Skip to main content
NCBI home page
JACS Au logoLink to JACS Au
. 2026 Feb 11;6(2):1347–1358. doi: 10.1021/jacsau.5c01683

Aqueous-Phase Polycondensation of Hydroxy Fatty Acids via a Whole-Cell CoA Activation–Acyltransferase Cascade

Shuming Jin †,‡,§,∥,⊥, Dong Lu †, Qiuyang Wu †, Yulu Zhang †, Ni Jiang †, Junfeng Liu †,‡,§,∥, Fang Wang †,‡,§,∥, Jan Baeyens #, Li Deng †,‡,§,∥,*, Kaili Nie †,‡,§,∥,⊥,*
PMCID: PMC12933353  PMID: 41755814

Abstract

Polyester formation by polycondensation in water is limited by an esterification–hydrolysis equilibrium that strongly favors hydrolysis under mild conditions. Although aqueous-phase esterification has been demonstrated with chemical catalysts and isolated enzymes, these systems typically target small-molecule esters or rely on preactivated donors. The polyhydroxyalkanoate (PHA) pathway is the only known native metabolic route for polyester biosynthesis, but it is essentially limited to the polymerization of 3-hydroxybutyrate and related short-chain hydroxyalkanoate monomers. Consequently, realizing efficient aqueous-phase polycondensation to high-molar-mass polyesters remains a major challenge. Here, an intracellular CoA activation/acyltransferase cascade (ACOS5 At -WS2 Mh ) for poly­(hydroxy fatty acid) (PHFA) biosynthesis is constructed within a whole-cell catalyst. Mechanistic analysis indicates that a mildly hydrophilic region lining the ACOS5 At substrate tunnel is critical for ω-hydroxy fatty acid (ωHFA) recognition and activation in water. The substantial steric bulk of ω-hydroxyacyl-CoA (ωHFA-CoA) hinders its entry into the WS2 Mh hydroxyl-donor channel, enforcing a hydroxyl-terminal chain-growth mode, whereas the higher diffusivity and lower steric hindrance of short oligomers underlie the low dispersity of the resulting PHFA. Chassis engineering, catalytic optimization, and a substrate-channeling fusion-protein strategy collectively increase the cascade flux and the product titer. Under optimized conditions, a PHFA titer of 1.87 g L–1 (M n,app = 10.8 kDa; Đ = 1.05) was obtained in a 3 L bioreactor, yielding polymers with heterotelechelic hydroxyl/carboxyl chain ends. This work establishes a green, fully aqueous, whole-cell route from unactivated ωHFAs to non-PHA polyesters and provides general design principles for engineering living catalysts capable of overcoming hydrolysis-limited equilibria in condensation polymerization.

Keywords: aqueous-phase polycondensation, polyester, whole-cell biocatalysis, CoA activation−transfer cascade, transfer cascade, ω-hydroxy fatty acid

graphic file with name au5c01683_0007.jpg

graphic file with name au5c01683_0006.jpg

Introduction

Polyesters underpin packaging, fibers, and coatings. Hydroxy fatty acids (HFAs), obtainable from fatty acids via biotransformation, , integrate a hydroxyl group with long aliphatic chains, making them attractive monomers for poly­(hydroxy fatty acid) (PHFA) synthesis. − Structure–property relationship studies demonstrate that increasing the length of methylene segments in aliphatic polyesters reduces ester-bond density, enhances crystallinity, and leads to improved mechanical performance. , However, conventional chemical or enzymatic methods often require dehydration, organic media, or metal catalysts and often suffer from side reactions and poorly controlled molecular-weight distributions. From a catalysis perspective, polyester polycondensation in water is fundamentally constrained by hydrolysis-driven equilibria and competing ester hydrolysis. A key unsolved challenge is therefore to drive polycondensation of unactivated hydroxy acids directly in water, without resorting to monomer preactivation or removal of water.

Aqueous-phase esterification has been successfully achieved using chemical catalysis, enzymatic catalysis, and whole-cell biocatalysis. Porous polymer acids enable in-water Fischer esterification, but the process requires strongly acidic conditions and elevated temperatures. Enzymatically, lipases typically require biphasic or nonaqueous environments; Thomas et al. have achieved fully aqueous ester formation by Novozym 435 (immobilized CALB) using buffered aqueous micelles. Kobayashi et al. demonstrated early on that lipase PC could catalyze the polycondensation of dicarboxylic acids and diols in aqueous systems; however, this was only feasible at relatively low monomer concentrations (≤0.1 M), which significantly limits its applicability for large-scale production. In addition, promiscuous hydrolases/acyltransferases and carboxylic acid reductases have been reported to catalyze ester-bond formation in aqueous media. However, these strategies generally rely on preactivated donors (vinyl or ethyl esters) and predominantly generate dimers or short oligomers rather than high-degree polyesters.

For hydroxy carboxylic acid-based polyesters, the industrial process typically involves preparing the monomer as an activated ester, such as hydroxy acid methyl/ethyl esters or lactones, to reduce water formation during subsequent polymerization. These precursors entail corrosive activating agents, extensive solvent use, or costly enzymes, which complicate scale-up and diminish greenness.

Recent studies have explored radical polymerization inside living cells. However, these mechanisms are not suitable for the polycondensation of HFA, which requires dedicated intracellular condensation pathways; native coenzyme A (CoA)-dependent pathways enable the biosynthesis of esters and polyesters. , For aliphatic polyesters, polyhydroxyalkanoates (PHAs) remain the only native polymeric compounds reported to be synthesized directly through cellular metabolism. Recent work has leveraged the PHA pathway for intracellular amidation, producing polyester-polyamide copolymers. However, native PHA synthases (PhaC) predominantly polymerize hydroxyacyl-CoA monomers in which the hydroxyl lies 2–4 backbone carbons from the thioester carbonyl, whereas 6-hydroxyacyl-CoA are incorporated only weakly. This spacing constraint makes HFAs (>C6) mechanistically incompatible as direct PHFA monomers. More broadly, although CoA ligase-acyltransferase pairs provide a versatile platform for synthesizing small-molecule esters, their catalytic scope does not extend to enzymatic, chain-growth elongation polycondensation of long-chain hydroxy acids in water.

An aqueous-phase whole-cell biocatalytic route for PHFA synthesis, leveraging intracellular CoA activation and acyl transfer pathways, is established herein (Figure ). ACOS5 At (Arabidopsis thaliana) activates HFA to HFA-CoA in vivo, and WS2 Mh (Marinobacter hydrocarbonoclasticus) iteratively transfers these acyl units to hydroxyl-terminated chains, enabling chain-growth elongation polycondensation of unactivated ωHFAs directly in aqueous solution. By combining chassis engineering with an activation–transfer cascade, we achieve scalable PHFA formation under fully aqueous conditions. Mechanistic analysis uncovers catalyst-level design principles linking substrate-tunnel microhydrophilicity and hydroxyl-terminal transacylation to polymer growth and dispersity control.

1.

1

Open in a new tab

Intracellular activation–transfer cascade in aqueous medium. (a) The “activation–transfer” cascade mechanism for PHFA formation. ACOS5 At (long-chain acyl-CoA synthetase, LACS) activates HFA to hydroxyacyl-CoA (HFA-CoA) in an ATP-dependent manner. The activated HFA-CoA then undergoes acyl transfer, catalyzed by WS2 Mh (acyltransferase, AT), forming ester bonds with another HFA molecule, leading to polyester chain elongation. (b) Schematic of the aqueous-phase PHFA polymerization using a whole-cell biocatalyst. The system features engineered metabolic channeling between CoA activation (LACS) and acyl transfer (AT). Polymer elongation proceeds by iterative acyl transfer from ωHFA-CoA to a hydroxyl-terminated chain, preserving the ω-hydroxyl as the new chain end.

