The section above described in general that biosynthesis of PHAs is often carried out using sugars, plant oils, and fatty acids. However, usage of these expensive carbon sources are not feasible for scaling up PHA production as it may add a significant cost burden to production. With the current growing interest in utilizing PHA as an alternative plastic material to replace some petro-plastics, it is crucial to source cheaper carbon substrates that can markedly decrease the production cost. Waste plant oils, be it from the edible oil industry or other agricultural sectors, stand as inexpensive and excellent carbon substrates to substitute the commonly used carbon sources in PHA biosynthesis. WCO, for example, is one of the major wastes produced from the human food processing industry. It is estimated that over 29 million tons of WCO is produced annually around the globe (Maddikeri et al., 2012), giving it high potential as an inexpensive carbon substrate for PHA production. The high content of free fatty acids (FFA) in WCO gives it an added advantage to be used as a carbon source for PHA biosynthesis. Song et al. have reported the utilization of waste vegetable oil in the synthesis of various types of mcl-PHAs (C6-16) from Pseudomonas sp. strain DR2 (Song et al., 2008). Rao et al. have successfully biosynthesized a P(3HB-co-4HB) copolymer using spent palm oil after frying as the main carbon source while co-supplementing with 1,4-butanediol as a precursor compound (Rao et al., 2010). Another study has reported the usage of waste frying oil (rapeseed oil) for the production of P(3HB) in C. necator strain H16 (Verlinden et al., 2011). Another team of researchers have also explored the utilization of waste rapeseed oil obtained after frying for PHA biosynthesis via Pseudomonas sp. strains G101 and G106 (Mozejko et al., 2011). This study reported the production of mcl-PHAs with Pseudomonas sp. strain G101 producing predominantly 3-hydroxyoctanoic acid (3HO) and 3-hydroxydecanoic acid (3HD) monomers, while strain G106 produced significant amounts of 3-hydroxyhexanoate (3HHx) monomers (Mozejko et al., 2011). In a later study by the same group, the production of mcl-PHAs by Pseudomonas sp. strain G101 was markedly improved using a pulse-feeding method of waste rapeseed oil, whereby 44% of the total cell biomass was found to contain mcl-PHAs after 41 h of fermentation (Mozejko and Ciesielski, 2014). In a separate study by Mozejko and Ciesielski, saponified waste palm oil (SWPO) was used as the sole carbon source for the production of mcl-PHA using Pseudomonas sp. strain G101. It was found that under fed-batch fermentation conditions with a total SWPO fed at 15 g/L, the bacterial strain in the study could produce up to 43% of mcl-PHA by dry cell weight (Mozejko and Ciesielski, 2013).

Kamilah and co-workers have also shown that wild type and engineered strains of C. necator were able to utilize WCO (palm oil) in the production of P(3HB) and P(3HB-co-3HHx), respectively (Kamilah et al., 2013, 2018). In a separate study, it was reported that P. putida strain KT2440 was able to produce high contents of mcl-PHAs when grown via high cell density fermentation using hydrolyzed WCO as the sole carbon source (Ruiz et al., 2019). Kourmentza et al., on the other hand, have reported co-production of P(3HB) polymers along with rhamnolipids when Burkholderia thailandensis was employed for the bioconversion of used cooking oil. It was reported that up to 60% of the cell biomass consisted of P(3HB) while 2.2 g/L of rhamnolipids were produced simultaneously (Kourmentza et al., 2018).

Besides WCO, there are other categories of plant oils that are of no value for human consumption that may be used for PHA production. One example is the use of jatropha oil. This oil is extracted from the seeds of the plant Jatropha curcas L, which has no food value. It has been shown that by utilizing jatropha oil, C. necator strain H16 was able to successfully synthesize P(3HB) homopolymers, which shared similar physicochemical attributes with homopolymers produced via palm oil or other carbon sources. While jatropha oil is known for its toxicity to humans and animals, it did not show any inhibition to the growth of C. necator during PHA biosynthesis (Ng et al., 2010). In another study, it was shown that Pseudomonas oleovorans (ATCC 29347) was able to assimilate saponified jatropha oil as the sole carbon source to produce the copolymer P(3HB-co-3HV) (Allen et al., 2010). In a more recent report, Zainab et al. have explored a series of underutilized inedible oils obtained from Africa and some parts of Asia as sole carbon substrate for PHA biosynthesis. These oils include African elemi oil, bitter apple oil, desert date oil, and Amygdalus pedunculata oil (Zainab-L et al., 2018). It was found that wild type and engineered C. necator strains were able to metabolize these oils into P(3HB) homopolymer and P(3HB-co-3HHx) copolymer, respectively, with these two polymers having high molecular weights (M_w) (500,000–2,400,000 Da). The 3HHx proportions in the copolymers were reported to be as high as 31 mol% when biosynthesized using these oils (Zainab-L et al., 2018).

