JOURNAL OF ViRoLoGy, Apr. 1976, p. 42-47 Copyright CD 1976 American Society for Microbiology
Vol. 18, No. 1 Printed in U.S.A.
Reverse Transcriptase Activity per Virion for Avian Myeloblastosis Virus and Rauscher Murine Leukemia Virus LEONARD F. LIEBES, MARVIN A. RICH, J. JUSTIN McCORMICK,* IRVING SALMEEN, AND LAJOS RIMAI Department of Biology, Michigan Cancer Foundation, Detroit, Michigan 48201,* and Scientific Research Staff, Ford Motor Company, Dearborn, Michigan 48121 Received for publication 30 October 1975
We have measured reverse transcriptase enzyme activity per virus particle for samples of avian myeloblastosis virus (BAI strain) and murine leukemia virus (Rauscher) using the synthetic template poly(rC)-oligo(dG). Absolute virus concentrations were determined directly by laser beat frequency spectroscopy. Enzyme activity per virion was determined from the slope of the activity plotted as a function of virus concentration. With this reverse transcriptase assay, the minimum activity (expressed as picomoles of dGTP incorporated/ virion per hour) is estimated at (28.1 ± 4.2) x 10-7 for avian myeloblastosis virus and (1.1 + 0.2) x 10-7 for murine leukemia virus. The sensitivity of this assay, which is determined by the level of incorporated radioactivity measurable above background, is 2.5 x 10-4 virions for avian myeloblastosis virus (with dGTP specific activity of 8.9 Ci/mmol) and 88 x 10-4 virions for murine leukemia virus (with dGTP specific activity of 6.52 CI/mmol). These results show that although reverse transcriptase assays can obviously be used to measure relative virus concentrations of equally purified samples of the same virus, they can be very misleading when used to compare the concentrations of different virus species. Since the discovery of RNA-directed DNA polymerase ("reverse transcriptase") by Temin and Mizutani (11) and Baltimore (1), virologists frequently have used this enzyme to identify and quantify RNA tumor viruses. The work of Filbert et al. (3), for example, demonstrates that the reverse transcriptase activity of the RD-114 virus parallels the induction of virusspecific gs antigen, and thus it can be used as a relative measure of the amount of virus in a preparation. However, the use of this enzymatic assay as a relative measure of the amount of virus gives misleading results if inhibitors of the reaction, such as RNase, are present in varying amounts. This was observed in assays for the reverse transcriptase activity of mouse leukemia virus added to human milk samples by McCormick et al. (6). It was also observed that EDTA treatment of mouse leukemia virus preparations completely destroyed reverse transcriptase activity, and that it could not be restored by addition of Mg2+ or Mn2+ (6). Therefore, although a positive reverse transcriptase test indicates the presence of virus, a negative test is inconclusive because it may result either from the inhibition of the assay by an inhibitor or from too low a concentration of virus particles present in the assayed solution. Thus it is
useful to know the reverse transcriptase activity per virion in pure preparations which minimize interference by inhibitors. In this paper we describe experiments that measure the reverse transcriptase activity per virion for two representative RNA tumor viruses. MATERIALS AND METHODS Virus purification. Avian myeloblastosis virus (AMV, BAI strain) and murine leukemia virus (MuLV, Rauscher) in plasma citrate (a 1:1 dilution of plasma with 0.306 M sodium citrate) were supplied by Joseph and Dorothy Beard and University Laboratories, respectively, through the cooperation of the Virus Cancer Program of the National Cancer Institute. For both MuLV and AMV, two independent preparations of purified virus were made from plasma citrate. Using each of these virus preparations, reverse transcriptase assays were carried out at two dilutions. Virus purification was carried out by discontinuous sucrose gradient centrifugation through a 20% (wt/wt) sucrose cushion onto a 65% (wt/wt) sucrose pad (Spinco SW50.1 rotor, 98,000 x g, 1 h). The viral light scattering band was collected and dialyzed at 4 C against four changes (over a 3-h period) of 0.005-ionic-strength phosphate buffer (Na2HPO4-KH2PO4). The samples were then centrifuged at 21,000 x g for 10 min to remove any aggregated material and dust. At this point, aliquots were removed for quantification by laser beat frequency spectroscopy as described below. The remaining ma42
VOL. 18, 1976
AMV AND MuLV REVERSE TRANSCRIPTASE ACTIVITY
