The abundance of mRNA transcripts of bacteroidetal polyethylene terephthalate (PET) esterase genes may indicate a role in marine plastic degradation

PET-active Hidden-Markov-Model (HMM)-based search and sheds new light on their potential impact on plastics degradation in the environment.


27
PET is one of the most common plastics used in many consumer products. The worldwide PET 28 resin production amounted to 27.8 million tons in 2015 1,2 . However, only a small fraction of PET is recycled, 29 Zhang et al., 2021; PET hydrolases affiliated with the phylum Bacteroidetes 3 and it is estimated that 58 % ends up in the landfills and in the ocean 3,4 . Our knowledge of microbial 30 degradation of most plastics is rather limited. Degradation is, however, initiated by UV light and or 31 mechanical grinding through waves and other movements generating microplastics (< 5mm) 5 . Thereby, it 32 can be assumed that the microparticles allow better microbial attachment 6,7 . In the case of PET, recent 33 research has demonstrated that some bacteria are able to degrade the polymer. Although it is unclear the 34 larger crystalline fibers are degraded by bacteria, it is well known that cutinases (EC 3.1.1.74), lipases (EC

46
More recently, the complete degradation of amorphous PET materials was described for the Gram-47 negative Betaproteobacterium Ideonella sakaiensis 201-F6, which is capable of using PET as a major 48 energy and carbon source 18 . I. sakaiensis' genome also encodes a tannase that appears to be unique, and 49 which is designated MHETase as it is capable to degrade MHET. Besides these, a number of other PETases 50 affiliated with the Proteobacterial phylum have been identified [19][20][21][22] .

51
In a previous study, we identified potential PET esterases affiliated with the Bacteroidetes phylum 52 using HMM profile database searches 9, 19 . These enzymes were mainly recovered from marine 53 environments and annotated solely on the basis of homology. However, we had not verified the enzymatic 54 function, the environmental distributions and expression of the predicted enzymes within that framework, 55 and because of their global occurrence, we now sought to determine, if the predicted enzymes are indeed 56 acting on the PET polymer. Bacteroidetes representatives can be found in nearly all ecological niches 57 including soils, oceans and fresh water and are part of the microbiome of many animals, especially as inhabitants of the intestinal tract [23][24][25][26] . The Bacteroidetes phylum, however, is highly heterogenous and 59 contains at least four classes of bacteria (e.g. Bacteroidia, Flavobacteria, Sphingobacteria, and Cytophagia) 60 with each class having several thousand described species. The phylum contains non-spore forming and 61 rod shaped aerobic but often anerobic microorganisms with an enormous metabolic diversity 24  Previously, we identified PET-active genes and enzyme candidates affiliated with phylum 77 Bacteroidetes 19 . In this study, we initiated work to enrich the diversity of these genes encoding PET-active 78 enzymes and to validate their catalytic function using experimental approaches. To achieve these goals, we 79 performed global database searches using publicly available data from single bacterial genomes and 80 different metagenomes available through NCBI GenBank. In addition, we searched several private datasets 81 harboring human-and environmental-affiliated Bacteroidetes sequences (TABLE 1)

103
In line with these observations, each candidate contained N-terminal signal domains for protein 104 transport to the periplasm as predicted with SignalP 5.0 35 , further supporting the notion that these are 105 secreted proteins (TABLE 2). Further analyses of the amino acid sequences identified a G-x-S-x-G motif 106 which is typical for α/β serine hydrolases 36 (FIGUREs 1&S1). The catalytic triad consists of the residues 107 Asp-His-Ser and a potential substrate binding site was identified containing the aa Phe-Met-(Trp/Tyr/Ala).

108
The latter differed from the known IsPETase, the LCC and PET2 binding sites in which a Tyr was reported 109 in the first position and position 3 was occupied by a Trp (TABLE 2). PET57 is the only exception with a Trp-110 Met-Tyr binding site.

111
For a more detailed structural inspection, we modeled the structures of all predicted PETases using 112 the IsPETase (PDB code 6QGC) as backbone. These modeling experiments suggested that catalytic parts 113 of the predicted PETases have minor differences in their 3D structures (FIGURE 1 and FIGURE S2).

