Poly(3-hydroxybutyrate) is associated with specific proteins in the cytoplasm and membranes of Escherichia coli.

Poly(3-hydroxybutyrate) (PHB) is well-known as a high molecular weight homopolymer of R-3-hydroxybutyrate which accumulates in storage granules within the cytosols of certain bacteria. Escherichia coli does not amass these granules; however, small amounts of low molecular weight PHB (<0.02% of dry weight) have been found in the plasma membranes in complexes with calcium polyphosphate; the complexes serve as voltage-activated calcium channels. Here we report that polyphosphate-complexed PHB is only a minor fraction of the polyester in E. coli. PHB comprises 0.36 to 0. 55% of the dry weight of log-phase cells, depending on culture medium, and this amount increases by 15 to 20% when the cells are made genetically competent. The PHB is widely distributed throughout the cell, wherein it is primarily associated with proteins. The identity of protein-associated PHB was established by antibody reaction, chemical assay, and 1H NMR spectroscopy. As expected, the physical and chemical properties of protein-associated PHB were found to be considerably different from those of the bulk polymer or granule PHB, e.g. protein-PHB complexes are normally insoluble in chloroform, soluble in water and alkaline hypochlorite, and are converted to crotonic acid more slowly on heating in concentrated sulfuric acid. Our studies indicate that the majority of cellular PHB (over 80%) is located in cytoplasmic proteins, especially proteins of the ribosomal fraction. Western immunoblots, probed with polyclonal anti-PHB IgG, revealed a number of PHB-polypeptides having a wide range of molecular weights in all cell fractions. These results suggest that PHB is a fundamental constituent of cells that may have physiological functions in addition to facilitating ion transmembrane transport or serving as a carbon reserve.

There is increasing evidence that low molecular weight poly(3-hydroxybutyrate) (PHB) 1 (Ͻ15,000) is a ubiquitous constituent of prokaryotes and eukaryotes Sadoff, 1983, 1988;Reusch et al., 1986Reusch et al., , 1992Reusch, 1989Reusch, , 1992Seebach et al., 1994). Unlike the high molecular mass polymer (Ͼ60,000), sequestered in cytoplasmic inclusion bodies of certain eubacteria and archaebacteria (Anderson and Dawes, 1990; Dawes and Senior, 1973), the low molecular weight poly-mer has always been found in association with other macromolecules, hence it has been called complexed-PHB (cPHB) (Reusch, 1992). This term is now used more generally to refer to all cellular PHB other than the high molecular weight polymer found in inclusion granules. cPHB was first discovered in the plasma membranes of genetically competent bacteria, complexed with calcium polyphosphate (PP i ) (Reusch and Sadoff, 1988). These cPHB-polyphosphate complexes function as voltage-activated calcium channels , and they have also been postulated to serve in coexport of calcium and phosphate, in DNA transmembrane transfer, and as calcium stores (Reusch and Sadoff, 1988). The association of cPHB with proteins has been reported previously in plasma .
Current protocols for determining PHB in bacteria were designed to measure the high molecular weight uncomplexed polyester found in inclusion granules (Holmes, 1987); consequently, they exploit one or more of the physical and chemical properties of the bulk polymer, solubility in chloroform, insolubility in water, methanol, acetone, and ether, resistance to degradation by alkaline hypochlorite, and acid-catalyzed degradation to crotonic acid. We became aware of the inadequacies of these methods when they failed to detect cPHB in samples of bovine serum albumin that displayed strong positive reactions to polyclonal anti-PHB IgG . Subsequently, cPHB was extracted from bovine serum albumin with a mixed solvent system, and its identity was confirmed by chemical assay and 1 H NMR (Reusch, 1992;Seebach et al., 1994). The failure of traditional procedures could in part be attributed to their insensitivity to the comparatively low concentrations of cPHB in the samples, but could also be related to changes in the physical and chemical properties of the polyester effected by its association with protein.
This study was undertaken to determine whether cPHBprotein complexes exist in Escherichia coli. We investigated several common procedures for assaying PHB in bacteria and developed a sensitive protocol for determination of cPHB. Using this improved assay and anti-PHB IgG, we detected significant quantities of cPHB in E. coli and found that the polyester is widely distributed throughout the cell, wherein it is largely associated with proteins.