terial was used in the reverse transcriptase reaction. Reverse transcriptase assay. The synthetic template reverse transcriptase reactions were done by the procedure of Sarngadharan et al. (9) with the following modifications. A measured volume of each purified virus sample was diluted to 5.5 ml with 0.1 M Tris-hydrochloride buffer (pH 8.3). The samples were then centrifuged at 4 C for 1 h (Spinco SW50.1 rotor, 139,000 x g). The pellet was resuspended in 0.050 ml of a lysing solution (0.01 M Tris-hydrochloride buffer [pH 8.31-50 ,ug of bovine serum albumin0.4% Nonidet P-40 detergent-77 mM dithiothreitol) and incubated at 4 C for 10 min. Then 0.050 ml of a standard reverse transcriptase reaction mixture {40 mM KCl, 5 ,ug of poly(rC)-oligo(dG), 2.5 ,.Ci of [3H]dGTP (6.51 or 8.90 Ci/mmol; New England Nuclear), and 12 mM magnesium chloride for AMV or 0.5 mM manganese chloride for MuLV} were added to bring the 0.050-ml of sample of lysed virus solution to a final volume of 0.100 ml. All reactions were carried out at 37 C. Aliquots of the reaction mixture were removed at 15, 30 and 60 min, and the reaction was terminated with the addition of 1 ml of 10% trichloroacetic acid. After the addition of 0.05 ml of 1.00% yeast RNA as carrier, the samples were kept overnight at 4 C, precipitated on Whatman GF/C glass-fiber filters, washed with 10% trichloroacetic acid, dried at 60 C, and then counted in a xylene2,5-diphenyloxazole (PPO)-1,4-bis-(5-phenyloxazolyl)benzene (POPOP) scintillation fluid containing 3% NCS solubilizer using a Packard liquid scintillation spectrometer. In all cases the incorporation of [3H]dGTP was linear in time. The data reported below are for 60-min reaction times. Since metal ions are known to have a particularly large effect on reverse transcriptase activity (13), particular care was taken to optimize the assays with respect to Mg2+ or Mn2+. We then compared these optimized activities with those which we obtained with the metal ion concentrations reported by Waters and Yang (13) for the purified DNA polymerases of these two respective viruses, Verma and Baltimore (12) for the purified polymerase of AMV, and Dion et al. (2) for lysed MuLV viral preparations. In the case of MuLV, we found that 0.5 mM Mn2+ was optimum with a lysed virus preparation as opposed to the use of 8 mM Mg2+ recommended by Waters and Yang for purified MuLV reverse transcriptase enzyme (8 mM Mg2+ resulted in 50% less incorporation than that with 0.5 mM Mn2+). However, we found that 12 mM Mg2+ was optimum for the assay of lysed AMV virus preparations, which is identical to the optimum Mg2+ concentration reported by Waters and Yang for purified AMV reverse transcriptase. A 12 mM Mg2+ concentration resulted in a twofold greater rate of incorporation than the 6 mM Mg2+ concentration recommended by Verma and Baltimore (12). Quantification by laser beat frequency light scattering spectroscopy. We have recently demonstrated the use of laser beat frequency light scattering for particle quantification (8). In brief, laser light is scattered from a suspension of particles which is contained in a rectangular cuvet (5 by 10
43
mm). Because of the random diffusional motion of the particles, the intensity of the scattered light is time dependent. The random time dependence of the scattered light is reflected in the output photocurrent of the photomultiplier tube which is used to detect the scattered light. This photocurrent is analyzed in terms of its time autocorrelation function, which is obtained as the output of a Saicor model SAI-42 real time correlator. This autocorrelation function generally consists of two exponentially decaying components. (i) Component A (Fig. lb and c) reflects fluctuations in the total number of particles in the scattering volume. The amplitude of this term is proportional to No, the average number of particles in the scattering volume, and the decay time is related to the diffusion time of the particles through the scattering volume (10). (ii) Component B (Fig. la-c) reflects the time dependence of relative particle positions due to random diffusional motion. This time dependence in relative positions causes fluctuations in the interference intensity of the scattered light reaching the photodetector. The strength (B) of the component is proportional to No2, and the decay time corresponds to a diffusion time over distances of the order of the wavelength of light. For a monodisperse sample, this decay time yields directly the particle diffusion constant. Under typical experimental conditions, the decay of component A is so much longer than that of component B that the former behaves as an additive constant to the latter. These considerations show that the ratio of the intensities of the two components is proportional to the number concentration of the particles, i.e., BIA = 13n0, where no is the number of particles per milliliter of solution and the constant 18 takes into account the actual size of the scattering volume, which relates no to No, as well as the geometrical and optical characteristics of the scattering spectrometer (5). Although 13 can be calculated, it is best to determine it experimentally using samples of particles for which the concentration can be determined directly by dry weight. We determined 13 using samples of 176-nm polystyrene latex spheres. The experimental BIA ratio, combined with the known latex sphere concentration, yields 13. Once the constant 1S was known, virus concentrations were obtained by serially diluting a virus solution until the number fluctuations in the scattering volume contributed sufficiently to the autocorrelation function, so as to give a measurable ratio of B/A. Figure 1 illustrates these considerations with data obtained on MuLV. The proportionality between B/A and concentration is shown in Fig. 2.
