108, 164-176 (1981)
Capsid-DNA Complexes in the DNA Packaging Pathway of Bacteriophage T7: Characterization of the Capsids Bound to Monomeric and Concatemeric DNA PHILIP Department
SERWER The University
of Texas Health Science Center, San Antonio,
Accepted August 11, 1980 Complexes of capsids with monomeric bacteriophage T7 DNA and complexes of capsids with concatemeric bacteriophage T7 DNA have been isolated from lysates of T’7-infected Escherichia coli and the capsids of these complexes have been characterized. In electron micrographs capsids bound to either monomeric DNA or concatemeric DNA have envelopes which are: (1) significantly less thick than the envelope of a DNA-free capsid (capsid I) isolated from T7-infected E. coli, (2) indistinguishable from the envelope of bacteriophage T7 and the envelope of a second DNA-free capsid (capsid II) isolated from T7-infected E. coli. By agarose gel electrophoresis (under nondenaturing conditions) and SDS-polyacrylamide gel electrophoresis the capsids released from the above capsid-DNA complexes are similar to capsid II and different from capsid I. The physical characteristics and radiolabeling kinetics of these capsid-DNA complexes suggest that they are DNA packaging intermediates; the results suggest a pathway for the packaging of DNA by bacteriophage T7.
Studies of the packaging of duplex DNA by bacteriophages T3, T4, T’7, P22, and h have revealed pathways of DNA packaging which all have the following features: (1) a DNA-free capsid is assembled and subsequently packages DNA; structural alteration of the capsid occurs just before or (and) during packaging of DNA. (2) When packaging is initiated the DNA is longer than the mature DNA of the bacteriophage (mature length bacteriophage DNA will be referred to as monomeric DNA; linear bacteriophage DNA with a length greater than monomeric DNA will be referred to as concatemeric DNA); the concatemeric DNA is cut to monomeric DNA during or after packaging (reviewed in Casjens and King, 1975; Murialdo and Becker, 1978; Wood and King, 1979; Earnshaw and Casjens, 1980; details for bacteriophage T7 are in the legend to Fig. 6). These observations indicate that capsid-DNA complexes with DNA outside of the capsid should be in the DNA packaging pathways of the above bacteriophages 0042~6822B1/010164-13$02.00/O Copyright 0 1981 by Academic Press, Inc. AU rights of reproduction in any form reserved.
and that at least some of these complexes should have concatemeric DNA. Electron microscopy of lysates of bacteriophage-infected cells has produced evidence of capsid-DNA complexes in the DNA packaging pathways of bacteriophage A (Syvanen and Yin, 1978; Yamagishi and Okamoto, 1978) and T4 (Kemper and Brown, 1976). Capsid-DNA complexes formed during in vitro DNA packaging by bacteriophage A have been partially isolated (Becker et al., 1977). Complexes of bacteriophage T7 capsids with monomeric, concatemeric, and possibly replicating T7 DNA have been isolated (Serwer, 1974a, b). However, none of the capsids of the above capsid-DNA complexes has been sufficiently characterized to determine which of the structural transitions thought to occur during DNA packaging have occurred in these capsids. Subsequent to the above studies of T7 capsid-DNA complexes, the following techniques have been developed for’characterizing capsids of the capsid-DNA complexes of bacteriophage T7: (1) a negativestaining technique for electron microscopy
which reveals DNA of capsid-DNA complexes in an extended configuration (Serwer, 19’78); (2) agarose gel electrophoresis (after release of capsids from DNA) (Serwer and F’ichler, 1978); (3) buoyant density sedimentation in density gradients of metrizamide (after release of capsids from DNA) (Serwer, 1980). In the present study these and other techniques have been used to characterize capsids bound to monomeric and concatemeric DNA, isolated from lysates of bacteriophage T7-infected Escherichia coli. Implications of the results for understanding the DNA packaging pathway of T7 are discussed. MATERIALS
of Bacteria and Bacteriophages
Wild-type bacteriophage T7 and T7 amber mutants (Studier, 1969) were received from Dr. F. W. Studier. The following amber mutants were used: gene 5-am 28; gene 19-am 10. The host for bacteriophage T7 and the nonpermissive host for T7 amber mutants was E. coli BBIl; the permissive host for amber mutants was E. coli O-l 1’ . Intermediates isolated from lysates of the nonpermissive host infected with an amber mutant will be referred to by the number of the mutant gene; for instance, capsid I from a lysate of the nonpermissive host infected with the 5am mutant is referred to as 5 urn capsid I. Buffers and Media Standard/G buffer was 0.15 M NaCl, 0.05 M Tris-Cl, pH 7.4, 0.005 M EDTA, 100 pg/cm” gelatin. Cultures of E. coli BBIl were grown in M9 medium (Serwer, 1976). Infection, Radiolabeling, and Artifiial Lysis of T?‘-Infected E. Coli A log phase culture of E. coli BBIl was infected at 4.0 x 108/cm3 with bacteriophage T7 or a T7 amber mutant using a multiplicity of 15. The culture was labeled for the time indicated with either 3H or 14C-algal hydrolysate (ICN or Schwarz/Mann); the culture was quenched and lysed with the nonionic detergent Brijs8 as previously described (Serwer, 1974a).
