Cell, Vol. 15, 1427-1437,
December
1978,
Copyright
0 1978 by MIT
Three Species of Polyoma Virus Tumor Antigens Share Common Peptides Probably near the Amino Termini of the Proteins John E. Smart and Yoshiaki Ito Imperial Cancer Research Fund Lincoln’s Inn Fields London WC2A 3PX, England
Summary Detergent extracts of polyoma virus-infected mouse cells contain three major proteins of approximately lOO,OOO-108,000 (lOOK), 55,000 (55K) and 21,500 (22K) daltons, which react with sera obtained from rats carrying tumors induced by the virus. A comparison of the %-methlonine-, 3H-leucine- and 3H-prollne-labeled tryptic peptides of each of these proteins by cation-exchange chromatography followed by descending paper chromatography has shown that: at least five peptides are shared by all three l-reactive proteins; at least three peptides are shared by the 55K and 22K proteins, but not by the 1OOK protein; at least three peptides are found only in the 22K protein; at least six peptides are found only in the 55K protein; and at least sixteen peptides are found only In the 1OOK protein. The results are consistent with the hypothesis that the polypeptide chains of the lOOK, 55K and 22K dalton tumor antigens of polyoma virus share a common virus-coded amino terminal region. The data also suggest that there is a portion of the polypeptide chains (probably immediately adjacent to the common amino terminal region of the molecules) that is shared by the 55K and 22K proteins, but not by the 100K protein (perhaps because this portion of the genetic information is spliced out of the messenger RNA coding for the 100K protein). The facts that all the peptides common to the 100K and 55K proteins are also found in the 22K protein and are thus assigned to the common amino terminal region of the molecules, and that there are several peptides unique to the 100K protein, as well as several peptides unique to the 55K protein, suggest that the presumed carboxy terminal portion of the poly peptide chain of the 100K protein is considerably, if not entirely, different from that of the 55K protein. Introduction The early region of polyoma virus DNA, which extends clockwise from 71 to 26 map units and comprises about 55% of the genome (see Figure I), encodes gene products which are important in DNA replication, control of RNA transcription and cell transformation (Fried and Griffin, 1976). This region contains two known complementation
groups which are defined by the temperature-sensitive tsA mutants and the host-range transformation-defective (hr-t) mutants (Eckhart, 1977; Fluck, Staneloni and Benjamin, 1977). Viruses with mutations in either complementation group are defective for cell transformation (Fried, 1965; Benjamin, 1970). Detergent extracts of polyoma virus-infected mouse cells contain several proteins which react with sera obtained from rats or hamsters carrying tumors induced by the virus (anti-T sera). The major species of these T-reactive proteins, or tumor (T) antigens, has an apparent molecular weight of 97,000-108,000 daltons upon sodium dodecylsulfate polyacrylamide gel electrophoresis (SDS-PAGE). Several smaller T-reactive proteins, which are variable in number and amount, have been reported (Turler and Salomon, 1977; Ito, Brocklehurst and Dulbecco, 1977a; Ito, Spurr and Dulbecco, 1977b; Ito et al., 1977c; Schaffhausen, Silver and Benjamin, 1978). The three major species of T-reactive proteins obtained under the experimental conditions used in the present study have apparent molecular weights of lOO,OOO108,000 (lOOK), 55,000 (55K) and 21,500 (22K) daltons. The 1OOK protein is identified as the product of the tsA cistron because the tsA mutations, which map between 1.4 and 26 map units (Miller and Fried, 1976), code for a thermolabile 1OOK protein (Ito et al., 1977b), and because deletions, which remove viral sequences between 93 and 23 map units, reduce the size of the 100K protein (Ito et al., 1977c). These observations are consistent with the earlier studies of Paulin and Cuzin (1975), who showed by complement fixation that the T-reactive proteins produced by mouse cells transformed by the tsA mutant of polyoma virus were inactivated at a faster rate at the restrictive temperature (39°C) than those produced by wild-type-transformed cells. The size of the 100K protein is not affected by the hr-t deletion mutations (Ito et al., 1977b; Schaffhausen et al., 1978), which occur between 80 and 85 map units (Fried and Griffin, 1976; Feunteun et al., 1977). On the other hand, in cells infected with these mutants, the 22K protein is either absent or present in a truncated form (Ito et al., 1977c), thus making it the most likely candidate for the hr-t cistron. The 55K protein is not detected in cells infected with the hr-t mutant, NG18 (Y. Ito, manuscript in preparation), and is the same size as the main T-reactive protein which is associated with the plasma membranes of infected cells. The plasma membrane-associated 55K protein is not detected either in cells infected with NG18 (Ito et al., 1977a). The maximum molecular weight of the polypeptide chain which the early region of the genome
Cell 1428
which have been metabolically labeled with 35Smethionine, 3H-leucine and 3H-proline. The results are compatible with the final possibility listed above and suggest a novel structural organization for the early region of the virus.
Results Isolation of T-Reactive Proteins
Figure 1. Circular Physical Fried and Cowie. 1974)
Map of
Polyoma
Virus
DNA (Griffin,
Early and late regions are indicated with arrows showing the 5’ to 3’ direction of the mRNA.s transcribed from those regions (Kamen et al., 1974). The area where tsA mutants have been mapped (Miller and Fried, 1976) and the area where the deletions of the hr-t mutant have been mapped (Feunteun et al., 1977; Ito et al., 1977~) are indicated.
could encode in a single reading frame is approximately 110,000 daltons. If the three proteins are encoded by three independent segments of DNA, this coding capacity is clearly insufficient, since the sum of the molecular weights of the three proteins is estimated to be 176,500 daltons. Using guanidine/Sepharose chromatography, Schaffhausen et al. (1978) have estimated the molecular weight of the largest T-reactive protein to be 81,000 daltons. Even allowing for the possibility that the molecular weight of each of the T-reactive proteins is overestimated -20% by SDS-PAGE, the predicted sum of their molecular weights would be approximately 140K daltons. Thus it seemed probable that one or more of the proteins was derived from host-coded information. If, on the other hand, these three proteins contained information derived from the early region of the viral genome, the following possibilities had to be considered: the molecular weights of the proteins had been grossly overestimated; one or more of the smaller proteins are proteolytic cleavage products of the larger proteins; one or more of the proteins are partly encoded in the late region of the viral genome, using anti-late codons; one or more of the proteins are partly virus- and partly host cell-coded; or the proteins are translated from three different messenger RNAs which, although transcribed from the same portion of the virus genome, are distinct due to a post-transcriptional splicing event. In this case, the reading frame of one molecule need not be the same as that of the others. In an attempt to determine the relatedness of the lOOK, 55K and 22K dalton proteins, we have compared the tryptic peptides of these T-reactive proteins from polyoma virus-infected mouse cells
Previous work has shown that mouse 3T6 cells infected with the tsA mutant of polyoma virus “overproduced” the T-reactive proteins when incubated at the restrictive temperature, 39°C (Y. Ito, unpublished observations). This result is presumably due to “overproduction” of the early mRNAs under these conditions (Cogen, 1978). Thus to obtain larger amounts of the radiolabeled tumor antigens, 3T6 cells were infected with the tsA mutant and incubated at 32°C for 42 hr, followed by 4 hr at 39°C. Detergent extracts of these cells were reacted with anti-tumor sera obtained from rats carrying tumors induced by the virus. The antigen-antibody complexes were isolated by using formalin-fixed, protein A-producing Staphylococcus aureus. Figure 2 shows an autoradiograph of the 35S-methionine-labeled T-reactive proteins which have been separated by SDS-PAGE. The major species of T-reactive proteins migrate with apparent molecular weights of lOO,OOO-108,000 (lOOK), 55,000 (55K) and 21,500 (22K) daltons. Parallel extracts of cells metabolically labeled with 3H-proline and 3H-leucine yielded comparable fluorographs (data not shown). After localization by autoradiography or fluorography, these major bands were excised, performic acid-oxidized and digested with trypsin. The resulting tryptic peptides were then analyzed by cation-exchange chromatography followed by descending paper chromatography.
Comparison of the Tryptic Peptides of lOOK, 55K and 22K Dalton Proteins by Cation-Exchange Column Chromatography Figure 3a shows the elution profiles of the 35Smethionine-labeled tryptic peptides of the lOOK, 55K and 22K dalton T-reactive proteins of polyoma virus-infected mouse cells. The peptides eluting in fractions 7-12 represent those peptides which failed to bind to the column. The second cluster of peptides at positions 47-51 appears to be due to an incompletely understood solvent front artifact: for example, when an amino acid mixture is chromatographed under the same conditions, a cluster of amino acids (including the most acidic ones) is seen at this position. The remaining peaks represent single or co-chromatographing groups of peptides that lie well within the pH gradient (see
Peptide 1429
Analysis
of Three
Polyoma
Virus
T Antigens
. OrI in 100 55
22
ries of smaller peaks at 152-157,159-164 and 166168 for the 55K protein; and at positions 51-55 for the 22K protein. Although the relative scarcity of peaks makes it difficult to determine the relatedness of the proteins, the results suggest that the three proteins may be structurally related. At the same time, it is clear that the 1OOK and the 55K proteins are distinctly different; in fact, there seem to be more methionine-containing peptides in the 55K protein than in the 100K protein. Figure 3b shows the elution profiles of the 3Hleucine-labeled peptides of the three proteins. The large number of peaks is consistent with the fact that leucine is one of the more abundant amino acids in most proteins. Furthermore, the general complexity of the leucine-containing peptides seems to represent more closely what might be expected on the basis of the molecular weight of the proteins. The occurrence of these peptides in the three proteins is made clearer in the subsequent analysis of the cation-exchange column peaks by paper chromatography. The earlier elution of the “solvent front” peak in the 100K protein profile was due to an error in the loading procedure (both the 3H-leucine and 35S-methionine counts, which normally elute at positions 47-51, appeared at the earlier position in this experiment). The elution profiles of the 3H-proline-labeled peptides of the three proteins are shown in Figure 3c. The peptide eluting at position 190-193 would appear to be shared by all three antigens, whereas the peptide at position 194-l 97 would appear to be unique to the 22K protein. The occurrence of the proline-containing peptides in the three proteins is made clearer in the following section.
