JOURNAL OF VIROLOGY, Dec. 1978, p. 917-928 0022-538X/78/0028-0917$02.00/0 Copyright X 1978 American Society for Microbiology
Vol. 28, No. 3 Printed in U.S.A.
Electrophoresis of Bacteriophage T7 and T7 Capsids in Agarose Gels PHILIP SERWER* AND MARY E. PICHLER Department of Biochemistry, The University of Texas Health Science Center, San Antonio, Texas 78284 Received for publication 12 June 1978
Agarose gel electrophoresis of the following was performed in 0.05 M sodium phosphate-0.001 M MgCl2 (pH 7.4): (i) bacteriophage T7; (ii) a T7 precursor capsid (capsid I), isolated from T7-infected Escherichia coli, which has a thicker and less angular envelope than bacteriophage T7; (iii) a second capsid (capsid II), isolated from T7-infected E. coli, which has a bacteriophage-like envelope; and (iv) capsids (capsid IV) produced by temperature shock of bacteriophage T7. Bacteriophage T7 and all of the above capsids migrated towards the anode. In a 0.9% agarose gel, capsid I had an electrophoretic mobility of 9.1 ± 0.4 x 10-5 cm2/V.s; bacteriophage T7 migrated 0.31 ± 0.02 times as fast as capsid I. The mobilities of different preparations of capsid II varied in such gels: the fastestmigrating capsid II preparation was 0.51 ± 0.03 times as fast as capsid I and the slowest was 0.37 ± 0.02 times as fast as capsid I. Capsid IV with and without the phage tail migrated 0.29 ± 0.02 and 0.42 ± 0.02 times as fast as capsid I. The results of the extrapolation of bacteriophage and capsid mobilities to 0% agarose concentration indicated that the above differences in mobility are caused by differences in average surface charge density. To increase the accuracy of mobility comparisons and to increase the number of samples that could be simultaneously analyzed, multisample horizontal slab gels were used. Treatment with the ionic detergent sodium dodecyl sulfate converted capsid I to a capsid that migrated in the capsid II region during electrophoresis through agarose gels. In the electron microscope, most of the envelopes of these latter capsids resembled the capsid II envelope, but some envelope regions were thicker than the capsid II envelope.
Current evidence indicates that DNA enters perature-shocked bacteriophage T7; capsid IVA a preassembled capsid (capsid I) during the in has no tail, and capsid IVB has a tail (11). To determine the mechanism by which bacvivo assembly of bacteriophage T7 (8, 11, 14). In the electron microscope the envelopes of capsid teriophage T7 DNA is packaged, the following I and bacteriophage T7 both appear roughly data are useful: (i) physical characterization of spherical, but the capsid I envelope appears less bacteriophage T7 and T7 capsids; (ii) measureangular and thicker than the bacteriophage en- ments of the amounts and time course of T7 velope (11). This greater thickness is probably capsids in cells infected with bacteriophage T7 caused by the presence of larger amounts of P9 and T7 mutants; (iii) induction in a cell-free (T7 proteins are referred to by P followed by the system of the events occurring during DNA protein's gene number [14]) in the capsid I en- packaging. In the work described here, agarose velope than in the bacteriophage envelope (11). gel electrophoresis of bacteriophage T7 and T7 A capsid with an envelope that has a phage-like capsids was developed as a technique to help protein composition, angularity, and thickness obtain these data. Implications of the results in (capsid II) has also been isolated from T7-in- understanding the mechanism of T7 DNA packfected cells; capsid I may convert to capsid II aging are discussed. during an abortive or incomplete DNA packaging attempt (11, 14). Capsids I and II and bacMATERIALS AND METHODS teriophage T7 all have an internal, roughly cyand bacterial strains. BacterioBacteriophage lindrical core with an axial hole (11). The capsid phage T7 was received F. W. Studier. The host I and T7 phage cores appear to be attached to for T7 was Escherichia from coli BB/1. the capsid envelope, and DNA may enter capsid Media and buffers. Bacterial cultures were grown I through the axial hole. Capsids with envelopes and infected in M9 medium (11). Tris/Mg buffer is 0.2 resembling the envelope of capsid II in the elec- M NaCl-0.01 M Tris-chloride (pH 7.4)-0.001 M MgCl2; tron microscope have been isolated from tem- for the purification and storage of radiolabeled bacte917
