P(3HB-co-4HB) samples were produced at the Laboratory of Chemoautotrophic Biosynthesis at the Institute of Biophysics of the SB RAS (Siberian Branch of the Russian Academy of Sciences). The wild-type strain C. necator B-10646 capable of synthesizing copolymers of various monomer compositions was used as a producer using the author’s cultivation technology [49].

The results of C. necator B-10646 cultivation and P(3HB-co-4HB) synthesis using WFO, a fat-containing substrate of complex composition, compared with the control (sole substrate–butyric acid) are presented in Fig. 2 and Table 2. The more affordable ɛ-caprolactone was used as a precursor of 4-hydroxybutyrate [59,60]. The precursor was introduced into the growing bacterial culture at 12 h from the start of the process, fractionally by means of two additions (at 12 and 24 h of growth) in a concentration that did not significantly inhibit bacterial growth (one dose contained 0.5–2.0 g/L of ε-caprolactone). The duration of the process was 48 h, which is significantly shorter than in the case of P(3HB) homopolymer synthesis (60–70 h). This is due to the fact that the highest rate of PHA synthesis in a batch culture of PHA-producer strains, including C. necator B-10646, and the activity of the PHA-synthase enzyme is characteristic of the exponential growth phase (the first 15–20 h from the beginning of the cultivation). It was shown that the inclusion of 4HB monomers into the polymer chain is usually observed 12–24 h after the precursor was added to the culture [61]. Further, as the culture ages and the activity of intracellular PHA depolymerases increases, endogenous breakdown of PHA molecules occurs. Thus, the content of polymer and monomers other than 3HB decreases against the background of an increase in the content of 3HB monomers. Therefore, in the case of the synthesis of copolymer PHAs, it is necessary to sacrifice the total PHA and the cell biomass yield, reducing the duration of the bacterial growth process.

The bacterial culture parameters in the control and in the experiment without the use of the precursor were comparable. The precursor dosing mode affected the bacterial biomass yield, P(3HB-co-4HB) yields, and the 3HB/4HB monomer ratio. At the same time, with the same precursor dosing mode in the bacterial culture, the incorporation of 4HB into the polymer chain of 3-hydroxybutyrate occurred differently. At a ε-caprolactone concentration of 1.0 g/L fed to the culture, the cell biomass yield, intracellular content of the copolymer, and 4HB monomers (6.1–7.4 mol.%) did not differ statistically. In the case of a higher precursor concentration (2.0 g/L) in the control (butyric acid) and in the treatment (WFO), the copolymer and 4HB monomer contents were comparable (increasing to 9.6–9.7 mol.%). However, the inhibitory effect of the precursor was evident. That negatively affected the biomass yield in the treatment, in contrast to the control. A further increase in the precursor concentration (up to 3.0 g/L) enhanced the inhibitory effect in the treatment. The value of X (g/L) and the yield of the copolymer decreased. The content of 4HB in the control (19.0 mol.%) statistically significantly exceeded this parameter in the treatment (11.6 mol.%). A further increase in the precursor concentration (up to 4.0 g/L) dramatically inhibited the growth of cells in the treatment. By contrast, in the control, the 4HB content and the total yield of the copolymer, as well as the yield of bacterial biomass, did not change. The use of several more variants of the dosing regimens of ε-caprolactone in the culture of C. necator B-10646 did not allow us to obtain copolymer samples in which the 4HB content would be higher than 11.6 mol.%.

Previously, in the culture of C. necator B-10646, using glucose as the main carbon substrate and γ-butyrolactone as a precursor (the number of additions varied from one to 5-6), samples of P(3HB-co-4HB) with a significantly higher content of 4HB, from 10 to 56 mol.% and 75 mol.%, as well as 3- and 4-component copolymers, respectively, P(3HB-co-4HB-co-3HV) and P(3HB-co-4HB-co-3HV-co-3HHx), in which the 4HB content varied from 9 to 55 mol.% (valeric acid or valeric + hexanoic acids were additionally used as precursors) were obtained [49]. In this paper, the authors did not discuss or take into account the production parameters of the strain in terms of the total bacterial biomass yield, since they mainly considered the composition of the copolymers and their 4HB content. At the same time, it was shown that the bacterial biomass yield with a large amount of precursor additions and high 4HB content was extremely low, about 0.30–0.50 g/L (unpublished data of the authors of the present article).

The production parameters of C. necator B-10646 culture grown on WFO with the addition of ε-caprolactone in different concentrations are given in Table 2.