Results and Discussion

Catalytic Selectivity of ACOS5 At and WS2 Mh

The efficient substrate recognition of HFA by LACS and of HFA-CoA by AT is critical for functional reconstitution of the intracellular activation–transfer pathway involved in PHFA synthesis. To identify a suitable enzyme for the PHFA cascade, the apparent catalytic activity of four LACSs from diverse sources was evaluated: ACOS5 At , FadD Bt (Bacillus thuringiensis), ACS2 Mn (Marinobacter nauticus), and FadD Ec (Escherichia coli). In preliminary experiments, the ACOS5 At activity toward C12–C16 HFAs and the whole-cell formation of HFA ethyl esters were evaluated (Figure S1). Although ω-hydroxy palmitic acid (ωH-PA) showed the highest activity in the cell-free system, the greatest ethyl-ester formation efficiency in whole cells was obtained with ω-hydroxy lauric acid (ωH-LA), likely owing to reduced transmembrane transport limitations compared to ωH-PA. Accordingly, ωH-LA was selected as the substrate for this study.

For the remaining three LACSs, after purification (Figure S2), dual-substrate reactions with equimolar lauric acid (LA) and ωH-LA were performed to evaluate the catalytic selectivity of these LACSs (Figure a,b). Only ACOS5 At exhibited catalytic activity toward ωH-LA. Structure-guided analysis rationalized the substrate preference of ACOS5 At . The overall architecture of the cavity that forms the catalytic pocket enabled substrate binding and subsequent catalysis. Crystallographic analyses of LACSs (Mycobacterium smegmatis FadD32, PDB 5D6N; Mycobacterium tuberculosis FadD8, PDB 1V25) indicate that substrate specificity is largely dictated by the chemical and geometric properties of residues lining the substrate channel. Compared with the other three LACSs, ACOS5 At contains multiple hydrophilic residues in both the principal substrate tunnel and the binding pocket and features a larger substrate entry portal, as supported by CAVER-based tunnel radius profiling in the fatty acyl-AMP-bound conformation, potentially broadening the substrate selectivity (Figure S3).

2.

2

Open in a new tab

Substrate activation and acyl transfer characteristics of ACOS5 At and WS2 Mh in PHFA biosynthesis. (a) Substrate specificity of the four LACSs toward lauric acid (LA) and ω-hydroxy lauric acid (ωH-LA) under cell-free assay conditions. (b) Substrate selectivity of LACSs revealed by high-performance liquid chromatography, ω-hydroxy lauroyl-CoA (ωH-L-CoA) and lauroyl-CoA (L-CoA). (c) Results of hydrophilic mutagenesis in the ACOS5 At substrate tunnel. Catalytic selectivity was defined as the selectivity index PR = [ωH-LA-CoA]/[LA-CoA] (right y-axis). (d) Effects of mutating selected hydrophobic residues lining the ACOS5 At substrate tunnel on catalytic apparent activity and substrate preference. (e) Structural analysis of ACOS5 At . The substrate entrance and binding pocket are highlighted. Key regions include the substrate channel (blue), the C-terminal domain (residues 439–542, orange), and gating residue Tyr237 (green); the model presented was generated by AlphaFold 3. (f) Hydroxyl position selectivity of WS2 Mh . The reactivities of 1-, 2-, 3-, and 4-decanol with fatty acids were tested. The reaction between oleic acid and 1-decanol was normalized to 1.0 for comparison. (g) Substrate-binding model of WS2 Mh , highlighting the spatial accommodation of ω-positioned hydroxyl groups; the model presented was generated by AlphaFold 3.

To assess whether increasing tunnel hydrophilicity could shift catalytic selectivity toward ωH-LA, residues located 4 to 6 Å from the terminal hydroxyl were substituted with hydrophilic side chains of varying lengths. Dual-substrate reactions (LA/ωH-LA) were performed. The selectivity index, defined as the ωH-L-CoA/L-CoA product ratio (PR), increased for several variants, with L214Q and I225E showing the most significant shifts toward ωH-LA (PR of WT ≈ 0.9). Although total activities toward FA and HFA substrates declined in these mutants, the altered product ratios are consistent with a contributing role of local tunnel hydrophilicity in shaping substrate preference (Figures c–e and S4), indicating that a mildly hydrophilic substrate tunnel plays a critical role in tuning the substrate-activation selectivity of ACOS5 At toward ωHFAs.

Acyltransferases demonstrate stringent substrate specificity, which critically governs both the catalytic efficiency and the final molecular weight of the product. Moreover, this selectivity directly dictates structural features of the HFA substrate, such as the acyl chain length and regioselective positioning of the hydroxyl group. In this study, WS2 Mh was selected owing to its capacity to utilize long-chain acyl-CoAs and fatty alcohols, which makes it well-suited for polyester synthesis. This enzyme lacks diacylglycerol acyltransferase activity, thus minimizing possible interference from bypass metabolism during catalysis. Although the chain length preferences of WS2 Mh substrates have been documented, the catalytic efficiency of this enzyme concerning the precise location of the hydroxyl group on the alkyl chain, particularly its position relative to the terminal carbon (e.g., ω or subterminal regions), remains poorly characterized. The results show that WS2 Mh activity is strongly dependent on the hydroxyl position (Figures f and S5). Terminal alcohols (e.g., 1-decanol) give the highest activity, whereas activity decreases as the hydroxyl moves inward and is nearly undetectable for 4-decanol. Structural modeling supports this trend, indicating that internal hydroxyls face steric and conformational constraints that prevent productive positioning at the catalytic center (Figure g). Extending this to polymerization, ω-hydroxyl is most readily captured and engaged as the chain-end acceptor and shows the highest efficiency. In contrast, ω-1 or ω-2 hydroxyls likely require additional conformational adjustment and stricter orientation matching to enter the donor tunnel, reducing the effective transfer rate, and are thus unfavorable for chain growth. Therefore, intracellular polymerization is expected to be most efficient with ωHFAs, and it may extend to ω-1/ω-2-HFAs with reduced efficiency. To overcome the hydroxyl position specificity imposed by WS2 Mh , directed evolution of this acyltransferase may be required. This could involve expanding and remodeling the acceptor binding pocket to enhance the conformational tolerance for substrate entry and positioning, thereby relaxing its regioselectivity and improving acyl transfer and chain growth with more internally hydroxylated substrates.

Whole-Cell Polycondensation of ωHFAs into PHFA

The recombinant plasmid pCOLADuet-ACOS5 At -WS2 Mh was transformed into E. coli BL21­(DE3) for the T7 promoter-driven overexpression of ACOS5 At and WS2 Mh . The catalytic performance of the ACOS5 At /WS2 Mh coexpression system was subsequently evaluated using ωH-LA as substrate at a concentration of 5 mM. Based on the ethanol-precipitation property of high-molecular-weight polyesters, the product was isolated from the concentrated intracellular extract by precipitation with cold ethanol (Figure a). Flocculent precipitates were observed specifically under the conditions of induction with substrate addition (Figure b). To verify that the precipitate contained the target polyester, it was collected and subjected to alkaline hydrolysis, followed by gas-chromatography (GC) analysis of the released monomers. In contrast, the high-molecular-weight polyester in the untreated precipitate was not directly detected by GC analysis due to its low volatility (Figure c). The structure of the intracellularly synthesized polymer was clearly confirmed by 1H NMR spectroscopy. Key characteristic resonances aligned with the expected structure of poly­(ω-hydroxy lauric acid) and enabled calculation of the number-average degree of polymerization (DPn), which ranged from 31–34 repeat units (Figure d, S6). GPC analysis corroborated the number-average molecular weight (M n,app) = 7.0 kDa, the weight-average molecular weight (M w,app) = 9.8 kDa, and the dispersity (Đ) = 1.40 (Figure e). Notably, a distinct resonance at δ 3.57 ppm in the 1H NMR spectrum was assigned to the methylene protons of the ω-hydroxy group, confirming the preservation of hydroxyl end-groups in the synthesized polyester. This end-group functionality not only enhances downstream processability but also expands the material’s potential application in sustainable polymers and biobased coatings, , where functional end-groups are crucial for further modification or controlled degradation.

