Introduction

Plastic pollution is an environmental threat that continues to expand in scope. In 2015, more than 320 million tons of plastic products were produced worldwide, up from 180 million tons in 2000 (Prof_Research_Reports 2014). With an average compound annual growth rate (CAGR) of 8.6% since 1950 (Geyer et al. 2017), combined with the strong correlation between global plastic waste generation and gross national income per capita (Geyer et al. 2017), plastic production and waste generation are likely to continue to rise as developing nations worldwide continue to grow economically (Geyer et al. 2017; Markets 2018; Prof_Research_Reports 2014). Of concern, if historic trends continue, a significant fraction of plastic can be expected to reach the environment; of the estimated 6.3 billion tons of plastic waste generated since 1950, 12% was incinerated and 9% was recycled, leaving approximately 4.9 billion tons of plastic accumulating in landfills and marine and soil ecosystems (Geyer et al. 2017).

One facet of a solution to plastic pollution is to reduce the amount of postconsumer plastic entering the environment. An important element in accomplishing this goal is to give postconsumer plastic waste economic value, i.e., to valorize it, so it is not readily discarded. Standard plastic recycling techniques generally yield materials that are of lower quality than the initial product, resulting in “downcycling,” or a loss of value (La Mantia 2004). In contrast, a biological approach to recycling plastic waste can potentially be economically sustainable: plastic waste can be used as feedstocks for bioconversion processes to make biochemicals and other value-added biological products, resulting in “upcycling” (Guzik et al. 2014; Kenny et al. 2008; Narancic and O'Connor 2017; Ward et al. 2006; Wierckx et al. 2015; Xu et al. 2016). The rationale for this concept has two parts: first, several kinds of postconsumer plastics can be depolymerized into mixtures of labile molecules and second, microbes can synthesize triacylglycerols, organic acids, enzymes, and other biological products as they grow on plastic-derived media. This approach can be used to make products that justify the costs associated with recycling plastic waste (Beopoulos et al. 2008; Dahlbo et al. 2018; Fickers et al. 2005; Guzik et al. 2014; Kenny et al. 2008; Ward et al. 2006).

Polypropylene (PP) comprises nearly 25% of the global plastics market share (Geyer et al. 2017), yet to date, there are no reported microbial bioconversion processes for PP. A challenge for microbial metabolism of PP is its low bioavailability, resulting from its high molecular weight, hydrophobicity, and a molecular structure that resists enzymatic attack (Arutchelvi et al. 2008; Jeyakumar et al. 2013; Longo et al. 2011). We hypothesized that a depolymerization pretreatment to generate PP-derived oligomers could aid in the production of a growth medium suitable for microbial bioconversion. We used pyrolysis to depolymerize and oxidize PP and then used biodegradable surfactants to disperse the resulting products in an aqueous medium suitable for microbial growth. The PP-derived growth medium was used for cultivating the oleaginous yeast Yarrowia lipolytica. Y. lipolytica grows on diverse substrates and produces a wide variety of intracellular and extracellular products, including fatty acids, organic acids, extracellular enzymes, and other proteins (Abghari and Chen 2014; Ageitos et al. 2011; Aggelis 2002; Beopoulos et al. 2008; Bialy et al. 2011; Fickers et al. 2005; Rakicka et al. 2015; Xu et al. 2016; Xue et al. 2013; Zhang et al. 2014). In this work, Y. lipolytica was cultivated in a PP-derived medium in order to produce FAs. The following report summarizes and discusses our findings.

Methods

Polypropylene growth medium preparation

OP4 medium, a polypropylene-derived medium, was prepared by pyrolyzing virgin amorphous polypropylene (MW = 14,000) in 3-g batches at 540 °C in 125-mL borosilicate flat-bottomed flasks for 190 min. The resulting pyrolysis oil (at 15 g L−1) was combined with additional compounds as follows: 5.4 g L−1 Tween 80®, 4.5 g L−1 oleic acid, 1.25 g L−1 (NH4)2SO4, 2.5 g L−1 KH2PO4, and 0.830 g L−1 MgSO4·7H2O. The mixture was emulsified with a handheld food-grade homogenizer and autoclaved for 70 min at 121 °C and 15 psi.

Chemicals and reagents

Virgin polypropylene pellets and oleic acid were purchased from Sigma Aldrich (Millipore Sigma, USA). Tween 80®, chloroform, methanol, cyclohexane, hexane, and all culturing compounds used were of research grade and purchased from Fisher Chemicals (Fisher Scientific, USA).

