[CANCER RESEARCH 34, 2585-2593,
October 1974]
Comparative Biochemical Properties of RNA-directed DNA
Polymerases from Rauscher Murine Leukemia Virus and
Avian Myeloblastosis Virus1
Larry C. Waters and Wen-Kuang Yang
Carcinogenesis Program, Biology Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37830
SUMMARY
Various biochemical properties of the RNA-directed DNA
polymerases of avian myeloblastosis virus and Rauscher
murine leukemia virus were studied. Effects of divalent
cations, pH, and temperature on the ability of the two
polymerases to utilize various template primers are signifi
cantly different. The template primers tested include DNaseactivated calf thymus DNA,d(A-T)n,
(rA)n-(rU)n, (rA)n(dT)n, (rA)n-(dT)6, (dA)n-(dT)6,
(rC)n-(dG)6, and (rl)n(dC)6. All showed individual characteristic requirements of
divalent cations, pH, and temperature for optimum activity
with the two polymerases. A minor change in the optimal
conditions usually resulted in a marked decrease in the
detected polymerase activity. In kinetic experiments, initial
deoxythymidine 5'-monophosphate incorporation appeared to
be the rate-limiting step in the (rA)n-(rU)n-dependent
activity
of the murine virus polymerase, whereas this was less apparent
with the avian virus polymerase. The d(A-T)n-dependent
activity of both purified virus polymerases is competitively
inhibited by (rU)n or (rG)n.
INTRODUCTION
RNA-directed DNA polymerase activities have been shown
in all the known RNA oncogenic viruses (11). It is almost
certain that this polymerase activity is required for viral
infection and transformation of susceptible cells (13, 27).
Because of the unique capacity to transcribe natural RNA
sequences to DNA and the obligatory role in viral oncogenesis,
a thorough knowledge of the properties and characteristics of
these enzymes is important. Justifiably, much effort has been
applied in the use of synthetic polynucleotide
template
primers to increase the sensitivity of detecting viral polym
erases (18, 26) and also in defining assay conditions that will
selectively differentiate the viral enzyme from normal cell
DNA polymerases (2, 12, 22, 30). Relatively less consideration
(9, 30) has been given concerning the manner in which known
viral RNA-dependent DNA polymerases from different sources
react under similar reaction conditions.
We previously reported (29) that the response of DNA
1Research jointly sponsored by the Cancer Virus Program of the
National Cancer Institute and the United States Atomic Energy
Commission under contract with Union Carbide Corporation.
Received March 27, 1974; accepted June 25,1974.
polymerase
from RLV2
for a particular
template
primer
depends strongly on a variety of controllable conditions in the
assay. The present study compares the RNA-directed DNA
polymerases from RLV and AMV, with emphasis on the
biochemical properties important for an effective polymerase
assay. The results showed that these 2 enzymes are markedly
different on the basis of several of the properties compared
here.
MATERIALS AND METHODS
Materials
RLV from infected mouse plasma was obtained as a 10-fold
concentrate from the National Cancer Institute (Lot RPV-HL684, Hazleton Laboratory).
In some experiments RLV
prepared in our laboratory from the plasma of infected
BALB/c mice was used. Plasma containing approximately
IO12 AMV particles/ml was kindly provided by Dr. Anne
Deeney, Oregon State University. Tritiated dNTP's and
polyribonucleotides
were obtained
from Schwarz/Mann,
Orangeburg, N. Y. Specific activities (Ci/mmole and mCi/
mmole polynucleotide phosphorus) were: dTTP, 17.4;dATP,
18.5;dCTP, 27.4; dGTP, 4.0; (rA)n, 51.0 (M.W. ~ 2.4 X 10s);
(rU)n, 7.0 (M.W. > 100,000); and (rC)n, 35.0 (M.W. >
100,000). Unlabeled dNTP's were obtained from P-L Biochemicals, Milwaukee, Wis. Nonionic detergent NP-40 was
obtained from the Shell Chemical Co., New York, N. Y. Calf
thymus DNA, DNase I, and pancreatic RNase were from
Worthington Biochemical Corp., Freehold, N. J. Nonlabeled
synthetic polynucleotides of sizes ranging from 4 to IOS were
obtained from Miles Laboratories, Kankakee, 111.Oligodeoxynucleotides were products of Collaborative Research, Waltham, Mass. Hydroxylapatite (Hypatite C) was purchased from
Clarkson Chemical Co., Williamsport, Pa., and Whatman
phosphocellulose, P-l, was from H. Reeve Angel & Co., Inc.,
Clifton, N.J.
Methods
Virus Preparation. RLV obtained from the National Cancer
Institute was pelleted by centrifugation at 164,000 X g for 30
JThe abbreviations used are: RLV, Rauscher murine leukemia virus;
AMV, avian myeloblastosis virus; dNTP, deoxynucleotide triphosphate;
NP-40, Nonidet P-40.
OCTOBER 1974
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1974 American Association for Cancer Research.
2585
L. C. Watersand W.-K. Yang
min and suspended in 50 mM NaCl and 15% glycerol to a
protein concentration of 1.5 mg/ml. The viruses were isolated
from plasma essentially according to the method of Duesberg
and Robinson (7). They were sedimented through a 25%
sucrose layer onto a 60% sucrose cushion, containing 50 mM
Tris-Cl (pH 7.5), 100 mM NaCl, and 1 mM EDTA, at 25,000
rpm for 90 min in a SW27 Spinco rotor. The banded virus was
then sedimented to equilibrium in a 25 to 60% sucrose
gradient in the same buffer. The band that formed at a density
of about 1.16 g/ml was pooled, pelleted, and suspended as
described above. The final protein concentrations
in 2
different AMV preparations were 3.4 and 5.2 mg/ml, and in
another RLV preparation was 5 mg/ml, as determined by the
method of Lowry et al. (17). Fifty-j/1 aliquots of the virus
preparations were stored in liquid nitrogen until used.
Determination of Polymerase Activity. A typical reaction
mixture contained 30 mM Tris-Cl (pH 8.0 at 22U):50 to 100
M KC1 to contain 500 mM KC1 and then frozen and thawed 3
times in liquid nitrogen to ensure complete disruption.
