the orientation of the major coat protein of bacteriophage ... · pdf filethe orientation of...

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 256. No. 19. Issue of October 10, pp. 9951-9958, 1981 Prmted in U.S.A.

The Orientation of the Major Coat Protein of Bacteriophage f l in the Cytoplasmic Membrane of Eschen’chia C O W

(Received for publication, May 18, 1981)

Izumi Ohkawa and Robert E. Webstert From the Department of Biochemistry, Duke University Medical Center, Durham, North Carolina 27710

The orientation of the major coat (B) protein of the bacteriophage f l , an integral membrane protein in the cytoplasmic membrane of infected Escherichia CoZi, was examined. Pyridoxal 5’-phosphate and [‘H]NaB& were used to label the cytoplasmic membrane proteins in spheroplasts and membrane vesicles of E. coli infected with bacteriophage fl. Under the conditions described, tritium incorporation was almost completely dependent on the presence of pyridoxal 6’-phosphate and little if any of the cytoplasmic proteins were labeled when the reaction was applied to intact spheroplasts. The major coat protein was isolated from the cytoplasmic membranes labeled in this manner and the chymotryptic peptides were analyzed for the presence of tritium in the pyridoxamine 5’-phosphate conjugate. When the proteins were labeled in the intact spheroplast, only the MIz-terminal chymotryptic peptide of the coat protein was labeled. If the proteins were labeled during osmotic lysis of the spheroplasts or in isolated vesicles, the chymotryptic peptide containing the COOH terminus of the coat protein as well as the Mia-terminal peptide was labeled. The NHz-terminal peptide was labeled to approximately the same extent as occured in the intact spheroplast. These results are consistent with the hypothesis that the mature fl coat protein asymmetrically spans the cytoplasmic membrane of the infected host with its NHz terminus exposed on the outside and COOH terminus exposed on the cytoplasmic surface.

The fl filamentous bacteriophage consists of a circular single-stranded DNA molecule covered along its length with approximately 2700 molecules of the phage-encoded major coat protein (1). At the ends of the virion are small amounts of four other minor coat proteins coded by the phage genome (2-4). Within the infected Escherichia coli, a large amount of the major coat protein is synthesized as precursor molecules containing an additional 23 amino acid residues at the NH2- terminal end (1, 5). After insertion of the precursor molecule into the cytoplasmic membrane, the NHt-terminal portion is removed by specific proteolytic cleavage (6,7) and the mature coat protein remains as an integral membrane protein until it is assembled into a mature phage particle (8, 9). This mode of synthesis of the major coat protein together with its chemical (10-13) and physical (14, 15) characteristics in detergent micelles and phospholipid vesicles suggests that the coat protein behaves as a typical membrane protein. Therefore, the study

* This work was supported by Public Health Service Grant GM 19305. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

8 Recipient of an American Cancer Society faculty research award.

of this protein has been used as a model system to understand better the biosynthesis of membrane proteins in procaryotes. Some of these studies have suggested that synthesis of the coat protein involves co-translational insertion of the precursor coat protein into the membrane followed by cleavage of the precursor portion (6, 16, 17), a mechanism which is consistent with the signal hypothesis of Blobel and Dobberstein (18). Other workers have concluded that the precursor molecule is completely synthesized in the cytoplasm and then is inserted into the membrane followed by proteolytic cleavage (7, 19, 20). This is in accord with the membrane trigger hypothesis put forth by Wickner (21).

Both of these hypotheses assume that, following synthesis, insertion, and processing, the mature coat protein asymmetrically spans the cytoplasmic membrane with the NHz-ter- mind 20 amino acids exposed on the outside and the COOH- terminal 11 amino acids on the cytoplasmic surface (22, 23). The observation that antibodies directed against the NHt- terminal portion of the coat protein bind to cells or spheroplasts of infected bacteria is supportive of this hypothesis (24). In addition, it recently has been shown that the NHt-terminal portion of the mature coat protein can be removed from sucrose/EDTA-treated fl-infected bacteria by treatment with proteases (25). However, there is no definitive in vivo data on the location of the COOH-terminal portion of the coat protein in infected bacteria; it has been inferred, based on a number of in vitro observations by various groups, that it is exposed on the cytoplasmic surface of the membrane. Analysis of purified coat protein in detergent micelles or artificially formed phospholipid vesicles showed that the coat protein could assume an orientation which spanned the membrane but did not rule out any U-shape orientation of the protein where both the COOH and NHz termini would be present on the same membrane surface (10-13). Also, synthesis of the coat protein in an in vitro system in the presence of inverted membrane vesicles resulted in vesicles containing the coat protein where only the COOH-terminal region appeared susceptable to proteolysis (16).

In order to understand better the orientation of membrane- associated coat protein, we decided to react intact and lysed spheroplasts with a membrane-impermeable reagent and then to analyze the coat protein from each reaction for the location of modified residues. A number of such reagents have been used to assess the orientation of membrane proteins in various systems (26). The only study with E. coli used a~e ty l [~~S] methionyl methylphosphate sulfone to label extracellularly the nacent polypeptides traversing the cytoplasmic membrane (27). Another reagent is pyridoxal 5’-phosphate, which has been shown to form stable pyridoxine 5”phosphate conjugates with a- and €-amino groups in proteins following reduction with NaBH4 (28, 29). Rifkin et al. (30) fist made use of this observation to locate the proteins on the exterior of the influenza virus membrane. A number of groups now have used

9951

9952 Membrane-associated Bacteriophage fl Coat Protein

this reagent to determine the orientation of membrane proteins with various degrees of success (31-36).

In this paper, we report a procedure for labeling the proteins exposed on the outside of the cytoplasmic membrane of fl- infected E. coli spheroplasts, using pyridoxal Ei'-phosphate and [3H]NaBHr. Analysis of the fl major coat protein from these membranes showed that only the NHz-terminal portion of the protein had been labeled. The COOH-terminal portion can ody be labeled in membrane vesicles where both sides of the membrane are exposed. These results are consistent with the hypothesis that the mature coat protein asymmetrically spans the cytoplasmic membrane of the infected host.

