Isolation and Characterization of a Bacteriochlorophyll- containing Protein from Rhodopseudornonas spheroidese

Polyacrylamide gel electrophoresis of the major proteincontaining fraction of Rhodopseudomonas spheroides chromatophores reveals the presence of five electrophoretic components. The predominant component has associated with it, nearly all of the chlorophyll and phospholipid initially present in the chromatophore. Because of its potential structural and functional importance, this fraction has been isolated and purified by sodium dodecyl sulfate (SDS) preparative acrylamide gel electrophoresis. As isolated from preparative gels, this protein is at least 95% pure by virtue of the following criteria: protein chemistry, analytical gel electrophoresis, immunochemical reactivity, and a linear sedimentation-equilibrium plot. Chemical characterization of the protein-lipid-pigment complex reveals the presence of approximately 59% protein, 35% phospholipid, and 6% chlorophyll. Amino acid analysis of the purified protein shows a high content of nonpolar residues, no histidine or cysteine. Minimum molecular weights were estimated as 7,000 by amino acid analysis, 10,000 by SDS-acrylamide gel analysis, and 15,000 by sedimentation equilibrium analysis. Immunochemical characterization indicates that this protein is specific to only the photosynthetic membrane system derived from phototrophically grown cells.

SUMMARY Polyacrylamide gel electrophoresis of the major proteincontaining fraction of Rhodopseudomonas spheroides chromatophores reveals the presence of five electrophoretic components. The predominant component has associated with it, nearly all of the chlorophyll and phospholipid initially present in the chromatophore. Because of its potential structural and functional importance, this fraction has been isolated and purified by sodium dodecyl sulfate (SDS) preparative acrylamide gel electrophoresis. As isolated from preparative gels, this protein is at least 95% pure by virtue of the following criteria: protein chemistry, analytical gel electrophoresis, immunochemical reactivity, and a linear sedimentation-equilibrium plot. Chemical characterization of the protein-lipid-pigment complex reveals the presence of approximately 59% protein, 35% phospholipid, and 6% chlorophyll. Amino acid analysis of the purified protein shows a high content of nonpolar residues, no histidine or cysteine. Minimum molecular weights were estimated as 7,000 by amino acid analysis, 10,000 by SDS-acrylamide gel analysis, and 15,000 by sedimentation equilibrium analysis. Immunochemical characterization indicates that this protein is specific to only the photosynthetic membrane system derived from phototrophically grown cells.
Rhodopseudomonas spheroides is a facultative photosynthetic bacterium.
When grown phototrophically, the light-harvesting machinery is localized in a continuous vesicular structure which can be isolated as a homogeneous, cornminuted membranous fraction of molecular size, 30 million daltons.
These particulate vesicles are termed chromatophores and are believed to be derived from the peripheral cell membrane (1). As described previously (2) bulk quantities of purified chromatophores can be isolated which contain less than lo/, nonchromatophore protein as well as no detectable RNA, DNA, or  composed of protein (640/,), phospholipid (25%), and bacteriochlorophyll (5%). Purified chromatophores can be fractionated into two major protein Fractions PI and PII. PI is insoluble in chloroethanol and represents less than 5% of the total chromatophore protein.
Electrophoresis of PI on SDS* polyacrylamide gels reveals ten electrophoretic components whose general pattern was similar to that of the peripheral cell membrane.2 The chloroethanol-soluble fraction, PII, contains 95% of the chromatophore protein as well as most of the pigment and phospholipid.
SDS gel electrophoresis of PII indicates the presence of five components.
The fast migrating major component (Band 15) was found to have most of the bacteriochlorophyll and phospholipid associated with it. The work described here is concerned with the purification of this protein-chlorophyll complex (Band 15) as well as its chemical and physical analysis.
Also, described are some preliminary immunochemical studies which show the reIationship of this protein to other cellular components derived from both phototrophic and chemotrophic cells.

MATERIALS AND METHODS
Growth of Organisms-R. spheroides strain 2.4.1 was grown both chemotrophically and phototrophically as previously described (2).
Amino Terminus-NHz-terminal analysis was performed with the dansyl chloride methods of Gray (3) and Gros and Labouesse (4) employing 15 nmoles of protein.
The molecular weight of 14,000 as indicated by sedimentation-equilibrium anaIyses was assumed throughout.
Thin layer chromatography was performed on polyamide sheets (15 x 15 cm). The first dimension was developed with 1.5% formic acid (v/v) and the second with benzene-acetic acid (9:1, v/v). Chromatograms were examined with both short and long wave ultraviolet light. Horse heart cytochrome c was used as a standard.