Thus, the production parameters of the C. necator B-10646 bacterial culture in terms of bacterial biomass yield (X (g/L)), total copolymer yield (P(3HB-co-4HB) (g/L), as well as the process productivity in terms of biomass and copolymer (P_X and P_P(3/4HB)) during growth on WFO decreased with an increase in the precursor concentration to 3.0 g/L (Table 2). At the same time, in general, the possibility of synthesizing P(3HB-co-4HB) copolymers with the inclusion of 4HB monomers from 7.4 to 11.6 mol.% with satisfactory parameters in terms of the total bacterial biomass yield (X (g/L)) was shown, using WFO derived from accessible and renewable fish processing waste.

The review of the literature showed that the ability of wild-type and recombinant Cupriavidus strains to synthesize P(3HB-co-4HB) with a 4HB monomer content from units to 90–96 mol.% when they were grown on sugars and γ-butyrolactone as a precursor has been studied and discussed in many publications, including reviews [14,15,17]. However, data on the synthesis of P(3HB-co-4HB) on fat-containing substrates, mainly on vegetable oils and low-grade animal fats and their waste, are very limited. At the same time, the synthesis of P(3HB-co-4HB) copolymers on fatty waste from fish processing is reported in only one work. The bacterium Salinivibrio sp. M318 (VTCC910086) grown on fish processing waste and glycerol with the addition of 1,4-butanediol, γ-butyrolactone, or sodium 4-hydroxybutyrate synthesized copolymers P(3HB-co-4HB) with yields of bacterial biomass and copolymer, respectively, up to 9.8–10.2 g/L and 49.2%–54.6%, but the content of 4HB did not exceed 5.2–5.9 mol.% [62].

In the present work, P(3HB-co-4HB) productivity in the experiment with C. necator B-10646 grown on WFO in flasks was close to the values found for wild-type Cupriavidus strains, but was somewhat inferior to recombinant strains [27,63]. This is significantly inferior to the productivity parameters for the synthesis of the P(3HB) homopolymer (3.14 g/L·h) [5]. Thus, in general, the synthesis of PHA copolymers is less productive.

In synthesis of copolymers, the affinity of PHA synthases to certain monomers is of great importance for the production of PHA copolymers and the content of monomers of different structures [64]. The results of the P(3HB-co-4HB) synthesis in the Cupriavidus sp. culture demonstrate that the activity of PHA synthase is the “bottleneck” in increasing the 4HB content of copolymers [65]. However, this aspect of PHA biosynthesis still represents an insufficiently studied niche today. Therefore, the study of PHA synthases in new wild-type and engineered PHA producers, as well as in new strategies for conducting biosynthesis processes on complex C-substrates, is important for planning and organizing efficient processes for the production of technological PHAs [14,15,17]. The detected decrease in the incorporation of the studied precursors into 4HB in the case of P(3HB-co-4HB) synthesis by C. necator B-10646 cells during growth on WFO (a complex fatty substrate) compared to a sole substrate (butyric acid) may also be associated with the mechanism of precursor transport into cells. However, this complex issue requires additional and special studies.

PHA properties (temperature characteristics, molecular weight, degree of crystallinity) vary significantly depending on the composition and ratio of monomers; this makes it possible to obtain products with different physical and mechanical characteristics. The chemical composition of polyhydroxyalkanoates and the resulting properties are determined by the physiological and biochemical characteristics of the PHA producer strains and the substrate specificity of PHA synthases responsible for the assembly of monomers into PHAs, as well as the conditions of carbon nutrition and the biochemical pathways by which the C-substrate is metabolized [66].

A few publications show that the polymer usually synthesized from fat-containing substrates, including WFO, is the poly(3-hydroxybutyrate) homopolymer, although isolated cases of copolymer formation have been reported. For example, Salinivibrio sp. M318 (VTCC910086) grown on WFO and glycerol synthesized only poly(3-hydroxybutyrate), but it synthesized poly(3-hydroxybutyrate-co-3-hydroxyvalerate) [P(3HB-co-3HV)] copolymers when propionate was added and P(3HB-co-4HB) copolymers when 1,4-butanediol, γ-butyrolactone, or sodium 4-hydroxybutyrate was added [62]. Some Pseudomonas strains synthesized polymers containing, in addition to 3HB, medium-chain monomers, with 2–3 mol.% of 3-hydroxyhexanoate, and 3-hydroxyoctanoate and 3-hydroxyhexanoate dominated when they were grown on the fatty waste of the Alaska pollock fish [41]. Synthesis of P(3HB-co-3HV) copolymers was detected in C. necator TISTR 1095 cells grown on the wastewater of a tuna cannery [44]. When WFO was used as a carbon substrate for the cultivation of the C. necator B-10646 strain studied in this work, it mainly synthesized P(3HB), but in some cases it synthesized a three-component polymer with a dominant fraction of 3HB monomers and a minor content of 3-hydroxyvalerate and 3-hydroxyhexanoate monomers [48].