EXPERIMENTAL PROCEDURES

Materials-Pyridoxal 5'-phosphate, NaBH4, egg white lysozyme, and bovine albumin were obtained from Sigma. [3H]NaBH4 (64 and 7.6 Ci/mmol), ~-['~C]lysine (342 mCi/mmol), ~-[4,5-~H]lysine (78.1 Ci/mmol), ~-['~C]proline (288 mCi/mmol), and Protosol were purchased from New England Nuclear. Both the tritiated and nonradioactive NaBH, were dissolved in 10 mM NaOH and used fresh or stored at -70 "C. Triton X-100 was from Research Product Interna- tional Corp. and SDS' was from BDH Biochemicals, Ltd. Deoxycho- late was purchased from Fisher Scientific and recrystallized from acetone according to Makino et al. (IO). Deoxyribonuclease 1 and a- chymotrypsin were from Worthington. Dialysis tubing was purchased from Spectrum Medical Industries, Inc., and Bio-Gel P-2 was from Bio-Rad Laboratories. Sources of all other materials have been given by Lin et al. (3).

Growth of Bacteria and Bacteriophage-E. coli strains K38, a nonsuppressing strain, and K37, a supD suppressor strain, were used (37). R12 phage, which contains an amber mutation in gene IV (37), was grown on K37, labeled with specific amino acids, and purified according to Lin et al. (3). Only stocks of R12 phage were used which had reversions of less than 2 X

Preparation of Spheroplasts from Infected Bacteria Labeled with CL4C]Lysine--K38 was grown at 37 "C in 250 ml of MTPA (38) supplemented with 0.2% glucose and a 1 mM concentration of each of amino acids. When the culture was 2 X 10' cells/ml, the bacteria were collected on an Amicon Microporous filter (0.45 pm pore size), resuspended in 250 ml of supplemented MTPA containing 20 pM L-['~C] lysine (12 pCi/pmoI), and incubated at 37 "C with aeration.Five min later, the culture was infected with R12 at a multiplicity of infection of 90 and, after an additional 17 min, was quickly chilled to 4 "C. All subsequent operations were done at 4 "C unless otherwise stated. The bacteria were couected by Ntration, washed, and resuspended in 225 ml of supplemented MTPA. After blending for 1 min in a Sorvall Omni-Mix or at setting 5, the bacteria were collected by centrifugation at 10,OOO X g for 7 min. The filtering and blending steps removed about 95% of the input bacteriophage.

The washed bacteria were resuspended in 15 ml of 0.75 M sucrose, 10 mM Tris HCI, pH 7.8, and converted to spheroplasts by the method of Osborn et al. (39). The formation of spheroplasts was monitored by phase contrast microscopy and almost all bacteria appeared to be converted to round spheroplasts. Equal amounts of the spheroplast suspension (approximately IO9 spheroplasts/ml) were layered on each of four tubes containing 9 ml of S buffer (0.5 M sucrose, 50 mM NaC1, 25 mM Na2B407, pH 8.0) and the spheroplasts were pelleted at 6900 x g for 10 min in a swinging bucket rotor. Two of the four pellets, samples A and B, were gently resuspended in 0.3 ml of S buffer and kept as spheroplasts for reaction with [3H]NaBH4 in the presence and absence of PLP. The other two pellets, samples C and D, were first lysed, so that both sides of the membrane could be reacted with [3H] N a B R in the presence and absence of PLP (see below).

Reaction of Intact Spheroplasts and Membranes with PLP and [3N]NaBH4-Samples A and B were pretreated with 2 pl of 0.1 M NaBHd for 10 rnin at 22 "C to reduce any reactive aldehydes or ketones in the solutions. Then 50 pl of a 40 mM PLP solution in S buffer was added to sample A, to give a final concentration of approximately 5 mM PLP. As a control, 50 pl of S buffer was added to sample B. After 10 min incubation at room temperature, 20 p1 of 0.05 M I3H]NaBH4 was added to each sample (64 mCi of 3H/sample), the

I The abbreviations used are: SDS, sodium dodecyl sulfate; PLP, pyridoxal 5-phosphate; DNase I, deoxyribonuclease I; MTPA, a min- imal amino acid media defined by Vfiuela et al. (38).

incubation was continued for another 10 min, and the reactions were terminated by the addition of 10 ml of ice-cold S buffer containing 1 mM PLP, 10 mM lysine, 2 IDM MgClz, and 25 pg/ml of DNase 1. The DNase I was present to hydrolyze any DNA resulting from spheroplast breakage. The small amount of aggregates was allowed to settle a t 0 "C and the supernatants, containing the majority of the spheroplast, were layered onto 20 ml of S buffer containing 0.6 M sucrose and the above concentrations of PLP, lysine, MgClz, and DNase I; the intact spheroplasts were collected by centrifugation for 10 min at 8000 X g. Approximately 60% of the spheroplasts survived the reaction and subsequent centrifugation. The spheroplasts were lysed by resuspension in 3 ml of 1 mM PLP, 10 mM lysine, 25 mM Na2B407, pH 8.0, and debris was removed by centrifugation for 20 min at loo0 x g. The membranes were then Collected by centrifugation at 240,000 x g for 90 min through 1 ml of S buffer on a 0.1-ml shelf of S buffer containing 55% (w/v) sucrose.

Spheroplasts in the C and D pellets were lysed by resuspension in 3 ml of 25 mM NazB407, pH 8.0, the unbroken cells were removed by centrifugation at lo00 X g for 20 min, and the membranes were collected by centrifugation as described above. The bottom 0.35 ml was collected from each sample and treated with 2 pl of 0.1 M NaBH4 to reduce endogenous aldehydes and ketones. Then the membranes in sample C were reacted with PLP and ['HINaBHr and those in sample D were reacted with buffer and [3H]NaBH4 exactly as described above for the intact spheroplasts. Each suspension was pushed through a 23-gauge needle after the addition of each reagent to ensure the availability of the reactants to both sides of the membrane.

Isolation of the fl Major Coat Protein from the Cytoplasmic Membrane-The membranes from samples A-D were made up to 3 d with 1 mM PLP, 10 mM lysine, 0.2 M NaCl, 0.1 m g / d of lysozyme, 25 RIM NazB407, pH 8.0, and passed through a 23-gauge needle several times. Lysozyme and NaCl were present to remove most of the lysozyme which was labeled in the above reaction. The membranes were collected by centrifugation at 300,000 X g for 90 rnin through 1 ml of 15% sucrose (w/v) 0.2 M NaCl, 5 m~ EDTA, 10 mM Tris-HCl, pH 7.6, onto a 0.2-ml shelf of 55% (w/v) sucrose in the same buffer. The membranes on the shelf were washed again by suspension in 3.6 ml of 10 mM Tris-HC1, pH 7.6, using a 23-gauge needle, collected by centrifugation at 92,000 X g for 90 min, and allowed to resuspend overnight at 4 "C in 50 pl of 2 mM MgClz, 10 mM Tris-HC1, pH 7.6. The next morning, 10 pl of the same buffer containing 20% (w/v) Triton X-100 was added and the cytoplasmic membrane proteins were allowed to solubilize a t room temperature for 30 min (40). An additional 1 ml of buffer without Triton X-100 was added, and the insoluble outer membrane proteins were removed by centrifugation at 208,000 X g for 3 h. Carrier bovine serum albumin (30 pg) was added and the cytoplasmic proteins were precipitated from the su- pernatant by the addition of 0.2 volumes of 50% trichloroacetic acid. The precipitate from each sample was collected by centrifugation and dissolved in 0.1 mI of 6 M urea, 8% SDS, 5% P-mercaptoethanol, 5% glycerol, 0.125 M Tris-HC1, pH 6.8, and the residual trichloroacetic acid was neutralized with NaOH.