(Lysates prepared in this manner can be stored for 6 to 8
weeks in liquid nitrogen without substantial loss in activity.) If
the lysate was to be directly subjected to phosphocellulose
chromatography, it was appropriately diluted to contain 50
mM KC1.
Initial purification was by glycerol gradient centrifugation
on 10 to 30% glycerol (4 ml) containing 50 mM Tris-Cl (pH
8.0):5 mM MgCl2:5 mM dithiothreitol:0.1%
NP-40:500 mM
KC1. The gradients were centrifuged at 55,000 rpm at 4°for
12 hr in a SW56 titanium Spinco rotor. Fractions were
collected from the bottom of the tube and assayed for
polymerase activity. Both the RLV and AMV polymerases
sedimented at positions consistent with their reported molecu
lar weights (15, 22). Characteristically, the RLV enzyme
sedimented as a sharp band whereas the AMV enzyme, at
mM KChbovine serum albumin, 0.17 mg/ml:2 mM dithioequivalent activity, sedimented as a broader band. Thus, the
threitol:appropriate
dNTP-3H, 17.5 ¿iCi/ml. The labeled
AMV lysate may contain a polymerasemucleic acid complex,
dNTP's were diluted by the addition of the same unlabeled
similar to that demonstrated in the Rous sarcoma virus
dNTP at 0.1 mM. When required, the other unlabeled dNTP's
preparations (6, 8).
were present at 0.2 mM. Contents and concentrations of
Further purification was attained by applying the active
template primers and divalent cations, as well as temperatures,
fractions from glycerol gradient centrifugation, adjusted to
for the incubations are indicated in the figure legends. Tris contain 0.1 M KC1, to a hydroxylapatite column that had been
buffers shift considerably in pH at different temperatures,
equilibrated with 10 mM potassium phosphate buffer (pH 7.2)
therefore, in the study of pH optima, pH measurements were containing 5 mM MgCl2:5 mM dithiothreitol:0.2% NP-40:0.1
performed with 30 mM Tris-Cl buffer at the indicated
M KC1:15% glycerol. The enzyme was eluted with a linear
temperature of the reaction. Usually, 5 to 10 ¡uof enzyme
gradient from 10 to 150 mM phosphate (pH 7.2). Either of the
fraction were assayed in a total volume of 25 to 100 ¡i\. 2 viral polymerases eluted from hydroxylapatite as a single
Twenty-jul aliquots were taken at various time intervals of peak.
incubation, and the incorporated radioactivity was measured
The pooled enzyme fractions from hydroxylapatite were
by the paper disc method of Bollum (3) with slight diluted 4 times with a medium [50 mM Tris-Cl (pH 8.0):3 mM
NPmodifications. The modifications involved washing the discs 4 MgCl2 or 0.4 mM MnCl2:5 mM dithiothreitol:0.2%
times in an ice-cold acid solution (5% glacial acetic acid:0.7%
40:15% glycerol] containing 50 mM KC1 and applied to a
HC1:1% sodium pyrophosphate), twice in ethanohether (1:1),
phosphocellulose
column preequilibrated
with the same
and finally once in ether. The discs were dried under a heat dilution medium. The enzyme was eluted with a linear KC1
lamp and placed in vials with 5 ml of 0.4% 2,5-bis[2-(5-fmgradient (50 to 500 mM) in the same medium. The RLV
butylbenzoxazoyl)]thiophene
in toluene. The incorporation
enzyme eluted as a single peak whereas AMV was resolved into
2 peaks, as shown in "Results." The 2nd AMV enzyme peak
of 1 pmole of dNMP registered 45 cpm under our conditions
of scintillation counting.
was that used in most of these studies. Active fractions from
Template Primer Preparations. Calf thymus DNA was the phosphocellulose
columns were pooled, adjusted to
activated with DNase I by the method of Aposhian and contain 0.5 M KC1, and used immediately for the experiments.
Kornberg (l) prior to use. (rA)n'(rU)n, with a molar A:U ratio
Based on enzyme activity with activated calf thymus DNA
of 0.77 as obtained commercially, was adjusted to an A:U and protein concentration
determined by the method of
ratio of 2.47 with (rA)n. The mixture, at a total polyribonuBramhall et al. (4), purification of the polymerase preparations
cleotide concentration of 0.4 mg/ml in 10 mM NaCl, was was usually in the range of 50 to 100 times for AMV and 100
heated to 45°for 10 min and then slowly cooled and stored
to 150 times for RLV (Table 1). Recovery of the polymerase
frozen at —¿20°
until used. Preliminary experiments indicated
activity usually ranged from 200 to 400% (AMV) and 100 to
that the same result was obtained whether the poly ribonucleo- 200% (RLV) of that of the virus lysate, presumably due to
removal of inhibitors (14).
tide:oligodeoxyribonucleotide
complexes used were formed
RNase Determination. The purified polymerase preparation
by such preannealing or by simply mixing them together in the
(final concentration of 1.0 Mg/ml) or the crude virus lysate
reaction mixture.
(10.0 jug/ml) was incubated at 37°up to 12 hr in a solution
Enzyme Purification. Viral lysates were made by diluting
the virus preparation with an equal volume of 50 mM Tris-Cl containing labeled polyribonucleotide,
2.5 jug/ml:50 mM
Tris-Cl (pH 8.0):2.0 mM MnCl2:5 mM dithiothreitol:50 mM
(pH 8.0) containing 100 mM NaCl. This was then diluted
2-fold by the addition of 40 mM MgCl2 (or4mM MnCl2):20
KC1. RNase H activity was determined using a similar reaction
(l:l)
(19). The
mM dithiothreitol:0.8%
NP-40. This mixture, which resulted mixture with 5.0 jug/ml (rA3H)-(dT)n
in a visible clearing of the virus suspension, was incubated in extent of hydrolysis was measured by the paper disc
an ice bath for 15 min. The sample was then adjusted with 2.5 method (3).
2586
CANCER RESEARCH VOL. 34
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1974 American Association for Cancer Research.
DNA Polymerases from 2 Oncogene
Table 1
Polymerasc and nuclease activities of RL V and AMV enzyme
preparations
RLV enzyme
Activity
Lysate
Purified"
AMV enzyme
Lysate
Purified"
Viruses
containing 0.5 M KC1. This difference in stability has more
recently been shown to be a result of the greater instability of
the AMV enzyme in the presence of low levels of dithiothreitol, the oxidation of which is considerable under those
storage conditions. Both enzymes can be stably stored, for
long periods, at —¿180°.