In the production of palm oil in the palm oil mills in Malaysia, Indonesia, and Thailand, a wet milling method is used, whereby the fresh fruit bunch (FFB) is subjected to various stages of treatment before oil is pressed, purified, and refined. These processes require the usage of large volumes of water. It is estimated that for every ton of FFB, 1.5 m³ of water is used. Half of this water ends up as palm oil mill effluent (POME) and is collected in a pond outside of the mill (Mumtaz et al., 2010). During the initial stage of POME discharge, the floating residual oil that is collected is known as sludge palm oil (SPO), which exists as a dark brown solid at room temperature (25°C) and exudes a strong odor. This low-grade oil that is rich in FFA is usually used to make cheap laundry soaps, candles, and animal feed additives. SPO has also been explored as a carbon feedstock for the production of PHA due to the high content of FFAs. In a report published by Kang et al., it was observed that SPO could be utilized by several Pseudomonas species to produce a myriad of mcl-PHAs. Among all the strains tested, P. putida strain S12 was found to produce the highest yield of elastomeric mcl-PHAs (Kang et al., 2017). A recent study by Thinagaran and Sudesh has revealed that emulsified SPO, when supplemented at 10 g/L to the engineered C. necator strain Re2058/pCB113, produced as high as 9.7 g/L of dried cells mass in which 74 wt% contained P(3HB-co-21 mol% 3HHx). Further improvements in the yield were achieved by employing a fed-batch fermentation strategy, which resulted in biomass productivity of 1.9 g/L/h and PHA productivity of 1.1 g/L/h (Thinagaran and Sudesh, 2019). Table 1 summarizes all the bacterial strains and the respective waste plant oil feedstocks that were used for PHA production.

In most cases, scl-PHA such as P(3HB) is highly crystalline with poor tensile strength while mcl-PHA tends to be amorphous and very elastomeric. The former is too brittle whereas the latter can be too sticky to be processed conveniently. However, a mcl-PHA, P(2.1 mol% 3HD-co-97.9 mol% 3HDD), produced by Pseudomonas entomophila LAC25 containing almost pure 3-hydroxydodecanoate (3HDD) and a small fraction of 3HD, has the widest difference between T_g (−49.3°C) and T_m (82.4°C) (Chung et al., 2011). High contents of long-chain monomers favor side-chain crystallization, unlike the amorphous conventional mcl-PHA, which generally have a high composition of 3HO and 3HD (Ouyang et al., 2007; Ma et al., 2009; Chung et al., 2011). Due to this unusual effect, Chung et al. (2011) proved that the polymer produced by P. entomophila LAC25 consisting of almost pure 3HDD could lead to improved mcl-PHA properties.

Alternatively, incorporation of a minor quantity of 3HHx (5 mol%) into P(3HB) can reduce the melting point to below 155°C (Matsusaki et al., 2000; Loo et al., 2005). In fact, Grigore et al. (2019) revealed that P(3HB-co-5.9 mol% 3HHx) had the highest elongation to break value of 163%. When discussing about the molecular weight of P(3HB-co-3HHx), Doi (1990) and Riedel et al. (2012) revealed that a decrement in molecular weight (2–11 × 10⁵ Da) was observed when the 3HHx molar fraction became higher. Unexpectedly, a slight increase in the molecular weight was discovered by Murugan et al. (2017) using polymers with 4–15 mol% 3HHx in their studies. The molecular weight was between 5.47 × 10⁵ and 6.85 × 10⁵ Da and the viscoelastic behavior further supported their findings using a rheometer (Murugan et al., 2017). Therefore, developing combinations of scl-mcl monomers can produce a polymer that overcomes the defects of each monomer to possess soft and flexible properties.