Each sample was heated at 100 "C for 5 min and then subjected to electrophoresis on a 16.8% polyacrylamide gel in the presence of 8 M urea and SDS as described by Li et al. (3). The regions containing the coat protein were excised and placed in 1 ml of deoxycholate buffer (15 mM deoxycholate, 50 mM H3B03, 12.5 IDM Na2B407, pH 8.5), and the coat protein was eluted by incubation with shaking for 2 days at room temperature. The eluted coat protein was dialyzed 3 days against deoxycholate buffer using Spectrapor 6 dialysis tubing. Throughout the labeling and isolation procedure, samples were kept in the dark as much as possible to avoid any photochemical release of the pyridoxyl residue from the protein (41).

To analyze the overall pattern of labeling of cytoplasmic membrane proteins, the membranes were extracted with Triton/Mg2' as above and the solubilized cytoplasmic membrane proteins were dialyzed against 0.1% SDS, 10 mM Tris-HC1, pH 6.8, and concentrated by Iyophylization. The samples were resuspended in 0.1 ml of 5% p- mercaptoethanol, 5% glycerol and a portion was subjected to electm- phoresis on a 10% acrylamide, 0.15% bkacrylamide slab gel (1.5 mm thick x 30 cm long) under the same conditions as previously described (3). Each lane was divided into 3-mm slices and the radioactivity in each slice was determined using the Protosol method of L h et at. (3).

Reaction of Purified Coat Protein with PLP and f3HJiVaBH4- Radioactive bacteriophage labeled with specific amino acids were grown and purified as described (3). The coat protein was solubilized and isolated by gel filtration on Sephadex G-150 in the presence of deoxycholate according to Woolford and Webster (42), and its con-

Membrane-associated Bacteriophage fl Coat Protein 9953

Ala-Glu-Gly-Asp-Asp-Pro-Ala-~~~-Ala-Ala-Phe-Asp-Ser-Leu-~ln-Ala-Ser- fl major coat protein (46, 46). Most FIG. 1. Amino acid sequence of the

of all of the underlined region interacts with the hydrocarbon portion of the bi-

30 layer (12,141. Lysine residues in boldface Ala-Thr-Glu-Tyr-Ile-Gly-Tyr-Ala-Trp-Ala-Met-Val-Val-Val-Ile-Val-Gly- type are probably available to react with

PLP. The arrows mark the positions most susceptable to cleavage by chymo-

40 + * trypsin in the presence of deoxycholate

10 +

20

Ala-Thr-Ile-Gly-Ile-Lys-Leu-Phe-~ys-Ly~-Phe-Thr-Ser-~y~-Ala-Ser 50 (12,42).

centration was determined by absorption a t 280 mm (43). The purified [“C]lysine coat protein at a concentration of 3.8 mg/ml (0.72 mM) was reacted with 10 m~ PLP in S buffer containing 15 mM deoxycholate. After 10 min a t room temperature, the solution was made 2.9 m~ C3H]NaBH4 (26 pCi/mmol), incubated for 10 additional min, and then dialyzed for 3 days against the deoxycholate buffer.

Chymotryptic Peptide Analysis of Coat Protein-Radioactive coat protein, either prepared from purified phage or from the polyacrylamide gel in deoxycholate buffer, was incubated with 0.2 m g / d of chymotrypsin at 37 “C. The same amount of chymotrypsin was added after 1 and 2 h of incubation and the digestion was terminated at the end of 3 h by lowering the pH to 5.0 by the addition of HC1. The resulting precipitate of deoxycholate and insoluble peptides was removed by filtration through a Whatman GFJC glass fiber filter. The filtrate, containing the soluble cleaved peptides, was adjusted to pH 7.6, and a 1.5-ml .sample was applied to a Blo-Gel P-2 column (0.5 X 62 cm) in 50 mM Tris HC1, pH 7.6. The peptides were eluted with the same buffer a t a flow rate of 6 ml/h and the radioactivity of each 0.56-ml fraction watt determined.

The NH2-terminal residue of each peptide of interest was determined to aid in their proper identiiication. Coat protein was treated with chymotrypsin as above, but the ratio of the final concentration of chymotrypsin to coat protein was 1:20. The peptides were isolated by gel filtration on Bio-Gel P-2 in 20 mM NaC1,lO mM H~B03.2.5 m~ Na2B407, pH 8.5, and the NHz-terminal residue of each peptide was determined by the technique of Weiner et al. (44).

RESULTS

Procedure for Modifying and Analyzing Membrane-associated Coat Protein-Numerous studies on the interaction of purified coat protein with detergent micelles and artifically formed phospholipid bilayers have defined three major do- mains (10-13, 42). There are two hydrophilic regions com- posed of the NHz-terminal 20 amino acids and the COOH- terminal 11 amino acids which interact with the aqueous environment outside the membrane and are susceptible to cleavage by proteolytic enzymes (Fig. 1). The remaining 19 amino acids (residues 21-39, Fig. 1) form a hydrophobic region of which most or all interact with the hydrocarbon part of the membrane bilayer. We attempted to determine the orientation of the two hydrophilic regions in the bacterial cytoplasmic membrane using the procedure outlined in Fig. 2 (see “Exper- imental Procedures” for details). K38 bacteria were infected with R12 in the presence of [’4C]lysine, washed to remove most of the free phage, and converted to spheroplasts. Bac- teria infected with R:12 mutant phage for only 17 min were used since they produce no phage particles, still synthesize coat protein, and produce no aberrant membrane structures (49,471. A portion of the spheroplasts was reacted with PLP and C3H]NaBH, in order to label the available lysines and NH2 termini of proteins only exposed on the outside of the cytoplasmic membrane (sample A). As a control for any non- PLP-dependent labeling of proteins by C3H]NaBH4, another sample of intact spheroplasts was treated with t3H1NaBH4 in the absence of PLP (sample B) . After these reactions, modified intact spheroplasts were recovered and the membranes were isolated from them. Another portion of the unreacted spheroplasts was lysed by osmotic shock and the resulting membranes were reacted with C3H]NaBH4 in the presence (sample

I

A D I FLF I I -Lyse- I

1 1 -PAGE- 1 1 -Chyliatrypsin- I 1 - I ” ? Coluwn- 1 1

MIDIFICI) 8. W n I F I E D PEPTIMS

COAlPROTUN

FIG. 2. Procedure for specific labeling and analysis of coat protein in the cytoplasmic membrane. Details are given in text under “Experimental Procedures” PAGE, SDS-polyacrylamide gel electrophoresis.

