The In Vitro Enzymatic Degradation of Cross-Linked Poly(trimethylene carbonate) Networks

The in vitro enzymatic degradation of cross-linked poly(trimethylene carbonate) networks (PTMC-Ns) was performed in lipase solutions at 37 °C, and the effect of the initial molecular weight and cross-linker amount as well as the cross-linker type on the degradation rate of PTMC-Ns was investigated. Due to their denser structure and more hydrophobic surface as well as the higher glass transition temperature, a slower degradation rate was seen for PTMC-Ns with high initial molecular weight at a given cross-linker amount. Similar results could be observed as the cross-linker amount increased, and cross-linker type also influenced the degradation rate of PTMC-Ns. Furthermore, the enzymatic degradation of PTMC-Ns was accelerated by the surfactants role of lipase via surface erosion mechanism, the enzymatic degradation rate was higher than that of hydrolysis case. The results indicated that PTMC-Ns were promising candidates for clinical subcutaneous implants, especially due to their tunable degradation rate and enhanced form-stability as well as no acidic degradation products.


Introduction
As one of the most important biodegradable polymers, poly(trimethylene carbonate) (PTMC) has attracted considerable attention in recent years due to its favorable characteristics, such as well-documented biocompatibility, excellent biodegradability and rubber-like properties [1][2][3]. On top of that, the degradation products of PTMC are not acidic, which is much better than polyesters to avoid inflammation [4][5][6][7] led by acidic degradation products. Therefore, PTMC has been one of the promising candidate materials for biomedical and pharmaceutical applications including drug delivery systems [8][9][10][11][12], nerve guides [13][14][15] and temporary three-dimensional (3D) scaffolds in tissue engineering [16][17][18][19][20]. However, further study showed poor compatibility between degradation rate and form-stability of PTMC during in vivo degradation [21]; for example, although the degradation rate was slow, the form-stability for low-molecular-weight PTMC was poor; while the form-stability was enhanced, high-molecular-weight PTMC had an undesired faster degradation rate in vivo. It would greatly hinder the extensive application of PTMC in biomedical fields. The exact combination of good form-stability and low degradation rate of PTMC needs to be achieved urgently.

Materials
TMC was purchased from Daigang Biomaterial Co., Ltd (Jinan, China), recrystallized twice from acetic ether and dried for 24 h at 37 • C under reduced pressure before polymerization. Stannous octoate [Sn(Oct) 2 ] (95%) and lipase solutions (from Thermomyces lanuginosus, ≥100,000 U/g) were used as received from Sigma-Aldrich. All other solvents and reagents used were of analytical grades and purified by standard methods.

Measurements
The initial molecular weight (Mn) was determined by GPC (Waters, Milford, MA, USA) with a Waters Model 1515 isocratic high-performance liquid chromatography pump, a Waters Model 2414 differential refractive index detector and a Waters Styragel HT4 chromatographic column using THF as the solvent and a flow rate of 1 mL/min at 35 • C. The water contact angles of the different PTMC-Ns were determined using a contact angle-measuring instrument (Kruss, Hamburg, Germany) equipped with a video measuring system was used to determine static contact angles of sessile drops (each based on 6 different films) at room temperature. The phase transition temperature of PTMC-Ns was determined with a Netzsch DSC 200 F3 (Netzsch, Serb, Germany) equipped with a liquid nitrogen cooling system. The measured temperature range was between −100 and 100 • C, and the heating rate was 10 • C/min under nitrogen atmosphere. The thermal stability of the polymers under nitrogen atmosphere used a Netzsch TGA 209 F3 (Netzsch, Serb, Germany) from room temperature to 550 • C Polymers 2017, 9,605 3 of 16 and a heating rate of 10 • C/min. The mechanical properties of the networks were measured using an Instron 1121 universal testing machine (Instron, Boston, MA, USA) with a crosshead speed of 50 mm/min in accordance with GB/T 1040. 1-2006. The tests were done on triplicate samples, and the results were presented as an average of the tested samples. The scanning electron microscopy (SEM) images were obtained with a XL30ESEM-FEG microscope (FEI-Philips, Amsterdam, The Netherlands). The sample surfaces were coated with Au to avoid charging. The pH of the media before and after degradation was monitored with a combination microelectrode (InLabMicro™, Mettler-Toledo, Zurich, Switzerland,) and a digital pH meter (FiveEasy™, Mettler-Toledo, Zurich, Switzerland).