Paper Chromatography of the Cation-Exchange Chromatography Peptides
ab Figure 2. T-Reactive Proteins Cells Infected with tsA Mutant T-reactive proteins labeled separated by SDS-PAGE dures. The gel was dried exposed to Kodak X-Omat normal rat sera.
of Polyoma
Virus
Induced
in 3T6
with YS-methionine were isolated and as described in Experimental Proceimmediately after electrophoresis and H film for 16 hr. (a) Rat anti-T sera; (b)
Experimental Procedures). The peaks at positions 78-82 and 170-l 78 are present in all three proteins, thus suggesting the existence of commonly shared methionine-containing tryptic peptides in all three antigens. On the other hand, unique methionine peaks occur at positions 183-188 for the 100K protein; at positions 135-143, 194-198, plus a se-
As stated in Experimental Procedures, peptide peaks from the cation-exchange analysis were appropriately pooled and then analyzed in a second dimension by descending paper chromatography. The results are summarized in Table 1, which lists the more clearly separated peptides with respect to their occurrence in the three forms of the tumor antigens; their position of elution from the cation exchange (top line); the position of the pooled cation-exchange peaks upon reanalysis by descending paper chromatography (bottom line); and the labeled amino acids which were contained in the peptides chromatographing at the given positions in both dimensions (in parentheses immediately below the positional information). It should be mentioned that we have listed only the more clearly separated peptides, and thus have necessarily underestimated the total number of methionine and/or leucineand/or proline-containing peptides of each of the proteins.
Gel I 1430
T-------1 a! I ntai ?wL‘ia
a 1OOK Mst
Fraction Figure
b
3. Cation-Exchange
1OOK LeU
Number Column
Fraction Chromatography
of the Tryptic
Number
Peptides
of the T-Reactive
Fraction Proteins
of Polyoma
Number Virus
(a) 3%S-methionine-labeled, (b) 3H-leucine-labeled and (c) 3H-proline-labeled tryptic peptides. Although equal portions Of the “Smethionine-labeled tryptic peptides were appropriately mixed with both the 3H-leucineand 3H-proline-labeled peptides before CatiOnexchange chromatography, the profiles of each amino acid are presented separately. The molecular weight of the tumor antigen and the amino acid with which it was labeled are given in each panel.
To illustrate the type of data from which Table 1 is constructed, some examples from each of the columns of Table 1 are presented in the remainder of this section. It is worth stating at this point that, although we have separated the 3H-leucine and 3H-proline profiles from those of %-methionine, in the actual experiment each of the 3H-amino acid preparations was chromatographed with, and therefore counted with, an equivalent ?S-methionine preparation (see Experimental Procedures for further details). Three examples of peptides which are common to all three forms of the tumor antigens are shown in Figure 4. The left-hand panel shows the results of paper chromatography of the peak eluting from the cation-exchange column at position -78-88. It appears that this fraction from the 55K and 22K proteins contains only a single species of methionine-containing peptide, while that from the 1OOK protein appears to contain an additional methionine-containing peptide which migrates slightly behind (peaking at paper position -13) the major common peptide. Inspection of the proline profile of this column fraction indicates that the slower migrating peptide unique to the 100K protein also contains a proline residue. In other experiments, the two methionine peptides of the 100K protein have been clearly resolved (data not shown). A
proline-containing peptide unique to the 22K protein is also observed in this fraction [at positions -78-84 (column), 19-21 (paper)]. The middle panel of Figure 4 shows that the common proline peak at column position -189193 migrates as a homogeneous peptide upon paper chromatography at position -19-21. Furthermore, it would appear that this peptide, which is shared by all three proteins, also contains leutine. The right-hand panel shows the second dimension analysis of the final major methioninecontaining peak, which appears to be shared by all three forms of polyoma virus tumor antigens (column position -170-177). Although the quantitative yields are at variance, it appears that the two methionine peaks at paper positions -7-9 and 1 l14 are shared by all three proteins. In separate mixing experiments, these two peaks were found to be indistinguishable in 35S-methionine-labeled 100K and 22K dalton proteins (data not shown). The methionine-labeled peptide (which also appears to contain a leucine residue) at paper position -17-20 and the methionine-containing peptide at paper position -21-22 seem to be common to the 55K and 22K proteins, but absent in the 100K protein. This fraction also contains a leucine-labeled peptide (paper position -23-26) unique to the 1 OOK protein.
Peptide 1431
Table
Analysis
of Three
1. The More
Common to WOK, 55K, and 22K Proteins 76-02 14-16 (M) 170-177 -..__ 7-9
Clearly
Polyoma
Separated
Virus
T Antigens
Tryptic
Common to 55K and 22K Proteins
Peptides
of the IOOK, 55K and 22K Dalton
Unique Protein
40-52 21-23 (M, L, P)
51-55 16-19
170-173
70-02
17-20
to 22K
W)
Unique Protein
Tumor
Antigens
of Polyoma
to 55K
46-52 2-4 (MS P) 135-143
Unique
to 1OOK Protein
40-52 14-17
165-169 2-3
m
P)
70-82
168-169
27-31
2-3
11-15
W)
(M, L)
P)
WV 4
(Ma P)
170-l 77 11-14
170-l 73 21-23
194-197 14-17
152-I 57 6-8
143-l 46 5-7
W)
w
P)
177-102 35-40
W) 159-165
9-11
m 143-l
22-25 (M, L. P)
22-25
191-200 19-22
166-l 66 4-6
152-156 2-3
P)
170-173 22-23
(La P)
04 161-165 13-15
(W
P)
(P)
196-l 99 10-11
159-162 2-5
163-168 19-21
(W
(W
(4 159-l
62
21-22 (4
205-220 4-7 0-j
165-167
205-220
6-9
15-17
m Tryptic peptides which are clearly distinguishable by the combination of cation-exchange column chromatography paper chromatography are listed with respect to their occurrence in the three forms of the tumor antigens, their the cation-exchange column (top line), the position of the selected cation-exchange peaks upon reanalysis chromatography (bottom line), and the labeled amino acids which were contained in the peptide chromatography both dimensions (in parentheses immediately below the positional information).
Figure 5 shows that paper chromatography of the “solvent front” peak (column position -47-52) resolves a methionine-, leucineand proline-containing peptide which is shared by the 55K and 22K proteins, but absent in the 100K protein. A methionine- and proline-containing peptide which appears to be unique to the 55K protein (paper positon -2-4) also occurs in this column fraction. The small shoulder of methionine seen only in the 22K protein (column position -51-55) is clearly separated by paper chromatography and thus would appear to be unique to the 22K protein. A proline-containing peptide unique to the IOOK protein also occurs in this fraction at paper position -14-17. Figure 6 shows the second dimension analysis of a series of peptides which were unique to the 55K dalton protein in the first dimension. Since the broad peak at column position 135-143 appeared to be somewhat heterogeneous, it was split into two fractions. The fact that neither fraction moves significantly from the origin in the second dimension makes interpretation questionable; however, if this peptide were a single species, it would
(R 170-I 73 2-3 (P)
46
(L)
L
Virus
w followed by descending position of elution from by descending paper at the given positions in
appear to contain both methionine and proline residues. A methionine-, leucine- and proline-containing peptide unique to the 55K protein is seen at position -159-165 (column), 22-25 (paper). Two more methionine-containing peptides found only in the 55K protein are observed at positions -152157 (column), 5-8 (paper) and 166-168 (column), 5-8 (paper). The methionine-containing peak at column position -196-199, which appeared to be homogeneous (sharply eluting) in the cation-exchange dimension, is resolved into two species upon paper chromatography. However, since separate experiments have shown that the amount of the faster migrating species at paper position -1619 varies considerably (approaching zero in some experiments), it would seem that it may originate from a modification of the slower migrating species (paper position -g-10), which occurs after cationexchange chromatography. Figure 7 shows a series of peptides which appear to be unique to the 100K protein. Among these are proline-containing peptides at positions -143-146 (column), 5-7 (paper); 152-156 (column), 2-3 (paper); 165-167 (column), 8-9 (paper); 165-169 (col-
Cell 1432
Fraction Figure 4. Paper Chromatographs Three Tumor Antigens
of Selected
The molecular cation-exchange
antigen, the amino in each panel.
weight of the tumor column are given
Peaks
from acid
Number
the Cation-Exchange with
umn), 2-3 (paper); 168-169 (column), 9-l 1 (paper); 170-175 (column), 2-4 (paper); and 181-185 (column), 13-15 (paper). Leucine-containing peptides, which are unique to the 100K protein, occur at positions -159-162 (column), 2-5 (paper); 159-162 (column), 21-22 (paper); 205-220 (column), 3-4 (paper); and 205-220 (column), 15-17 (paper). A prolineand leucine-containing peptide, which is found only in the 1OOK protein, is seen at position -143-147 (column), 22-25 (paper). The methionine-containing peptide, which was unique to the
which
Column
it was labeled
Which
Contain
and the position
Tryptic of elution
Peptides
Shared
of the peptide
by All from
the
100K protein and appeared to be homogeneous in the cation-exchange dimension (column position -183-188), is resolved into two peaks in the second dimension chromatography. However, since separate experiments have shown that the quantity of the faster migrating peak (paper position -2528) varies considerably, only the peak at paper position -19-21 is considered to be unique to the 100K protein. The lower right-hand section of Figure 7 shows the second dimension separation of the proline-
Paptide
Analysis
of Three
Polyoma
Virus
T Antigens
1433
2?.
