Polyethylene terephthalate.
Polyethylene terephthalate (PET) is mainly used for production of PET bottles, PET foil, and fibers in the textile industry. PET is a polar, linear polymer of repeating units of the aromatic terephthalic acid and ethylene glycol. The PET monomer is designated bis(2-hydroxyethyl) terephthalate (BHET) (
14). PET is a thermoplast and partly crystalline. The annual production of PET exceeded 30 million tons in 2017 (
https://www.plasticsinsight.com/global-pet-resin-production-capacity).
Currently, only a few bacteria and fungi have been described for the partial degradation of PET to oligomers or monomers (
8). All known PET hydrolases have relatively low turnover rates. Intriguingly, the trait for PET degradation appears to be limited to a few bacterial phyla, and most bacterial isolates with the potential for PET degradation are members of the Gram-positive phylum
Actinobacteria (
15). The best characterized examples originate from the genera
Thermobifida and
Thermomonospora (
16–23). The enzymes involved in the degradation (e.g., PET hydrolase and tannase, MHETase) are typical serine hydrolases, e.g., cutinases (EC 3.1.1.74), lipases (EC 3.1.1.3), and carboxylesterases (EC 3.1.1.1). These enzymes possess a typical α/β-hydrolase fold, and the catalytic triad is composed of a serine, a histidine, and an aspartate residue (
18,
24). They can also contain several disulfide bonds caused by cysteine residues, which promote thermal stability and specific binding to PET, as shown by the example of PETase from
Ideonella sakaiensis 201-F6 (
25).
Also, for the bacterium
I. sakaiensis, usage of PET as a major energy and carbon source has been described (
25). In addition to the PET hydrolase, the
I. sakaiensis genome codes for a second enzyme that appears to be unique so far and which shares high similarity to the group of tannases, capable of degrading mono(2-hydroxyethyl) terephthalic acid. PET hydrolase as a secreted enzyme produces the intermediate mono(2-hydroxyethyl) terephthalic acid (MHET). MHET is internalized by the cell and hydrolyzed by MHETase. The resulting monomers are then used for bacterial metabolism.
I. sakaiensis is affiliated with the phylum
Betaproteobacteria and belongs to the order
Burkholderiales.
The
I. sakaiensis PETase three-dimensional (3D) structure was elucidated recently (
26). The overall structure most resembles the structures of cutinases. Austin et al. showed that a double mutation (S238F/W159H), which narrows the active site of the enzyme and makes the protein even more like a cutinase resembling the enzyme from
Thermobifida fusca, leads to an improved variant. The majority of the functionally verified PET hydrolases contain a C-terminal disulfide bond, promoting thermal and also kinetic stability (
27–29). The only exception from this so far is a
para-nitrobenzylesterase from
Bacillus subtilis (
30). An additional disulfide bond can be found in
I. sakaiensis PETase, as well as in structural models of the functionally tested PET hydrolases described by Danso et al. (
31). The structural data indicate that PETases bind the polymer with the hydrophobic surface and the substrate-binding cleft. In total, 4 MHET moieties are bound to the protein (one to subsite I and three to subsite II), whereby the ester bond to be cleaved is located between both subsites next to the catalytic serine. The MHETase from
I. sakaiensis that further hydrolyzes MHET to ethylene glycol and terephthalic acid has been recently crystallized ligand free (2.05 Å) and with a nonhydrolyzable MHET analogue bound (2.1 Å). The enzyme possesses a lid domain that almost exclusively confers substrate specificity and activity toward MHET, with a
kcat of 11.1 ± 1.4 s
−1 (
32).
While the
I. sakaiensis enzymes are the best-studied models, other enzymes and organisms have been identified as potent PET degraders. Currently, four enzymes from
Thermobifida species, one from
Saccharomonospora, and one from the phylum
Thermomonospora are known to act on PET. These actinobacterial enzymes are often Ca
2+-dependent, especially in terms of their thermal stability (
33), and they are partially inhibited by their released hydrolysis products MHET and BHET (
33). Therefore, efforts have been made to overcome this limitation; one approach lies in the combination of polyester hydrolases with other enzymes to improve substrate binding and catalytic properties (
26,
34,
35).
Besides the actinobacterial PET hydrolases, fungal cutinases showed activity on PET substrates as well. The most prominent examples are cutinases of the phyla
Fusarium and
Humicola. The latter was also used together with the lipase CalB from
Candida antarctica in order to circumvent the previously mentioned product inhibition by BHET and MHET (
34). While CalB completely converted to terephthalic acid, the
Humicola-derived enzyme was limited in the last reaction step and accumulated the intermediate MHET.