918
SERWER AND PICHLER
J. VIROL.
riophage T7 and capsids, 100 ,ug of boiled gelatin was strengths are described in the figure legends. During included per ml (Tris/Mg/G buffer). Electrophoresis electrophoresis, buffer was circulated between the resbuffer is 0.05 M sodium phosphate (pH 7.4)-0.001 M ervoirs at a rate of 50 to 200 ml/min. At the end of an electrophoresis, the pH values of buffer in both tanks MgCl2. Preparation and purification of protein-radio- of a horizontal slab gel were within 0.1 pH unit of the labeled T7 bacteriophage and capsids. A station- starting pH. The concentrations of bacteriophage T7 ary-phase bacterial culture in M9 medium was diluted samples were determined by optical density (1); the into fresh M9 medium and was grown overnight at concentrations of capsid samples were determined by 300C with aeration until there were 3 x 108 bacteria the dye-binding method of Bradford (2). Detection of bacteriophage and capsids in agaper ml. The culture was infected with bacteriophage T7 (multiplicity of infection, 10), and at 10 min after rose gels. Radiolabeled bacteriophage T7 and capsids infection radioactive amino acids were added (either in cylindrical gels were detected by autoradiography 3H- or 14C-labeled algal hydrolysate or [35S]methionine after slicing the gels in a sagittal plane and drying was used at a final concentration of 20 to 40 ttCi/ml, under vacuum onto chromatography paper. Kodak purchased from Schwarz-Mann or ICN). After spon- Royal X-Omat film was used. Radiolabel in agarose taneous lysis (at 30 to 35 min after infection) the lysate slab gels can also be detected by autoradiography. To detect unlabeled DNA, gels were soaked for 45 was clarified by centrifuging at 10,000 rpm for 10 min, 40C, in a Beckman JA-21 rotor. The lysate was then min in electrophoresis buffer containing 1 Mug of ethidsedimented through a cesium chloride preformed step ium bromide per ml. The gels were then illuminated gradient containing 100 ,tg of gelatin per ml as previ- with UV light and photographed as previously deously described (11). Bacteriophage particles, which scribed (12). The presence of DNA is revealed by the banded at a density of 1.51 g/ml, were collected and ethidium bromide fluorescence enhancement induced dialyzed into Tris/Mg/G buffer. Capsids I and II, by the DNA. DNA packaged inside bacteriophage T7 which together banded in a single peak at 1.28 g/ml, induces ethidium bromide fluorescence enhancement, were collected, dialyzed into Tris/Mg/G buffer, and although the amount is less than induced by unpackthen isolated by sedimentation through an 11.0-ml, 5 aged DNA (12). To detect unlabeled protein, gels were stained in to 25% sucrose gradient supported by a 0.8-ml 62% sucrose shelf, all in Tris/Mg/G buffer (SW41 rotor, 0.05% Coomassie blue-10% acetic acid for 1.5 h at 25,000 rpm, 120 min, 180C). Capsid I traversed 70 to room temperature. The gels were destained by diffu80% of the gradient and traveled 1.36 to 1.40 times as sion into 10% acetic acid at room temperature. Relative amounts of protein in capsid bands were deterfar as capsid II. Preparation of unlabeled bacteriophage T7 mined by quantitative densitometry of the gels with a and T7 capsids. Fifteen- or 30-liter lysates of T7- Helena Quick Scan Densitometer. Changes in gel diinfected E. coli were prepared; T7 phage and capsids mensions during staining were less than 0.5%. Elution from gels. To recover unlabeled capsids were purified as previously described (11), except that (i) the sucrose gradient used in the final step of the from agarose cylindrical gels, the gels were sliced in a capsid purification had Tris/Mg buffer and (ii) the sagittal plane, and one half of the gel was stained with sucrose from this final step was not removed by di- Coomassie blue. Using the stained half of the gel as a template, the capsid regions of the unstained half were alysis. Electrophoresis in agarose gels. Agarose excised. Unlabeled bacteriophage T7 was recovered by (Seakem, ME) was dissolved in glass-distilled water observing light scattering from the bacteriophage band by boiling. Buffer was then added by diluting 10-fold and excising the band region of the gel. Material was into the agarose solution, which was maintained at 80 eluted from gel slices by one of two methods: (i) the to 900C. Cylindrical gels were poured in glass tubes gel slice was incubated with 20 to 50 pl of Tris/Mg (6-mm ID, 8.5 cm long); horizontal slab gels (6 mm buffer for 2 to 3 days at 40C (diffusion elution); or (ii) thick, 15.2 cm in the direction of the electrical field, the gel slice was placed in a dialysis bag with 100 to 13.2 cm wide) were poured in an apparatus previously 500M of electrophoresis buffer, then placed in a chamdescribed (6). Samples in Tris/Mg or Tris/Mg/G ber with two electrodes; after sufficient electrophoresis buffer were diluted into 5 volumes of 2% sucrose-0.005 buffer was added to