3.

3

Open in a new tab

Characterization of PHFA biosynthesis and chassis engineering for expanded substrate applicability in whole-cell catalysis. (a) Schematic depiction of the procedure for intracellular PHFA polymerization. (b) Ethanol precipitates under four conditions: 1) induced with substrate 2), ωH-LA dissolved in ethanol, 3) induced without substrate, and 4) uninduced with substrate. (c) GC traces of ethanol-precipitated materials before and after alkaline hydrolysis; ω-hydroxy lauric acid standard shown for RT matching. (d) 1H NMR spectrum of poly­(ω-hydroxy lauric acid). (e) Gel permeation chromatography (GPC) analysis of PHFA synthesized from 5 mM ωH-LA. (f) Schematic of cellular fatty acid metabolism during polymerization. Target LACS-AT route (green) vs β-oxidation bypass (red). FadL facilitates uptake of long-chain HFAs. (g) Polyester titers and ωH-LA conversion in different chassis configurations: A+W (coexpression of ACOS5 At and WS2 Mh ); A+W/FadL (coexpression of ACOS5 At , WS2 Mh and FadL); DE (fadD/fadE knockout strain). (h) Catalytic applicability of HFAs with different chain lengths (C12–C16).ωH-MA: ω-hydroxy myristic acid. (*) Titer increase upon FadL overexpression.

The whole-cell catalytic efficiency was significantly influenced by intracellular metabolic pathways and transmembrane substrate transport. To minimize competition with polyester biosynthesis, fadD was deleted to reduce the conversion of endogenous free fatty acids to acyl-CoAs, and fadE was knocked out to prevent β-oxidation of the HFA-CoA intermediates. In addition, overexpression of the endogenous long-chain fatty acid transporter FadL enhanced the uptake of HFAs (Figure f). Whole-cell catalysis was therefore conducted in an E. coli strain deficient in fadD and fadE and overexpressing FadL (Figure g). Compared with the wild-type (WT) strain, the ΔfadD/ΔfadE double knockout strain (hereafter DE) increased the polyester titer from 66 mg L–1 to 107 mg L–1 (∼1.62-fold improvement).

The substrate scope was evaluated by using C12–C16 ωHFAs (Figure h). The titers obtained with ωH-MA and ωH-PA were lower than that with ωH-LA, and the benefit of FadL overexpression became more pronounced with increasing chain length (ωH-LA, +20.9% vs ωH-PA, +72.1%). These results indicate that the enhanced hydrophobicity of substrates beyond C12 increasingly imposes severe uptake and mass-transfer limitations. The NMR and GPC results for the PHFAs derived from ωH-MA and ωH-PA are presented in Figure S7. Importantly, this chain-length-dependent performance in the whole-cell system should be distinguished from the intrinsic catalytic activity of ACOS5 At . Although ACOS5 At exhibits the highest in vitro activity toward ωH-PA (Figure S1), its efficiency decreases under whole-cell conditions. This contrast suggests that the reduced titer with longer-chain substrates is primarily attributed to their limited solubility and inefficient membrane uptake in the fully aqueous environment, rather than the enzyme’s inherent substrate preference. While extending this system to longer-chain ωHFAs is theoretically feasible, its practical efficiency in an exogenous addition system is constrained by the limitations. For such systems, these challenges could be mitigated by strategies aimed at improving substrate solubility or enhancing membrane uptake, such as increasing the membrane permeability or introducing specific transporters. Alternatively, a more integrated approach would involve developing an endogenous ωH-PA production pathway coupled with the acyl-transfer-based chain-growth elongation system for polyester synthesis.

Intracellular Substrate Channeling System

Whole-cell catalysis functions as an intracellular multienzyme system, whose efficiency depends on both the intrinsic enzyme activities and spatiotemporal dynamics, particularly substrate diffusion and metabolic flux distribution. To overcome diffusion limitations of high-energy intermediates in a two-enzyme cascade, an intracellular substrate channeling system was employed to enable direct intermediate transfer between sequential enzymes, thus improving catalytic efficiency and reducing intermediate loss. , Here, ACOS5 At and WS2 Mh were engineered as a fusion protein to spatially restrict the diffusion of the intermediate ωH-L-CoA (Figure a). The influence of the enzyme orientation, linker length, and rigidity on the catalytic efficiency of the fusion enzyme was systematically investigated (Table S1). The fusion protein engineered with a GGGGS linker (N–C orientation) demonstrated a significant enhancement in polyester titer, achieving 0.66 g L–1, which substantially exceeded the titer obtained with individually expressed enzymes or other fusion constructs (Figure b). Interestingly, although the fusion-protein strategy led to a higher product titer, it concurrently resulted in an approximately 20% decrease in the molecular weight of the polymer at the same substrate concentration (Figure c).

4.

4

Open in a new tab

Mechanistic insights into the intracellular substrate channeling system and hydroxyl-terminal transacylation in PHFA biosynthesis. (a) Schematic illustration of cascade catalysis mediated by ACOS5 At -WS2 Mh fusion constructs. (b) Effect of different fusion protein architectures on polyester titers and substrate conversion. Photographs above show the extent of ethanol precipitation of polymer products. (c) Molecular weights of PHFA synthesized using various fusion configurations. (d) Proposed mechanisms for chain extension during polymerization: (i) hydroxyl-terminal transacylation, (ii) CoA-terminal transacylation, and (iii) oligomer coupling via intermolecular transacylation. (e) AlphaFold 3-based model of WS2 Mh with docked HFA (yellow) and HFA-CoA (green): the hydroxyl-donor region (blue) and the acyl-chain region (orange) form a narrow channel. (f) Mechanism of polyester chain elongation and termination. (g) Chain-growth PHFA synthesis through iterative HFA-CoA activation and hydroxyl-terminus acyl transfer.

Chain-Growth Mechanism

To elucidate the enzymatic mechanism of polyester chain elongation, we systematically evaluated the catalytic behavior of WS2 Mh . Three potential mechanisms were considered: (i) hydroxyl-terminal transacylation, involving sequential acyl transfer from ωHFA-CoA to the terminal ω-hydroxyl of the growing chain; (ii) CoA-terminal transacylation, requiring a CoA-terminated polyester to react with −OH from another ωHFA-CoA; and (iii) oligomer coupling via intermolecular transacylation (Figure d). In the latter two mechanisms, the retention of a CoA-terminated chain end is essential for continued polyester elongation.

For the CoA-activation catalyzed by ACOS5 At , substrate profiling has identified ACOS5 At as a medium- to long-chain acyl-CoA synthetase, with both shorter and longer saturated fatty acids competing much less efficiently in CoA-formation assays. Consistent with this, structural analysis indicates that a branching pocket within the substrate tunnel further restricts the accessible chain length, so that ACOS5 At efficiently activates only monomeric ωHFAs, whereas extended hydroxyacyl oligomers with backbones longer than C24 cannot be productively accommodated. Consequently, ACOS5 At effectively supplies ωHFA-CoA only from monomeric ωHFAs, so that initiation of mechanisms (ii) and (iii) can occur solely between two ωHFA-CoA molecules. This requirement makes the identity of the hydroxyl donor (ωHFA or ωHFA-CoA) in the first WS2 Mh transacylation step a critical determinant of the overall pathway.

