Purification and characterization of the protein product of gene 11 of bacteriophage T4D.

Abstract The protein product of gene 11 of bacteriophage T4D was purified to apparent homogeneity from a bacteriophage culture in which this particular structural protein had not yet become incorporated into the mature bacteriophage particle. Monomer and apparent trimer molecular weights were determined to be approximately 24,000 and 70,000, respectively, and an amino acid analysis was performed. The kinetics and stoichiometry of interaction between the product of gene 11 and 11- defective bacteriophage particles were examined.

The protein product of gene 11 of bacteriophage T4D was purified to apparent homogeneity from a bacteriophage culture in which this particular structural protein had not yet become incorporated into the mature bacteriophage particle.
Monomer and apparent trimer molecular weights were determined to be approximately 24,000 and 70,000, respectively, and an amino acid analysis was performed. The kinetics and stoichiometry of interaction between the product of gene 11 and ll-defective bacteriophage particles were examined.
In the study of the molecular details of the assembly of biological structures, a potentially useful approach is to isolate the relevant macromolecular components after they have been synthesized but prior to incorporation.
With such material in hand, one is then in a position to study, by means of the many physical and chemical techniques currently available, the precise mode of interaction which leads to the biologically meaningful macromolecular complex. The coliphage T4 system is particularly well suited for this sort of an endeavor because it is of an appropriate degree of structural complexity, its genetics have been very extensively studied, and the structural function of many genetically identified gene products have been described (1) as have the pathways by which gene products sequentially interact with one another to give the final virus structure (2). From the point of view of purification of structural proteins, again the T4 system is well suited because judicious use of amber mutations permits the production of specifiable gene products prior to incorporation into larger structures. That many of the phage assembly steps can occur in titro (24), as seen by the production of active phage particles following mixing of suitably matched amber defective lysates, allows the development of an assay system for proteins whose only detectable function is structural, as opposed to biochemical catalysis. * This work was supported in part by the Swiss National Foundation for Scientific Research.
$ Supported by an American Heart Association Advanced Research Fellowship.
The protein product of gene 11 (Pll) was chosen for an initial trial run of structural protein purification on the grounds that after a preliminary screening of a number of gene products, PI1 appeared most amenable, biochemically, to isolation. Investigations of Simon et al. (5) have shown that Pll function plays a critical role in irreversible adsorption or sheath contraction, or both, but their electron micrographs of defective particles did not reveal any visible morphological defect associable with the absence of Pll. Edgar and Lielausis (2) showed that PI1 is associated with tail assembly and that in the normal course of events contributes its function rather early in the assembly of this structure. However, unlike most other gene products, PI1 can also be added after the phage is completely assembled except for the addition of P12 (gene 12 product).
The latter can be added only after Pll has been attached to the phage particle.
P12 has been partially purified' (6). Hosoda and Levinthal (7), by means of changes in gel electrophoresis banding patterns of appropriate defective lysates, have shown that the products of genes 7, 10, and 11, all of which are known to contribute to base plate assembly (8,9), manifest some sort of interaction, which is yet to be clarified.
The genetic map position of gene 11 is known (1), and it has been shown (10) that this gene is part of a transcription which includes at least genes 9, 10, and 11 and is transcribed in that direction.
In this communication, the purification and some of the characteristics of PI1 are described.

MATERIALS AND METHODS
Bacteria and Bacteriophage-Escherichia coli Bep and CR63 were used, respectively, as the restrictive and permissive hosts in all the experiments-reported here. The following bacteriophage strains were used: T4D lO(amB255) and T4D 7(amB16), amber mutants of genes 10 and 7, respectively, which (12) and the half-cystine content was examined by amino acid analysis of a sample of Pll that had been performic oxidized at -10" according to the method of Hirs (13).