cover the dialysis tubing, the gel M sodium phosphate (pH 7.4)-0.001 M MgC12-400 ,ug was subjected to electrophoresis (2.9 V/cm) for 2 to 3 of bromophenol blue per ml prior to layering on the h at 4°C (electrophoretic elution). To recover radiolabeled bacteriophage or capsids gels. Samples can also be dialyzed into electrophoresis buffer before electrophoresis, but this usually resulted from cylindrical gels, the gels were sliced into disks in the loss of 30 to 60% of a radiolabeled capsid sample and the disks were subjected to diffusion elution into on the dialysis tubing. The bottoms of cylindrical gels Tris/Mg/G buffer. The fractions containing radiolabel were covered with nylon mesh or dialysis tubing to were determined by diluting portions of the eluates prevent the agarose from falling out of the tube. The into toluene fluor containing Triton X-100. Recoveries tubes were placed in a commercial electrophoresis of radiolabeled capsids I and II from capsid peak apparatus and were filled with buffer; the samples regions of 0.9% agarose gels were 30 to 55%. Sucrose gradient sedimentation. For determiwere layered underneath the buffer. For horizontal slab gels, samples were placed in wells (2 by 6 mm) nation of the sedimentation rates of eluted capsids, capsid samples were layered on 4.8-ml, linear 5 to 25% previously filled with electrophoresis buffer. Electrophoresis was done at room temperature (22 sucrose gradients containing Tris/Mg/G buffer, to 2500); however, the temperature of horizontal slab poured over a 0.35-ml layer of sodium diatrizoate gels using a field strength of 2.1 V/cm was 3 to 50C (density = 1.30 g/ml). The gradients were centrifuged above room temperature. The times and field in a Beckman SW50.1 rotor for 60 min at 28,000 rpm,
ELECTROPHORESIS OF PHAGE T7
VOL. 28, 1978
18°C, and were fractionated by tube puncture. Capsid I sedimented through 70 to 80% of the sucrose gradient. Electron microscopy. Carbon support films for electron microscopy were prepared as previously described (11). A drop of sample was placed on a support film and was incubated for 1 min at room temperature. The film was washed with 4 drops of water followed by 2 drops of 1% sodium methyl phosphotungstate (pH 7.6) (7) (a gift of Glenn Williams). The film was dried with filter paper. To determine the degree of aggregation of bacteriophage T7 or T7 capsids embedded in negative stain, particles separated by 5 nm or less were assumed to be attached to each other. Bacteriophages or capsids scored for their degree of aggregation were selected at random except that all those selected were more than two bacteriophage diameters from the edge of a stain droplet. This restriction was applied because aggregation induced by drying of the specimen appeared to occur most readily at stain droplet edges. SDS-polyacrylamide gel electrophoresis of material in agarose cylindrical gels. A sample was subjected to electrophoresis in an agarose cylindrical gel, 3 mm in diameter. The gel was soaked in 1% sodium dodecyl sulfate (SDS)-0.05 M Tris-chloride (pH 6.8)-1% fi-mercaptoethanol for 15 min at room temperature and was then dropped into a tube of the same buffer maintained at 100°C by boiling. The gel was boiled until it started to become transparent (for 0.9% gels this occurs at 20 to 35 s after the start of boiling). The gel was then chilled on ice, layered across a vertical SDS slab gel similar to those used previously (11), and sealed with agarose (the same percentage used in the tube gel) in the above buffer. A well was made in the agarose for a standard used to identify protein bands. Electrophoresis and staining of the SDS gel were performed as previously described (11). Theoretical considerations. The electrophoretic mobility of a particle, u, is defined as VIE; V is the velocity of the particle in an electric field of strength E. The magnitude of u in buffer without agarose, uo, for a spherical particle depends on (i) the average density of electrical charge per unit of surface area on the particle, a-, (ii) the radius of the particle, r, (iii) the viscosity, 71, of the solution in which the particle is suspended; (iv) the Debye-Huckel constant, k, which is proportional to the square root of the ionic strength (the ionic strength of electrophoresis buffer is calculated to be 0.116). An approximate determination of u° in terms of the above parameters has been obtained (13) (this equation has been previously used in a study of bacteriophage T4 [5]): 0
2arf(kr)
(1)
3iq(l + kr)(1
f(kr) is a function that varies in value between 1.0 and 1.5 as kr varies from 0 to values above 100. An additional parameter of interest is the total number of electron equivalents of charge on a particle, z, calculated by multiplying a by the surface area of the
particle. If, to prevent convection, electrophoresis is done in the presence of a supporting gel, the gel may affect u by (i) forming channels with a pore size small enough