The purified RLV polymerase and
Peak 2 of AMV sedimented in the glycerol gradient
centrifugation at ~4 S and ~6 S, respectively, similar to the
previous reports (15, 22).
with(rA*)n-(dT)n(rA*)n(rC*)n(rU*)n86372280459,13059,26015600044154165314213,62712,9798,944000
Nuclease Activities. Table 1 shows the polymerase and
nuclease activities of the 2 viral enzyme preparations. Low
levels of RNase activity were detected in the crude virus lysate,
particularly that of AMV. After purification, both AMV and
RLV polymerase preparations snowed no nuclease activity
0 Prepared by the sequential steps of glycerol gradient centrifuga- with radioactive polyribonucleotides
as substrates, which gave
tion, hydroxylapatite chromatography, and phosphocellulose chroma
no sign of decreased sedimentation rate by glycerol gradient
tography (Peak 2 of AMV).
centrifugation following 8 hr of incubation. Most of the RNase
Expressed as pmoles dTMP incorporated per jug protein during
30-min incubation at 37° (activated DNA], at 30°[(rA)n-(dT)6,
H activity present in the crude RLV lysate appeared to be
AMV], and at 20°|(rA)n-(dT)6, RLV].
removed after purification of this polymerase; in contrast, the
c Expressed as pmoles nucleotide monophosphate hydrolyzed per ¿ig RNase H activity remained associated with the polymerase
protein in a 2-hr incubation at 37°.
activity after application of the purification procedures to
AMV. The RNase H activities of the 2 viruses functioned
optimally at 2 mM Mn2+ using (rA)n-(dT)n as substrate.
RESULTS
Polymeraseb
withActivated
DNA(rA)n-(dT)6Nuclease6'
Hydroxylapatite
and Phosphocellulose Chromatography.
RLV polymerase readily adsorbs to hydroxylapatite
in a
solution of 0.5 M KC1 and 10 mM potassium phosphate buffer
at pH 7.2 and, with a linear gradient of phosphate solution,
élûtes
at about 25 mM phosphate concentration (31). In
contrast, the AMV polymerase does not bind to hydroxylapa
tite in 10 mM phosphate unless the KC1 concentration of the
medium is lowered to 0.25 M, and then it élûtes
as a single
peak at about 40 mM phosphate concentration.
Phosphocellulose column chromatography of RLV polym
erase showed a single peak (data not shown). In contrast, AMV
polymerase gave 2 distinct peaks, 1 of which bound only
slightly, if at all, to phosphocellulose in the initial 50 mM
KC1 buffer solution (Chart 1). This was observed with crude
AMV lysates as well as with preparations from glycerol
gradient centrifugation and hydroxylapatite chromatography.
This observation of 2 polymerase activity peaks from
phosphocellulose chromatography is similar to that of Hurwitz
and Leis (14), although we usually obtained a larger
percentage of the enzyme as Peak 1 (~45%). Rechromatography of Peak 1 resulted in 30 to 40% of the activity eluting in
the region of Peak 2. Also, pretreatment of crude AMV lysate
with pancreatic RNase (1 M8/mO caused a quantitative shift of
polymerase activity from Peak 1 to Peak 2. Both Peaks 1 and 2
and the crude lysate polymerase preparations showed similar
sedimentation rates, although Peak 1 and the crude virus lysate
gave a broader and more diffuse peak. These results, with the
observation on polyribonucleotide
inhibition experiments
described later, would suggest that the 2 Chromatographie
peaks of AMV represent 2 forms of the same polymerase and
that Peak 1 may be a complex of Peak 2 with RNA (6).
Both Peak 2 polymerase of AMV and the single polymerase
peak of RLV were similarly eluted from phosphocellulose at
0.25 M KC1. However, the RLV enzyme appeared to be more
stable to storage at 4°or —¿20°
in Chromatographie medium
OCTOBER
Although there was evidence of inhibitors in our viral lysate
preparations, polymerase inhibition of the potency described
by other workers (14) was not observed in our study; such
inhibitors were completely removed from our purified
polymerase preparations.
Metal Requirements for Various Template Primers. Scolnick
et al. (25) have shown that the preference of viral polymerase
for Mn2* or Mg2* depends on the template primer that is being
used. We have shown further (29) that certain combinations of
Mn2+ and Mg2+ are better than either alone for certain
12-1-3
-0.6
-0.4
-0.2
10
20
FRACTION NUMBER
Chart 1. Phosphocellulose chromatography of AMV lysate. E"ive-/j|
aliquots of each fraction were assayed in a 25-/ul reaction volume at
40°.Radioactive incorporation in 20 >ilwas measured at 60 min. The
activated calf thymus DNA concentration was 100 jug/ml, and the
(rA)n-(rU)n concentration was 20 ¿jg/ml.Metal concentrations were:
for DNA, 1.7 mM Mg2*; and for (rA)n-(rU)n, 0.5 mM Mn2*:6 mM
Mg2*.A, (rA)n-(rU)n; *, activated DNA.
1974
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1974 American Association for Cancer Research.
2587
L. C. Waters and W.-K. Yang
template primers. These studies suggest that it is the particular
template primer that dictates the particular optimal metal
requirement.
Chart 2 illustrates the response of the 2 viral polymerases to
Mg2+ concentration,
using activated calf thymus DNA as
template primer. The RLV enzyme responds maximally at 10
to 12 mM. The AMV enzyme, although responsive at the
higher concentrations, is stimulated maximally at 1 to 2 mM.
Mn2+ alone at the optimal concentration (0.5 mM) is less than
20% as effective as Mg2+with either enzyme. Furthermore, the
addition of Mn''* at optimal Mg2+ concentration
causes
inhibition with both enzymes. The template primer d(A-T)n
has metal requirements very similar to those shown for
activated calf thymus DNA (Chart 2).