Cultures and growth conditions

Yarrowia lipolytica strain 78-003 (ATCC strain 46483) was the sole strain used in this work. Glycerol frozen stocks of Y. lipolytica were prepared (1 mL, OD600 = 20) and were stored at − 80 °C. For all experiments, a frozen aliquot was thawed and added into a 250-mL Erlenmeyer flask containing 50 mL of 5% glucose medium (consisting of 50 g L−1 glucose and 3 g L−1 yeast extract). Each inoculated flask was incubated overnight at 30 °C with shaking at 200 rpm. After incubation, 1-mL samples were withdrawn and centrifuged at 10,000 rpm for 2 min, and the pellets were washed twice with 50 mM phosphate-buffered saline (PBS) solution. Y. lipolytica was cultured using 500-mL Erlenmeyer flasks containing 50 mL OP4 medium; the flasks were inoculated with an overnight culture of Y. lipolytica at an inoculation density of 0.3 (OD600 mL−1) and incubated at 30 °C with shaking at 200 rpm.

Several versions of OP4 were prepared to evaluate the impact of medium components on lipid and biomass yields. “Nitrogen and trace minerals only” medium consisted of 1.25 g L−1 (NH4)2SO4, 1.25 g L−1 yeast extract, 2.5 g L−1 KH2PO4, and 0.830 g L−1 MgSO4·7H2O. “Surfactant only” medium was comprised of 5.4 g L−1 Tween 80® and 4.5 g L−1 oleic acid, without trace minerals or nitrogen. “PP only” medium, which was OP4 medium without any surfactants, was comprised of 12 g L−1 pyrolyzed polypropylene, 1.25 g L−1 (NH4)2SO4, 1.25 g L−1 yeast extract, 2.5 g L−1 KH2PO4, and 0.830 g L−1 MgSO4·7H2O. “Surfactant control” medium was comprised of 5.4 g L−1 Tween 80®, 4.5 g L−1 oleic acid, 0.25 g L−1 (NH4)2SO4, 1.25 g L−1 yeast extract, 2.5 g L−1 KH2PO4, and 0.830 g L−1 MgSO4·7H2O.

The hexadecane medium used for FA profile experiments consisted of 5% v/v hexadecane, 5 g L−1 yeast nitrogen broth (w/amino acids), 0.5 g L−1 KH2PO4, and 0.25 g L−1 MG2SO4. The starch medium used for FA profile experiments consisted of 50 g L−1 hydrolyzed starch, 2 g L−1 yeast nitrogen broth (w/amino acids), 0.5 g L−1 KH2PO4, and 0.25 g L−1 MG2SO4. OP5 medium consisted of 5.4 g L−1 Tween 80®, 1.25 g L−1 (NH4)2SO4, 2.5 g L−1 KH2PO4, and 0.830 g L−1 MgSO4. Glucose (5%) medium used for FA profile experiments consisted of 50 g L−1 glucose and 3 g L−1 yeast extract.

Microscopy for imaging oil droplet production

A modified Nile Red staining method (Rostron and Lawrence 2017) was utilized to stain Y. lipolytica intracellular neutral lipids. Sample aliquots (1 mL) of post-fermentation culture were withdrawn and centrifuged, and pellets were washed twice with 0.9% NaCl. Pellets were suspended in 1.0 mL of 10 mM PBS with 0.15 M KCl and the OD600 was adjusted to 10. Technical grade Nile Red powder (Millipore Sigma, USA) was dissolved in acetone to make a 10 mg mL−1 solution and 10 μL was added to 1.0-mL cell samples (suspended in 50 mM PBS and calibrated to an OD of ~ 5.0) in dark conditions, and samples were vortexed and incubated at room temperature (in dark conditions) for approximately 15 min. A Zeiss LSM 510 confocal microscope (Zeiss, USA) was used to visualize stained cells. Ten microliters of sample was used for confocal analysis. Cells were excited at 488 nm, and emissions were imaged at 543 nm.

