Eur. J. Biochem. 53, 343-348 (1975)
Structure and Synthesis of a Lipid-Containing Bacteriophage The Molecular Weight and Other Physical Properties of Bacteriophage P M 2 Rafael D. CAMERINI-OTERO and Richard M . FRANKLIN The Public Health Research Institute of the City of New York (Received November 7, 1974/January 16, 1975)
Several physical and chemical parameters of bacteriophage PM2 have been measured. The sedimentation constant was determined to be = 293 S. The buoyant density in sucrose at 20 "C was 1.24 g cm-3 and in CsCl at 25 "C was 1.29 g cmP3.The high-speed equilibrium centrifugation method of Yphantis (1964) was used to measure the molecular weight of PM2. The necessary auxiliary parameters were also determined. A value of 0,771 0.005 cm3 8-l for the apparent specific volume at constant chemical potential in 1 M sodium chloride has been obtained by pycnometry; the viral concentration was determined using the absorption coefficient at 260 nm (4.60 ? 0.10 cm2 mg-I), which in turn was calculated from the phosphorous content of the virus (17.89 k 0.28 pg of P per mg dry weight of virus). The molecular weight of PM2 determined with these parameters is (44.1 & 1.2 x lo6). From the phosphorous content of the virus, the percentage of phosphorus known to be in its DNA (Camerini-Otero and Franklin, 1972), and the molecular weight of the bacteriophage, we have calculated a molecular weight for PM2 DNA of 6.26 x lo6, which confirms values determined using empirical relationships.
The lipid-containing bacteriophage PM 2 has been used as a model for studying the structure of membranes and the interactions between proteins and lipids 11 -41. Crucial to these investigations is a knowledge of the viral mass in terms of density, specific volume, and molecular weight. In the original report on P M 2 by Espejo and Canelo [5] several physical and chemical properties were described. The density of the virus in cesium chloride was reported and a rough estimate of the molecular weight was made. No data was presented on the specific volume. Since large quantities [6 - 81 of chemically homogeneous virus, free of contaminating host proteins 17-91 and host antigens [lo] can now be prepared, we have undertaken a more extensive study of these physical properties. Central to the problem of determining the apparent specific volume, and therefore the molecular weight of a macromolecule, is the determination of macromolecular concentration. We have used the absorption coefficient at 260 nm to measure the concentration of PM2. Althg-u~h Espejo and Canelo [ 5 ] reported a value for A?:? 'at 260 nm, their data did not include an estimate of the error and therefore we have reThis is paper number XVII in the series.
Eur. J. Biochem. 53 (1975)
determined this coefficient and analyzed its uncertainty. Since an appreciable quantity of phosphorus is present in the phospholipids of PM 2 [2,5], it was not possible to obtain an absorption coefficient assuming that all of the phosphorus is derived from its DNA (cf [ll]). Unfortunately P M 2 is not stable in low ionic strength solutions and exhibits spectral changes upon disruption (see below) and therefore it is difficult to dialyze the virus against water, record its spectrum and then perform a dry weight measurement, as was apparently done by Espejo and Canelo [ 5 ] , and obtain an accurate value for A;:&. Rather than deal with large amounts of salt in dried samples, the obvious alternative to the procedure above, we have determined the absorption coefficient from the ratio of the phosphorous content of the virus, dried after extensive dialysis against water, to the phosphorous content per absorbance unit at 260 nm.
MATERIALS AND METHODS
Preparation of Purified Bacteriophage PA42 Bacteriophage PM 2 was grown and purified according to a procedure published elsewhere [6].