For the transacylation catalyzed by WS2 Mh , AlphaFold 3-based structural modeling in complex with acyl-CoA revealed two narrow tunnels that specifically accommodate long-chain aliphatic substrates (Figure S8). Comparison with the Marinobacter VT8 wax synthase WS/DGAT crystal structure (Maqu_0168; 6CHJ) confirms congruent hydroxyl-donor and acyl-chain binding regions (Figure S9). These tunnels are functionally partitioned into a hydroxyl-donor binding region and an acyl-chain binding region for the acyl-CoA. To reach the catalytic center of WS2 Mh , a hydroxyl donor must traverse the hydroxyl-donor tunnel, which spans approximately 32.8 Å from the tunnel entrance to the catalytic HHXXXDG motif. However, because the acyl chain of ωHFA-CoA spans only ∼27.6 Å, the substantial steric hindrance caused by the bulky CoA headgroup, compounded by the confined and deep architecture of the hydroxyl-donor tunnel (Figure e, S10), prevents proper positioning of the ω-hydroxyl group in the catalytic center. These structural constraints indicate that during the transacylation step, free ωHFA rather than ωHFA-CoA is the preferred hydroxyl donor, implying that propagating oligomeric chains are predominantly maintained as carboxyl-terminated species.

To further probe which chain-elongation pathway operates under our conditions, a dual-substrate competition assay was designed by simultaneously introducing equimolar LA and ωH-LA into the reaction system. The results revealed a fatty-acid-mediated chain-termination effect: LA strongly inhibited the polymer elongation. Alcohol precipitation of cellular extracts yielded no high-molecular-weight polymer; only the dimeric product, laurate-ω-hydroxy lauric acid ester, was detected (Figure S11). This chain termination behavior is attributed to the substrate preference of ACOS5 At . Once the FA concentration surpasses a specific level in the reaction system containing both ωHFA and FA, ACOS5 At preferentially generates FA-CoA and uses HFA as a hydroxyl donor to form a dimeric ester. This dimer lacks a reactive hydroxyl group, thereby preventing further chain elongation (Figure f). These results further indicate that the carboxyl-terminal ends of the oligomers cannot be effectively reactivated as CoA thioesters by ACOS5 At .

In parallel, the tunnel architecture of WS2 Mh renders direct use of the terminal hydroxyl group of ωHFA-CoA as the hydroxyl donor in the transacylation step geometrically and sterically infeasible. Thus, models (ii) and (iii), which both require a persistently CoA-terminated chain end for further elongation, are unlikely to represent the major chain-elongation polycondensation pathway. Instead, the combined biochemical, structural, and competition data support a stepwise hydroxyl-terminal transacylation mechanism (i), in which ACOS5 At continually supplies ωHFA-CoA as the acyl donor, while free ωHFAs or hydroxyl-terminated oligomers serve as the hydroxyl donors for WS2 Mh -catalyzed chain elongation.

Based on these results, it is proposed that the “activation–transfer” cascade proceeds predominantly via chain-end elongation, consistent with a chain-growth mechanism. This process involves an activation step, in which ACOS5 At catalyzes successive activation of HFAs to generate HFA-CoA monomers, and a transfer step, in which WS2 Mh transfers these monomers sequentially to the hydroxyl terminus of the oligomer through acyl transfer. These two sequential reactions constitute the overall polymerization process (Figure g). Additionally, the native esterase of E. coli preferentially catalyzes short-chain or water-soluble ester substrates, whereas the lack of a true lipase within its genome significantly limits its degradation capacity for long-chain fatty acid esters (LCFAEs). This inherent inability to hydrolyze LCFAEs ensures that precursor molecules for PHFA synthesis can be accumulated intracellularly. Moreover, the acyl-transfer-based chain-growth elongation system for polyester synthesis, unlike conventional chemical or enzymatic methods, achieves high selectivity by utilizing HFA-CoA as the sole efficient acyl donor. This mechanism prevents polyester chain backbiting via WS2 Mh . As a result, this system generates fewer byproducts, yielding PHFA with a narrow Đ. In the fusion enzyme system, ωH-L-CoA generated by ACOS5 At is rapidly captured and transferred by the adjacent WS2 Mh domain to the hydroxyl terminus of the growing polyester chain. This local channeling minimizes diffusive loss or competing reactions of ωH-L-CoA and increases the number of initiating chains and allows more even distribution of the limited ωH-L-CoA substrate among multiple extending chains. Under this allocation pattern, the average degree of polymerization (and thus M n) per chain decreases, while the total amount of polyester formed increases, consistent with the observed combination of higher titer and reduced molecular weight in the fusion system. Subsequent optimization of substrate concentration or balancing acyl supply with chain elongation kinetics may further improve the degree of polymerization and product distribution.

From the perspective of the whole-cell PHFA polymerization pathway, the observed low Đ (Figures e and c) not only arises from the intrinsic advantage of the activation–transfer polymerization mechanism but also benefits from optimization of the chassis strain and catalytic process. First, polyester formation primarily occurs via terminal hydroxyl-directed transacylation rather than random hydroxyl–carboxyl coupling. In addition, because short-chain oligomers diffuse more rapidly and experience lower steric hindrance, the narrow hydroxyl binding tunnel of WS2 Mh preferentially accommodates the short-oligomer, whereas terminal hydroxyl groups on longer chains access this tunnel less efficiently and are therefore elongated more slowly. This dynamic bias tends to preferentially extend shorter chains while limiting overgrowth of longer ones, thereby passively compressing the chain-length distribution, reducing chain-to-chain variation. As a result, PHFA exhibits a relatively low Đ (∼1.40) even under unoptimized initial conditions. Building on this, further system-level optimizations reduced Đ to 1.05. First, using a fadD knockout chassis minimized the nonspecific conversion of intracellular FA to FA-CoAs, reducing premature chain termination by FA-CoA and resulting in more uniform chain lengths. Second, construction of a substrate channeling strategy accelerated the conversion of key intermediates (ωHFA-CoAs), reducing diffusional loss and side-pathway interference while improving the polymerization efficiency. Third, dispersity was found to be closely linked to the physiological state of the whole-cell catalyst. Upon scale-up to a 3 L bioreactor with optimized nutrient supply and dissolved oxygen conditions, intracellular energy and cofactor supply (ATP/CoA) were enhanced. This led to more robust cellular metabolism and energy balance, ultimately improving product uniformity and achieving a final Đ of 1.05.

5.

5

Open in a new tab

Optimization and scale-up of whole-cell catalytic reactions for PHFA production. (a) Comparison of PHFA titers and substrate conversion rates under different cultivation conditions (25 mM ωH-LA). (b) Dependence of polymer molar mass on the substrate supply. SQubstrate concentration denotes the cumulative ωH-LA added (5 mM × n feeds). (c) GPC characterization of PHFA (25 mM ωH-LA).

Optimization and Scale-up of the PHFA Biosynthesis

After optimization of the cellular chassis of the biocatalyst, further efforts were directed toward enhancing the catalytic process itself. Unlike chemical or in vitro enzymatic catalysis, this system relies heavily on intracellular cofactors such as ATP and CoA, requiring the host cells to maintain a high metabolic activity and stability. In addition, the substrate concentration needed to be carefully controlled to minimize enzyme inhibition and cellular toxicity. To identify optimal induction and reaction conditions, the DE (ACOS5 At -GGGGS-WS2 Mh + FadL) strain was employed as a whole-cell catalyst. Four key parameters were systematically optimized: IPTG concentration, catalytic temperature, substrate concentration, and carbon source. Using exogenously added ωH-LA as the substrate, single-factor optimization identified 0.05 mM IPTG, a reaction temperature of 30 °C, a one-time 5 mM substrate charge, and supplementation with 1 g L–1 glucose as the optimal conditions. Under these conditions, at a substrate concentration of 5 mM, 73.3% of the maximum intracellular product accumulation was reached after 8 h (Figure S12).