C) or absence (sample D) of PLP in order to label reactive groups of proteins exposed on both sides of the membrane.

The cytoplasmic membrane proteins were extracted from all four samples into a Triton X-100 solution containing Mg2+ (40) and the coat protein was separated from other membrane proteins by SDS-polyacrylamide gel electrophoresis. The coat protein was eluted from the gel into deoxycholate buffer and subjected to the action of chymotrypsin (12,42). The solution was made acidic to precipitate the deoxycholate and the peptide containing the hydrophobic region (Fig. 1). The acid- soluble hydrophilic chymotryptic peptides from the NH2 and COOH termini were identified by gel fitration on a Bio-Gel P-2 column. The difference in the tritium radioactivity of peptides in samples A and B and samples C and D was the amount of PLP-dependent modification which occured to the protein in intact spheroplasts and membrane vesicles, respectively.

The efficacy of this procedure requires that four criteria be met. First, reaction conditions must be determined which &ow PLP to react with a- and €-amino groups of coat protein in the spheroplasts. Second, it must be possible to identify chymotryptic peptides containitig residues 1-11, 43-45, and 46-50 so that the extent of the labeling at the NH2 and COOH ends of the protein can be analyzed (Fig. I). Third, conditions of the reaction should allow labeling of only proteins exposed on the outside surface of the spheroplasts. Last, all intact

9954 Membrane-associated Bacteriophage f l Coat Protein

phage particles which might have been present during the labeling procedure must be removed so that only membrane- associated coat protein will be analyzed. These criteria will be addressed in the following sections along with the results of the outlined procedure.

Determination of Reaction Conditions-We made use of the spectral characteristics of PLP (28, 48) to determine the optimal conditions for modifying proteins and keeping spheroplasts stable. In the presence of S buffer at 22 "C, conditions where spheroplasts are quite stable, we found that PLP has an absorption peak at 390 nm with a molar extinction coefficient of approximately 230 M"/cm. The addition of excess lysine shifted the peak to 410 nm and gave an extinction coefficient of 4,200 M"/Cm at this wavelength. Mixtures of PLP with lysine, polylysine, or glycine gave the same spectrum, indicating that Shiff base formation between PLP and either an a- or e-amino group resulted in the same spectral change under these conditions. However, a 10-fold higher concentration of glycine was required to give the same absorbance, indicating the lower reactivity of PLP for a-amino rather than e-amino groups.

The reaction of PLP with lysine in S buffer at 22 "C is shown in Fig. 3. The formation of the Schiff base between PLP and lysine, as measured by the increase in absorbance at 410 nm, reached equilibrium within 10 min over a large range of lysine concentrations (Fig. 3, curves 1 and 2). From this kind of data, we estimated the association constant of PLP with lysine to be about 340 M". Although PLP appeared to

1

rnin

FIG. 3. Reaction of PLP with lysine, or lysozyme followed by reduction with NaB&. PLP (0.2 mM) was mixed with 40 mM lysine (curue I ) or 2 mM lysine (curue 2) and 2 mM PLP was mixed with 1 mg/ml of lysozyme (curue 3), and the absorbance at 410 nm was measured as a function of time. At the time indicated by the arrow, NaBH4 was added to give a final concentration of either 0.5 mM (curue I ) or 1 mM (curue 3). All reactions were carried out in S buffer at 22 o c :

N I" - 1 i

have the same reactivity for lysine residues contained in polylysine (data not shown), it had an apparent lower reactivity for amino groups in proteins. Based on the data in Fig. 3 (curve 3), we calculated that only 16% of the amino groups in lysozyme were modified using the extinction coefficient found for the lysine. PLP complex. However, the reaction with proteins still reached equilibrium within 10 min. Addition of NaBH, to these reaction mixtures resulted in the reduction of PLP in both free and Schiff base form in 10 min (Fig. 3). Based on these observations, we decided to react the membranes with 5-10 mM PLP for 10 min at 22 "C followed by reduction for 10 min with 2.5-5 mM C3H]NaBH4. It was felt that this concentration of PLP was high enough to compen- sate for the apparent low reactivity of PLP for proteins but low enough to allow complete reduction by reasonable concentrations of r3H]NaBH4.

Analysis of Chymotryptic Peptides of Coat Protein-When coat protein is associated with deoxycholate micelles, the hydrophilic portions of the molecule have been shown to be susceptible to cleavage by proteolytic enzymes (12, 42). In order to identify the soluble chymotryptic peptides, coat protein labeled with various radioactive amino acids was subjected to chymotryptic digestion at pH 8.5 in the presence of 15 mM deoxycholate. The deoxycholate and insoluble peptides were removed by precipitation at pH 5.0 and the soluble peptides were separated by gel filtration as shown in Fig. 4a. The three ['4C]lysine-labeled peaks eluted at the positions expected for the three chymotryptic peptides containing residues 1-11, (peak I ) , 46-50 (peak 11) and 43-45 (peak 111) (see Fig. 1). The following observations c o n f i e d this iden- tification. First, only peak I contained proline while all three peaks had lysine (Fig. 4a) and the molar ratio of proline to lysine was 1, as expected. Similar experiments show that peak I1 contained serine while phenylalanine was only present in peaks I and 111. Second, the NH2 terminus of peak I was alanine, peak I1 was threonine, and peak I11 was lysine as determined using the microdansylation technique of Weiner et al. (44). In a typical experiment, 77, 85, and 50% of the theoretical yield of peaks I, 11, and I11 were obtained.

