Engineering microbial division of labor for plastic upcycling

The study addresses the growing global problem of plastic pollution and explores biological upcycling of polymers, focusing on polyethylene terephthalate (PET). While microbial and enzymatic approaches have advanced PET depolymerization and monomer assimilation, monoculture biotransformations face key challenges: mixed monomers (terephthalic acid, TPA, and ethylene glycol, EG) trigger catabolite repression and hinder simultaneous utilization; crude hydrolysates can be inhibitory; and complex multienzyme product pathways require system-level tuning that is difficult to implement in a single strain. The authors hypothesize that engineered microbial consortia employing division of labor (DOL) can reduce catabolic interference, distribute metabolic load, mitigate toxicity, and provide tunable modules for efficient conversion and upcycling of PET hydrolysate.

Prior work has identified PET-degrading microbes (e.g., Ideonella sakaiensis, Thermobifida fusca, Saccharomonospora viridis) and improved hydrolases via mutagenesis and machine learning to enhance depolymerization efficiency. Engineered bacteria and fungi have been used to hydrolyze PET and convert its monomers to value-added products. However, monoculture systems often exhibit carbon catabolite repression in mixed substrates and sensitivity to inhibitory intermediates. Previous reports achieved mcl-PHA production from TPA or EG, but titers dropped using crude PET hydrolysate. For muconic acid (MA), multi-step processes or glucose-supplemented fermentations improved titers but relied on additional substrates or faced intermediate accumulation (catechol) and toxicity. Synthetic microbial consortia and DOL are emerging strategies that can modularize tasks, reduce cross-talk, and enable tunable ecosystem-level optimization.

Chassis and strain design: Pseudomonas putida was selected for its metabolic versatility and engineering amenability. Two specialists were constructed: Pp-T (TPA specialist) from EM42 by deleting the ped operon to block EG oxidation and introducing the Rhodococcus jostii RHA1 tpa cluster (tpaAa, tpaAb, tpaB, tpaC, tpaK) under Ptac to convert TPA to protocatechuate (PCA). Pp-E (EG specialist) from M31 deleted gclR (to derepress the Gcl pathway) and replaced the native glcDEF promoter/RBS with Ptac and an engineered RBS to enhance glycolate oxidation and prevent glycolate accumulation. A generalist co-utilizer Pp-TE was created by introducing the tpa cluster into Pp-E. Batch fermentations were conducted in modified M9 at 30 °C, 250 rpm, with TPA (as Na2TPA), EG, or mixtures; co-cultures were inoculated at defined OD600 and Pp-T:Pp-E ratios. PET hydrolysis: PET powder was depolymerized by alkaline hydrolysis in an autoclave reactor at 130 °C for 2 h with 1–20% NaOH. Optimal conditions used 5% NaOH plus benzalkonium chloride (BKC) as a phase transfer catalyst, yielding >97% molar yields of Na2TPA and EG. The crude hydrolysate was mixed with M9, neutralized, filtered, and used directly for fermentations. mcl-PHA engineering: For PHA production, Pp-T and Pp-E were engineered with pSEVA421-phaG-alkK-phaC1-phaC2 under PEM7 (yielding Pp-TP, Pp-EP). Further deletions in fatty acid β-oxidation (fadBA, fadBAxE) and depolymerase phaZ generated PpΔ-TP and PpΔ-EP. The generalist Pp-TE was similarly engineered to Pp-TEP and PpΔ-TEP. Controls (Pp-TS, Pp-ES, Pp-TES) harbored empty vectors. Fermentations with pure substrates or PET hydrolysate were performed under nitrogen-surplus or nitrogen-limiting conditions; fed-batch with pulse feeding was also tested. PHA was extracted after substrate depletion and quantified by GC-MS; monomer composition was determined. MA pathway and DOL: For MA production, the pathway was split at catechol (CAT). Upstream (TPA→CAT) was implemented in Pp-TC by deleting catRBCA and introducing codon-optimized aroY and ecdB under Ptac, together with the tpa cluster. Downstream (CAT→MA) was implemented in Pp-EM by deleting catRBC and replacing the native catA promoter with Ptac to increase catechol 1,2-dioxygenase activity. A single-strain control Pp-TEM combined both modules. TC-EM co-cultures were tested on PET hydrolysate across varied initial Pp-TC:Pp-EM ratios and with controlled inoculation time lags to modulate CAT flux and mitigate toxicity. Analytics and composition: TPA, CAT, and MA were quantified by HPLC-UV using a C18 column; EG, glycolate, glyoxylate, and succinate by HPLC-RID on a Rezex ROA column. Community composition dynamics were determined by CFU counts on selective media. Statistical analyses used one-way ANOVA with Dunnett’s multiple comparisons.