On the other hand, when the 3HHx composition increased from 0 to 17 mol%, the tensile strength decreased from 43 to 20 MPa while the elongation to break improved from 6 to 850% (Doi et al., 1995). The degree of X-ray crystallinity of solvent-cast copolymer films decreased from 60 to 18% as the 3HHx fraction increased from 0 to 25 mol% (Doi et al., 1995). This finding was also supported by Volova et al. (2016), who used X-ray structure analysis to demonstrate a significant reduction in the crystalline to amorphous ratio of different P(3HB-co-3HHx) samples by increasing the 3HHx monomer fraction. This phenomenon suggests that 3HHx monomers are excluded from the P(3HB) crystalline phase and therefore reduce the overall crystallinity (Doi et al., 1995). Yu (2007) reported that the crystal lattice of P(3HB-co-3HHx) is reduced to about half that of the P(3HB) lattice. The T_m also decreased steeply from 180 to 52°C with similar 3HHx molar fractions (0 to 25 mol%) and T_g was reduced from 4 to −4°C (Doi et al., 1995). Interestingly, with 3HHx fractions >60 mol%, the elongation was almost 40–50 times higher than with 12 mol% 3HHx monomer. The Young's modulus also decreases with greater proportions of 3HHx, from 1,286.4 (12 mol% of 3HHx) to 217.0 MPa (68 mol% of 3HHx) (Volova et al., 2016).

Waste oils are indeed an attractive carbon feedstock for PHA production due to its inexpensive cost and availability. However, the challenge remains on how efficiently these waste feedstocks can be used by the bacterial strains for PHA production. One of the major factors that determine efficient utilization of these waste oils is its assimilability by the microbes for PHA biosynthesis. Although feedstocks such as SPO are rich in FFAs, it was found to solidify instantaneously when added to the culture medium (Thinagaran and Sudesh, 2019) causing the oil to be not available homogenously in the bacterial culture. This has resulted in the conversion yield of SPO to biomass to be lower than conversion yield of CPKO to biomass (Thinagaran and Sudesh, 2019). SPO was made readily available for the bacterial strains when emulsified using surfactants. It also ensured that the oil droplets were dispersed more homogenously in the culture medium resulting in improved cell biomass compared to the non-emulsified SPO. Similar strategies were also employed by Riedel et al. when biosynthesis of PHA was performed using waste animal fats and tallow, which exist as solid in room temperature (Riedel et al., 2015). Another important factor in determining efficient utilization of waste oils is the choice of bacterial strain for PHA biosynthesis. While many Pseudomonas strains are able to efficiently utilize the waste oils to produce mcl PHAs as shown in the previous section, strains such as C. necator, which is also the model PHA producing strain is able to utilize oils to produce only scl-PHA [P(3HB)] in its wild-type form. This homopolymer is less favorable for industrial applications due to its brittle nature. As such, the wild-type C. necator has been extensively engineered by various genetic tools to enable the biosynthesis of scl-mcl PHA [e.g., P(3HB-co-3HHx)] using waste plant and animal oils. The scl-mcl PHAs are more durable for industrial applications due to their softness and high flexural strength. Besides that, the PHA production efficiency using waste oil feedstocks can be enhanced by altering the fermentation strategy that is involved in the PHA biosynthesis. More often, the biosynthesis of PHA are conducted as batch fermentations in shake flasks to determine the ability of bacterial strains to utilize certain carbon feedstocks for PHA production. Based on the observations in shake flask experiments, the biosynthesis is usually scaled up in a bioreactor system and this is where many parameters have to be carefully studied and manipulated to generate a maximum conversion yield of carbon feedstock to biomass. Based on the reported studies using bioreactors for PHA production, it was found that fed-batch fermentation strategy and/or pulse-feeding of carbon feedstocks are the preferred methods for efficient PHA production using waste plant or animal oils (Mozejko and Ciesielski, 2013, 2014; Riedel et al., 2015; Zainab-L et al., 2018; Thinagaran and Sudesh, 2019). Employing a suitable fermentation strategy also goes hand in hand with a proper control of some of the key bioreactor conditions such as dissolved oxygen (DO), impeller design, agitation speed of impeller, aeration rate, pH, and temperature control for an efficient utilization of waste oil feedstocks for PHA production. While foaming indicates a good bacterial growth during fermentation process, it may pose a serious problem in smaller scale bioreactors if not handled properly. The bacterial cultures may flow out continuously causing dilution of the fermentation medium and inefficient utilization of carbon feedstocks. Proper control of foaming is necessary either by using a suitable anti-foam compound or a mechanical foam breaker. This is especially important when high cell density fermentations are performed. All these factors are interrelated for an efficient PHA production process and therefore cannot be omitted. Figure 1 shows the involvement of these factors in the entire PHA biosynthesis process. Answering the question posed in the title of this review, “Can polyhydroxyalkanoates be produced efficiently from waste plant and animal oil?,” the answer is a yes. These waste plant and animal oil feedstocks can be definitely used for an efficient PHA production, providing the three major factors proposed in this section are properly optimized. In fact, waste plant and animal oils are a potential good source of carbon feedstocks for an economical and sustainable PHA production commercially.