In order to determine the elution positions of these peptides with pyridoxine 5'-phosphate attached, ['4C]lysine-labeled coat protein was reacted with PLP and [3H]NaBH4 in S buffer containing deoxycholate as described under "Experimental Procedures." The results of the peptide analysis on this labeled coat protein is shown in Fig. 4b. Peaks 1-111, labeled with ['4C]lysine, represent the nonmodified peptides identified in Fig. 4a, while the tritium-labeled peaks 1'-111' represent the modified peptides which contain an additional 231 daltons

40 50 60 70 80 90 40 50 60 70 80 90

FIG. 4. Gel filtration of the soluble chymotryptic peptides. a, purified coat protein labeled with ['4C]proline and C3H]lysine 'was digested with chymotrypsin in deoxycholate buffer and the soluble peptides were analyzed by gel filtration through a Bio-Gel P-2 column as described under "Experimental Procedures." b, ['4C]lysine-labeled coat protein (3.8 mg/ml) in deoxycholate buffer was reacted with 10 mM PLP for 10 min at 22 "C, followed by reduction with 2.9 mM [3H]NaBH4 for 10 min. After dialysis against deoxycholate buffer, a portion was digested with chymotrypsin and the peptides analyzed as in a.

FRACTION FRACTION

Membrane-associated Bacteriophage f l Coat Protein 9955

7.5

5 0

c? -

2 5 X

E 9. v u o - 7 5

L2J I

1 5.0

2.5

0 10 20 30 40 50 10 20 30 40 50

FIG. 5. Labeling of membrane proteins with PLP and [3HJNaB&. R12-infected [14C]lysine-labeled K38 bacteria were converted to spheroplasts and divided into samples A-D, and each was treated with [3H]NaBH4 with or without PLP as described under Fig. 2 and “Experimental Procedures.” The modified membranes were isolated and the Triton/Mg*+-soluble proteins were prepared and subjected to electrophoresis on a SDS-10% polyacrylamide gel as described under “Experimental Proce- dures.” A, membrane proteins from intact spheroplasts reacted with PLP (sample A); B, membrane proteins from intact spheroplasts treated in the absence of PLP (sample B); C, proteins from isolated membranes reacted with PLP (sample C); D, proteins from isolated membranes treated in the absence of PLP (sample D) . The arrows from left to right are the respective positions of the following proteins: phosphorylase, 94,000 daltons; bovine serum albumin, 67,000 daltons; ovalbumin, 43,000 daltons; lysozyme, 14,300 daltons.

SLICE NUMBER

due to the coupling of pyridoxine 5’-phosphate. Based on these data, peptide I’ represents the modified NH2-terminal peptide containing residues 1-11, peptide 11’ represents the modified COOH-terminal peptide containing residues 46-50, and peptide 111’ represents the modified peptide containing residues 43-45. The extent of modification of peptides in peaks I and I1 was about 5 and 6%, respectively, based on the specific activity of C3H]NaBH4 and [‘4C]lysine in the coat protein.

Modification of Membrane Proteins with PLP and r3H] NuBH,”Spheroplasts were formed from [14C]lysine-labeled K38 bacteria infected with R12, divided into four equal portions, and treated as described in Fig. 2 and under “Experi- mental Procedures.” Two of the portions were reacted with [‘HH]NaBH4 after treatment with PLP (sample A) or without PLP (sample B) and the intact spheroplasts were isolated again. Membranes were isolated from the other two portions of spheroplasts and then were reacted in the same manner with [“H]NaBH4 after treatment with PLP (sample C) or in the absence of PLP (sample D). The cytoplasmic membrane proteins from each sample were isolated and analyzed by SDS-polyacrylamide gel electrophoresis (Fig. 5). Comparison of the tritium patterns shows that incorporation of tritium into proteins was dependent on the presence of PLP in both intact spheroplasts (Fig. 5, compare A and B ) and isolated membranes (Fig. 5, compare C and D), indicating that most of the tritium is associated with the modified a- and e-amino groups on these proteins. It is also evident that the ratio of tritium to [’4C]lysine is much higher in the proteins from the isolated membrane fraction (Fig. 4C) than from the spheroplast fraction (Fig. 44). In addition, there are tritium peaks present in the proteins from sample C that are absent in sample A. These results are consistent with the idea that only the proteins on the outside of the cytoplasmic membrane of spheroplasts were available for modification with PLP. To test this, Rl2-infected spheroplasts labeled with [14C]lysine were reacted with PLP and [3H]NaBH, before or during lysis by osmotic shock. The cytoplasmic proteins were then isolated and analyzed by SDS-polyacrylamide gel electrophoresis (Fig. 6). The lack of any appreciable tritium-labeled proteins in the cytoplasm isolated from the modified intact spheroplasts (Fig. SA) shows that the spheroplasts are quite impermeable to PLP under these conditions.

0P0 S L I C E NUMBER

FIG. 6. Exclusion of PLP from the cytoplasm of intact spheroplasts. Spheroplasts from R12-infected K38 bacteria labeled with [14C]lysine were divided into two portions. One was reacted with 8 mM PLP and then 4 mM [3H]NaBH4 in S buffer, the intact spheroplasts were isolated and lysed by osmotic shock; and the cytoplasmic proteins were isolated and concentrated by precipitation with 5% trichloroacetic acid. The other portion was lysed in the presence of 8 mM PLP and then reduced with 4 mM [’H]NaBH+ and the cytoplasmic proteins were isolated and concentrated by precipitation with 5% trichloroacetic acid. The resulting proteins from both preparations were subjected to SDS-polyacrylamide gel electrophoresis. a, c-yto- plasmic proteins from intact spheroplasts isolated from reaction with PLP and [3H]NaBH4; b, cytoplasmic proteins of spheroplasts which were lysed in the presence of PLP and labeled with [3H]NaBH4.

Isolation of Membrane-associated Coat Protein-Any analysis of the orientation of the coat protein in the membrane requires that there are no phage particles present, either from infecting or newly packaged virions. The use of R12-infected

9956 Membrane-associated Bacteriophage f l Coat Protein

K38 precluded the assembly of any newly synthesized coat protein into phage. To ensure that all exogenous phage particles were removed before analyzing the modified coat protein, we infected [3H]lysine-labeled K38 with ‘*C-labeled phage particles and determined the amount of phage present at each step of the procedure (see Fig. 2). Table I shows that the purified spheroplasts still contained about three infectious particles/spheroplast, and all of these subsequently were found in the Triton/M$+-insoluble fraction and not in the Triton/Mg2”soluble fraction, which contained the cytoplasmic membrane proteins.