Complementary to the above outlined activity-based approaches, a hidden Markov model (HMM) motif-based large-scale global search of existing genome and metagenome databases has been developed for the presence of potential PET hydrolases (
31). Using this approach, >800 potential PET hydrolases were identified in bacterial and archaeal genomes and metagenomes, and several enzymes were functionally verified (e.g., PET2, PET4, PET6, and PET12). These findings imply that PET hydrolase-encoding genes are globally distributed in marine and terrestrial metagenomes (
31).
Using an
in silico genome mining approach, a cutinase from
Pseudomonas pseudoalcaligenes (PpCutA) and a putative lipase from
Pseudomonas pelagia (PpelaLip) were identified as potential enzymes acting on polyesters in general. Further experimental work using recombinant enzymes of PpCutA and PpelaLip verified the hydrolytic activities of both enzymes on different types of polyesters, including the hydrolysis of polyoxyethylene terephthalate (
36). In their study, the authors used structurally different ionic phthalic acid-based polyesters with an average molecular weight ranging from 1,770 to 10,000 g/mol and semicrystalline polyesters with crystallinity below 1% to test and verify the microbial degradation. Notably, the identified organism belongs to a biotechnologically important novel species within the genus
Pseudomonas, which was designated
Pseudomonas pertucinogena (
37).
In addition to the metagenome-derived PET esterases described above, colleagues recently reported on the functional screening of metagenomes and the characterization of selected enzymes. Among those were the metagenome-derived esterases MGS0156 and GEN0105, which hydrolyzed polylactic acid (PLA) and polycaprolactone, as well as bis(benzoyloxyethyl)-terephthalate. For MGS0156, 3D structural data at 1.95 Å indicate a modified α/β-hydrolase fold with a lid domain and a highly hydrophobic active site (
38). The closest homologue to MGS0156 is an enzyme from
Desulfovibrio fructosivorans with 70% sequence similarity.
In summary, PETases represent the best-explored and -studied class of enzymes with respect to the hydrolysis of synthetic polymers.
Polyurethanes.
Polyurethanes (PUR) can be synthesized by using different polyether or polyester polyols. PUR is a polymer of organic units connected by carbamate. The additional incorporation of aromatic ring structures has further impact on the physical and chemical properties of the polymer. PUR is a widely used synthetic polymer for the production of foams, insulation materials, textile coatings, and paint to prevent corrosion (
39). With over 27 tons produced annually (
2), it ranks fifth among the most often produced synthetic polymers.
To date, only bioactivities that act on the ester-based PUR have been reported (
40,
41). Biodegradation was achieved by either bacteria or fungi. With respect to bacteria capable of degrading PUR, Gram-negative
Betaproteobacteria from the genus
Pseudomonas have been most frequently linked with PUR activities. One of the first enzymes identified to act on PUR was the PueB lipase from
Pseudomonas chlororaphis (
42,
43). This organism codes for at least one additional enzyme active on PUR, which was designated PueA (
44). Both enzymes are lipases; PUR is degraded by the secreted hydrolases, and the degradation is tightly regulated. Their respective genes are part of a larger gene cluster encompassing seven open reading frames (ORFs) (
45).
Pseudomonas protegens strain Pf-5 uses a similar mechanism to degrade dispersions of the polyester PUR. In this strain, however, it was shown that PUR degradation is tightly regulated by mechanisms of carbon catabolite control and that both lipase genes,
pueE and
pueB, appear to be essential for growth on PUR dispersions (
46). In a similar manner,
Pseudomonas putida was reported to degrade PUR at relatively high rates (
47). The bacterium needed 4 days to grow and consume the added colloidal PUR. Yet another example comes from
Comamonas acidovorans TB-35. This strain produces a PUR-active enzyme that is an esterase and which was designated PudA (
48,
49). PudA shows a hydrophobic PUR-surface-binding domain and a distinct catalytic domain, and its surface-binding domain is considered to be essential for PUR degradation. PudA acts as a 62-kDa monomer, and it releases diethylene glycol and adipic acid at an optimum temperature of 45°C and an optimum pH of 6.5.
Within this context, it is perhaps notable that often enzyme activities that are reported are based on clearing zones in agar plates. However, these assays are not fully reliable. For instance, different enzymes from
Pseudomonas spp. and
Bacillus spp. showed significant esterase activities and partially or even completely cleared plates containing colloidal PUR. However, only the
Pseudomonas sp. lipase significantly degraded the added PUR based on nuclear magnetic resonance (NMR) and infrared (IR) data (
50). Furthermore, there is strong evidence that some
B. subtilis and
Alicycliphilus sp. isolates are able to degrade PUR (
51–53).
