Telogen and taxogen control of enzymatic degradation of polymers formed by TBRTThe TBRT of dimethacrylate monomers yields a branched polyester backbone with pendant telogen-derived groups, Fig. 1B. The enzymatic degradation of polyester backbones is not, however, a given feature of such polymers; substitution neighbouring ester carbonyls may yield poor degradation behaviour in the presence of water-borne enzymes, especially for hydrophobic polymers. In recent years, the isolation of new enzymes capable of degrading p(ethylene terephthalate), PET, from marine microorganisms has opened new avenues of investigation for bioremediation of waste27,28; however, we selected to study the action of esterase, lipase and protease enzymes that are relatively commonly used in industrial products or processes.Lauryl dimethacrylate (LDMA, 1) is a commercially available bifunctional monomer that has been employed previously as an MVT in TBRT studies22. The synthesis of three TBRT homopolymers containing the LDMA backbone was readily accomplished using either 1-dodecanethiol (DDT, 3), cyclohexanethiol (CHT, 4), or 1-thioglycerol (TG, 5), Fig. 2A, leading to high molecular weight polymers varying only in their side chain chemistry, Fig. 2B (Supplementary Figs. S1-13; Table S1). The synthesis of the three polymers, p(DDT-LDMA), p(CHT-LDMA), and p(TG-LDMA) was studied using varying [MVT]0/[Telogen]0 molar ratios within the reactions and optimised to provide samples with similar molecular weights and dispersities (Ð). As seen previously, high molecular weight and soluble p(DDT-LDMA) was synthesised at an [MVT]0/[DDT]0 molar ratio of 0.54, yielding samples with a number average molecular weight (Mn) of 19,480 g mol–1 and a weight average molecular weight (Mw) of 249,900 g mol–1 using 2,2’-azobis(isobutyronitrile) (AIBN) as the free radical source and 50 wt% solids concentration in ethyl acetate at 70 °C. Due to the relatively poor chain transfer from the secondary thiol of CHT, the highest [MVT]0/[Telogen]0 molar ratio achieved yielding soluble polymer was 0.34 (p(CHT-LDMA): Mn = 10,490 g mol–1 ; Mw = 561,090 g mol–1); similarly, reactions with [MVT]0/[Telogen]0 molar ratios >0.37 were unable to control the TBRT reaction when using TG, but soluble p(TG-LDMA) homopolymer was obtainable at this ratio (Mn = 21,220 g mol–1; Mw = 893,850 g mol–1).Fig. 2: Starting materials and resulting TBRT polymers.A Multi-vinyl taxogens: lauryl dimethacrylate, 1, and bis-HEMA glutarate, 2. Telogens: 1-dodecanethiol, 3, cyclohexanethiol, 4, and 1-thioglycerol, 5. B\ Resulting branched polymers derived from lauryl dimethacrylate polymerisation with each telogen (6, 7, and 8), and the polymerisation of bis-HEMA glutarate with each telogen (9, 10, and 11).In all cases, complete consumption of vinyl groups was observed within the 1H nuclear magnetic resonance (NMR) spectra of the crude reaction mixtures. As we have reported, the [MVT]F/[Telogen]F ratio within the purified polymers is highly indictive of the presence of cycles within these branched polymers24, but each of these polymers showed values close to unity, indicating a near ideal branched structure (p(DDT-LDMA) = 1.01; p(CHT-LDMA) = 1.06; p(TG-LDMA) = 1.04). The polymers were hydrophobic, fully soluble in ethyl acetate, and exhibited glass transition temperatures (Tg) which varied with telogen, as seen in our previous reports of TBRT materials (p(TG-LDMA = -67 °C; p(DDT-LDMA) = –24 °C; p(CHT-LDMA) = –6 °C).An initial degradation screen (Supplementary Figs. S14-15, Equation S1) of the three LDMA-derived polymers utilised pig liver esterase, proteinase K, bovine pancreatic protease, porcine pancreatic lipase, and Pseudomonas lipase; the absence of enzyme was employed as a negative control and a p(ε-caprolactone) sample chosen as a positive control (p(ε-CL)). Mass loss was used to establish the extent of enzyme catalysed hydrolysis and indicate susceptibility for degradation. Each polymer was placed in a vial with an aqueous solution of enzyme in phosphate buffered saline (PBS; pH = 7.2) and incubated for a period of 4 weeks at 37 °C with each enzyme being regularly replenished during the study. For the p(ε-CL) control polymer, there was no discernible difference in polymer degradation in the absence of enzyme and when using pig liver esterase, proteinase K, and bovine pancreatic protease, with no mass loss observed during this study. This was also seen for all enzyme and control conditions when studying both p(DDT-LDMA) and p(CHT-LDMA), Fig. 3A.Fig. 3: Evaluation of enzymatic hydrolysis of TBRT polymers using five enzymes (p(ε-caprolactone) as