50-53
t
55K LW
.’
\
rjoj
II
49-50
I
22K Leu
/
27..
21"
36..
.'
29..
II-
22K
22K Met 51-55
10
SO
Fraction
Number
Figure 5. Paper Chromatographs of Selected Peaks from the Cation-Exchange 55K and 22K Dalton Proteins, but Not by the lOOK Dalton Protein The molecular cation-exchange
weight of the tumor column are given
antigen, the amino in each panel.
containing peptide that appeared the 22K protein by cation-exchange phy (column position -194-197). migrates as a fairly homogeneous paper chromatography, it would peptide unique to the 22K protein.
acid
with
to be unique to chromatograSince it also peptide upon appear to be a
Discussion We have investigated the relatedness of three major forms of the tumor antigens of polyoma virus by comparative tryptic peptide analysis. The antigens were isolated by immunoprecipitation of detergent extracts of lytically infected mouse cells which had been metabolically labeled with 35S-methionine, 3H-leucine and 3H-proline. Table 1, which lists only the more clearly separated labeled tryptic peptides of the lOOK, 55K and 22K dalton tumor antigens, shows that there are at least five peptides which are shared by all three proteins; three peptides which are shared by the 55K and 22K proteins, but not by the 100K protein; three peptides which are found only in the 22K protein; six peptides which are found only in the 55K protein; and sixteen peptides which are found only in the 100K protein. Comparative peptide mapping of three distinct
which
Column
it was labeled
’ Which
n w)
Contain
and the position
?o
a Tryptic of elution
i
a Peptide
*
KJ
Shared
by the
of the peptide
from
the
proteins can yield only seven possible modes in which a given peptide can occur. In the present study, the possibilities were that a given peptide could be shared by the lOOK, 55K and 22K dalton proteins; shared by the 100K and 55K proteins, but not by the 22K protein; shared by the 100K and 22K proteins, but not by the 55K protein; shared by the 55K and 22K protein, but not by the 1OOK protein; unique to the 100K protein; unique to the 55K protein; and unique to the 22K protein. Of these seven possibilities, the only two which do not occur in this study are the second (shared by the 1OOK and 55K proteins, but not by the 22K protein) and third (shared by the 100K and 22K protein, but not by the 55K protein). This distribution of the tryptic peptides of the lOOK, 55K and 22K dalton forms of the tumor antigens of polyoma virus has distinct implications with regard to the organization of their polypeptide chains. The fact that the hr-t deletions affect the 22K protein is consistent with the interpretation that the genetic information between 80 and 85 map units codes for a part of the 22K protein. Indeed, recently determined nucleotide sequences of polyoma virus DNA show that there is one open reading frame (containing no termination codons) between -74
Cell 1434
Fraction
Number
Figure 6. Paper Chromatographs of Selected Peaks from the Cation-Exchange Column Which Contain Dalton Protein The molecular weight of the tumor antigen, the amino acid with which it was labeled and the position cation-exchange column are given in each panel.
and 85 map units which could code for a polypeptide of about 23,000 daltons (E. Soeda, J. R. Arrand, N. Smolar and B. E. Griffin, manuscript in preparation). Thus we have assumed that the 22K protein has an amino terminus near 74 map units and a carboxyl terminus near 85 map units. Table 1 shows that eight of the eleven peptides found in the 22K protein are also found in the 55K protein. We have therefore assigned the peptides which are shared by the 22K and 55K proteins to the amino terminal regions of the two proteins. Table 1 also shows that the 100K protein shares at least five peptides with both the 55K and 22K proteins. Since the region between -80 and 85 map units (hr-t deletion mutations) does not encode the 1OOK protein, any segment shared by the 100K and 22K proteins would have to be derived from between about 74 and 80 map units. The data are therefore consistent with the hypothesis that this portion of the viral genome codes for a common amino terminal portion for all three forms of the tumor antigens. Table 1 shows that the 55K and 22K proteins share at least three peptides that are not found in the 1OOK protein. Since we have assigned the region between -74 and 80 map units to the amino
Tryptic
Peptides
of elution
Unique
of the peptide
to the 55K from
the
terminal portion which is shared by all three proteins, these peptides are most probably derived from the region of the viral genome that lies between -80 and 85 map units. The fact that NG18, which has a 3% deletion between 80 and 85 map units (Feunteun et al., 1977; Ito et al., 1977c), affects both the 55K and 22K proteins (Ito et al., 1977a, 1977c) is consistent with the assignment of these peptides to this region of the genome. The -15,000 dalton T-reactive protein observed in NG18-infected cells (Ito et al., 1977c) could be a truncated form of the 22K and/or 55K proteins. As mentioned in the Introduction, mutations between 93 and 26 map units affect the 1OOK protein, thus suggesting that this portion of the viral genome codes for the 100K protein. The assignment of the amino terminal portion of the 100K protein between -74 and 80 map units therefore implies that it is encoded by at least two noncontiguous segments of viral DNA. This interpretation is consistent with the observation that one class of cytoplasmic messenger RNA from the early region of polyoma virus does not contain the nucleotide sequences which occur between about 80 and 85 map units (Ft. Kamen, unpublished observations). By comparison, SV40, which has a genetic orga-
Peptide 1435
Analysis
of Three
Polyoma
Virus
T Antigens
90.
100 K
100K
f
m
-7
:,
Met
183-188
50.
,,
30.. 10..
36.
h
100K Pro 165-167
26., 20 ”
c_
100K Leu
205-209
2?
1OOK
f
152-156
:
100K LeU
! 1
216-220
i
22K Pro a-197
1
A 0
lo
20
m
10
50
Fraction Figure 7. Paper Protein The molecular cation-exchange
Chromatographs weight of the tumor column are given
of Selected
Peaks
from
antigen, the amino in each panel.
Number
the Cation-Exchange acid
with
nization very similar to that of polyoma virus (Fried and Griffin, 1976), induces two early proteins which are equivalent to the 1OOK and 22K dalton proteins of polyoma virus (Crawford et al., 1978; Sleigh et al., 1978). It has recently been shown that the amino terminal portions of these two SV40 proteins are identical (Paucha et al., 1978). Comparison of the nucleotide sequences of the equivalent regions of the DNA of polyoma virus and SV40 (which presumably code for the 22K proteins and the amino terminal regions of the 1OOK proteins) reveals a striking similarity in the amino acid sequences deduced from the nucleotide sequences of both viruses (Friedmann, Doolittle and Walter, 1978; E. Soeda et al., manuscript in preparation). Table 1 lists sixteen peptides unique to the 100K protein, six peptides unique to the 55K protein and three peptides unique to the 22K protein. Since we have assigned a common amino terminal portion to the three proteins, the unique peptides of each
which
Column
it was labeled
Which
Contain
and the position
Peptides of elution
Unique
to the ICrCIK Dalton
of the peptide
from
the
protein are assigned to the carboxyterminal portion of the polypeptide chains. As stated above, the maximum size of the polypeptide chain which would be derived from the viral information between -74 and 85 map units is approximately 23,000 daltons (E. Soeda et al., manuscript in preparation). This implies that the information for the remainder of the 1OOK and 55K proteins (70,000-90,000 daltons and 25,000-35,000 daltons, respectively) must come from that portion of the mRNA which codes for the carboxyterminal portions of these molecules. If the genetic information for the carboxyterminal regions of the 100K and 55K proteins is contained in the early region of the viral DNA, and if the mRNA is read in the same coding frame, then one would expect to find a series of peptides which are shared by the 100K and 55K proteins, but not by the 22K protein. Table 1 reveals no peptides that are shared by the 100K and 55K proteins which are not also shared by the
call 1436
22K protein. Furthermore, the presumed carboxyterminal portion of the 55K protein is rich in methionine-containing peptides (all six peptides unique to the 55K protein contain methionine residues), whereas the presumed carboxyportion of the 1OOK protein is rich in proline-containing and deficient in methionine-containing peptides (ten of the sixteen 100K protein unique peptides contain proline residues, whereas only two contain methionine residues). These observations allow one to consider the following possibilities for the origin of the unique portion of the 55K protein. First, if the genetic information for the presumed carboxyterminal portion of the 55K protein is derived from the early region of the viral DNA, then most, if not all, of this information is translated from a reading frame other than that used for the 1OOKprotein; second, the genetic information for the presumed carboxyterminal region of the 55K protein is derived from host-coded sequences; or third, the genetic information for the presumed carboxyterminal region of the 55K protein is derived from the late region of viral DNA using the anti-late strand Although we have listed three peptides which are unique to the 22K protein and tentatively assigned them to the carboxyterminal region of this molecule, we believe that further characterization of these peptides is required. The present study suggests that polyoma virus codes for at least three early proteins. It is possible, therefore, that polyoma virus may have three distinct early functions. If this is the situation, then it should be possible to separate these functions genetically by inducing the appropriate mutations. Whether or not the two apparently different genetic defects of the hr-t mutant-the inability to grow in certain types of cells and the inability to transform hamster and rat cells (Benjamin, 1970)-can be directly attributed to a functional defect in the 55K or 22K proteins remains to be determined. Experimental
Procedures
Cells and Virus Mouse 3T6 cells (Todaro and Green, 1963) were used as host cells throughout the experiments. The cells were infected with the tsA mutant of polyoma virus (Fried, 1965) and incubated at the permissive temperature (32°C) until late in infection, followed by shift-up to the nonpermissive temperature (39°C). The multiplicity of infection was 10-20 pfu per cell. The culture medium was Dulbecco’s modified Eagle’s medium (DMEM) supplemented by 5% fetal calf serum for cell growth and 3% horse serum after virus infection.
