Introduction

In 2011, nearly 280 million tons of petrochemicals-based polymers were produced with expected increased in 4 % per annum to 2016 [1]. Production of synthetic polymers is expected to increase to around 810 million tons by 2050 [2]. A strong interest exists in attempting to replace petrochemicals-derived plastics with biologically produced alternatives. Here we review the production, recovery and applications of polyhydroxyalkanoate (PHA) biopolymers.

PHAs are a class of biopolymers with useful physicochemical properties for diverse industrial and biomedical applications. PHAs are biocompatible and biodegradable. PHAs can be produced sustainably using renewable resources and biological methods. Accumulation of PHAs in the bacterium Bacillus megaterium was first reported in 1926 [3]. Since then, many microorganisms have been shown to accumulate PHAs as intracellular granules [4–6] or secrete the polymer extracellularly [3, 7, 8]. Good yields of PHAs have been reported from certain genetically engineered plants [9–12].

Synthetically produced polymers are generally inexpensive, but their persistence in the environment poses a significant problem. Furthermore, production of fossil-based polymers has a significant environmental impact [13] as a net contributor to the level of atmospheric carbon dioxide. Processes for making synthetic polymers often use hazardous materials. In contrast to most petropolymers, biologically produced polymers are generally biodegradable, biocompatible [14, 15] and may be produced sustainably using processes with a reduced environmental impact. Technology for producing various biopolymers is developing rapidly, but because of their relatively high cost, they are used mostly in specialty applications. By 2018, the production of biopolymers is expected to grow by nearly 35 % and around $5 billion worth of biopolymers are expected to be produced [16, 17].

Continuing advances in genetic engineering, metabolic engineering and enzymology are improving accessibility of an increasing number of biopolymers. In addition to production in microorganisms and plants, in vitro enzymatic production of biopolymers is attracting much attention [18–23]. PHAs are commercially produced and are perhaps the best studied of the biopolymers [24, 25]. Nevertheless, they are still relatively expensive and this hinders their wider commercial use.

PHA Production

Production by Microbial Fermentation

In bacteria, PHAs accumulate in the presence of an excess of a carbon source coupled to a deprivation of nutrients such as nitrogen [26]. All metabolizable carbon sources can be used for the production of PHAs, including fatty acids and carbohydrates. PHA polymers accumulate as intracellular inclusions in bacteria of the genera such as Alcaligenes, Pseudomonas, Enterobacter, Necator, Rhodobacter, Ralstonia and Cupriavidus [27–32]. Literature on PHA production in bacteria is extensive [24, 25, 33, 34]. Some of the most recent studies are summarized in Table 1. Certain cyanobacteria also accumulate PHAs under suitable environmental conditions [35, 36] and so do some halophiles [33].

Table 1 PHA fermentation processes
Full size table

Random copolymers of PHA have been successfully produced in monoculture fermentations by controlling the type of carbon feed and composition [37]. This has been shown to be possible using both native and recombinant bacteria [38]. Using Cuprividus necator in fed-batch fermentations to produce the copolymer poly-3-hydroxybutyrate-co-4-hydroxybutyrate, Chanprateep et al. [27] attained a polymer level of 77 % (w/w) in the biomass. This was achieved with a C:N mole ratio of 200:1 with fructose as the precursor for 3-hydroxybutyrate (3HB) and 1,4-butanediol as the precursor for 4-hydroxybutyrate (4HB). Ammonium sulfate was the nitrogen source. The composition of the copolymer, i.e., the ratio of 3HB to 4HB in it, was affected by the molar ratio of the two carbon substrates in the culture medium. For example, if the carbon source contained 25 % (w/w) 1,4-butanediol, a copolymer with a 30 % mole fraction of 4HB was produced, but this could be increased to 80 % by increasing the 1,4-butandiol level to 75 % (w/w). Another commonly used precursor for 4HB is the sodium salt of γ-hyroxybutyrate, but this is more expensive than 1,4-butanediol. Although using too high a concentration of 1,4-butanediol has been found to be toxic to cells [27]. In the production of poly(3-hydroxybutyrate-co-3-hydroxyvalerate) (P3HB3HV) by a Cupriavidus sp. changes in culture pH were reported to affect the composition of the copolymer produced [31].

Accumulation of poly-3-hydroxyoctanoate (PHO) homopolymer has been recently shown to occur in a wild-type Pseudomonas mendocina [39]. Up to 31 % of the cell dry weight consisted of PHO, but the biomass concentration in this preliminary study was relatively low [39]. A two-stage approach was proposed for accumulating the polymer. This consisted of a first stage tailored to achieving a high concentration of the biomass, followed by a second stage tailored to accumulation of the polymer within the cells [39]. Using a controlled feeding strategy based on a metabolic flux balance analysis, Ramalingam et al. [40] achieved a PHA content in the biomass of 35.6 % (w/w) with a PHA concentration in the fermentation broth of 1.14 g L−1 in a continuous fermentation involving Pseudomonas putida MTCC 102 (Type B). Linoleic acid was used as the carbon source.

In some cases at least, the culture temperature can alter whether PHA or some other metabolite is produced preferentially. For example, using Pseudomonas aeruginosa IFO3924 both medium-chain-length PHA (mcl-PHA) and rhamnolipids were produced simultaneously [41]. Production of rhamnolipids was favored by changing the temperature from 30 to 28 °C [41]. P. aeruginosa IFO3924 fed with fatty acids having an even-number of carbons led to the production of 3-hydroxyalkanoates containing only an even-number of carbons in the in the polymer. In contrast, feeding of fatty acids with an odd-number of carbons resulted in 3-hydroxyalkanoates containing both odd- and even-numbered carbon chains. Of the different fatty acids fed, C11 and C12 fatty acids proved to be the best carbon sources for this microorganism [41]. The feeding of C11 substrate resulted in 504 mg L−1 PHA.

Recombinant Escherichia coli harboring phaABC and phaP of Azotobacter sp. has been reported to accumulate PHB (polyhydroxybutyrate) when fed with glycerol [42], a relatively inexpensive carbon source. Accumulation of PHB decreased with an improved supply of oxygen. Changing the carbon source to glucose also led to PHB accumulation, but in this case accumulation was enhanced by an improved supply of oxygen. Presumably, therefore, differences in the oxidation state of the cells differently affect PHB accumulation from carbon sources with different oxidation state [42]. Glycerol has a lower oxidation state of −2 compared to glucose (0). Depending on the oxidation state, catabolism of a carbon source would produce different ratios of NADH/NAD+ and this will determine how much of the carbon flows to the synthesis of a more reduced product [42].

