (1967) J. Mol. Biol. 26,365-369. 15. Sharp, P. A., Sugden, B. & Sambrook, J. (1973) Biochemistry 12, 3055-3063. 16. Yoshimori, R. N. (1971) PhD Dissertation, University of California at San Francisco. 17. Barzilai, R. (1973) J. Mol. Biol. 74, 739-742. 18. Hamer, D. & Thomas, C. (1976) Proc. Natl. Acad. Sci. USA 73, 1537-1541. 19. MacPherson, I. & Montagnier, L. (1964) Virology 23, 291294. 20. Miller, L. K. & Fried, M. (1976) Nature 259,598-601. 21. Kamen, R., Lindstrom, D. M., Shure, H. & Old, R. (1974) Cold Spring Harbor Symp. Quant. Biol. 39, 187-198. 22. Shenk, T. E., Carbon, J. & Berg, P. (1976) J. Virol. 18, 664671. 23. Oxman, M. N., Takemoto, K. K. & Eckhart, W. (1972) Virology 49, 675-682. 24. Paulin, D. & Cuzin, F. (1975) J. Virol. 15,393-397.