Sedimentation of Lysates in SucroseMetrizamide Gradients; Isolation of Radiolabeled Capsid I, Capsid II, and Bacteriophage T?’ Isolation of 14C-labeled capsid I and capsid II by sedimentation though a biphasic gradient of sucrose and metrizamide was performed as previously described (Serwer, 1980). Isolation
Unlabeled capsid I was isolated as described in Serwer (1976). Sedimentation of Capsid -DNA complexes in Sodium Iothalamate Density Gradients After partial isolation of 35-55 S capsidDNA complexes by sedimentation in a sucrose-met&amide gradient (see Fig. 1 of Serwer, 1980), the in viva capsid-monomerit DNA complex was sometimes further purified by either: (1) buoyant density sedimentation for 19.5 hr in a sodium iothaIamate density gradient, as described in Serwer (1980) (density = 1.174 g/cm3> or (2) layering of 1.70 cm3 of sample on a preformed, discontinuous sodium iothalamate gradient and centrifuging at 45 K, 25” for 4.5 hr. The layers of sodium iothalamate in the discontinuous gradient had the following volumes and densities: 0.50 cm3, 1.30 g/cm3; 1.00 cm3, 1.19 g/cm3; 2.00 cm3, 1.07 g/cm3; the buffer in the gradients was 0.01 M Tris-Cl, pH 7.4, 0.001 M EDTA, 100 pg/cm3 gelatin. Capsid-DNA complexes were stored and used without dialysis” Buoyant Density Sedimentation amide Density Gradients
Buoyant density sedimentation in metrizamide density gradients was performed as previously described (Serwer, 1980). Sedimentation of Capsid -DNA Complexes in Cesium Chloride Step Gradients Capsid-DNA complexes, partially purified by sedimentation in a sucrose-metrizamide gradient, were layered on a discontinuous cesium chloride gradient with layers
that had the following volumes and densities: 0.40 cm3, 1.72 g/cm”; 1.20 cm3, 1.60 g/cm”; 0.85
em3, 1.50 g/cm3; 0.90 cm3, 1.28 g/cm3; 0.80 cm3, 1.16 g/cm3. A sample with a volume of 1.0 cm3 was layered on top of the step gradient and was centrifuged at 40 K, 20”, for 6.5 hr. Fractions (five drops) were collected through an l&gauge needle after puncturing the bottom of the centrifuge tube; the speed of collection was less than one drop per second. Material to be used for further analysis was immediately dialyzed into standard/G buffer. The capsid-DNA complexes migrated after the above sedimentation to positions within 0.01 g/cm3 of their isodense position (determined by buoyant density sedimentation of the capsid-DNA complexes after isolation from the cesium chloride step gradient). Sedimentation in a step gradient was used in preference to the buoyant density sedimentation previously used (Serwer, 1974b) because the capsid-DNA complexes from the above sucrose-metrizamide gradient had not been as well separated from slowly sedimenting, rapidly diffusing contaminants as in the previous study, and sedimentation in a step gradient is a better technique than buoyant sedimentation for separating the capsid-DNA complexes from such contaminants. In the experiments performed here, a peak not observed in a previous study (Serwer, 19’74b) was observed at 1.555 g/cm3, the density of capsid-dimeric T7 DNA complexes. (In the previous study a plateau was observed in this region.) This difference in results may be caused by the difference in media for growth used in these two studies. Velocity Sedimentation crose Gradients
X-100 and then counting in a liquid scintillation spectrometer. Agarose Gel Electrophoresis
Electrophoresis in horizontal agarose slab gels was performed as previously described (Serwer, 1980) except that 0.001 M EDTA replaced the MgCl, in the buffer used for electrophoresis and in buffer used for diluting samples for electrophoresis. (Omitted in the description of methods in this previous study was reference to the recirculation of buffer at 50 cm3/min starting 0.75 hr after the commencement of electrophoresis.) The EDTA replaced the MgCI, to protect capsid-DNA complexes from the action of DNases. Within experimental error the ratio of the mobility of capsid II to the mobility of capsid I is the same in the buffer used here as in the buffer previously used. Electron
Capsid- DNA complexes were prepared for electron microscopy using either the aqueous method or formamide method previously described for negatively staining capsid- DNA complexes (Serwer, 1978; these procedures reveal DNA in an extended configuration). When using the formamide method, 20% formamide was in the mixture applied to the carbon support film. This amount of formamide does not affect the appearance of the envelope of capsids bound to DNA, but reduces the tangling of duplex DNA that occurs when using the aqueous method (Serwer, 1978). SDS-Polyacrylamide
Velocity sedimentation in sucrose gradients was performed as previously described (Serwer, 1980), except that a 0.35 cm3 sodium diatrizoate layer (1.30 g/cm3) was used beneath the sucrose gradient instead of the layer of met&amide.
The procedures described by Studier (1973) were used. The separating gel was a lo-20% gradient of acrylamide. Fluorography using the procedure of Bonner and Laskey (1973) and preflashed film (Laskey and Mills, 1975) was used to increase the efficiency of detection for 14C.
Assay for Radioactivity
Radioactivity in samples was assayed by diluting into toluene fluor containing Triton
Kinetic labeling of T7-infected E. coli and determination of the amounts of 14C in bac-
FIG. 1. Electron microscopy of the capsid region of T7 capsid-DNA complexes. In viva capsid-monomerit DNA complexes were isolated by sedimentation in a sucrose-metrizamide gradient, followed by either buoyant density sedimentation in a sodium iothalamate density gradient or sedimentation in a sodium iothalamate step gradient, as described in Materials and Methods. In viva capsid-concatemeric DNA complexes with a 8 of 1.555 g/cm3 during buoyant density sedimentation in a cesium chloride density gradient were prepared by sedimentation in a sucrose-metriaamide gradient followed by sedimentation in a cesium chloride step gradient, as described in Materials and Methods. (a)-(h) In viva eapsid-monomeric DNA complexes, prepared for electron microscopy using the aqueous technique (Materials and Methods), (i) capsid I, prepared for electron microscopy using the aqueous technique, (j-l) in viva capsid-concatemeric DNA complexes, prepared for electron microscopy using the formamide technique (Materials and Methods). 167
teriophage T7, capsid I, MLD capsid II, and MHD capsid II was performed as previously described (Serwer, 1980).