Reagents and Strains
Organisms used in this study were E. coli strains K12, DH5␣, JM101, and CA10 (kindly provided by A. Kornberg). SOB medium (Hanahan, 1987) was 2% Bacto-tryptone (Difco), 0.5% yeast extract (Difco), 10 mM NaCl, 2.5 mM KCl, 20 mM MgCl 2 . CB medium (competence buffer) was 10 mM CaCl 2 , 45 mM MnCl 2 , 100 mM KCl, in 10 mM MES, pH 6.3. MSG medium was minimal salts M9 medium, 6.0 g of Na 2 HPO 4 , 3.0 g of KH 2 PO 4 , 0.5 g of NaCl, 1.0 g of NH 4 Cl in 1 liter of distilled water with 0.5% glucose. All salts used in preparation of CB medium were high purity quality (Ͼ99%) from Aldrich. * This work was supported by National Institutes of Health Grant GM 33375 and National Science Foundation Grant MCB 9507282. 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.
‡ To whom correspondence should be addressed: Dept. of Microbiology,Giltner Hall,Michigan State University,East Lansing, The abbreviations used are: PHB, poly(3-hydroxybutyrate); cPHB, Preparation of Cells E. coli were cultured in SOB medium, or minimal salts M9 medium containing 0.5% glucose (MSG), at 37°C with shaking at 180 rpm to an A 550 of 0.4 -0.5. For JM101, MSG was supplemented with casamino acids at a final concentration of 0.02%. Cells were made genetically competent by a variation of the method of Hanahan (1987); cells in a log-phase culture in SOB medium were collected by centrifugation at low speed (800 ϫ g) at 4°C and were suspended in one-third volume of ice-cold CB medium for 30 min. The number of live cells was determined by plating cells on SOB agar; total cells were determined by counting in a Petroff-Hausser bacteria counter (C. A. Hausser & Sons); dry weight was determined by heating washed cell pellets in a 105°C oven to constant weight.

Determination of PHB
Method 1: Methanolysis in 3% Sulfuric Acid-The procedure used was that of Braunegg et al. (1978). Briefly, a mixture of methanol acidified with H 2 SO 4 (3% v/v) and chloroform (1:1) was added to lyophilized cells. The preparation was heated at 100°C for 4 h and cooled to ambient temperature. Water was added and the organic phase was analyzed with a Perkin Elmer AutoSystem Gas Chromatograph using a 30-m ϫ 0.32-mm inside diameter Carbowax PEG column. Quantitation was done by evaluating peak areas using granule PHB (Aldrich) as standards.
Method 2: Hydrolysis in 2 N NaOH-The procedure of Brandl (1992) was used. Briefly, lyophilized cells were heated in 2 N NaOH at 100°C for 30 min. The cooled samples were neutralized with 2 N HCl and analyzed by HPLC as described in Method 3 below.
Method 3: Acid-catalyzed ␤ Elimination to Form Crotonic Acid-The procedure used was an adaptation of the method of Karr et al. (1983). Concentrated sulfuric acid (0.5 ml) was added to a dry cell pellet (from 30 ml of culture), and the mixture was stirred and heated in a dry heating block (Pierce) at 92°C or other desired temperature. At specified time periods, the tube was cooled on ice, 1 ml of saturated sodium sulfate was added, and the solution was extracted 4 times with 3 ml of dichloromethane. Sodium hydroxide (100 l of 1 N) was added to the dichloromethane extract to convert crotonic acid to crotonate, and the dichloromethane was evaporated with a stream of nitrogen. The residue was acidified to pH 2 with 3 N sulfuric acid and filtered with a 0.2-m nylon syringe filter. The filtrate was chromatographed on an HPLC Aminex HPX-87H ion exclusion organic acid analysis column (Bio-Rad) with 0.014 N sulfuric acid as eluant at a flow rate of 1 ml/min. The crotonic acid peak was identified by comparison with known crotonic acid and by its UV absorption curve, and its identity was confirmed by mass spectrometry. Crotonic acid content was estimated by evaluation of peak area using granule PHB isolated from Azotobacter vinelandii as described previously (Reusch and Sadoff. 1983), or purchased from Aldrich, as standards.