Intracellular lipid quantification

A modified Bligh and Dyer extraction (Bligh and Dyer 1959) was used to extract lipids from yeast samples. Post-fermentation OP4 culture was withdrawn and aliquoted into pre-weighed 50-mL conical tubes, and tubes were centrifuged, and pellets washed twice with 0.9% NaCl. Pellets were lyophilized overnight using a BT3.3 EL Lyophilizer Tabletop (SP Scientific, USA), weighed, suspended in a 2:1 v/v chloroform:methanol mixture (5 mL per 50 mg cell dry weight) and sonicated using a Sonic Dismembrator (Fisher Scientific, USA) at 20 kHz, 20% amplitude with pulsing (40 s on and 20 s off for a total working time of 20 min). The chloroform layer was taken and dried under a nitrogen stream. Dried lipids were weighed and derivatized for GC/MS analysis via base-catalyzed esterification with 2.5 mL sodium methoxide (0.1 M) (Milanesio et al. 2013). The reaction was quenched using 200 μL sulfuric acid (> 95%), and 2.5 mL hexane was added to each sample, which was then vortexed and centrifuged (5 min at 5000 rpm). 1.5 mL from the top hexane layer was withdrawn for GC/MS analysis.

GC/MS analysis of lipid profile

The FA profile was characterized with an Agilent 7890A gas chromatograph (GC) attached to a 5977A mass spectrometer detector and equipped with an Agilent J&W HP-5ms UI capillary column (30 mm × 0.25 mm × 0.25 μm). One-microliter samples were injected via an Agilent 7693 Automatic Liquid Sampler in splitless mode with an inlet temperature of 275 °C, using helium as the carrier gas at a flow rate of 1 mL min−1. GC oven temperature was held at 60 °C for 1 min and ramped to 100 °C (rate 25 °C min−1, hold 1 min). The oven temperature was then increased to 200 °C (rate 25 °C min−1, hold 1 min). The oven temperature is then increased to 220 °C (rate 5 °C min−1, hold 7 min) and then increased to an ending temperature of 300 °C (rate 25 °C min−1, hold 2 min). A C8-C24 FAME analytical standard was used during sample analyses as an external standard (Sigma Aldrich, USA).

Pyrolysis oil characterization via GC/MS

Three grams of pyrolysis oil was dissolved in 200 mL chloroform (Sigma Aldrich, USA) and 1.5-μL aliquots were analyzed via GC/MS. A GC method (Guzik et al. 2014) was used to characterize the polypropylene pyrolysis oil. One microliter of the diluted pyrolysis oil was injected via automatic liquid sampler at an inlet temperature of 275 °C and a 2:1 split ratio. The oven method was 30 °C for 1 min, then ramping to 100 °C (rate 7.5 °C min−1), then ramping to 300 °C (rate 10 °C min−1, hold 2 min).

Substrate degradation analysis

Substrate degradation analysis was done both gravimetrically and via GC/MS. Ten-milliliter aliquots of OP4 medium pre- and post-fermentation were taken for analysis. Media was centrifuged (10 min at 7000 rpm) and 5 mL of supernatant transferred to pre-weighed conical tubes and wet weight of media documented. Samples were lyophilized and weighed, and the dry cell weight was suspended in a 1:1:1 v/v mixture of hexane, chloroform, and deionized water (15 mL), and vortexed until the sample was fully dissolved. Afterward, 1.5 mL of the top hexane layer was withdrawn and used for GC/MS analyses.

Growth measurements

One milliliter of Y. lipolytica overnight culture was pelleted (10,000 rpm, 10 min) and pellets were washed twice with 50 mM PBS, and then resuspended in 1 mL of 50 mM PBS to reach a final OD600 of 15. The inoculum was used to inoculate all experiments involving growing cells at the initial optical densities indicated in the text. At set timepoints, 1 mL aliquots were withdrawn and OD600 measurements were taken using an Eppendorf 6131 Biophotometer (Eppendorf, USA).

For gravimetric analysis of growth, 50-mL aliquots were withdrawn, centrifuged at 10,000 rpm for 10 min, the supernatant was discarded, and cell pellets were washed twice with 50 mM PBS and lyophilized overnight before being weighed.

Cyclohexane toxicity assay

Yeast malt broth (10 g L−1 dextrose, 5 g L−1 malt extract, 3 g L−1 peptone, and 5 g L−1 yeast extract) was used to cultivate Y. lipolytica cells. A one-milliliter aliquot of overnight culture (inoculation density = 0.30 (600 nm)) was used to inoculate 50 mL of either yeast malt broth only or yeast malt broth with 0.23% w/v cyclohexane. One mL samples were withdrawn at select time points, and growth was measured spectrophotometrically (600 nm).