Curboxy Terminus-The COOH terminus of the complex was determined with the hydrazinolysis method of Akabori as described by Narita (5)  ternal standard following hydrazinolysis to determine the recovery of free amino acids while lysozyme was used as a standard to determine the reliability of the reaction in our hands. At least 1.0 pmole of protein was used in all cases.
Amino Acid Analysis-Pigment and lipid was removed from the complex by extraction with acetone-methanol (7:2, v/v).
Samples of 0.418 mg of protein were added to each of six tear bulbs together with 3.6 ml of constant boiling HCl (5.7 N).
The tubes were evacuated, sealed, and hydrolysis of the protein was performed in duplicate for 24, 48, and 72 hours at 110". Separate analysis for methionine, cystine, cysteic acid, and tryptophan were conducted as previously described (2). The time-dependent loss of threonine and serine were corrected for by extrapolation to time zero by the method of least squares. Sedimentation Equilibrium-Sedimentation equilibrium analyses at 39,680 rpm were obtained with a Beckman Spinco model E ultracentrifuge equipped with interference optics. The protein (lipid and pigment removed) was dissolved in 4 M guanidine HCI (optically pure) in 0.01 M Tris buffer, pH 7.0, at concentrations of 3.0, 2.25, and 1.5 mg per ml. Lysozyme and ribonuclease were used as standard proteins in order to measure the degree of error caused by the use of this solvent system. The samples in a six-channel Kel F centerpiece with sapphire windows were centrifuged in an An-D rotor with counterbalance.
The high speed Yphantis method was used for these determinations (7). The contribution of both concentration and rotor velocity to the molecular weight was also determined.
The partial specific volume used (0.754 cm3 per g) was calculated from the amino acid composition of the protein.
Determination of Protein-bound XDS---The concentration of SDS in solution or bound to protein was determined colorimetrically by the method of Reynolds and Tanford (8).
Immunochemistry-Two rabbits were injected with approximately 3.5 mg of protein suspended in Freund's complete adjuvant.
One rabbit received only the protein derived from the complex (pigment and lipid were removed), the other received the whole complex.
Assays-Protein, phospholipid, and bacteriochlorophyll were determined as before (2). The absorption spectrum of the pigment-protein complex in water was obtained with a Cary model 14R recording spectrophotometer.
Cellulose thin layer chromatography sheets (6069) were purchased from Eastman Kodak.
Silica gel plates (5 x 20 cm) were obtained from EM Laboratories. The polyamide sheets are distributed by Gallard-Schlessinger Chemical Manufacturing Corp. SDS was purchased from Fisher and ultrapure guanidine-HCl from Schwarz.

RESULTS
Isolation of Pigmented Protein Complex-Because of the difficulties encountered in attempting to solubilize Fraction PII in aqueous solvent systems, we chose to use SDS preparative polyacrylamide gels for the fractionation. We used a preparative gel apparatus similar to the one designed by Fogel and Sypherd (11) and modified by the addition of a cooling jacket through which ethylene glycol at 15" was circulated. This served to reduce the heat produced by electrophoresis without causing precipitation of the SDS. A 250.ml, 6% polyacrylamide SDS gel was used under standard conditions devised for the analytical gel procedures previously described (2) with the exception that the concentration of EDTA was reduced IO-fold. Twenty-five milligrams of freshly prepared PII protein (2) were suspended in 5 to 6 ml of sample buffer containing 1% SDS, 6% glycerol, and the sample was layered onto the surface of the gel below the upper electrophoretic buffer. The gel was subjected to electrophoresis for approximately 44 hours at 20 ma per gel. The system was designed such that one to four gels could be accommodated at one time. After electrophoresis the gel was removed from the apparatus and a wedge-shaped slice was made along the length of the gel with pia,no wire. The slice was fixed and stained (2) and following destaining was placed back into the gel so tha,t the position of the various fractions could be determined.
A disc-shaped slice was made at the position of each protein.
This in turn was cut into about six pieces and placed in l-inch dialysis tubing with electrophoretic buffer (one-fifth the normal concentration). The dialysis tubing was then laid in a buffer chamber parallel to the electrodes which were oriented along the length of the tubing. After 12 hours of electrophoresis at 40 ma, approximately 90% of the protein eluted from the gel slices. This eluate was filtered through a Millipore filter in order to remove pieces of acrylamide. The filtrate was then lyophilized to dryness, resuspended in 5 ml of distilled water, and desalted over a 120-ml column of Sephadex G-25 in distilled water.