linear polymer standard).A Comparative hydrolysis of polymers synthesised using lauryl dimethacrylate; B Comparative hydrolysis of polymers synthesised using bis-HEMA glutarate; C Structure of bis-HEMA terephthalate, 12, branched polymer resulting from the TBRT of bis-HEMA terephthalate with 1-thioglycerol, 13, and comparison of mass loss via enzymatic hydrolysis of 1-thioglycerol containing TBRT polymers when exposed to and pseudomonas lipase for 28 days.Significant mass loss was observed when p(ε-CL) was treated with porcine pancreatic lipase and Pseudomonas lipase (Supplementary Fig. S15), but surprisingly, p(TG-LDMA) showed susceptibility to hydrolysis across all enzymes studied under these conditions, ranging from 10.7–19.5% mass loss. Given that the structural variation of these TBRT polymers is purely derived from telogen (pendant group), variation of the susceptibility towards degradation by the broad range of enzymes appears to be telogen driven, presenting a new structural feature to induce degradation within high molecular weight polymers, Fig. 3A.To study the impact of MVT chemistry on enzyme degradation, a novel dimethacrylate MVT was synthesised by coupling two 2-hydroxethyl methacrylate (HEMA) monomers using glutaryl chloride to form bis-HEMA glutarate (BHEMAG, 2), Fig. 2A (Supplementary Fig. S16–18; Table S2). LDMA and BHEMAG have just a single atom difference in chain length and were therefore expected to show similar reactivity under TBRT conditions22. The introduction of two unhindered esters within the MVT, and ultimately the nominal repeat unit within the resulting polymers, Fig. 2B, was expected to enable enzymatic degradation. This MVT was subjected to TBRT using the same conditions employed for LDMA and the same series of DDT, CHT and TG telogens (Supplementary Figs. S19–30; Table S3). The outcomes were remarkably similar, with soluble branched homopolymers synthesised at [MVT]0/[Telogen]0 molar ratios of 0.35 using CHT (p(CHT-BHEMAG): Mn = 12,200 g mol–1 ; Mw = 91,700 g mol–1), 0.40 in the presence of TG (p(TG-BHEMAG): Mn = 14,200 g mol–1; Mw = 489,400 g mol–1), and 0.55 for reactions containing DDT (p(DDT-BHEMAG): Mn = 19,620 g mol–1; Mw = 234,100 g mol–1). Values for Tg were established as –58 °C, –42 °C and -4 °C for p(TG-BHEMAG), p(DDT-BHEMAG), and p(CHT-BHEMAG), respectively, following the same telogen-derived trend as the LDMA-derived TBRT polymers. The [MVT]F/[Telogen]F molar ratios within the final purified polymer samples were determined as p(DDT-BHEMAG) = 1.03, p(CHT-BHEMAG) = 1.01, and p(TG-BHEMAG) = 1.07, also falling within a very similar range as the polymers employing LDMA.When subjected to the range of enzyme degradation studies described above, all of the BHEMAG-derived polymers showed considerably enhanced susceptibility to hydrolysis, Fig. 3B (Supplementary Fig. S31). p(DDT-BHEMAG) and p(CHT-BHEMAG) exhibited very similar rates of enzyme catalyzed hydrolysis across all enzymes, with mass losses ranging from 7.3–14.4%, except when proteinase K was used. Importantly, p(TG-BHEMAG) again displayed the telogen-directed enhanced susceptibility to enzymatic cleavage, leading to considerable mass losses (39.5–78.8%) across all enzymes. These results indicate that the branched TBRT polymers derived from the BHEMAG MVT are more widely susceptible to different enzymes than p(ε-CL), with p(TG-BHEMAG) outperforming this established degradable polymer in the presence of pig liver esterase, proteinase K, and bovine pancreatic protease, and rivalling degradation when subjected to Pseudomonas lipase under these conditions.Telogen (pendant group)-mediated enzymatic degradation has not been previously reported for branched polyesters and the impact of TG may be interpreted as enabling hydration of the polymer and, therefore, providing enzyme access to ester functional groups. Nevertheless, when BHEMAG is used within the backbone of the TBRT polymer, access is clearly still available in the presence of CHT and DDT, both of which are hydrophobic and appear to limit, but not fully prevent, the ability of each enzyme to reach the ester degradation sites (with the exception of proteinase K). The water-solubility of the hydrolysis products may also vary with varying telogen and higher degrees of hydrolysis may be required to yield fully soluble byproducts.The reason for the enhanced susceptivity of the backbone esters for these TBRT polymers, when compared to p(ε-CL), is not entirely clear. Although the physical form of the polymer samples may play a role (p(ε-CL): film; TBRT polymers: low Tg semi-solid), it is unlikely that the