Results

Polypropylene thermal depolymerization

Pyrolysis of virgin amorphous polypropylene (PP) pellets generated an oil-like fluid that turned waxy when cooled rapidly. GC/MS analysis of the PP oil identified approximately 18 different compounds across the batches that were prepared. More than 80% of PP oil was branched chain compounds, with branched fatty alcohols (50.9%) and branched alkenes (25.1%) making up approximately 75% of all available carbon sources (Fig. 1). The branched alkenes detected were all CnH2n compounds, with the most abundant compound being 2,4-dimethylhept-1-ene (~ 14%), followed by 2,4,6,8-tetramethyl-1-undecene (~ 6%), and 1,4-dimethyl-decene (~ 5%). The branched fatty alcohols detected were CnH2n+2 compounds, with 2-hexyl-1-decanol (~ 41%) being the predominant compound, followed by 2-methyl-1-decanol (~ 10%).

Fig. 1
figure 1

Carbon composition of PP oil after thermal depolymerization

Full size image

Analyzing Y. lipolytica growth and activity in OP4 medium: comparisons with a “surfactant control” medium

48.9% of the total carbon in OP4 medium, on a moles C basis, was derived from the biodegradable surfactant. The remainder of the carbon in the medium was found in the PP oil, with trace amounts derived from the supplemented yeast extract. To determine the extent to which the PP oil in OP4 contributed to Y. lipolytica growth and to the formation of biochemical products, a series of experiments were conducted comparing the activity of Y. lipolytica grown in OP4 medium versus a “surfactant control” medium. The “surfactant control” was identical in composition to OP4 medium except that it contained no PP oil.

To determine the extent that PP oil contributed to Y. lipolytica growth, cell growth was measured gravimetrically. The analysis indicated an average of 27% less biomass when cells were grown on OP4 medium (Fig. 2a). Biomass accumulation peaked after 72 h. The biomass remained at 2.4 g L−1 or greater through 192 h.

Fig. 2
figure 2

Growth of Y. lipolytica in OP4 medium. a Growth in OP4 medium versus a “surfactant control” medium. Growth was measured as cell dry weight. The “surfactant control” included all emulsifiers, nitrogen, phosphorous, and trace nutrient sources found in OP4 medium but did not include PP-derived compounds. b OP4 medium use during Y. lipolytica growth. The change in OP4 concentration was measured gravimetrically. Fatty alcohol consumption was measured by GC/MS. Mean ± SD plotted for each data set

Full size image

OP4 medium uptake during growth

To determine the extent of OP4 medium uptake during growth, the mass of medium remaining in solution was measured. More than 80% of OP4 medium was taken up after 13 days (312 h) (Fig. 2b). Approximately 51% of the PP oil in OP4 medium is comprised of fatty alcohols. To determine whether medium components originating from PP were taken up by Y. lipolytica, the change in fatty alcohol concentration in the medium was analyzed via GC/MS (Fig. 2b). 51% of the fatty alcohol fraction of OP4 medium was taken up by Y. lipolytica by 120 h, with 86% of the fatty alcohol fraction assimilated by 312 h. The uptake of total FAs and fatty alcohols occurred in a similar fashion over the course of experiments.

Impact of OP4 components on growth and FA production

OP4 medium contains several components that contribute to the overall growth of Y. lipolytica. To determine the effect of individual components of the medium on cell growth, a series of comparisons were made (Fig. 3). Yeast extract was used as a supplemental nitrogen source and contributed approximately 25% of the maximum measured biomass. A comparison of the “surfactant only” and “surfactant control” treatments confirmed that supplemental nitrogen and salts were necessary to increase the yield. Comparing the “PP only” and “OP4” treatments indicated that the presence of surfactant in the medium increased the yield more than six-fold. On the other hand, comparisons of “OP4” and “surfactant only” treatments or “PP only” and “nitrogen and trace minerals only” treatments determined that PP oil in the growth medium inhibited Y. lipolytica growth. Although PP oil in OP4 medium inhibited growth, it significantly increased the FA yield (Table 1).