Isolation of PHB
Alkaline Hypochlorite Digestion-The cell pellet was digested with 5.25% alkaline hypochlorite (Clorox) at 37°C for 30 min. The residue was collected by centrifugation and washed sequentially with water, methanol, and acetone (2ϫ each) and dried (Law and Slepecky, 1961). PHB was then converted to crotonic acid by heating in concentrated sulfuric acid at 92°C for 45 min. The crotonic acid was recovered and determined as described above (Method 1).
Chloroform Extraction-The cell pellet was washed sequentially with water, methanol, and acetone (2ϫ each) and dried. The residue was extracted with 2 ml of warm chloroform (50°C) four times. The extracts were combined and the chloroform evaporated with a stream of dry nitrogen gas. PHB was then determined as above (Method 1).

Fractionation of Cells
E. coli log-phase cells (10 liters) were cultured in SOB medium to an A 550 of 0.4 -0.5. The cells were pelleted by centrifugation at 1800 ϫ g, washed with 10 mM Hepes buffer, pH 7.3, and then suspended in the same buffer containing 20 g/ml RNase and 20 g/ml DNase. The cells were broken by passage through a French pressure cell at a pressure of 18,000 p.s.i. Unbroken cells were removed by centrifugation at 7000 rpm in a GSA Rotor (Beckman) for 10 min (2ϫ) at 4°C. The supernatant was layered over a pillow of 60% sucrose and centrifuged in a Beckman Type 50.2 Ti rotor at 35,000 rpm for 45 min at 4°C in a Beckman LP50 Ultracentrifuge. The membranes were recovered from the sucrose interface with a J-hook and washed twice with 10 mM Hepes, pH 7.4. The cytoplasm was centrifuged in a Type 50.2 Ti rotor at 200,000 g for 2.5 h (2ϫ) to collect the ribosomal fraction (Sykes, 1971).
The remaining supernatant was considered to be cytosol. Protein was determined with the Bio-Rad Protein Assay Kit and the Pierce BSA Kit.

H NMR Spectroscopy
1 H NMR spectra were recorded in CDCl 3 solution using a Varian 500 MHz spectrometer at 25°C with 2000 accumulations, 60°pulse, and a 1.5-s repetition rate.

SDS-Polyacrylamide Gel Electrophoresis-Western Blots
Proteins of cell fractions (4 g each fraction) were resolved on 10-cm minigels (Bio-Rad) of 10% acrylamide (2.6% bis) and transferred to supported nitrocellulose membranes (BRL) using a Mini Trans-Blot electrophoretic transfer cell (Bio-Rad). The membranes were probed with polyclonal antibodies to PHB raised in rabbit as described previously  that had been preadsorbed with PHB-free proteins (see below) of E. coli. The second antibody was goat anti-rabbit IgG-alkaline phosphatase conjugate (Life Technologies, Inc.). Color development was done with the alkaline phosphatase substrate kit (Bio-Rad).

Elimination of cPHB from Protein
A pellet of log-phase JM101 cells was suspended in ice-cold 5% trichloroacetic acid and incubated on ice for 1 h. The residue was collected by centrifugation, washed 2ϫ with cold 5% trichloroacetic acid, and then suspended in 5% trichloroacetic acid and heated in a water bath at 90°C for 15 min. The suspension was cooled in an ice bath for 1 h, and the acid-insoluble residue was collected by centrifugation. After lyophilization, the residue was assayed for cPHB content as described above. The supernatant was neutralized with NaOH, dried by lyophilization, and assayed for cPHB in the same manner.

Isolation and Determination of cPHB from E. coli by Conventional
Methods-Current protocols for determining high molecular weight PHB in inclusion bodies of bacteria can be divided into two categories: 1) methods in which PHB is isolated and purified before analysis; 2) methods in which PHB is determined directly in whole cells. The efficiencies of some of the most popular protocols of each class were examined to determine the most sensitive method(s) for measuring cPHB in E. coli.