919
to slow particles down because of multiple collisions with gel fibers (sieving); if the channels are narrow enough, migration may be totally inhibited; and (ii) adsorbing particles by chemical interaction. The latter effect will tend to spread out electrophoretic zones and is therefore undesirable. It has been assumed that u° can be obtained by extrapolating u measured in agarose gels of decreasing agarose concentration to an agarose concentration of 0. Values of u are readily measured on horizontal slab gels; E is measured by monitoring the voltage drop across two points on the gel. However, for 0.5% and less concentrated agarose gels it is extremely difficult to stain slab gels without gel breakage; 0.9% slab gels were used for most measurements. Also, bacteriophage T7 and capsid IVB bands, for unknown reasons, were three to five times broader in slab gels than they were in cylindrical gels. Therefore, extrapolation to 0% agarose concentration was performed by subjecting a sample to electrophoresis on agarose cylindrical gels (which are less fragile than slab gels) with agarose percentages of 0.3, 0.5, 0.9, 1.5, 2.0, and 2.5% (all gels were run simultaneously in the same apparatus). The distance migrated by the sample in each gel, divided by the distance migrated in the 0.9% gel (referred to as c), was extrapolated to 0% agarose (referred to as c°), and uo was calculated:
u0 u0.c =
(2)
Determinations of c were made at E values of 5.8 V/cm; determinations of u were made at 2.1 V/cm. Values of u measured for capsids I and II and bacteriophage T7 were independent of E ±10% between 1.8 and 6.8 V/cm in 0.9% gels; u for capsids I and II was also independent of time between 100 and 529 min at 2.1 V/cm. However, for unknown reasons, u for T7 phage was an increasing function of time using the conditions of Fig. 2, for times above 140 min. Values of c decreased linearly as a function of agarose concentration, A, below an A of 2.0%; By including ethidium bromide in the tube gel before and during electrophoresis, phage T7 was stained without having to remove the gel from the glass tube, and data taken down to 0.1% agarose were still linear, although -(dc/dA) was 1.5 times as high in the presence of ethidium bromide as in its absence [-(dc/dA) = 0.48 to 0.56 without ethidium bromide]. Assuming that the pore size of agarose gels decreases as A increases, - (dc/dA) should be an increasing function of r for spherical particles. Using bacteriophages T7 and R17 and tail-free bacteriophage T5 full capsids (12), this has been shown qualitatively to be the case (P. Serwer, unpublished data). Direct determinations of u have been made only for capsids I and II. For other structures the ratio of the distance migrated to the distance migrated by capsid I, R( CI), in the same gel was used to quantitate mobility.
RESULTS Electrophoresis of bacteriophage T7. Samples that contained 60, 30, 15, 6, and 3 ,ig of unlabeled bacteriophage T7 were each mixed with 35S-labeled bacteriophage T7 and subjected to electrophoresis through 0.9% cylindrical gels.
920
SERWER AND PICHLER
J. VIROL.
After ethidium bromide staining, two bands, both migrating toward the anode, were revealed on the gels with 60 to 15 pg of bacteriophage (Fig. lb, gels 1 to 3). Coomassie blue, however, stained only the slower band (not shown). This indicates that the faster band was T7 DNA (released from disrupted bacteriophages) and that the slower band had T7 DNA and capsid protein. To determine whether DNA from the slower band was packaged two experiments were performed. (i) Experiment 1. A 15-pug bacteriophage T7 sample was diluted for electrophoresis as described in Materials and Methods and was then subjected to electrophoresis with and without DNase I digestion (50 pg/ml, 30 min, 3000). The DNase eliminated the faster band but had no
detectable effect on the slower band (Fig. lb, gel 7); the undigested control is in Fig. lb, gel 6. (A comparatively small quantity of DNase-resistant, fluorescent material migrated slightly ahead of the major bacteriophage band in this bacteriophage preparation, a different preparation from the one used in Fig. lb, gels 1 to 6; the properties of the minor bacteriophage band have not yet been further investigated.) In addition, a cylindrical gel was soaked in DNase I (20 pug/ml, 1.0 h, 3000) after electrophoresis and before electrophoresis in a second dimension on a horizontal slab gel. This did not eliminate the slower band during electrophoresis in the second dimension, though the faster band was eliminated (demonstrating that material in the gel was accessible to the DNase). These data indi-
a
.
-PHAGCE
Wr
< OR I GI N
b
p
-
_
_e
8-DNAD
-
-
PHHAGE ORIGIN
1 2345 67 FIG. 1. Electrophoresis of bacteriophage T7 as a function of bacteriophage concentration and treatment with DNase. Samples containing 3S-labeled bacteriophage T7 (1,500 cpm,
Vol. 28, No. 3 Printed in U.S.A.