344
Purified virus was dialyzed against three changes of 1 M NaCl in either parafilm-sealed cylinders or stoppered bottles, the last change for at least 48 h ; the virus was stored in this solvent at 4 "C. The final outer dialysate was always saved for use as a blank. Plaque assays were carried out as previously published [12]. Due to the relative instability of the virus in the solvent used here [6], all experiments were performed from a week to 10 days after the final purification step. Preparation of Labeled Virus
Virus double-labeled with [32P]phosphoric acid and a mixture of 3H-labeled amino acids was prepared as previously described [9]. The 3H-labeled L-amino acid mixture and [32P]phosphoric acid were obtained from New England Nuclear Corp. (Boston, Mass.). Spectrophotometric Measurements
All dilutions for spectrophotometric measurements were gravimetric. The measurements themselves were performed on a Cary model 14 recording spectrophotometer, calibrated against the absolute transmittance scale of a Zeiss spectrophotometer; all absorbancies, unless otherwise reported, are without lightscattering corrections. Nitrogen and Phosphorous Content of P M 2
Preparations were dialyzed against six changes of double glass-distilled water. Nitrogen and phosphorous content on a dry weight basis were determined by Schwarzkopf Laboratories (New York). Total phosphorus per A,,, unit was determined according to Ames [I31 except that measurements of phosphorus were made at 660 nm. Isopycnic Centrifugation
Sucrose gradients were prepared by layering 2.2 ml of 30% (w/w) sucrose in 0.05 M Tris (pH 7.4) on top of 2.2 ml of 6 0 x , (w/w) sucrose in the same buffer, and allowing the solutions to equilibrate for 20 h at room temperature (23 "C); 10 to 50 PI of doublelabeled virus were layered on top of these gradients and the tubes were then filled with mineral oil. The gradients were run in a SW65 rotor (36000 rev./min, 20 T).The gradients were collected into 45 fractions, aliquots of each fraction were used for determining radioactivity and refractive index; the density of each fraction at 20 "C was obtained from the International Critical Tables.
Physical Properties of PM 2
Cesium chloride solutions were prepared with 0.420 g CsCl per g of 0.05 M Tris (pH 7.4), 50- 100 yl of double-labeled virus were layered on top of 5 ml of this solution. The solutions were run in an SW65 rotor at 50000 rev./min for 20 h at 25 "C and collected in 30 to 50 fractions; aliquots were taken for measuring the radioactivity and refractive index of each fraction. The density of the fractions was calculated from the empirical relation between n;5 and eZ5 [14]. A Bellingham and Stanley high-accuracy Abbe 60 refractometer, equipped with a constant-temperature circulating water bath, was used for the determination of n;' and nAo. Determination of'the Apparent Specific Volume
Apparent specific volumes using the equation:
(4') were
calculated
(es - e,>k = 1 - d'eo (1) where ,o,) is the density of the outer dialysate, eS is the density of the virus solution, and c is the solute concentration. Both Q, and e, were determined at 25 "C f 0.05 "C in a 5.2-cm3 Lipkin pycnometer [15]. For each determination of 4', three separate series of measurements were made. In determining the concentration of the virus solutions, corrections were applied for the difference between solution and solvent densities to calculate the gravimetric dilutions. Weights were known to within f 0.02mg and volumes to within &- 0.0002 cm3. Sedimentation Velocity Measurements
Virus solutions in 12-mm double-sector cells with epon-filled centerpieces and sapphire windows were run in an An-D rotor in a Beckman model E ultracentrifuge equipped with ultraviolet optics, photoelectric scanner, electronic speed control, and temperature control unit. The temperature of the runs was regulated at close to 25 "C and corrected to standard conditions, and the speed was close to 13000 rev./min and checked periodically using the revolution counter. The motion of the 50 %, concentration point in the boundary region was measured to determine the sedimentation velocity. Data reduction was performed on a General Electric Mark I1 timesharing system, using a Fortran IV program written by us. Equilibrium Sedimentation