capsid I are in Fig. 1, i). Thus, the capsids of the in vivo capsid-monomeric DNA complexes appear to be a form of capsid II. To further test the properties of the capsids of the in vivo capsid-monomeric DNA RESULTS complex a 14C-labeled preparation of in vivo Physical Characterization of Capsids Bound capsid-monomeric DNA complex less than to Monomeric T?’ DNA 20% contaminated with in vitro capsidDNA complexes (the degree of contaminaCapsid-monomeric DNA complexes from tion was assumed to be the percentage of bacteriophage T7-infected E. coli (in vivo 14C in capsids not released from DNA by capsid-monomeric DNA complexes; Sertreatment with Sarkosyl followed by analywer, 197413; see the legend to Fig. S), purisis using velocity sedimentation in sucrose fied by either buoyant density sedimentagradients; see the legend to Fig. 6) was dition in a sodium iothalamate density gradigested with DNase and subjected to agaent or sedimentation in a sodium iothalamate rose gel electrophoresis. The profile of the step gradient as described in Materials and released capsids (Fig. 2, slot c) had a band Methods, were prepared for electron microscopy. Of 68 capsids observed at random on migrating 1.07 times as fast as the band of the viewing screen, 64 appeared bound to capsid II (Fig. 2, slot b) and had no band migrating near the band of capsid I (Fig. 2, DNA; the envelopes of all of the DNA-bound slot a) (treatment of capsid I and capsid II capsids had a capsid II-like appearance. with DNase did not alter the mobility of Thirty of the DNA-bound capsids were atthese capsids; data not shown). Exposures tached to the DNA near one of the ends of four times as dark as the one in Fig. 2 also the DNA and are, therefore, presumed not revealed no band of capsid I in Fig. 2, slot c. to be capsid-DNA complexes from disThus, by agarose electrophoresis, as well as rupted bacteriophages (in vitro capsidby electron microscopy, the envelopes of the DNA complexes; Serwer, 19’7413)contamicapsids of this capsid-DNA complex renating the preparation. The remaining sembled the envelope of capsid II and were DNA-bound capsids were either attached to different from the envelope of capsid I. DNA that was too entagled to determine The small difference in the distances mithe distance from the nearest end (31 capgrated by capsid II and the capsid of the in sids) or were attached further from the DNA complex has nearest end than the capsid of the in vivo vivo cap&d-monomeric not always been observed; variability of the capsid-monomeric DNA complex (three electrophoretic mobility of capsid II has capsids; these may be in vitro capsid-DNA previously been reported (Serwer and Pichcomplexes from disrupted particles of bacler, 1978). Therefore, the above difference teriophage T7). To determine the characterin distance migrated does not demonstrate a istics of the capsids of in vivo capsid-monosignificant difference between the envelope merit DNA complexes, only capsids near of capsid II and the envelope of the capsid one end of the DNA will be considered of the in vivo capsid-monomeric DNA combelow. Micrographs were taken of 20 in vivo cap- plex. Electrophoresis of the in viva capsidsid-monomeric DNA complexes (the capsid monomeric DNA complex after treatment buffer used region of eight of these capsid-DNA com- with the magnesium-containing for digestion with DNase (Fig. 2, slot d) and plexes is shown in Fig. 1, a-h); the thickness of the envelopes of all 20 of these ca.psids after no treatment (Fig. 2, slot e) revealed: (1) a band migrating 0.48 times as fast as the was 2.0-3.0 nm, a thickness indistinguishable from the thickness of the envelope of band formed by capsid II, and (2) electroheterogeneous material beeither MHD capsid II, MLD capsid II, or phoretically bacteriophage T7 (Serwer, 1977, 1980) (see tween the origin and the band formed by the complex. Fig. 6 for a description of capsid I and capsid capsid-DNA II) and significantly smaller than the thickTo determine whether capsids of the in ness of the envelope of capsid I (particles of vivo capsid-monomeric DNA complex had
FIG. 2. Agarose gel electrophoresis of capsids released from the in viva capsid-monomeric DNA complex. ‘“C-labeled capsid I and capsid II were obtained as described in Materials and Methods using a culture labeled with ‘“C from 14 to 24 min after infection. In vivo capsid-monomeric DNA complexes with 14C-labeled capsids were prepared by subjecting 35-55 S capsidDNA complexes from the same lysate to buoyant density sedimentation in a sodium iothalamate density gradient, as described in Materials and Methods. Two 3~1 samples of the in vivo capsid-monomeric DNA complex were separately diluted into 32 ~1 of a buffer containing 0.010 M Tris-Cl, pH 7.4, 0.012 M MgCl,, 100 pgicm” gelatin; to one of these mixtures was added 2.0 ~1 of DNase I (1 mg/cm3, Worthington) and both mixtures were incubated at 30.0” for 1.0 hr before addition of 2 ~1 of 0.2 M sodium EDTA, pH 7.4 and 40 ClJof buffer containing 0.005 M sodium phosphate, pH 7.4,O.OOl M EDTA, 6% sucrose, 400 cLgIcm3bromophenol blue. A 50-~1 sample of each of these mixtures was layered in the sample wells of a horizontal agarose slab gel. Samples of capsid I and capsid II were also incubated, diluted, and layered on an agarose slab gel as aforementioned. The samples were subjected to electrophoresis in a 0.9% agarose gel for 16.0 hr and were dried for autoradiography as described in Materials and Methods. (a) Capsid I, (b) capsid II, (cl in vivo capsid-monomeric DNA complex digested with DNase, (d) in vivo capsid-monomeric DNA complex treated with the buffer used for the DNase digestion (no DNase), (e) in vivo capsid-monomeric DNA complex, no incubation.