Assay
When a dilution series of Pll, whether in the form of a crude defective lysate or in a relatively pure state, is assayed, an approximately second order relationship is found between concentration and titratable activated bacteriophage. For any given 1 l-defective lysate, there is a clearly defined Pll concentration at which point the system is saturated, and greater Pll concentrations do not lead to higher titers of activated bacteriophage particles ( Fig. 1). However, in order to keep the presentation of results as simple as possible, the "activity" portion of elution profiles shall be reported in terms of activated bacteriophage titers. As there are very substantial differences between different preparations in respect to the efficiency with which they are activated (particularly between fresh and frozen defective lysates), direct quantitative comparisons between different elution profiles may not be made.  To keep an approximate account of activity recovery, a sample of the starting material in which Pll appeared to be stable was assayed along with each set of PI1 assays; in accord with the apparent second order nature of the activation, the square root of the raw data was taken as a measure of the relative Pll activities of the samples assayed; the recovery was then calculated relative to the starting material.

Pll PuriJication
An ammonium sulfate fractionation of the PI1 starting material, described under "Materials and Methods," was followed by a series of chromatographic separations, first on DEAE-cellulose and then Sephadex G-200, resulting in a material which appeared homogeneous by the criteria of sedimentation velocity and SDS gel electrophoresis.
All steps in the purification were carried out at 4'.
For each chromatographic step, an elution profile is presented; the activity recovery from the various stages of purification is summarized in Table I. Step 1: Ammonium Sulfate Fractionation-The high speed supernatant (starting material) (2600 ml) was brought to 20% saturation, centrifuged, and the pellet was discarded.
The supernatant was then brought to 40% saturation, centrifuged, and the pellet was resuspended in 560 ml of 0.01 M potassium phosphate buffer (pH 7.5) and dialyzed three times against 5-liter portions of 0.01 M potassium phosphate buffer, pH 7.5.
Step d: First Column Chromatography on DEAE-cellulose-A column (3.9 x 20 cm) of DEAE-cellulose was prepared and equilibrated with 0.01 M potassium phosphate buffer, pH 7.5. To this column was applied the dialyzed ammonium sulfate precipitate at a flow rate of 180 ml per hour; the column was then washed with 500 ml of 0.01 M potassium phosphate buffer (pH 7.5) and eluted with a linear gradient of 2 liters of potassium phosphate buffer (pH 7.5) to 2 liters of 0.25 M potassium phosphate buffer (pH 7.5) at the same flow rate. The elution profile is shown in Fig. 2. Fractions corresponding to 1130 to 1410 ml were pooled and diluted with 3 volumes of HzO.
Step 3: Second Column Chromatography on DEAE-cellulose-A column (1.3 x 25 cm) of DEAE-cellulose was prepared and equilibrated with 0.01 M potassium phosphate buffer, pH 7.5. To this column was applied the diluted material from Step 2 at a flow rate of about 30 ml per hour, washed with 20 ml of 0.01 M potassium phosphate buffer (pH 7.5), and eluted with a linear gradient of 500 ml of 0.01 M potassium phosphate buffer (pH 7.5) to 500 ml of 0.1 M potassium phosphate buffer, pH 7.5. The elution profile is shown in Fig. 3A. Fractions corresponding to 440 to 620 ml were pooled and dialyzed against 0.01 M potassium phosphate buffer (pH 7.5). The shape of the activity curve suggests that the assay system was saturated by PI1 from the peak fractions.
Step 4: Third Column Chromatography on DEAE-cellulose-Precisely the same procedure was followed as in Step 3. The elution profile is shown in Fig. 3B. Fractions corresponding to 475 to 600 ml were pooled, diluted with 3 volumes of HzO, and used for Step 5.
Step 5: Fourth Column Chromatography on DEA.@cellulose-Precisely the same procedure was followed as in Step 3. The elution profile is shown in Fig. 3C. Fractions corresponding to 485 to 580 ml were pooled, diluted with 3 volumes of HzO, and used for Step 6.
Step 6: Concentration on DEAE-cellulose and Chromatography on Sephuok G-2m-A column (1 x 3 cm) of DEAE-cellulose was prepared and equilibrated with 0.01 M potassium phosphate buffer (pH 7.5). To this column was applied the diluted material from Step 5, and, after carefully washing with 10 ml of 0.01 M potassium phosphate buffer (pH 7.5), eluted with l-ml aliquots of 0.1 M buffer (pH 7.5). Fractions 3 to 7 (5 ml) containing the bulk of the material, were pooled and applied to a column (4 x 45 cm) of Sephadex G-200, previously equilibrated with 0.1 M potassium phosphate buffer (pH 7.5). The elution profile is shown in Fig. 4. The fractions corresponding to 255 to 325 ml were pooled and concentrated as described above, yielding 5 ml of material The second, third, and fourth chromatographic runs are shown in A, B, and C, respectively.