Equilibrium sedimentation was carried out according to the high-speed method of Yphantis [16]. The virus solutions, ranging in initial concentration Eur. J. Biochem. 53 (1975)
345
R. D. Camerini-Otero and R. M. Franklin
from 20 to 50 pg cmP3were centrifuged to equilibrium in an An-J rotor. The temperature was held at 20 "C and the rotor speed, either 800 or 861 rev./min, was accurately determined by means of the revolution counter. For bacteriophage PM2 these rotor speeds correspond to effective reduced molecular weights, 0,of 2.9 and 3.4 cm-, [16]. The distributions in the analytic cell, recorded at 265 nm, were independent of time over the last 48 h of centrifugation. A 1-cm deflection on the scan recording corresponded to 0.2 absorbance units ( A units) ; the estimated uncertainty in reading these scans was k 0.0035 A units. The weight-average molecular weight at radius r , Mw(r), was calculated from the slope of a leastsquare straight line through six adjacent points in a plot of In A,,, versus r-'; overlapping regions with four common points were evaluated throughout the concentration distribution at absorbancies from 0.05 to 1.0. In the range of concentrations used in the experiments with PM 2, M J r ) was independent of the virus concentration. M , was calculated from the slope of a least-squares straight line through all the points tabulated for a cell sector. Error- Analysis Whenever possible, the uncertainty, ax,associated with the mean value, 2,of a set of n observations x ihas been estimated as both the standard error of the mean and the 95% confidence interval for X according to the Student's t-distribution (cf. [17]). The uncertainty in a derived quantity of M , a function of experimental uncorrelated variables, x, y , . . . , was estimated from the uncertainties 6x,6y, . . . , according to the law of propagation of errors (c$ [17]) such that: ( 6 M ) 2 = (aM/ax)z
(ax),
+ (aM/ay)2 (6y)2 +
..
RESULTS
Some General Properties o f the Virus Representative ultraviolet absorption spectra of intact PM 2 bacteriophage and PM2 disrupted by 0.05 "/, Sarkosyl [5], an anionic detergent, are shown in Fig. 1. The spectra of whole virus solutions are characterized by a minimum at 244 nm and a maximum at 259 nm; the ratio of the absorbance at 259 nm to that at 280 nm is 1.43, that at 259 nm to that at 244 nm is 1.14. Disruption of the virion by Sarkosyl is accompanied by a lowering of the absorbance (see Fig.1 and cJ [5]). The nitrogen and phosphorous content per mg dry weight of PM 2, as well as the phosphorous content per absorbance unit at 260nm and the absorption Eur. J. Biochem. 53 (1975)
0.4 0.3 a,
c
8 0.2
D
D
Q
0.1
0 I
260
240
280
300
320 Waveiength (nm)
I
I
340
360
Fig. 1. Ultraviolet ahsorption spectrum qf hucteriophuge P M 2 I M NaCl, 0.05 M Tris, pH 7.4 at 23 ' C i - ----) ; same values 260, 270 und 280 nrn corrected for light scattering I@)I( P M 2 23 ,,C in 1 M NuCl. 0.05 M Tris, p H 7.4, 0.052, Suikosyl (-) Baseline (- . .-)
in at at ;
~
Table 1. Determinution qf' the ahsorption coefficient at 260 nm for a I mgjcm3 solution of virus Nitrogen and phosphorus were determined on the basis of dry weight of virus. 1 A260 unit = the amount of material contained in 1 rnl of a solution which has an absorbance of 1 at 260 nm, when measured in a 1 cm pathlength cell. Results are quoted & standard error of the mean; values in parentheses are the corresponding 95 yo confidence intervals Quantity
~
N P P A 260 p. 1 ",,
Value
Unit
Number of observations
pg/mg virus pg/mgvirus pg/A,,o unit A,,,unit/mgvirus
8 6 4 -
~~~~~~~~~
1.7 (4.0 ) 138.2 17.89 k 0.28 (0.72) 3.892 k 0.062 (0.20) 4.60 f 0.10 (0.30)
coefficient at 260 nm, both latter values uncorrected for light-scattering, are given in Table 1. Based on this absorption coefficient, the specific infectivity of our purified virus preparations is in the range of 6 to 7 x 10l2 p.f.u./mg. Sedimentation CoefJicient of P M 2 The sedimentation coefficient of P M 2 has been measured in 1 M NaCl at 25 "C at a pH of 6.3 to 6.6. The values of s, measured in three separate runs, were : 261.1, 261.7 and 262.5 S. The average value is 261.8 f 0.4 (1.7) S, where the quoted uncertainty is the standard error of the mean and the value in parentheses is the corresponding 95 confidence interval. Since all these runs were carried out at low virus concentration (