This step was repeated two times and resulted in less than 10% of the original SDS (w/w) remaining bound to the protein.
The remaining SDS could be removed by precipitation of the protein in cold acetone. Determination of Purity of Pigmented Profein Complex-The isolated pigmented protein was suspended in sample buffer and subjected to electrophoresis on analytical gels to determine its degree of purity; as can be seen in Fig. 1 a scan of the gel (2) at 365 nm (adsorption maximum of bacteriochlorophyll) prior to staining and at 550 nm after staining indicate that only a single component is present.
Furthermore, it can be seen from Fig. 1  Mes, dansyl methionine sulfone.
antisera developed against whole chromatophores, Fraction Prr and the complex itself, we were able to show that the complex consisted of only one immunological species (see Section A-l of Fig. 6) as judged by a single line of identity on Ouchterlony plates. Characterization of this protein, which will be described later, indicated that it had single NHz-and COOH-terminal amino acids and the plot of In J versus r2 from the sedimentation equilibrium centrifugation was linear. All of the above criteria indicate that the protein was at least 95% pure.
Determination of Amino Terminus-Dansylation of the protein seemed to be equally efficient regardless of the method used. Two-dimensional chromatography (Fig. 2) indicated that there were two spots present which matched in color and RF, standard dansylated methionine and methionine sulfone. As an additional test (Fig. 3) chromatography was performed with a chloroformbenzvl alcohol-acetic acid system (70:30:3, v/v/v) on silica  (12) which separates methionine and methionine sulfone from dansyl serine and threonine.
The unknown spots again migrated to the same position and showed a similar color to the dansyl methionine and methionine sulfone. When the unknown dansyl derivative of Band 15 protein was mixed with the two standard dansyl derivatives and chromatographed, only two spots were found indicating that they ran coincident with the standards.
Thus we conclude that the NHz terminus of this protein is methionine.
Determination of Carboxyl Terminus-Hydrazinolysis of the protein for 11 hours at 80" by the resin-catalyzed method of Schroeder as described (5) gave equimolar amounts of glycine and glutamic acid, representing a recovery of about 70%. This suggested that either the protein was composed of two different chains or that the glutamic acid could be derived from an internal residue. With the hydrazine sulfate-catalyzed method of hydrazinolysis for 10 hours at 60" we recovered twice as much glycine (50% recovery) as glutamic acid.
As an additional check the protein was digested for 0,4, 8, and 16 hours with carboxypeptidase A. As can be seen from Table  I, the hydrolysis, although inefficient under these conditions, yielded only glycine and no other amino acid. We therefore conclude that glycine is the carboxy terminus.
More recent experiments by Juinn Huangs utilizing carboxypeptidase A under more optimal conditions for its activity reveal the complete release of 1 pmole of glycine per pmole of protein within 35 min, at which time only 0.2 pmole of glutamic acid is released per pmole of protein.
Amino Acid Analysis of Pigment-Protein Complex-The data from duplicate 24-, 48-, and 72-hour hydrolysates were averaged and expressed in Table II as mole percent.
The recovery of amino acid residues was 79% based upon the recovery of norleucine.
The general composition is similar to that of the whole chromatophore protein (see Table 11). The isolated protein has neither histidine, nor cysteine, a generally low amount of basic and other polar amino acids and a high level of neutral and nonpolar amino acids. Assuming 2 moles of tryptophan per mole of protein the minimum molecular weight calculated from the composition was 6950. This calculation does not take into account amide content.
Determination of Molecular Weight of Protein of Complex-Because of the insolubility and small size of this protein it was difficult to obtain an accurate estimate of the molecular weight; therefore, determinations of molecular weight were performed with both SDS polyncrylamide gel electrophoresis (13) as previously described (2) and by sedimentation equilibrium. The results of the molecular weight determination by gel electrophoresis can be seen in Fig. 4. The molecular weight of the extracted protein by this method is 9900, however, the protein is barely on the linear portion of the curve and therefore the analysis probably is subject to more than the usual 10% error. In order to calculate the molecular weight from the plot of In J versus rZ obtained from the sedimentation equilibrium centrifugations, values for (&P:X) were calculated from t'he data given by Reisler and Eisenberg (14). From the expression (1 -+P") we calculated values of 0.24 for ribonuclease, 0.25 for lysozyme, and 0.20 for the polypeptide derived from the extracted complex.