First we examined two procedures of the first category: one based on the relative resistance of PHB to degradation by alkaline hypochlorite solution, and the other based on the solubility of PHB in chloroform (see "Experimental Procedures"). The alkaline hypochlorite method yielded Ͻ0.004 g of cPHB/mg dry weight in log-phase cells and 0.14 to 0.24 g/mg dry weight in competent cells (Table I). The recovery of cPHB by extraction with chloroform proved less efficient; this method yielded Ͻ0.001 g of PHB/mg dry weight from log-phase cells and 0.05-0.06 g/mg dry weight in competent cells (Table I).
Next we examined the efficiency of three procedures for direct determination of PHB in whole cells: 1) acid methanolysis with 3% sulfuric acid as described by Braunegg et al. (1978); 2) alkaline hydrolysis in 2 N NaOH as described by Brandl (1992); 3) conversion to crotonic acid by heating in concentrated sulfuric acid as described by Karr et al. (1983). As shown in Table II, the direct methods were generally more efficient in measuring cPHB. The latter two methods were about 10-fold more sensitive than the first, and the conversion to crotonic acid by hot concentrated sulfuric acid proved to be about one-third more sensitive than alkaline hydrolysis (after corrections were made for both procedures), yielding on average 0.4 g of cPHB/mg dry weight for log-phase cells.
Further Studies of the Conversion of cPHB to Crotonic Acid in Concentrated Sulfuric Acid-The protocol that was most efficient in determining cPHB, i.e. ␤ elimination of crotonic acid effected by concentrated sulfuric acid, was examined further. The reaction can be roughly described by Reaction 1.
where X refers to other degradation products of PHB and Y to degradation products of crotonic acid.
This reaction is usually carried out at temperatures of 85°C to 100°C and typically results in 82-85% conversion of PHB to crotonic acid in 20 to 30 min (Karr et al., 1983). In our hands, the degree of conversion was found to be 85% (determined by extrapolation of the curve for degradation of granule PHB in Fig. 1A to zero). Most of the loss is attributed to conversion of PHB to other products (X) by other pathways. The maximal yield was achieved in 20 min and remained at about the same level for over 1 h (Fig. 1A). Pure crotonic acid, treated in the same manner, showed little degradation over this time period (ϳ3%) (Fig. 1A). The net result was that a 1-g sample of granule PHB yielded 0.82 g of crotonic acid after 20 min at 92°C. In contrast, the cPHB in lyophilized whole cells of E. coli JM101 produced little crotonic acid in 20 min (Fig. 1B). In this case, the yield did not level off but continued to rise for ϳ12 h. Over this time period, crotonic acid degradation was significant (47%) (not shown). After 12 h at 92°C, only 0.45 g of crotonic acid was obtained from 1 g of granule PHB. If we assume the same efficiency for cPHB, then the log-phase JM101 cells which produced 2.20 g of crotonic acid/mg dry weight contained ϳ4.9 g of cPHB ( Fig. 1B and Table II). The true efficiency for cPHB is probably less (and the amount of cPHB greater), since protein-associated cPHB is more likely to be degraded by other pathways; however, this loss is at least partly offset by a reduction in crotonic acid degradation due to its slower rate of formation.
To determine the best conditions for measurement of cPHB in E. coli cells, the time required to obtain a maximal yield of crotonic acid was measured as a function of temperature. As can be seen in Fig. 2, the rate proved to be highly sensitive to Cell pellet was digested with 5.25% alkaline hypochlorite (1 ml/10 10 cells) at room temperature for 30 min. The residue was washed with water, methanol, and acetone and dissolved in chloroform. After filtration, the chloroform was evaporated, 1 ml of concentrated sulfuric acid was added, and the mixture was heated at 92°C for 45 min. Crotonic acid was then isolated and quantitated as described under ''Experimental Procedures.'' Values are an average of 3 determinations and are corrected for incomplete conversion to crotonic acid and degradation of crotonic acid as determined for PHB granules used as standards. Cell pellet was washed with methanol and acetone, and the dry residue was extracted with warm chloroform (50°C, 4ϫ). The extract was filtered and chloroform was evaporated. PHB content was determined by conversion to crotonic acid in concentrate sulfuric acid as in 1.