In contrast to the situation with DNA, Mn2+ rather than
Mg2+ is the preferred cation when (rA)n-(rU)n is the template
6-
ce
2p
5 O
CL
Chart 3. Effect of MgJ* and Mn2* on (rA)n-(dT)6-dependent
DNA
synthesis with (A) 0.29 »ig/ml,purified RLV polymerase, and (B) 0.2
/ag/ml, purified AMV polymerase. The enzyme activity measurements
were made at 25°.Incorporation data represent 20 ^1 after a 30-min
reaction. (rA)n concentration was 10 Mg/ml with 10 ^M (dT)6. Mg2*
and Mn2*concentrations were as indicated.
primer for dTMP incorporation. RLV polymerase shows no
activity when Mn2+ is absent from the reaction. Requirement
for Mn2+ is less absolute by the AMV polymerase, which gives
Other template primers for which we have determined
some activity with (rA)n-(rU)n by using 6 to 10 mM Mg2* optimal metal conditions (Table 2) include the following, (a)
dGMP incorporation with (rC)n-(dG)6:RLV,
8 mM Mg2+;
only. Although both polymerases show high activity with this
RNA template primer in the presence of Mn2+ alone, AMV, 12 mM Mg2+;Mn2+ substitutes but is less effective; there
combined use of Mn2+ and Mg2+ may give higher enzyme is no stimulation by combined Mg2* and Mn2+. (b) dAMP
activity than does the use of Mn2+ only. The optimal divalent incorporation with (rU)n-(dA)6:RLV, 1 to 2 mM Mn2+; AMV,
cation combinations are 0.5 to 1.0 mM Mn2+ plus 1 to 2 mM 2 mM Mn2+; Mg2+ is ineffective alone but does not seem to be
Mg2+ for RLV polymerase and 0.5 mM Mn2+ plus 2 to 6 mM inhibitory in combination with Mn2*. (c) dCMP incorporation
2 to 8 mM Mg2+; AMV, 1 to 2 mM
Mg2* for AMV polymerase. A significant difference is that, at with (r!)n-(dC)6:RLV,
Mg2+; Mn2+ is effective alone and inhibitory in combination
the optimal Mn2+ concentration, RLV polymerase-catalyzed
dTMP incorporation with (rA)n-(rU)n is markedly inhibited with Mg2+. (d) dTMP incorporation with (rA)n-(dT)n:RLV
by Mg2+ concentrations greater than 2 mM, whereas Mg2+ and AMV, both 5 to 12 mM Mg2+ concentration; the inclusion
of Mn2+ (0.5 to 1 mM) strongly inhibits the AMV activity but
begins to inhibit the AMV activity only at concentrations
only slightly inhibits the RLV activity. Although the relative
greater than 6 mM.
The metal requirements for the 2 enzymes using (rA)n- activities of the 2 enzymes are approximately equal when
(rC)n-(dG)6 or (rU)n'(dA)6 are used as template primers,
(dT)6 are shown in Chart 3. RLV polymerase has a metal
optimum of 0.5 mM Mn2+ plus 1 to 2 mM Mg2*, and Mg2+at there is a striking difference in the ability of the 2 enzymes to
utilize (r!)n-(dC)6. The (r!)n-(dC)6-dependent
activity with
concentrations higher than 4 mM becomes inhibitory (Chart
3A). In contrast, AMV polymerase prefers 8 to 10 mM Mg2+ RLV polymerase is barely detectable, whereas with AMV
alone over any Mn2+:Mg2+ combinations or Mn2+ alone (Chart polymerase it is comparable to the activity seen with
(rA)n-(rU)n or activated calf thymus DNA. Nuclease activity
3B).
against (rl)n was not specifically tested (Table 1). However,
12since AMV preparations generally had greater nuclease content
than did RLV preparations, and since purified RLV prepara
O totions also fail to utilize (r!)n-(dC)6 efficiently, we consider it
unlikely that this difference is due to nuclease.
I 8-1
We have used various metal conditions and temperatures in
an effort to get (rA)n-(rU)n-dependent
dAMP incorporation.
z
o
We have observed none with either enzyme, even in the
opresence of (dA)6. This observation probably relates to the
o
greater stability of (rA)n-(rU)n
relative to the expected
product (dA)n.(rU)n (21).
2Effect of Temperature. We reported previously (29) that the
optimum temperature for the RLV polymerase reaction varies
0
considerably, depending on the template primer used. The
246
10
8
12
14
observed optimal temperature values appeared to correlate
MgCI2 \mM\
with the thermal stability of the template primer. For
Chart 2. Effect of Mg2* concentration on DNA-dependent DNA well
example, (r!)n-(rC)n had a higher temperature optimum (47°)
synthesis with purified RLV (o) and AMV (A) enzyme preparations.
than did (rA)n-(rU)n (40°), and o(rC)n-(dG)6 was more
Enzyme activity was measured at 40°. Activated calf thymus DNA
effective at a higher temperature (40°)than was (rA)n-(dT)6
concentration was 100 ¿ig/ml.The final protein concentrations in the (20°).
assay were: RLV polymerase, 0.29 jig/ml; AMV polymerase, 0.2 fig/ml.
The data represent dTMP incorporation in 20 n\ of a 30-min reaction.
2588
The results shown in Chart 4 indicate that not only the
CANCER RESEARCH VOL. 34
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1974 American Association for Cancer Research.
DNA Polymerases from 2 Oncogenic Viruses
Table 2
Comparison of divalent metal requirements and pH and reaction temperature optima for various template primers with the
DNA polymerases fromRLVandAMV
DNase-activated calf thymus DNA was used at a concentration of 100 Mg/ml; d(A-T)n was used at 10 Mg/ml. The other
RNA templates were used at 10 Mg/ml. Oligomers at a concentration of 10 juMwere used in all cases except for the optimum
temperature determinations with (rA)n-(rU)n and (rA)n as templates, where 1 MMoligomer was used.
Polymer*DNAad(A-T)n°(rA)n.(rU)n"(rA)n.(dT)n°(rA)n-(rU)n(rA)n°(rU)n(rC)n(rl)nEnzymeRLV
incorporateddTTP
PH7.7-7.9
temperature42
(mM)10-12
1-210
AMVRLV dTTPdTTP
NoneNone
8.0-8.37.2
4540-42
or dATP
AMVRLV dATPdTTP
dTTP or
NoneNone
8.0-8.47.9
>4540>4520
AMVRLV dTTPdTTP
NoneNone
7.5-7.87.8-8.0
0.5000.5-100.5 2-65-12
AMVRLV dTTPdTTP
None(dT)6
8.1-8.37.8-8.0
5-121-2
AMVRLV dTTPdTTP
(dT)6CdT).(dT)6(dA)6
00.5-1
1-21-2
81-2
4015-20
7.7-7.9Optimum 30-3545>4537
AMVRLV dTTPdATP
00
01-220
AMVRLV dATPdGTP
(dA)6(dG)6
AMVRLV dGTPdCTP
(dG)6(dC)6
40-4535-40
AMVdNTP-3H
dCTPOligomerNone
(dC)6Optimum
45[Mn2*](mM)00[Mg"J
8-100
08
00
122-8
1-2
a Data were obtained using purified polymerase preparations also.