Isolatbn
of Radloactively
Labeled
T-Reactive
Protelnr
Briefly, infected cells were labeled at 32°C 24-42 hr after infection, followed by 4 hr of incubation at 39°C in medium containing 20% of the normal level of unlabeled methionine and leucine, 3% horse serum and 100 &i/ml of ‘8S-methionine (SW Ci/mmole), 500 &i/ml of 3H-Ieucine (40 Cilmmole) or 500 pCi/ml of SH-
proline (29 Ci/mmole) (Radiochemical Centre, Amersham. England). After labeling, cells were removed from culture dishes with versene and washed once with Tris-saline, and extracts were made with 2 ml per 2 x IO’ cells of an extraction buffer containing 100 mM Tris-HCI (pH &O), 100 mM NaCI, 0.5% NP40 and 200 pg/ ml of fresh phenylmethylsulfonyl fluoride (PMSF). Isolation of Treactive proteins from these extracts by the protein A antibody adsorbent method using rat anti-T serum followed by SDS-PAGE was essentially the same as described earlier (Ito et al., 1977b), except that the concentration of acrylamide was raised to 12%. A more detailed report of the extraction and isolation of the Treactive proteins will be given elsewhere (Y. Ito, manuscript in preparation).
Tryptlc
DlgesUon
oi 1fw T-Reactive
Proteins
5sS-labeled polypeptide bands were visualized by autoradiography of dried gels, whereas 3H-labeled peptide bands were visualized by fluorography after impregnation of the gels with 2,5diphenyloxazole (PPO) (Bonner and Laskey, 1974). The regions of the gel containing the polypeptides were excised and swollen in gel destain solution (520 ml H,O, 400 ml methanol, 53 ml acetic acid). PPO was removed from 3H-labeled gel slices by extraction with two to three changes of dimethylsulfoxide (DMSO). The DMSO was subsequently removed by two changes of gel destain solution. After the paper onto which the gels were dried was carefully removed, the slices were minced and completely dried under vacuum. The polypeptides were oxidized with performic acid (Hirs, 1967) while still in the gel slices. After lyophilization, the polypeptide-containing gel slices were suspended in 5 ml of 0.1 M NH,HCOI and digested for 16 hr at 37°C with 50 pg of TPCKtreated trypsin (Worthington Biochemicals). A further 50 Fg of TPCK-treated trypsin were added and the digestion was continued for 4 hr. The gel slices were then removed by low-speed centrifugation. and washed once with 3 ml of 0.1 M NH,HCO,, and the resulting supernatants were combined and lyophilized. The relative recovery of all the tryptic peptides of 35S-methionine-labeled 1 OOK protein is the same when the intact protein is eluted from the gel before digestion; when the protein is digested in the gel after fixation with gel destain solution; or when the protein is digested in the gel after incorporation and subsequent removal of PPO (data not shown).
Chromatography
of Tryptic
Peptidas
The peptides were dissolved for about IO min in 100 ~1 of formic acid, which was then diluted with 400 ~1 of H,O. The digest was applied to a 0.4 cm x 24 cm column of Chromobead P cationexchange resin (Technicon Chemicals S.A.). After following the sample with 4.5 ml of 2O%formic acid, the column wasdeveloped at 50°C with a gradient formed from three equal chambers containing 3 ml of formic acid + 0.3 ml pyridine + 56.7 ml H,O and a final equal chamber containing 15 ml of pyridine + 45 ml H,O, at a flow rate of -12 ml/hr. Fractions of 59 drops were collected. A 300 ~1 aliquot of each fraction was mixed with 3 ml of Aquasol (New England Nuclear), and the radioactivity was determined in a liquid scintillation counter. Fractions containing the labeled peptides were appropriately combined and dried under a stream of N,. Peptide fractions were dissolved in 100 ~1 of 20% formic acid and spotted onto Whatman IMM chromatography paper. Peptides were separated by descending chromatography in butanol:acetic acid:water (4:1:5) (Bennet, 1967) for 16-20 hr. After drying, the strips of paper were cut into 50 equal pieces (-1 cm) and elutad with 1 ml of 20% formic acid by gentle rocking overnight at room temperature. The paper strips were removed and the eluded peptides were mixed with 6 ml of Aquasol before counting in a liquid scintillation counter. To simplify the presentation of the column and paper chromatography profiles, each of the labeled amino acid profiles has been separated. In the actual experiment, however, the SH-Ieutine hyptic digest was mixed with half of the 35S-methionine
Peptide 1437
Analysis
of Three
Polyoma
Virus
T Antigens
tryptic digest before cation-exchange chromatography, while the 3H-proline tryptic digest was mixed with the other half of the %methionine tryptic digest. The mixing was carried out as follows: 1OOK leucine versus 22K methionine, 55K leucine versus IOOK methionine. 22K leucine versus 55K methionine, IOOK proline versus IOOK methionine, 55K proline versus 55K methionine and 22K proline versus 22K methionine.
Acknowledgments The authors gratefully acknowledge the technical assistance of Ms. Susan Crowhurst and Mr. N. Spurr. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Received
August
24, 1978; revised
October
References
Cogen, B. (1978). Virus-specific a tsA mutant of polyoma virus.
of polyoma
virus.
infected
temperature-sen(hr-t) mutants
Feunteun, J.. Sompayrac, L., Fluck. M. and Benjamin, Localization of gene functions in polyoma virus DNA. Acad. Sci. USA 73, 4169-4173.
of
T. (1977). Proc. Nat.
Hr-t and ts-a: 77, 810-624.
Fried, M. (1985). Cell transforming ability of a temperature-sensitive mutant of polyoma virus. Proc. Nat. Acad. Sci. USA 53, 488491. Fried, M. and Griffin, B. E. (1976). Organization of the genomes of polyoma virus and SV40. Adv. Cancer Res. 24, 87-103. Friadmann, T.. Doolittle. R. and Walter, G. (1978). Amino acid sequence homology between polyoma and SV40 tumor antigens deduced from nucleotide sequences. Nature 274, 291-293. B. E., Fried, M. and Cowie, A. (1974). Polyoma map. Proc. Nat. Acad. Sci. USA 71, 2077-2081.
Hirs, C. H. W. (1967). Performic acid Enzymology, 77, S. P. Colowick and York: Academic Press), pp. 198-199.
DNA-a
oxidation. In Methods in N. 0. Kaplan, eds. (New
Ito, Y., Brocklehurst, J. and Dulbecco, R. (1977a). Virus-specific proteins in the plasma membrane of cells lytically-infected or transformed by polyoma virus. Proc. Nat. Acad. Sci. USA 74, 4866-4670. Ito. Y.. Spurr. N. and Dulbecco, polyoma virus T-antigen. Proc. 1263. Ito, Y., Brocklehurst. J.. Spurr, Fried, M. (1977~). Polyoma virus
Miller, L. and Fried, M. (1978). the polyoma genome. J. Virol.
R. (1977b). Nat. Acad.
Construction 16, 824-832.
of the genetic
map of
Paucha, E., Mellor, A., Harvey, R., Smith, A. E., Hewick. R. M. and Waterfield, M. D. (1978). SV40 large and small T-antigens have identical amino terminal mapping at 0.65 map units. Proc. Nat. Acad. Sci. USA 75,2185-2169. Paulin, D. and Cuzin, F. (1975). Polyoma Synthesis of a modified heat-labile T antigen with the ts-a mutant. J. Virol. 15, 393-397.
virus T antigen. in cells transformed
Tilrler, H. and Salomon. C. (1977). Characterization antigen. In Early Proteins of Oncogenic Viruses INSERM-EMBO, 69, pp. 131-144.
I.
of polyoma TColloquium,
in Proof
by
Crawford, L., Cole, C.. Smith, A., Paucha. E.. Tegtmeyer, P.. Rundell, K. and Berg, P. (1978). Organization and expression of early genes of simian virus 40. Proc. Nat. Acad. Sci. USA 75. 117121.
Griffin, physical
R., Lindstrom, D., Shure, H. and Old, R. (1974). VirusRNA in cells productively infected or transformed by virus. Cold Spring Harbor Symp. Quant. Biol. 39, 187-
Note Added
early RNA in 3T6 cells Virology 85, 222-230.
Fluck, M., Staneloni, R. and Benjamin, T. (1977). two early gene functions of polyoma virus. Virology
INSERM-
Todaro, G. J. and Green, H. (1983). Quantitative studies on the growth of mouse embryo cells in culture and their development into established lines. J. Cell Biol. 17, 299-313.
A film detection method for acids in polyacrylamide gels.
Eckhart. W. (1977). Complementation between sitive (ts) and host range nontransforming polyoma virus. Virology 77, 589-597.
Colloquium,
Sleigh, M. J., Topp, W. C.. Hanich, R. and Sambrook, J. F. (1978). Mutants of SV40 with an altered small t protein are reduced in their ability to transform cells. Cell 74, 79-88.
Bennet, C. J. (1967). Paper chromatography and electrophoresis; special procedure for peptide maps. In Methods in Enzymology, 17, S. P. Colowick and N. 0. Kaplan, eds. (New York: Academic Press), pp. 330-339. Bonner, W. and Laskey. R. (1974). tritium-labeled proteins and nucleic Eur. J. Biochem. 46, 83-88.
Kamen, specific polyoma 198.
Viruses
Schatfhausen, B., Silver, J. and Benjamin, T. (1978). Tumor antigen(s) in cells productively infected by wild-type polyoma virus and mutant NG-18. Proc. Nat. Acad. Sci. USA 75, 79-83.
2, 1978
Benjamin, T. L. (1970). Host range mutants Proc. Nat. Acad. Sci. USA67, 394-399.
In Early Proteins of Oncogenesis EMBO, 69, pp. 145-151.
Characterization of Sci. USA 74, 1259-
N.. Griffith% M., Hurst, J. and wild-type and mutant T-antigens.
John tory,
E. Smart’s present P.O. Box 100, Cold
address is Cold Spring Harbor LaboraSpring Harbor, New York 11724.