The columns were prepared, the samples were applied, and gradient elution was carried out as described in the text. Approximately 26ml fractions were collected.
Pll activities are presented as the raw data of the complementation assays, corrected only by the appropriate dilution factors.
In C, the bacteriophage titer scale appears to be very high, which is primarily due to the use of a fresh gene 11 defective lysate for assaying, as opposed to the stored frozen preparations used hitherto.
The run was prematurely terminated by a power failure. The sample and column were prepared as described in the text. The sample was carefully applied to and washed into the top of the column, and elution was carried out at maximal flow rate with almost zero hydrostatic head above the surface of the gel bed; 5-ml fractions were collected.
Pll activities are presented as raw data of the complementation assays, corrected only by the .
. appropriate dilution factors. As in Fig. 3C, a fresh gene 11 defective lysate was used, yielding the high bacteriophage titers.
The excluded and total column volumes were measured to he 185 and 1330 ml, respectively.
with an Azso of 5.2. This final concentrated material was used for subsequent experiments.
Summary- Table  I summarizes the purification of Pll in terms of relative activities recovered at key steps in the purification scheme. There are several points which deserve mention.
(a) A very high recovery is recorded from the first DEAE-column, which may reflect either serious quantitative inadequacies of the Pll assay or purification away from inhibitory materials. This problem has not been further investigated.
(b) Judging from the specific activity figures, Steps 3,4, and 5 add little or nothing to the purity of Pll, although succeeding elution patterns improve in quality.
Accordingly, one of these steps would be abandoned in future preparations of this protein.

Analytical
Ultracentrijugation-An aliquot of the purified preparation at a concentration of approximately 7.3 mg per ml was examined in a sedimentation velocity nm in order to determine the homogeneity of the sample and to measure the sedimentation constant. The s20,W value was calculated to be 3.8 S. Runs done with material at 34 and s the concentration yielded the same value. Fig. 5 shows the frame taken at SO min.
Absorption Spectrum-PI 1 yields a characteristic protein spectrum in the region of 250 to 300 nm at pH 7.5, with a A2so:A260 ratio of 1.9. Amino acid analysis of samples of known absorb- The frame shown was taken at 80 min with a bar angle of 50". measure of the purity of the final material as well as an indication of its molecular weight, the final material was subjected to SDS gel electrophoresis, along with the molecular weight standards of egg white lysozyme, hemoglobin, trypsin, pepsin, and bovine serum albumin (14). All six were run separately to determine their mobilities under these conditions of electrophoresis and they were all run together on a single gel in order to measure more precisely the mobility of the PI1 relative to the others. Pll was found to migrate as a single homogeneous band (Fig. 6A) when run alone, and corn&rated with trypsin (Fig. 6B) when rnn along with the collection of molecular weight standards. The latter observation indicates a monomer molecular weight value of about 24,000.
Gel Filtration-To obtain a molecular weight estimate of PI1 in its native form (corresponding to the 3.8 S value observed in the sedimentation velocity analysis), a calibrated gel filtration was done on a column (2 x 46 cm) of Sephadex G-200 in 0.01 M potassium phosphate buffer (pH 7.5). Bovine serum albumin and trypsin, in addition to bromphenol blue and blue dextran, were used for calibration of the column. The elution positions of the standards were plotted against the log molecular weight (Fig.  6C) ; the peak of Pll was found to elute several milliliters in advance of bovine serum albumin, which indicates a native molecular weight in the neighborhood of 70,000 (15).
As Pll activity was not recoverable from the low molecluar appeared which corresponded to the bulk of 280 nm absorbing weight band seen in SDS electrophoresis, the question could be material eluting in the excluded volume (Fig. 7A). The active raised as to whether that band indeed represented Pll.
To in-fractions were pooled, concentrated, and run on a calibrated vestigate this point, a fresh small scale (8 liters) preparation of Bio-Gel P-300 column equilibrated with 6 M urea in 0.1 M potas-Pll wae made and carried through the ammonium sulfate stage sium phosphate at pH 6.3. Individual fractions were concenof purification.