The utilization of complex low cost carbon substrates and the elimination of sterilization energy cost made mixed microbial culture (MMC) to be a cost-effective PHA production process [43]. In fact, it has been suggested that based on life cycle analysis (LCA), PHA production by MMC could be more favorable compared to using pure cultures in both economic and environmental perspectives [44]. MMC have been successfully used in producing PHAs [43]. In such an operation with a 2-stage stirred tank reactor, the use of molasses as the carbon source provided a PHA content in the biomass of 61 % (w/w) [43]. In similar studies with continuous stirred tank reactors and sequencing batch reactors, the fermentation conditions were found to greatly influence the accumulation of PHA in the microbial cells growing on volatile fatty acids [44]. For example, a feast–famine operational regimen affected PHA content in the biomass [44]. The composition of the biopolymer formed in the mixed culture could be manipulated by changing the composition of the carbon source and by whether it was fed pulse-wise or continuously [44]. Such changes could be used to alter the ratio of hydroxybutyrate (HB) and hydroxyvelarate (HV) in the PHA copolymer to change its average molecular weight, degree of crystallinity and other physical properties [44].

The production economics of PHAs depend very much on the cost of the carbon feedstock and whether aseptic (monoculture) or open mixed culture production methods are used. In some applications, the use of mixed carbon sources and microbial cultures as found in municipal wastewaters, for example, may be acceptable for making PHAs. Use of palm oil for producing PHAs has been discussed [34].

In attempts to use a cheap feedstock for producing PHA, ruthenium (Ru) was used for catalytic hydrolysis of inexpensive cellulose to glucose [45]. The hydrolysis occurred at 220 °C and converted 15–20 % of cellulose to glucose and other products. The crude hydrolysate was then used to culture a recombinant E. coli. The bacterium accumulated 42 % (w/w) of its biomass as PHB in a 72 h fermentation at 30 °C, but the concentration of biomass was low because of Ru toxicity. The polymer produced had a number averaged molecular weight of 3.1 × 105 Da. For otherwise the same fermentation conditions, the use of pure analytical grade glucose as the feed yielded a polymer with a somewhat higher number averaged molecular weight of 4.3 × 105 Da.

Methane has been used as a carbon source for producing PHAs using consortia of methanotrophic bacteria as well as single strains. In one such study, Methylobacterium organophilum pure culture was used in a two-phase partitioning reactor [46]. The specific methane consumption rate of the isolate was 100 mg CH4 g−1 h−1 compared to much lower rates reported for consortia of methanotrophic bacteria. The isolate accumulated nearly 57 % (w/w) PHB in the biomass under nitrogen limiting conditions.

Exposure of the activated sludge bacteria to a 7 mT static magnetic field has been claimed to enhance PHB accumulation by the cells under conditions of elevated concentrations of acetate [47]. This effect is said to be a consequence of a reduced uptake of acetate which is toxic to cells in high concentration [47]. The imposed magnetic field alters acetate uptake by modifying the net-charge on the surface of the cell membrane [47].

Certain photosynthesizing microorganisms are known to accumulate PHAs [35, 36]. In cyanobacteria such as Spirulina subsalsa [48], accumulation of PHA has been shown to be influenced by the salinity of the culture medium. A two-fold increase in NaCl concentration enhanced PHA accumulation relative to control, although the PHA yield was relatively low. The mechanism of this ionic strength effect is not entirely clear. PHAs accumulate also in certain halophilic bacteria as reviewed elsewhere [33].

Metabolic Engineering

In bacteria, the synthesis of polyhydroxyalkanoates from sugars normally begins with glycolysis of the sugar to pyruvate. The latter is converted to acetyl-CoA via the pyruvate dehydrogenase (PDH) oxidation pathway. Two molecules of acetyl-CoA are then condensed to form acetoacetyl-CoA through the action of β-ketothiolase, an enzyme encoded by the phaA gene. Acetoacetyl-CoA is reduced by acetoacetyl-CoA reductase (phaB) to form the monomer of (R)-3-hydroxyacyl-CoA, the building block of PHAs. PHA synthase (phaC) finally polymerizes the monomers to PHA (Fig. 1).

Fig. 1
figure 1

General scheme of PHA biosynthesis from sugar catabolism, fatty acid β-oxidation and intermediary pathways

Full size image

Metabolic engineering of PHA biosynthesis pathways (Fig. 1) has been used to produce PHAs of different types and properties in various bacterial species. For example, Escherichia coli strains have been metabolically engineered to regulate the expression of short chain fatty acid catabolism operon to significantly enhance the expression of short chain complexed poly-(R)-3-hydroxybutyrate (cPHB) [49, 50]. Theodorou et al. [50] reported a 1.7-fold increase in accumulation of cPHB in a mutant E. coli compared to the wild-type.

Acetoacetate induction was used to regulate expression of the product [49–51]. Reducing the production of acetic acid capability is reported to improve carbon channeling to polymer biosynthesis. Recently, Jian et al. [52] reported a 2-fold increase in production of biomass and 3.5-fold increase in production of PHA in an E. coli mutant that had been engineered to reduce the excretion of acetate, lactate, ethanol and formate. The excretion of acetate by the mutant was 90 % lower compared to the parent strain [52]. Simultaneous production of succinate and PHA in a metabolically engineered Escherichia coli has also been reported [53, 54]. This was achieved by deleting ptsG, sdhA and pta genes and overexpressing the phaC1 gene of Pseudomonas aeruginosa. The engineered E. coli produced nearly 21 g L−1 succinate and 0.54 g L−1 PHA, equivalent to a polymer content of nearly 6 % in the biomass. The feed used was a mixture of glycerol and fatty acids. The PHA produced consisted of 3-hydroxyoctanoate (58.7 % by mol) and 3-hydroxydecanoate (41.3 % by mol).

Liu et al. [55] reported the use of β-oxidation inhibition to produce mcl-PHA homopolymer in a mutant Pseudomonas putida (KTQQ20) fed with the relevant fatty acid. Six genes of the β-oxidation pathway were knocked out to significantly reduce the fatty acid β-oxidation activity. Feeding dodecanoic acid to the mutant strain resulted in mcl-PHA accumulation at a 10 % (w/w) level in the biomass. The PHA homo-copolymer contained 3-hydroxydecanoate monomer (16 % by mol) and 3-hydroxydodecanoate monomer (84 % by mol). Changing the feed carbon source to decanoic acid resulted in a total PHA accumulation of about 5 % (w/w) in the biomass. The accumulated PHA was a pure homopolymer of 3-hydroxydecanoate. If the feed was changed to tetradecanoic acid, the PHA content of the biomass was higher at 78 % (w/w) and the product was a pure homopolymer of 3-hydroxytetradecanoate. The polymers produced in the β-oxidation inhibited mutant had improved mechanical properties compared to the polymers produced in the native bacterium [55].