difference is solely based on this given the comparative differences within the series of TBRT materials. It is also worth emphasising that cleavage of the more hindered backbone esters derived purely from the methacrylate functionality is occurring within p(TG-LDMA), as no additional ester groups are present within the polymer backbone. To study the impact of ester chemistry in more detail, HEMA was reacted with terepthaloyl chloride to form bis-HEMA terephthalate (BHEMAT; 12), Fig. 3C, which, again, closely resembled the molecular dimensions of LDMA and BHEMAG (Supplementary Fig. S32-33; Table S4). TBRT polymerisation of BHEMAT with TG yielded soluble branched polymers up to a [MVT]0/[Telogen]0 molar ratio of 0.42 (p(TG-BHEMAT); 13: Mn = 18,600 g mol–1; Mw = 236,900 g mol–1), and after purification the [MVT]f/[Telogen]f molar ratio within the final polymer was measured as 1.00, indicating an ideal branched structure with few, if any, cyclic substructures (Supplementary Fig. S34–36; Table S5).When subjected to Pseudomonas lipase under the same conditions described above, it was clear that p(TG-BHEMAT) showed no discernible mass loss over the 28-day period, Fig. 3C. Despite the presence of the additional backbone ester functionality, p(TG-BHEMAT) appears to not be susceptible to enzymatic cleavage through the same mechanisms acting on the other polymer samples studied. The additional impact of the inclusion of aromatic ester groups clearly indicates that TBRT offers the potential to mediate enzyme hydrolysis through telogen and taxogen selection/design, leading to polymers that are resistant or highly susceptible to degradation under these conditions.Controlling enzymatic degradation through TBRT copolymer designAs mentioned above, TBRT offers additional synthetic strategies for copolymer synthesis. Through mixing telogens, for example, TBRT polymers can be generated with a statistical mixture of telogen-derived side chains that mimic statistical copolymers synthesised by chain-growth polymerisation approaches. Additionally, by mixing MVTs it is possible to generate statistical copolymers that vary in their backbone chemistry, which are analogous to copolymers synthesised using mixed A2 or B2 monomers using step-growth polymerisation chemistries26.Given the impact of telogens and MVT structures on the enzymatic cleavage of the TBRT polymers, a series of copolymer structures was synthesised in order to investigate the potential to fine tune the observed degradation profiles. Within these studies, Pseudomonas lipase was selected as the sole study enzyme given the variable degradation seen within the p(Telogen-BHEMAG) polymers and the apparent enzymatic hydrolysis susceptibility of p(TG-LDMA). The two telogens selected for mixed telogen copolymer synthesis were DDT and TG, whilst LDMA and BHEMAG were the obvious choices for mixed MVT copolymer synthesis, Fig. 4A.Fig. 4: Schematic representation of the synthesis of TBRT copolymer structures.A Mixed telogen copolymer and structure of resulting polymer derived from the TBRT of bis-HEMA glutarate with a 1:1 mixture of 1-dodecanethiol and 1-thioglycerol; B) Mixed multi-vinyl taxogen copolymer and structure of resulting polymer derived from the TBRT of a 1:1 mixture of lauryl dimethacrylate and bis-HEMA glutarate with 1-thioglycerol.Mixed telogen copolymers with a range of compositions were synthesised simply by mixing TG and DDT at varying concentrations of the initial feedstock ratio within the TBRT reaction of BHEMAG, under the same conditions described for homopolymer synthesis (Supplementary Fig. S37, 38). To compensate for the different chain transfer coefficients of the telogens, the initial [MVT]0/[Telogen]0 molar ratios were varied to ensure high molecular weight soluble polymer was formed in each case with full consumption of vinyl functionality, Table 1. After purification, the mixed telogen copolymer samples varied in Mw from 54,100 g mol–1 to 636,800 g mol–1 as the composition of the copolymers changed from p([DDT-BHEMAG]25-stat-[TG-BHEMAG]75) to p([DDT-BHEMAG]75-stat-[TG-BHEMAG]25), with no indication of additional cyclisation when compared to the homopolymer syntheses.Table 1 1H nuclear magnetic resonance (NMR) and triple detection size exclusion chromatography (TD-SEC) analyses of mixed telogen and mixed multi-vinyl taxogen copolymers using synthesised using TBRTSimilarly, when polymerising a mixed MVT feedstock of LDMA and BHEMAG using TG as the telogen under TBRT conditions, the formation of copolymers ranging from p([TG-LDMA]75-stat-[TG-BHEMAG]25) to p([TG-LDMA]25-stat-[TG-BHEMAG]75), Fig. 4B, was