The use of WFO or butyric acid (control) as the main growth substrate and ε-caprolactone as the precursor made it possible to synthesize P(3HB-co-4HB) copolymers with different 4HB contents; their chemical composition and properties are illustrated in Table 3 and Fig. 3–7.

The inclusion of 4HB in the structure of P(3HB-co-4HB) was confirmed by NMR. Fig. 3 shows, as an example, the ¹H NMR spectra of P(3HB-co-4HB) samples with similar contents of 4HB (9.6 and 9.7 mol.%), synthesized on WFO and on butyric acid. Both spectra are equivalent and do not depend on the type of substrate used.

For a comparative study of the physicochemical properties of P(3HB-co-4HB) copolymers synthesized on WFO or butyric acid, samples with similar 4HB contents were selected (Table 3).

Analysis of the molecular weight characteristics of the copolymers produced in this study (Table 3, Fig. 4) showed that in the treatment and in the control, as the 4HB content increased, there was a decrease in the molecular weight values. This is typical for the PHA copolymers in contrast to P(3HB). When WFO was used, a decrease in M_n from 332 to 160 kDa (M_w from 593 to 330 kDa) was noted with an insignificant increase in polydispersity (Ð). A similar pattern is typical for the control. At the same time, in general, the molecular weight characteristics of the P(3HB-co-4HB) samples synthesized on WFO are 1.5–2.0 times higher than those of P(3HB-co-4HB) synthesized on butyric acid with a similar content of 4HB.

PHA molecular weight is a highly variable parameter, significantly depending on the cultivation mode and duration, the polymer extraction procedure and other factors; therefore, the data on the molecular weight may differ significantly [67]. The published data on the molecular weight characteristics of PHA containing 4HB synthesized from fat-containing substrates as the main carbon source vary widely [61,68].

On the thermograms double wide melting peaks were recorded for all samples (Fig. 5) and melting points lower compared to P(3HB). In the experimental group, an increase in the 4HB content leads to a decrease in the melting point from 167°C to 156°C and a decrease in the enthalpy of melting (Table 3). A decrease in the enthalpy of melting indicates a decrease in the degree of crystallinity and amorphization of the samples with an increase in the 4HB monomer content. In the control group, the sample with a 4HB content of 9.7 mol.% had the highest melting point, 167°C. A decrease in the glass transition temperature is also noted in both groups with an increase in the 4HB content. The lowest glass transition temperature of –4.1°C was recorded for the sample with a 4HB content of 11.6 mol.% in the treatment group and for the samples containing 9.7 and 19 mol.% (control group). All samples demonstrated good thermal stability; thermal degradation of the samples began when they were heated above 250°C, and the maximum rate of weight loss was reached in the temperature range of 274°C–302°C. In both groups, the best thermal stability was noted for samples with 4HB monomer contents of 9.6 and 9.7 mol.%; the thermal degradation temperature was 285°C and 275°C with a maximum weight loss at 302°C and 298°C, respectively.

Jo et al. points to a decrease in the degree of crystallinity with an increase in the 4HB content, as well as the appearance of cold crystallization peaks, which the authors associate with an increase in chain randomness [69]. Double melting peaks and lower melting point (158°C) than for P(3HB) (170°C) and crystallinity degree of 40% were observed for the copolymer with 4HB content of 19 mol.%, which was synthesized by Comamonas acidovorans on mixed carbon sources (n-butyric acid and 1,4-butanediol) [70].

Crystallization kinetics plays a central role in determining the mechanical properties of polymer products. This indicates the importance of studying this process in PHA. Fig. 6A,B illustrates the crystallization and melting of the obtained P(3HB-co-4HB) crystals with different 4HB contents, under isothermal conditions at different temperatures (60°C, 70°C, 80°C, 90°C, 100°C). For all samples, the crystallization enthalpy rises with increasing crystallization temperature, which may be due to the emergence of nucleation difficulties at higher temperatures. In the melting thermograms of crystals obtained by crystallization at 60°C, the melting peaks have a small shoulder on the side of low temperatures. With increasing crystallization temperature, this shoulder develops into a separate peak, the area of which increases. Thus, all thermograms of crystallization at high temperatures have two peaks: one in the region of low temperatures (138°C–157°C) and the other in the region of higher temperatures (154°C–169°C). The position of the first peak is not constant, and with an increase in the crystallization temperature it shifts towards higher temperatures; the location of the second melting peak on the thermogram does not differ significantly. In many studies, double melting peaks for copolymers of this type are associated with the formation of two phases: a phase enriched in molecules containing 4HB and forming defective crystals and a phase enriched in 3HB, in which more perfect crystals with a higher melting point are formed [71].