Similar observations have been reported in relation to the effect of inhibition of β-oxidation on production of PHA (Fig. 1) in a mutant Pseudomonas entomophila [56]. The β-oxidation inhibited mutant produced by knocking out some of the genes of the relevant pathway, accumulated PHA at the level of >90 % (w/w) in the biomass [56]. The product consisted of mainly (99 % by mol) 3-hydroxydodecanoates. The number averaged molecular weight of the polymer formed was as high as 39,000 Da and it had a polydispersity index of 2.1.

In the biosynthesis of PHA copolymers when the focus is on increasing the mole fraction of a particular monomer in the copolymer, the use of a high concentration of the precursor of the preferentially desired monomer is required in the culture medium. However, a high concentration of certain precursors can be quite toxic to cells. One possible way of overcoming this toxicity is to use a PHA synthase gene that has a high affinity for polymerizing the toxic co-monomer [57]. For example, if a copolymer with a high content of 3HV is wanted, the culture medium would need to be rich in a 3HV precursor such as valeric acid which may be toxic. The producing microorganism may be engineered to contain the PHA synthase gene such as phaC of Chromobacterium sp. USM2, which is reported to have a high affinity towards valeric acid. This strategy was used to accumulate poly(3HB-co-3HV-co-3HHx) (HHx = hydroxyhexanoate) to the level of 86 % (w/w) in the biomass of an engineered Cupriavidus necator [57]. The 3HV monomer content in the terpolymer was nearly 91 % (by mol) and the polymer had mechanical properties comparable to those of the common low density polyethylene [57].

An Aeromonas hydrophila mutant with two genes of the acetic acid metabolic pathway deleted, accumulated poly (3HB-co-3HHx) at the level of 47 % (w/w) in the dry biomass, corresponding to a polymer concentration of around 3 g L−1 in the broth [55]. This level of production was nearly 45 % greater compared to the native strain [55]. Further transformation of the already engineered bacterium to harbor genes relating to fatty acids biosynthesis increased production of PHBHHx by 63 % compared to the strain with only the genes of the acetic acid metabolism deleted.

Yeasts generally have a larger cell size than bacteria and consequently are comparatively easier to recover from the fermentation broth by processes such as centrifugation and filtration. Compared to bacteria, yeast cells are also easier to break for recovering intracellular products [58]. For these reasons, there is an interest in using yeasts to produce PHAs. PHA production by different metabolically engineered yeasts has been reported [7, 8, 59, 60]. For example, Buelhamd et al. [59] produced PHAs in a transgenic Saccharomyces pombe. The yeast was engineered to produce PHB by transfection with the plasmid pBHR68 harboring the PHB synthesis genes encoding β-ketothiolase (phbA Re ), acetoacetyl-CoA reductase (phbB Re ) and PHB synthase (phbC Re ) of Ralstonia eutropha. Under optimized conditions, the yeast accumulated PHB at the level of nearly 9 % (w/w) in the biomass.

Properties of the PHA synthesized in engineered Saccharomyces pombe and Saccharomyces cerevisiae have been discussed in detail [59]. PHA produced in these yeasts was found to have a melting temperature in the range of 153–171 °C. The degree of crystallinity of the product ranged from 27 to 32 % during heating and 36 to 50 % during cooling [7].

A metabolically engineered yeast of the genus Kloeckera accumulated PHA within the cells and secreted a water-soluble bioflocculant polymer in the extracellular medium [7]. The mutant accumulated PHA to the level of about 7 % (w/w) in the biomass. This yeast had been made by transfection with R. eutropha’s phaABC operon harboring genes encoding phaA, phaB and phaC [59].

PHAs accumulate as intracellular inclusions; therefore, their recovery from the cells is inherently expensive. The overall cost of producing PHAs would reduce greatly if the cells could be coaxed into secreting the polymer into the extracellular broth. Extracellular production may be made possible through genetic engineering of the producing cells. This approach is attracting attention. Extracellular secretion of PHA in an Alcanivorax borkumensis mutant specifically engineered for this purpose was observed when the microorganism was fed on either pyruvate or octadecane as the sole carbon source [61].

Recovery of PHAs

Scalable processes are needed for inexpensively recovering the intracellular PHAs from microorganisms. Examples of the recently published recovery methods are provided in Table 2. Although other recovery methods have been published [62], extraction of the polymer with organic solvents appears to be a commonly used approach and has been reported to have an undoubted advantages over the other PHA extraction methods [63]. Its simplicity and rapidity is reported to incur its frequent use in laboratory scale PHA extractions. Solvents extract the polymer without degrading it by improving the cellular membrane permeability and subsequent solubilization of the PHA [63]. Solvent extraction process was also reported to effectively prevent Gram-negative bacterial endotoxin contamination of the polymer, therefore improving the polymeric quality for biomedical applications [64].

Table 2 PHA recovery processes
Full size table

In some cases, the biomass recovered by centrifugation is first washed with an organic solvent to remove fatty acids and oils left from the culture medium. Subsequently, the biomass may be freeze-dried prior to extracting the polymer with a solvent such as chloroform [4]. Cold methanol is then generally used to precipitate the polymer from chloroform [65]. Alternatively, a freshly harvested biomass paste may be washed with acetone and dried under vacuum at ambient temperature prior to solvent extraction of the polymer [66]. PHB-co-PHV has been recovered from Halomonas campisalis cells using a similar process [65] (Table 2). P3HB-co-4HB has been similarly extracted [66] and a similar extraction has been reported for poly-3HB-co-3HV from Pseudomonas oleovorans cells [67]. Variations of solvent extraction methods have been used by others [27, 44, 68] (Table 2). In some processes dichloromethane has been used for extraction instead of chloroform [69].

Some studies have used mechanical disruption by ultrasonication of cells in combination with solvent extraction [70]. Use of other mechanical disruption methods has been reported [62]. Use of chemicals to digest the cell envelope to facilitate solvent extraction has been reported for PHA recovery from Enterobacter cloacae cells [71] as well as from other microorganisms.