December
1978,
Copyright
0 1978 by MIT
Three Species of Polyoma Virus Tumor Antigens Share Common Peptides Probably near the Amino Termini of the Proteins John E. Smart and Yoshiaki Ito Imperial Cancer Research Fund Lincoln’s Inn Fields London WC2A 3PX, England
Summary Detergent extracts of polyoma virus-infected mouse cells contain three major proteins of approximately lOO,OOO-108,000 (lOOK), 55,000 (55K) and 21,500 (22K) daltons, which react with sera obtained from rats carrying tumors induced by the virus. A comparison of the %-methlonine-, 3H-leucine- and 3H-prollne-labeled tryptic peptides of each of these proteins by cation-exchange chromatography followed by descending paper chromatography has shown that: at least five peptides are shared by all three l-reactive proteins; at least three peptides are shared by the 55K and 22K proteins, but not by the 1OOK protein; at least three peptides are found only in the 22K protein; at least six peptides are found only in the 55K protein; and at least sixteen peptides are found only In the 1OOK protein. The results are consistent with the hypothesis that the polypeptide chains of the lOOK, 55K and 22K dalton tumor antigens of polyoma virus share a common virus-coded amino terminal region. The data also suggest that there is a portion of the polypeptide chains (probably immediately adjacent to the common amino terminal region of the molecules) that is shared by the 55K and 22K proteins, but not by the 100K protein (perhaps because this portion of the genetic information is spliced out of the messenger RNA coding for the 100K protein). The facts that all the peptides common to the 100K and 55K proteins are also found in the 22K protein and are thus assigned to the common amino terminal region of the molecules, and that there are several peptides unique to the 100K protein, as well as several peptides unique to the 55K protein, suggest that the presumed carboxy terminal portion of the poly peptide chain of the 100K protein is considerably, if not entirely, different from that of the 55K protein. Introduction The early region of polyoma virus DNA, which extends clockwise from 71 to 26 map units and comprises about 55% of the genome (see Figure I), encodes gene products which are important in DNA replication, control of RNA transcription and cell transformation (Fried and Griffin, 1976). This region contains two known complementation
groups which are defined by the temperature-sensitive tsA mutants and the host-range transformation-defective (hr-t) mutants (Eckhart, 1977; Fluck, Staneloni and Benjamin, 1977). Viruses with mutations in either complementation group are defective for cell transformation (Fried, 1965; Benjamin, 1970). Detergent extracts of polyoma virus-infected mouse cells contain several proteins which react with sera obtained from rats or hamsters carrying tumors induced by the virus (anti-T sera). The major species of these T-reactive proteins, or tumor (T) antigens, has an apparent molecular weight of 97,000-108,000 daltons upon sodium dodecylsulfate polyacrylamide gel electrophoresis (SDS-PAGE). Several smaller T-reactive proteins, which are variable in number and amount, have been reported (Turler and Salomon, 1977; Ito, Brocklehurst and Dulbecco, 1977a; Ito, Spurr and Dulbecco, 1977b; Ito et al., 1977c; Schaffhausen, Silver and Benjamin, 1978). The three major species of T-reactive proteins obtained under the experimental conditions used in the present study have apparent molecular weights of lOO,OOO108,000 (lOOK), 55,000 (55K) and 21,500 (22K) daltons. The 1OOK protein is identified as the product of the tsA cistron because the tsA mutations, which map between 1.4 and 26 map units (Miller and Fried, 1976), code for a thermolabile 1OOK protein (Ito et al., 1977b), and because deletions, which remove viral sequences between 93 and 23 map units, reduce the size of the 100K protein (Ito et al., 1977c). These observations are consistent with the earlier studies of Paulin and Cuzin (1975), who showed by complement fixation that the T-reactive proteins produced by mouse cells transformed by the tsA mutant of polyoma virus were inactivated at a faster rate at the restrictive temperature (39°C) than those produced by wild-type-transformed cells. The size of the 100K protein is not affected by the hr-t deletion mutations (Ito et al., 1977b; Schaffhausen et al., 1978), which occur between 80 and 85 map units (Fried and Griffin, 1976; Feunteun et al., 1977). On the other hand, in cells infected with these mutants, the 22K protein is either absent or present in a truncated form (Ito et al., 1977c), thus making it the most likely candidate for the hr-t cistron. The 55K protein is not detected in cells infected with the hr-t mutant, NG18 (Y. Ito, manuscript in preparation), and is the same size as the main T-reactive protein which is associated with the plasma membranes of infected cells. The plasma membrane-associated 55K protein is not detected either in cells infected with NG18 (Ito et al., 1977a). The maximum molecular weight of the polypeptide chain which the early region of the genome
Cell 1428
which have been metabolically labeled with 35Smethionine, 3H-leucine and 3H-proline. The results are compatible with the final possibility listed above and suggest a novel structural organization for the early region of the virus.
Results Isolation of T-Reactive Proteins
Figure 1. Circular Physical Fried and Cowie. 1974)
Map of
Polyoma
Virus
DNA (Griffin,
Early and late regions are indicated with arrows showing the 5’ to 3’ direction of the mRNA.s transcribed from those regions (Kamen et al., 1974). The area where tsA mutants have been mapped (Miller and Fried, 1976) and the area where the deletions of the hr-t mutant have been mapped (Feunteun et al., 1977; Ito et al., 1977~) are indicated.
could encode in a single reading frame is approximately 110,000 daltons. If the three proteins are encoded by three independent segments of DNA, this coding capacity is clearly insufficient, since the sum of the molecular weights of the three proteins is estimated to be 176,500 daltons. Using guanidine/Sepharose chromatography, Schaffhausen et al. (1978) have estimated the molecular weight of the largest T-reactive protein to be 81,000 daltons. Even allowing for the possibility that the molecular weight of each of the T-reactive proteins is overestimated -20% by SDS-PAGE, the predicted sum of their molecular weights would be approximately 140K daltons. Thus it seemed probable that one or more of the proteins was derived from host-coded information. If, on the other hand, these three proteins contained information derived from the early region of the viral genome, the following possibilities had to be considered: the molecular weights of the proteins had been grossly overestimated; one or more of the smaller proteins are proteolytic cleavage products of the larger proteins; one or more of the proteins are partly encoded in the late region of the viral genome, using anti-late codons; one or more of the proteins are partly virus- and partly host cell-coded; or the proteins are translated from three different messenger RNAs which, although transcribed from the same portion of the virus genome, are distinct due to a post-transcriptional splicing event. In this case, the reading frame of one molecule need not be the same as that of the others. In an attempt to determine the relatedness of the lOOK, 55K and 22K dalton proteins, we have compared the tryptic peptides of these T-reactive proteins from polyoma virus-infected mouse cells
Previous work has shown that mouse 3T6 cells infected with the tsA mutant of polyoma virus “overproduced” the T-reactive proteins when incubated at the restrictive temperature, 39°C (Y. Ito, unpublished observations). This result is presumably due to “overproduction” of the early mRNAs under these conditions (Cogen, 1978). Thus to obtain larger amounts of the radiolabeled tumor antigens, 3T6 cells were infected with the tsA mutant and incubated at 32°C for 42 hr, followed by 4 hr at 39°C. Detergent extracts of these cells were reacted with anti-tumor sera obtained from rats carrying tumors induced by the virus. The antigen-antibody complexes were isolated by using formalin-fixed, protein A-producing Staphylococcus aureus. Figure 2 shows an autoradiograph of the 35S-methionine-labeled T-reactive proteins which have been separated by SDS-PAGE. The major species of T-reactive proteins migrate with apparent molecular weights of lOO,OOO-108,000 (lOOK), 55,000 (55K) and 21,500 (22K) daltons. Parallel extracts of cells metabolically labeled with 3H-proline and 3H-leucine yielded comparable fluorographs (data not shown). After localization by autoradiography or fluorography, these major bands were excised, performic acid-oxidized and digested with trypsin. The resulting tryptic peptides were then analyzed by cation-exchange chromatography followed by descending paper chromatography.
Comparison of the Tryptic Peptides of lOOK, 55K and 22K Dalton Proteins by Cation-Exchange Column Chromatography Figure 3a shows the elution profiles of the 35Smethionine-labeled tryptic peptides of the lOOK, 55K and 22K dalton T-reactive proteins of polyoma virus-infected mouse cells. The peptides eluting in fractions 7-12 represent those peptides which failed to bind to the column. The second cluster of peptides at positions 47-51 appears to be due to an incompletely understood solvent front artifact: for example, when an amino acid mixture is chromatographed under the same conditions, a cluster of amino acids (including the most acidic ones) is seen at this position. The remaining peaks represent single or co-chromatographing groups of peptides that lie well within the pH gradient (see
Peptide 1429
Analysis
of Three
Polyoma
Virus
T Antigens
. OrI in 100 55
22
ries of smaller peaks at 152-157,159-164 and 166168 for the 55K protein; and at positions 51-55 for the 22K protein. Although the relative scarcity of peaks makes it difficult to determine the relatedness of the proteins, the results suggest that the three proteins may be structurally related. At the same time, it is clear that the 1OOK and the 55K proteins are distinctly different; in fact, there seem to be more methionine-containing peptides in the 55K protein than in the 100K protein. Figure 3b shows the elution profiles of the 3Hleucine-labeled peptides of the three proteins. The large number of peaks is consistent with the fact that leucine is one of the more abundant amino acids in most proteins. Furthermore, the general complexity of the leucine-containing peptides seems to represent more closely what might be expected on the basis of the molecular weight of the proteins. The occurrence of these peptides in the three proteins is made clearer in the subsequent analysis of the cation-exchange column peaks by paper chromatography. The earlier elution of the “solvent front” peak in the 100K protein profile was due to an error in the loading procedure (both the 3H-leucine and 35S-methionine counts, which normally elute at positions 47-51, appeared at the earlier position in this experiment). The elution profiles of the 3H-proline-labeled peptides of the three proteins are shown in Figure 3c. The peptide eluting at position 190-193 would appear to be shared by all three antigens, whereas the peptide at position 194-l 97 would appear to be unique to the 22K protein. The occurrence of the proline-containing peptides in the three proteins is made clearer in the following section.