Electrophoresis of Bacteriophage T7 and T7 Capsids in Agarose Gels PHILIP SERWER* AND MARY E. PICHLER Department of Biochemistry, The University of Texas Health Science Center, San Antonio, Texas 78284 Received for publication 12 June 1978
Agarose gel electrophoresis of the following was performed in 0.05 M sodium phosphate-0.001 M MgCl2 (pH 7.4): (i) bacteriophage T7; (ii) a T7 precursor capsid (capsid I), isolated from T7-infected Escherichia coli, which has a thicker and less angular envelope than bacteriophage T7; (iii) a second capsid (capsid II), isolated from T7-infected E. coli, which has a bacteriophage-like envelope; and (iv) capsids (capsid IV) produced by temperature shock of bacteriophage T7. Bacteriophage T7 and all of the above capsids migrated towards the anode. In a 0.9% agarose gel, capsid I had an electrophoretic mobility of 9.1 ± 0.4 x 10-5 cm2/V.s; bacteriophage T7 migrated 0.31 ± 0.02 times as fast as capsid I. The mobilities of different preparations of capsid II varied in such gels: the fastestmigrating capsid II preparation was 0.51 ± 0.03 times as fast as capsid I and the slowest was 0.37 ± 0.02 times as fast as capsid I. Capsid IV with and without the phage tail migrated 0.29 ± 0.02 and 0.42 ± 0.02 times as fast as capsid I. The results of the extrapolation of bacteriophage and capsid mobilities to 0% agarose concentration indicated that the above differences in mobility are caused by differences in average surface charge density. To increase the accuracy of mobility comparisons and to increase the number of samples that could be simultaneously analyzed, multisample horizontal slab gels were used. Treatment with the ionic detergent sodium dodecyl sulfate converted capsid I to a capsid that migrated in the capsid II region during electrophoresis through agarose gels. In the electron microscope, most of the envelopes of these latter capsids resembled the capsid II envelope, but some envelope regions were thicker than the capsid II envelope.
Current evidence indicates that DNA enters perature-shocked bacteriophage T7; capsid IVA a preassembled capsid (capsid I) during the in has no tail, and capsid IVB has a tail (11). To determine the mechanism by which bacvivo assembly of bacteriophage T7 (8, 11, 14). In the electron microscope the envelopes of capsid teriophage T7 DNA is packaged, the following I and bacteriophage T7 both appear roughly data are useful: (i) physical characterization of spherical, but the capsid I envelope appears less bacteriophage T7 and T7 capsids; (ii) measureangular and thicker than the bacteriophage en- ments of the amounts and time course of T7 velope (11). This greater thickness is probably capsids in cells infected with bacteriophage T7 caused by the presence of larger amounts of P9 and T7 mutants; (iii) induction in a cell-free (T7 proteins are referred to by P followed by the system of the events occurring during DNA protein's gene number [14]) in the capsid I en- packaging. In the work described here, agarose velope than in the bacteriophage envelope (11). gel electrophoresis of bacteriophage T7 and T7 A capsid with an envelope that has a phage-like capsids was developed as a technique to help protein composition, angularity, and thickness obtain these data. Implications of the results in (capsid II) has also been isolated from T7-in- understanding the mechanism of T7 DNA packfected cells; capsid I may convert to capsid II aging are discussed. during an abortive or incomplete DNA packaging attempt (11, 14). Capsids I and II and bacMATERIALS AND METHODS teriophage T7 all have an internal, roughly cyand bacterial strains. BacterioBacteriophage lindrical core with an axial hole (11). The capsid phage T7 was received F. W. Studier. The host I and T7 phage cores appear to be attached to for T7 was Escherichia from coli BB/1. the capsid envelope, and DNA may enter capsid Media and buffers. Bacterial cultures were grown I through the axial hole. Capsids with envelopes and infected in M9 medium (11). Tris/Mg buffer is 0.2 resembling the envelope of capsid II in the elec- M NaCl-0.01 M Tris-chloride (pH 7.4)-0.001 M MgCl2; tron microscope have been isolated from tem- for the purification and storage of radiolabeled bacte917