A variety of digestive detergents have been evaluated for PHA recovery [72] with varying efficacy. Presumably, the effectiveness of a particular detergent in dissolving the cell envelope depends on the microbial species, but there is no information on this. Detergents such as sodium dodecylsulfate (SDS) have been used to rupture cell membranes to recover granules of the crude PHA polymers [62]. Incubation of R. eutropha and E. coli cells with 5 % (w/v) SDS for 3 to 6 h has allowed recovery of 95 % of the intracellular PHA [72], but the purity of the recovered polymer was improved by extending the detergent treatment to longer than 6 h. Digestion of the cells was enhanced by increasing the incubation temperature. Digestion of cells with alkali (NaOH) has been used for PHA recovery [62, 72] and there is evidence that overzealous treatment with alkali damages the polymer. A treatment regimen for PHA recovery from E. coli may be the use of 1 M NaOH at 50 °C with gentle mixing for 10 min [73].

A comparison of chemical digestion and solvent extraction suggest the latter to afford a greater purity of the PHA product, but if the cells contain a high level of PHA (e.g. >80 % (w/w) of cell mass) digestion with chemicals may give as high a purity as solvent extraction [72]. A case-based evaluation is always necessary in selecting a preferred recovery method.

Enzymatic digestion of Pseudomonas putida cells to recover mcl-PHA with a purity of nearly 93 % was reported by Kathiraser et al. [74]. Solvent extraction gave a product of a somewhat higher purity (~96 % pure). The recovery of mcl-PHA from P. putida cells by combined enzymatic-chemical digestion has been reported [75]. Alcalase and lysozyme enzymes were effective in digesting the cellular material. Chemicals such as sodium dodecylsulfate (SDS) and ethylenediamine tetra-acetic acid (EDTA) helped in solubilizing the non-PHA materials. This combined treatment allowed nearly 90 % of the PHA to be recovered with a purity of nearly 93 % [75]. Enzymes in general may be too expensive for use in a large-scale extraction process. Some enzymes may digest the PHA polymer [74]. Also, cell wall digesting enzymes tend to the microorganism specific [58]; therefore, a case-based evaluation is necessary.

Combinations of enzymatic and mechanical cell disruption treatments have also been used for recovering the intracellular PHAs [76]. Bacteria contain a significant proportion of their biomass as DNA, a jelly-like polymer, and disruption of the cells releases this DNA in the cell homogenate. Therefore, the homogenate can be quite viscous and difficult to process. Thus, enzymes such as benzonase (a commercial nuclease) may need to be added, for example at a concentration of 10 μL/L cell broth at pH 10, to reduce viscosity by digesting DNA and ease processing through certain mechanical cell disruption devices [76]. In some cases, the PHA pellet recovered from the digested cells has been further treated with ozone or peroxide to remove contaminants [76].

Supercritical fluids have attracted attention for PHA recovery [64, 77–79] (Table 2). Of particular interest is supercritical CO2 as it is inexpensive, readily available, does not leave behind a toxic residue, has a low reactivity, is nonflammable and has a moderate critical temperature (31 °C) and pressure (7.29 MPa). Supercritical fluids have been used to extract nearly 90 % of the PHA in the biomass at purities ranging from 86 to 99 % [64, 79] (Table 2). A variety of other relatively less used methods of recovering PHA exist [64].

PHA Production in Genetically Modified Plants

Cost of the carbon source is a substantial contributor to the cost of producing PHAs by microbial processes [80]. Furthermore, recovering the microbial biomass from the fermentation broth and further processing to extract the PHA are expensive. A potentially cheaper production option is to use atmospheric carbon dioxide and sunlight to produce PHAs in genetically modified plants [9, 81, 82]. Plants are easily harvested and a large amount of water does not need to be removed from plant biomass for extracting PHAs. Plant platforms for producing PHA have been extensively reviewed [80, 83, 84]. Some recent studies on PHA biosynthesis in plants are summarized in Table 3.

Table 3 PHA biosynthesis in plants
Full size table

Synthesis of PHB in genetically modified tobacco plant (Nicotiana tabacum) has been reported [85]. The plant had been transformed with a plasmid construct containing genes from Acinetobacter sp. and Bacillus megaterium to code the enzymes required for PHA synthesis. The modified tobacco produced between 17 and 19 % (w/w) PHB in leaf tissue and nearly 9 % in the total plant biomass.

Matsumoto et al. [86] used the codon optimization method to improve expression of PHA in tobacco inserted with the PHA synthesis genes of R. eutropha. The codon-optimization of phaB gene resulted in a two-fold increase in PHB content of the plant tissue compared to the case for the non-optimized gene (Table 3) [86]. In contrast, the codon-optimization of phaC gene had no significant effect on PHB accumulation [86]. This led to the conclusion that phaB gene product had a rate determining influence on PHB production in tobacco leaves [86].

Ariffin et al. [81] reported the transfection of immature palm oil seedlings with Agrobacterium tumefaciens carrying pRMIN and pLMIN plasmids harboring the various PHA synthesis genes. Nearly 90 % of the transfected calli were successfully transformed (Table 3). A PHB yield of 3.7 % of leaf dry weight and 1.2 % of stalk dry weight has been reported for production in genetically engineered switchgrass Panicum virgatum L harboring genes of Ralstonia eutropha and a hybrid gene construct from Pseudomonas oleovorans/Zoogloea ramigera [87] (Table 3).

A high level of PHB expression in plants causes chlorosis, a condition characterized by a reduced chlorophyll production, consequently reduced production of carbohydrate and reduced plant growth [80]. This problem may be alleviated by delaying the synthesis of PHB until the photosynthetic tissues of the plant are well developed. This may be achieved, for example, by using a chemically induced gene-switch ([88]. The gene switching approach has been demonstrated in Arabidopsis thaliana [88]. With this approach, a PHB level of around 14 % (w/w) was obtained in younger leaves and 7 % in older ones [89] (Table 3).

Gene switching is not the only strategy for reducing or overcoming chlorosis. An alternative approach is to produce PHA within the plant peroxisomes [11]. Peroxisomal production of PHA has been reported in A. thaliana and Saccarum sp. (sugarcane) using R. eutropha genes [11]. PHB yields of 1.6 % (w/w) and 1.8 % (w/w) based on dry biomass were obtained in sugarcane leaves and A. thaliana seedlings, respectively (Table 3). Although, peroxisomes were the targeted production sites, in sugarcane PHB accumulated throughout most of the leaf cell including in the peroxisomes and the vacuoles.