Fabrication of PTMC-Ns Rods
The PTMC-Ns rods were directly synthesized via bulk ring-opening polymerization reaction of TMC with cross-linker bis(trimethylene carbonate) (BTB) or bis(ε-caprolactone) (BCP) in the presence of Sn(Oct) 2 as the catalyst. Under nitrogen atmosphere, dried TMC and cross-linker with anhydrous toluene solution of Sn(Oct) 2 were added to glass ampoules containing clean and dry Teflon tubes with an inner diameter of 3 mm; The toluene was removed afterwards by evacuation. The ampoules were purged with dry nitrogen, heat-sealed under vacuum and vigorously shaken to obtain a homogeneous mixture of the monomers and the catalyst. Then, the ampoules were conditioned in an oil bath pre-heated to 130 ± 2 • C for 24 h. After the reaction, the ampoules were quenched at room temperature, and the PTMC-Ns rods were easily removed from the Teflon tubes and used without any further purification.
To determine the gel percentage and swelling degree of the resulting PTMC-Ns, a piece of network sample was weighed (w 0 ) and kept in a sealed flask containing chloroform. At regular intervals, the sample was removed and the excess solvent was blotted. The sample was then weighed and returned to the medium. This procedure was continued until a constant weight was attained (w). The gel percentages were determined gravimetrically after the swollen gels were dried to constant weights (w 1 ). The gel percentage and swelling degree were calculated according to Equations (1) and (2), respectively.
where w 0 is the initial weight of the samples before swelling, w 1 is the final weight of the dried (extracted) samples, and w is the weight of the swollen samples. These measurements for gel percentage and swelling degree were done in triplicate for each network sample.

In Vitro Enzymatic Degradation
According to previous reports [2,21,31], lipase can effectively degrade PTMC. PTMC-Ns rods with a length of 10 mm were weighed and conditioned in lipase solutions (from Thermomyces lanuginosus, ≥100,000 U/g), which were refreshed twice a week. The degradation experiments were performed in triplicate at 37 • C with gentle shaking. At regular time intervals, the PTMC-Ns rods were removed from the degradation media and washed with distilled water. After blotted with a tissue, the specimens were weighted and then vacuum dried at 37 • C to a constant weight for further analysis, and the diameter of the dried specimens was recorded. The pH of the media containing degradation products was also measured. The mass loss and loss in diameter were calculated via the following equations: where w i , w w and w d represent the initial weight, wet weight and dry weight of the samples, respectively; and d i and d d represent the initial diameter and final diameter of the dry samples after degradation, respectively. As a control, the hydrolytic degradation of the PTMC-Ns rods was carried out similarly with the enzymatic degradation, except that the medium was pH = 7.4 PBS solutions in absence of lipase.

Synthesis and Properties of PTMC-Ns
To investigate the effect of initial molecular weight and cross-linking amount as well as the cross-linker type on degradation, PTMC-Ns were fabricated in different cross-linking conditions, as described in our previous works [25]. Briefly, linear PTMC with four different molecular weights in the range of 72-329 kg/mol was cross-linked by 0.1 mol % BTB to prepare PTMC-Ns, and the molecular weight of the linear PTMC was regarded as the initial molecular weight of the PTMC-Ns. Besides, PTMC-Ns were obtained via the cross-linking of PTMC (Mn = 256 kg/mol) with varied BTB amount from 0 to 0.5 mol %. To investigate the effect of cross-linker type, PTMC (Mn = 256 kg/mol) was cross-linked by 0.1 mol % BTB and BCP, respectively. The properties of the obtained PTMC-Ns are shown in Table 1.
As shown in Table 1, the cross-linking conditions greatly influenced the properties of the resulting PTMC-Ns. By increasing the initial molecular weight and by increasing the cross-linker amount, higher gel percentage and lower swelling degree as well as enhanced thermal and mechanical properties were recorded for the obtained PTMC-Ns. For example, at a given initial molecular weight of 256 kg/mol, the gel percentage increased from 80.94 ± 3.23% to 97.89 ± 1.1% and the swelling degree decreased from 3943 ± 103% to 498 ± 16% as the BTB amount increased from 0.05 to 0.5 mol %, while, at the given BTB amount of 0.1 mol %, the gel percentage increased from 39.85 ± 1.98% to 97.96 ± 1.29% and the swelling degree decreased from 9916 ± 270% to 567 ± 25% with the increase of initial molecular weight from 72 to 329 kg/mol. Similar tendency could be observed for the thermal and the mechanical properties of the PTMC-Ns. Table 1 also showed that the PTMC-Ns cross-linked by BTB had slightly stronger physical properties than that of counterparts cross-linked by BCP. It was because of the more similar structure and reactivity of BTB than that of BCP to TMC monomer.

Mass Loss and Degradation Rate
It has been reported that lipase is one of the most efficient enzymes in eroding PTMC polymers [32]. The enzymatic degradation of the resulting PTMC-Ns was performed in vitro in the presence of the lipase (from Thermomyces lanuginosus, ≥100,000 U/g) at 37 • C. Figure 1 shows the profiles of the mass loss of the PTMC-Ns as a function of the degradation time.