918
SERWER AND PICHLER
J. VIROL.
riophage T7 and capsids, 100 ,ug of boiled gelatin was strengths are described in the figure legends. During included per ml (Tris/Mg/G buffer). Electrophoresis electrophoresis, buffer was circulated between the resbuffer is 0.05 M sodium phosphate (pH 7.4)-0.001 M ervoirs at a rate of 50 to 200 ml/min. At the end of an electrophoresis, the pH values of buffer in both tanks MgCl2. Preparation and purification of protein-radio- of a horizontal slab gel were within 0.1 pH unit of the labeled T7 bacteriophage and capsids. A station- starting pH. The concentrations of bacteriophage T7 ary-phase bacterial culture in M9 medium was diluted samples were determined by optical density (1); the into fresh M9 medium and was grown overnight at concentrations of capsid samples were determined by 300C with aeration until there were 3 x 108 bacteria the dye-binding method of Bradford (2). Detection of bacteriophage and capsids in agaper ml. The culture was infected with bacteriophage T7 (multiplicity of infection, 10), and at 10 min after rose gels. Radiolabeled bacteriophage T7 and capsids infection radioactive amino acids were added (either in cylindrical gels were detected by autoradiography 3H- or 14C-labeled algal hydrolysate or [35S]methionine after slicing the gels in a sagittal plane and drying was used at a final concentration of 20 to 40 ttCi/ml, under vacuum onto chromatography paper. Kodak purchased from Schwarz-Mann or ICN). After spon- Royal X-Omat film was used. Radiolabel in agarose taneous lysis (at 30 to 35 min after infection) the lysate slab gels can also be detected by autoradiography. To detect unlabeled DNA, gels were soaked for 45 was clarified by centrifuging at 10,000 rpm for 10 min, 40C, in a Beckman JA-21 rotor. The lysate was then min in electrophoresis buffer containing 1 Mug of ethidsedimented through a cesium chloride preformed step ium bromide per ml. The gels were then illuminated gradient containing 100 ,tg of gelatin per ml as previ- with UV light and photographed as previously deously described (11). Bacteriophage particles, which scribed (12). The presence of DNA is revealed by the banded at a density of 1.51 g/ml, were collected and ethidium bromide fluorescence enhancement induced dialyzed into Tris/Mg/G buffer. Capsids I and II, by the DNA. DNA packaged inside bacteriophage T7 which together banded in a single peak at 1.28 g/ml, induces ethidium bromide fluorescence enhancement, were collected, dialyzed into Tris/Mg/G buffer, and although the amount is less than induced by unpackthen isolated by sedimentation through an 11.0-ml, 5 aged DNA (12). To detect unlabeled protein, gels were stained in to 25% sucrose gradient supported by a 0.8-ml 62% sucrose shelf, all in Tris/Mg/G buffer (SW41 rotor, 0.05% Coomassie blue-10% acetic acid for 1.5 h at 25,000 rpm, 120 min, 180C). Capsid I traversed 70 to room temperature. The gels were destained by diffu80% of the gradient and traveled 1.36 to 1.40 times as sion into 10% acetic acid at room temperature. Relative amounts of protein in capsid bands were deterfar as capsid II. Preparation of unlabeled bacteriophage T7 mined by quantitative densitometry of the gels with a and T7 capsids. Fifteen- or 30-liter lysates of T7- Helena Quick Scan Densitometer. Changes in gel diinfected E. coli were prepared; T7 phage and capsids mensions during staining were less than 0.5%. Elution from gels. To recover unlabeled capsids were purified as previously described (11), except that (i) the sucrose gradient used in the final step of the from agarose cylindrical gels, the gels were sliced in a capsid purification had Tris/Mg buffer and (ii) the sagittal plane, and one half of the gel was stained with sucrose from this final step was not removed by di- Coomassie blue. Using the stained half of the gel as a template, the capsid regions of the unstained half were alysis. Electrophoresis in agarose gels. Agarose excised. Unlabeled bacteriophage T7 was recovered by (Seakem, ME) was dissolved in glass-distilled water observing light scattering from the bacteriophage band by boiling. Buffer was then added by diluting 10-fold and excising the band region of the gel. Material was into the agarose solution, which was maintained at 80 eluted from gel slices by one of two methods: (i) the to 900C. Cylindrical gels were poured in glass tubes gel slice was incubated with 20 to 50 pl of Tris/Mg (6-mm ID, 8.5 cm long); horizontal slab gels (6 mm buffer for 2 to 3 days at 40C (diffusion elution); or (ii) thick, 15.2 cm in the direction of the electrical field, the gel slice was placed in a dialysis bag with 100 to 13.2 cm wide) were poured in an apparatus previously 500M of electrophoresis buffer, then placed in a chamdescribed (6). Samples in Tris/Mg or Tris/Mg/G ber with two electrodes; after sufficient electrophoresis buffer were diluted into 5 volumes of 2% sucrose-0.005 buffer was