the MHD property previously described for capsid II (see the legend to Fig. S), capsids from this complex were subjected to buoyant density sedimentation in a metrizamide density gradient after being released from DNA by treatment with DNase I. All cap-
sids released had a 8 of 1.268 g/cm”, a 8 indistinguishable from that of MHD capsid II. (Treatment with DNase was not necessary to release the eapsid of the in vivo capsidmonomeric DNA complex; buoyant density sedimentation in a met&amide gradient without treatment with DNase released this capsid, but not the capsid of the in vitro capsid-DNA complex.) Because the capsid of the in vivo capsid - monomeric DNA complex is by the above criteria indistinguishable from MHD capsid II, it will be referred to as MHD capsid II (the nomenclature used here replaces a less descriptive nomenclature previously proposed; Serwer, 1976). Of the 30 capsids in in vivo MHD eapsid II-monomeric DNA complexes observed in the electron microscope, 6 had an internal object resembling the core (see Fig. 6) or a piece of the core of capsid I and capsid II (Fig. 1, particles a and b); the remainder of these capsids did not have a core (Fig. 1, particles c-h). The DNA bound to 5 of the core-containing capsids joined the envelope at a point indistinguishable from the point at which the core joined the envelope (Fig. 1, particles a, b). It is likely, therefore, that the DNA is bound to the envelope at the base of the core. (The one core that was not joined to the envelope at the DNA-joining point may have been a core detached from the envelope.) It has been hypothesized that an external projection from capsid I binds to DNA during T’7 DNA packaging (Serwer and Pichler, 1978; Serwer, 1980). Therefore, evidence for an external projection on the capsid of the in vivo MHD capsid II-monomeric DNA complexes (possibly the same projection proposed for capsid I) was sought. There was no evidence of an external projection on the capsids of 25 of the in vho MHD capsid II-monomeric DNA complexes, including all of the complexes with corecontaining capsids (capsids without cores or external projections are in Fig. 1, c-e). However, there was evidence of a DNAbound external projection on the envelopes of the remaining 5 capsid-DNA complexes. This projection varied in size from 11- 13 nm (Fig. 1, f)to44-48nm(Fig. l,g, h). Thetail of bacteriophage T7 is 15-20 nm long (Luftig and Haselkorn, 1968); therefore, the
FIG. 3. Electron microscopy of an in viva capsid-concatemeric DNA complex. In vivo capsid-concatemeric complexes, isolated by centrifugation in a cesium chloride step gradient as described in Materials and Methods (0 = 1.555 g/cm3), prepared for electron microscopy using the formamide technique. The arrow indicates the capsid.
longer projections cannot be tails, although they may contain proteins of the tail. Physical Characterization of Capsids Bound to Concatemeric T7 DNA Capsid-concatemeric DNA complexes, isolated by sedimentation in a cesium chloride step gradient as described in Materials and Methods, were also subjected to agarose gel electrophoresis after digestion of the DNA with DNase I. The profile of 14C from all capsid-DNA complex-containing regions of the step gradient was the profile of capsid II; no band at the position of eapsid I was observed (data not shown). The slowest sedimenting form of MLD capsid II (see Serwer, 1980) cosedimented with 14C-labeled capsids released from capsid-concatemeric DNA complexes by DNase I (data not shown). By electron microscopy also the envelopes of the capsids bound to concatemeric DNA were indistinguishable
from the envelope of capsid II (Fig. 1, particles j-l; an entire capsid-concatemeric DNA complex is in Fig. 3). Of 25 capsids bound to concatemeric DNA (0 in cesium chloride of the complexes = 1.555 g/cm3) all had envelopes with a thickness of 2.0-3.0 nm, the thickness of the envelope of capsid II. There was no evidence of an external projection or core on 24 of these capsids (Fig. 1, j, k); one DNA-bound capsid had an internal core joined to the envelope of the capsid at the same location as the DNA (Fig. 1, 1). By the above criteria, the capsids bound to concatemeric T7 DNA are indistinguishable from capsid II. The density of these capsids during buoyant density sedimentation in metrizamide density gradients was the density of MHD capsid II (data not shown) and these capsids will be referred to as MHD capsid II. All capsids released from in viva capsidmonomeric DNA complexes and in vivo cap-
i j k Im
FIG. 4. SDS-polyacrylamide gel electrophoresis of cap&d-DNA complexes from a cesium chloride step gradient. Samples of fractions of a cesium chloride step gradient used to fractionate capsid-DNA complexes, as described in Materials and Methods, were subjected to SDS-polyacrylamide gel electrophoresis. The volume of all samples was 40 4, except for sample g (capsid-monomeric DNA complex) of which only 16 d was used. Also included are wild-type capsid I, 5 am capsid I, 19 am capsid I. The samples from the step gradient are identified by their 0 in g/cm3. (a) 1.630, (b) 1.609, (c) 1.582, (d) 1.555, (e) 1.529, (f) 1.510, (g) 1.490, (h) 1.468, (i) 1.447, (j) wild-type capsid I, (k) blank, (1) 5 am capsid I, (m) 19 am capsid I. The arrowhead indicates the origin of electrophoresis. The arrow indicates an unidentified protein that migrates close to Pll.