IWWcp5^"^O
'\/\
80-h-<Iccoè
AI
\0
»'
M
>' \
GO-oozCLS
A'
''
'A
1/
«;
A/
20-iLurrnöl
l'
'/
\'
40-!^Lu|
nature of the template primer but also the source of the
enzyme is undoubtedly very important in determining the
optimum reaction temperature. The RLV enzyme shows
optimal activity at 25 when (rA)n-(dT)6 is the template
primer, whereas the AMV enzyme is maximally active at 35°
.\ '
'o/'
>u
iY
\A
with the same template primer (Chart 4). The optimum
temperature for both enzymes increases slightly when the
concentration of (dT)6 is increased from 1 to 10 t¿M.The
higher temperature optima of the reaction of the AMV
enzyme with (rA)n-(dT)6 is not due to the higher levels of
Mg2* in the assays. With all of the template primers studied,
\
\
o
\
\
\
\
\
higher optimum reaction temperatures have been observed for
the AMV enzyme than for the RLV enzyme (Table 2). Since
the same Tris-Cl buffer (pH 8.0 at 22°)was used throughout
\
\
\Oo\>
10
30
TEMPERATURE
50
Chart 4. Optimal temperature for (dT)6-stimulated (rA)n-dependent
DNA synthesis with RLV (o) and AMV (A) polymerases. The data are
expressed as percentages of maximal activity at the indicated
temperatures and represent dTMP incorporation in 20 p\ after a 30-min
incubation. Metal conditions were: RLV, 0.5 mM Mn2+:2 mM Mg2*;
AMV, 8 mM Mg2*. (rA)n was at a final concentration of 10 Mg/mlwith
1 MM(dT)6. Final protein concentrations were: RLV polymerase, 0.29
Mg/ml, and AMV polymerase, 0.2 Mg/ml.
OCTOBER
these experiments for optimum temperature determination,
the question should be asked whether the observed differences
in the optimum temperatures for the 2 polymerases reflect
effects of pH shift caused by temperature change on the
polymerase activity (see "Effect of pH"). A careful examina
tion showed that the effect of temperature and the effect of
pH on the polymerase activity are independent.
Effect of pH. Using the respective optimal metal conditions
at various temperatures, we investigated the effects of pH on
the activities of the 2 viral polymerases in 30 mM Tris-Cl
buffers. These results can be summarized as follows: with
(rA)n-(dT)6 at 25°,pH 7.8 to 8.0 for RLV and pH 7.7 to 7.9
for AMV; with d(A-T)n at 40°,pH 7.2 for RLV and pH 8.0 to
1974
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1974 American Association for Cancer Research.
2589
L. C. Waters and W.-K. Yang
8.4 for AMV; with (rA)n-(rU)n
at 40°,pH 7.9 for RLV and
pH 7.5 to 7.8 for AMV; with activated calf thymus DNA at
40°,pH 7.7 to 7.9 for RLV and pH 8.0 to 8.3 for AMV; and
with (rA)n-(dT)n at 40°,pH 7.8 to 8.0 for RLV and pH 8.1 to
8.3 for AMV. The pH measurement varies considerably with
temperature for a single Tris-Cl buffer, so that a buffer
measuring pH 8.0 at 22°gives pH 7.6 at 40°. A control
experiment indicated that the pH optimum for the polymerase
with a given template primer remains constant at different
reaction temperatures, although it takes different Tris-Cl
buffers to maintain this constant pH. When the 2 enzymes
were assayed under identical instead of their respective
optimal metal conditions, the same results were obtained,
indicating that the pH optimum properties of the 2 enzymes
are independent of the metal contents in the reaction.
Kinetic Observations. Characteristically,
DNA synthesis,
with activated calf thymus DNA as template primer, is linear
with time for either enzyme (Chart 5), and the same is true
when d(A-T)n is the template primer. Linear kinetics with the
RLV enzyme is most obvious at enzyme concentrations lower
than that shown in this chart. However, the kinetics of the 2
enzymes with (rA)n-(rU)n as template primer are quite
different. The lag observed with the RLV polymerase is not
seen with the AMV enzyme (Chart 5). The lag seen with the
RLV enzyme appears to reflect a difficulty in initial dTMP
incorporation, presumably onto the (rU)n primer, and it is
eliminated when oligomers of deoxythymidine are added. We
must emphasize that the (rA)n-(rU)n used in all these studies
was adjusted to an A:U ratio of 2.47, the ratio empirically
determined to give maximal template primer activity. A linear
rate of synthesis is observed with the AMV enzyme. It would
appear that initiation is rate limiting to a much greater degree
40
TIME (min)
80
Chart 5. Kinetics of dTMP incorporation with DNA and (rA)n-(rU)n
as template primer with purified RLV and AMV preparations. Data
represent incorporation of dTMP in 10 [RLV with (rA)n-(rll)nl or 20
(RLV with DNA and AMV with both template primers) ti\ of the
reaction at 40°.Metals and template primers were as described in Chart
1. The final protein concentrations were: RLV, 0.58 Mg/m'>and AMV,
0.2 Mg/ml. o, RLV, (rA)n-(rU)n; •¿,
RLV, DNA; A, AMV, (rA)n-(rU)n;
*, AMV, DNA.
2590
1:32 1:161:8 1:4 1:2
Undil.
ENZYME DILUTION
Chart 6. Effect of protein concentration on the rate of DNA- and
(rA)n-(rU)n-dependent DNA synthesis with RLV and AMV polymerases. Temperature, template primer, and metals were as in Chart 1.