Fig. 3
figure 3

Growth on OP4 medium components. Growth was measured gravimetrically. Statistical significance was determined by ANOVA. Samples were considered significantly different if p < 0.05. Treatments that share the same lower case or upper case letter are not significantly different than one another

Full size image
Table 1 Intracellular lipid yields by Y. lipolytica during growth in OP4 medium or its componentsab
Full size table

Effect of cycloalkanes on Y. lipolytica growth

Cyclic alkanes have been reported elsewhere to be poor growth substrates for Y. lipolytica and other industrial yeasts (Beam and Perry 1974; Das and Chandran 2011; Mauersberger et al. 1996). OP4 contains 0.15 ± 0.08% (w/v) cyclic alkanes. To determine whether cyclic alkanes inhibited growth of Y. lipolytica at the concentration that they are found in OP4 medium, we compared growth in a rich medium with or without 0.23% (w/v) cyclohexane added (Fig. 4). Cyclohexane was selected as a representative cyclic alkane. The presence of cyclohexane reduced cell growth relative to control cells by 72 h, a difference which persisted through 120 h. The extent of growth inhibition ranged from 6 to 32% over the course of the experiment.

Fig. 4
figure 4

Growth of Y. lipolytica on rich medium with and without 0.3% w/v cyclohexane. Growth was measured as optical density (600 nm). Statistical significance determined by Student’s t test. Asterisk (**) indicates a significant difference (p < 0.01) between treated samples and the untreated control

Full size image

Microscopy analysis of lipid accumulation

Lipid accumulation by Y. lipolytica cells grown in OP4 medium or “surfactant control” medium was analyzed by Nile Red staining and confocal microscopy followed by quantitative image analysis. Differences in the cellular lipid content were visibly noticeable in compositely stained cells (Fig. 5a). Cells grown in OP4 medium tended to aggregate compared with those grown in a surfactant-based medium. The mean fluorescence for cells grown in OP4 medium was nearly twice that of cells growing in surfactant-based medium (p < 0.05), indicating greater lipid accumulation when cells were grown in OP4 medium (Fig. 5b).

Fig. 5
figure 5

FA production during growth in OP4 medium. Cells were imaged after 96 h. Magnification × 400. a Left, OP4 medium; right and surfactant control. Note lipid inclusion bodies are in yellow. b Mean Nile Red fluorescence quantified by digital image analysis

Full size image

Fatty acid yields during growth in OP4 medium versus surfactant control

To determine the extent that PP contributed to the production of fatty acids by Y. lipolytica, a comparison of FA production between cells grown in OP4 medium versus surfactant-based medium was conducted. Lipid yields were determined via GC/MS analysis. FA yields were significantly higher when Y. lipolytica grew on OP4 medium compared with the “surfactant control” medium (Fig. 6a). This was true at each of the measured time points over the course of experiments. Y. lipolytica generated most of its lipid bulk between 72 and 120 h. The bulk of the FAs produced were C 18, followed by C 16, with small amounts of C 20 and trace amounts of C 14 FAs at 240 h (Fig. 6b). The bulk of the FAs produced were unsaturated; there was no significant difference in the proportion of unsaturated to saturated FAs between the “surfactant control” and the OP4 medium (Fig. 6c). However, there was a significant increase in the percentage of saturated FAs produced as the experiment progressed: between 72 and 240 h the fraction of saturated FAs increased from 10 ± 1 to 35 ± 8% (p < 0.05). The majority of FAs produced was monounsaturated C18:1 FAs (Fig. 6d).

Fig. 6
figure 6

Analysis of fatty acid production by Y. lipolytica during growth on OP4 medium vs surfactant control. a total FA yield. b Distribution of FAs by carbon number. c Distribution of saturated and unsaturated FAs. d Distribution of saturated and unsaturated FAs