sid-concatemeric DNA complexes after isolation of these complexes using only sedimentation in a sucrose-metrizamide gradient had the electrophoretic profile of capsid II (data not shown). This indicates that the further purification of the capsid-DNA complexes used above has not caused a capsid I + capsid II conversion in the capsid of these capsid-DNA complexes, a conversion that sometimes occurs (Serwer and Pichler, 1978). Sodium Dodecyl Sulfate -Polyacrylamide Gel Electrophoresis To determine the proteins in in vivo capsid-DNA complexes, fractionated on a cesium chloride step gradient as described in Materials and Methods, these complexes were subjected to SDS-polyacrylamide gel electrophoresis with fluorographic detection. The proteins present in the fraction at
the peak formed by the in vivo MHD capsid II-monomeric DNA complex (Fig. 4, slot g) are: proteins of the T7 envelope, PlO and Q (T7 proteins with identified genes are indicated by P followed by the number of the protein’s gene (Studier, 1972); other T7 proteins are indicated by letters (Serwer, 1976)); a protein of the T7 tail, P17; and proteins of the internal, cylindrical T7 core, Pl5 and P16. After sedimentation in sucrose gradients and before sedimentation in cesium chloride gradients P13, P14, and sometimes P12 appear to be associated with the in vivo MHD capsid II-monomeric DNA complex (data not shown), suggesting that these proteins detach from the capsid of this complex during banding in the cesium chloride step gradient. In the gel of Fig. 4, P9 (a major protein of the envelope of capsid I, but not capsid II or bacteriophage T7; Serwer, 1976; see Fig. 6) has migrated too close to PlO to be seen as a separate band in the more intense samples (for instance, P9 is not resolved from PlO in the samples of capsid I in slots j, 1, and m of Fig. 4). However, lesser exposure of the gel in Fig. 4 and independent SDS gel profiles revealed no or only trace amounts of P9 in the in vivo MHD capsid IImonomeric DNA complex or the in viva MHD capsid II-concatemeric DNA complexes. The proteins present in the in U~ZYI MHD capsid II-concatemeric DNA complexes (Fig. 4, slots b-f) are the same as those present in the in vivo MHD capsid IImonomeric DNA complex, except that the relative amount of l*C in P17 decreases with increasing 0 (increasing length of DNA; Fig. 4, slots f-d) of the capsid-DNA complexes until 0 = 1.582 g/cm3 (Fig. 4, slot c). Evidence exists suggesting that P19, a relatively minor protein of capsid I, is necessary for the initial binding of capsid I to DNA (Roeder and Sadowski, 1977); P19 may, therefore, be the protein that binds capsid I (and possibly the MHD capsid II of the in vivo MHD capsid II-DNA complexes) to DNA. To identify P19 in Fig. 4 a profile of capsid I with P19 (from a 5am lysate; Fig. 4, slot I) was compared to a profile of capsid I without P19 (from a 19 am lysate; Fig. 4, slot m); this was necessary because the relative mobility of P19 and P17 is variable and the two proteins can, therefore, be confused. No P19 was found in any of the in viuo cap-
sid-DNA complexes in Fig. 4; exposing the same gel for four times as long as the exposure used for Fig. 4 also did not reveal P19 in any of the capsid-DNA complexes. Fluorograms such as the one in Fig. 4 reveal qualitative, but not quantitative, information concerning the proteins in T7 capsids. This is because the proteins incorporate 14Cwith different kinetics. For instance, 14C is incorporated into the PlO of capsid I more rapidly than it is incorporated into the core proteins of capsid I (P. Serwer, unpublished observations). Because all preparations of the in viva MHD capsid II-monomeric DNA complex are contaminated with Sarkosyl-resistant capsid-DNA complexes, probably related to the in vitro capsid-DNA complex, it is possible that some of the comparatively minor proteins in the capsid-DNA complexes of Fig. 4 are present primarily or totally in contaminating particles of in vitro capsid-DNA complex-like material. The capsid-monomeric DNA complexes must be separated from each other to be sure of the proteins present in each capsid-DNA complex. A technique for performing this separation has not yet been developed. Kinetics
Data obtained in a previous study (Serwer, 1980) indicated the following temporal sequence for DNA-free capsids isolated from lysates of T7-infected E. coli: capsid I + MLD capsid II + MHD capsid II. Because capsids bound to concatemeric DNA and monomeric DNA have several physical properties indistinguishable from the properties of MHD capsid II, it is likely that these capsid- DNA complexes occur after MLD capsid II in the T7 DNA packaging pathway. To determine experimentally the sequence of the capsid-DNA complexes in the T7 assembly pathway, T7-infected E. coli were labeled with 14C-labeled amino acids from 13.7 to 15.0 min after infection (the pulse); a sample of the culture was then taken and an excess of unlabeled amino acids was added to the culture (the chase); additional samples were taken at 15.7, 17.5, 19.0,21.0, and 25.1 min after the chase. All samples were lysed with Brij58 as described
FIG. 5. Time course of DNA-free capsids and capsidDNA complexes. A kinetic labeling experiment (pulsechase) was performed and the amounts of 14Cin capsid I, MLD capsid II, MHD capsid II, and bacteriophage T7 at each time were determined as described in Materials and Methods. The amount of 14Cin Sarkosylresistant capsid-DNA complexes at each time was determined by subjecting Sarkosyl-treated (1% Sarkosyl in standard/G buffer, 1.0 hr, 30.0”)capsid-DNA complexes from the 35 to 55 S region of suerose-metrizamide gradients of the labeled lysates to sedimentation through a cesium chloride step gradient as described in Materials and Methods. The amount of 14C in the in viva MHD capsid II-monomeric DNA complex of each of the labeled lysates was determined by subtracting the amount of “C in the Sarkosyl-resistant capsidDNA complex from the amount of 14C in all forms of capsid-monomeric DNA complex, the latter determined by subjecting untreated capsid-DNA complexes from the 35 to 55 S region of sucrose-metrisamide gradients to sedimentation in cesium chloride step gradients, as described in Materials and Methods. The amount of 14Cin the MHD capsid II-concatemeric DNA complexes of each of the labeled lysates was determined by sedimentation in a cesium chloride step gradient, performed as described in Materials and Methods. - A -, Capsid I, maximum amount of 14C = 122,305cpm; - 0 -, MLD capsid II, maximum amount of “C = 8642 cpm; - A -, MHD capsid II, maximum amount of 14C = 16,014 cpm; - 0 -, in vivo MHD capsid II-concatemeric DNA complex, maximum amount of 14C = 1988 cpm; - n -, in vivo MHD capsid II-monomeric DNA complex; maximum amount of 14C = 4589 cpm; - l -, Sarkosyl-resistant capsid-DNA complex, maximum amount of 14C = 2366 cpm; - 0 -, bacteriophage T?, maximum amount of 14C = 181,374 cpm.