Paper Chromatography of the Cation-Exchange Chromatography Peptides
ab Figure 2. T-Reactive Proteins Cells Infected with tsA Mutant T-reactive proteins labeled separated by SDS-PAGE dures. The gel was dried exposed to Kodak X-Omat normal rat sera.
of Polyoma
Virus
Induced
in 3T6
with YS-methionine were isolated and as described in Experimental Proceimmediately after electrophoresis and H film for 16 hr. (a) Rat anti-T sera; (b)
Experimental Procedures). The peaks at positions 78-82 and 170-l 78 are present in all three proteins, thus suggesting the existence of commonly shared methionine-containing tryptic peptides in all three antigens. On the other hand, unique methionine peaks occur at positions 183-188 for the 100K protein; at positions 135-143, 194-198, plus a se-
As stated in Experimental Procedures, peptide peaks from the cation-exchange analysis were appropriately pooled and then analyzed in a second dimension by descending paper chromatography. The results are summarized in Table 1, which lists the more clearly separated peptides with respect to their occurrence in the three forms of the tumor antigens; their position of elution from the cation exchange (top line); the position of the pooled cation-exchange peaks upon reanalysis by descending paper chromatography (bottom line); and the labeled amino acids which were contained in the peptides chromatographing at the given positions in both dimensions (in parentheses immediately below the positional information). It should be mentioned that we have listed only the more clearly separated peptides, and thus have necessarily underestimated the total number of methionine and/or leucineand/or proline-containing peptides of each of the proteins.
Gel I 1430
T-------1 a! I ntai ?wL‘ia
a 1OOK Mst
Fraction Figure
b
3. Cation-Exchange
1OOK LeU
Number Column
Fraction Chromatography
of the Tryptic
Number
Peptides
of the T-Reactive
Fraction Proteins
of Polyoma
Number Virus
(a) 3%S-methionine-labeled, (b) 3H-leucine-labeled and (c) 3H-proline-labeled tryptic peptides. Although equal portions Of the “Smethionine-labeled tryptic peptides were appropriately mixed with both the 3H-leucineand 3H-proline-labeled peptides before CatiOnexchange chromatography, the profiles of each amino acid are presented separately. The molecular weight of the tumor antigen and the amino acid with which it was labeled are given in each panel.
To illustrate the type of data from which Table 1 is constructed, some examples from each of the columns of Table 1 are presented in the remainder of this section. It is worth stating at this point that, although we have separated the 3H-leucine and 3H-proline profiles from those of %-methionine, in the actual experiment each of the 3H-amino acid preparations was chromatographed with, and therefore counted with, an equivalent ?S-methionine preparation (see Experimental Procedures for further details). Three examples of peptides which are common to all three forms of the tumor antigens are shown in Figure 4. The left-hand panel shows the results of paper chromatography of the peak eluting from the cation-exchange column at position -78-88. It appears that this fraction from the 55K and 22K proteins contains only a single species of methionine-containing peptide, while that from the 1OOK protein appears to contain an additional methionine-containing peptide which migrates slightly behind (peaking at paper position -13) the major common peptide. Inspection of the proline profile of this column fraction indicates that the slower migrating peptide unique to the 100K protein also contains a proline residue. In other experiments, the two methionine peptides of the 100K protein have been clearly resolved (data not shown). A
proline-containing peptide unique to the 22K protein is also observed in this fraction [at positions -78-84 (column), 19-21 (paper)]. The middle panel of Figure 4 shows that the common proline peak at column position -189193 migrates as a homogeneous peptide upon paper chromatography at position -19-21. Furthermore, it would appear that this peptide, which is shared by all three proteins, also contains leutine. The right-hand panel shows the second dimension analysis of the final major methioninecontaining peak, which appears to be shared by all three forms of polyoma virus tumor antigens (column position -170-177). Although the quantitative yields are at variance, it appears that the two methionine peaks at paper positions -7-9 and 1 l14 are shared by all three proteins. In separate mixing experiments, these two peaks were found to be indistinguishable in 35S-methionine-labeled 100K and 22K dalton proteins (data not shown). The methionine-labeled peptide (which also appears to contain a leucine residue) at paper position -17-20 and the methionine-containing peptide at paper position -21-22 seem to be common to the 55K and 22K proteins, but absent in the 100K protein. This fraction also contains a leucine-labeled peptide (paper position -23-26) unique to the 1 OOK protein.
Peptide 1431
Table
Analysis
of Three
1. The More
Common to WOK, 55K, and 22K Proteins 76-02 14-16 (M) 170-177 -..__ 7-9
Clearly
Polyoma
Separated
Virus
T Antigens
Tryptic
Common to 55K and 22K Proteins
Peptides
of the IOOK, 55K and 22K Dalton
Unique Protein
40-52 21-23 (M, L, P)
51-55 16-19
170-173
70-02
17-20
to 22K
W)
Unique Protein
Tumor
Antigens
of Polyoma
to 55K
46-52 2-4 (MS P) 135-143
Unique
to 1OOK Protein
40-52 14-17
165-169 2-3
m
P)
70-82
168-169
27-31
2-3
11-15
W)
(M, L)
P)
WV 4
(Ma P)
170-l 77 11-14
170-l 73 21-23
194-197 14-17
152-I 57 6-8
143-l 46 5-7
W)
w
P)
177-102 35-40
W) 159-165
9-11
m 143-l
22-25 (M, L. P)
22-25
191-200 19-22
166-l 66 4-6
152-156 2-3
P)
170-173 22-23
(La P)
04 161-165 13-15
(W
P)
(P)
196-l 99 10-11
159-162 2-5
163-168 19-21
(W
(W
(4 159-l
62
21-22 (4
205-220 4-7 0-j
165-167
205-220
6-9
15-17
m Tryptic peptides which are clearly distinguishable by the combination of cation-exchange column chromatography paper chromatography are listed with respect to their occurrence in the three forms of the tumor antigens, their the cation-exchange column (top line), the position of the selected cation-exchange peaks upon reanalysis chromatography (bottom line), and the labeled amino acids which were contained in the peptide chromatography both dimensions (in parentheses immediately below the positional information).
Figure 5 shows that paper chromatography of the “solvent front” peak (column position -47-52) resolves a methionine-, leucineand proline-containing peptide which is shared by the 55K and 22K proteins, but absent in the 100K protein. A methionine- and proline-containing peptide which appears to be unique to the 55K protein (paper positon -2-4) also occurs in this column fraction. The small shoulder of methionine seen only in the 22K protein (column position -51-55) is clearly separated by paper chromatography and thus would appear to be unique to the 22K protein. A proline-containing peptide unique to the IOOK protein also occurs in this fraction at paper position -14-17. Figure 6 shows the second dimension analysis of a series of peptides which were unique to the 55K dalton protein in the first dimension. Since the broad peak at column position 135-143 appeared to be somewhat heterogeneous, it was split into two fractions. The fact that neither fraction moves significantly from the origin in the second dimension makes interpretation questionable; however, if this peptide were a single species, it would
(R 170-I 73 2-3 (P)
46
(L)
L
Virus
w followed by descending position of elution from by descending paper at the given positions in
appear to contain both methionine and proline residues. A methionine-, leucine- and proline-containing peptide unique to the 55K protein is seen at position -159-165 (column), 22-25 (paper). Two more methionine-containing peptides found only in the 55K protein are observed at positions -152157 (column), 5-8 (paper) and 166-168 (column), 5-8 (paper). The methionine-containing peak at column position -196-199, which appeared to be homogeneous (sharply eluting) in the cation-exchange dimension, is resolved into two species upon paper chromatography. However, since separate experiments have shown that the amount of the faster migrating species at paper position -1619 varies considerably (approaching zero in some experiments), it would seem that it may originate from a modification of the slower migrating species (paper position -g-10), which occurs after cationexchange chromatography. Figure 7 shows a series of peptides which appear to be unique to the 100K protein. Among these are proline-containing peptides at positions -143-146 (column), 5-7 (paper); 152-156 (column), 2-3 (paper); 165-167 (column), 8-9 (paper); 165-169 (col-
Cell 1432
Fraction Figure 4. Paper Chromatographs Three Tumor Antigens
of Selected
The molecular cation-exchange
antigen, the amino in each panel.
weight of the tumor column are given
Peaks
from acid
Number
the Cation-Exchange with
umn), 2-3 (paper); 168-169 (column), 9-l 1 (paper); 170-175 (column), 2-4 (paper); and 181-185 (column), 13-15 (paper). Leucine-containing peptides, which are unique to the 100K protein, occur at positions -159-162 (column), 2-5 (paper); 159-162 (column), 21-22 (paper); 205-220 (column), 3-4 (paper); and 205-220 (column), 15-17 (paper). A prolineand leucine-containing peptide, which is found only in the 1OOK protein, is seen at position -143-147 (column), 22-25 (paper). The methionine-containing peptide, which was unique to the
which
Column
it was labeled
Which
Contain
and the position
Tryptic of elution
Peptides
Shared
of the peptide
by All from
the
100K protein and appeared to be homogeneous in the cation-exchange dimension (column position -183-188), is resolved into two peaks in the second dimension chromatography. However, since separate experiments have shown that the quantity of the faster migrating peak (paper position -2528) varies considerably, only the peak at paper position -19-21 is considered to be unique to the 100K protein. The lower right-hand section of Figure 7 shows the second dimension separation of the proline-
Paptide
Analysis
of Three
Polyoma
Virus
T Antigens
1433
2?.