accomplished as previously, using [MVT]0/[Telogen]0 molar ratios within the range 0.36–0.40. This is almost identical to the conditions used for the synthesis of the two homopolymers p(TG-LDMA) and p(TG-BHEMAG), as might be expected (Supplementary Fig. S39-40). In our previous studies of mixed telogen and mixed taxogen branched copolymer synthesis using TBRT, there were clear indications that telogen and taxogen incorporation was relatively homogeneous throughout the branched polymer structures26. This is not surprising given the statistical nature of intermolecular branching reactions at high vinyl group consumption.The enzymatic degradation conditions used previously were again employed to study the impact of Pseudomonas lipase on the enzyme catalysed hydrolysis of different copolymers over a 28-day incubation period but including multiple repeats (n = 3), Fig. 5A. Importantly, the extremes of each copolymer series are different when considering the two studies. The homopolymers for the mixed-telogen statistical copolymers are p(DDT-BHEMAG) and p(TG-BHEMAG) with variation of DDT/TG molar ratios within the systematically varying structures; the homopolymer extremes for the mixed-MVT study are p(TG-LDMA) and p(TG-BHEMAG) with LDMA/BHEMAG varying within the copolymer series.Fig. 5: Evaluation of enzymatic hydrolysis of TBRT copolymer series using and pseudomonas lipase.A Comparison of weight loss for mixed telogen copolymers (solid blue bars) and mixed multi-vinyl taxogen (hashed blue bars) after 28 days exposure to enzyme. TBRT homopolymers shown for each comparative series are p(DDT-BHEMAG) and p(TG-BHEMAG) (mixed telogen copolymer series), and p(TG-LDMA) and p(TG-BHEMAG) (mixed multi-vinyl taxogen copolymer series); B Monitoring mass loss for the mixed telogen copolymer series over 42 days; C Monitoring mass loss for the mixed multi-vinyl taxogen copolymer series over 42 days.The mixed telogen copolymers clearly demonstrated a variation in enzymatic degradation that correlates to increasing TG within the statistical copolymer composition, Fig. 5A. The inclusion of 25 mol% of TG had a relatively minor impact with a small decrease in the measured mass remaining in the sample (from 86.2% to 82.4%) after 28 days. The formation of p([DDT-BHEMAG]50-stat-[TG-BHEMAG]50) led to a larger decrease in mass during the experiment, to 73.1% of the starting sample, with p([DDT-BHEMAG]25-stat-[TG-BHEMAG]75) leading to nearly a 50% decrease over this timescale, Fig. 5A.The inclusion of 25 mol% of BHEMAG into the mixed MVT structures was more impactful than a 25 mol% TG addition within the mixed telogen copolymers; the p([TG-LDMA]75-stat-[TG-BHEMAG]25) showed almost double the mass loss observed in the p(TG-LDMA) homopolymer. Further inclusion of BHEMAG into the backbone of the statistical copolymers led to increasing mass losses. In both cases of mixed telogen and mixed MVT statistical copolymers, it is expected that the inclusion of more TG or increasing BHEMAG above the respective 75 mol% studied here, will have a dramatic impact on enzymatic degradation, as the mass remaining after degrading p(TG-BHEMAG) under these conditions is approximately 40% of these values. Fig.5A (Supplementary Tables S6–8).The two statistical copolymer series were also subjected to a 42-day degradation study using Pseudomonas lipase, with samples taken at regular intervals to investigate the different hydrolysis kinetics within the two copolymer strategies, Fig. 5B, C (Supplementary Tables S6-8). Within the mixed telogen copolymer series, the p(DDT-BHEMAG) homopolymer showed an initial steady decrease in mass until 14 days, after which the sample mass stabilised, suggesting no further degradation over the following 28 days, Fig. 5B.This may suggest that regions within the p(DDT-BHEMAG) homopolymer sample are highly accessible to the enzyme and once these have been degraded, access to the branched polymer architecture becomes increasingly difficult. The most intuitive areas that may provide accessibility are the backbone esters near the DP1 terminal groups, although it is also possible that the lower molecular weight species within the distribution are responsible for this initial mass loss. The degradation profile of p([DDT-BHEMAG]75-stat-[TG-BHEMAG]25) appears to follow a similar trajectory to the p(DDT-BHEMAG) homopolymer but rather than a plateau, the degradation appears to continue slowly from 14 days through to the 42-day endpoint (approximately additional 9.5% mass loss during this time), Fig. 5B. The equimolar mixed telogen copolymer p([DDT-BHEMAG]50-stat-[TG-BHEMAG]50) displayed a much