RESULTS Reverse transcriptase data for one set of assays are shown in Table 1. The enzyme activity per virion, expressed as picomoles of dGTP incorporated/virion per hour, was determined from the slope of the enzyme activity measured as a function of virus concentration, as shown in Fig. 3. The error bars in activity are the standard deviations calculated from data such
CORRELATOR OUTPUT (Arbitrary Units) ,60
a
'40 120
80
60 40-
20-
0
1
3
2
TIME (msec) 44
45
AMV AND MuLV REVERSE TRANSCRIPTASE ACTIVITY
VOL. 18, 1976
that in Table 1. The error bars in particle verse transcriptase activity per virion for AMV concentration are intrinsic to the light scatter- is (28.1 4.2) x 10-7 pmol of dGTP incorpoing measurement (8) and are ca. ±15%. The rated/virion per h, and for MuLV it is (1.1 straight line through the data in Fig. 3 is the 0.2) x 10-7 pmol of dGTP incorporated/virion average of two least squares straight lines, one per h. assuming no experimental uncertainties in the DISCUSSION concentrations and the other assuming no experimental uncertainties in the activities. The The reverse transcriptase activity per virion slope of this average straight line is taken as for AMV is comparable to that which can be the estimate of the enzyme activity per vision. calculated from data, reported by Fine et al. (4), The estimated experimental uncertainty in this for an assay of murine mammary tumor virus slope is ca. 15%. From these plots, the re- (MuMTV) in tissue culture fluid. These workers used poly(rG) oligo(dG), a reverse transcripMuLV (Rauscher) tase assay similar to that described here, and 20 electron microscopy for particle counting. The reverse activities per virion for four independent assays, calculated from Table 1 of Fine et 15 al. (4), are 29 x 10-7, 34 x 10-7, 44 x 10-7 and 75 x 10-7 pmol of dGTP incorporated/virion per h. 4~ The differences among the activities of the 10 reverse transcriptase assays for AMV, MuLV, and MuMTV could be caused by different num-
as
±
±
±
,
bers of enzyme molecules per virion,
5
inherently
different enzyme activities, different amounts of unknown inhibitors which co-purify with the virions, inactivation of enzymatic activity | | | z z| caused by virus purification, or by exposure to .8 .6 .4 I.O .2 o high temperatures (e.g., 39 C in the bird) or RELATIVE CONCENTRATION optimization conditions for the enzyme activity FIG. 2. Plot of B/A as a function of dilution of which have not been met. As for any enzyme MuLV. T'he B/A values were measured from the ex- assay, it is impossible to prove that these assays have been absolutely optimized. However, --perimentc21 correlation Ifunctions shown in Fig. 1. --
Virus
AMV
---
No. of virions in sample
MuLV
dGTP incorporated (p
VPCC
67,847
55,754
50,556
58,052
±8,872
2.9 ± 0.4
0.095
6.5)
237,772
235,476
228,059
236,624
+1,624
12.0 + 0.10
0.38
0.2)
22,402
25,588
18,434
22,141
+3,584
1.5 + 0.3
24
(15.6 + 2.0) x 107
237,490
274,722
215,206
242,473
+30,067
2.0
240
(43.4
+
105 +
X
105
(1.6
+
x 107
MuLV
(AL)
1.6)
(10.8 X
AMV
TABLE 1. Typical data for reverse transcriptase assay dpm deviation Avg Sample 1 Sample 2 Sample 3
16.8
+
Virus concentration was determined by laser beat frequency spectroscopy. b Specific activity, 8.90 Ci/mmol for AMV assay and 6.51 Ci/mmol for MuLV assay. c VPC, volume of viral plasma which would contain the amount of purified virus used in the assay assuming no particle loss during preparation. a
FIG. 1. Experimentally observed autocorrelation function for MuLV (Rauscher) at three concentrations showing the progressive appearance of number fluctuations in the scattering volume. The solid dots lie on a calculated exponential with a time constant determined from high concentration data. (a) High concentration sample where the correlator output is dominated entirely by the B component. (b) Dilution of the MuLV sample in (a) where BIA = 6.5. This corresponds to a concentration of ca. 3.6 x 108 particleslml. (c) Further dilution of the MuLV sample where B/A = 4 and corresponds to ca. 2 x 108 particleslml.
46
J. VIROL.
LIEBES ET AL. 15
a-
A
AMV
10 I
0 I-
E
0.
5
10
20
40
30
50
PARTICLE NUMBER (xIO 5) 20
'5
a. e0
0
0
E 0.
5
5
go
15
20
PARTICLE NUMBER (x 107) FIG. 3. Plot of the picomoles of dGTP incorporated as a function of virus particle number for AMV (A) and MuLV (B). The error bars are discussed in the text. The dashed line in B represents the AMV data plotted for comparison on the same scale with the MuLV data.
with respect to these arbitrarily standardized conditions, the activity of the AMV assay is ca. 24 times greater than that of MuLV. Because of this difference, the reverse transcriptase assays can be very misleading when used to compare virus concentrations of different virus species. These data can be used to estimate the sensitivity of the reverse transcriptase assays. The sensitivity limit is determined by the level of
incorporated radioactivity measurable above background. The background, which was determined by adding lysed virus preparation to the assay mixture after the addition of trichloroacetic acid, is No 2,800 dpm. If we accept No/2 = 1,400 as the minimum significant number of dpm above background (which is substantially larger than the statistical criterion of 2 No = =
75), then we find a sensitivity of 2.5 x 104
VOL. 18, 1976
AMV AND MuLV REVERSE TRANSCRIPTASE ACTIVITY
TABLE 2. Reverse transcriptase activitylvirion Virus
AMV MuLV MuMTVb
dGTP incorporated virion/h) (pmol/
(28.1 ± 4.2) (1.1 + 0.2) 29 34 44 75
x x x x x x
10-7 10-7 10-7 10-7 10-7
Sensitivity (vi-
rions)a
3.4 x 88 x 6.7 x 5.6 x 4.3 x 2.6 x
104 104 104b
104 104 10-7 104 a Normalized for comparison purposes to dGTP with specific activity of 6.51 Ci/mmol (See text). Minimum detectable signal above background of 1,400 dpm was assumed. b Calculated from Table 1 of reference 4.