50-53
t
55K LW
.’
\
rjoj
II
49-50
I
22K Leu
/
27..
21"
36..
.'
29..
II-
22K
22K Met 51-55
10
SO
Fraction
Number
Figure 5. Paper Chromatographs of Selected Peaks from the Cation-Exchange 55K and 22K Dalton Proteins, but Not by the lOOK Dalton Protein The molecular cation-exchange
weight of the tumor column are given
antigen, the amino in each panel.
containing peptide that appeared the 22K protein by cation-exchange phy (column position -194-197). migrates as a fairly homogeneous paper chromatography, it would peptide unique to the 22K protein.
acid
with
to be unique to chromatograSince it also peptide upon appear to be a
Discussion We have investigated the relatedness of three major forms of the tumor antigens of polyoma virus by comparative tryptic peptide analysis. The antigens were isolated by immunoprecipitation of detergent extracts of lytically infected mouse cells which had been metabolically labeled with 35S-methionine, 3H-leucine and 3H-proline. Table 1, which lists only the more clearly separated labeled tryptic peptides of the lOOK, 55K and 22K dalton tumor antigens, shows that there are at least five peptides which are shared by all three proteins; three peptides which are shared by the 55K and 22K proteins, but not by the 100K protein; three peptides which are found only in the 22K protein; six peptides which are found only in the 55K protein; and sixteen peptides which are found only in the 100K protein. Comparative peptide mapping of three distinct
which
Column
it was labeled
’ Which
n w)
Contain
and the position
?o
a Tryptic of elution
i
a Peptide
*
KJ
Shared
by the
of the peptide
from
the
proteins can yield only seven possible modes in which a given peptide can occur. In the present study, the possibilities were that a given peptide could be shared by the lOOK, 55K and 22K dalton proteins; shared by the 100K and 55K proteins, but not by the 22K protein; shared by the 100K and 22K proteins, but not by the 55K protein; shared by the 55K and 22K protein, but not by the 1OOK protein; unique to the 100K protein; unique to the 55K protein; and unique to the 22K protein. Of these seven possibilities, the only two which do not occur in this study are the second (shared by the 1OOK and 55K proteins, but not by the 22K protein) and third (shared by the 100K and 22K protein, but not by the 55K protein). This distribution of the tryptic peptides of the lOOK, 55K and 22K dalton forms of the tumor antigens of polyoma virus has distinct implications with regard to the organization of their polypeptide chains. The fact that the hr-t deletions affect the 22K protein is consistent with the interpretation that the genetic information between 80 and 85 map units codes for a part of the 22K protein. Indeed, recently determined nucleotide sequences of polyoma virus DNA show that there is one open reading frame (containing no termination codons) between -74
Cell 1434
Fraction
Number
Figure 6. Paper Chromatographs of Selected Peaks from the Cation-Exchange Column Which Contain Dalton Protein The molecular weight of the tumor antigen, the amino acid with which it was labeled and the position cation-exchange column are given in each panel.
and 85 map units which could code for a polypeptide of about 23,000 daltons (E. Soeda, J. R. Arrand, N. Smolar and B. E. Griffin, manuscript in preparation). Thus we have assumed that the 22K protein has an amino terminus near 74 map units and a carboxyl terminus near 85 map units. Table 1 shows that eight of the eleven peptides found in the 22K protein are also found in the 55K protein. We have therefore assigned the peptides which are shared by the 22K and 55K proteins to the amino terminal regions of the two proteins. Table 1 also shows that the 100K protein shares at least five peptides with both the 55K and 22K proteins. Since the region between -80 and 85 map units (hr-t deletion mutations) does not encode the 1OOK protein, any segment shared by the 100K and 22K proteins would have to be derived from between about 74 and 80 map units. The data are therefore consistent with the hypothesis that this portion of the viral genome codes for a common amino terminal portion for all three forms of the tumor antigens. Table 1 shows that the 55K and 22K proteins share at least three peptides that are not found in the 1OOK protein. Since we have assigned the region between -74 and 80 map units to the amino
Tryptic
Peptides
of elution
Unique
of the peptide
to the 55K from
the
terminal portion which is shared by all three proteins, these peptides are most probably derived from the region of the viral genome that lies between -80 and 85 map units. The fact that NG18, which has a 3% deletion between 80 and 85 map units (Feunteun et al., 1977; Ito et al., 1977c), affects both the 55K and 22K proteins (Ito et al., 1977a, 1977c) is consistent with the assignment of these peptides to this region of the genome. The -15,000 dalton T-reactive protein observed in NG18-infected cells (Ito et al., 1977c) could be a truncated form of the 22K and/or 55K proteins. As mentioned in the Introduction, mutations between 93 and 26 map units affect the 1OOK protein, thus suggesting that this portion of the viral genome codes for the 100K protein. The assignment of the amino terminal portion of the 100K protein between -74 and 80 map units therefore implies that it is encoded by at least two noncontiguous segments of viral DNA. This interpretation is consistent with the observation that one class of cytoplasmic messenger RNA from the early region of polyoma virus does not contain the nucleotide sequences which occur between about 80 and 85 map units (Ft. Kamen, unpublished observations). By comparison, SV40, which has a genetic orga-
Peptide 1435
Analysis
of Three
Polyoma
Virus
T Antigens
90.
100 K
100K
f
m
-7
:,
Met
183-188
50.
,,
30.. 10..
36.
h
100K Pro 165-167
26., 20 ”
c_
100K Leu
205-209
2?
1OOK
f
152-156
:
100K LeU
! 1
216-220
i
22K Pro a-197
1
A 0
lo
20
m
10
50
Fraction Figure 7. Paper Protein The molecular cation-exchange
Chromatographs weight of the tumor column are given
of Selected
Peaks
from
antigen, the amino in each panel.
Number
the Cation-Exchange acid
with
nization very similar to that of polyoma virus (Fried and Griffin, 1976), induces two early proteins which are equivalent to the 1OOK and 22K dalton proteins of polyoma virus (Crawford et al., 1978; Sleigh et al., 1978). It has recently been shown that the amino terminal portions of these two SV40 proteins are identical (Paucha et al., 1978). Comparison of the nucleotide sequences of the equivalent regions of the DNA of polyoma virus and SV40 (which presumably code for the 22K proteins and the amino terminal regions of the 1OOK proteins) reveals a striking similarity in the amino acid sequences deduced from the nucleotide sequences of both viruses (Friedmann, Doolittle and Walter, 1978; E. Soeda et al., manuscript in preparation). Table 1 lists sixteen peptides unique to the 100K protein, six peptides unique to the 55K protein and three peptides unique to the 22K protein. Since we have assigned a common amino terminal portion to the three proteins, the unique peptides of each
which
Column
it was labeled
Which
Contain
and the position
Peptides of elution
Unique
to the ICrCIK Dalton
of the peptide
from
the
protein are assigned to the carboxyterminal portion of the polypeptide chains. As stated above, the maximum size of the polypeptide chain which would be derived from the viral information between -74 and 85 map units is approximately 23,000 daltons (E. Soeda et al., manuscript in preparation). This implies that the information for the remainder of the 1OOK and 55K proteins (70,000-90,000 daltons and 25,000-35,000 daltons, respectively) must come from that portion of the mRNA which codes for the carboxyterminal portions of these molecules. If the genetic information for the carboxyterminal regions of the 100K and 55K proteins is contained in the early region of the viral DNA, and if the mRNA is read in the same coding frame, then one would expect to find a series of peptides which are shared by the 100K and 55K proteins, but not by the 22K protein. Table 1 reveals no peptides that are shared by the 100K and 55K proteins which are not also shared by the
call 1436
22K protein. Furthermore, the presumed carboxyterminal portion of the 55K protein is rich in methionine-containing peptides (all six peptides unique to the 55K protein contain methionine residues), whereas the presumed carboxyportion of the 1OOK protein is rich in proline-containing and deficient in methionine-containing peptides (ten of the sixteen 100K protein unique peptides contain proline residues, whereas only two contain methionine residues). These observations allow one to consider the following possibilities for the origin of the unique portion of the 55K protein. First, if the genetic information for the presumed carboxyterminal portion of the 55K protein is derived from the early region of the viral DNA, then most, if not all, of this information is translated from a reading frame other than that used for the 1OOKprotein; second, the genetic information for the presumed carboxyterminal region of the 55K protein is derived from host-coded sequences; or third, the genetic information for the presumed carboxyterminal region of the 55K protein is derived from the late region of viral DNA using the anti-late strand Although we have listed three peptides which are unique to the 22K protein and tentatively assigned them to the carboxyterminal region of this molecule, we believe that further characterization of these peptides is required. The present study suggests that polyoma virus codes for at least three early proteins. It is possible, therefore, that polyoma virus may have three distinct early functions. If this is the situation, then it should be possible to separate these functions genetically by inducing the appropriate mutations. Whether or not the two apparently different genetic defects of the hr-t mutant-the inability to grow in certain types of cells and the inability to transform hamster and rat cells (Benjamin, 1970)-can be directly attributed to a functional defect in the 55K or 22K proteins remains to be determined. Experimental
Procedures
Cells and Virus Mouse 3T6 cells (Todaro and Green, 1963) were used as host cells throughout the experiments. The cells were infected with the tsA mutant of polyoma virus (Fried, 1965) and incubated at the permissive temperature (32°C) until late in infection, followed by shift-up to the nonpermissive temperature (39°C). The multiplicity of infection was 10-20 pfu per cell. The culture medium was Dulbecco’s modified Eagle’s medium (DMEM) supplemented by 5% fetal calf serum for cell growth and 3% horse serum after virus infection.