In J versus r2 resulted in slopes of 1.51, 1.52, and 1.07 for lysozyme, ribonuclease, and extracted Band 15, respectively.
With these values the calculated molecular weights are 17,000 + 400 for lysozyme, 17,800 f 400 for ribonuclease, and 15,100 =t 100 for extracted Band 15. Lysozyme and ribonuclease were used as internal controls in order to assess the methodology and the uncertainty involved. As indicat'ed by Reisler and Eisenberg these calculations are subject to more than the usual error.
However, our values for lysozyme and ribonuclease are within acceptable limits if one considers that the values of (6P : SC) were calculated from those values obtained by Reisler and Eisenberg for rabbit muscle aldolase and bovine serum albumin.
Because of the different molecular weight values determined from SDS gel electrophoresis and sedimentation equilibrium we have routinely used the value of 14,000 which we have derived from the minimum molecular weight of 7,000 which has been determined from the amino acid analysis considering 1 residue of L-tryptophan per 7,000 daltons. Composition of Complex-An absorption spectrum of the complex indicated that the bacteriochlorophyll had been pheophytinized (see Fig. 5). This is not surprising since the material had been suspended in acidified chloroethanol during the puri- conditions we believe the 690~nm peak to be due to oxidized bacteriopheophytin and not oxidized bacteriochlorophyll because we have been unable to shift this peak under reducing conditions. However, until a difference spectrum has been performed, these results remain inconclusive. Similarly, the broad peak at 900 nm is probably due to aggregated complex containing oxidized bacteriopheophytin.
Composition studies indicate that 59% of the dry weight of the complex is protein, 35% phospholipid and 6% bacteriochlorophyll (see Table III).
To determine if any carotenoids were associated with this complex, extracts of the complex were applied to silica gel plates and chromatographed in petroleum ether-ethylacetate (10 : 1, v/v) (15). Both No. 9 the yellow and red carotenoids were present in the PII fraction, but were not apparently associated with t,he complex. However, addition of SDS to PII caused immediate decolorization of the carotenoids. Therefore, it is not possible to say at this time where the carotenoids are located.
However, visual observation of the complex does suggest that at least a portion of the carotenoids do appear to be present.
Phospholipid Composition-The phospholipid composition of Band 15 is presented in Table IV. The major components, representing greater than 94% of the total lipid phosphorus are similar in type and quantity to those observed by Gorchein (16) namely: phosphotidylcholine, phosphotidylethanolamine, and phosphotidylglycerol.
Ornithylphosphotidylglycerol was not determined separately and is included in the value presented for phosphotidylglycerol.
These results further suggest the integral structure of the protein-lipid-chlorophyll complex in the chromatophore structure of R. spheroides.
Tryptic Digest-The tryptic digest of this protein produced 11 to 13 peptides, most of which did not move from the origin during electrophoresis at pH 6.7 and are therefore considered to be neutral.
Further analysis of the composition of each of these peptides is in progress.
Based upon the estimate of molecular ureight as determined from the sedimentation-equilibrium analy- sis, and the amino acid composition we would expect 12 tryptic peptides.
Immunochemistry-In Fig. 6, Sections A-l and A-2 we see that the pigment-protein complex (15) cross-reacts well with antisera prepared against whole chromatophores (AC), Fraction PII (APTI), the pigment protein complex (A15), and the protein from the complex (A15X) in each instance forming a single line of identity.
It does not react, however, with antiserum developed against the PI (APJ membrane-like proteins. Of special interest are Sections A-3 and A-4 which show that there is no cross-reactivity between any of the above antisera and the pigment.ed protein isolated by Thornber and Olson (OP) (17) from t'he green bacterium Chloropseudomonas ethylicum.
In Sections B-l through B-4 of Fig. 6, we see t'hat there is no crossreactivity between either the antipigment protein complex (Al5) or the antiprotein (M5X) (from the complex) sera and the soluble cell supernatant (CS, PX) or sonicated membrane fraction (CM, PM) from either chemotrophic or phototrophic grown cells. The same was true for anti-PI1 serum, while antiserum prepared against whole chromatophores cross-reacts weakly with the cell membrane and cell supernatant from phototrophic cells. Anti-P1 sermn cross-reacted only with the cell membrane from phototrophic cells. u1scuss10iY The pigment-protein complex derived from R. spheroides chromatophores appears to be at least 95'$$ pure by immunological, chemical, and physical criteria.
As mentioned, 65% of its amino acids are made up of the neutral and hydrophobic amino acids and it is very close in composition to the whole chromatorhore protein.