The proteins from the Triton/Mg”-insoluble and -soluble fractions were then analyzed on a 16.8% polyacrylamide gel in the presence of SDS (Fig. 7). While the insoluble fraction contained the [‘*C]lysine-labeled coat protein from the adsorbed phage particles, it contained only a small amount of newly synthesized [3H]lysine-labeled coat protein (Fig. 7a). Almost all of the newly synthesized coat protein was present in the Triton-soluble fraction together with two phage equivalents of [’4C]lysine-labeled coat protein which probably en- tered the membrane during infection (Fig. 7b). This latter protein probably is in the same orientation in the cytoplasmic membrane as newly synthesized coat protein since coat protein of infecting phage has been shown to be reincorporated into newly assembled phage particles by Smilowitz (49). Based on the amount of 3H and 14C radioactivity in the coat peaks in Fig. 7B, at least 97% of the coat protein in the membrane was synthesized following infection.

Analysis of Coat Protein Modified with PLP and f 3 H j NaBH4 in Spheroplasts and Membrane Vesicles-The above data show that the procedure outlined in Fig. 2 should give accurate information concerning the orientation of coat protein in the cytoplasmic membrane. Therefore, spheroplasts and membranes were prepared from [14C]lysine-labeled bacteria infected with R12, treated with low levels of r3H)NaBH4 to reduce background labeling, and then reacted with PLP and [3H]NaBH4 as described under “Experimental Proce- dures” and outlined in Fig. 2. After isolation of the intact spheroplasts in samples A and B, the membranes from all four samples (A-D) were washed, the coat protein was isolated, and the soluble chymotryptic peptides from each sample were analyzed as described (Figs. 2 and 4). As shown in Fig. 8, the

TABLE I Rernoual of adsorbed R12 phaRe particles

~~

Bacteriophage/ceU

Infectious parti- C~~cCILysinec Sample“ des*

Infected cells 91 270 Spheroplasts 2.8 11 Membrane fraction 2.3 4.8 Triton/Mg2’-insoluble 3.7 1.8 Triton/Mg*’-soluble 0.002 2.5

“[3H]Lysine-labeled K38 (2 X 10’ cells/ml) were infected with [“C]lysine-labeled R12, the spheroplasts formed were reacted with PLP and NaBH4, and the membrane proteins were fractionated as described under “Experimental Procedures. ’ The infectious R12 in each sample was determined by plating on K37 bacteria. Bacteriophage/cell for the infected cell sample is the number of infectious particles/initial number of bacteria. The bacteriophage/ceU for all other samples is the number of infectious particles/intact spheroplasts, assuming no losses during fractionation.

‘The number of equivalent phage particles/sample was determined from the amount of “C radioactivity in each sample based on the specific activity of the R12 phage preparation. Bacteriophagelcell for the infected cell sample is the phage equivalents/the initial number of bacteria. Bacteriophage/cell for all other samples is the phage equivalent/intact spheroplasts, assuming no losses during fractionation.

SLICE NUMBER FIG. 7. Solubilization of coat protein from the membranes.

Spheroplasts of [3H]lysine-labeled K38 bacteria infected with [“CJ lysine-labeled R12 bacteriophage were treated with PLP and NaBH4 and the membranes were isolated and treated with Triton/Mg2+ as described under “Experimental Procedures.” The proteins of each fraction were subjected to electrophoresis on a 16.8% polyacrylamide gel in the presence of SDS. a, insoluble fraction of the Triton/Mg2’ extraction; b, soluble fraction of the Triton/Mg2’ extraction. Arrow marks the position where purified coat protein migrated in a parallel gel.

[‘4C]ly~ine was only present in positions representative of chymotryptic peptides containing residues 1-11, 46-50, and 43-45, independent of whether the coat protein was isolated from membranes which were obtained from intact spheroplasts (sample A) or from lysed spheroplasts (sample C) (Fig. 8, a and b). This confiied that only pure coat protein was subjected to digestion with chymotrypsin.

The protein isolated from membranes which were reacted with PLP and r3H]NaBH4 as intact spheroplasts (sample A) yielded only the NH2-terminal peptide labeled with tritium (Fig. 8a, peak I‘). However, tritiated chymotryptic peptide consisting of residues 46-50 (peak II‘) together with the tritiated peptide consisting of residues 1-11 (peak If) were obtained upon digestion of coat protein isolated from mern- branes which were modified as lysed vesicles (sample C, Fig. 86). In both cases, treatment of chymotrypsin released 60% of the tritium contained in the isolated coat protein.

The amount of tritium in each peak was then normalized to the amount of [‘4C]lysine radioactivity in the corresponding unmodified peptide peaks for each sample, A-D. This cor- rected for slight differences in the recovery of peptides from each sample. Then the small amount of non-PLP-dependent background labeling (samples B and D) were subtracted from their respective counterparts (samples A and C) to obtain the tritium profiles solely dependent on PLP as shown in Fig. 8c. There was not a great difference in the extent of modification of the NHz-terminal peptide (I’) in coat protein from samples A and C. The only condition in which the lysine in the COOH- terminal peptide containing residues 46-50 could be modified was when both sides of the membrane from the lysed spheroplasts were exposed to PLP. Essentidy the same result was obtained in two other similar experiments. In one of these experiments, the membrane sampIes C and D were labeled during lysis of the spheroplasts. Based on the specific activity of the [3HJNaBH4 and the [“C]lysine used in these experiments, we estimate that approximately 0.2% of lysine 48 was modified. This low amount of modification of the COOH-

Membrane-associated Bacteriophage fl Coat Protein 9957

FRACTION

FIG. 8. Gel filtration of the soluble chymotryptic peptides of coat protein isolated from the cytoplasmic membrane. Spher- oplasts and membranes from [‘4C]lysine-labeled K38 infected with R12 were reacted with PLP and [3H]NaBH4, the coat protein was isolated, and the chymotryptic peptides were analyzed as outlined in Fig. 2 and described under “Experimental Procedures.” a, the chymotryptic peptides of coat protein isolated from PLP- and [3H] NaBH4-reacted intact spheroplasts as described for sample A b, the chymotryptic peptides of coat protein isolated from PLP- and [’HI NaBH4-reacted membranes of lysed vesicles as described for sample C; c, the peptide pattern of PLP-dependent 3H labeling (sample A) minus the non-PLP-dependent labeling (sample B) from coat protein labeled in intact spheroplasts (H) and the pattern of PLP- dependent 3H labeling (sample C) minus the non-PLP-dependent labeling (sample D) from coat protein labeled in membranes from lysed spheroplasts (0”U). The arrow marks the position of tritiated pyridoxine 5”phosphate.

terminal peptide might be a result of steric hindrance from other molecules since the 4 lysines resides in the short 11 amino acid hydrophilic COOH-terminal portion of the molecule. Any interaction of the lysines with negative charged phospholipid head groups might prevent Schiff base formation of PLP with these lysines.