This behavior during isothermal crystallization is typical for all PHA copolymers. Increasing the content of monomers other than 3HB disrupts the regularity of the P(3HB) chain, which hinders crystallization, but can also promote crystallization due to the higher flexibility of some monomers compared to 3HB at the corresponding isothermal temperatures. The presence of other monomers in the P(3HB) chain reduces the rate of crystallization; however, an increase in the flexibility of the copolymer chain above a certain content of other monomers can lead to an increase in the rate of crystallization and the formation of defective crystals [67,71–73].

Spherulites are the most common crystalline form of PHAs, formed upon rapid cooling of melts or upon precipitation from a concentrated polymer solution [74]. Typically, PHAs are characterized by the formation of extremely large spherulites under certain crystallization conditions [75]. In semicrystalline PHAs, crystallization occurs near the equilibrium state, at which crystalline lamellae develop into spherulites [76]. It was found that PHAs can form non-striped and striped spherulites by isothermal crystallization from a thin melt at both low and higher crystallization temperatures [77]. The present study shows that the change in the ratio of crystallization and diffusion rates is the reason for the change in spherulite morphology. The shape and size of spherulites in PHA are influenced by many factors, with the key factor being the temperature at which crystallization occurs. The periodicity and regularity of the spherulite band structure can change depending on the crystallization conditions, as well as the molecular weight of the polymer [78].

The morphology of spherulites of P(3HB-co-4HB) samples with different 4HB contents synthesized by C. necator B-10646 cells grown on WFO and butyric acid (control) is shown in Fig. 7.

A typical pattern–“Maltese cross” and concentric bands—was recorded for all the studied P(3HB-co-4HB) samples during crystallization, regardless of the type of the main carbon substrate used and the monomer ratio in the entire studied crystallization temperature range of the copolymer melts (70°C, 80°C and 90°C). The dependences of the change in the spherulite sizes of the studied P(3HB-co-4HB) samples on the crystallization temperature at all values of the content of the molar fraction of 4HB, as well as in the treatment on WFO and in the control (butyric acid), were comparable. Larger spherulites were formed at higher crystallization temperatures and a higher 4HB content of the P(3HB-co-4HB) samples. Thus, at a temperature of 70°C, the size of spherulites varied from 200–300 µm to 1–2 mm, depending on the 4HB content. At a temperature of 90°C, the range of changes in the size of spherulites was significantly wider, ranging between 1.5 and 3.0 mm or more. This is presumably associated with the observed lower nucleation density at a higher crystallization temperature and is consistent with the results obtained by Hobbs et al. [78].

The size of the spherulites increased linearly throughout the process and regardless of the crystallization temperature until the moment of collision. This indicates a constant spherulite growth rate for P(3HB-co-4HB) throughout the crystallization process. The nature of the temperature dependence of the growth rate of spherulites of the P(3HB-co-4HB) samples was practically independent of the 4HB content in the copolymer. However, minor differences were revealed in the maximum value of the spherulite formation rate at similar 4HB contents for samples synthesized on WFO and on butyric acid (control). The maximum values of the rate were recorded at a temperature close to 85°C. For example, for samples with similar 4HB monomer contents (9.6 and 9.7 mol.%), the values were 1.32 and 0.95 μm/min for the samples produced on butyric acid and WFO, respectively. Thus, the type of the main carbon substrate did not have a significant effect on this parameter.

Regardless of the type of the main substrate, the maximum value of the spherulite growth rate was proportional to the 4HB content. A similar nature of spherulite formation was found for both the highly crystalline P(3HB) homopolymer and the PHA copolymers containing 3-hydroxyvalerate (P(3HB-co-3HV)) [73] and 3-hydroxyhexanoate-P(3HB-co-3HHx) [72]. In that case, the maximum spherulite growth rate for P(3HB-co-4HB) significantly differed from this parameter (more than 3 μm/min) for P(3HB). A similar dependence was also described by You [79].

This fact allows us to state that for the P(3HB-co-4HB) copolymers, the effect of 4HB on the crystallization behavior is complex and is determined by the competition between two factors. On the one hand, the crystalline structure of the P(3HB) fraction is disrupted by the inclusion of 4HB in its chain. This leads to a decrease in the crystallization ability of P(3HB). On the other hand, the introduction of 4HB into the 3HB monomer chain can improve the mobility of macromolecular chains. This was also shown by Wen et al. [80]. However, this effect, favorable for crystallization, apparently does not depend on the type of C-substrate used.

In general, the obtained results are consistent with the few studies performed by other authors on the crystallization and formation of spherulites during isothermal crystallization of melts of P(3HB-co-4HB) samples synthesized by other strains on glucose as the main substrate [80].