Considering that the polyester product accumulates intracellularly, the cell density per unit volume may affect the final polyester titer. In a 3 L bioreactor scale-up experiment, glycerol feeding was used to increase cell density; the fed culture reached an OD600 2.33-fold higher than that of the unfed control (Figure S13). Under these conditions, total cell dry weight increased 1.96-fold, and polyester titer rose 1.73-fold, reaching 1.87 g L–1 (Figure a). The corresponding space–time yield improved from 27.1 mg L–1 h–1 in batch operation to 46.8 mg L–1 h–1 under fed-batch conditions (Table S2). This indicates a positive correlation between polyester titer and cell density: under nutrient-sufficient conditions, highly active, high-density cells markedly enhance catalytic efficiency. In terms of process performance, this cell-density-dependent increase in titer directly translates into a higher space–time yield, providing a simple handle for intensifying volumetric productivity in whole-cell polycondensation without changing temperature, pH, or solvent composition. In addition, the effect of the substrate concentration on polyester biosynthesis was investigated. With ωH-LA added at concentrations of 5, 10, 15, 20, and 25 mM, the results showed that substrate concentration was positively correlated with the degree of polymerization of the product (Figure b). This demonstrates that the DPn of the polyester product could be tuned by adjusting the substrate concentration. At a total substrate concentration of 25 mM (added in five batches of 5 mM each), 1H NMR analysis showed that the polyester comprised 50–51 monomer units. Assuming a repeating unit of (−O–(CH2)11–CO−) derived from ωH-LA with a molecular weight of ∼198 g mol–1, the M n (NMR) was calculated to be about 10.0 kDa (Figure S14). Further GPC analysis showed the polyester had a Đ of 1.05 and M n,app of 10.8 kDa (Figure c), consistent with NMR results. The Đ of PHFA samples in this study ranged from 1.05–1.40, lower than values commonly observed for chain-growth polyesters, which exhibit Đ > 1.5, indicating well-controlled polymerization with a narrow molecular weight distribution, beneficial for subsequent processing and performance stability.

Compared with other reported polyester production strategies (Table S3), both conventional enzymatic and chemical synthesis methods generally cannot be performed in a pure aqueous phase. These processes typically require organic solvents or excess substrates to enhance mass transfer efficiency. Moreover, the chain-extension step in these systems is often poorly controlled, leading to polymers with broad and difficult-to-regulate Đ. Biological strategies, primarily those based on the native PHA biosynthesis pathway, benefit from inherent metabolic advantages and can achieve relatively high product titers. Nevertheless, the Đ of PHA polymers is generally broader than that observed in our system. The titer of our method is currently lower than that of the PHA pathway. However, it possesses significant potential for future optimization. This can be achieved by engineering the substrate preference and enhancing the performance of the key enzyme ACOS5 At . Mechanistically, a further comparison reveals that the PHFA synthesis pathway operates via a distinct intracellular activation–transfer cascade. Here, ωHFA is first activated into high-energy ωHFA-CoA intermediates, which are then iteratively transferred by WS2 Mh to extend the polymer chain. The energy released during the conversion from thioester to ester bonds provides a strong thermodynamic driving force, enabling efficient polyester synthesis in fully aqueous, cell-based systems. This approach differs fundamentally from the PHA pathway, in which monomeric CoA donors are primarily derived from acetyl-CoA without requiring additional activation. Moreover, polymerization catalyzed by PhaC favors short-chain monomers and is largely ineffective toward long-chain substrates such as ωHFAs. Additionally, traditional enzymatic esterification (lipase or esterase catalyzed) in polyester synthesis relies on a serine-based catalytic triad, making it highly susceptible to hydrolysis in aqueous environments.

On the other hand, although continuous supplementation of the HFA substrate in the medium led to increased PHFA titers and higher molecular weights, a notable decline in substrate conversion efficiency was observed with increasing substrate concentrations (34.61% at 25 mM compared to 64.09% at 5 mM). A similar trend was already evident in the single-factor optimization experiments, where single-batch addition of high substrate concentrations was less favorable for PHFA formation than for lower loadings. Part of this effect can be attributed to substrate-induced cellular stress. To directly probe the catalytic contribution, the apparent kinetics of ACOS5 At toward ωH-LA in this study was determined (Table , Table S4) and ACOS5 At exhibited pronounced substrate inhibition at elevated ωH-LA concentrations (>230 mM). Although most kinetic measurements on long-chain acyl-CoA synthetases have been carried out at relatively low substrate concentrations and therefore have not revealed pronounced substrate inhibition, , a few studies have documented similar inhibitory behavior at elevated fatty acid levels for specific LACS family members (such as the Haemophilus parasuis acyl-CoA synthetase FadD2, and rat ACSL4), in line with the behavior observed for ACOS5 At in this work. Taken together, these data indicate that the productivity–conversion trade-off at high ωH-LA loadings arises primarily from kinetic substrate inhibition of ACOS5 At , and they underscore the importance of decoupling total substrate input from instantaneous extracellular concentration when designing whole-cell polycondensation processes.

1. Apparent Kinetic Parameters of ACOS5 At for LA and ωH-LA.

Substrate K m (μM) k cat (s–1) K i (μM) k cat /K m (s–1 μM–1)
LA 52 ± 9 2.9 ± 0.3 261 ± 49 0.056
ωH-LA 112 ± 21 3.9 ± 0.5 231 ± 42 0.035
Open in a new tab
a

Initial velocities were measured at 20–800 μM substrate and fitted to a substrate-inhibition model v = V max[S]/(K m + [S]­(1 + [S]/K i)) by nonlinear regression; apparent kinetic parameters are reported as mean ± SD (n = 3 independent experiments, Figure S15).

To improve the substrate utilization efficiency, simple recovery or recycling strategies were explored. After the reaction, a portion of the unconverted substrate was removed, along with the cells during centrifugation. The majority remained in the extracellular fermentation broth and could be recovered either by direct acidification–extraction or by broth recycling (Table S5). In the first strategy, acidification with dilute hydrochloric acid facilitated the precipitation of most of the substrate dissolved in the fermentation broth. At a substrate concentration of 25 mM, up to 2.33 g L–1 of ωH-LA could be recovered. Alternatively, reinoculating induced, metabolically active cells into the supernatant from the previous batch, supplemented with 1 g L–1 glucose, gave a PHFA titer of 0.799 g L–1. This strategy increased the overall substrate conversion efficiency to 49% (Table S5), demonstrating the feasibility of improving the substrate economy through simple process integration.

Fed-batch feeding and substrate recycling improved ωHFA utilization and PHFA titers, but the current system still has certain limitations. First, ACOS5 At exhibits substrate inhibition at high ωHFA levels, reducing the level of ωHFA-CoA formation and limiting chain elongation. Second, ωHFA activation is ATP- and CoA-dependent, increasing energy demand and relying on a limited intracellular free-CoA pool. Third, ACOS5 At shows higher affinity for FA than HFA (Table ). While FA levels are low under current feeding strategies, future integration of endogenous ωHFA biosynthesis may elevate FA levels, leading to competition for activation and hindering polymerization. To overcome these challenges, future work should focus on: (i) engineering ACOS5 At ’s substrate tunnel to enhance ωHFA selectivity and relieve substrate inhibition; (ii) improving cofactor supply through enhanced CoA biosynthesis/recycling and ATP regeneration systems; and (iii) optimizing endogenous ωHFA production by tuning FA generation/release and improving ωHFA-specific channeling to reduce FA competition and maintain a stable activation–transfer flux under scaled conditions.