It is approximately 40 to 50% by weight of the whole chromatophore protein.
The composition differs markedly from the pigment-protein complexes isolated from C. ethylicum FIG. 6. ImmunodifftLsion test of cross-reactivity. The symbols trophic cells, CM; sonic&ted membrane from phototrophic cells, are: chromatophores, C; pigmented protein (Band 15), 15; pro-PM; pigmented protein isolated by Olson from C. ethylicum, OP; tein with pigment-lipid extracted, 15X ; Fraction PI, PI; Fraction antisera against any of the above fractions is designated by an A PII, Prl; supernatant from chemotrophic cells, CS; supernatant preceding the fraction symbol. Antigens were used at a protein from phototrophic cells, PS; sonic&ted membrane from chemoconcentration of 1 mg per ml. Antisera were used undiluted. and spinach chloroplasts which also differ from one another. These data together with the fact that there was no immunological cross-reactivity between the two proteins suggest that the pigmented proteins are not phylogenetically related. This of course is not conclusive evidence of a lack of relationship since the amino acid composition of the complex from R. spheroides is very similar to that of impure pigmented proteins isolated from chard and corn chloroplasts (18).
The lack of cross-reactivity of anti-pigment-protein serum with either chemotrophic or phototrophic membrane is of great interest, for it suggests that the pigmented protein is not a normal component of the peripheral membrane (at least not in significant amounts).
The additional lack of cross-reactivity between the two types of membranes and serum developed against only the protein of the complex seems to preclude the possibiIity that during the induction period pigment and lipid are simply added onto a previously existing membrane protein.
Therefore, it appears that this particular chromatophore protein must be newly synthesized during the induction period of chromatophore synthesis.
This would also seem to be the case for all of the PII proteins, suggesting that at least 95y0 of the chromatophore proteins must be synthesized de novo following chromatophore induction. This might also suggest that the chromatophore does not arise from the peripheral cell membrane although other interpretations are possible. If the ratio of bacteriochlorophyll to protein present in the complex is fixed, then one could assume, according t,o the suggestions of Holt and Marr (19) that the bacteriochlorophyll content of the chromatophore increases only as the mass of the chromatophore increases. This in fact would seem to be the case since the specific chlorophyll content of purified chromatophores is about the same whether they are isolated by us (2) or others (19)(20)(21).
It is noteworthy that the molar ratio of bacteriochlorophyll to protein in the complex is 1.4 : 1. This may reflect inaccuracies in the determination (or determinations) of bacteriochlorophyll, molecular weight of the protein, or both. On the other hand, it is worthwhile entertaining the possibility that in the organization of the chromatophore the sharing of a molecule of bacteriochlorophyll between 2 molecules of protein is important in determining the three-dimensional structure of this organelle.
Recent investigations of Loach et al. (22), Segen and Gibson (23), and Clayton and Wang (24) all working with R. spheroides seem to yield results similar to those found earlier (2) and to those presented in this investigation.
The d:fficulty in comparing these results is primarily due to different methods of preparation of the materials under investigation and to differences in the interests of the investigators.
Very recent experiments of Clayton and Haselkorn4 strongly suggest that their low molecular weight, protein-light-harvesting-chlorophyll fraction, is comparable to our Band 15. Similar interpretations may apply to t#he fast migrat,ing component of Segen and Gibson (23).
Likewise, several of our PII proteins other than Band 15, have molecular weights and fractionation characteristics similar to the "reaction-center" complexes described by Loach et al. (22), Segen and Gibson (23), and Clayton and Wang (24). It would seem that some unified system of nomenclature and recognition is now highly desirable.
However, the fact that methods as 4 Clayton and Haselkorn, personal communication.
diverse as those used in the investigations cited (22)(23)(24), seem to yield strikingly similar results, strongly suggests the validity of the results. Questions as to how the chromatophore is formed remain unanswered.
It is reasonable to assume that new proteins are assembled at points along the peripheral membrane (l), perhaps at special sites. Whether the chromatophore proteins are synthesized synchronously and are assembled at the membrane in a one-step mechanism is also open to question.
With the isolation and purification of other chromatophore proteins, which is in progress, we feel that many of the above questions can be answered; additionally it becomes possible to determine, both immunologically and biochemically, the degree of homology between the chromatophore proteins.
With more sophisticated immunological techniques we should be able to detect the synthesis of each of these proteins and trace their development during the period when cells are in transition from chemotrophic to phototrophic growth conditions resulting in the induction of chromatophore synthesis.