Full size image

Effect of lipids or carbohydrates on FA production

Oleaginous yeast are often grown on carbohydrates and produce FAs by de novo biosynthesis; in contrast, OP4 medium contains hydrocarbons derived from PP depolymerization and from biodegradable surfactants, likely inducing ex novo FA biosynthesis. To help understand how the composition of OP4 medium affected FA production, Y. lipolytica was grown in media containing lipids or carbohydrates as the carbon source (Fig. 7). The substrates that were examined were as follows: glucose (5%), starch (2%), hexadecane (5%), and OP4 medium (1.5% PP-derived compounds). Hexadecane was selected as a representative hydrophobic substrate found in OP4 medium. The C 18 FA oleic acid was the dominant FA produced by Y. lipolytica during growth in each medium except hexadecane. C 16 and C 18 FA distribution was consistent among cells grown on both carbohydrate-based substrates and OP4 medium, with deviation occurring on hexadecane-grown cells. Palmitic or palmitoleic acid, C 16 FAs, was the second most abundant, followed by C 18 stearic acid. In a hexadecane-containing medium, saturated palmitic acid was the main FA produced. Only 5 FAs were produced when Y. lipolytica was grown in OP4 or hexadecane-containing medium. In contrast, the FA profiles of Y. lipolytica grown on glucose and starch were more complex, with 10 different types of FAs produced, including the shorter chain FAs myristic, dodecanoic, pentadecanoic, and tridecanoic acids.

Fig. 7
figure 7

Effect of lipid- versus carbohydrate-containing media on fatty acid production. Cells were grown for 120 h prior to analysis

Full size image

To determine whether the PP content of OP4 influenced the type of FA products produced, the FA profiles from growth in OP4 medium versus in the “surfactant control” medium were compared (Table 2). Several features were evident. First, with the exception of 192 h, palmitoleic acid was a greater fraction of the FAs produced in OP4 medium than the “surfactant control.” Conversely, at each timepoint except 120 h, the palmitic acid fraction was larger in the “surfactant control.” Second, there was a spike in the fraction of oleic acid in the OP4 medium with respect to the “surfactant control” at 192 h. Lastly, very long-chain fatty acids (≥ C 20) were only found in cells grown in OP4 medium.

Table 2 Fatty acid profiles of Y. lipolytica grown in OP4 medium or surfactant control mediuma
Full size table

Discussion

If plastic waste can be used in bioconversion processes, then an incentive will exist to keep it out of the environment. In this work, we demonstrated that compounds generated from PP pyrolysis can be transformed by Y. lipolytica into fatty acids suitable for use in industry or other diverse applications. We are not aware of work by others reporting PP as a growth substrate for microbial bioconversion. The yield of FAs produced by Y. lipolytica during growth in OP4 medium was comparable to related bioconversion processes. There are limited examples of bioconversion processes where plastic was used as the feedstock, but in the most related process to the work presented here, Guzik used polyethylene (PE) pyrolysis oil for polyhydroxyalkanoate (PHA; “bioplastic”) production by Pseudomonas sp. (Guzik et al. 2014). After 48 h, 84.4% of the PE oil was metabolized, producing 0.23 g L−1 CDW, of which 9.8% was PHA, a biomass to substrate yield of 0.14 g gC−1 and a PHA to substrate yield of 0.01 g gC−1 (Guzik et al. 2014). In comparison, the growth of Y. lipolytica 78-003 in OP4 medium yielded 2.34 g L−1 CDW, a biomass to substrate yield of 0.13 g gC−1, and a FA to substrate yield of 0.03 g gC−1 (0.54 g L−1 FAs). Similarly, when Y. lipolytica was grown on 5 g L−1 food oil waste, a FA-rich feedstock similar in hydrophobicity and carbon content to OP4 medium, the yield was 3 g L−1 CDW and 0.75 g L−1 FAs after 6 days (Bialy et al. 2011). In general, lower FA accumulation is observed during growth of oleaginous yeast on hydrophobic substrates and additional measures must be taken to increase the yield of FAs (Papanikolaou and Aggelis 2011).

Did cultivating Y. lipolytica on a PP-derived growth substrate impact FA production? By comparing FA production in OP4 medium versus a “surfactant control,” a few differences in the FA profile were evident, notably an increase the palmitoleic acid fraction, the presence of C 20 compounds, and a larger product yield. It remains to be seen in future experiments whether specific PP-derived compounds directly influenced the observed FA product profile. On the other hand, a comparison of the C 16 and C 18 FA profile measured in this work during growth in OP4 medium with FA profiles reported by others revealed a remarkable similarity (Table 3). These data suggest that the Y. lipolytica FA profile is influenced to a significant extent by cellular metabolism rather than substrate characteristics. Additionally, by comparing FA profiles for cells grown on hydrophobic substrates (OP4 medium, hexadecane) and cells grown on hydrophilic substrates (glucose, starch), some differences were evident, notably the greater diversity of FAs produced during growth on the carbohydrate media. In general, the observed FA profiles were consistent with de novo lipid synthesis during growth on glucose and starch, where FAs are formed as secondary metabolites after nitrogen depletion warrants carbon storage and ex novo lipid synthesis during growth on OP4 medium and hexadecane, where FA synthesis is a growth-coupled process that assimilates hydrophobic substrates into lipids while simultaneously utilizing them for growth and maintenance (Papanikolaou and Aggelis 2011). Overall, the data support the view that FA production by Y. lipolytica in a PP-derived medium is similar to FA production during growth on naturally occurring substrates.