After fermentation, the downstream processing of PHA plays a noteworthy role in production cost. Unlike upstream processing, reports on the downstream processing and purification of PHAs are scarce. PHAs are present as amorphous solid granules suspended in the cytoplasm of the cells with 5–10 wt% of water along with a protein component containing PHA synthase and PHA depolymerase that binds strongly to the PHA granule (McCool et al., 1996; Bresan et al., 2016). The separation of PHA from non-PHA cell mass (NPCM) is technically challenging because both PHA and NPCM are in the solid phase. Some downstream processing strategies are discussed in this section (Figure 1).

To date, various methods have been established to separate and purify PHAs from NPCM with varying efficiencies. More often than not, the PHA is purified by either solubilizing the PHA or NPCM dissolution. Chemical treatments can be broadly divided into two categories. One is solvent extraction, where PHA molecules are dissolved in appropriate solvents and later the PHA is precipitated. The second method is digestion. While the solvent extraction techniques involve the solubilization of PHA granules, in the digestion technique, NPCM is dissolved.

The most commonly used solvents are chlorinated hydrocarbons, namely, chloroform, 1,2-dichloroethane, methyl ethyl ketone (MEK), and methyl isobutyl ketone (MIBK), or some cyclic carbonates such as ethylene carbonate and 1,2-propylene carbonate (Lafferty and Heinzle, 1978; Ramsay et al., 1994; Jiang et al., 2006). PHA is precipitated using non-solvents like methanol or ethanol (Kunasundari and Sudesh, 2011). The solvent extraction method is considered to be the most effective in terms of purity. Optimized PHA extraction in C. necator with chloroform was able to recover 94% of the polymer with 98% purity (Fiorese et al., 2009). This method is commonly used on the laboratory scale but has not been widely adopted on the pilot scale and for commercial processing. This is due to its high operational and capital cost along with waste disposal issues (Tamer et al., 1998; Yu and Chen, 2006; McChalicher et al., 2010).

The digestion method can be further sub-classified into chemical digestion and enzymatic digestion. In chemical digestion, the NPCM is digested by sodium hypochloride or surfactants like sodium dodecyl sulfate (SDS), Triton X-100, palmitoyl carnitine, or betaine. Sodium hypochloride has a strong non-selective oxidizing property that can digest the NPCM (Yu and Chen, 2006). Sequential treatment with sodium hypochloride and a mixture of surfactants promoted the recovery of PHA with 50% reduced cost when compared to solvent extraction (Yu, 2009; Divyashree and Shamala, 2010). Even though chemical digestion has a low operating cost, it is considered unattractive due to complications in wastewater treatment and also the cost involved in the utilization of surfactants (Kunasundari and Sudesh, 2011).

The recovery of PHA via enzymatic degradation is a complex procedure. In general, the dissolution of NPCM is followed by heat treatment, enzymatic hydrolysis, and surfactant washing. Various types of enzymes have been studied for their activity to digest NPCM, but the most commonly employed is protease. The enzymatic degradation technique is highly efficient and yields up to 92% purity because the enzymes are specific in their action. However, the cost of enzymes and the complexity of the procedure make it unattractive for industry (Middelberg, 1995; Kapritchkoff et al., 2006).