DISCUSSION

A procedure is described to label specifically amino groups of proteins on the membrane surface of E. coli infected with f l bacteriophage which involves reaction of PLP with spheroplasts followed by reduction with C3H]NaBH4. It is an adap- tation of the method first described by Rifkin et al. (30) to label the proteins on the outside of influenza virus. Under the conditions described in this paper, tritium incorporation is almost completely dependent on the presence of PLP, and only those proteins on the outside of the spheroplasts will be labeled. We found that three procedures are essential to obtain good results with this method. First, the spheroplast suspen- sions have to be prereduced with nonradioactive NaBH4 before reaction with PLP. This serves to reduce the level of nonspecific labeling by r3H]NaBH4, presumably due to the

presence of low levels of aldehydes or ketones which might have higher reactivity than PLP for amino groups of proteins. Such prereduction reduced by 90% the level of non-PLP- dependent labeling of the membrane proteins in the experiments presented in this paper. Also, a control reaction always should be run without PLP so that the remaining non-PLP- dependent labeling can be accounted for. Part of this nonspecific labeling results from penetration of [3H]NaBH4 into the spheroplast. Second, excess PLP should be present until the reduction by C3H]NaBH4 since the formation of the Schiff base is reversible. Last, following the labeling reaction, the intact spheroplasts should be reisolated. This ensures against analyzing any membrane proteins from spheroplasts which might have lysed during the labeling procedure.

When the coat protein was purified from spheroplasts labeled in this manner and the chymotryptic peptides were analyzed for the presence of tritium, only the NH2-terminal peptide containing residues 1-11 appeared to be modified. No PLP-dependent tritium labeling of any chymotryptic peptide originating from the COOH-terminal end of the coat protein could be detected. If the lysines at the COOH-terminal end of the protein were exposed, they should have been amenable to reaction with PLP since lysine 48 and either 43 or 44 can be modified in deoxycholate-solubilized purified coat protein (Figs. 1 and 4). Such data suggest that the COOH-terminal portion of the coat protein is only exposed on the cytoplasmic surface of the membrane and would be labeled under conditions where both sides of the membrane could interact with the reagents. This was confirmed by the observation that, when the labeling procedure was applied to isolated membrane vesicles or membranes during osmotic lysis of the spheroplasts, the COOH-terminal chymotryptic peptide containing residues 46-50 also was labeled (Fig. 8). Approximately the same amount of NH2-terminal peptide was labeled per equivalent spheroplast independent of whether the reaction was carried out on membrane vesicles or intact spheroplasts. These data strongly suggest that the NHn-terminal portion of the coat protein only is exposed on the outside of the cytoplasmic membrane, a conclusion in agreement with previous results (24, 25).

The fact that the chymotryptic peptide containing lysine 48 only can be labeled in membrane vesicles from lysed spheroplasts would indicate that the COOH-terminal portion of coat protein is exposed only on the cytoplasmic side of the membrane. This is based on the reasonable assumptions that gentle lysis by osmotic shock does not extensively perturb the rela- tive orientation of membrane proteins and that the coat protein molecules are labeled at random. With regard to the latter point, we noticed that the coat protein appeared to be labeled to a lesser extent than other membrane proteins. This probably was a result of the small size of the hydrophilic regions of the coat protein, especially at the COOH terminus which is 11 amino acids long and contains four lysines. These lysines might be expected to interact to some extent with the negatively charged head groups of the phospholipids (50,51), making it difficult for the PLP to form a Schiff base with them. This may be the reason why deoxycholate-solubilized purified coat protein can be modified to a greater extent than can membrane-associated coat protein.

The results presented in this paper give the fist in vivo evidence that the COOH-terminal portion of the coat protein is exposed only on the cytoplasmic side of the membrane and thus is in agreement with the idea that the mature coat protein asymmetrically spans the cytoplasmic membrane of the infected host. The data from experiments with detergent miscelles and phospholipid vesicles suggest that the region containing residues 40-50 interacts with the hydrophilic rni-

9958 Membrane-associated Bacteriophage fl Coat Protein

leau of the cytoplasm (11,12). Most or all of the residues 21- 39 appear to interact with the hydrocarbon bilayer of the membrane. Although there are some experiments indicating that tyrosine 21 might be exposed for iodination by lactoper- oxidase in intact spheroplasts (23), most recent evidence using nuclear magnetic resonance strongly suggests that both tyrosine residues are not accessible to solvent when the coat protein is in phospholipid vesicles (52) and thus are buried in the hydrophobic portion of the bilayer. This would leave residues 1-20 exposed to the outside of the cell membrane.

It should be possible to apply this technique to ascertain the orientation of other proteins in the E. coli cytoplasmic membrane. This is especially true for those membrane proteins which can be synthesized in large amounts under certain conditions or to which antibodies are available to aid in the isolation of the protein. One improvement might be to use radioactive PLP, thus removing the need for monitoring for non-PLP-dependent labeling with C3H]NaBH4. One difficulty with using radioactive PLP is that a rather high concentration is required for adequate labeling of membrane proteins, as pointed out in this paper. Reagents other than PLP (this paper) or a~etyl[~~S]methionyl methylphosphate sulfone (27) which can react with proteins under mild conditions in which spheroplasts remain intact may also be of value. In this regard, we found that Bolton-Hunter (53) reagent was unsatisfactory in that it penetrated into the spheroplasts within 1 min. Diazosulfanilic acid (54) was impermeable to the membrane but showed extremely poor reactivity to the lysine of coat protein even though this reagent has been reported to have some reactivity with the €-amino group of lysine (55). Knowl- edge obtained with in uiuo experiments using group specific reagents combined with those with detergents and model membrane systems should add to our understanding of the complex bacterial membrane.

Acknowledgments-We thank Ray Grant and T. S. Benedict Yen for heipfid discussions, Lucinda Hamilton for technical assistance, and Mary Jo Outlaw for assistance in preparing this manuscript.

1.

2.

3.

4.

5.

6.

7.

8.

9. 10.

11. 12.

13.

14.

REFERENCES Denhardt, D. T., Dressler, D., and Ray, D. S., eds (1978) The

Single-Stranded DNA Phages, Cold Spring Harbor Laborato- ries, New York

Simons, G. F. M., Konings, R. N. H., and Schoenmakers, J. G. G. (1979) FEBS Lett. 106, 8-12

Lin, T.-C., Webster, R. E., and Konigsberg, W. (1980) J. Biol. Chem. 255,10331-10337

Grant, R. A., Lin, T.-C., Konigsberg, W., and Webster, R. E. (1981) J. Bio2. Chem. 256,539-546

Sugimoto, K., Sugisaki, H., Okamoto, T., and Takanami, M. (1977) J. Mol. Biol. 110,487-507

Chang, C. N., Blobel, G., and Model, P. (1978) Proc. Natl. Acad.