Table 3 C 16 and C 18 FA profiles of Y. lipolytica growing on various media in this and other work. Values are reported as percentages. After (Papanikolaou and Aggelis 2011)
Full size table

Two additional points regarding the presented data warrant further comment. First, Y. lipolytica did not grow as extensively in OP4 medium as it did in the “surfactant control.” The most likely explanation is that there were growth-inhibitory compounds associated with PP that were not present in the control, mainly cyclic alkanes. Cyclic alkanes are not an adequate growth substrate for Y. lipolytica, as the P450 monooxygenases mainly responsible for hydrophobic substrate oxidation during assimilation are not able to oxidize cyclic alkanes (Beam and Perry 1974; Das and Chandran 2011; Mauersberger et al. 1996). Our work shows that concentrations of cyclic alkanes as minimal as 0.23% w/v impeded growth. Notably, in spite of the lower biomass yield during growth on OP4 medium, the FA yield was significantly higher. Cellular stress, including nutrient limitations (Aggelis 2002; André et al. 2009; Beopoulos et al. 2008; Kitcha and Cheirsilp 2011; Klug and Daum 2014; Kuttiraja et al. 2016), increases FA storage by Y. lipolytica; chemical toxicity may have caused a similar stress response. Second, the PP component of OP4 medium contributed to Y. lipolytica growth and FA production. This result was evident in three ways. (1) Gravimetric analysis of changes in the growth medium mass during experiments determined that 81% of the medium was taken up by Y. lipolytica. Even if all the non-PP components of OP4 medium were consumed first, these were less than half of the medium, meaning that at least 30% of the substrate taken up by Y. lipolytica was PP-derived. (2) The concentration of branched fatty alcohols in OP4 medium declined by 86% over the course of experiments. These compounds were the most abundant constituent of PP oil and were not part of the biodegradable surfactant in OP4 medium; the reduction in concentration indicates that Y. lipolytica assimilated at a minimum one major component of PP oil. (3) Y. lipolytica produced considerably more FAs when grown on OP4 medium than when grown on the “surfactant control,” indicating that PP-derived compounds contributed to FA production.

The production of FAs by Y. lipolytica can be optimized by altering fermentation conditions and by metabolic engineering. For example, supplementing a food oil waste-derived growth medium with 10 g L−1 glucose increased the biomass yield from 3 to 13 g L−1 with a concomitant increase in the FA yield from 0.75 to 7.3 g L−1 (Patel and Matsakas 2018). Additionally, the bioavailability of the growth substrate can be increased, for example, bioconversion of waste cooking oil was enhanced by ultrasonication of the growth medium, leading to an increase in FA production (Patel and Matsakas 2018). When the FA degradation and remobilization genes, Pox1-6 and TGL4, were inhibited, Y. lipolytica grown on 250 g L−1 glycerol was able to obtain a lipid yield of 15.5 g L−1, with lipid content constituting 31% CDW (Rakicka et al. 2015). Qiao et al. (2015) determined that simultaneous overexpression of Y. lipolytica stearyl CoA desaturase, acetyl-CoA carboxylase, and diacylglyceride acyl-transferase genes yielded a strain with fast cell growth and a high lipid titer (55 g L−1) (Qiao et al. 2015). Based on these results, it is likely that a strategy can be found to increase the yield of FAs for cells grown in OP4 medium. More specifically, we hypothesize that adding carbon sources such as glucose or glycerol will favor biomass accumulation during the growth of the inoculum, and the larger microbial population should lead to a quicker uptake of PP-derived compounds and greater FA accumulation. Alternatively, we hypothesize that overexpressing enzymes responsible for ex novo FA biosynthesis including fatty alcohol dehydrogenase, fatty alcohol oxidase, and fatty aldehyde dehydrogenase will favor FA accumulation over catabolism by peripheral pathways, leading to increased yields. Overall, we think that bioconversion can be part of a terminal recycling solution for plastic waste.