Enzymatic Synthesis of PHA In Vitro

Enzyme catalyzed synthesis of PHAs in vitro without involving any microorganisms is an alternative production method [21]. Enzymatic syntheses are generally highly stereoselective, chemoselectvie, regioselective and enantioselective. This ensures a well-defined structure of the synthesized polymer. Furthermore, enzyme catalyzed reactions typically occur under ambient reaction conditions and the separation of the polymer from the reaction mixture is straightforward.

PHA polymers can be produced in vitro from a wide range of substrates including cyclic lactones and carbonates (Table 4). PHAs such as poly-3-hydroxypropionate, poly-4-hydroxybutyrate and poly-6-hydroxyhexanoate have been produced this way. Enzymes used in PHA synthesis include cutinases [20], oxidoreductases [90–92] and hydrolases, especially the serine hydrolases such as lipase (EC 3.1.1.3) [20, 93]. Lipases function as hydrolases in aqueous media to hydrolyze ester bonds [94]; however, in microaqeuous media, lipases can catalyze the formation of ester bonds (Fig. 2). Immobilized lipase B of Candida antarctica has proved to be especially effective in production of PHAs via ring opening polymerization of cyclic lactones [91, 95, 96].

Table 4 Enzyme-catalyzed polymerization
Full size table
Fig. 2
figure 2

Substrate–enzyme interaction at the active site during enzyme-catalyzed ring opening polymerization

Full size image

Cutinase of the soft-rot fungus Humicola insolens has been used to catalyze polycondensations and ring-opening polymerizations of lactones [97]. The optimal activity temperature for this enzyme was 70 °C. The enzyme, immobilized on beads of an ion exchange resin, catalyzed the ring-opening polymerization of both ε-caprolactone (PCL) and ω-pentadecalactone (PDL) in toluene to poly(ε-caprolactone) and poly(ω-pentadecalactone), respectively [97]. Polymerization did not occur if the monomers were changed to (R,S)-β-butyrolactone and L-lactide. In ring-opening polymerization of ε-caprolactone at 70 °C, a monomer conversion of nearly 99 % was achieved (Table 4). The number averaged molecular weight and the polydispersity index of the product depended on whether the polymerization occurred in toluene or in the bulk substrate as the solvent [97].

Using ω-pentadecalactone monomer in toluene, a polymer with a high molecular weight of 44,600 Da could be produced [97]. Lactones with 7- and 16-carbon rings were found to be good monomers for ring-opening polymerization with the immobilized cutinase of H. insolens [97]. The enzyme also catalyzed polycondensation of diols and diacids with good activity depending on the chain length of the substrate [97].

Feder and Gross [98] explored the chain length selectivity of H. insolens cutinase immobilized on Amberzyme oxirane resin, in polycondensation of ω-hydroxyalkanoic acids having 6, 10, 12 and 16 carbons. The catalytic activity increased as the chain length increased from C12 to C16. No polymerization occurred with C6 and C10 substrates. The C16 substrate could be polymerized to a product with a number averaged molecular weight of 40,400 Da. Using the C16 substrate and immobilized C. antarctica lipase B (Novozym 435) as the catalyst, the polymer produced had a number averaged molecular weight of only 25,500 Da [98]. The lipase catalyzed the polymerization of the C12 substrate as effectively as it did that of the C16 substrate, but had no polymerization activity with the C6 substrate [98].

Advances in enzyme catalysis have made possible the production of functionalized PHAs without the need for protection/deprotection of the functional group during the polymerization reaction. Takwa et al. [99] reported the synthesis of ω-functionalized polypentadecalctone containing dithiol, thiol acrylate, diacrylate or dimethacrylate end groups. A single-step solvent-free lipase catalyzed polymerization process was used (Table 4). Two approaches were explored for the synthesis. In the first approach, a difuntionalized polymer with dithiol or thiolacrylate end groups was obtained by mixing the lipase, the lactone and an equimolar mixture of the functional initiator (6-mercapto-1-hexanol) and terminator (11-mercapto-1-undecanoic acid or vinyl acrylate). This way, about 96 % of the polymer was functionalized with dithiol or thiolmethacrylate end groups. In the second approach, a functional diester (ethylene glycol diacrylate or ethylene glycol dimethacrylate) was mixed with lactone and non-dehydrated lipase, allowing the enzymatic water content to serve as initiator. After 2 h of reaction, the pressure was reduced to evaporate the accumulated water so that the reaction could be driven further towards polymerization. This way, more than 96 % of the polymer product could be functionalized with diacrylate or dimethacrylate end groups.

Lipase-catalyzed synthesis of functionalized poly-ω-pentadecalactone-co-butylene-co-carbonate has been reported [19]. A two-stage synthesis was used (Table 4). In the first stage, oligomerization was carried out under a low vacuum (600 mmHg, 18 h, 80 °C). This was followed by the second stage in which polymerization occurred under a high vacuum (2.4 mmHg, 48 h, 80 °C). The copolymer yield from the monomers was nearly 92 % and the product had a high molecular weight of nearly 33,000 Da. The composition of the copolymer could be influence by varying the ratio of the monomers in the feed [19].

Lipase catalyzed synthesis of a copolymer made of 3-hydroxybutyric acid (3HA) and d-glucono-δ-lactone monomers, has been reported [100] (Table 4). Of the several lipases tested for this reaction, the Novozym 435 lipase B from Candida antarctica proved to be the best. The reaction was carried out at 80 °C and did not require a prior derivatization of the monomers. The composition of the product was strongly influenced by the nature of the reaction medium: using a blended tert-butanol/dimethylsulfoxide solvent or a medium of the ionic liquid [Bmim]PF6, around 99 % of the 3HA monomer was converted to product. However, the same reaction carried out in the bulk substrate (i.e., without any solvent) resulted in an almost 1.5 fold reduced conversion of the 3HA monomer [100]. The product had a low molecular weight (≤470 Da) and this was attributed to a poor specificity of the lipase towards 3-hydroxybutyric acid [100].

Controlling the concentration of the initiator in the reaction medium in an enzyme catalyzed polymerization can provide some control over the molecular weight of the polymer formed [101]. For example, in the absence of water (an initiator), the polymer formed tends to have a high molecular weight, but the molecular weight is reduced as the concentration of water in the solvent increases. The use of specific initiators may allow an improved tailoring of the polymerization with a better defined outcome in molecular weight and improved fidelity in the terminating end.