in Materials and Methods; the normalized amounts of the following were determined for each lysate and are plotted as a function of time in Fig. 5: capsid I, MLD capsid II, MHD capsid II, Sarkosyl-sensitive MHD capsid II-monomeric DNA complex (the in vivo MHD capsid II-monomeric DNA complex), Sarkosyl-resistant capsid-mono-
-P9 -PI9 Capsid I + Concaiimric F8
(Attached to loos+ DNA?)
Capsid I-IConcatemeric cc%Yp?ex
t MLD Capid II-0 Conca~NmAeric Complex
2s PI4 PI5 PI6 PI9 o*
1 “In Vivd + “In Vivd’ MHD Copsid II- MHD Copsid IIConca;aNr7pric Monomeric DNA Complex Complex
E;o* PII PI2 Pi3 22 Pi6 PI7 Pi
FIG. 6. The proposed sequence of intermediates in the DNA packaging pathway of bacteriophage T7. The T7 DNA packaging pathway, deduced from the intermediates isolated, is shown schematically. Capsid I (sometimes referred to as prohead (Roeder and Sadowski, 1977)) has an envelope d%erent in thickness, angularity, composition and average electrical surface charge density from the envelope of bacteriophage T7 (Serwer, 1976; Serwer and Pichler, 1978; Roeder and Sadowski, 197’7). After binding DNA capsid I converts first to a capsid with a bacteriophage-like, metrizamide-impermeable envelope (metrizamide low density eapsid II or MLD capsid II) and then to a eapsid with a bacteriophage-like, metrizamide-permeable envelope (met&amide high density capsid II or MHD capsid II) (Serwer, 1989). All capsid I, all MLD capsid II, and some MHD capsid II isolated have been DNA-free; complexes of these capsids with DNA are apparently unstable in the conditions used thus far. The earliest, thus far isolated capsid-DNA complex of the pathway is the in Go MHD capsid II-concatemeric DNA complex described in the Results and Serwer (19’7413). Evidence presented in the Results indicates that DNA was at least partially packaged in the capsid of this complex prior to the lysis of T7-infected cells. In the latter stages of packaging, DNA is cut to mature size producing a particle that has monomeric, packaged DNA (though pictured as fully packaged, the extent of packaging in this latter particle is not known) and that empties its DNA during lysis to form the in wivo MHD capsid II-monomeric DNA complex. All T7 capsids in the above pathway have an internal, cylindrical core attached to the envelope of the capsid. During DNA packaging, DNA may enter the capsid through an axial hole in this core (Serwer, 1976,198o and the Results). A capsid-DNA complex is also produced by disrupting baeteriophage T7 with high temperature (in vitro capsid-DNA complex) (Serwer, 197413, 1976). The in vitro complex has its capsid bound at a random position on the DNA and is insensitive to the ionic detergent Sarkosyl; the in viva complexes all have their capsids near an end of the DNA and are sensitive to Sarkosyl (Serwer, 1974b). (Most capsid proteins released by Sarkosyl sediment and migrate in agarose geb as intact capsids; P. Serwer and R. H. Watson, unpublished data.) A question mark indicates a capsidDNA complex whose existence is indicated by the data, but which has not yet been isolated. The proteins present in capsid I and bacteriophage T7 are indicated; the symbol + indicates variable presence (see Serwer, 1976).