steeper decline in mass during the first 7 days of the degradation study and a similar slowing of mass loss for the remainder of the study, whilst leading to a larger overall mass loss. p([DDT-BHEMAG]25-stat-[TG-BHEMAG]75) led to a considerable change in degradation within the first 7 days of the study, matching the behaviour of the p(TG-BHEMAG) homopolymer. The slowing of degradation from 7 days through to the study endpoint was seen for this copolymer but was absent from the homopolymer sample, presumably due to decreasing access to the backbone as sections bearing TG side chains are cleaved, leaving DDT side chains as the predominant chemistry. The p(TG-BHEMAG) homopolymer showed steady mass loss across the 42-day study with only 14.6% of the initial sample mass remaining at the end of the study and no sign of the onset of a plateau at this point, Fig. 5B.The mixed MVT statistical copolymers showed similar behaviour, as the LDMA within the backbone was decreased from 100 mol% through to 0 mol%. Interestingly, the degradation of p(TG-LDMA) appears to be faster than p(DDT-BHEMAG) over the first 14 days, but again a plateau was observed where no further degradation was measured over the next 28 days, Fig. 5C. Inclusion of 25 mol% BHEMAG into the statistical mixed MVT copolymer led to a noticeably faster decrease in sample mass than the inclusion of 25 mol% of TG within the mixed telogen samples (7-day mass loss: p([TG-LDMA]75-stat-[TG-BHEMAG]25) = 17.6%; p([DDT-BHEMAG]75-stat-[TG-BHEMAG]25) = 8.3%). In general, the mixed MVT statistical copolymers led to a greater mass loss than the corresponding mixed telogen samples, probably due to the hydrophilic nature of the single telogen used and the formation of soluble cleavage products even when LDMA segments had not fully degraded.Characterisation of enzymatic degradation products and mechanism elucidationIdentifying the byproducts of polymer degradation can be extremely challenging; however, the TBRT polymers derived from BHEMAG should preferentially undergo enzymatic cleavage at the glutaric acid residues within the polyester backbone and yield the distribution of telomers that act as structural subunits (linear, branching and terminal) within the high molecular weight branched polyester products14,21. When using BHEMAG, the telomer distribution should, therefore, resemble the products of a conventional telomerisation of HEMA.To establish the conditions for characterising the theoretical degradation products, a HEMA telomerisation was carried out using TG as the telogen and under conditions similar to those used to form p(TG-BHEMAG), Figure 6Ai&B (Supplementary Figs. 41–43, Table S9). Time of flight matrix assisted laser desorption ionisation (MALDI-TOF) mass spectrometry was used to study the p(HEMA)-TG linear telomerisation, and a distribution of ions was observed, with a clear repeating pattern of species separated by the mass of additional HEMA repeat units. Figure 6Aii.Fig. 6: Elucidating the mechanism of enzymatic hydrolysis.A i) Schematic representation of the linear telomerisation of 2-hydroxyethyl methacrylate using 1-thioglycerol, and ii) the resulting MALDI-TOF mass spectrum of the telomer sample; B Structures of the telomers formed from the telomerisation of 2-hydroxyethyl methacrylate (DP1-4 shown); and C) i) Schematic representation of p(TG-BHEMAG), ii) the distribution of telomers that may be formed (DP1-4 shown) and how they mirror the telomer distribution from the telomerisation of 2-hydroxyethyl methacrylate, and iii) the MALDI-TOF mass spectrum of the byproducts of enzymatic hydrolysis showing telomer substructures with masses < 1600 g mol-1.The MALDI-TOF analysis of the isolated byproducts of the enzymatic hydrolysis of p(TG-BHEMAG) after exposure to Pseudomonas lipase, Figure 6Ci-ii, showed a distribution of species complicated by the oxidation of the thioether units within the polymer structure, Figure 6Ciii (Supplementary Figs. S44–46, Table S10). It is known that under mass spectrometry conditions, different degrees of oxidation may be seen within sulphur containing species29,30. Close examination of the species present clearly showed repeating patterns of low molecular weight materials separated by the HEMA repeat unit mass and equating to TG-terminated telomers with chain lengths up to 11 HEMA units (DP11). Under these conditions, no species with a mass of > 1600 g mol–1 were observed. Identification of DP1 structures was accomplished using a porphyrin-based matrix and the intensity of detected telomers > DP4, was significantly reduced, potentially due to ionisation differences.