Mass Loss and Degradation Rate
It has been reported that lipase is one of the most efficient enzymes in eroding PTMC polymers [32]. The enzymatic degradation of the resulting PTMC-Ns was performed in vitro in the presence of the lipase (from Thermomyces lanuginosus, ≥100,000 U/g) at 37 °C. Figure 1 shows the profiles of the mass loss of the PTMC-Ns as a function of the degradation time. It was clearly seen that the mass loss of the PTMC-Ns was dependent on the cross-linker amount and the initial molecular weight. Figure 1A showed that the mass loss of PTMC-Ns cross-linked by higher cross-linker amount was lower, whereas the networks cross-linked by lower cross-linker amount lost their mass in a much faster way. The mass loss of the PTMC256-0, PTMC256-0.05, PTMC256-0.1 and PTMC256-0.5 was 46.3 ± 0.4, 33.6 ± 1.8, 30.4 ± 0.4 and 29.8 ± 0.03%, respectively, after 12 weeks. Furthermore, the PTMC initial molecular weight also had conspicuous effect on the enzymatic degradation behavior of the PTMC-Ns. As shown in Figure 1B, the mass loss of PTMC-Ns degraded in lipase increased with the initial molecular weight. For instance, at a given BTB amount of 0.1 mol %, the PTMC72-0.1 and PTMC135-0.1 displayed a mass loss of 26.6 ± 0.2% and 29.6 ± 1.6%, respectively, after 12 weeks, whereas the mass loss for PTMC329-0.1 was 33.7 ± 0.05% at the same time point.
To quantitatively elucidate the degradation behavior, the degradation rate constant of the PTMC-Ns was calculated from the mass loss based on a first-order kinetic model, and the following equation was adopted: where k is the degradation rate constant and Mt is mass loss of PTMC-Ns at the degradation time of t.
The results determined from linear fitting of the data in the plots of mass loss versus the degradation time ( Figure 1) are listed in Tables 2 and 3. Apparently, the degradation rate of the PTMC-Ns was lower versus that of PTMC and decreased with increasing the cross-linker amount, as shown in Table 2. The degradation rate of the non-crosslinked PTMC256-0 was 3.81 %/w, which was higher than 2.76 %/w of the 0.05 mol % BTB cross-linked PTMC network (PTMC256-0.05). Then, the degradation rate of the PTMC256-Ns further decreased to 2.48 %/w as the cross-linker amount reached to 0.5 mol %. The results indicate that cross-linking could result in networks with much reduced degradation rate [33], as the cross-linker amount increased from 0 to 0.5 mol %. It was due to that gel with three-dimensional structures was insensitive to the enzymes attack, the PTMC-Ns with higher gel percentage have higher anti-degradation ability. The gel percentage for the PTMC256-Ns crosslinker by 0.05-0.5 mol % BTB was in a range of 80.94 ± 3.23-97.89 ± 1.1%, indicating the formation of much denser networks that make the carbonate bonds less accessible for enzymatic hydrolysis. It was clearly seen that the mass loss of the PTMC-Ns was dependent on the cross-linker amount and the initial molecular weight. Figure 1A showed that the mass loss of PTMC-Ns cross-linked by higher cross-linker amount was lower, whereas the networks cross-linked by lower cross-linker amount lost their mass in a much faster way. The mass loss of the PTMC 256 -0, PTMC 256 -0.05, PTMC 256 -0.1 and PTMC 256 -0.5 was 46.3 ± 0.4%, 33.6 ± 1.8%, 30.4 ± 0.4% and 29.8 ± 0.03%, respectively, after 12 weeks. Furthermore, the PTMC initial molecular weight also had conspicuous effect on the enzymatic degradation behavior of the PTMC-Ns. As shown in Figure 1B, the mass loss of PTMC-Ns degraded in lipase increased with the initial molecular weight. For instance, at a given BTB amount of 0.1 mol %, the PTMC 72 -0.1 and PTMC 135 -0.1 displayed a mass loss of 26.6 ± 0.2% and 29.6 ± 1.6%, respectively, after 12 weeks, whereas the mass loss for PTMC 329 -0.1 was 33.7 ± 0.05% at the same time point.
To quantitatively elucidate the degradation behavior, the degradation rate constant of the PTMC-Ns was calculated from the mass loss based on a first-order kinetic model, and the following equation was adopted: where k is the degradation rate constant and M t is mass loss of PTMC-Ns at the degradation time of t. The results determined from linear fitting of the data in the plots of mass loss versus the degradation time ( Figure 1) are listed in Tables 2 and 3. Apparently, the degradation rate of the PTMC-Ns was lower versus that of PTMC and decreased with increasing the cross-linker amount, as shown in Table 2. The degradation rate of the non-crosslinked PTMC 256 -0 was 3.81 %/w, which was higher than 2.76 %/w of the 0.05 mol % BTB cross-linked PTMC network (PTMC 256 -0.05). Then, the degradation rate of the PTMC 256 -Ns further decreased to 2.48 %/w as the cross-linker amount reached to 0.5 mol %. The results indicate that cross-linking could result in networks with much reduced degradation rate [33], as the cross-linker amount increased from 0 to 0.5 mol %. It was due to that gel with three-dimensional structures was insensitive to the enzymes attack, the PTMC-Ns with higher gel percentage have higher anti-degradation ability. The gel percentage for the PTMC 256 -Ns cross-linker by 0.05-0.5 mol % BTB was in a range of 80.94 ± 3.23%-97.89 ± 1.1%, indicating the formation of much denser networks that make the carbonate bonds less accessible for enzymatic hydrolysis. It is well known that the mobility of the polymer chain is an important factor that influences the polymers degradation [34]. The dense network formation could weaken the chain mobility of PTMC-Ns, it would make the chains more inaccessible to the active site of the enzymes, and then reduce the degradation rate. On the other hand, water uptake occurred during the degradation and appeared to be an important factor in the enzymatic degradation of PTMC-Ns. The permeated water molecules could lead to plasticization of the networks, which would enhance the polymer chain mobility and thus facilitate enzymatic attack. As mentioned earlier, the swelling degree of PTMC 256 -Ns in chloroform decreased from 3943 ± 103% to 498 ± 16% with the increase of BTB amount from 0.05 to 0.5 mol %. This indicated that the formed networks were densely cross-linked, almost no space so that water could easily penetrate into the networks. Thus, the water uptake of the PTMC 256 -Ns decreased with the increase of BTB amount ( Figure 2). As a result, we concluded that a higher cross-linking density in the PTMC 256 -Ns cross-linked by higher BTB amount enabled it to be degraded slower.   Table 3. The enzymatic degradation rate of BTB cross-linked PTMC-Ns with different initial molecular weight determined from the increase of mass loss.