Structure and Synthesis of a Lipid-Containing Bacteriophage The Molecular Weight and Other Physical Properties of Bacteriophage P M 2 Rafael D. CAMERINI-OTERO and Richard M . FRANKLIN The Public Health Research Institute of the City of New York (Received November 7, 1974/January 16, 1975)
Several physical and chemical parameters of bacteriophage PM2 have been measured. The sedimentation constant was determined to be = 293 S. The buoyant density in sucrose at 20 "C was 1.24 g cm-3 and in CsCl at 25 "C was 1.29 g cmP3.The high-speed equilibrium centrifugation method of Yphantis (1964) was used to measure the molecular weight of PM2. The necessary auxiliary parameters were also determined. A value of 0,771 0.005 cm3 8-l for the apparent specific volume at constant chemical potential in 1 M sodium chloride has been obtained by pycnometry; the viral concentration was determined using the absorption coefficient at 260 nm (4.60 ? 0.10 cm2 mg-I), which in turn was calculated from the phosphorous content of the virus (17.89 k 0.28 pg of P per mg dry weight of virus). The molecular weight of PM2 determined with these parameters is (44.1 & 1.2 x lo6). From the phosphorous content of the virus, the percentage of phosphorus known to be in its DNA (Camerini-Otero and Franklin, 1972), and the molecular weight of the bacteriophage, we have calculated a molecular weight for PM2 DNA of 6.26 x lo6, which confirms values determined using empirical relationships.
The lipid-containing bacteriophage PM 2 has been used as a model for studying the structure of membranes and the interactions between proteins and lipids 11 -41. Crucial to these investigations is a knowledge of the viral mass in terms of density, specific volume, and molecular weight. In the original report on P M 2 by Espejo and Canelo [5] several physical and chemical properties were described. The density of the virus in cesium chloride was reported and a rough estimate of the molecular weight was made. No data was presented on the specific volume. Since large quantities [6 - 81 of chemically homogeneous virus, free of contaminating host proteins 17-91 and host antigens [lo] can now be prepared, we have undertaken a more extensive study of these physical properties. Central to the problem of determining the apparent specific volume, and therefore the molecular weight of a macromolecule, is the determination of macromolecular concentration. We have used the absorption coefficient at 260 nm to measure the concentration of PM2. Althg-u~h Espejo and Canelo [ 5 ] reported a value for A?:? 'at 260 nm, their data did not include an estimate of the error and therefore we have reThis is paper number XVII in the series.
Eur. J. Biochem. 53 (1975)
determined this coefficient and analyzed its uncertainty. Since an appreciable quantity of phosphorus is present in the phospholipids of PM 2 [2,5], it was not possible to obtain an absorption coefficient assuming that all of the phosphorus is derived from its DNA (cf [ll]). Unfortunately P M 2 is not stable in low ionic strength solutions and exhibits spectral changes upon disruption (see below) and therefore it is difficult to dialyze the virus against water, record its spectrum and then perform a dry weight measurement, as was apparently done by Espejo and Canelo [ 5 ] , and obtain an accurate value for A;:&. Rather than deal with large amounts of salt in dried samples, the obvious alternative to the procedure above, we have determined the absorption coefficient from the ratio of the phosphorous content of the virus, dried after extensive dialysis against water, to the phosphorous content per absorbance unit at 260 nm.
MATERIALS AND METHODS
Preparation of Purified Bacteriophage PA42 Bacteriophage PM 2 was grown and purified according to a procedure published elsewhere [6].