merit DNA complex (presumed to be a capsid-DNA complex related to the in vitro capsid-DNA complex), MHD capsid IIconcatemeric DNA complex, and bacteriophage T7. This experiment is the same experiment reported in Fig. 12 of Serwer (1980) and shows that the complexes of MHD capsid II with monomeric DNA and with eoncatemeric DNA both appear after MLD capsid II and DNA-free MHD capsid II. The amount of 14C in the in wivo MHD capsid II-monomeric DNA complex increased as a function of time in Fig. 5 until it reached a maximum at 21.0 min and then
decreased. This is a kinetic behavior similar to that previously described for the in viva MHD capsid II-monomeric DNA complex (Serwer, 1974b) and is the kinetic behavior expected of a DNA packaging interrne&ate. In contrast, the amount of 14C in the Sarkosyl-resistant capsid-monomeric DNA complex and the amount of 14C in bacteriqhage T7 increased monotonically as a function of time; the Sarkosyl-resistant capsid- DNA complex appeared after the in vivo MHD capsid II-monomeric DNA complex aml before bacteriophage T7. The monotonic increase in the amount of 14C in the Sarkosyl-
resistant capsid-monomeric DNA complex suggests that this complex is either: (1) an abortive end-product of assembly or a breakdown product of such an end-product or (2) an intermediate in assembly that appears too late in assembly to observe at the later times of Fig. 5 a decrease in the amount of 14C it has incorporated. The appearance of 14C in the Sarkosyl-resistant capsid-monomeric DNA complex before its appearance in bacteriophage T7 indicates that some 14C-labeled particles of this capsid-DNA complex are not products of the degradation of completed bacteriophage T7. Some Sarkosyl-resistant capsid- monomeric DNA complexes might be products of the breakdown of a particle that had packaged DNA and that had completed some, but not all, post-DNA packaging events in the assembly of T7. The in viva MHD capsid II-concatemeric DNA complexes of Fig. 5 also appeared with the kinetics of a DNA packaging intermediate; they appeared before the in vivo MHD capsid II-monomeric DNA complexes and after the DNA-free MHD capsid II. However, the difference in the time of appearance of the in vivo MHD capsid IImonomeric DNA complex and the in vivo MHD capsid II-concatemeric DNA complex is probably too small to be sure this difference is significant. Because mature bacteriophage T7 has monomeric DNA, it is assumed that the in vivo MHD capsid IIconcatemeric DNA complex appears before the in vivo MHD capsid II-monomeric DNA complex in the DNA packaging pathway of bacteriophage T7. Previous data suggest that the DNA-free particles of MHD capsid II had packaged at least some DNA prior to lysis (Serwer, 1980). The appearance in Fig. 5 of the in vivo MHD capsid II-concatemeric DNA complex after DNA-free MHD capsid II suggests that the MHD capsid II bound to concatemeric DNA has at least partially packaged the DNA to which it is bound. DISCUSSION
Characterization of the capsids in the T7 in vivo capsid-concatemeric DNA complex and the capsids in the T7 in vivo capsid-
monomeric DNA complex revealed the envelopes of these capsids to be indistinguishable from the envelope of MHD capsid II. The term MHD capsid II does not indicate the presence or absence of components internal or external to the envelope of the capsid. Electron micrographs suggest the variable presence of: (1) the internal core, and (2) an external projection on the envelopes of the capsids of the above capsidDNA complexes. Because capsids are not necessary for the production of T7 concatemers (Studier, 1972; Hausman and LaRue, 1969), it is likely that concatemers are made before capsids are bound to them. Therefore, the data indicating that in vivo MHD capsid II-concatemerit DNA complexes are DNA-packaging intermediates (Results) and that capsid I binds DNA during the initial stages of DNA packaging (Serwer, 1980) also suggest that the DNA bound by capsid I is a concatemeric DNA. The data indicating that MLD capsid II occurs before MHD capsid II in the T7 DNA packaging pathway (Serwer, 1980; Results) indicate that the DNA bound by MLD capsid II is also a concatemeric DNA (see Fig. 6). None of the capsid-concatemeric DNA complexes (or capsid-monomeric DNA complexes) isolated here had any detectable capsid I or MLD capsid II. The following are possible explanations: (1) Concatemers that have bound these capsids are part of a larger complex, possibly the rapidly sedimenting, replicating DNA (100 S+ DNA) described in Serwer (1974a). (2) Complexes of capsid I and MLD capsid II with DNA dissociate during lysis or fractionation using the conditions in Materials and Methods. ‘(3) For capsid I it is possible that the conversion to MLD capsid II is too fast to observe the capsid I-DNA complex. It has been shown that capsids are bound to 100 S+ DNA; more than 95% of the capsids bound to 100 S+ DNA have capsid II-like envelopes by electron microscopy and agarose gel electrophoresis; no detectable proteins (~5%) bound to this DNA had the 8in metrizamide density gradients of MLD capsid II (P. Serwer, unpublished observations). Explanation 2 or (and) 3 above is, therefore, probably correct. Attempts are currently being
made to isolate the complex of capsid I with DNA and the complex of MLD capsid II with DNA. The data suggesting that the in viva MHD capsid II-concatemeric DNA complexes are DNA packaging intermediates which had packaged DNA prior to lysis suggest that at least some T7 DNA is packaged before it is cut to monomeric size; this proposal was previously made (Serwer, 19’7413) using a less complete experimental justification than used here. It is possible, therefore, that the signal for initiating the final cut in T7 DNA is the packaging of some amount of DNA. It has previously been proposed that the packaging of a monomeric length of bacteriophage T4 DNA is the signal for cutting T4 DNA to monomeric size from a concatemerit DNA (Streisinger et al., 1967). The existence of viable T7 deletion mutants with less DNA than wild-type bacteriophage Ti’ (Simon and Studier, 1973) indicates that less than a wild-type monomeric amount of packaged DNA is sufficient to signal the final cut in T7 DNA, although the making of the final cut may be slowed if a wild-type monomeric amount of DNA is not packaged. Experimental