The data represent dTMP incorporation in 20 >il at 45 min (purified
RLV polymerase) and 30 ^1 at 40 min (purified AMV polymerase). The
protein concentrations in the undiluted purified enzyme samples were:
RLV, 0.58 Mg/â„¢';AMV, 0.7 Mg/ml- The symbols are the same as in
Chart 5.
with the RLV polymerase than with the AMV enzyme. In this
regard, when tested under optimal conditions for (rA)n-(rU)ndependent DNA synthesis, 10 juM(dT)s stimulates the rate by
approximately 20-fold with the RLV enzyme. However, under
conditions optimized for (rA)n-(rU)n-dependent
synthesis, the
AMV enzyme is stimulated only about 2-fold by the addition
of 10/jM (dT)g.
A linear relationship is observed with increasing AMV
polymerase concentration when DNA is the template primer
(Chart 6). A slightly sigmoidal curve is seen when (rA)n-(rU)n
is the template primer. DNA-dependent synthesis with the
RLV polymerase is linear with increasing protein concentra
tion. In contrast, the increase in the rate of (rA)n-(rU)ndependent DNA synthesis is enhanced with increasing protein
concentration. A highly purified preparation of RNA-directed
DNA polymerase from RLV-infected mouse spleens (31)
exhibits the same properties with DNA or (rA)n-(rU)n as
template primers, as is shown in Chart 6 for the RLV purified
polymerase.
Polyribonucleotide
Inhibition. DNA-directed polymerase
activity of RLV lysates was shown to be extremely sensitive to
inhibition by single-stranded polyribonucleotides (28). Chart 7
illustrates this observation with (rG)n. It also shows that the
AMV polymerase, prepared by glycerol gradient centrifugation, appears less sensitive to this type of inhibition. This is
particularly true of AMV lysates, which sometimes showed
almost complete resistance to the inhibition as reported by
Gallo (10). When the AMV polymerase was isolated in a highly
purified state, such as Peak 2, from phosphocellulose
chromatography,
it became susceptible to (rG)n or (rU)n
CANCER RESEARCH
VOL. 34
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1974 American Association for Cancer Research.
DNA Polymerases from 2 Oncogene
Viruses
inhibition to the same extent as did the RLV polymerase
(Chart 7). Although pretreatment of an AMV lysate signifi
cantly increased the proportion of the polymerase that eluted
as Peak 2, such pretreatment did not increase the sensitivity of
an AMV lysate to (rG)n inhibition. Thus, this marked
difference of the 2 viruses could have been due to factors
other than the polymerase molecules per se in the virus lysate.
Therefore, we performed subsequent comparative studies with
polymerase preparations of highest purity in our stock. The
data of inhibition experiments were analyzed by Cleland's
kinetic method (5). With both RLV and AMV polymerases,
plots of \¡vversus \/A (reciprocal plots) indicate competitive
inhibition of the d(A-T)n-directed
activity, namely that
polyribonucleotides alter the slope but not the intercept of the
reciprocal plots. Replots of the slope versus the inhibitor (rU)n
concentration,
however, indicate that the 2 enzymes are
inhibited in a different manner, Le., a linear curve with K¡of 3
to 4 fig/ml for the RLV polymerase and a hyperbolic curve for
the AMV polymerase (Chart 8).
DISCUSSION
Reported molecular weights of the 2 enzymes (14, 15, 22)
indicate that proteins with the same function but from
different oncogenic RNA viruses can be markedly different in
molecular weight. Similar findings with other structural
proteins components of the oncogenic RNA viruses have been
observed. Completely different protein patterns have been
demonstrated for AMV and RLV lysates by sodium dodecyl
14-1
200
Chart 7. Inhibition of RLV and AMV DNA polymerases with (rG)n.
The reaction (40°)was done in a volume of 50 julcontaining d(A-T)n,
20 fig/ml, and the data are based on DNA synthesis in 40-fil samples at
40 min. Peak polymerase fractions obtained by centrifuging a viral
lysate on a glycerol gradient were used. »,RLV; *, AMV; f-, data
obtained using highly purified AMV polymerase Peak 2 from
phosphocellulose.
OCTOBER
-Ki
25
50
(rU)n (/ig/ml)
75
100
Chart 8. Replots of slopes of reciprocal plots versus inhibitor (rU)n
concentration for the RLV polymerase (•)and the AMV polymerase
(*). d(A-T)n template primer concentration used for obtaining the
reciprocal plots at various (rU)n levels were 20, 10, 5 and 2.5 jjg/ml in
50->il reaction at 41°for 20 min. The enzymes had been purified by
procedures described in "Methods," followed by an additional step of
polyguanidylate:Sepharose
communication).
affinity Chromatograph y (D. Joseph, private
sulfate polyacrylamide gel electrophoresis (23). Even with
phylogenetically more closely related viruses, such as feline
leukemia virus and RLV, a careful comparison by the same
electrophoretic technique showed that some protein com
ponents migrate differently, possibly suggesting different
molecular weights (24).
Two facts are obvious from our data on the divalent metal
requirements for the DNA polymerase assays: (a) for a single
DNA polymerase from an oncogenic RNA virus the metal
requirements for optimal activity differ greatly, depending on
the template primers used; (b) for 2 different polymerases the
optimum metal requirements for the same template primer
may also differ greatly. Similar conclusions can be drawn from
studies on pH effects, namely, that the same viral polymerase
may require different pH conditions with different template
primers and that different viral polymerases may require
different pH conditions with the same template primer. The
effect of divalent cations and the effect of pH seem to be
independent events, since no correlation can be established.
However, both metal and pH conditions exert profound
influences on the outcome of polymerase assays. It is not
uncommon to obtain less than 50% of the maximal
polymerase activity as a result of a minor change of these
conditions, such as assaying (rA)n-(rU)n-dependent
activity of
the RLV polymerase at pH 7.6 (optimal for AMV) instead of
pH 7.9 (optimal for RLV), or using a combination of 0.5 mM
Mn2* and 6.0 mM Mg2+ instead of 0.5 mM Mn2+ and 2.0 mM
Mg2+.
In view of these results for metal and pH requirements, the
experimental data commonly found in the current literature
concerning oncogenic RNA virus polymerase, in which the
relative efficiencies of various template primers are compared
1974
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1974 American Association for Cancer Research.