PTMC-Ns
It is well known that the mobility of the polymer chain is an important factor that influences the polymers degradation [34]. The dense network formation could weaken the chain mobility of PTMC-Ns, it would make the chains more inaccessible to the active site of the enzymes, and then reduce the degradation rate. On the other hand, water uptake occurred during the degradation and appeared to be an important factor in the enzymatic degradation of PTMC-Ns. The permeated water molecules could lead to plasticization of the networks, which would enhance the polymer chain mobility and thus facilitate enzymatic attack. As mentioned earlier, the swelling degree of PTMC256-Ns in chloroform decreased from 3943 ± 103 to 498 ± 16% with the increase of BTB amount from 0.05 to 0.5 mol %. This indicated that the formed networks were densely cross-linked, almost no space so that water could easily penetrate into the networks. Thus, the water uptake of the PTMC256-Ns decreased with the increase of BTB amount ( Figure 2). As a result, we concluded that a higher cross-linking density in the PTMC256-Ns cross-linked by higher BTB amount enabled it to be degraded slower. The in vitro enzymatic degradation rate of the PTMC-Ns seemed to be initial molecular weight dependent. As shown in Table 3, at a given BTB amount of 0.1 mol %, the degradation rate of the PTMC-Ns with low initial molecular weight of 72 kg/mol (PTMC72-0.1) was 2.16 %/w, and then it was increased to 2.44 %/w for PTMC135-0.1 as the initial molecular weight reached to 135 kg/mol. As the The in vitro enzymatic degradation rate of the PTMC-Ns seemed to be initial molecular weight dependent. As shown in Table 3, at a given BTB amount of 0.1 mol %, the degradation rate of the PTMC-Ns with low initial molecular weight of 72 kg/mol (PTMC 72 -0.1) was 2.16 %/w, and then it was increased to 2.44 %/w for PTMC 135 -0.1 as the initial molecular weight reached to 135 kg/mol. As the initial molecular weight further increased to 256 and 329 kg/mol, the degradation rate of PTMC 256 -0.1 and PTMC 329 -0.1 was 2.52 %/w and 2.78 %/w, respectively; obviously, the higher the initial molecular weight, the greater the degradation rate of the PTMC-Ns.
This phenomenon was related to the hydrophilicity, which has an important effect on the polymer degradation behavior [35]. As shown in Figure 3, the static contact angle of the 0.1 mol % BTB cross-linked PTMC-Ns increased with the increase of the initial molecular weight, reaching a maximum of 91.9 • at the 326 kg/mol. The results illustrated that PTMC samples became more hydrophobic as the initial molecular weight increased. It has been reported that the lipase activity is exhibited much higher on the hydrophobic surface [36,37]. Hence, the lipase activity is high in degrading the PTMC-Ns with high initial molecular weight, owing to their hydrophobic surfaces. The result was in coincidence with that of the in vitro enzymatic degradation of the pure linear PTMC with different initial molecular weight in our previous works [30]. initial molecular weight further increased to 256 and 329 kg/mol, the degradation rate of PTMC256-0.1 and PTMC329-0.1 was 2.52 %/w and 2.78 %/w, respectively; obviously, the higher the initial molecular weight, the greater the degradation rate of the PTMC-Ns. This phenomenon was related to the hydrophilicity, which has an important effect on the polymer degradation behavior [35]. As shown in Figure 3, the static contact angle of the 0.1 mol % BTB cross-linked PTMC-Ns increased with the increase of the initial molecular weight, reaching a maximum of 91.9° at the 326 kg/mol. The results illustrated that PTMC samples became more hydrophobic as the initial molecular weight increased. It has been reported that the lipase activity is exhibited much higher on the hydrophobic surface [36,37]. Hence, the lipase activity is high in degrading the PTMC-Ns with high initial molecular weight, owing to their hydrophobic surfaces. The result was in coincidence with that of the in vitro enzymatic degradation of the pure linear PTMC with different initial molecular weight in our previous works [30]. The effect of the cross-linker type on the degradation behavior of the resulting PTMC-Ns was also investigated, and the results were displayed in Figure 4. It was found that the mass loss of PTMC256-0.1 cross-linked by BTB or BCP increased gradually with degradation. The linear relationships between the mass loss and degradation time were determined for the both cases, and the degradation rate calculated based on Equation (4) was 2.52 %/w (R = 0.996) and 2.67 %/w (R = 0.998), respectively. The degradation rate of the BTB cross-linked PTMC256-0.1 was slightly lower than that of the BCP cross-linked ones. It was attributed to the difference in the gel percentages of the two groups as shown in Table 1. Similar to the results stated above, higher gel percentage and Tg induced reduction in the degradation rate. Here, we concluded that the BTB was more suitable to be utilized widely to fabricate the strong elastomeric polymer network with low degradation rate for biomedical applications. The effect of the cross-linker type on the degradation behavior of the resulting PTMC-Ns was also investigated, and the results were displayed in Figure 4. It was found that the mass loss of PTMC 256 -0.1 cross-linked by BTB or BCP increased gradually with degradation. The linear relationships between the mass loss and degradation time were determined for the both cases, and the degradation rate calculated based on Equation (4) was 2.52 %/w (R = 0.996) and 2.67 %/w (R = 0.998), respectively. The degradation rate of the BTB cross-linked PTMC 256 -0.1 was slightly lower than that of the BCP cross-linked ones. It was attributed to the difference in the gel percentages of the two groups as shown in Table 1. Similar to the results stated above, higher gel percentage and T g induced reduction in the degradation rate. Here, we concluded that the BTB was more suitable to be utilized widely to fabricate the strong elastomeric polymer network with low degradation rate for biomedical applications.
The results of the degradation rate of the PTMC-Ns stated above strongly demonstrated that the chemical cross-linking was an efficient strategy to reduce the degradation rate of PTMC via the manipulation of the cross-linker amount, initial molecular weight as well as the cross-linker type, allowing their extensive use in the long-term biomedical applications. The results of the degradation rate of the PTMC-Ns stated above strongly demonstrated that the chemical cross-linking was an efficient strategy to reduce the degradation rate of PTMC via the manipulation of the cross-linker amount, initial molecular weight as well as the cross-linker type, allowing their extensive use in the long-term biomedical applications.