Mechanical disruption is the most common method for the removal of intracellular protein. Bead milling and high-pressure homogenization are the most commonly used mechanical disruption techniques for PHA recovery. Mechanical disruption is the most favored method for PHA recovery due to the minimal damage caused to the polymer during the recovery process (Kunasundari and Sudesh, 2011). Additionally, mechanical disruption is cost-effective and environmentally friendly.

Biological recovery is a novel method that bypasses most of the complicated processes mentioned above. The concept is based on utilizing animals, which are fed with dried bacterial cells containing PHA. The animals will digest the NPCM and defecate the PHA (Murugan et al., 2016b; Kunasundari et al., 2017). Mealworms and rats are attractive animal models to be used for PHA purification (Murugan et al., 2016b; Ong et al., 2018a). A detailed study was conducted to evaluate the efficiency of PHA recovery from C. necator cells by mealworms as well as the purity of the recovered PHA (Murugan et al., 2016b). The purity of PHA recovered from mealworms was lower than that recovered by rats (Ong et al., 2018b). However, washing the fecal pellets with water and 1% sodium dodecyl sulfate (SDS) was able to improve the purity up to 100% (Murugan et al., 2016b). Later, the 1% SDS was replaced with 0.1 M sodium hydroxide (Ong et al., 2018a). PHA recovered by this method was close to its native morphology in the bacterial cells, while the molecular weights of PHA were not affected (Kunasundari et al., 2017). Currently, this biological recovery method is being scaled-up at a mealworm farm in Malaysia. The entire biological recovery process using mealworms is summarized in Figure 2.

Apart from the abovementioned recovery methods, there are several other methods that are still at the laboratory scale. Some of these methods are briefly described and discussed below.

Super critical fluids (SCF) are considered to be a promising future technique to extract PHA from bacterial cells. SCF are solvents with high density and low viscosity. Supercritical-carbon dioxide is the most common SCF solvent for PHA due to its low toxicity, low reactivity, low cost, inflammable temperature and pressure (31°C and 73 atm), and ready availability (Hejazi et al., 2003; Khosravi-Darani et al., 2003; Darani and Mozafari, 2010). Many parameters have to be optimized for this method. However, 89% of PHA has been recovered from C. necator using SCF at 200 atm, 40°C, and 0.2 ml of methanol (Hejazi et al., 2003).

Cell fragility is another recovery method, in which the composition of the growth medium is altered to increase the osmotic fragility of the cells after the PHA production phase. Cells grown in inorganic salt media have a deficiency in diaminopimelic acid (DAP) and decreased concentrations of amino acids. DAP is a key component in the bridging of peptidoglycans in the cell wall and hence it plays a crucial role in cell wall stability. About 86–100% of PHA was extracted using hot chloroform. The crucial step in this process is to maintain a balance between cell wall softening when desired and cell wall integrity while PHA is still accumulating (Kunasundari and Sudesh, 2011).

Recovery using gamma radiation provides optimal cell disruption at lower dosage. Following disruption, PHAs from the cells were extracted using organic solvents, alkaline or acidic solutions, surfactants, or enzyme digestion (Bhattacharya, 2000). However, the initial investment cost for this method is extremely high.

In “Agenda 21” adopted at the special United Nations Conference on Environment and Development in 1992, the need for the development and application of new, environmentally friendly processes and materials was emphasized (Sitarz, 1993). Hence, the focus was laid on finding an alternative to petro-plastics due to their non-degradable nature that contributes to the accumulation of solid waste around the world. PHAs hold an exclusive position among biodegradable polyesters with a number of useful properties and a wide range of applications (Cornell, 2007). The attractive properties of PHAs have contributed to its continued studies by both industry and academia.

As the volume of PHA production increases and its application widens, it is important to study the mechanism of biodegradation of PHAs in various environmental conditions. PHAs are degraded by various microorganisms to produce CO₂ and water in aerobic conditions, and methane and water in anaerobic conditions (Hodzic, 2004; Volova et al., 2010). PHA is able to degrade in most environments that have microbial activities, including soil (Mergaert et al., 1993), compost (Weng et al., 2011), sewage sludge (Lee and Choi, 1999), fresh water, and sea water (Ohura et al., 1999). PHA depolymerases are the key enzymes in the degradation of PHA. The main function of the PHA depolymerase is to hydrolyze the long, water-insoluble PHAs into shorter, water-soluble forms so that microbes can uptake and metabolize the monomers as nutrients (Sudesh and Abe, 2010). The rate of degradation of PHAs is dependent on their characteristics such as monomeric composition, side chain, and crystallinity. The degradation of a particular polymer is also affected by various aspects such as the type of depolymerase, temperature, nutrient availability, and moisture content (Mergaert et al., 1993).