added to cover the dialysis tubing, the gel M sodium phosphate (pH 7.4)-0.001 M MgC12-400 ,ug was subjected to electrophoresis (2.9 V/cm) for 2 to 3 of bromophenol blue per ml prior to layering on the h at 4°C (electrophoretic elution). To recover radiolabeled bacteriophage or capsids gels. Samples can also be dialyzed into electrophoresis buffer before electrophoresis, but this usually resulted from cylindrical gels, the gels were sliced into disks in the loss of 30 to 60% of a radiolabeled capsid sample and the disks were subjected to diffusion elution into on the dialysis tubing. The bottoms of cylindrical gels Tris/Mg/G buffer. The fractions containing radiolabel were covered with nylon mesh or dialysis tubing to were determined by diluting portions of the eluates prevent the agarose from falling out of the tube. The into toluene fluor containing Triton X-100. Recoveries tubes were placed in a commercial electrophoresis of radiolabeled capsids I and II from capsid peak apparatus and were filled with buffer; the samples regions of 0.9% agarose gels were 30 to 55%. Sucrose gradient sedimentation. For determiwere layered underneath the buffer. For horizontal slab gels, samples were placed in wells (2 by 6 mm) nation of the sedimentation rates of eluted capsids, capsid samples were layered on 4.8-ml, linear 5 to 25% previously filled with electrophoresis buffer. Electrophoresis was done at room temperature (22 sucrose gradients containing Tris/Mg/G buffer, to 2500); however, the temperature of horizontal slab poured over a 0.35-ml layer of sodium diatrizoate gels using a field strength of 2.1 V/cm was 3 to 50C (density = 1.30 g/ml). The gradients were centrifuged above room temperature. The times and field in a Beckman SW50.1 rotor for 60 min at 28,000 rpm,
ELECTROPHORESIS OF PHAGE T7
VOL. 28, 1978
18°C, and were fractionated by tube puncture. Capsid I sedimented through 70 to 80% of the sucrose gradient. Electron microscopy. Carbon support films for electron microscopy were prepared as previously described (11). A drop of sample was placed on a support film and was incubated for 1 min at room temperature. The film was washed with 4 drops of water followed by 2 drops of 1% sodium methyl phosphotungstate (pH 7.6) (7) (a gift of Glenn Williams). The film was dried with filter paper. To determine the degree of aggregation of bacteriophage T7 or T7 capsids embedded in negative stain, particles separated by 5 nm or less were assumed to be attached to each other. Bacteriophages or capsids scored for their degree of aggregation were selected at random except that all those selected were more than two bacteriophage diameters from the edge of a stain droplet. This restriction was applied because aggregation induced by drying of the specimen appeared to occur most readily at stain droplet edges. SDS-polyacrylamide gel electrophoresis of material in agarose cylindrical gels. A sample was subjected to electrophoresis in an agarose cylindrical gel, 3 mm in diameter. The gel was soaked in 1% sodium dodecyl sulfate (SDS)-0.05 M Tris-chloride (pH 6.8)-1% fi-mercaptoethanol for 15 min at room temperature and was then dropped into a tube of the same buffer maintained at 100°C by boiling. The gel was boiled until it started to become transparent (for 0.9% gels this occurs at 20 to 35 s after the start of boiling). The gel was then chilled on ice, layered across a vertical SDS slab gel similar to those used previously (11), and sealed with agarose (the same percentage used in the tube gel) in the above buffer. A well was made in the agarose for a standard used to identify protein bands. Electrophoresis and staining of the SDS gel were performed as previously described (11). Theoretical considerations. The electrophoretic mobility of a particle, u, is defined as VIE; V is the velocity of the particle in an electric field of strength E. The magnitude of u in buffer without agarose, uo, for a spherical particle depends on (i) the average density of electrical charge per unit of surface area on the particle, a-, (ii) the radius of the particle, r, (iii) the viscosity, 71, of the solution in which the particle is suspended; (iv) the Debye-Huckel constant, k, which is proportional to the square root of the ionic strength (the ionic strength of electrophoresis buffer is calculated to be 0.116). An approximate determination of u° in terms of the above parameters has been obtained (13) (this equation has been previously used in a study of bacteriophage T4 [5]): 0
2arf(kr)
(1)
3iq(l + kr)(1
f(kr) is a function that varies in value between 1.0 and 1.5 as kr varies from 0 to values above 100. An additional parameter of interest is the total number of electron equivalents of charge on a particle, z, calculated by multiplying a by the surface area of the
particle. If, to prevent convection, electrophoresis is done in the presence of a supporting gel, the gel may affect u by (i) forming channels with a pore size small enough