Values are an average of 3 determinations.  1. Rate of conversion of PHB to crotonic acid in concentrated sulfuric acid at 92°C. Samples were removed at the indicated times, and the crotonic acid content was measured as described under "Experimental Procedures." A, granule PHB (1 g) (É). The rate of degradation of crotonic acid (Ç) in concentrated sulfuric acid at this temperature is also shown. B, cPHB in lyophilized whole cells of E. coli JM101. EOOE, log-phase cells; ⅪOOⅪ, competent cells. temperature. For routine measurements, we adopted a temperature of 120°C, at which the recovery of crotonic acid is maximal at 40 min. At this temperature, we found that only 69% of granule PHB was converted to crotonic acid (Fig. 3A). After 40 min at 120°C, pure crotonic acid was 28% degraded (not shown), so that the yield of crotonic acid from 1 g of granule PHB was 0.5 g. If we again assume the same efficiency for cPHB, then the crotonic acid measured in lyophilized E. coli cells after 40 min at 120°C represents about 50% of actual cPHB ( Fig. 3B and Table II). Despite the inexactness of the measurements of cPHB, due to the lack of an appropriate standard, values obtained under the same assay conditions can provide useful comparisons of cPHB content.
cPHB in E. coli Whole Cells-Using these conditions (120°C for 40 min), we compared cPHB levels for several strains of E. coli cells in log-phase growth (A 550 ϳ 0.4). The results, shown in Table III, indicate that cPHB constitutes 0.36 -0.55% of the dry weight of log-phase cells, depending on the nature of the growth medium. When cells were made competent, they produced 15-20% more cPHB (0.61-0.65% of dry weight); 10 -15% of the additional polyester was isolated from the competent cell pellets, and 5% was recovered from lysed cells in the competence buffer. About 10% of this newly synthesized cPHB was chloroform-soluble. Chloroform-soluble cPHB is barely detectable in E. coli log-phase cells (Table I), and in competent cells it constitutes ϳ0.01% of the dry weight and ϳ1.5% of cPHB.
Distribution of cPHB-The cellular location(s) of cPHB in log-phase and competent cells was next determined. Cells were disrupted and separated by centrifugation into cytoplasmic and membrane fractions, and the cytoplasms were further separated by centrifugation into cytosol and ribosomal fractions. These fractions were then assayed for cPHB content. As shown in Table IV, over 90% of this highly water-insoluble polyester was found in the cytoplasm, with the largest concentration in the ribosomal fraction. The cPHB synthesized de novo during the development of competence was found mainly in the ribosomal and membrane fractions.
Identity of cPHB-The chemical assay for PHB is based on the assumption that crotonic acid is a unique product of PHB degradation in concentrated sulfuric acid. To rule out the possibility that crotonic acid could also be formed by reactions of amino acids under the assay conditions, we examined homopolymers of those amino acids whose structure suggested this capability, polythreonine, polyglutamate, polyglutamine, polymethionine, and polyarginine. None were found to produce detectable amounts (Ͼ0.001%) of crotonic acid.
The identity of cPHB in the ribosomal fraction was confirmed by 500-MHz 1 H NMR spectroscopy. Although cPHB could not be extracted from the acetone-dried ribosomes by warm chloroform, small amounts of cPHB were dissolved in refluxing CHCl 3 :MeOH (9:1) in 1 h. This procedure had been used previously to extract cPHB from bovine serum albumin, spinach, beef heart mitochondria, and human aortal tissue (Reusch et al., 1992;Seebach et al., 1994). As shown in Fig. 4, the 1 H NMR spectra of the extract includes resonances with the characteristic chemical shifts and coupling constants of the methylene and methine protons of PHB (Reusch, 1992;Seebach et al., 1994). The assignments were confirmed by selective decoupling. The methyl resonances were hidden under other signals.