This material was run directly on a calibrated trated, dialyzed to free them of urea, and assayed. The main Bio-Gel P-150 column (2 x 40 cm) equilibrated with 0. (The broad low "peak" is an artifact of destaining encountered with all of the gels in this series of runs.) B, SDS gel electrophoresis (as described in A) of Pll and the molecular weight standards of egg white lysozyme,

104
Molecular wetght hemoglobin, trypsin, pepsin, and bovine serum albumin, were done individually and in combination. Log molecular weights of the standards are plotted against their relative mobilities.
C, a column (2 X 40 cm) of Sephadex G-260 was equilibrated with 0.01 M phosphate buffer (pH 7.5). The elution positions of the molecular weight standards trypsin and bovine serum albumin were separately determined and plotted against log molecular weight.
The elution positions of Pll, blue dextran and bromphenol blue were also determined and are indicated on the standard curve. A, a column (2 X 40 cm) of Bio-Gel P-150 was equilibrated with 0.1 M potassium phosphate buffer at pH 6.0 and 10". The elution positions of the molecular weight standards myoglobin and bovine serum albumin were separately determined and plotted against log molecular weight.
The elution positions of blue dextran (excluded volume) and bromphenol blue (included volume) are also shown. A crude preparation (26 to 46% ammonium sulfate cut of a lo-defective lysate) was chromatographed on this column and the fractions assayed for Pll activity.
B, a column (2 X 40 cm) of Bio-Gel P-300 was equilibrated with 6 M The elution positions of the dye markers are also shown. The pooled active samples from the column run described above (A) were concentrated and chromatographed.
Fractions were concentrated and dialyzed against the standard 0.1 M potassium phosphate buffer (PH 7.5) and assayed in the usual manner.
The major peak of activity eluted in the position indicated. again came through in the excluded volume (Fig. 7B).
Amino Acid Analysis-In Table II is shown the approximate amino acid composition of purified PI 1. The estimates are based on the average or extrapolated molar ratios for each amino acid from the samples subjected to, respectively, 48, 72, and 96 hours of hydrolysis.
The normalization value of 19.2 for glycine was chosen on the grounds that the corresponding sum of residue weights is appruximately the 24,000 estimated from the analytical SDS gel electrophoresis for the monomer. The one striking feature of the amino acid composition of this protein is the apparent complete absence of the sulfur-containing amino acids, cysteine and methionine. This observation was confirmed in an analysis of a performic-oxidized sample of this protein.
Kinetics and Stoichiometry of Interaction of Pll and 1 l-Defective Particles-A a-fold dilution series of Pll was made, covering a concentration range of greater than IO*, and the kinetics of activation was measured with each aliquot in the series. In Fig.  8A are presented three characteristic curves observed with, respectively, concentrated, moderately dilute, and very dilute Pll.
In Fig. 8B are plotted (as a function of relative PI1 concentration) the maximum rates of activation and the end points, respectively.
In the lower portion (more dilute Pll) of both curves, a 1.5 order concentration dependence is observed. A, an aliquot of purified Pll (O---O) 6.5 X 10eZ mg per ml, (X---X) 5.2 X 10e4 mg per ml, (A---A) 7.2 X lO+ mg per ml was mixed with an equal volume of an ll-defective lysate in which the particle titer was approximately 2.6 X 10" and incubated at 37". At the times indicated, 0.05-ml samples were taken into 5 ml of buffer and assayed.
The 108). Suitable calculation shows that approximately 430 Pll monomer units were required to activate a single bacteriophage particle.

DISCUSSION
The purification of Pll of bacteriophage T4 has been described, as has a few of its physical and biological properties.
That the material isolated is physically homogeneous is supported by the production of a single peak on DEAELcellulose chromatography and Sephadex G-200 chromatography, a single moving boundary in a sedimentation velocity analysis and a single band in SDS gel electrophoresis.
The identification of this material as PI1 is shown by the congruence of the peak of PI1 activity and the peak of optical density in both types of chromatography, and a molecu-lar weight of about 70,000 has been tentatively assigned to this species on the basis of its elution position in calibrated Sephadex G-200 runs. The sedimentation value of 3.8 S is not inconsistent with that assignment.