SEM Observation
The SEM measurements were performed to monitor the degradation of the resulting PTMC-Ns. Figure 5 shows the surface morphology of PTMC329-0.1 before and after degradation. The original samples of PTMC329-0.1 exhibited a smooth surface before enzymatic degradation. However, the surface became rough after 1 week of degradation, and a highly hollow structure was detected after four weeks, the size and deepness of the hollows observed on the surface increased with the incubation time, as illustrated in Figure 5. The results were similar to the findings reported in our previous works [30]. Figure 6 shows the surface morphology of the PTMC-Ns with different cross-linker amount and initial molecular weight after four weeks degradation in 37 °C lipase solutions. Obviously, the surface morphology varied with the formulation of the resulting PTMC-Ns. After four weeks of degradation, pits with greater size and deepness were observed on the surface of PTMC256-0, indicating that the surfaces had been extensively eroded. The pits on the surface of PTMC256-0.1 and PTMC256-0.5 became much shallower as indicated by the smaller roughness. That indicates the reduction in the erosion extent of the resulting networks with increasing cross-linker amount is very clear. Similar results were observed as the initial molecular weight decreased. The SEM results were coincidence with that of the mass loss.

SEM Observation
The SEM measurements were performed to monitor the degradation of the resulting PTMC-Ns. Figure 5 shows the surface morphology of PTMC 329 -0.1 before and after degradation. The original samples of PTMC 329 -0.1 exhibited a smooth surface before enzymatic degradation. However, the surface became rough after 1 week of degradation, and a highly hollow structure was detected after four weeks, the size and deepness of the hollows observed on the surface increased with the incubation time, as illustrated in Figure 5. The results were similar to the findings reported in our previous works [30].   Figure 6 shows the surface morphology of the PTMC-Ns with different cross-linker amount and initial molecular weight after four weeks degradation in 37 • C lipase solutions. Obviously, the surface morphology varied with the formulation of the resulting PTMC-Ns. After four weeks of degradation, pits with greater size and deepness were observed on the surface of PTMC 256 -0, indicating that the surfaces had been extensively eroded. The pits on the surface of PTMC 256 -0.1 and PTMC 256 -0.5 became much shallower as indicated by the smaller roughness. That indicates the reduction in the erosion extent of the resulting networks with increasing cross-linker amount is very clear. Similar results were observed as the initial molecular weight decreased. The SEM results were coincidence with that of the mass loss.

Form-Stability and Degradation Mechanism
The form-stability of biodegradable polymers greatly influences their utility in the final biomedical devices, especially in those where deformation is not desired, e.g., the implant carriers for controlled drug release. Here, the form-stability of the resulting PTMC-Ns in vitro enzymatic degradation was determined via macroscopic observation. Figure 7 shows that the uncross-linked low-molecular-weight PTMC rods (PTMC72-0 and PTMC135-0) splayed poor form-stability, the shape of the polymer rods was not cylindrical anymore, Figure 6. SEM micrographs of the PTMC-Ns surface with different cross-linker amount (Rank) and different initial molecular weight (Column) after four weeks degradation in lipase solution.