virions for the AMV assay (with dGTP specific activity of 8.9 Ci/mmol) and 88 x 104 virions for the MuLV assay (with dGTP specific activity of 6.51 Ci/mmol). Assuming the same counting background, the sensitivity of the assay for MuMTV reported by Fine et al. (4) ranges from 45 x 104 to 116 x 104 virions with dGTP specific activity of 0.377 Ci/mmol. For purposes of comparison, these sensitivities are normalized to dGTP specific activity of 6.51 Ci/mmol and tabulated in Table 2. Our reverse transcriptase assays for AMV and MuLV and that for MuMTV derived from the data of Fine et al. (4) are four orders of magnitude greater than those stated by Ringold et al. (7) (See Table 3 of reference 7) for B77 strain of avian sarcoma virus (1.6 x 108 virions) and MuMTV virus (1.4 x 108 virions). Ringold et al. used poly(rA).oligo(dT) as the synthetic template, "since it confers greater sensitivity to the assay than does the more specific poly(rC)oligo(dG)" (7). The reason for the relative insensitivity reported by Ringold et al. is not readily apparent. One obvious possibility is that we and Fine et al. (4) determined the virus concentrations by direct methods, whereas Ringold et al. (7) relied on an indirect method which required knowledge of the number of RNA molecules per virion. In any event, the agreement between our data and that of Fine et al. (4) suggests that the sensitivity estimate of Ringold et al. (7) is in error. ACKNOWLEDGMENTS This work was supported by Public Health Service contracts N01-CP-33226 and N01-CP-33347 and grants CA
47
14680 and CA 14100 from the National Cancer Institute, as well as by an institutional grant from the United Foundation of Greater Detroit. We thank Ernest Retzel for technical assistance with part of this work. LITERATURE CITED 1. Baltimore, D. 1970. RNA-dependent DNA polymerase in virions of RNA tumor viruses. Nature (London) 226:1209-1211. 2. Dion, A. S., A. B. Vaidya, and S. F. Garland. 1974. Cation preferences for poly(rC)-oligo(dG)-directed DNA syntheses by RNA tumor viruses and human milk particulates. Cancer Res. 34:3509-3515. 3. Filbert, J. E., R. M. McAllister, M. 0. Nicolson, and R. V. Gilden. 1974. RD 114 virus infectivity assay by measurements of DNA polymerase activity and virus group specific antigen. Proc. Soc. Exp. Biol. Med. 145:366-370. 4. Fine, D. L., L. 0. Arthur, J. K. Plowman, E. A. Hillman, and F. Klein. 1974. In vitro system for production of mouse mammary tumor virus. Apple. Microbiol. 28:1040-1046. 5. Koppel, D. W. 1971. Analysis of gaussian light by clipped photocount autocorrelation: the effect of finite sampling times and incomplete spatial coherence. J. Appl. Phys. 42:3216-3225. 6. McCormick, J. J., L. J. Larson, and M. A. Rich. 1974. RNase inhibition of reverse transcriptase activity in human milk. Nature (London) 251:737-740. 7. Ringold, G., E. Y. Lasfragues, J. M. Bishor, and H. E. Varmus. 1975. Production of mouse mammary tumor virus by cultured cells in the absence and presence of hormones: assay by molecular hybridization. Virology 65:135-147. 8. Salmeen, I., L. Rimai, L. Liebes, M. A. Rich, and J. J. McCormick. 1975. Hydrodynamic diameters of RNA tumor viruses: studies by laser beat frequency light scattering spectroscopy of avian myeloblastosis and Rauscher murine leukemia viruses. Biochemistry 14:134-141. 9. Sarngadharan, M. G., P. S. Sarin, and R. C. Gallo. 1972. Reverse transcriptase activity of human acute leukemic cells: purification of the enzyme, response to AMV 70S RNA, and characterization of the DNA products. Nature (London) New Biol. 240:67-72. 10. Schaefer, D. W., and B. J. Berne. 1972. Light scattering from nongaussian concentration fluctuations. Phys. Rev. Lett. 28:475-477. 11. Temin, H. M., and S. Mizutani. 1970. RNA-dependent DNA polymerase in virions of Rous sarcoma virus. Nature (London) 226:1211-1213. 12. Verma, I. and D. Baltimore. 1973. Purification of the RNA-directed DNA polymerase from avian myeloblastosis virus and its assay with polynucleotide templates, p. 125-130. In K. Moldave and L. Grossman (ed.), Methods in enzymology, vol. 29. Academic Press Inc., New York. 13. Waters, L. C., and W. Yang. 1974. Comparative biochemical properties of RNA-directed DNA polymerase from Rauscher murine leukemia virus and avian myeloblastosis virus. Cancer Res. 34:2585-2593.