Lipases and cutinases are not the only enzymes capable of polymerizing substrates in vitro. Purified PHA synthase has been shown to polymerize substrates in vitro [102]. Using class II PHA synthase (PhaC1Pp) from Pseudomonas putida and class III PHA synthase (PhaECAv) from Allochromatium vinosum, polyhydroxyalkanoates could be synthesized on a hydrophobic support of highly oriented pyrolytic graphite (HOPG) [102]. A poly-3-hydroxyoctanoate film of a few nanometers thickness was formed on the HOPG support when PhaC1Pp and 3-hydroxyoctanoyl-CoA were used. Using the synthase PhaECAv and 3-hydroxybutyryl-CoA, a homogenous poly-3-hydroxybutyrate was formed on the support. This technology provides a method of forming an ultra-thin PHA film on a hydrophobic support and may have other industrial applications in surface-coatings.

Although enzyme-mediated in vitro polymerization has important advantages, problems remain. Hazardous organic solvents are generally required for achieving high activity with enzymes such as lipases [18, 103]. Also, as polymerization progresses, the concentration of the dissolved polymer increases and so does the viscosity of medium. This imposes diffusion limitations and leads to a polymer with a characteristically low molecular weight [15]. In vitro polymerization with enzymes may be improved by some of the following approaches: (1) microwave irradiation of the reaction mixture [23, 104, 105]; (2) ultrasonic irradiation [96]; (3) replacement of conventional organic solvents with supercritical fluids [106–108] and ionic liquids [104, 109, 110]; (4) use of co-solvent blends [18]; and (5) the use of continuous flow microreactors [111]. In addition, enzyme catalysts themselves may be improved, for example, through molecular engineering [112, 113], modification of immobilization methods [114] and co-lyophilization with non-buffer salts.

Functionalized linear hyperbranched polymers have been produced in supercritical fluids from lactones using lipase-catalyzed synthesis [115]. Supercritical CO2 (scCO2), the bulk substrates (i.e., lactones) and 1,1,1,2-tetrafluoroethane (R-134a) were compared as media for this reaction. Polymerization occurred in both solvents, but was faster in supercritical carbon dioxide. The maximum yield in scCO2 was 89.4 % at 120 h compared to 71.2 % in R-134a and 70.4 % in the bulk substrate. Beyond 120 h of reaction, the yield began to decline in both solvents, but particularly strongly in scCO2 [115]. This was attributed to polymer degradation presumably through hydrolysis.

Enzyme-catalyzed polymerization in a continuous flow microreactor has been reported [111]. The reactor was configured to provide a high catalyst surface area compared to the reactor volume, to improve contact of the substrate with the immobilized enzyme and mass transfer was enhanced through flow. The reaction medium was toluene and the catalyst was Novozym 435 lipase. The reaction occurred at 70 °C. A monomer conversion of >90 % was achieved in less than 5 min in the microreactor whereas a batch process took 30 min to attain a monomer conversion of about 70 % [111].

Molecular modeling techniques have been used to design a mutant lipase of Candida antarctica with a 90-fold increased activity relative to the wild-type enzyme during ring-opening polymerization of d,d-lactide [113]. Simulations of molecular dynamics were used to identify steric hindrances preventing effective catalysis at the active site of the native enzyme. Site-specific mutagenesis was then used to delete three amino acids at the entrance of the enzyme’s active site. This modification caused the aforementioned increase in enzyme activity relative to the wild-type enzyme and improved monomer conversion. Similar approaches have been used to improve the performance of Rhizomucor miehei lipase [116]

A thermophilic lipase from Fervidobacterium nodosum has been reported to catalyze the ring opening polymerization of ε-caprolactone [38]. This enzyme had optimal activity at 90 °C compared to an optimum temperature of 60 °C for most other lipases. The thermophilic enzyme achieved a near 100 % conversion of the monomer and yielded a product with a number averaged molecular weight of 2,340 Da. The F. nodosum lipase (FNL) had a higher affinity towards ε-caprolactone monomer compared to the commonly used Candida antarctica lipase B (CALB). However, in ring-opening polymerization of ε-caprolactone, the specificity and selectivity of FNL were far below those of CALB [38].

A naturally occurring alkaline lipase isolated from Acinetobacter sp. has been reported to be stable in a variety of solvents (ethanol, methanol, isopropyl alcohol, dimethylformamide, dimethylsulfoxide, n-hexane, acetone), retaining 80 % of its initial activity after 90 min at pH 10 and 50 °C [112].

The selectivity of Novozym 435 in ring-opening polymerization of lactones has been found to depend on the conformation of the substrate: in cisoid lactones, the enzyme shows S-selectivity whereas in transoid lactones it has a pronounced R-selectivity. Iterative tandem catalysis has been used to polymerize 6-methyl-ε-caprolactone [117]. In iterative tandem catalysis, two different catalysts work together to accomplish polymer propagation. For example, combining Novozym 435 with a racemization catalyst results in turning the unreactive terminal alcohols with an S-configuration into reactive ones with an R-configuration that can be propagated. Using this approach, a racemic monomer could be quantitatively converted into a homochiral polymer [117]. In vitro production of PHAs has been further reviewed [21].

Chemo-Biosynthesis

Chemical, morphological and physical properties of polymers can be usefully modified by functionalization with different structural and chemical motifs. Synthesis of novel functionalized polymers has been shown to be possible by using a combination of chemical and enzymatic processes [91, 118].

Synthesis of symmetric quintuplet CBABC-type pentablock copolymers has been achieved with a combination of lipase catalysis and atom transfer radical polymerization (ATRP) [118]. The lipase Novozym 435 and ε-caprolactone were used in a first step to produce a tri-block copolymer of di-hydroxyl terminated polycaprolactone block polyethylene oxide (PCL-b-PEO-b-PCL) using the terminal hydroxyl of di-hydroxyl-capped polyethylene oxide (PEO) as initiator (Fig. 3). Further chemical esterification of this tri-block copolymer with α-bromopropionyl bromide in the presence of dichloromethane gave bromine ended tri-block microinitiator that accepted ATRP of styrene in the presence of copper (I) chloride and 2,2′-bipyridine, forming quintuplate pentablock copolymer (PSt-b-PCL-b-PEO-b-PCL-b-PSt). The number averaged molecular weight of this product was around 38,900 Da. The polymer was capable of assuming different self assembled aggregate morphologies in aqueous media. Similar synthetic processes have been reported for H-shaped block copolymers (Fig. 4) [119] and a Y-shaped ABA2-type tri-block copolymer (Fig. 5) [120]. Chemo-enzymatic synthesis of biodegradable PHA copolymers with an excellent-shape memory has been reported [121].