Form-Stability and Degradation Mechanism
The form-stability of biodegradable polymers greatly influences their utility in the final biomedical devices, especially in those where deformation is not desired, e.g., the implant carriers for controlled drug release. Here, the form-stability of the resulting PTMC-Ns in vitro enzymatic degradation was determined via macroscopic observation. Figure 7 shows that the uncross-linked low-molecular-weight PTMC rods (PTMC 72 -0 and PTMC 135 -0) splayed poor form-stability, the shape of the polymer rods was not cylindrical anymore, especially for the PTMC 72 -0 that underwent the worst deformation, and began changing its shape from cylinder to a highly sphere at week 1. Obviously, the corresponding cross-linked PTMC-Ns presented much better form-stability than the uncross-linked ones. As shown in Figure 7, the resulting networks PTMC 72 -0.1 and PTMC 135 -0.1 retained their initial appearance and no deformation was observed during the degradation process. Therefore, we deduced that cross-linking plays an important role in enhancing the form-stability of the PTMC-based polymers. In our previous works [21], we found that the uncross-linked high-molecular-weight PTMC (PTMC 256 -0 and PTMC 329 -0) could sustain their forms during the in vitro enzymatic degradation process, indicating their good form-stability. In this study, no difference was determined in the form-stability between the uncross-linked and cross-linked PTMC-Ns with high initial molecular weight (the figures were not displayed). The results illustrated that the cross-linking influenced the form-stability of the low-molecular-weight PTMC more significantly.
resulting networks PTMC72-0.1 and PTMC135-0.1 retained their initial appearance and no deformation was observed during the degradation process. Therefore, we deduced that cross-linking plays an important role in enhancing the form-stability of the PTMC-based polymers. In our previous works [21], we found that the uncross-linked high-molecular-weight PTMC (PTMC256-0 and PTMC329-0) could sustain their forms during the in vitro enzymatic degradation process, indicating their good form-stability. In this study, no difference was determined in the form-stability between the uncrosslinked and cross-linked PTMC-Ns with high initial molecular weight (the figures were not displayed). The results illustrated that the cross-linking influenced the form-stability of the lowmolecular-weight PTMC more significantly. As seen in Figure 7, although the shape of PTMC-Ns was still rod-like independent of their initial molecular weight or the cross-linker amount, the diameter decreased upon enzymatic degradation. Figure 8A,B clearly shows the loss in diameter of the PTMC-Ns as a function of the degradation time. Apparently, the loss in diameter was dependent on the cross-linker amount and the initial molecular weight of PTMC-Ns. The higher the BTB amount, the lower the loss in diameter ( Figure 8A), and the lower the initial molecular weight, the lower the loss in diameter ( Figure 8B). The results were in a good agreement with that of the mass loss and degradation rates of PTMC-Ns. As seen in Figure 7, although the shape of PTMC-Ns was still rod-like independent of their initial molecular weight or the cross-linker amount, the diameter decreased upon enzymatic degradation. Figure 8A,B clearly shows the loss in diameter of the PTMC-Ns as a function of the degradation time. Apparently, the loss in diameter was dependent on the cross-linker amount and the initial molecular weight of PTMC-Ns. The higher the BTB amount, the lower the loss in diameter ( Figure 8A), and the lower the initial molecular weight, the lower the loss in diameter ( Figure 8B). The results were in a good agreement with that of the mass loss and degradation rates of PTMC-Ns.
The linear relationship between mass loss and loss in diameter, shown in Figure 8C,D, implied that the decrease in mass occurred simultaneously with a decrease in diameter, indicating that the in vitro enzymatic degradation of PTMC-Ns occurred via surface erosion mechanism. Similar results were observed for the BCP cross-linked PTMC 256 -0.1.