Vol. 18, No. 1 Printed in U.S.A.
Reverse Transcriptase Activity per Virion for Avian Myeloblastosis Virus and Rauscher Murine Leukemia Virus LEONARD F. LIEBES, MARVIN A. RICH, J. JUSTIN McCORMICK,* IRVING SALMEEN, AND LAJOS RIMAI Department of Biology, Michigan Cancer Foundation, Detroit, Michigan 48201,* and Scientific Research Staff, Ford Motor Company, Dearborn, Michigan 48121 Received for publication 30 October 1975
We have measured reverse transcriptase enzyme activity per virus particle for samples of avian myeloblastosis virus (BAI strain) and murine leukemia virus (Rauscher) using the synthetic template poly(rC)-oligo(dG). Absolute virus concentrations were determined directly by laser beat frequency spectroscopy. Enzyme activity per virion was determined from the slope of the activity plotted as a function of virus concentration. With this reverse transcriptase assay, the minimum activity (expressed as picomoles of dGTP incorporated/ virion per hour) is estimated at (28.1 ± 4.2) x 10-7 for avian myeloblastosis virus and (1.1 + 0.2) x 10-7 for murine leukemia virus. The sensitivity of this assay, which is determined by the level of incorporated radioactivity measurable above background, is 2.5 x 10-4 virions for avian myeloblastosis virus (with dGTP specific activity of 8.9 Ci/mmol) and 88 x 10-4 virions for murine leukemia virus (with dGTP specific activity of 6.52 CI/mmol). These results show that although reverse transcriptase assays can obviously be used to measure relative virus concentrations of equally purified samples of the same virus, they can be very misleading when used to compare the concentrations of different virus species. Since the discovery of RNA-directed DNA polymerase ("reverse transcriptase") by Temin and Mizutani (11) and Baltimore (1), virologists frequently have used this enzyme to identify and quantify RNA tumor viruses. The work of Filbert et al. (3), for example, demonstrates that the reverse transcriptase activity of the RD-114 virus parallels the induction of virusspecific gs antigen, and thus it can be used as a relative measure of the amount of virus in a preparation. However, the use of this enzymatic assay as a relative measure of the amount of virus gives misleading results if inhibitors of the reaction, such as RNase, are present in varying amounts. This was observed in assays for the reverse transcriptase activity of mouse leukemia virus added to human milk samples by McCormick et al. (6). It was also observed that EDTA treatment of mouse leukemia virus preparations completely destroyed reverse transcriptase activity, and that it could not be restored by addition of Mg2+ or Mn2+ (6). Therefore, although a positive reverse transcriptase test indicates the presence of virus, a negative test is inconclusive because it may result either from the inhibition of the assay by an inhibitor or from too low a concentration of virus particles present in the assayed solution. Thus it is
useful to know the reverse transcriptase activity per virion in pure preparations which minimize interference by inhibitors. In this paper we describe experiments that measure the reverse transcriptase activity per virion for two representative RNA tumor viruses. MATERIALS AND METHODS Virus purification. Avian myeloblastosis virus (AMV, BAI strain) and murine leukemia virus (MuLV, Rauscher) in plasma citrate (a 1:1 dilution of plasma with 0.306 M sodium citrate) were supplied by Joseph and Dorothy Beard and University Laboratories, respectively, through the cooperation of the Virus Cancer Program of the National Cancer Institute. For both MuLV and AMV, two independent preparations of purified virus were made from plasma citrate. Using each of these virus preparations, reverse transcriptase assays were carried out at two dilutions. Virus purification was carried out by discontinuous sucrose gradient centrifugation through a 20% (wt/wt) sucrose cushion onto a 65% (wt/wt) sucrose pad (Spinco SW50.1 rotor, 98,000 x g, 1 h). The viral light scattering band was collected and dialyzed at 4 C against four changes (over a 3-h period) of 0.005-ionic-strength phosphate buffer (Na2HPO4-KH2PO4). The samples were then centrifuged at 21,000 x g for 10 min to remove any aggregated material and dust. At this point, aliquots were removed for quantification by laser beat frequency spectroscopy as described below. The remaining ma42
VOL. 18, 1976
AMV AND MuLV REVERSE TRANSCRIPTASE ACTIVITY