Fig. 3
figure 3

Chemo-enzymatic synthesis of a symmetric quintuplet CBABC-type pentablock copolymer. Adapted from [118]

Full size image
Fig. 4
figure 4

Chemo-enzymatic synthesis of H-shaped block copolymer. Adapted from [119]

Full size image
Fig. 5
figure 5

Chemo-enzymatic synthesis of Y-shaped block copolymer. Adapted from [120]

Full size image

Most of the copolymerization processes require two consecutive steps. After the formation of the first block, an intermediate transformation step is used to convert the end-groups of the block into active micro-initiators for the next block. An alternative one-pot cascade synthesis has been reported for making block copolymers using a bifunctional initiator to allow consecutive polymerization without the need for intermediate transformation steps [122]. However, the one-pot cascade approach does present major challenges in production of copolymers with a high molecular weight [120, 122].

Tajima et al. [123] reported a chemo-enzymatic synthesis of poly-lactate-co-3-hydroxybutyrate in a water–organic solvent two-phase reaction system (Fig. 6). Chemically synthesized thiophenyl (R)-lactate [(R)-LATP] and thiophenyl (R)-3-hydroxylbutyrate [(R)-3HBTP] were used as substrate precursors to first produce hydroxyalkanoylCoA (HACoA) by the ester exchange reaction between the thiophenyl alkanoate and CoA. Then an engineered lactate-polymerizing PHA synthase was used to polymerize the hydroxyalkanoyl-CoA (HACoA) to PHA. This resulted in a copolymer with a number averaged molecular weight of 11,000 Da and a polydispersity index of 1.4. The ratio of the monomers in the copolymer could be controlled by varying the molar ratio of (R)-LATP and (R)-3HBTP fed to the process. Other similar schemes involving a two-phase reaction system have been reported [124].

Fig. 6
figure 6

Chemo-enzymatic synthesis of polylactate-co-3-hydroxybutyrate in a two-phase reaction system. Adapted from [123]

Full size image

Functionalized Biopolymers

Biodegradable polymers are continuously finding applications in numerous fields especially biomedical. However, most of the synthesized biopolymers lack biological stimulus found in either intra or extra cellular matrix, thus specialized biopolymers are needed to be applied for this purpose. Recent advances in biopolymer engineering resulted in significant efforts towards the synthesis of biopolymers with specific functional groups capable of coupling bioactive ligands (Table 5). For example, stimuli responsive biopolymers having an ability to mimic the cellular response process were reported [125–127]. It is not surprising that these biopolymers have attracted much attention recently due to their ability to respond to specific changes in basic environmental stimuli such as temperature [128], pH [129–131], photo [132] and eletro [133] stimuli while others were reported to respond to multiple stimuli [134]. Recently, a thermo-sensitive triblock of PLA-b-PNIPAAm-b-PLA having low critical solution temperature (LSCT) of 31.15–32.62 °C has been reported [128]. It has been reported that the thermal stimuli of these polymers to have arisen as a result of the hydrophobic interactions among PNIPAAm molecular chains, the intermolecular hydrogen bonding between the PNIPAAm chains, water molecules, and the intramolecular hydrogen bonding between the –CONH2 groups [125]. Poly-3-hydroxybutyrate macro initiator was used in ATRP to initiate the synthesis of a novel thermo sensitive amphiphilic triblock hydrophobic PHB flanks by hydrophilic PNIPAAm (Table 5) [135]. Zhu et al. [136] reported the synthesis of polycaprolactone based temperature sensitive polymer (PNIPAM-b-(HEMA-PCL)) using PNIPAAm as the macroinitiator in RAFT polymerization process. Differences in cellular pH have been utilized in designing novel drug delivery devices (Table 5). For instance, Bawa et al. [130] reported the general blood and tissue pH to be about 7.4 while in carcinogenic cells the pH was found to be about 1.0. Yin et al. [137] recently observed the physico-chemical characteristics of pH sensitive PHF-b-PEG micelles (Table 5). The researchers reported that the pK a value of the copolymer can generally be controlled by changing the ratio of the amino acid residue to that of the lactide and ethylene glycol. Taking these biopolymers as drug delivery devices for example, differences in stomach acidic pH to that of intestinal basic pH could determine their target site for delivering the drug. Thus, serve as pH dependent specific delivery devices. Despite their poor biodegradability, poor polymer-cellular interaction and low solubility in most organic solvents; electro conductive polymers were reported to be used as scaffold in nerve regeneration culture and other biomedical fields [138]. Wei et al. [139] reported the synthesis of specialty polymer with enhanced PC-12 cellular attachment and differentiation using a film of PANi funtionalized with bioactive laminin-derived adhesion peptide. Plant bioactive coumarin is reported to be used as photoinduced cross linker in the synthesis of photosensitive polymers [140]. Coumarin encapped PCL-co-TMC were use as photocurable precursors in biomedical devices fabrication and drug encapsulation devices [125, 141]. Polyethyleneglycole functionalized gadolinium ion (Gd3+) has been reported to be used as in vivo paramagnetic probe for magnetic resonance imaging [142].

Table 5 Functionalized biopolymers
Full size table

Sugars such as galactose and mannose were reported to be specific ligands to the ASGPR receptor, which is overexpressed in hepatocellular carcinoma [143]. Jiang et al. [144] observed the solution behavior of PCL functionalized hydroxyethyl cellulose (HEC). Previously, Lu et al. [145] reported the synthesis of poly(1′-O-vinyladipoyl-sucrose) in chemo-enzymatic process resulting in polymer with molecular weight as high as 53 kDa having improved solvation properties that can be explored as promising biomaterial. In general, polymer functionalization has resulted in recent increasing demand of biodegrabable polymers in diverse industrial applications. Functionalization of polymer has opened a new avenue for the production of novel polymers with specific application that were not possible earlier.

Applications of Polyhydroxyalkanoates

Biomedical Applications

PHAs have attracted much attention as materials for biodegradable implants in biomedical and tissue engineering applications. Specifically, PHB has been reported to be biocompatible with various kinds of cells [146] (Table 6).

Table 6 Applications of polyhydroxyalkanoates
Full size table

Use of PHAs as drug delivery systems (Table 6) for prolonged release of therapeutics into systemic circulation is receiving attention [147]. Polyhydroxyesters such as block copolymer of PEG-b-PCL in the form of micelles and nanoparticles have been used for parenteral delivery of taxanes [148] (Table 6). A matrix of nanoparticles of poly-3-hydroxybutyrate-co-3-hydroxyhexanoate (PHBHHx) has been used to deliver antineoplastic agents to cancer cells [149] (Table 6). PHB nanoparticles functionalized with a tumor-specific ligand have been examined for specifically targeting certain breast cancer cells [150]. The non-steroidal anti-inflammatory drug ibuprofen has been conjugated to nontoxic oligo(3-hydroxybutyrate), in attempts to improve drug delivery but this novel formulation remains to be thoroughly assessed [151].