919
to slow particles down because of multiple collisions with gel fibers (sieving); if the channels are narrow enough, migration may be totally inhibited; and (ii) adsorbing particles by chemical interaction. The latter effect will tend to spread out electrophoretic zones and is therefore undesirable. It has been assumed that u° can be obtained by extrapolating u measured in agarose gels of decreasing agarose concentration to an agarose concentration of 0. Values of u are readily measured on horizontal slab gels; E is measured by monitoring the voltage drop across two points on the gel. However, for 0.5% and less concentrated agarose gels it is extremely difficult to stain slab gels without gel breakage; 0.9% slab gels were used for most measurements. Also, bacteriophage T7 and capsid IVB bands, for unknown reasons, were three to five times broader in slab gels than they were in cylindrical gels. Therefore, extrapolation to 0% agarose concentration was performed by subjecting a sample to electrophoresis on agarose cylindrical gels (which are less fragile than slab gels) with agarose percentages of 0.3, 0.5, 0.9, 1.5, 2.0, and 2.5% (all gels were run simultaneously in the same apparatus). The distance migrated by the sample in each gel, divided by the distance migrated in the 0.9% gel (referred to as c), was extrapolated to 0% agarose (referred to as c°), and uo was calculated:
u0 u0.c =
(2)
Determinations of c were made at E values of 5.8 V/cm; determinations of u were made at 2.1 V/cm. Values of u measured for capsids I and II and bacteriophage T7 were independent of E ±10% between 1.8 and 6.8 V/cm in 0.9% gels; u for capsids I and II was also independent of time between 100 and 529 min at 2.1 V/cm. However, for unknown reasons, u for T7 phage was an increasing function of time using the conditions of Fig. 2, for times above 140 min. Values of c decreased linearly as a function of agarose concentration, A, below an A of 2.0%; By including ethidium bromide in the tube gel before and during electrophoresis, phage T7 was stained without having to remove the gel from the glass tube, and data taken down to 0.1% agarose were still linear, although -(dc/dA) was 1.5 times as high in the presence of ethidium bromide as in its absence [-(dc/dA) = 0.48 to 0.56 without ethidium bromide]. Assuming that the pore size of agarose gels decreases as A increases, - (dc/dA) should be an increasing function of r for spherical particles. Using bacteriophages T7 and R17 and tail-free bacteriophage T5 full capsids (12), this has been shown qualitatively to be the case (P. Serwer, unpublished data). Direct determinations of u have been made only for capsids I and II. For other structures the ratio of the distance migrated to the distance migrated by capsid I, R( CI), in the same gel was used to quantitate mobility.
RESULTS Electrophoresis of bacteriophage T7. Samples that contained 60, 30, 15, 6, and 3 ,ig of unlabeled bacteriophage T7 were each mixed with 35S-labeled bacteriophage T7 and subjected to electrophoresis through 0.9% cylindrical gels.
920
SERWER AND PICHLER
J. VIROL.
After ethidium bromide staining, two bands, both migrating toward the anode, were revealed on the gels with 60 to 15 pg of bacteriophage (Fig. lb, gels 1 to 3). Coomassie blue, however, stained only the slower band (not shown). This indicates that the faster band was T7 DNA (released from disrupted bacteriophages) and that the slower band had T7 DNA and capsid protein. To determine whether DNA from the slower band was packaged two experiments were performed. (i) Experiment 1. A 15-pug bacteriophage T7 sample was diluted for electrophoresis as described in Materials and Methods and was then subjected to electrophoresis with and without DNase I digestion (50 pg/ml, 30 min, 3000). The DNase eliminated the faster band but had no
detectable effect on the slower band (Fig. lb, gel 7); the undigested control is in Fig. lb, gel 6. (A comparatively small quantity of DNase-resistant, fluorescent material migrated slightly ahead of the major bacteriophage band in this bacteriophage preparation, a different preparation from the one used in Fig. lb, gels 1 to 6; the properties of the minor bacteriophage band have not yet been further investigated.) In addition, a cylindrical gel was soaked in DNase I (20 pug/ml, 1.0 h, 3000) after electrophoresis and before electrophoresis in a second dimension on a horizontal slab gel. This did not eliminate the slower band during electrophoresis in the second dimension, though the faster band was eliminated (demonstrating that material in the gel was accessible to the DNase). These data indi-
a
.
-PHAGCE
Wr
< OR I GI N
b
p
-
_
_e
8-DNAD
-
-
PHHAGE ORIGIN
1 2345 67 FIG. 1. Electrophoresis of bacteriophage T7 as a function of bacteriophage concentration and treatment with DNase. Samples containing 3S-labeled bacteriophage T7 (1,500 cpm,