114
However, the C-terminal part affiliated with the T9SS domain differed largely. It is not present in the IsPETase and was in some cases 100 aa in length (TABLE 2) and consisted of up to seven predicted β-116 sheets and, occasionally, a few α-helices. the Active Site (AS), PET27 and PET30 (orange and purple, respectively) present a C-terminal PorC domain (7x β-strands) for protein 121 secretion via the Bacteroidetes-specific Type IX Secretion System (T9SS) that is not present in the IsPETase (light yellow). b All three 122 enzymes present the typical residues of Ser-hydrolases at the catalytically active positions (Ser, His and Asp), but PET27 and PET30 123 differ in some of the amino acids associated with PET binding. The residues of IsPETase are indicated in black. They also lack a 124 disulfide bridge in the proximity of a catalytic loop. 3D structures were modeled using the Robetta server 37 using the IsPETase crystal 125 structure (6QGC) as a backbone. Figure S1 provides the position of these residues in details on the amino acids level.

128
To expand on our bioinformatic analysis, we cloned and expressed the predicted PETases in 129 Escherichia coli for functional testing. The nine candidate genes were synthesized and cloned into the

149
Additional tests with PET27 and PET30 indicated that these enzymes hydrolyzed the esters para-   , TABLE 3). Surprisingly, under the same conditions, PET30 released only 15.9 ± 9.5 µM TPA 158 (FIGURE 2b, TABLE 3). When we benchmarked these data with self-produced recombinant IsPETase, 1 159 mg ml -1 of IsPETase released under the same conditions 4055.7 ± 516.9 µM of TPA. Thus, the IsPETase 160 is 4.7-fold more active compared to PET27 and approximately 253-fold more active compared to PET30.
While these data clearly demonstrate the capability of both enzymes to act on amorphous PET, the 162 observed differences may be related to a single amino acid substitution in the predicted substrate binding 163 pocket of PET27 and PET30 (TABLE 2). Notably, the IsPETase carries a Tyr-Met-Trp motif in the known 164 and experimentally verified PET binding site. PET27, however, has the Tyr replaced with a Phe in its 165 predicted binding site and PET 30 has in addition the Trp in position 3 replaced with a Tyr (TABLE 2)

201
The presence of DTT and PMSF (1 and 10 mM) inactivated PET30 almost completely. Whereas EDTA at 202 Finally, we asked if the C-terminal sorting domain is of importance for its catalytic activities. To 206 answer this question, we constructed a deletion mutant designated PET30_Δ300-366 that lacked the sorting 207 sequence. Biochemical tests implied that it was not affected in its activities using pNP-hexanoate or PET 208 foil (TABLEs 2 & 3). The enzyme released similar amounts of TPA as it was observed for the wildtype 209 enzyme (TABLE 3).

210
In summary, these data imply that PET30 is a promiscuous mesophilic esterase with highest 211 activities on carboxylic esters with C6-acyl chains, which get increased in the presence Co 2+ , Zn 2+ and Ni 2+ .

224
The observation here that both enzymes were still active at lower temperatures raised the question 225 whether they would also turn over PET foil at these low temperatures. Therefore, TPA release on foil was      TABLE 1 and TABLE S1, and colors were assigned to the tree at the phylum 426 level using the treesapp colour command. This reference tree was also built for and used with TreeSAPP-

490
SI072 medium plus quality MAG sequences were placed in the reference tree using the same methodology 491 as for the contigs described above to better understand and interpret phylogenetic placements from Saanich 492 Inlet. All trees were then visualized in iTOL 78 . It is important to note TreeSAPP abundance calculations as visualized in iTOL maps read abundance at the tips of the reference tree. If a sequence maps closer to the 494 root, that abundance gets split evenly among the children in that cluster, and abundances shown represent 495 a combination of reads that mapped to all nodes feeding into that particular leaf. Multiple sequences can 496 also be placed at the same location in the tree, and reads mapping to all sequences will contribute to the 497 abundance calculation.

526
The impact of cofactors, solvents, detergents, and inhibitors was assayed at different concentration levels.

527
The possible cofactors Ca 2+ , Co 2+ , Cu 2+ , Fe 3+ , Mg 2+ , Mn 2+ , Rb 2+ and Zn 2+ with a final concentration of 1 and 528 10 mM were used. Detergent stability was assayed with SDS, Triton X-100 and Tween 80 at 1 % and 5 % 529 (w/v, v/v) concentration. The inhibitory effect of EDTA, DTT and PMSF was tested at 1 and 10 mM 530 concentration. After 1 h incubation in the presence of these substances, the residual activity was determined 531 after 10 min incubation at the optimal temperature with pNP-hexanoate and at the optimal pH.