determination of the extent to which DNA was packaged (prior to lysis) in the capsids of the in vivo capsid-DNA complexes cannot yet be made. Assuming that the time of the final cut in T7 DNA and the time of the completion of DNA packaging are indistinguishable, and assuming that MLD capsid II has never packaged DNA (Serwer, 1980), the difference between the time of appearance of the in Gvo MHD capsid II-monomeric DNA complex and the time of appearance of MLD capsid II (1.5 min; Fig. 5) is an upper limit for the time required to package T7 DNA. If the above assumptions are correct, the data in Fig. 5 also indicate that the time used to accomplish post-DNA packaging steps in T7 assembly is greater than the time used to package T7 DNA. The hole in the envelope of MHD capsid II through which DNA enters the capsid may be contiguous with an axial hole of the T7 core observed in particles of capsid I (Serwer, 1976) (the data indicate that the core of capsid I is present in MHD capsid II during assembly and that this core disag-
gregates and possibly leaks out of most particles of MHD capsid II during lysis or fractionation; Serwer, 1980). The first demonstrated example of a preferential binding of T7 DNA near the core of a capsid is the observation in the Results section of monomeric and concatemeric DNA molecules apparently attached to the envelope of MHD capsid II at the same site as the internal core. DNA may also bind capsid I at the core-joining region of the envelope of capsid I during the initial stage of DNA packaging. However, the protein(s) of capsid I that binds DNA are not known and are not necessarily the same as the protein(s) that binds DNA in the in vivo capsid-DNA complexes. For instance P19, a protein which appears necessary for binding DNA to capsid I (Roeder and Sadowski, 1977). has not been found in the in vivo capsidDNA complexes studied here (Results). The external projections observed on some of the capsids of the in vivo MHD capsid II-monomeric DNA complex are too long to be the T7 tail and might have components that are part of the DNA-binding apparatus of capsid I. For instance, these projections might contain proteins that: (1) are found in the internal core of capsid I, and (2) have moved to the exterior of the envelope of the MHD capsid II in the Gr. vivo MHD capsid II-monomeric DNA complex. Such a retractable region of the internal eore of capsid I may be necessary for the binding of capsid I to DNA during the initial steps of T7 DNA packaging. ACKNOWLEDGMENTS For assistance with some of the experiments reported here we thank Nancy 1,. Smith. For provision of the facilities used for electron microscopy we thank Dr. Edward G. Rennels, Department of Anatomy, The University of Texas Health Science Center at San Antonio. Support was received from grant GM24365 of the National Institutes of Health, and from a grant of the Robert A. Welch Foundation. REFERENCES BECKER,
events in the “in vitro” packaging of bacteriophage A DNA. Virology 78,291-305. BONNER,
detection method for tritium-labeled
nucleic acids in polyacrylamide gels. Eur. J. Biochim. 46,83-88. CASJENS, S., and KING, J. (1975). Virus assembly. Annu. Rev. Biochem. 44, 555-611. EARNSHAW, W. C., and CASJENS, S. (1980). DNA packaging by the double-stranded DNA bacteriophages. Cell, in press. HAUSMAN, R., and LARUE, K. (1969). Variations in sedimentation patterns among deoxyribonucleic acids synthesized after infection of Escher&a coli by different amber mutants of bacteriophage T7. J. Viral. 3,278~281. KEMPER, B., and BROWN, D. T. (1976). Function of gene 49 of bacteriophage T4 II. Analysis of intracellular development and the structure of very fastsedimenting DNA. J. Viral. 18, 1000-1015. LASKEY, R. A., and MILLS, A. D. (1975). Quantitative film detection of3H and 14Cin polyacrylamide gels by fluorography. Eur. J. Biochem. 56.335-341. LUFTIG, R., and HASELKORN, R. (1968). Comparison of blue-green algae virus LPP-1 and the morphologically related viruses G,,, and coliphage T7. Virology 34,675-678. MURIALDO, H., and BECKER, A. (1978). Head morphogenesis of complex double-stranded deoxyribonucleic acid bacteriophages. Microbial. Rev. 42, 529-576. ROEDER, G. S., and SADOWSKI, P. D. (1977). Bacteriophage T7 morphogenesis: phage-related particles in cells infected with wild-type and mutant T7 phage. Virology 76,263~285. SERWER, P. (1974a). Fast sedimenting bacteriophage T7 DNA from T’I-infected Escherichia coli. Virology 59, 70-88. SERWER, P. (1974b). Complexes between bacteriophage T7 capsids and T7 DNA. Virology 59.89-107. SERWER, P. (1976). Internal proteins of bacteriophage T7. J. Mol. Bial. 107, 271-291. SERWER, P. (1977). Flattening and shrinkage of bac-
teriophage T7 after preparation for electron microscopy by negative staining. J. Ultrastruct. Res. 58,235-243. SERWER, P. (1978). A technique for observing extended DNA in negatively stained specimens: Observation of bacteriophage T7 capsid-DNA complexes. J. Ultrustruct. Res. 65, 112-118. SERWER, P. (1980). A metrizamide-impermeable capsid in the DNA packaging pathway of bacteriophage T7. J. Mol. Biol. 138,65-91. SERWER, P., and PICHLER, M. E. (1978). Electrophoresis of bacteriophage T7 and T7 capsids in agarose gels. J. Viral. 28, 917-928. SIMON, M. N., and STUDIER, F. W. (1973). Physical mapping of the early region of bacteriophage T7 DNA. J. Mol. Biol. 79,249-265. STREISINGER, G., EMRICH, J., and STAHL, M. M. (1967). Chromosome structure in Phage T4 III. Terminal redundancy and length determination. Proc. Nat. Acad. Sci. USA 57,292-295. STUDIER, F. W. (1969). The genetics and physiology of bacteriophage T7. Virology 39,562-574. STUDIER, F. W. (1972). Bacteriophage T7. Science 176,367-376. STUDIER, F. W. (1973). Analysis of bacteriophage T7 early RNAs and proteins on slab gels. J. Mol. Biol. 79,237-248. SWANEN, M., and YIN, J. (1978). Studies of DNA packaging into the heads of bacteriophage lambda. J. Mol. Biol. 126, 333-346. WOOD, W. B., and KING, J. (1979). Genetic control of complex bacteriophage assembly. In “Comprehensive Virology” (H. Fraenkel-Conrat and R. R. Wagner, eds.), Vol13, pp. 581-633. Plenum, New York. YAMAGISHI, H., and OKAMOTO, M. (1978). Visualization of the intracellular development of bacteriophage A, with special reference to DNA packaging. Proc. Nat. Acad. Sci. USA 75, 3206-3210.