Elimination of cPHB from Cell Proteins-cPHB was removed from proteins (and other macromolecules) of JM101 log-phase cells by taking advantage of the greater lability of ester bonds, as compared with amide bonds, to dilute acid hydrolysis. Cel- lular macromolecules were precipitated with cold 5% trichloroacetic acid, and the cPHB content of a sample of the cell residue was determined. Fig. 5B shows the HPLC chromatogram of crotonic acid produced by this sample, and Fig. 5A the chromatogram of a crotonic acid standard. The remainder of the cell residue was suspended in 5% trichloroacetic acid and heated at 90°C for 15 min. After cooling to allow reprecipitation of acidinsoluble material, the residue and supernatant were analyzed for cPHB content. As can be seen in Fig. 5C, this treatment removed Ͼ97% of cPHB from the acid-precipitable fraction. The process resulted in only slight degradation of the proteins as indicated by the appearance of polypeptide bands on SDSpolyacrylamide gel electrophoresis gels (not shown). The proteins in this residue contain very little cPHB and are referred to as cPHB-free proteins. Continuing the reaction for longer periods did not release additional amounts of cPHB and resulted in increased degradation of the proteins. The lyophilized supernatant (5% trichloroacetic acid-soluble fraction) yielded crotonic acid when heated in concentrated sulfuric acid (Fig.  5D), indicating that the trichloroacetic acid hydrolysis yielded small oligomers of PHB or the monomer, 3-hydroxybutyrate.
cPHB Proteins-Considering the insolubility of PHB in water, its concentration in the cytoplasm suggests it is sequestered within hydrophobic pockets of other macromolecules, most probably protein. To detect specific polypeptides associated with cPHB, Western blots of SDS-PAGE gels of JM101 cell fractions were probed with polyclonal anti-PHB IgG that had been preadsorbed with cPHB-free proteins (see above). The adsorbed antisera gave a strong positive reaction to synthetic PHB 96 (not shown), but did not react significantly to cPHB-free proteins, with the exception of a diffuse band at approximately 12,000 Da (Fig. 6, lane 1). All cell fractions contained a number of polypeptides that reacted to the antisera, with the highest concentration of cPHB polypeptides appearing in the ribosomal fraction (lane 4). Cell fractions of competent cells had essentially the same distribution of cPHB polypeptides as log-phase cells, and no new polypeptides were observed (not shown).
When the cytoplasmic fraction was treated with proteinase K (200 g/ml) for 2 h at 37°C before SDS-gel electrophoresis and Western blotting, no polypeptides or cPHB-containing bands a Values are estimates based on the assumption that the yield of crotonic acid from cPHB under the assay conditions is approximately the same as that for PHB granules used as standards (50%). a Values are estimates based on the presumption that the yield of crotonic acid from cPHB under the assay conditions is the same as that for PHB granules used as standards (50%). were observed at M r Ͼ 16,000; however, there was a diffuse band of degraded proteinaceous material in the molecular weight region of ϳ2,000 -15,000 which reacted positively to the adsorbed cPHB antisera (not shown). It is possible that the actual M r range of these presumed cPHB-peptides or polypeptides is less, since the uncharged polyester may have a conspicuous influence on the electrophoretic mobility of small molecules. cPHB was not released by this enzymatic proteolysis; less than 1% of the polyester in the hydrolysate was chloroform-soluble. This result suggests that cPHB is covalently bound to protein, and that it protects associated peptides from hydrolytic attack. DISCUSSION PHB is best known as a high molecular weight polymer sequestered in inclusion granules in the cytoplasm of certain prokaryotes (Anderson and Dawes, 1990;Dawes and Senior, 1973). The granules are not easily overlooked as they are phase-bright and may constitute up to 80% of the cell dry weight. Granule PHB can be extracted and measured using a number of standard protocols which are based on the physical and chemical properties of the bulk polyester (Holmes, 1987;Karr et al., 1983). By contrast, other cellular PHB, designated cPHB, which is complexed and of relatively low molecular weight, is sparse and obscure. The association of cPHB with other macromolecules significantly modifies its physical and chemical properties, making prevailing analytical methods ineffectual. Not only are the properties of cPHB different from those of granule PHB, but they vary with the nature of the complexant. The solubility characteristics of the cPHB complexes depend on a number of factors, including the size and location (inside or outside) of the complexant relative to cPHB and the hydrophobicity of residues at the surface of the complex. Protein-associated cPHB is converted to crotonic acid in hot concentrated sulfuric acid at a much slower rate than granule PHB, but its rate of hydrolysis in dilute acid is not significantly affected. This may be attributed to the more stringent conformational requirements of ␤-elimination compared with acid hydrolysis, and the degree to which ␣,␤-unsaturated intermediates react with component amino acid units.
