WO2012065745A1

WO2012065745A1 - Process for preparing copolyesters, copolyesters and their medical uses

Info

Publication number: WO2012065745A1
Application number: PCT/EP2011/005791
Authority: WO
Inventors: I. Van Der Meulen; R. Deumens; A. Heise; C. E. Koning; M. A. E. Marcus; E. A. J. Joosten
Original assignee: Stichting Dutch Polymer Institute
Priority date: 2010-11-18
Filing date: 2011-11-17
Publication date: 2012-05-24

Abstract

The invention relates to a copolymer from at least an unsaturated macrolactone A and a lactone B, wherein the lactone B is chosen from the group consisting of ε-caprolactone (CL), substituted lactones and lactones that contain one or more etherlinkages in the lactone ring. The invention further relates to porous structures of the biodegradable copolymer, which can be used for in example in medical applications.

Description

PROCESS FOR PREPARING COPOLYESTERS, COPOLYESTERS AND THEIR MEDICAL USES

The present invention relates to copolymers of a macrolactone and a comonomer. The invention also relates to a process for preparing such a copolymer, in particular to an enzyme catalyzed process for preparing a copolymer using ring-opening polymerization. The invention also relates to medical uses of such copolymers.

Polyesters are ery interesting materials because of the properties that these materials can exhibit. These properties, for instance, include biocompatibility, biodegradability and drug permeability. Therefore, polyesters are of great interest for medical and food packaging applications. For these purposes materials with an engineered structure are desired, which implies the need for a high level of control over the polymerization reaction. In addition, with the right properties, polyesters can form an interesting biodegradable alternative for polyethylene in many applications.

Traditional polyester synthesis strategies, using e.g. polycondensation, give rise to fundamental problems that can make the controlled synthesis of these materials a tedious process. For example, the preparation of polyesters by polycondensation can be accompanied by stoichiometric problems, the need for high conversion and the removal of small molecules formed during the reaction.

A suitable replacement for these conventional strategies is the ring-opening polymerization of lactones. This polymerization is based on the fact that cyclic monomers "open up" and form a polymer chain by means of a chain-growth process.

It is known that ring-opening homo-polymerization reactions can be performed with enzymes with satisfactory conversion under mild polymerization conditions (in particular at low temperatures). For example, lipases such as Candida Antarctica Lipase B (CALB) are highly active in the ring-opening polymerization of lactones and show exceptionally high polymerization rates for macrolactones. The reactivity of lactones in this process is not governed by the high ring-strain of small lactones (cisoid ester bonds) but by the preference of the lipase for transoid ester bond conformation present in large ring lactones. Macrolactones (i.e. lactones having a large number of atoms in the ring, preferably more than 15) can thus easily be polymerized by CALB. For example, poly(pentadecalactone) with a number average molecular weight up to 150 000 g/mol has been reported (Focarete et al., J. Polym. Sci. B: Polym. Phys. 2001 , 39, 1721 and De Geus et a/., Polym. Chem. 2010. 1, 525).

Cross-linked polyesters are usually made with α,ω-functionalized macromonomers of the polyester used (see: T. Aoyagi et al., J. Control. Release 1994, 32, 87; M. Takwa et al., Macromol. Rapid Comm. 2006, 27, 1932; and N. Simpson et al., Macromolecules 2008, 41, 3613).

For example this can be done by using acrylates as end groups which can be polymerized via radical polymerization (see: T. R. Thatiparti et al., J. Biomed. Mater. Res. B: Appl. Biomat. 2009, 111). The formed network contains a polyacrylate with polyester crosslinks.

An alternative to radical polymerization for the cross-linking is thiol-ene chemistry, using macromonomers end-functionalized with thiols or acrylates (see: . Takwa et al., Macromolecules 2008, 41, 5230). These groups can react with each other, resulting in the formation of thioether cross-links.

A disadvantage of homopolymers of macrolactones is that they are not bio-degradable, in particular non-degradable under physiological conditions (i.e. conditions existing inside the human or animal body).

Another disadvantage of the known polymers and copolymers of macrolactones is that crosslinking them is a very laborious procedure: first the obtained or synthesized macromonomer has to be functionalized (introducing cross-linkable end groups; secondly the monomer has to be polymerized into a cross-linkable polymer or copolymer; and finally the obtained polymer has to be cross-linked via one of the above described routes. It is an object of the present invention to provide a cross-linkable biodegradable polymer based on macrolactone monomers. It is another object of the present invention to provide a simple process for preparing a cross-linkable polymer based on macrolactone monomers.