In a recent publication, Schmidt and colleagues reported on microbial degradation of PUR (i.e., Impranil DLN). The authors of this study employed the known polyester hydrolases LC-cutinase, TfCut2, Tcur1278, and Tcur0390 in their assays and observed significant weight loss of the tested foils when incubated for extended time periods (200 h) at a temperature of 70°C (
54). The observation that cutinases, otherwise known to degrade polyethylene terephthalate, also act on PUR could be attributed to the promiscuous nature of the
Thermobifida-derived cutinases. Recent research on promiscuity of enzymes implies that lipolytic enzymes such as cutinases are very often highly promiscuous and can convert up to 78 different substrates (
55).
While the list of PUR-active bacteria is steadily increasing, a larger number of fungi have also been reported to degrade polyurethane (
41). Notably, the authors of that study identified a 21-kDa metallo-hydrolase from
Pestalotiopsis microspora as a responsible enzyme in PUR degradation.
Additional studies identified
Fusarium solani,
Candida ethanolica (
56), and
Candida rugosa (
57) as PUR degraders. While for
C. rugosa, a lipase has been identified as the key enzyme involved in PUR metabolism, no enzymes were yet identified for
C. ethanolica and
F. solani. Other fungi reported belong to the
Cladosporium cladosporioides complex, including the species
Cladosporium pseudocladosporioides,
Cladosporium tenuissimum,
Cladosporium asperulatum, and
Cladosporium montecillanum, and three others were identified as
Aspergillus fumigatus,
Penicillium chrysogenum (
58), and
Aspergillus flavus (
59). In the case of
A. flavus, it is assumed that secreted esterases are responsible for the degradation. However, no defined enzyme has yet been linked to the observed activities. In a similar study, it was recently reported that
Aspergillus tubingensis colonizes PUR and acts on the surface of films made of PUR. However, no enzyme was linked with the PUR activities (
60).
It is noteworthy that the above-mentioned PUR-active enzymes and organisms were all acting on ester-linked PUR. However, to the best of our knowledge, no enzymes have yet been described acting on polyurethane ethers.
Polyethylene.
Polyethylene (PE) consists of long-chain polymers of ethylene, and it is produced as either high-density (HD-PE) or low-density (LD-PE) polyethylene. PE is chemically synthesized by polymerization of ethane and is highly variable, since side chains can be obtained depending on the manufacturing process. Such modifications mainly have influence on crystallinity and molecular weight. The polymer is most frequently used in the packaging industry as one of the main packaging materials, and more than 100 million tons of PE are produced globally per year (
2,
61) (
Fig. 2).
Possible PE degradation has been affiliated with a surprisingly large number of bacterial genera. Among those were Gram-negative species affiliated with the genera
Pseudomonas,
Ralstonia, and
Stenotrophomonas but also many Gram-positive taxa (e.g.,
Rhodococcus,
Staphylococcus,
Streptomyces,
Bacillus, and others) (see references in Sen and Raut [
62] and Restrepo-Florez et al. [
63]). In addition, fungal genera affiliated with assumed PE degradation were reported; these included
Aspergillus,
Cladosporium,
Penicillium, and others (see references in references
62,
63, and
64–69). In addition, a few studies linked the PE-degrading microbes with the complex gut microbiomes of invertebrates (
70,
71).
It is notable that in almost all the above-mentioned studies on PE-degrading microorganisms, the authors reported on degradation of the polymers using commercial polymers that possibly contained chemical additives, and degradation was determined by measuring weight loss and by Fourier transform infrared spectroscopy (FTIR). Since weight loss and surface structure changes are most likely attributed to the degradation of chemical additives, which often make up a significant fraction of the polymer, the results in these studies need to be verified using more advanced technologies. None of these studies reveled biochemical mechanisms and enzymes involved in PE breakdown. Within this framework, a more recent publication identified a
Penicillium-derived laccase as potentially involved in PE breakdown (
72). Unfortunately, no detailed biochemical characterization was performed, and no sequence of the protein or the corresponding gene was deposited.
Polyamide.
Polyamide (PA) is a polymer of repeating units of aliphatic, semiaromatic, or aromatic molecules linked via amide bonds. Since the monomers for making this polymer can be very versatile, there are many different types of synthetic polyamides, with the most popular being nylon and Kevlar. Synthetic polyamides are mainly used in textiles, automotive applications, carpets, and sportswear (
73).