Specialist construction and co-utilization: Pp-T consumed 56.2 mM TPA completely within 36 h and tolerated up to 316 mM TPA, but did not use EG; specialization reduced EG-induced catabolite interference compared to a parental strain with ped. Pp-E efficiently consumed EG with negligible intermediate accumulation, did not use TPA, and showed sensitivity to TPA at increasing levels. In co-culture (T-E consortium), TPA and EG were simultaneously and fully catabolized within 48 h, unlike either monoculture. Consortium vs single-strain generalist: The generalist Pp-TE co-utilized TPA and EG but showed delayed TPA assimilation when EG was present, indicative of catabolic crosstalk. Across 16 substrate combinations, the T-E consortium consistently degraded mixtures faster, especially at high concentrations. For 100 mM TPA + 100 mM EG, the consortium completed in 84 h vs 132 h for Pp-TE. The consortium also showed higher transient glycolate and final succinate accumulation, suggesting opportunities to redirect succinate to products. PET hydrolysis and crude hydrolysate fermentation: 5% NaOH + BKC yielded >97% molar yields of Na2TPA and EG. Using crude hydrolysate (~31.6–56.2 mM each), the T-E consortium completed fermentations faster than Pp-TE at both levels. At 100 mM hydrolysate, neither system functioned, indicating increased hydrolysate toxicity compared to pure mixtures. mcl-PHA upcycling: Engineered strains increased PHA production from pure substrates (e.g., Pp-TP and PpΔ-TP outperformed controls on TPA; Pp-EP and PpΔ-EP outperformed controls on EG). For mixed pure substrates, Pp-TEP achieved 392.6 ± 15.7 mg/L PHA, higher than control Pp-TES. On PET hydrolysate under nitrogen limitation, the TP-EP consortium simultaneously depleted TPA and EG in ~65 h and outperformed Pp-TEP, which left 82% EG unconsumed. In fed-batch on hydrolysate, the consortium reached 637.30 ± 10.14 mg/L mcl-PHA, ~92% higher than Pp-TEP; major monomers were C8 and C10 (>75%). Muconic acid (MA) via DOL optimization: Pp-TC (upstream) produced and transiently accumulated CAT from TPA, with a safety-valve MA formation via catA2; Pp-EM (downstream) converted CAT to MA efficiently but was growth-inhibited by high initial CAT. The TC-EM consortium enabled tuning by inoculation ratio and timing. At an optimal initial ratio (Pp-TC:Pp-EM = 1:5), the consortium fully consumed TPA/EG without CAT accumulation or culture blackening and produced 33.28 ± 1.34 mM MA (4.73 ± 0.19 g/L), a 2.84-fold increase over the single-strain Pp-TEM (12.74 ± 0.85 mM). Introducing a 12 h inoculation lag for Pp-TC similarly optimized performance, mirroring the optimal ratio condition.

The engineered division of labor reduced catabolic interference between TPA and EG pathways, enabling simultaneous co-utilization and faster conversion than a generalist strain. Compartmentalizing tasks across specialists improved tolerance to mixed substrates and crude hydrolysate, facilitated modular tuning of product pathways, and allowed ecosystem-level optimization via inoculation ratio and timing. For mcl-PHA, the consortium avoided EG accumulation seen in the generalist and delivered higher titers, particularly in fed-batch from crude hydrolysate. For muconic acid, splitting the pathway at catechol separated production from detoxification, enabling control over intermediate levels and mitigating catechol toxicity through population modulation. These results emphasize the advantages of synthetic consortia for complex substrate processing and upcycling, mirroring the functional roles of multispecies communities observed in environmental plastic degradation.

This work establishes a synthetic microbial consortia platform that fully assimilates PET hydrolysate and efficiently upcycles it into valuable products through engineered division of labor. The T-E consortium co-utilizes TPA and EG faster than a single-strain generalist, especially at high substrate loads and with crude hydrolysate. The TP-EP consortium achieves superior mcl-PHA titers in batch and fed-batch, while the TC-EM consortium enables high-yield muconic acid production by tuning inoculation ratio and timing to avoid catechol toxicity. The modular, orthogonal design and ecosystem-level tunability highlight engineered consortia as a promising route for scalable, robust plastic upcycling. Future work could minimize remaining cross-talk, enhance tolerance to crude hydrolysates, redirect byproduct fluxes (e.g., succinate) to products, and adapt the approach to other polymers and product pathways.

Crude PET hydrolysate exhibited higher toxicity than pure TPA/EG mixtures; at ~100 mM hydrolysate, neither the consortium nor single strains functioned. The EG specialist was sensitive to TPA and crude hydrolysate components, slowing EG assimilation. In some conditions, the consortium accumulated intermediates (glycolate transiently; succinate finally), reducing biomass yield and suggesting incomplete carbon funneling. For MA production, catechol accumulation caused culture darkening and growth inhibition unless population balance and timing were optimized. Engineered production strains sometimes showed slower substrate depletion due to pathway burdens. Further reduction of metabolic cross-talk and improved tolerance to hydrolysate impurities are needed to enhance robustness and scalability.