terial was used in the reverse transcriptase reaction. Reverse transcriptase assay. The synthetic template reverse transcriptase reactions were done by the procedure of Sarngadharan et al. (9) with the following modifications. A measured volume of each purified virus sample was diluted to 5.5 ml with 0.1 M Tris-hydrochloride buffer (pH 8.3). The samples were then centrifuged at 4 C for 1 h (Spinco SW50.1 rotor, 139,000 x g). The pellet was resuspended in 0.050 ml of a lysing solution (0.01 M Tris-hydrochloride buffer [pH 8.31-50 ,ug of bovine serum albumin0.4% Nonidet P-40 detergent-77 mM dithiothreitol) and incubated at 4 C for 10 min. Then 0.050 ml of a standard reverse transcriptase reaction mixture {40 mM KCl, 5 ,ug of poly(rC)-oligo(dG), 2.5 ,.Ci of [3H]dGTP (6.51 or 8.90 Ci/mmol; New England Nuclear), and 12 mM magnesium chloride for AMV or 0.5 mM manganese chloride for MuLV} were added to bring the 0.050-ml of sample of lysed virus solution to a final volume of 0.100 ml. All reactions were carried out at 37 C. Aliquots of the reaction mixture were removed at 15, 30 and 60 min, and the reaction was terminated with the addition of 1 ml of 10% trichloroacetic acid. After the addition of 0.05 ml of 1.00% yeast RNA as carrier, the samples were kept overnight at 4 C, precipitated on Whatman GF/C glass-fiber filters, washed with 10% trichloroacetic acid, dried at 60 C, and then counted in a xylene2,5-diphenyloxazole (PPO)-1,4-bis-(5-phenyloxazolyl)benzene (POPOP) scintillation fluid containing 3% NCS solubilizer using a Packard liquid scintillation spectrometer. In all cases the incorporation of [3H]dGTP was linear in time. The data reported below are for 60-min reaction times. Since metal ions are known to have a particularly large effect on reverse transcriptase activity (13), particular care was taken to optimize the assays with respect to Mg2+ or Mn2+. We then compared these optimized activities with those which we obtained with the metal ion concentrations reported by Waters and Yang (13) for the purified DNA polymerases of these two respective viruses, Verma and Baltimore (12) for the purified polymerase of AMV, and Dion et al. (2) for lysed MuLV viral preparations. In the case of MuLV, we found that 0.5 mM Mn2+ was optimum with a lysed virus preparation as opposed to the use of 8 mM Mg2+ recommended by Waters and Yang for purified MuLV reverse transcriptase enzyme (8 mM Mg2+ resulted in 50% less incorporation than that with 0.5 mM Mn2+). However, we found that 12 mM Mg2+ was optimum for the assay of lysed AMV virus preparations, which is identical to the optimum Mg2+ concentration reported by Waters and Yang for purified AMV reverse transcriptase. A 12 mM Mg2+ concentration resulted in a twofold greater rate of incorporation than the 6 mM Mg2+ concentration recommended by Verma and Baltimore (12). Quantification by laser beat frequency light scattering spectroscopy. We have recently demonstrated the use of laser beat frequency light scattering for particle quantification (8). In brief, laser light is scattered from a suspension of particles which is contained in a rectangular cuvet (5 by 10
43
mm). Because of the random diffusional motion of the particles, the intensity of the scattered light is time dependent. The random time dependence of the scattered light is reflected in the output photocurrent of the photomultiplier tube which is used to detect the scattered light. This photocurrent is analyzed in terms of its time autocorrelation function, which is obtained as the output of a Saicor model SAI-42 real time correlator. This autocorrelation function generally consists of two exponentially decaying components. (i) Component A (Fig. lb and c) reflects fluctuations in the total number of particles in the scattering volume. The amplitude of this term is proportional to No, the average number of particles in the scattering volume, and the decay time is related to the diffusion time of the particles through the scattering volume (10). (ii) Component B (Fig. la-c) reflects the time dependence of relative particle positions due to random diffusional motion. This time dependence in relative positions causes fluctuations in the interference intensity of the scattered light reaching the photodetector. The strength (B) of the component is proportional to No2, and the decay time corresponds to a diffusion time over distances of the order of the wavelength of light. For a monodisperse sample, this decay time yields directly the particle diffusion constant. Under typical experimental conditions, the decay of component A is so much longer than that of component B that the former behaves as an additive constant to the latter. These considerations show that the ratio of the intensities of the two components is proportional to the number concentration of the particles, i.e., BIA = 13n0, where no is the number of particles per milliliter of solution and the constant 18 takes into account the actual size of the scattering volume, which relates no to No, as well as the geometrical and optical characteristics of the scattering spectrometer (5). Although 13 can be calculated, it is best to determine it experimentally using samples of particles for which the concentration can be determined directly by dry weight. We determined 13 using samples of 176-nm polystyrene latex spheres. The experimental BIA ratio, combined with the known latex sphere concentration, yields 13. Once the constant 1S was known, virus concentrations were obtained by serially diluting a virus solution until the number fluctuations in the scattering volume contributed sufficiently to the autocorrelation function, so as to give a measurable ratio of B/A. Figure 1 illustrates these considerations with data obtained on MuLV. The proportionality between B/A and concentration is shown in Fig. 2.