It is increasingly apparent that cPHB is ubiquitous in biological cells, both prokaryotic and eukaryotic (Reusch, 1989(Reusch, , 1992Reusch et al., 1992;Seebach et al., 1994). With the methods used in this study, we find that cPHB comprises 0.36 -0.55% of the dry weight of log-phase E. coli cells. However, these values are based on assumptions drawn from the behavior of granule PHB (Fig. 3), and it is very probable that we are still underestimating cPHB. Our studies show that cPHB is widely distributed within the cell, wherein it is largely associated with proteins. Surprisingly, the majority of this lipophilic polyester is associated with proteins in the cell cytoplasm, particularly proteins of the ribosomal fraction ( Fig. 6 and Table IV). Studies are now in progress to identify the proteins associated with cPHB.
When cells are made genetically competent, there is de novo synthesis of cPHB amounting to 15-20% of the extant cPHB ( Fig. 1B and Table III Reusch and Sadoff, 1988). Only a portion of the newly synthesized cPHB is chloroform-extractable (ϳ10%). Most if not all of this chloroform-soluble fraction of cPHB, until recently believed to be the only cPHB in E. coli, forms complexes with calcium polyphosphate in the plasma membranes (Reusch and Sadoff, 1988) that function as voltage-activated calcium channels . We surmise that most of the remainder of cPHB formed on induction of competence is incorporated into existing PHB proteins, since no new polypeptides were observed and earlier studies showed that there is no de novo synthesis of protein during the development of competence (Reusch et al., 1986). In this study, we have shown that the great majority of cPHB in both log-phase and competent cells is chloroform-insoluble and associated with proteins.
There are a number of ways in which cPHB may associate with protein. As an amphiphilic molecule, PHB can form hydrophobic bonds via its methyl groups or act as a hydrogenbond acceptor or form coordinate bonds to bridging cations via its carbonyl ester oxygens. Individually, such noncovalent bonds are weak, but the frequency at which the methyl groups and ester carbonyl oxygens repeat along the PHB backbone allows each polymer molecule to have many such interactions. The number and strength of these bonds would be determined by the primary structure and three-dimensional geometry of the protein at the binding site(s). The tenacity of the bonding, as evidenced by its ability to withstand heating in sodium lauryl sulfate, extraction with hot chloroform, and proteolysis, suggests that there is also a covalent bond between the protein and the terminal carboxy or hydroxy group of cPHB. Thus far, we have succeeded in removing PHB from its complexes with proteins only by treatments which effect hydrolysis of the ester bonds, i.e. heating in dilute acid or methanol:chloroform mixtures.
PHB is a linear, flexible, polymer chain with a high density of "sticky" sites, alternating polar and nonpolar. This amphiphilic nature may enable cPHB to adjust or reverse the polarity of specific regions of the protein. When the methyl groups of cPHB are bound to nonpolar residues, the ester carbonyl oxygens form a hydrophilic surface which may associate with polar molecules. Conversely, when the carbonyl oxygens bond to polar amino acids or their salts, the hydrophobic coat of cPHB can attract hydrophobic molecules or promote insertion of the peptide into a bilayer.
In summary, we find that E. coli cells contain significant amounts of cPHB. In log-phase cells, most cPHB is associated with specific proteins in both the cytoplasm and membranes, but particularly with proteins of the ribosomal fraction. It seems probable that the cPHB is covalently bound and that cPHB proteins constitute yet another class of protein conjugates. When cells are made genetically competent, there is de novo synthesis of cPHB; about 10% of the newly synthesized polyester forms complexes with calcium polyphosphate in the plasma membranes Reusch and FIG. 6. Western immunoblot of cell fractions of E. coli DH5␣ log-phase cells probed with anti-PHB IgG preadsorbed with PHB-free proteins (see "Experimental Procedures"). The same amount of protein (10 g) was loaded in each lane. Lanes 1-5, total cPHB-free protein, cytoplasm, membranes, ribosomes, and cytosol. Sadoff, 1988;Reusch et al., 1995;Castuma et al., 1995), and the remainder is incorporated into some extant cPHB proteins. These findings imply that PHB is a fundamental constituent of biological cells with important physiological roles in addition to its function in ion transport. Considering the solvent and saltcomplexing properties of this polyester, we suggest it is involved in the inter-and intracellular transport of macromolecules.