Properties of PHFA from Whole-Cell Biosynthesis

The PHFA (M n,app ≈ 10.8 kDa) exhibits a combination of a low melting temperature and high thermal stability. Thermogravimetric analysis (TGA) revealed an onset decomposition temperature (T onset) of 398.3 °C and a maximum degradation temperature of 423 °C (Figure S16), suggesting safe thermal processing well below 300 °C and the feasibility of thermal recycling. Differential scanning calorimetry (DSC) measurements showed a melting temperature (T m) of 85.2 °C and a crystallization temperature (T c) of 70.0 °C (Figure S16), suggesting moderate processability within a midrange temperature window. The supercooling gap (ΔT = T m – T c = 15.2 °C) implies relatively fast crystallization kinetics, which is advantageous for high-cycle molding and extrusion operations. The measured melting enthalpy (ΔH m) of 56.8 J g–1 is consistent with a moderately crystalline polymer structure.

A salient feature of this material is the preservation of both hydroxyl and carboxyl functional groups at the polymer chain ends, which are crucial to its functional versatility. They enable further modification or the incorporation of bioactive agents, making it a promising candidate for biomedical applications, including drug-delivery systems. , Compared with PLA, which undergoes autocatalytic hydrolytic erosion in aqueous and physiological environments, long-chain polyesters are more hydrophobic and degrade more slowly, thereby providing extended release profiles., By analogy, PHFA is expected to provide a more tunable and slower hydrolysis behavior. This controlled degradation behavior is particularly advantageous for applications requiring sustained release, such as in long-term implantable drug delivery devices. Furthermore, PHFA’s intrinsic hydrophobicity and mechanical robustness may support applications including tissue-engineering scaffolds, protective coatings, and biodegradable implants.

For the mechanical properties of PHFA-type polyesters, several studies have provided benchmark values using chemically synthesized polyesters derived from long-chain ωHFA. High-molecular-weight poly­(ω-pentadecalactone) combines high tensile strength (≈60.8 MPa) with high elongation (≈650%), and similar toughening has been reported for poly­(ω-hydroxyl tetradecanoic acid) (≈700% elongation; true tensile strength ≈50 MPa) and poly­(10-hydroxydecanoate) (>1400% elongation with ∼20 MPa strength). In contrast, PLA is typically stiff but brittle (elongation at break often <10%), whereas PCL is ductile but limited by lower strength (16–24 MPa), lower modulus (240–420 MPa), and a low melting point (∼60 °C). Collectively, these comparisons suggest that, once produced at higher molecular weight, PHFA-type long-chain polyesters could provide improved strength–toughness relative to PLA and enhanced thermomechanical stability relative to PCL, while retaining biodegradability and reactive end groups. Future work will focus on increasing molecular weight and leveraging end-group chemistry for chain extension, cross-linking, or network formation to broaden the accessible mechanical-property space. ,

Conclusion

This work establishes a whole-cell biocatalytic system for aqueous-phase polymerization of PHFA and demonstrates the microbial synthesis of non-native polyesters in living cells. This intracellular pathway effectively addresses key limitations of traditional chemical and enzymatic synthesis routes, which typically rely on monomer preactivation, organic or nonaqueous media, and dehydrating conditions. Through systematic chassis engineering, fusion-protein design, and catalytic optimization, fed-batch operation afforded PHFA with a M n,app of up to 10.8 kDa and a titer of 1.87 g L–1. The narrow dispersity (Đ = 1.05–1.40) indicates robust synthetic stability and reproducibility. Mechanistic analysis reveals that (i) hydrophilic amino acid residues lining the ACOS5 At substrate tunnel contribute significantly to ωHFA recognition and activation, thereby providing a basis for extending the acceptable ωHFA chain-length range and tuning polymer properties; (ii) in the ACOS5 At -WS2 Mh cascade, intracellular polymerization proceeds predominantly via a chain-growth elongation polycondensation based on hydroxyl-terminal transacylation in which efficient chain growth requires that the terminal hydroxyl group remain accessible to the hydroxyl-donor tunnel of WS2 Mh ; and (iii) dynamic “short-oligomer-preferred” regulation during chain elongation and acyl transfer likely contributes to the formation of low-dispersity PHFA. Future work should therefore focus on engineering ACOS5 At to enhance selectivity for ωHFA and attenuate substrate inhibition. In summary, this work demonstrates that intracellular polyester biosynthesis can be extended beyond native PHA pathways, opening new avenues for the microbial production of structurally diverse and high-performance biodegradable polymers. We provide a green and scalable route for synthesizing PHFA from renewable feedstocks and lay the groundwork for engineering living cells capable of producing next-generation polyesters with tailored material properties.

Methods

All information describing experimental details can be found in the Supporting Information.

Supplementary Material

au5c01683_si_001.pdf (1.6MB, pdf)

Acknowledgments

This work was supported by the Interdisciplinary Research Center of Beijing University of Chemical Technology (No. XK2025-05).

Glossary

Abbreviations

HFA

hydroxy fatty acid

PHFA

poly­(hydroxy fatty acid)

PLA

polylactic acid

PCL

polycaprolactone

PHAs

polyhydroxyalkanoates

LACS

long-chain acyl-CoA synthetase

AT

acyltransferase

CoA

coenzyme A

NMR

nuclear magnetic resonance

GPC

gel permeation chromatography

TGA

thermogravimetric analysis

DSC

differential scanning calorimetry

GC

gas chromatography

PDB

Protein Data Bank

IPTG

isopropyl β-D-1-thiogalactopyranoside

OD600

optical density at 600 nm

WT

wild type

M n,app/M w,app

apparent number-/weight-average molecular weight

Đ

dispersity

DPn

number-average degree of polymerization

T m/T c

melting/crystallization temperature

T onset

onset decomposition temperature

ΔH m

melting enthalpy

ΔT

T m–T c

LA

lauric acid

ωH-LA/ωH-PA/ωH-MA

ω-hydroxy lauric/palmitic/myristic acid

DE

ΔfadD/ΔfadE double knockout strain

WS/DGAT

wax synthase/diacylglycerol acyltransferase

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/jacsau.5c01683.

  • Experimental procedures; strains/plasmids; cultivation and whole-cell catalysis; enzyme purification and activity assays; PHFA production and analysis; analytical methods (HPLC, GC); scale-up strategy with supernatant reuse; HFA recovery; characterization protocols (NMR, GPC, TGA, DSC); AlphaFold 3 modeling; Supporting Figures S1–S16; Supporting Tables S1–S6; references (PDF)

All authors have given approval to the final version of the manuscript.

The authors declare no competing financial interest.