RESULTS Reverse transcriptase data for one set of assays are shown in Table 1. The enzyme activity per virion, expressed as picomoles of dGTP incorporated/virion per hour, was determined from the slope of the enzyme activity measured as a function of virus concentration, as shown in Fig. 3. The error bars in activity are the standard deviations calculated from data such
CORRELATOR OUTPUT (Arbitrary Units) ,60
a
'40 120
80
60 40-
20-
0
1
3
2
TIME (msec) 44
45
AMV AND MuLV REVERSE TRANSCRIPTASE ACTIVITY
VOL. 18, 1976
that in Table 1. The error bars in particle verse transcriptase activity per virion for AMV concentration are intrinsic to the light scatter- is (28.1 4.2) x 10-7 pmol of dGTP incorpoing measurement (8) and are ca. ±15%. The rated/virion per h, and for MuLV it is (1.1 straight line through the data in Fig. 3 is the 0.2) x 10-7 pmol of dGTP incorporated/virion average of two least squares straight lines, one per h. assuming no experimental uncertainties in the DISCUSSION concentrations and the other assuming no experimental uncertainties in the activities. The The reverse transcriptase activity per virion slope of this average straight line is taken as for AMV is comparable to that which can be the estimate of the enzyme activity per vision. calculated from data, reported by Fine et al. (4), The estimated experimental uncertainty in this for an assay of murine mammary tumor virus slope is ca. 15%. From these plots, the re- (MuMTV) in tissue culture fluid. These workers used poly(rG) oligo(dG), a reverse transcripMuLV (Rauscher) tase assay similar to that described here, and 20 electron microscopy for particle counting. The reverse activities per virion for four independent assays, calculated from Table 1 of Fine et 15 al. (4), are 29 x 10-7, 34 x 10-7, 44 x 10-7 and 75 x 10-7 pmol of dGTP incorporated/virion per h. 4~ The differences among the activities of the 10 reverse transcriptase assays for AMV, MuLV, and MuMTV could be caused by different num-
as
±
±
±
,
bers of enzyme molecules per virion,
5
inherently
different enzyme activities, different amounts of unknown inhibitors which co-purify with the virions, inactivation of enzymatic activity | | | z z| caused by virus purification, or by exposure to .8 .6 .4 I.O .2 o high temperatures (e.g., 39 C in the bird) or RELATIVE CONCENTRATION optimization conditions for the enzyme activity FIG. 2. Plot of B/A as a function of dilution of which have not been met. As for any enzyme MuLV. T'he B/A values were measured from the ex- assay, it is impossible to prove that these assays have been absolutely optimized. However, --perimentc21 correlation Ifunctions shown in Fig. 1. --
Virus
AMV
---
No. of virions in sample
MuLV
dGTP incorporated (p
VPCC
67,847
55,754
50,556
58,052
±8,872
2.9 ± 0.4
0.095
6.5)
237,772
235,476
228,059
236,624
+1,624
12.0 + 0.10
0.38
0.2)
22,402
25,588
18,434
22,141
+3,584
1.5 + 0.3
24
(15.6 + 2.0) x 107
237,490
274,722
215,206
242,473
+30,067
2.0
240
(43.4
+
105 +
X
105
(1.6
+
x 107
MuLV
(AL)
1.6)
(10.8 X
AMV
TABLE 1. Typical data for reverse transcriptase assay dpm deviation Avg Sample 1 Sample 2 Sample 3
16.8
+
Virus concentration was determined by laser beat frequency spectroscopy. b Specific activity, 8.90 Ci/mmol for AMV assay and 6.51 Ci/mmol for MuLV assay. c VPC, volume of viral plasma which would contain the amount of purified virus used in the assay assuming no particle loss during preparation. a
FIG. 1. Experimentally observed autocorrelation function for MuLV (Rauscher) at three concentrations showing the progressive appearance of number fluctuations in the scattering volume. The solid dots lie on a calculated exponential with a time constant determined from high concentration data. (a) High concentration sample where the correlator output is dominated entirely by the B component. (b) Dilution of the MuLV sample in (a) where BIA = 6.5. This corresponds to a concentration of ca. 3.6 x 108 particleslml. (c) Further dilution of the MuLV sample where B/A = 4 and corresponds to ca. 2 x 108 particleslml.
46
J. VIROL.
LIEBES ET AL. 15
a-
A
AMV
10 I
0 I-
E
0.
5
10
20
40
30
50
PARTICLE NUMBER (xIO 5) 20
'5
a. e0
0
0
E 0.
5
5
go
15
20
PARTICLE NUMBER (x 107) FIG. 3. Plot of the picomoles of dGTP incorporated as a function of virus particle number for AMV (A) and MuLV (B). The error bars are discussed in the text. The dashed line in B represents the AMV data plotted for comparison on the same scale with the MuLV data.
with respect to these arbitrarily standardized conditions, the activity of the AMV assay is ca. 24 times greater than that of MuLV. Because of this difference, the reverse transcriptase assays can be very misleading when used to compare virus concentrations of different virus species. These data can be used to estimate the sensitivity of the reverse transcriptase assays. The sensitivity limit is determined by the level of
incorporated radioactivity measurable above background. The background, which was determined by adding lysed virus preparation to the assay mixture after the addition of trichloroacetic acid, is No 2,800 dpm. If we accept No/2 = 1,400 as the minimum significant number of dpm above background (which is substantially larger than the statistical criterion of 2 No = =
75), then we find a sensitivity of 2.5 x 104
VOL. 18, 1976
AMV AND MuLV REVERSE TRANSCRIPTASE ACTIVITY
TABLE 2. Reverse transcriptase activitylvirion Virus
AMV MuLV MuMTVb
dGTP incorporated virion/h) (pmol/
(28.1 ± 4.2) (1.1 + 0.2) 29 34 44 75
x x x x x x
10-7 10-7 10-7 10-7 10-7
Sensitivity (vi-
rions)a
3.4 x 88 x 6.7 x 5.6 x 4.3 x 2.6 x
104 104 104b
104 104 10-7 104 a Normalized for comparison purposes to dGTP with specific activity of 6.51 Ci/mmol (See text). Minimum detectable signal above background of 1,400 dpm was assumed. b Calculated from Table 1 of reference 4.