344
Purified virus was dialyzed against three changes of 1 M NaCl in either parafilm-sealed cylinders or stoppered bottles, the last change for at least 48 h ; the virus was stored in this solvent at 4 "C. The final outer dialysate was always saved for use as a blank. Plaque assays were carried out as previously published [12]. Due to the relative instability of the virus in the solvent used here [6], all experiments were performed from a week to 10 days after the final purification step. Preparation of Labeled Virus
Virus double-labeled with [32P]phosphoric acid and a mixture of 3H-labeled amino acids was prepared as previously described [9]. The 3H-labeled L-amino acid mixture and [32P]phosphoric acid were obtained from New England Nuclear Corp. (Boston, Mass.). Spectrophotometric Measurements
All dilutions for spectrophotometric measurements were gravimetric. The measurements themselves were performed on a Cary model 14 recording spectrophotometer, calibrated against the absolute transmittance scale of a Zeiss spectrophotometer; all absorbancies, unless otherwise reported, are without lightscattering corrections. Nitrogen and Phosphorous Content of P M 2
Preparations were dialyzed against six changes of double glass-distilled water. Nitrogen and phosphorous content on a dry weight basis were determined by Schwarzkopf Laboratories (New York). Total phosphorus per A,,, unit was determined according to Ames [I31 except that measurements of phosphorus were made at 660 nm. Isopycnic Centrifugation
Sucrose gradients were prepared by layering 2.2 ml of 30% (w/w) sucrose in 0.05 M Tris (pH 7.4) on top of 2.2 ml of 6 0 x , (w/w) sucrose in the same buffer, and allowing the solutions to equilibrate for 20 h at room temperature (23 "C); 10 to 50 PI of doublelabeled virus were layered on top of these gradients and the tubes were then filled with mineral oil. The gradients were run in a SW65 rotor (36000 rev./min, 20 T).The gradients were collected into 45 fractions, aliquots of each fraction were used for determining radioactivity and refractive index; the density of each fraction at 20 "C was obtained from the International Critical Tables.
Physical Properties of PM 2
Cesium chloride solutions were prepared with 0.420 g CsCl per g of 0.05 M Tris (pH 7.4), 50- 100 yl of double-labeled virus were layered on top of 5 ml of this solution. The solutions were run in an SW65 rotor at 50000 rev./min for 20 h at 25 "C and collected in 30 to 50 fractions; aliquots were taken for measuring the radioactivity and refractive index of each fraction. The density of the fractions was calculated from the empirical relation between n;5 and eZ5 [14]. A Bellingham and Stanley high-accuracy Abbe 60 refractometer, equipped with a constant-temperature circulating water bath, was used for the determination of n;' and nAo. Determination of'the Apparent Specific Volume
Apparent specific volumes using the equation:
(4') were
calculated
(es - e,>k = 1 - d'eo (1) where ,o,) is the density of the outer dialysate, eS is the density of the virus solution, and c is the solute concentration. Both Q, and e, were determined at 25 "C f 0.05 "C in a 5.2-cm3 Lipkin pycnometer [15]. For each determination of 4', three separate series of measurements were made. In determining the concentration of the virus solutions, corrections were applied for the difference between solution and solvent densities to calculate the gravimetric dilutions. Weights were known to within f 0.02mg and volumes to within &- 0.0002 cm3. Sedimentation Velocity Measurements
Virus solutions in 12-mm double-sector cells with epon-filled centerpieces and sapphire windows were run in an An-D rotor in a Beckman model E ultracentrifuge equipped with ultraviolet optics, photoelectric scanner, electronic speed control, and temperature control unit. The temperature of the runs was regulated at close to 25 "C and corrected to standard conditions, and the speed was close to 13000 rev./min and checked periodically using the revolution counter. The motion of the 50 %, concentration point in the boundary region was measured to determine the sedimentation velocity. Data reduction was performed on a General Electric Mark I1 timesharing system, using a Fortran IV program written by us. Equilibrium Sedimentation
Equilibrium sedimentation was carried out according to the high-speed method of Yphantis [16]. The virus solutions, ranging in initial concentration Eur. J. Biochem. 53 (1975)
345
R. D. Camerini-Otero and R. M. Franklin
from 20 to 50 pg cmP3were centrifuged to equilibrium in an An-J rotor. The temperature was held at 20 "C and the rotor speed, either 800 or 861 rev./min, was accurately determined by means of the revolution counter. For bacteriophage PM2 these rotor speeds correspond to effective reduced molecular weights, 0,of 2.9 and 3.4 cm-, [16]. The distributions in the analytic cell, recorded at 265 nm, were independent of time over the last 48 h of centrifugation. A 1-cm deflection on the scan recording corresponded to 0.2 absorbance units ( A units) ; the estimated uncertainty in reading these scans was k 0.0035 A units. The weight-average molecular weight at radius r , Mw(r), was calculated from the slope of a leastsquare straight line through six adjacent points in a plot of In A,,, versus r-'; overlapping regions with four common points were evaluated throughout the concentration distribution at absorbancies from 0.05 to 1.0. In the range of concentrations used in the experiments with PM 2, M J r ) was independent of the virus concentration. M , was calculated from the slope of a least-squares straight line through all the points tabulated for a cell sector. Error- Analysis Whenever possible, the uncertainty, ax,associated with the mean value, 2,of a set of n observations x ihas been estimated as both the standard error of the mean and the 95% confidence interval for X according to the Student's t-distribution (cf. [17]). The uncertainty in a derived quantity of M , a function of experimental uncorrelated variables, x, y , . . . , was estimated from the uncertainties 6x,6y, . . . , according to the law of propagation of errors (c$ [17]) such that: ( 6 M ) 2 = (aM/ax)z
(ax),
+ (aM/ay)2 (6y)2 +
..
