Some mathematical calculations were done that provided information about the structure and biochemistry of polyhydroxyalkanoic acid (PHA) granules and about the amounts of the different constituents that contribute to the PHA granules. The data obtained from these calculations are compared with data from the literature, which show that PHA granules consist not only of the polyester but also of phospholipids and proteins. The latter are referred to as granule-associated proteins, and they are always located at the surface of the PHA granules. A concept is proposed that distinguishes four classes of structurally and functionally different granule-associated proteins: (i) class I comprises the PHA synthases, which catalyze the formation of ester linkages between the constituents; (ii) class II comprises the PHA depolymerases, which are responsible for the intracellular degradation of PHA, (iii) class III comprises a new type of protein, which is referred to as phasins and which has most probably a function analogous to that of oleosins in oilseed plants, and (iv) class IV comprises all other proteins, which have been found to be associated with the granules but do not belong to classes I-III. Particular emphasis is placed on the phasins, which constitute a significant fraction of the total cellular protein. Phasins are assumed to form a close protein layer at the surface of the granules, providing the interface between the hydrophilic cytoplasm and the much more hydrophobic core of the PHA inclusion.

Regulation of expression of the phasin PhaP, which is the major protein at the surface of polyhydroxyalkanoate (PHA) granules in Ralstonia eutropha H16, was studied and analysed at the molecular level. The regulation of PhaP expression is achieved by an autoregulated repressor, which is encoded by phaR in R. eutropha. The occurrence of PhaR homologues and the organization of phaR genes was analysed in detail in 29 different bacteria. Three kinds of molecule to which PhaR binds were identified in cells of R. eutropha, as revealed by gel-mobility-shift assays, DNaseI footprinting, cell fractionation, immunoelectron microscopy studies employing anti-PhaR antibodies raised against purified N-terminal hexahistidine-tagged PhaR and in vitro binding studies employing artificial PHA granules. PhaR binds upstream of phaP at two sites comprising the transcriptional start site plus the -10 region and a region immediately upstream of the -35 region of the sigma(70) promoter of phaP, where two imperfect 12 bp repeat sequences (GCAMMAAWTMMD) were identified on the sense and anti-sense strands. PhaR also binds 86 bp upstream of the phaR translational start codon, where the sigma(54)-dependent promoter was identified. PhaR also binds to the surface of PHA granules. In the cytoplasm of a phaROmegaKm mutant of R. eutropha H16, increased quantities of PhaP were detected and the cells formed by this strain were much smaller and had many more PHA granules present than the wild-type. These data support the following model for the regulation of phaP expression. Under cultivation conditions not permissive for PHA biosynthesis or in mutants defective in PHA biosynthesis, PhaR binds to the phaP promoter region and represses transcription of this gene. After the onset of PHA biosynthesis, under conditions that are permissive for the formation of nascent granules, PhaR binds to PHA granules and phaP is transcribed. At the later stages of PHA accumulation, PhaR no longer binds to the granules and the transcription of phaP is again repressed. In addition to this, phaR expression is subject to autoregulation. Excess PhaR that has not bound to the phaP upstream region or to PHA granules binds to the phaR upstream region, thereby repressing its own transcription.

The N-terminal amino acid sequence of the polyhydroxyalkanoic acid (PHA) granule-associated 500 protein of Rhodococcus ruber (the GA14 protein) was analyzed.

The function of the polyhydroxyalkanoic acid (PHA) granule-associated GA14 protein of Rhodococcus ruber was investigated in Escherichia coli XL1-Blue, which coexpressed this protein with the polyhydroxybutyric acid (PHB) biosynthesis operon of Alcaligenes eutrophus. The GA14 protein had no influence on the biosynthesis rate of PHB in E. coli XL1-Blue(pSKCO7), but this recombinant E. coli strain formed smaller PHB granules than were formed by an E. coli strain that expressed only the PHB operon. Immunoelectron microscopy with GA14-specific antibodies demonstrated the binding of GA14 protein to these mini granules. In a previous study, two hydrophobic domains close to the C terminus of the GA14 protein were analyzed, and a working hypothesis that suggested an anchoring of the GA14 protein in the phospholipid monolayer surrounding the PHA granule core by these hydrophobic domains was developed (U. Pieper-Fürst, M. H. Madkour, F. Mayer, and A. Steinbüchel, J. Bacteriol. 176:4328-4337, 1994). This hypothesis was confirmed by the construction of C-terminally truncated variants of the GA14 protein lacking the second or both hydrophobic domains and by the demonstration of their inability to bind to PHB granules. Further confirmation of the hypothesis was obtained by the construction of a fusion protein composed of the acetaldehyde dehydrogenase II of A. eutrophus and the C terminus of the GA14 protein containing both hydrophobic domains and by its affinity to native and artificial PHB granules.

The boundary layer of PHA inclusion bodies isolated from Chromatium vinosum, Pseudomonas oleovorans, and Rhodococcus Tuber was investigated by negative staining and transmission electron microscopy. A typical boundary layer exhibits a basic lattice composed of regularly arranged phasin molecules and phospholipids, forming a thin, nevertheless mechanically stable cover surrounding the content of the inclusion body. Depending on the bacterial strain under investigation, the lattice parameters of the boundary layer may vary. Usually, values between 3.3 and 9.6 nm are observed for the spacing, and the lattice is rectangular. Enzyme particles interpreted as PHA synthase particles are attached to, or inserted into the basic lattice. They cover not more than 20% of the total surface of the inclusion body.These enzyme particles measure between 8 and 12 nm in diameter, are made up of subunits and occur as single units or small aggregates. No indications have been obtained which would support the view that the boundary layer is a double-layer of proteins with phospholipids in between.Rather, a visual inspection of detached boundary layers revealed that the boundary layer is a monolayer exhibiting only one kind of basic lattice. In regions where this monolayer had been artificially removed from the inclusion body, the surface of the contents of the inclusion body was exposed.