Remarkably, proteins as well as natural silk are polyamides
per se. Based on this, it should be expected that nature has evolved enzymes that act on these nonnative polymers. However, to date, there is no microorganism known that is able to fully degrade the intact high-molecular-weight polymer. In contrast, several studies are available on bacteria acting on either linear or cyclic nylon oligomers with rather short chain lengths. In one of the first studies, different bacteria were described to grow on various oligomers derived from nylon production (
74). In wastewater of nylon factories, 8-caprolactam, 6-aminohexanoic acid, 6-aminohexanoic acid cyclic dimer, and 6-aminohexanoic acid oligomers accumulate. These compounds can serve as the carbon and nitrogen source for specially adapted bacteria. One of the first bacteria described growing on these mixtures of oligomers was
Flavobacterium sp. strain KI72, which was later renamed
Achromobacter guttatus KI72 and then recently named
Arthrobacter sp. strain KI72 (
74,
75). Nylon oligomer-degrading
Arthrobacter isolates code in their genomes for different hydrolases and several aminotransferases involved in the initial degradation of the oligomers and the subsequent metabolism. In the case of strain KI72, the respective genes are located on an accessory plasmid, pOAD2 (
76–78).
Three main enzymes are essential for the initial hydrolysis of cyclic and linear 6-aminohexanoate oligomers. The first one is a cyclic-dimer hydrolase (NylA), the second a dimer hydrolase (NylB), and the third an endo-type oligomer hydrolase (NylC). NylC is a typical esterase, but its 3D structure also reveals motifs with β-lactamase folds (
79–87). Once the oligomers are hydrolyzed, the monomers are metabolized by different aminotransferases. The draft genome of
Arthrobacter sp. KI72 carries, among others, two genes, designated
nylD1 and
nylE1, that are responsible for the secondary 6-aminohexanoate metabolism. The 6-aminohexanoate aminotransferase (NylD1) catalyzes the reaction of 6-aminohexanoate to adipate semialdehyde. It uses α-ketoglutarate, pyruvate, and glyoxylate as amino acceptors and generates glutamate, alanine, and glycine, respectively. The reaction relies on pyridoxal phosphate as a cofactor. The second enzyme, the adipate semialdehyde dehydrogenase (NylE1), catalyzes the reaction, leading from adipate semialdehyde to adipate. This enzyme requires NADP
+ as a cofactor and is an oxidoreductase (
88,
89).
More recently, diverse marine bacteria were reported to act on nylon. The authors of this study reported a significant weight loss over a time period of 3 months. In their study,
Bacillus cereus,
Bacillus sphaericus,
Vibrio furnissii, and
Brevundimonas vesicularis were identified as potential nylon degraders (
90). The genes and enzymes associated with the nylon degradation, however, remain to be identified, and the possibility cannot be excluded that the weight loss observed was primarily linked to the degradation of chemical additives, as outlined above.
Rather than using the synthetic polymer, Oppermann and colleagues reported on 12 bacterial species capable of degrading the natural polymer poly-γ-glutamic acid. The high-molecular-weight polymer is synthesized by many Gram-positive bacteria as a major component of capsules and slime. In contrast to the synthetic polymer, however, it is a water-soluble molecule and is thus more easily accessible to microbial degradation (
91).
The only enzyme that has so far been reported to act on high-molecular-weight nylon fibers was classified as a manganese-dependent peroxidase and originated from a white rot fungus. The activity of the native and purified enzyme, however, differed from that of lignolytic enzymes. Nylon-degrading activity was quantified by measuring the structural disintegration of nylon-66 membranes. The enzyme had a molecular weight of 43 kDa and was dependent on the presence of lactate and other alpha-hydroxy acids. Unfortunately, no gene or protein sequence was determined (
92).
While the first reports were published in 1965 stating that, among others,
Pseudomonas aeruginosa is able to convert oligomeric nylon, further studies have confirmed that
P. aeruginosa and evolved strain PAO1 are able to efficiently degrade 6-aminohexanoate linear dimers (
74,
93). The main enzymatic activities were assigned to a 6-aminohexanoate cyclic-dimer hydrolase and a 6-aminohexanoate dimer hydrolase. Other
Pseudomonas species have, however, also been reported to utilize 6-aminohexanoate-dimers as a sole carbon and nitrogen source (
94).
Polystyrene.
Polystyrene (PS) [poly(1-phenylethene)] polymer consists of styrene monomers. PS is a widely used synthetic polymer for packaging industries but many daily use articles (CD cases, plastic cutlery, petri dishes, etc.) are also produced from this polymer (
95). In 2016, about 14 million tons were produced (
https://www.plasticsinsight.com/global-pet-resin-production-capacity).