2591
L. C. Waters and W.-K. Yang
under a single set of assay conditions, probably do not give a
true indication of the transcribing potential of the polymerase
tested. Similarly, the responses of different oncogenic RNA
virus polymerases to a given template primer under a single set
of uniform reaction conditions should not be taken as
indicative of the maximal potential response.
In a study of RNA synthesis by Escherichia coli RNA
polymerase, using polyribonucleotides as templates and com
plementary oligoribonucleotides as primers, Niyogi and Wilton
(20) have demonstrated a definite correlation between the
chain length of the oligomer and the optimum temperature of
the reaction and have shown that this is a reflection of the
melting temperature of the oligomer: polymer complex. Study
ing the ability of RLV polymerase to utilize various
poly(rA)-oligo(dT)
complexes for dTMP incorporation, we
have observed the same phenomenon, namely, the optimum
temperature for the reaction increases with increasing chain
length of the oligo(dT) used (W.-K. Yang and L. C. Waters,
unpublished results). If these results suggest a particular
preferred degree of oligomer:polymer association for optimal
activity of the polymerase, then the present results indicate
that the nature of the polymerase is also very important in
determining the utilization of the complex. This is illustrated
by the marked differences in temperature optima for the 2
polymerases with regard to the priming activity of (dT)6
(Chart 4). The AMV enzyme apparently uses (rA)n-(dT)6
most effectively near its melting temperature, whereas the
RLV polymerase may require a more strongly associated form
of this complex at a lower temperature. For all the template
primers tested, the AMV polymerase has higher temperature
optima than the RLV polymerase (Table 2). These results
emphasize that it is important to consider the reaction
temperature
when potential template primers are being
selected for the polymerase of a particular oncogenic RNA
virus.
Several observations with the RLV polymerase suggest that
initiation of synthesis is the rate-limiting step in RNA-primed
DNA synthesis but not in DNA-primed DNA synthesis (L. C.
Waters and W.-K. Yang, unpublished results). However, there is
little, if any, lag in the rate of DNA synthesis with the AMV
polymerase using either DNA or (rA)n-(rU)n as template
primers (Chart 5). Related to this is the comparison of the 2
polymerases with regard to the kinetic curve correlating the
rate of DNA synthesis and the enzyme concentration in the
reaction mixture. For RLV polymerase, a sigmoidal curve is
obtained with (rA)n-(rU)n as template primer but not with
DNA template primers. Similar curves are found for AMV
polymerase, although the sigmoidal relation with (rA)n-(rU)n
is less pronounced (Chart 6). In this regard, Leis and Hurwitz
(16) have also reported a sigmoidal relationship between the
rate of DNA synthesis and AMV enzyme concentration, using
RNA isolated from AMV or calf thymus DNA as template
primer. All these kinetic observations present interesting
problems regarding à possible difference in subunit structure
for the RLV and AMV polymerase and the relationship of
enzyme structure to template primer preference.
Tuominen and Kenney (28) showed that the DNA-directed
polymerase activity of RLV lysates was very sensitive to
inhibition by single-stranded polyribonucleotides. Gallo (10)
maintained that some RNA tumor virus preparations could be
2592
relatively insensitive to such inhibition. Our results with the
lysate polymerase preparations of RLV and AMV confirm
both reports. However, the fact that an extensive purification
of the AMV polymerase rendered it highly sensitive to the
inhibition indicates that the observed difference may be due to
factors other than polymerase per se in the virion. For this
reason, we have used highly purified preparations and
demonstrated a hyperbolic competitive inhibition with AMV
polymerase and a linear competitive inhibition with RLV
polymerase. It remains, however, to be determined whether
such differences in enzyme kinetic behavior represent different
complex structures of the 2 virus polymerases, which are
obvious from analyzing molecular weight and the associated
RNase H activity, or still insufficient purity of the AMV
polymerase preparation.
The experimental results presented above demonstrated that
the RNA-directed DNA polymerases from RLV and AMV are
definitely very different by biochemical parameters. Many
essentially identical results, like those described, were obtained
using crude viral lysates. We should emphasize, however, that
biologically highly active plasma viruses were used in our
study. Perhaps studies like those reported will be useful in the
definitive diagnosis of a presumptive human tumor virus.
ACKNOWLEDGMENTS
We appreciate the technical assistance of Leroy G. Hardin and C. K.
Koh. We thank Dr. Anne Deeney, Oregon State University, for
providing the AMV-infected chick plasma, and Dr. L. R. Sibal and Dr.
J. B. Moloney of the National Cancer Institute for supplying the RLV
concentrate. Dr. G. David Novelli's interest and encouragement during
the study are appreciated.
REFERENCES
1. Aposhian, H. V., and Kornberg, A. Enzymatic Synthesis of
Deoxyribonucleic Acid. IX. The Polymerase Formed after T2
Bacteriophage Infection of Escherichia coli: A New Enzyme. J.
Biol. Chem.,237: 519-525,1962.
2. Baltimore, D., and Smoler, D. Primer Requirement and Template
Specificity of the DNA Polymerase of RNA Tumor Viruses. Proc.
Nati. Acad. Sei. U. S., 68: 1507-1511, 1971.
3. Bollimi, F. J. Thermal Conversion of Nonpriming Deoxyribo
nucleic Acid to Primer. J. Biol. Chem., 234: 2733-2734, 1959.
4. Bramhall, S., Noack, N., Wu, M., and Loewenberg, J. R. A Simple
Colorimetrie Method for Determination of Protein. Anal. Biochem., 31: 146-148,1969.
5. Cleland, W. W. Steady State Kinetics. In: P. D. Boyer (ed.), The
Enzymes, Vol. 2, pp. 1-65. New York: Academic Press, Inc.,
1970.
6. Duesberg, P., Helm, K. V. D., and Canaani, E. Properties of a
Soluble DNA Polymerase Isolated from Rous Sarcoma Virus. Proc.
Nati. Acad. Sci. U. S.,68: 747-751, 1971.
7. Duesberg, P. H., and Robinson, W. S. Nucleic Acid and Proteins
Isolated from the Rauscher Mouse Leukemia Virus (MLV). Proc.
Nati. Acad. Sei. U. S., 55: 219-227, 1966.
8. Paras, A. J., Taylor, J. M., McDonnell, J. P., Levinson, W. E.,and
Bishop, J. M. Purification and Characterization of the Deoxyribo
nucleic Acid Polymerase Associated with Rous Sarcoma Virus.