Mandel, G., and Wickner, W. (1979) Proc. Natl. Acad. Sci. U. S.

Smilowitz, H., Carson, J., and Robbins, P. W. (1972) J. Supramol.

Webster, R. E., and Cashman, J . S. (1973) Virology 55,20-38 Makino, S., Woolford, J. L., Jr., Tanford, C., and Webster, R. E.

Wickner, W. (1976) Proc. Natl. Acad. Sci. U. S. A . 73, 1159-1163 Chamberlain, B. K., Nozaki, Y., Tanford, C., and Webster, R. E.

Nozaki, Y., Reynolds, J. A., and Tanford, C. (1978) Biochemistry

Hagen, D. S., Weiner, J. H., and Sykes, B. D. (1978) Biochemistry

Ski. U. S. A. 75,361-365

A. 76,236-240

Struct. 1, 8-18

(1975) J. Biol. Chem. 250,4327-4332

(1978) Biochem. Biophys. Acta 510,18-37

17, 1239-1246

17, 3860-3866

15. Smith, L. M., Smith, B. A., and McConnell, H. M. (1979) Bio-

16. Chang, C. N., Model, P., and Blobel, G. (1979) Proc. Natl. Acad.

17. Russel, M., and Model, P. (1981) Proc. Natl. Acad. Sci. U. S. A .

18. Blobel, G., and Dobberstein, B. (1975) J. Cell Biol. 67,835-851 19. Wickner, W., Mandel, G., Zwizinski. C.. Bates. M.. and Killick. T.

chemistry 18,2256-2259

Sci. U. S. A. 76, 1251-1255

78, 1717-1721

20.

21. 22. 23

24. 25.

26. 27.

28. 29. 30.

31. 32.

33.

3 4 .

35.

36.

37. 38.

39.

40. 41.

42.

43.

44.

45.

46.

47.

48.

49. 50.

51.

52.

53.

54. 55.

(1978) Proc. Natl. Acad. Sci. U. S . A. 75,175411758 Ito, K., Date, T., and Wickner, W. (1980) J. Bwl. Chem. 255,

Wickner, W. (1979) Annu. Reu. Biochem. 48,23-45 Marvin, D. A., and Wachtel, E. J . (1975) Nature 253, 19-23 Webster, R. E., and Cashman, J . S. (1978) in The Single-Stranded

DNA Phages (Denhardt, D. T., Dressler, D., and Ray, D., eds) pp. 557-569, Cold Spring Harbor Laboratories, New York

Wickner, W. (1975) Proc. Natl. Acad. Sci. U. S. A . 72,4749-4753 Date, T., Goodman, J . M., and Wickner, W. T. (1980) Proc. Natl.

Etemadi, A.-H. (1980) Biochem. Biophys. Acta 604,347-422 Smith, W. P., Tai, P.-C., Thompson, R. C., and Davis, B. D. (1977)

Proc. Natl. Acad. Sci. U. S. A. 74,2830-2834 Christensen, H. N. (1958) J. Am. Chem. SOC. 80.99-109 Churchich, J . E. (1965) Biochemistry 4, 1405-1410 Rifkin, D. B., Compans, R. W., and Reich, E. (1972) J. Biol.

Hunt, R. C., and Brown, J. C. (1974) Biochemistry 13,22-28 Cabantchik, I. Z., Balshin, M., Breuer, W., and Rothstein, A.

Juliano, R. L., and Behar-Bannelier, M. (1975) Biochem. Biophys.

Golds, E. E., and Braun, P. E. (1976) J. Biol. Chem. 251, 4729-

Garadi, R., and Babitch, J. A. (1980) Biochem. Biophys. Acta

Francis, G. A., and Richards, W. R. (1980) Biochemistry 19,5104-

Lyons, L. B., and Zinder, N. D. (1972) Virology 49.45-60 Viiiuela, E., Algranati, J. D., and Ochoa, S. (1967) Eur. J. Bio-

Osborn, M. J., Gander, J. E., Parisi, E., and Carson, J. (1972) J.

Schnaitman, C. (1971) J. Bacteriol. 108, 545-552 Ritchey, J. M., Gibbons, I., and Schachman, H. K. (1977) Bio-

Woolford, J. L., Jr., and Webster, R. E. (1975) J. Biol. Chem. 250,

Nozaki, Y., Chamberlain, B. K., Webster, R. E., and Tanford, C.

Weiner, A. M., Platt, T., and Weber, K. (1972) J. Bwl. Chem.

Asbeck, F., Beyreuther, K., Kohler, H., von Wettstein, G., and Braunitzer, G. (1969) Hoppe-Seyler’s 2. Physiol. Chem. 350, 1047-1066

Nakashima, Y., and Konigsberg, W. (1974) J. Mol. Biol. 88, 598- 600

Webster, R. E., and Rementer, M. (1980) J. Mol. Biol. 139,393- 405

Anderson, B. M., Anderson, C. D., and Churchich, J . E. (1966) Biochemistry 5,2893-2900

Smilowitz, H. (1974) J. Virol. 13,94-99 Chamberlain, B. K., and Webster, R. E. (1976) J. Biol. Chem.

Pluschke, G., Hirota, Y., and Overath, P. (1978) J. Biol. Chem.

Hagen, D. S., Weiner, J. H., and Sykes, B. D. (1979) Biochemistry

Bolton, A. E., and Hunter, W. M. (1973) Biochem. J. 133, 529-

Berg, H. C., and Hirsh, D. (1975) Anal. Biochem. 66,629-631 Pressman, D., and Grossberg, A. L. (1968) Structural Basis of

2123-2130

Acad. Sci. U. S. A. 77,4669-4673

Chem. 247,6432-6437

(1975) J. Biol. Chem. 250, 5130-5136

Acta 375,249-267

4735

595, 31-40

5111

chem. 1, 3-11

Biol. Chem. 241,3962-3972

chemistry 16,4584-4590

4333-4339

(1976) Nature 259,335-337

247,3242-3251

251, 7739-7745

253,5048-5055

18,2007-2012

539

Antibody Specificity, Benjamin, Inc., New York

the orientation of the major coat protein of bacteriophage ... · pdf filethe orientation of...

Documents