As the surface erosion mechanism worked on the enzymatic degradation of PTMC-Ns, the effective surface area of the degraded samples crucially influenced the degradation rate. Due to the poor form-stability, the effective surface area of the low-molecular-weight PTMC rods (PTMC 72 -0 and PTMC 135 -0) became smaller as they presented an appearance of oblate or sphere (Figure 7). Consequently, the degradation rate decreased with decreasing the surface area. The degradation rate of PTMC 72 -0 and PTMC 135 -0 calculated based on the increase of the mass loss was 0.94 and 1.46 %/w [21], respectively, much lower than that of the corresponding cross-linked ones. The results once again confirmed that the form-stability of biodegradable implants was an important property that should be paid high attention to, because it not only greatly influenced the final utility of the biomedical devices, but also played a significant role in the degradation behavior of the implants. The linear relationship between mass loss and loss in diameter, shown in Figure 8C,D, implied that the decrease in mass occurred simultaneously with a decrease in diameter, indicating that the in vitro enzymatic degradation of PTMC-Ns occurred via surface erosion mechanism. Similar results were observed for the BCP cross-linked PTMC256-0.1.
As the surface erosion mechanism worked on the enzymatic degradation of PTMC-Ns, the effective surface area of the degraded samples crucially influenced the degradation rate. Due to the poor form-stability, the effective surface area of the low-molecular-weight PTMC rods (PTMC72-0 and PTMC135-0) became smaller as they presented an appearance of oblate or sphere ( Figure 7). Consequently, the degradation rate decreased with decreasing the surface area. The degradation rate of PTMC72-0 and PTMC135-0 calculated based on the increase of the mass loss was 0.94 and 1.46 %/w [21], respectively, much lower than that of the corresponding cross-linked ones. The results once again confirmed that the form-stability of biodegradable implants was an important property that should be paid high attention to, because it not only greatly influenced the final utility of the biomedical devices, but also played a significant role in the degradation behavior of the implants.

The Role of Lipase on Degradation of PTMC-Ns
To elucidate the role of lipase in the degradation behavior of PTMC-Ns, the control experiment (hydrolytic degradation) without enzyme was also performed. The hydrolytic degradation of PTMC256-0.1 was investigated in pH 7.4 PBS. Figure 9 shows the mass loss of the PTMC256-0.1 degraded in the absence of lipase as a function of the degradation time, in comparison with that of the enzymolysis case.

The Role of Lipase on Degradation of PTMC-Ns
To elucidate the role of lipase in the degradation behavior of PTMC-Ns, the control experiment (hydrolytic degradation) without enzyme was also performed. The hydrolytic degradation of PTMC 256 -0.1 was investigated in pH 7.4 PBS. Figure 9 shows the mass loss of the PTMC 256 -0.1 degraded in the absence of lipase as a function of the degradation time, in comparison with that of the enzymolysis case. It is clearly seen that the PTMC256-0.1 lost its mass much faster with lipase than without lipase. After 12 weeks, the mass loss of PTMC256-0.1 incubated in lipase was 30.4 ± 0.37%, whereas it only lost 5.20 ± 0.67% of its mass after 50 weeks of degradation without lipase. The degradation rate constant of PTMC256-0.1 in the presence and in the absence of the lipase were 2.52 and 0.098 %/w, respectively. The former was more than 25 times higher than the latter. The rapid surface erosion of PTMC-Ns in enzymatic conditions as well as stable under hydrolytic conditions suggested that enzymes played an important role in the degradation process. It was attributed to the fact that the lipase could act as a surfactant to interact with the degradation products of PTMC-Ns and accelerate It is clearly seen that the PTMC 256 -0.1 lost its mass much faster with lipase than without lipase. After 12 weeks, the mass loss of PTMC 256 -0.1 incubated in lipase was 30.4 ± 0.37%, whereas it only lost 5.20 ± 0.67% of its mass after 50 weeks of degradation without lipase. The degradation rate constant of PTMC 256 -0.1 in the presence and in the absence of the lipase were 2.52 and 0.098 %/w, respectively. The former was more than 25 times higher than the latter. The rapid surface erosion of PTMC-Ns in enzymatic conditions as well as stable under hydrolytic conditions suggested that enzymes played an important role in the degradation process. It was attributed to the fact that the lipase could act as a surfactant to interact with the degradation products of PTMC-Ns and accelerate them dispersing into the surrounding media. We concluded that the lipase could accelerate the degradation of PTMC-Ns by enhancing the loss of the degradation products, as observed in the in vitro degradation of the PTMC in our separate studies [21].