P(3HB) is the most commonly synthesized PHA in most wild-type bacteria. During intracellular degradation, P(3HB) is oxidized by a dehydrogenase and converted into acetyl-CoA by β-ketothiolase and no toxic substances are formed in the cell (Lemes et al., 2015). This contributes to the bio-compatibility of PHAs. Microorganisms capable of producing PHAs express both PHA synthase and a complementary PHA depolymerase.

In general, the degradation rate of PHA will increase with the density of the microbial population. In a study of the degradation of P(3HB-co-3HV) by a microbial community, the degrading microbes were found to attach and colonize the polymer before the degrading enzymes were secreted (Sang et al., 2000). Most model microorganisms for PHA research express enzymes with high specificity toward P(3HB). However, various other microbes have also been identified that show a wide substrate specificity (Jendrossek and Handrick, 2002). For instance, genera such as Xanthomonas and Comamonas are able to degrade PHAs with aromatic side chains (Doi et al., 1992). Later, it was found that clear zones were observed on plates overlaid with P(3HB), P(3HO), and P(3HPV). Hence, Quinteros and co-workers concluded that there is a correlation between the length, composition, and side chains of the polymer and speed of degradation, where longer side-chain PHAs are degraded faster than those with shorter side chains (Quinteros et al., 1999).

Apart from the side chains, composition of the polymer also plays a significant role in degradation. Manna and Paul (2000) evaluated the ability of microbes to degrade P(3HB) and P(3HB-co-3HV) in different environments such as soil, water, compost, and sewage sludge. They found that the rate of degradation of homopolymers is higher than for copolymers (Manna and Paul, 2000). These results were in agreement with Doi (1990), who used purified enzymes from Alcaligenes faecalis to degrade P(3HB) and P(3HB-co-3HV). However, contrasting results were observed in the studies conducted by Mergaert et al. (1992, 1993, 1994), wherein P(3HB-co-3HV) degraded at higher rates than P(3HB) in soil and compost environments. Large variations in the rate of degradation are always observed when studies are carried out in natural environments. This is due to variations in conditions observed in different locations as well as the depolymerases present in the resident microorganisms and their substrate specificities (Manna and Paul, 2000).

The PHA degradation ability of microbes is inversely proportional to the crystallinity and the molecular weight of the polymer. This was determined when Kusaka et al. examined the degradation rate of high-molecular-weight 3HB with different crystallinities (65–85%). A similar trend was also observed in another study, which investigated the degradation rate of P(3HB-co-3HV) and P3HB with crystallinity of about 50–65% and 65–80%, respectively. The degradation rate of P(3HB-co-3HV) was 20–30% higher than that of P3HB, when degraded in soil (Kusaka et al., 1999). As the molecular weight of the polymer decreases, the monomers, dimers, and oligomers of the PHA become more available to the microbes (Gu, 2003).

In addition to the abovementioned characteristics of PHAs, the shape also plays a significant role in degradation. For example, thin films of PHA degrade faster than thicker films. A study conducted by Volova et al. (2010) to evaluate the degradation of PHA films and pellets in seawater revealed that the films degraded more compared to the pellets. The higher degradation rate is due to not only the thickness of the PHAs but also the availability of the surface of the PHAs to the microbes (Volova et al., 2010).

The environmental conditions in which PHAs are degraded play a significant role in the rate of degradation. Factors such as soil and climatic conditions influence the PHA degradation rate. A study conducted by Mergaert et al. (1993) stated that the degradation of PHAs are dependent on temperature. They also stated that the degradation of P(3HB) and P(3HB-co-3HV) is similar in both sterile buffer as well as in soil at 40°C. Further, Boyandin et al. (2012) evaluated the degradation ability of PHA film in different locations in Vietnam. They concluded that the hot and humid climate of Vietnam facilitated PHA degradation.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.