Biochemistry, 11: 2334-2342, 1972.
9. Fridlender, B., and Weissbach, A. DNA Polymerases of Tumor
Viruses: Specific Effect of Ethidium Bromide on the Use of
CANCER RESEARCH VOL. 34
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1974 American Association for Cancer Research.
DNA Polymerases from 2 Oncogene
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
Different Synthetic Templates. Proc. Nati. Acad. Sei. U. S., 68:
3116-3119,1971.
Gallo, R. C. Reverse Transcriptase, the DNA Polymerase of
Oncogenic RNA Viruses. Nature, 234: 194-198, 1971.
Gallo, R. C. RNA-Dependent DNA Polymerase in Viruses and
Cells: Views on the Current State. Blood, 39: 117-137,1972.
Goodman, N. C., and Spiegelman, S. Distinguishing Reverse
Transcriptase of an RNA Tumor Virus from Other Known DNA
Polymerases. Proc. Nati. Acad. Sci. U.S., 68: 2203-2206, 1971.
Hanafusa, H., and Hanafusa, T. Noninfectious RSV Deficient in
DNA Polymerase. Virology, 43: 313-316, 1971.
Hurwitz, J., and Leis, J. P. RNA-Dependent DNA Polymerase
Activity of RNA Tumor Viruses. I. Directing Influence of DNA in
the Reaction. J. Virol., 9: 116-129, 1972.
(vician, D. L., Watson, K. F., Burny, A., and Spiegelman, S.
Purification of the DNA Polymerase of Avian Myeloblastosis Virus.
Biochim. Biophys. Acta, 246: 365-385, 1971.
Leis, J. P., and Hurwitz, J. RNA-Dependent DNA Polymerase
Activity of RNA Tumor Viruses. II. Directing Influence of RNA in
the Reaction. J. Virol., 9: 130-142, 1972.
Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J.
Protein Measurement with the Folin Phenol Reagent. J. Biol.
Chem., 193: 265-275, 1951.
Mizutani, S., Boettiger, D., and Temin, H. M. A DNA-Dependent
DNA Polymerase and a DNA Endonuclease in Virions of Rous
Sarcoma Virus. Nature, 228: 424-427, 1970.
Moiling, K., Bolognesi, D. P., Bauer, H., Busen, W., Plassmann, H.
W., and Hausen, P. Association of Viral Reverse Transcriptase with
an Enzyme Degrading the RNA Moiety of RNA-DNA Hybrids.
Nature, 234: 240-243, 1971.
Niyogi, S. K., and Wilton, P. E. Effect of Temperature on the
Oligonucleotide Stimulation of the RNA Polymerase Reaction. J.
Mol. Biol., 41: 149-153, 1969.
Riley, M., Maling, B., and Chamberlain, M. J. Physical and
Chemical Characterization of Two- and Three-stranded AdenineThymine and Adenine-Uracil Homopolymer Complexes. J. Mol.
Biol., 20: 359-389,1966.
Ross, J., Scolnick, E. M., Todaro, G. J., and Aaronson, S. A.
Separation of Murine Cellular and Murine Leukemia Virus DNA
OCTOBER
Viruses
Polymerase. Nature, 231: 163-167,1971.
23. Schäfer,W., Bauer, H., Bolognesi, D. P., Fischinger, P., Frank, H.,
Gelderblom, H., Lange, J., and Nermut, M. V. Studies on
Structural and Antigenic Properties of C-Type Viruses. In:
Molecular Studies in Viral Neoplasia. M. D. Anderson Hospital and
Tumor Institute at Houston, pp. 115-140. Baltimore: The
Williams & Wilkins Co., 1974.
24. Schafer, W., De Noronha, F., Lange, J., and Bolognesi, D. P.
Comparative Studies on Group-Specific Antigens of RNA-Leukemia Viruses. In: L. G. Silverstri (ed.), Biology of Oncogenic
Viruses, pp. 116-123. New York: American Eisevier Publishing
Co., Inc., 1971.
25. Scolnick, E., Rands, E., Aaronson, S. A., and Todaro, G. J.
RNA-Dependent DNA Polymerase Activity in Five RNA Viruses:
Divalent Cation Requirements. Proc. Nati. Acad. Sei. U. S., 67:
1789-1796, 1970.
26. Spiegelman, S., Burny, A., Das, M. R., Keydar, J., Schlom, J.,
Travnicek, M., and Watson, K. Synthetic DNA-RNA Hybrids and
RNA-DNA Duplexes as Templates for the polymerases of the
Oncogenic RNA Viruses. Nature, 228: 430-432, 1970.
27. Ting, R. C., Yang, S. S., and Gallo, R. C. Reverse Transcriptase,
RNA Tumor Virus Transformation and Derivatives of Rifamycin
SV. Nature, 236: 163-166, 1972.
28. Tuominen, F. W., and Kenney, F. T. Inhibition of the DNA
Polymerase of Rauscher Leukemia Virus by Single-Stranded
Polyribonucleotides. Proc. Nati. Acad. Sei. U. S., 68: 2198-2202,
1971.
29. Waters, L. C., and Yang, W. K. Factors Influencing the Activities of
the RNA-Dependent DNA Polymerase from Rauscher Leukemia
Virus. Federation Proc., 30: 1163, 1971.
30. Wells, R. D., Fluegel, R. M., Larson, J. E., Schendel, P. F., and
Sweet, R. W. Comparison of Some Reactions Catalyzed by
Deoxyribonucleic Acid Polymerase from Avian Myeloblastosis
Virus, Escherichia coli, and Micrococcus luteus. Biochemistry, 11:
621-629,1972.
31. Yang, W. K., Koh, C. K., and Waters, L. C. Preparation of
RNA-Directed DNA Polymerase from Spleens of Balb/c Mice
Infected with Rauscher Leukemia Virus. Biochem. Biophys. Res.
Commun., 47: 505-511, 1972.
1974
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1974 American Association for Cancer Research.
2593
Comparative Biochemical Properties of RNA-directed DNA
Polymerases from Rauscher Murine Leukemia Virus and Avian
Myeloblastosis Virus
Larry C. Waters and Wen-Kuang Yang
Cancer Res 1974;34:2585-2593.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/34/10/2585
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1974 American Association for Cancer Research.