References

  1. Gomez de Santos P., González-Benjumea A., Fernandez-Garcia A., Aranda C., Wu Y., But A., Molina-Espeja P., Maté D. M., Gonzalez-Perez D., Zhang W.. Engineering a highly regioselective fungal peroxygenase for the synthesis of hydroxy fatty acids. Angew. Chem. 2023;135(9):e202217372. doi: 10.1002/ange.202217372. [DOI] [PubMed] [Google Scholar]
  2. Becker A., Böttcher D., Katzer W., Siems K., Müller-Kuhrt L., Bornscheuer U. T.. A Biocatalytic Cascade Reaction to Access a Valuable Long-Chain ω-Hydroxy Fatty Acid. ChemCatChem. 2020;12(16):4084–4089. doi: 10.1002/cctc.202000745. [DOI] [Google Scholar]
  3. Biermann U., Bornscheuer U. T., Feussner I., Meier M. A., Metzger J. O.. Fatty acids and their derivatives as renewable platform molecules for the chemical industry. Angew. Chem., Int. Ed. 2021;60(37):20144–20165. doi: 10.1002/anie.202100778. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Petrovic Z. S., Milic J., Xu Y., Cvetkovic I.. A chemical route to high molecular weight vegetable oil-based polyhydroxyalkanoate. Macromolecules. 2010;43(9):4120–4125. doi: 10.1021/ma100294r. [DOI] [Google Scholar]
  5. Gao M., Leng X., Zhang W., Wei Z., Li Y.. A biobased aliphatic polyester derived from 10-hydroxydecanoic acid: Molecular weight dependence of physical properties. Polym. Test. 2020;82:106295. doi: 10.1016/j.polymertesting.2019.106295. [DOI] [Google Scholar]
  6. Liu C., Liu F., Cai J., Xie W., Long T. E., Turner S. R., Lyons A., Gross R. A.. Polymers from fatty acids: Poly (ω-hydroxyl tetradecanoic acid) synthesis and physico-mechanical studies. Biomacromolecules. 2011;12(9):3291–3298. doi: 10.1021/bm2007554. [DOI] [PubMed] [Google Scholar]
  7. Zhang Q., Song M., Xu Y., Wang W., Wang Z., Zhang L.. Bio-based polyesters: Recent progress and future prospects. Prog. Polym. Sci. 2021;120:101430. doi: 10.1016/j.progpolymsci.2021.101430. [DOI] [Google Scholar]
  8. Nelson T. F., Rothauer D., Sander M., Mecking S.. Degradable and Recyclable Polyesters from Multiple Chain Length Bio-and Waste-Sourceable Monomers. Angew. Chem. 2023;135(43):e202310729. doi: 10.1002/ange.202310729. [DOI] [PubMed] [Google Scholar]
  9. Yu J., Xu S., Liu B., Wang H., Qiao F., Ren X., Wei Q.. PLA Bioplastic production: From monomer to the polymer. Eur. Polym. J. 2023;193:112076. doi: 10.1016/j.eurpolymj.2023.112076. [DOI] [Google Scholar]
  10. Park E. J., Bae C.. Superacidic porous polymer catalyst and its application in esterification of carboxylic acid. Structural Chemistry. 2017;28(2):493–500. doi: 10.1007/s11224-016-0879-6. [DOI] [Google Scholar]
  11. Thomas R. M., Lemus M. S. L., Ganesh K., Obbard D. B., Sivashanmugam K., Sambasivam G., Yang Y., Lipshutz B. H.. Dehydration in water: solid-supported lipases as green catalysts for esterification. Green Chem. 2024;26(17):9320–9329. doi: 10.1039/D4GC02904F. [DOI] [Google Scholar]
  12. Kobayashi S., Uyama H., Namekawa S.. In vitro biosynthesis of polyesters with isolated enzymes in aqueous systems and organic solvents. Polymer degradation and stability. 1998;59(1–3):195–201. doi: 10.1016/S0141-3910(97)00178-X. [DOI] [Google Scholar]
  13. Müller H., Terholsen H., Godehard S. P., Badenhorst C. P., Bornscheuer U. T.. Recent insights and future perspectives on promiscuous hydrolases/acyltransferases. ACS Catal. 2021;11(24):14906–14915. doi: 10.1021/acscatal.1c04543. [DOI] [Google Scholar]
  14. Pongpamorn P., Kiattisewee C., Kittipanukul N., Jaroensuk J., Trisrivirat D., Maenpuen S., Chaiyen P.. Carboxylic acid reductase can catalyze ester synthesis in aqueous environments. Angew. Chem. 2021;133(11):5813–5817. doi: 10.1002/ange.202013962. [DOI] [PubMed] [Google Scholar]
  15. Biundo A., Stamm A., Gorgoglione R., Syrén P.-O., Curia S., Hauer B., Capriati V., Vitale P., Perna F., Agrimi G.. Regio-and stereoselective biocatalytic hydration of fatty acids from waste cooking oils en route to hydroxy fatty acids and bio-based polyesters. Enzyme Microb. Technol. 2023;163:110164. doi: 10.1016/j.enzmictec.2022.110164. [DOI] [PubMed] [Google Scholar]
  16. Li Z., Shen Y., Li Z.. Ring-opening polymerization of lactones to prepare closed-loop recyclable polyesters. Macromolecules. 2024;57(5):1919–1940. doi: 10.1021/acs.macromol.3c01912. [DOI] [Google Scholar]
  17. Ornati E., Perrard J., Hoffmann T. A., Bonon R., Bruns N.. Bacteria-mediated intracellular radical polymerizations. J. Am. Chem. Soc. 2025;147(11):9496–9504. doi: 10.1021/jacs.4c17257. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Feng J., Zhang J., Ma Y., Feng Y., Wang S., Guo N., Wang H., Wang P., Jiménez-Bonilla P., Gu Y.. Renewable fatty acid ester production in Clostridium. Nat. Commun. 2021;12(1):4368. doi: 10.1038/s41467-021-24038-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Menendez-Bravo S., Comba S., Gramajo H., Arabolaza A.. Metabolic engineering of microorganisms for the production of structurally diverse esters. Applied microbiology and biotechnology. 2017;101(8):3043–3053. doi: 10.1007/s00253-017-8179-7. [DOI] [PubMed] [Google Scholar]
  20. Choi S. Y., Cho I. J., Lee Y., Kim Y. J., Kim K. J., Lee S. Y.. Microbial polyhydroxyalkanoates and nonnatural polyesters. Adv. Mater. 2020;32(35):1907138. doi: 10.1002/adma.202070264. [DOI] [PubMed] [Google Scholar]
  21. Chae T. U., Choi S. Y., Ahn D.-H., Jang W. D., Jeong H., Shin J., Lee S. Y.. Biosynthesis of poly (ester amide) s in engineered Escherichia coli. Nat. Chem. Biol. 2025;21:1171. doi: 10.1038/s41589-025-01842-2. [DOI] [PubMed] [Google Scholar]
  22. Satoh K., Kawakami T., Isobe N., Pasquier L., Tomita H., Zinn M., Matsumoto K. i.. Versatile aliphatic polyester biosynthesis system for producing random and block copolymers composed of 2-, 3-, 4-, 5-, and 6-hydroxyalkanoates using the sequence-regulating polyhydroxyalkanoate synthase PhaCAR. Microbial Cell Factories. 2022;21(1):84. doi: 10.1186/s12934-022-01811-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. de Azevedo Souza C., Kim S. S., Koch S., Kienow L., Schneider K., McKim S. M., Haughn G. W., Kombrink E., Douglas C. J.. A novel fatty Acyl-CoA Synthetase is required for pollen development and sporopollenin biosynthesis in Arabidopsis. Plant Cell. 2009;21(2):507–525. doi: 10.1105/tpc.108.062513. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Li W., Gu S., Fleming J., Bi L.. Crystal structure of FadD32, an enzyme essential for mycolic acid biosynthesis in mycobacteria. Sci. Rep. 2015;5(1):15493. doi: 10.1038/srep15493. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Li S., Qu Y.. Structural study of medium-long chain fatty acyl-CoA ligase FadD8 from Mycobacterium tuberculosis. Biochem. Biophys. Res. Commun. 2023;672:65–71. doi: 10.1016/j.bbrc.2023.06.024. [DOI] [PubMed] [Google Scholar]
  26. Kumari P., Sharma J., Singh A. K., Pandey A. K., Yusuf F., Kumar S., Gaur N. A.. Tailored designing of a diploid S. cerevisiae natural isolate for increased production of fatty acid ethyl ester. Chemical Engineering Journal. 2023;453:139852. doi: 10.1016/j.cej.2022.139852. [DOI] [Google Scholar]
  27. Miklaszewska M., Dittrich-Domergue F., Banaś A., Domergue F.. Wax synthase MhWS2 from Marinobacter hydrocarbonoclasticus: substrate specificity and biotechnological potential for wax ester production. Appl. Microbiol. Biotechnol. 2018;102(9):4063–4074. doi: 10.1007/s00253-018-8878-8. [DOI] [PubMed] [Google Scholar]