These objects are at least partly achieved by providing a copolymer from comprising a first monomeric unit (A) derivable from an unsaturated macrolactone monomer and a lactone B; and by providing an enzyme catalyzed ring-opening polymerization process using a lipase type enzyme as a catalyst for making the same. The copolymers according to the present invention are, contrary to homo-polymers of macrolactones, biodegradable; Without wanting to be bound to any theory, it is expected that this is due to their decreased crystallinity and decreased hydrophobiclty, when compared to homo-polymers of macrolactones.

The copolymers according to the present invention are unsaturated (i.e. contain -C=C- bonds), which makes them cross-linkable.

The copolymers according to the present invention are prepared from at least an unsaturated macrolactone A and a lactone B, wherein the lacton B are chosen from the group consisting of ε-caprolactone (CL), substituted lactones and lactones that contain one or more etherlinkages in the lactone ring.

Preferably, the copolymers according to the present invention are prepared from at least an unsaturated macrolactone A and a lactone B, wherein the lacton B are chosen from the group consisting of substituted lactones and lactones that contain one or more

etherlinkages in the lactone ring.

Macrolactone A is optionally substituted and may have a ring size ranging from 10 to 50, preferably between 12 and 40, more preferably between 14 and 25 carbon atoms, and contains at least one unsaturated C=C group. Examples of suitable macrolactone A are 5-tetradecene-1 -olide, 1 1-pentadecene-15-olide, 12-pentadecene-15-olide (also known as globalide), 7-hexadecene-16-olide (also known as ambrettolide), and

9- hexadecene- 16-olide .

Preferred unsaturated macrolactones A are globalide (Gl) or ambrettolide (Am).

Suitable lactones B are lactones which are substituted lactones or lactones that contain one or more etherlinkages in the lactone ring. Lacton B does not contain an unsaturated C=C group in the ring. The substituted lactones preferably have a small ring size, preferably comprising 7, 8, or 9 carbon atoms in the ring. Examples of substituted lactones B are 4-methyl caprolactone (4MeCL) and 4-ethyicaprolactone. Examples of lactones B that contain ether linkages are 1 ,5-dioxepan-2-one (DXO), 1 ,4-dioxan-2-one,

10- oxahexadecanolide, 11-oxahexadecanolide, 12-oxahexadecanolide,

12-oxahexadecen-16-olide and 2-oxo-12-crown-4-ether (OC). Most preferred are lactones B which are selected from the group consisting of 1,5-dioxepan-2-one (DXO) and 2-oxo-12-crown-4-ether (OC).

In a preferred embodiment of the invention the copolymer comprises a rriacrolactone A which is globalide (Gl) or ambrettolide (Am) and the lactone B is selected from the group consisting of 1,5-dioxepan-2-one (DXO) and 2-oxo-12-crown-4-ether (OC).

The copolymer according to the invention can also contain further lactones. Examples of such further lactones include lactones having a ring size of 6 to 40 carbon atoms. Ring sizes of less than 6 carbon atoms result in unacceptable low conversion and very low molecular weight. Preferably, the lactone is selected from the group consisting of n-valerolactone, 7-heptanolactone, 8-octalactone, 9-nonalactone, 10-decalactbne, 11-undecalactone, 12-dodecalactone, 13-tridecalactone, 14-tetradecalactone,

15- pentadecalactone, and 16-hexadecalactone. In each of these notations, the prefix specifies the number of carbons in the heterocycle (i.e. the distance between the relevant ester groups along the backbone). Therefore, the prefixes also indicate the size of the lactone ring. Preferably, the lactone used in the process of the invention has a ring size of 9-40 carbon atoms, even more preferably a ring size of 10-40 carbon atoms, such as a ring size of 12-40 carbon atoms. When using such lactones with relatively large ring sizes, the polymerization rate is relatively high.

Good results have, for instance, been obtained with lactones selected from the group consisting of 10-decalactone, 11-undecafactone, 15-pentadecalactone, and

16- hexadecalactone.

The copolymer may be in the form of e.g. a random copolymer, a statistical copolymer, a block copolymer or a gradient copolymer. Preferably the copolymer is a random copolymer.

The copolymer according to the invention is prepared from between 10 and 90 mol% of macromonomer A, preferably between 15 and 60 mol%, more preferably between 20 and 50 mol%; and from 10 to 90 mol% of lactone B, preferably between 40 and 85 mol%, most preferably between 50 and 80 mol%. The copolymer may have a Mn (number average molecular weight) between 10.000 and 50.000 g/mol (measured with GPC), preferably between 15.000 and 40.000, more preferably between 20.000 and 35.000. The MWD (molecular weight distribution Mw/Mn) as determined with GPC typically ranges between 1,8 and 3, more preferably between 2 and 2,5.