Isolatbn
of Radloactively
Labeled
T-Reactive
Protelnr
Briefly, infected cells were labeled at 32°C 24-42 hr after infection, followed by 4 hr of incubation at 39°C in medium containing 20% of the normal level of unlabeled methionine and leucine, 3% horse serum and 100 &i/ml of ‘8S-methionine (SW Ci/mmole), 500 &i/ml of 3H-Ieucine (40 Cilmmole) or 500 pCi/ml of SH-
proline (29 Ci/mmole) (Radiochemical Centre, Amersham. England). After labeling, cells were removed from culture dishes with versene and washed once with Tris-saline, and extracts were made with 2 ml per 2 x IO’ cells of an extraction buffer containing 100 mM Tris-HCI (pH &O), 100 mM NaCI, 0.5% NP40 and 200 pg/ ml of fresh phenylmethylsulfonyl fluoride (PMSF). Isolation of Treactive proteins from these extracts by the protein A antibody adsorbent method using rat anti-T serum followed by SDS-PAGE was essentially the same as described earlier (Ito et al., 1977b), except that the concentration of acrylamide was raised to 12%. A more detailed report of the extraction and isolation of the Treactive proteins will be given elsewhere (Y. Ito, manuscript in preparation).
Tryptlc
DlgesUon
oi 1fw T-Reactive
Proteins
5sS-labeled polypeptide bands were visualized by autoradiography of dried gels, whereas 3H-labeled peptide bands were visualized by fluorography after impregnation of the gels with 2,5diphenyloxazole (PPO) (Bonner and Laskey, 1974). The regions of the gel containing the polypeptides were excised and swollen in gel destain solution (520 ml H,O, 400 ml methanol, 53 ml acetic acid). PPO was removed from 3H-labeled gel slices by extraction with two to three changes of dimethylsulfoxide (DMSO). The DMSO was subsequently removed by two changes of gel destain solution. After the paper onto which the gels were dried was carefully removed, the slices were minced and completely dried under vacuum. The polypeptides were oxidized with performic acid (Hirs, 1967) while still in the gel slices. After lyophilization, the polypeptide-containing gel slices were suspended in 5 ml of 0.1 M NH,HCOI and digested for 16 hr at 37°C with 50 pg of TPCKtreated trypsin (Worthington Biochemicals). A further 50 Fg of TPCK-treated trypsin were added and the digestion was continued for 4 hr. The gel slices were then removed by low-speed centrifugation. and washed once with 3 ml of 0.1 M NH,HCO,, and the resulting supernatants were combined and lyophilized. The relative recovery of all the tryptic peptides of 35S-methionine-labeled 1 OOK protein is the same when the intact protein is eluted from the gel before digestion; when the protein is digested in the gel after fixation with gel destain solution; or when the protein is digested in the gel after incorporation and subsequent removal of PPO (data not shown).
Chromatography
of Tryptic
Peptidas
The peptides were dissolved for about IO min in 100 ~1 of formic acid, which was then diluted with 400 ~1 of H,O. The digest was applied to a 0.4 cm x 24 cm column of Chromobead P cationexchange resin (Technicon Chemicals S.A.). After following the sample with 4.5 ml of 2O%formic acid, the column wasdeveloped at 50°C with a gradient formed from three equal chambers containing 3 ml of formic acid + 0.3 ml pyridine + 56.7 ml H,O and a final equal chamber containing 15 ml of pyridine + 45 ml H,O, at a flow rate of -12 ml/hr. Fractions of 59 drops were collected. A 300 ~1 aliquot of each fraction was mixed with 3 ml of Aquasol (New England Nuclear), and the radioactivity was determined in a liquid scintillation counter. Fractions containing the labeled peptides were appropriately combined and dried under a stream of N,. Peptide fractions were dissolved in 100 ~1 of 20% formic acid and spotted onto Whatman IMM chromatography paper. Peptides were separated by descending chromatography in butanol:acetic acid:water (4:1:5) (Bennet, 1967) for 16-20 hr. After drying, the strips of paper were cut into 50 equal pieces (-1 cm) and elutad with 1 ml of 20% formic acid by gentle rocking overnight at room temperature. The paper strips were removed and the eluded peptides were mixed with 6 ml of Aquasol before counting in a liquid scintillation counter. To simplify the presentation of the column and paper chromatography profiles, each of the labeled amino acid profiles has been separated. In the actual experiment, however, the SH-Ieutine hyptic digest was mixed with half of the 35S-methionine
Peptide 1437
Analysis
of Three
Polyoma
Virus
T Antigens
tryptic digest before cation-exchange chromatography, while the 3H-proline tryptic digest was mixed with the other half of the %methionine tryptic digest. The mixing was carried out as follows: 1OOK leucine versus 22K methionine, 55K leucine versus IOOK methionine. 22K leucine versus 55K methionine, IOOK proline versus IOOK methionine, 55K proline versus 55K methionine and 22K proline versus 22K methionine.
Acknowledgments The authors gratefully acknowledge the technical assistance of Ms. Susan Crowhurst and Mr. N. Spurr. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Received
August
24, 1978; revised
October
References
Cogen, B. (1978). Virus-specific a tsA mutant of polyoma virus.
of polyoma
virus.
infected
temperature-sen(hr-t) mutants
Feunteun, J.. Sompayrac, L., Fluck. M. and Benjamin, Localization of gene functions in polyoma virus DNA. Acad. Sci. USA 73, 4169-4173.
of
T. (1977). Proc. Nat.
Hr-t and ts-a: 77, 810-624.
Fried, M. (1985). Cell transforming ability of a temperature-sensitive mutant of polyoma virus. Proc. Nat. Acad. Sci. USA 53, 488491. Fried, M. and Griffin, B. E. (1976). Organization of the genomes of polyoma virus and SV40. Adv. Cancer Res. 24, 87-103. Friadmann, T.. Doolittle. R. and Walter, G. (1978). Amino acid sequence homology between polyoma and SV40 tumor antigens deduced from nucleotide sequences. Nature 274, 291-293. B. E., Fried, M. and Cowie, A. (1974). Polyoma map. Proc. Nat. Acad. Sci. USA 71, 2077-2081.
Hirs, C. H. W. (1967). Performic acid Enzymology, 77, S. P. Colowick and York: Academic Press), pp. 198-199.
DNA-a
oxidation. In Methods in N. 0. Kaplan, eds. (New
Ito, Y., Brocklehurst, J. and Dulbecco, R. (1977a). Virus-specific proteins in the plasma membrane of cells lytically-infected or transformed by polyoma virus. Proc. Nat. Acad. Sci. USA 74, 4866-4670. Ito. Y.. Spurr. N. and Dulbecco, polyoma virus T-antigen. Proc. 1263. Ito, Y., Brocklehurst. J.. Spurr, Fried, M. (1977~). Polyoma virus
Miller, L. and Fried, M. (1978). the polyoma genome. J. Virol.
R. (1977b). Nat. Acad.
Construction 16, 824-832.
of the genetic
map of
Paucha, E., Mellor, A., Harvey, R., Smith, A. E., Hewick. R. M. and Waterfield, M. D. (1978). SV40 large and small T-antigens have identical amino terminal mapping at 0.65 map units. Proc. Nat. Acad. Sci. USA 75,2185-2169. Paulin, D. and Cuzin, F. (1975). Polyoma Synthesis of a modified heat-labile T antigen with the ts-a mutant. J. Virol. 15, 393-397.
virus T antigen. in cells transformed
Tilrler, H. and Salomon. C. (1977). Characterization antigen. In Early Proteins of Oncogenic Viruses INSERM-EMBO, 69, pp. 131-144.
I.
of polyoma TColloquium,
in Proof
by
Crawford, L., Cole, C.. Smith, A., Paucha. E.. Tegtmeyer, P.. Rundell, K. and Berg, P. (1978). Organization and expression of early genes of simian virus 40. Proc. Nat. Acad. Sci. USA 75. 117121.
Griffin, physical
R., Lindstrom, D., Shure, H. and Old, R. (1974). VirusRNA in cells productively infected or transformed by virus. Cold Spring Harbor Symp. Quant. Biol. 39, 187-
Note Added
early RNA in 3T6 cells Virology 85, 222-230.
Fluck, M., Staneloni, R. and Benjamin, T. (1977). two early gene functions of polyoma virus. Virology
INSERM-
Todaro, G. J. and Green, H. (1983). Quantitative studies on the growth of mouse embryo cells in culture and their development into established lines. J. Cell Biol. 17, 299-313.
A film detection method for acids in polyacrylamide gels.
Eckhart. W. (1977). Complementation between sitive (ts) and host range nontransforming polyoma virus. Virology 77, 589-597.
Colloquium,
Sleigh, M. J., Topp, W. C.. Hanich, R. and Sambrook, J. F. (1978). Mutants of SV40 with an altered small t protein are reduced in their ability to transform cells. Cell 74, 79-88.
Bennet, C. J. (1967). Paper chromatography and electrophoresis; special procedure for peptide maps. In Methods in Enzymology, 17, S. P. Colowick and N. 0. Kaplan, eds. (New York: Academic Press), pp. 330-339. Bonner, W. and Laskey. R. (1974). tritium-labeled proteins and nucleic Eur. J. Biochem. 46, 83-88.
Kamen, specific polyoma 198.
Viruses
Schatfhausen, B., Silver, J. and Benjamin, T. (1978). Tumor antigen(s) in cells productively infected by wild-type polyoma virus and mutant NG-18. Proc. Nat. Acad. Sci. USA 75, 79-83.
2, 1978
Benjamin, T. L. (1970). Host range mutants Proc. Nat. Acad. Sci. USA67, 394-399.
In Early Proteins of Oncogenesis EMBO, 69, pp. 145-151.
Characterization of Sci. USA 74, 1259-
N.. Griffith% M., Hurst, J. and wild-type and mutant T-antigens.
John tory,
E. Smart’s present P.O. Box 100, Cold
address is Cold Spring Harbor LaboraSpring Harbor, New York 11724.