That the molecular weight deduced from the calibrated SDS gel electrophoresis is 24,000 suggests that the material originally isolated was multimeric, perhaps a trimer. Although we have not yet been able to demonstrate definitively a congruence of Pll activity with purified material behaving physically as 24,000 daltons, two lines of evidence support this contention. Using a crude preparation of Pll, we were able to show that, when run on a Bio-Gel P-300 column in 6 M urea, there is a peak of PI1 activity corresponding to a molecular weight of about 27,000.
Furthermore, experiments of others (16) have demonstrated that when the SDS gel patterns of a W-labeled nondefective bacteriophage lysate are compared with those derived from an 1 l-defective lysate, the latter differs from the former only in the lack of a single band at a position corresponding to a molecular weight of approximately 26,000. Thus we are confident that the single low molecular weight band observed with the purified material in SDS gel electrophoresis is indeed Pll.
One chemical feature of this protein worthy of note is the failure to detect the sulfur-containing amino acids, cysteine and methionine.
The functional significance of this observation is unknown, but it may be of some practical interest in further work with this protein.
The data (presented in Fig. 8B) relating to the kinetics and stoichiometry of interaction of Pll and ll-defective bacteriophage particles present some problems in interpretation. That at relatively low PI 1 concentrations the maximum rate of activation and end point titers manifest a similar Pll concentration dependence (formally, 1.5 order) is a satisfying internal consistency. The end point titer curve shows a sharp break at the point which represents an equivalence of PI1 and ll-defective bacteriophage particles.
The maximum rate of activation curve continues to rise, however, with approximately the same order of concentration dependence to a Pll concentration several times equivalency, where there is a marked change toward Pll concentration independence. This probably indicates that another component of the system is becoming rate-limiting.
P12, which is known to attach obligatorily to the bacteriophage particle after PI1 has become attached (2,6), is a likely candidate for the third component.
This has not been investigated further. Using the established molecular weight of 24,000 and the observation that a solution of PI 1 with an Azso of 1 .O has a concentration of 1.3 mg per ml, it is possible to calculate that one bacteriophage equivalent of PI1 is 430 monomer units (or about 140 trimer units).
This figure is grossly out of line, since ll-defective particles are physically indistinguishable from complete particles (5). Furthermore, the reaction order of 1.5 is difficult to reconcile with this large a number of molecules.
That the end point titer of activated bacteriophage is in accord with electron microscope counts eliminates the possibility that a large number of "dead" particles are nonproductively adsorbing Pll. The most likely explanation is that only a small fraction of the isolated Pll is biologically active, yet physically indistinguishable (by the limited criteria employed in this investigation) from the bulk inactive form. The observation (7) that Pll interacts with PI0 5125 during the course of bacteriophage development may be of some relevance, since, in the isolation reported here, the starting material was a lo-defective lysate.
Although, as of this writing, this has not been investigated further, it is anticipated that Pll will be isolated from a defective lysate other than lo-, as well as from complete bacteriophage particles. Examination of the chemical and biological properties of PI1 isolated from these alternate sources may be expected to shed some light on the high equivalence number observed in the experiments reported in this communication.
It is anticipated further that such experiments shall yield information concerning changes in protein structure upon incorporation of the nascent polypeptide chain into the bacteriophage structure.
The assay system for PI1 described in this report was necessarily crude and qualitative.
As shown by an investigation of optimal pH and ionic concentration, 0.1 M phosphate buffer at pH 6.2 would have been slightly preferable to the 0.07 M phosphate buffer at pH 7.5 actually used. Furthermore, the overnight (12-to 18-hour) incubation was not necessarily sufliciently long to give an end point titer of activated bacteriophage, particularly at the more dilute end of the range.
However, even as performed, there was sufficient internal consistency to make this a valid and valuable qualitative assay for the purposes of identification of Pll in the separation procedures.
That the assay as performed manifested a second order PI1 concentration dependence (which was used in recovery and purification calculations), compared with the 1.5th order observed in the careful end point titrations, is attributable to the relatively short term incubations performed in the former case.