The copolymer according to the present invention can be prepared in an enzymatic polymerization. It has been found that polymerization reactions with lipase enzyms proceed favourably. Examples of such enzymes are Lipase CA, Lipase PS, Lipase CR, Lipase PC, Lipase PPL, Lipase PF. Preferred is the Candida Antarctica Lipase B. One example of the Candida Antarctica Lipase B is commercially available as CALB immobilized on a macroporous resin, (Novozym 435).

Preferably a solvent is used during the polymerization of the monomers. Suitable solvents that can be used in the polymerization should be able to dissolve the monomers, the polymer and the initiator (if necessary). Preferred solvents are toluene, THF, xylene, 1,2- dichloroethane, benzene, cyclohexane, chloroform, dioxane, heptanes and acetonitrile. Most preferred solvents are toluene, THF, xylene and cyclohexane.

Water present in the enzyme may be used as initiator for the polymerization reaction. The polymerization temperature typically ranges between 40 and 110 C, preferably between 50 and 90 C. Preferably the polymerization temperature is chosen above the melting temperature of the copolymer, in order to avoid high viscosities and heterogeneity of the polymerization mixture.

In an embodiment of the present invention, the copolymers of the invention are used to make porous structures like for example scaffolds. Such porous structures can be used in different application, like for example medical applications.

Porous structures can be made in different ways. In one example the copolymer is mixed with a porogen, which can be extracted (or leached) from the mixture. Particulate leaching is a very straightforward technique based on differences in solubility of the components used. Salt and sugar are two of the widest used porogeris in particulate leaching. The polymer solution is mixed with the porogen after which the mixture is lyophilized or cured. Afterwards the porogen is leached out by a good solvent for the porogen and a non-solvent for porous structure (like for example a polymeric scaffold). By altering the crystal size and the porogen weight fraction, the porous structure can be designed. Not only solids can be utilized as particulate, also other polymers can be used, such as PEG for example in a system where the polymer matrix is not soluble in water. Moreover, gasses can be used as porogen. An example of a suitable gas is supercritical carbon dioxide (scC0₂). In that case, a polymer sample is exposed to high pressure gas to saturate the sample. Afterwards the gas pressure is slowly decreased causing nucleation and pore formation in the sample. Preferred porogens contain sugars (like dextrose and glucose) or salts (like NaCI, or KCI). The sugar or salts are particulates having particles between 1 mu and 150 mu, preferably between 10 and 125 mu.

The weight ratio of polymer to porogen ranges between 1 :1 and 1 :15, preferably between 1 :5 and 1 :13, more preferably between 1 :7 and 1:12. Preferably the porous structures contain crosslinks. Crosslinks can be made with for example thermal curing agents and with UV-radiation activated curing agents. The skilled man is aware of these types of curing agents. A preferred example of a curing agent is dicumyl peroxide (DCP), which is known to be a very reactive thermal curing agent.

In another embodiment the cross-linking of the copolymer with thiol-ene chemistry can be used. This crosslinking route is preferred, since biodegradable crosslinks are formed upon curing the copolymer with thiol-ene chemistry.

In an example the copolymer was molten and mixed with a dithiol (for example ethylene glycol bis(3-mercaptopropionate)). It is known that the reaction between a thiol and an unsaturated carbon-carbon bond is very fast. Upon addition of a UV-initiator (for example 4-hydroxybenzophenone) and when exposed to UV light cross-linking takes place. No elevated temperatures are needed for curing and no byproducts are formed. Unreacted dithiol can be washed out after curing.

The porous copolymer preferably comprises pores that are interconnected with eachother. The sizes of the pores are not well defined, but in a scanning electron microscopy (SEM) picture pores can be identified that have sizes preferably ranging between 20 and 80 micro meter.

The invention also relates to a process for preparing the porous structure comprising the steps of a) Providing a copolymer

b) Mixing the copolymer with a porogen and crosslinking agent c) Crosslinking the copolymer

d) Leaching the porogen with a leaching solvent

.wherein the leaching solvent dissolves the porogen but not the crosslinked copolymer.

In one embodiment a solvent is used to dissolve the copolymer In this embodiment the process comprises the steps of a) Providing a copolymer dissolved in a first solvent

b) Mixing the copolymer with a porogen and crosslinking agent followed by

removing the first solvent

c) crosslinking the copolymer

d) Leaching the porogen with a second solvent

.wherein the first solvent does not substantially dissolve the porogen, and wherein the second solvent dissolves the porogen but not the crosslinked copolymer.