RESULTS
Some General Properties o f the Virus Representative ultraviolet absorption spectra of intact PM 2 bacteriophage and PM2 disrupted by 0.05 "/, Sarkosyl [5], an anionic detergent, are shown in Fig. 1. The spectra of whole virus solutions are characterized by a minimum at 244 nm and a maximum at 259 nm; the ratio of the absorbance at 259 nm to that at 280 nm is 1.43, that at 259 nm to that at 244 nm is 1.14. Disruption of the virion by Sarkosyl is accompanied by a lowering of the absorbance (see Fig.1 and cJ [5]). The nitrogen and phosphorous content per mg dry weight of PM 2, as well as the phosphorous content per absorbance unit at 260nm and the absorption Eur. J. Biochem. 53 (1975)
0.4 0.3 a,
c
8 0.2
D
D
Q
0.1
0 I
260
240
280
300
320 Waveiength (nm)
I
I
340
360
Fig. 1. Ultraviolet ahsorption spectrum qf hucteriophuge P M 2 I M NaCl, 0.05 M Tris, pH 7.4 at 23 ' C i - ----) ; same values 260, 270 und 280 nrn corrected for light scattering I@)I( P M 2 23 ,,C in 1 M NuCl. 0.05 M Tris, p H 7.4, 0.052, Suikosyl (-) Baseline (- . .-)
in at at ;
~
Table 1. Determinution qf' the ahsorption coefficient at 260 nm for a I mgjcm3 solution of virus Nitrogen and phosphorus were determined on the basis of dry weight of virus. 1 A260 unit = the amount of material contained in 1 rnl of a solution which has an absorbance of 1 at 260 nm, when measured in a 1 cm pathlength cell. Results are quoted & standard error of the mean; values in parentheses are the corresponding 95 yo confidence intervals Quantity
~
N P P A 260 p. 1 ",,
Value
Unit
Number of observations
pg/mg virus pg/mgvirus pg/A,,o unit A,,,unit/mgvirus
8 6 4 -
~~~~~~~~~
1.7 (4.0 ) 138.2 17.89 k 0.28 (0.72) 3.892 k 0.062 (0.20) 4.60 f 0.10 (0.30)
coefficient at 260 nm, both latter values uncorrected for light-scattering, are given in Table 1. Based on this absorption coefficient, the specific infectivity of our purified virus preparations is in the range of 6 to 7 x 10l2 p.f.u./mg. Sedimentation CoefJicient of P M 2 The sedimentation coefficient of P M 2 has been measured in 1 M NaCl at 25 "C at a pH of 6.3 to 6.6. The values of s, measured in three separate runs, were : 261.1, 261.7 and 262.5 S. The average value is 261.8 f 0.4 (1.7) S, where the quoted uncertainty is the standard error of the mean and the value in parentheses is the corresponding 95 confidence interval. Since all these runs were carried out at low virus concentration (