Unfortunately, there is no enzyme known today that can degrade the high-molecular-weight polymer. However, a first report was published recently by Krueger and colleagues on the identification of brown rot fungi able to attack polystyrol by employing hydroquinone-driven Fenton reactions. In this preliminary study,
Gloeophyllum striatum DSM 9592 and
Gloeophyllum trabeum DSM 1398 caused substantial depolymerization after 20 days of incubation. The most active
Gloeophyllum strains caused almost 50% reductions in molecular weight (
96). In an earlier study, the white rot fungi
Pleurotus ostreatus,
Phanerochaete chrysosporium, and
Trametes versicolor and the brown rot fungus
Gloeophyllum trabeum were affiliated with the depolymerization of polystyrene when coincubation together with lignin was performed (
97). While these are first and promising reports on the degradation of the high-molecular-weight polymer, the enzymes involved in the depolymerizing reaction remain to be elucidated. As already outlined above, weight loss may have been caused by the degradation of chemical additives.
Similarly, several bacteria have been reported to form either alone or as members of consortium biofilms on polystyrene films and particles, thereby degrading the polymer. In these studies, mainly weight loss has been assayed. Unfortunately, in none of these studies were enzymes linked to the assumed depolymerization (
98,
99).
While not a single bacterium is known to degrade the polymer, a larger number of bacterial genera that are capable of metabolizing the monomer styrene as a sole source of carbon are known. The biochemistry of styrene metabolism is well understood, and for more detailed reviews, see references
98 and
100–103 and references therein. Styrene degradation in bacteria is well studied in
Pseudomonas,
Xanthobacter,
Rhodococcus,
Corynebacterium, and others. It appears to be a widespread metabolism. Under aerobic conditions, styrene is oxidized by two different pathways, namely, (i) attacking the vinyl side chain and (ii) a rather unspecific aromatic ring, thereby forming primarily the intermediates 3-vinylcatechol, phenylacetic acid, and 2-phenylethanol. These intermediates are channeled into the Krebs cycle after ring cleavage. The degradation of the vinyl side chain involves the action of three key enzymes, a styrene monooxygenase, a styrene oxide isomerase, and a phenylacetaldehyde dehydrogenase (
104). The styrene monooxygenase attacks the vinyl side chain to release epoxystyrene, which is then subjected to isomerization to form phenylacetaldehyde. The latter is oxidized to phenylacetic acid though the involvement of a dehydrogenase. In
P. putida, the phenylacetic acid is activated to phenylacetyl-coenzyme A (CoA) and then subjected to β-oxidation to yield acetyl-CoA, which is directly fed into the Krebs cycle. The respective genes for side-chain oxygenation are frequently located in a single conserved gene cluster, often designated
styABC(D) (
105). Thereby, the
styA and
styB genes code for the styrene monooxygenase complex. The styrene monooxygenase is a two-component flavoprotein that catalyzes the NADH- and FAD-dependent epoxidation of styrene to styrene oxide. StyA is the actual monooxygenase, and StyB functions as flavin adenine dinucleotide (FAD) reductase, which transfers the electrons from NADH to FAD
+ to supply StyA with the required electrons (
106). The
styC gene codes for the styrene isomerase (
107), and
styD is a phenylacetaldehyde dehydrogenase gene (
108). The expression of the conserved cluster is regulated through either a two-component regulatory system or LysR-type regulators (
109–111).
The direct ring cleavage of styrene is initiated by a dihydroxylation of the aromatic ring. This reaction is catalyzed by a 2,3-dioxygenase and followed by a 2,3-dihydrodiol dehydrogenase. The two key products that are formed are styrene
cis-glycol and 3-vinylcatechol. The latter can then be degraded by subsequent meta- or orthocleavage to form acrylic acid, acetaldehyde, and pyruvate. The pathway is rather unspecific for the general degradation of various aromatic compounds, such as phenol or toluene (
100–102).
The produced phenylacetaldehydes are of interest to different industries, as they can be considered building blocks for the production of different fine chemicals or pharmaceutical compounds. They can serve as the starting material to synthesize fragrances, flavors, pharmaceuticals, insecticides, fungicides, or herbicides (
112). Recent studies have also shown that
Pseudomonas putida,
Rhodococcus zopfii, and other Gram-negative species can convert polystyrene (i.e., styrene oil) into the biodegradable polymer polyhydroxyalkanoate or other valuable compounds. The approach involves as a first step the pyrolysis of polystyrene to styrene oil. The styrene oil is then converted in a second step to polyhydroxyalkanoate or other compounds. While the overall concept of this two-step process is intriguing, it may not be feasible on a large scale, as the pyrolysis is a process that runs at 520°C and this is energetically very demanding (
113–115).