  28. Valle-Rodríguez J. O., Siewers V., Nielsen J., Shi S.. Directed evolution of a wax ester synthase for production of fatty acid ethyl esters in Saccharomyces cerevisiae. Appl. Microbiol. Biotechnol. 2023;107(9):2921–2932. doi: 10.1007/s00253-023-12466-8. [DOI] [PubMed] [Google Scholar]
  29. Aarsen C. V., Liguori A., Mattsson R., Sipponen M. H., Hakkarainen M.. Designed to degrade: tailoring polyesters for circularity. Chem. Rev. 2024;124(13):8473–8515. doi: 10.1021/acs.chemrev.4c00032. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Borrelli M. A., Warunek J. J., Ravikumar T., Balmert S. C., Little S. R.. End group chemistry modulates physical properties and biomolecule release from biodegradable polyesters. J. Mater. Chem. B. 2025;13(34):10621–10634. doi: 10.1039/D5TB00816F. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Somboon K., Doble A., Bulmer D., Baslé A., Khalid S., van den Berg B.. Uptake of monoaromatic hydrocarbons during biodegradation by FadL channel-mediated lateral diffusion. Nat. Commun. 2020;11(1):6331. doi: 10.1038/s41467-020-20126-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Hearn E. M., Patel D. R., Lepore B. W., Indic M., Van den Berg B.. Transmembrane passage of hydrophobic compounds through a protein channel wall. Nature. 2009;458(7236):367–370. doi: 10.1038/nature07678. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Dobovišek, A. ; Blaževič, T. ; Kralj, S. ; Fajmut, A. . Enzyme cascade to enzyme complex phase-transition-like transformation studied by the maximum entropy production principle. Cell Reports Physical Science 2025, 6 (2), 102400. 10.1016/j.xcrp.2024.102400 [DOI] [Google Scholar]
  34. Zhang Y., Fernie A. R.. Dynamically regulating metabolic fluxes with synthetic metabolons. Trends Biotechnol. 2022;40(9):1019–1020. doi: 10.1016/j.tibtech.2022.06.005. [DOI] [PubMed] [Google Scholar]
  35. Kummer M. J., Lee Y. S., Yuan M., Alkotaini B., Zhao J., Blumenthal E., Minteer S. D.. Substrate channeling by a rationally designed fusion protein in a biocatalytic cascade. Jacs Au. 2021;1(8):1187–1197. doi: 10.1021/jacsau.1c00180. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Vollheyde K., Kühnel K., Lambrecht F., Kawelke S., Herrfurth C., Feussner I.. Crystal structure of the bifunctional wax synthase 1 from Acinetobacter baylyi suggests a conformational change upon substrate binding and formation of additional substrate binding sites. ACS Catal. 2022;12(15):9753–9765. doi: 10.1021/acscatal.2c01712. [DOI] [Google Scholar]
  37. Petronikolou N., Nair S. K.. Structural and biochemical studies of a biocatalyst for the enzymatic production of wax esters. ACS catalysis. 2018;8(7):6334–6344. doi: 10.1021/acscatal.8b00787. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Viora L., Combeau M., Pucci M. F., Perrin D., Liotier P.-J., Bouvard J.-L., Combeaud C.. A comparative study on crystallisation for virgin and recycled polyethylene terephthalate (PET): multiscale effects on physico-mechanical properties. Polymers. 2023;15(23):4613. doi: 10.3390/polym15234613. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Khare G., Gupta V., Gupta R. K., Gupta R., Bhat R., Tyagi A. K.. Dissecting the role of critical residues and substrate preference of a Fatty Acyl-CoA Synthetase (FadD13) of Mycobacterium tuberculosis. PLoS One. 2009;4(12):e8387. doi: 10.1371/journal.pone.0008387. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Feng S., Xu C., Yang K., Wang H., Fan H., Liao M.. Either fadD1 or fadD2, which encode acyl-CoA synthetase, is essential for the survival of Haemophilus parasuis SC096. Frontiers in Cellular and Infection Microbiology. 2017;7:72. doi: 10.3389/fcimb.2017.00072. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Shimshoni J. A., Basselin M., Li L. O., Coleman R. A., Rapoport S. I., Modi H. R.. Valproate uncompetitively inhibits arachidonic acid acylation by rat acyl-CoA synthetase 4: relevance to valproate’s efficacy against bipolar disorder. Biochimica et Biophysica Acta (BBA)-Molecular and Cell Biology of Lipids. 2011;1811(3):163–169. doi: 10.1016/j.bbalip.2010.12.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Concilio M., Sulley G. S., Vidal F., Brown S., Williams C. K.. Precise Carboxylic Acid-Functionalized Polyesters in Reprocessable Vitrimers. J. Am. Chem. Soc. 2025;147(8):6492–6502. doi: 10.1021/jacs.4c14032. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Kost B., Basko M., Bednarek M., Socka M., Kopka B., Łapienis G., Biela T., Kubisa P., Brzeziński M.. The influence of the functional end groups on the properties of polylactide-based materials. Prog. Polym. Sci. 2022;130:101556. doi: 10.1016/j.progpolymsci.2022.101556. [DOI] [Google Scholar]
  44. Slomkowski S., Basinska T., Gadzinowski M., Mickiewicz D.. Polyesters and polyester nano-and microcarriers for drug delivery. Polymers. 2024;16(17):2503. doi: 10.3390/polym16172503. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Feng P., Jia J., Liu M., Peng S., Zhao Z., Shuai C.. Degradation mechanisms and acceleration strategies of poly (lactic acid) scaffold for bone regeneration. Materials & Design. 2021;210:110066. doi: 10.1016/j.matdes.2021.110066. [DOI] [Google Scholar]
  46. Woodruff M. A., Hutmacher D. W.. The return of a forgotten polymerPolycaprolactone in the 21st century. Prog. Polym. Sci. 2010;35(10):1217–1256. doi: 10.1016/j.progpolymsci.2010.04.002. [DOI] [Google Scholar]
  47. Versypt A. N. F., Pack D. W., Braatz R. D.. Mathematical modeling of drug delivery from autocatalytically degradable PLGA microspheresA review. Journal of controlled release. 2013;165(1):29–37. doi: 10.1016/j.jconrel.2012.10.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Miu D.-M., Eremia M. C., Moscovici M.. Polyhydroxyalkanoates (PHAs) as biomaterials in tissue engineering: production, isolation, characterization. Materials. 2022;15(4):1410. doi: 10.3390/ma15041410. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Cai J., Liu C., Cai M., Zhu J., Zuo F., Hsiao B. S., Gross R. A.. Effects of molecular weight on poly (ω-pentadecalactone) mechanical and thermal properties. Polymer. 2010;51(5):1088–1099. doi: 10.1016/j.polymer.2010.01.007. [DOI] [Google Scholar]
  50. Zhao X., Hu H., Wang X., Yu X., Zhou W., Peng S.. Super tough poly (lactic acid) blends: A comprehensive review. RSC Adv. 2020;10(22):13316–13368. doi: 10.1039/D0RA01801E. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Lo Re G., Spinella S., Boujemaoui A., Vilaseca F., Larsson P. T., Adås F., Berglund L. A.. Poly (ε-caprolactone) biocomposites based on acetylated cellulose fibers and wet compounding for improved mechanical performance. ACS Sustainable Chem. Eng. 2018;6(5):6753–6760. doi: 10.1021/acssuschemeng.8b00551. [DOI] [Google Scholar]
  52. Dworakowska S., Le Coz C., Chollet G., Grau E., Cramail H.. Cross-linking of polyesters based on fatty acids. European Journal of Lipid Science and Technology. 2019;121(11):1900264. doi: 10.1002/ejlt.201900264. [DOI] [Google Scholar]
  53. Czifrák K., Lakatos C., Árpád Kordován M., Nagy L., Daróczi L., Zsuga M., Kéki S.. Block copolymers of poly (ω-pentadecalactone) in segmented polyurethanes: novel biodegradable shape memory polyurethanes. Polymers. 2020;12(9):1928. doi: 10.3390/polym12091928. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

au5c01683_si_001.pdf (1.6MB, pdf)

Articles from JACS Au are provided here courtesy of American Chemical Society

RESOURCES