virions for the AMV assay (with dGTP specific activity of 8.9 Ci/mmol) and 88 x 104 virions for the MuLV assay (with dGTP specific activity of 6.51 Ci/mmol). Assuming the same counting background, the sensitivity of the assay for MuMTV reported by Fine et al. (4) ranges from 45 x 104 to 116 x 104 virions with dGTP specific activity of 0.377 Ci/mmol. For purposes of comparison, these sensitivities are normalized to dGTP specific activity of 6.51 Ci/mmol and tabulated in Table 2. Our reverse transcriptase assays for AMV and MuLV and that for MuMTV derived from the data of Fine et al. (4) are four orders of magnitude greater than those stated by Ringold et al. (7) (See Table 3 of reference 7) for B77 strain of avian sarcoma virus (1.6 x 108 virions) and MuMTV virus (1.4 x 108 virions). Ringold et al. used poly(rA).oligo(dT) as the synthetic template, "since it confers greater sensitivity to the assay than does the more specific poly(rC)oligo(dG)" (7). The reason for the relative insensitivity reported by Ringold et al. is not readily apparent. One obvious possibility is that we and Fine et al. (4) determined the virus concentrations by direct methods, whereas Ringold et al. (7) relied on an indirect method which required knowledge of the number of RNA molecules per virion. In any event, the agreement between our data and that of Fine et al. (4) suggests that the sensitivity estimate of Ringold et al. (7) is in error. ACKNOWLEDGMENTS This work was supported by Public Health Service contracts N01-CP-33226 and N01-CP-33347 and grants CA
47
14680 and CA 14100 from the National Cancer Institute, as well as by an institutional grant from the United Foundation of Greater Detroit. We thank Ernest Retzel for technical assistance with part of this work. LITERATURE CITED 1. Baltimore, D. 1970. RNA-dependent DNA polymerase in virions of RNA tumor viruses. Nature (London) 226:1209-1211. 2. Dion, A. S., A. B. Vaidya, and S. F. Garland. 1974. Cation preferences for poly(rC)-oligo(dG)-directed DNA syntheses by RNA tumor viruses and human milk particulates. Cancer Res. 34:3509-3515. 3. Filbert, J. E., R. M. McAllister, M. 0. Nicolson, and R. V. Gilden. 1974. RD 114 virus infectivity assay by measurements of DNA polymerase activity and virus group specific antigen. Proc. Soc. Exp. Biol. Med. 145:366-370. 4. Fine, D. L., L. 0. Arthur, J. K. Plowman, E. A. Hillman, and F. Klein. 1974. In vitro system for production of mouse mammary tumor virus. Apple. Microbiol. 28:1040-1046. 5. Koppel, D. W. 1971. Analysis of gaussian light by clipped photocount autocorrelation: the effect of finite sampling times and incomplete spatial coherence. J. Appl. Phys. 42:3216-3225. 6. McCormick, J. J., L. J. Larson, and M. A. Rich. 1974. RNase inhibition of reverse transcriptase activity in human milk. Nature (London) 251:737-740. 7. Ringold, G., E. Y. Lasfragues, J. M. Bishor, and H. E. Varmus. 1975. Production of mouse mammary tumor virus by cultured cells in the absence and presence of hormones: assay by molecular hybridization. Virology 65:135-147. 8. Salmeen, I., L. Rimai, L. Liebes, M. A. Rich, and J. J. McCormick. 1975. Hydrodynamic diameters of RNA tumor viruses: studies by laser beat frequency light scattering spectroscopy of avian myeloblastosis and Rauscher murine leukemia viruses. Biochemistry 14:134-141. 9. Sarngadharan, M. G., P. S. Sarin, and R. C. Gallo. 1972. Reverse transcriptase activity of human acute leukemic cells: purification of the enzyme, response to AMV 70S RNA, and characterization of the DNA products. Nature (London) New Biol. 240:67-72. 10. Schaefer, D. W., and B. J. Berne. 1972. Light scattering from nongaussian concentration fluctuations. Phys. Rev. Lett. 28:475-477. 11. Temin, H. M., and S. Mizutani. 1970. RNA-dependent DNA polymerase in virions of Rous sarcoma virus. Nature (London) 226:1211-1213. 12. Verma, I. and D. Baltimore. 1973. Purification of the RNA-directed DNA polymerase from avian myeloblastosis virus and its assay with polynucleotide templates, p. 125-130. In K. Moldave and L. Grossman (ed.), Methods in enzymology, vol. 29. Academic Press Inc., New York. 13. Waters, L. C., and W. Yang. 1974. Comparative biochemical properties of RNA-directed DNA polymerase from Rauscher murine leukemia virus and avian myeloblastosis virus. Cancer Res. 34:2585-2593.