pH
It has been reported that the acidic degradation compounds are not produced upon the degradation of PTMC [1,2,21]. As far as PTMC-Ns are concerned, the degradation products should be non-acidic because of the polycarbonate structure of PTMC-Ns. There are fair evidences to support this view from the changes in pH of the degradation media. During erosion studies, the enzyme solutions remained clear indicating that degradation products were water soluble. In Figure 10, the pH of the lipase solution was monitored with the degradation time. As expected, no acidification was seen when the networks degradation occurred. The pH of lipase was almost unchanged was close to the initial value of 6.02, strongly demonstrating that PTMC-Ns degraded without acidic products. Importantly, the property of releasing no acidic products was independent of the cross-linking amount and the initial molecular weight. The same result was observed for the BCP cross-linked PTMC-Ns. It is very important to eliminate the inflammatory reaction evoked by the acidic degradation products as observed in the cases of PLA, PGA and their copolymers [5][6][7]. In addition, no accelerated degradation induced by the autocatalysis [38] would take place during the degradation of PTMC-Ns, allowing them to be applied more widely in the biomedical fields.

Conclusions
The in vitro enzymatic degradation behavior of PTMC-Ns was quantitatively studied. The results illustrated that the degradation of PTMC-Ns was dependent on the initial molecular weight. Higher mass loss and subsequent higher degradation rate were detected for PTMC-Ns with higher initial molecular weight at a given cross-linker amount. The degradation rate of PTMC-Ns could also be tailored by the adjustment of cross-linker amount: the higher the cross-linker amount, the slower the degradation rate. Furthermore, the cross-linker type was also a valid means to tune the degradation rate of PTMC-Ns. The degradation rate of PTMC-Ns was slower than that of the linear ones, indicating the enhanced resistance to degradation of PTMC via cross-linking.
PTMC-Ns maintained better form-stability versus the non-crosslinked ones during the enzymatic degradation, indicating that cross-linking was an effective strategy to enhance the formstability of PTMC. With the comparison of an in vitro hydrolysis case, the much faster degradation rate in lipase was attributed to the essential role of enzymes in the erosion of PTMC-Ns. We concluded that the lipase acted as a surfactant to promote the dissolution of the degradation products into the surrounding media and thus accelerate the degradation via surface erosion mechanism.
As stated above, PTMC-Ns with enhanced form-stability and lowered degradation rate given by cross-linking are the promising candidates for potential clinical application in subcutaneous

Conclusions
The in vitro enzymatic degradation behavior of PTMC-Ns was quantitatively studied. The results illustrated that the degradation of PTMC-Ns was dependent on the initial molecular weight. Higher mass loss and subsequent higher degradation rate were detected for PTMC-Ns with higher initial molecular weight at a given cross-linker amount. The degradation rate of PTMC-Ns could also be tailored by the adjustment of cross-linker amount: the higher the cross-linker amount, the slower the degradation rate. Furthermore, the cross-linker type was also a valid means to tune the degradation rate of PTMC-Ns. The degradation rate of PTMC-Ns was slower than that of the linear ones, indicating the enhanced resistance to degradation of PTMC via cross-linking.
PTMC-Ns maintained better form-stability versus the non-crosslinked ones during the enzymatic degradation, indicating that cross-linking was an effective strategy to enhance the form-stability of PTMC. With the comparison of an in vitro hydrolysis case, the much faster degradation rate in lipase was attributed to the essential role of enzymes in the erosion of PTMC-Ns. We concluded that the lipase acted as a surfactant to promote the dissolution of the degradation products into the surrounding media and thus accelerate the degradation via surface erosion mechanism.
As stated above, PTMC-Ns with enhanced form-stability and lowered degradation rate given by cross-linking are the promising candidates for potential clinical application in subcutaneous implants. The degradation behavior with no acidic degradation products would donate well biocompatibility to PTMC-Ns. Future studies will focus on the in vivo degradation behaviors of and the tissue response to PTMC-Ns.