Poly-3-hydroxybutyrate microsphere have been tested in vitro for releasing the antibiotics gentamycin and tetracycline [152]. Multifunctional PHB/45S5Bioglass composite system has been discussed as drug delivery agents and for use in certain bone tissue engineering applications [152].

Polycaprolactone (PCL) tends to be highly permeable and this is an attractive feature in some drug delivery applications. Use of PCL in certain drug delivery applications has been approved by the US Food and Drug Administration [146]. PCL degrades slowly (2–4 years) in vivo and is therefore useful for developing drug release implants for long-term use [153].

Silver nanoparticles have attracted much attention because of their antibacterial properties [154–156], but slurries of such particles tend to be unstable. Use of polyhydroxyalkanoates in prolonging stability of such slurries has been reported [157].

Poly-3-hydroxybutyrate-co-3-hydroxyhexanoate scaffolds have been evaluated for use in eyelid reconstruction in experimental animals [158] (Table 6). Although the scaffold performed satisfactorily, it produced some inflammation that took about 2 weeks to clear. Poly-3-hydroxybutyrate-co-3-hydroxyhexanoate was found to induce cartilage development from mouse mesenchymal stem cells and preserve the chondrocytic phenotype of the cells [159] (Table 6).

The use of a terpolyester of 3-hydroxybutyrate-b-3-hydroxyvalerate-b-3-hydroxyhexanoate (PHBVHHx) as scaffold for promoting differentiation of human bone marrow mesenchymal stem cell into nerve cells (Table 6) has been reported [160]. PHBVHHx scaffolds with a pore size of 30–60 μm were found to be best.

PCL reinforced with phosphate glass fibers has been used to make fixation pins for intramedullary fractures, craniofacial repairs and general bone repair [146]. PCL can be modified in various ways to improve mechanical strength and alter properties such as degradability, compatibility, hydrophilicity and crystallinity. For example, poly-ε-caprolactone functionalized polyethylene glycol copolymer has been reported to have a strong amphiphilicity, a controlled biodegradability and excellent biocompatability.

PCL blended with poly-glycerol sebacate has been used as a fibrous scaffold for aortic valve regeneration [161]. Use of both PCL and PHA has been reported as substrates for cardiovascular tissue engineering [162] (Table 6). The diverse biomedical applications of PCL have been further reviewed by others [163, 164]. Among the various other biomedical applications, PHA is being used in tablet formulations [165, 166], surgical sutures [39], wound dressings [167, 168], surgical implants to join tubular body parts [169, 170], controlled release contraceptive devices [171–173], lubricating powders, blood vessels, tissue scaffolds, and bone fracture fixation plates [39, 174–177]. Of these applications, surgical sutures constitute perhaps the largest use category with a 2010 market value exceeding US$1.3 billion annually [178].

Industrial Applications

In addition to their biomedical applications, PHAs can potentially replace petrochemicals-based plastics in diverse other applications (Table 6).

A recently suggested application is the use of PHAs as precursors of biofuels. Like bioethanol from sugars, PHAs can be made into renewable biofuels. Hydrolysis of PHAs followed by methyl esterification provides 3-hydroxyalkanoates methyl esters with an energy content that is comparable to that of bioethanol [53, 179, 180]. Whether hydroxyalkanoate esters would be as cheap as bioethanol in fuel blends is debatable. This is because in making a PHA-based fuel, a carbon source first needs to be polymerized, the polymer then needs to be hydrolyzed and a subsequent methylation step is required. In contrast, glucose and other sugars can be directly fermented to bioethanol.

Polyhydroxyalkanoate latexes have been used in the paper industry for surface coating of paper and as sizing agents [181] (Table 6). PHB has had limited applications in packaging because of its high glass transition temperature which results in brittleness under typical use conditions. PHB’s utility in packaging has been improved by copolymerizing with various levels (5–20 %) of valerate to produce PHBV polymers with glass transition temperatures of 4 °C at 5 % and 1 °C at 20 % valerate content [182] (Table 6). Novel polymeric materials for food packaging consisting of PCL blends have been developed [183]. They have been reported to have an excellent durability and a remarkable tensile strength [183] (Table 6). Uses of PHAs in the food industry have been discussed elsewhere [184, 185].

PHAs appear to be potentially useful in controlling bacterial pathogens in certain aquaculture applications [186–188]. For example, administration in the feed of 1,000 mg L−1 of PHB particles of an average diameter of 30 μm, or addition of inactivated cells (107 cells mL−1) of PHB-containing Brachymonas bacteria (equivalent to ~10 mg L−1 PHB) to the culture water of brine shrimp (Artemia nauplii) larvae, conferred a complete protection from a virulent strain of the intestinal pathogen Vibrio campbellii [187]. Other similar reports have claimed an inhibitory effect of PHB on certain gut microflora of the giant freshwater prawn (Macrobrachium rosenbergii) larvae [188]. Administration of PHB (5 g L−1) in the feed significantly increased the survival of the prawn larvae and improved their development. The total bacterial counts and Vibrio spp. counts were significantly reduced in PHB-fed larvae compared to the control larvae.

Polyhydroxyalkanoates have been used as controlled release agents for herbicides in agricultural. Controlled release potentially reduces the impact of the herbicides on nontarget species and reduces the need for repeated applications [189]. Micro- and nanoparticles of PHB and PHBV were used in a controlled release formulation of the herbicide ametryn [189] (Table 6). PHB has been successfully tested for removing lipid-soluble organic pollutants from water by adsorption [190] (Table 6).

Concluding Remarks

PHAs are versatile biopolymers with diverse applications. PHAs are biodegradable and they may be produced in a sustainable way using renewable feedstocks. PHAs can be produced in vitro using enzymes without involving microbial cells. Alternatively, they may be produced by microbial fermentation processes and using recombinant plants. Production of PHAs using microorganisms and enzymes remains relatively expensive compared to plastics derived from petroleum. Production in plants is likely to establish itself as the least expensive option for certain PHAs, but does not currently offer the molecular versatility of the products that can be made using microbial fermentations and enzymes. In view of their relatively high cost, PHAs are likely to be used first in high-value niche applications, particularly in biomedicine, but their broader industrial use will increase as the cost of production declines.