The first solvent should be able to dissolve the crosslinker (the dithiol), the initiator (UV or thermal) and the polymer. Examples of first solvents are toluene, THF, chloroform, dichloromethane, xylene, mesitylene, and diethylether.

Removal of the first solvent can be done by methods known to the skilled man. Examples of suitable methods are drying under vacuum (for example between 1 and 100 mbar),and for example while heating the sample to a temperature between 30 and 80 C.

The second solvents should be able to dissolve the porogen, and preferably unreacted thiol compound, but not the crosslinked polymer, Examples of second solvents are water, methanol and ethanol.

Polyesters and copolymers obtained with the process of the invention can be used in a wide variety of applications depending on their respective properties, such as number average molecular weight, polydispersity index, etc. Some non-limitative exemplary applications include the following. The polyesters and copolymers may be comprised in the fabrication of fibers with high mechanical strength. Especially polyesters and copolymers with high molecular weight are suitable for this purpose. For fiber applications it is further preferred that the polymers have, a relatively low polydispersity index. Furthermore, the polyesters and copolymers may be used for biomedical applications. In this respect it is highly advantageous that the degradability of the copolymers can be tuned by the choice of comonomer. Examples of biomedical applications include screws (such as for bone), scaffolding, sutures, drug delivery devices, nerve conduits etc. In addition, the polyesters and copolymers obtained by the process of the invention may be used as a general altemative for polyethylene. In contrast to polyethylene, however, the polyesters and copolymers of the invention are

advantageously biodegradable (rate of biodegradability can optionally be tuned by choosing one or more appropriate comonomers lactone B) and biocompatible. Hence, litter of the applied polymer will eventually completely degrade in a time span of months to years as compared to a time span of ages for polyethylene. In one preferred embodiment the copolymer of the invention is used to make a porous scaffold suitable as a nerve guide tube. Vital requirements of nerve guide tube (NGT) in regard of peripheral nerve regeneration include the presence of [1] a tubular permeable structure having sufficient strength and flexibility to withstand the stresses the nerve is exposed to, [2] a filling of an inner matrix material that consists of an interconnected porous network and contains nutrients and other growth factors or even Schwann cells. The interconnected porous network enables nerve fibers to grow through the bridge. Both the tubular structure and the inner matrix of a NGT can be made of the same material if this material fulfils all demands. Property requirements are degradability, biocompatibility, flexibility, strength (up to 7 Pa) and permeability. The present invention provides a crosslinked porous structure comprising a copolymer according to the present invention, for use as nerve guide tube.

The present invention further provides nerve guide tubes comprising a crosslinked porous structure of a copolymer according to the present invention.

EXAMPLES

Materials

Novozym 435 (Candida antarctica Lipase B immobilized on cross-linked polyacrylate beads) was purchased from Novozymes A/S and dried following a literature procedure (see M. de Geus et al., Macromolecules 2005, 38, 4220).

Globalide and ambrettolide were obtained from Symrise AG.

Toluene was dried over aluminum oxide and stored over molecular sieves. All other solvents used were purchased from Biosolve and used without further purification. 1,5-dioxepan-2-one (DXO), 4-methyl caprolactone (4MeCL) and 2-oxo-12-crown-4-ether (OC) were synthesized following literature procedures (see T. Mathisen et al.,

Macromolecutes 1989, 22, 3842; P. Vangeyte et al., Polym. Set. A: Polym. Chem. 2004, 42, 1132; and L. van der Mee et al., J. Polym. Sci. A: Polym. Chem. 2006, 44, 2166, respectively).

Dicumyl peroxide was purchased from Aldrich, 4-hydroxybenzophenone from Fluka and ethylene glycol bis(3-mercaptopropionate) was from Wako.

Methods

Karl-Fischer Coulometry

The water content of all reaction mixtures was measured with Karl-Fischer Coulometry using a Mettler Toledo DL32 Coulometer and Apura CombiCoulomat fritless (Merck) as electrolyte. Size exclusion chromatography (SEC)

Size exclusion chromatography (SEC) was performed on a Waters Alliance system equipped with a Waters 2695 separation module, a Waters 2414 refractive index detector (40 ^eC), a Waters 2487 dual absorbance detector and a Polymer Laboratories PLgel guard column followed by 2 PLgel 5mm Mixed-C columns in series at 40°C. Tetrahydrofuran (THF, Biosolve), stabilized with BHT, was used as eluent at a flow rate of 1 mL/min. The molecular weights were calculated against polystyrene standards (Polymer Laboratories, M_p = 580 Da up to M_p = 7.1 ^χ 10⁶ Da). All samples were filtered through a 0.2 pm PTFE filter (13 mm, PP housing, Alltech) before analysis. ¹H and ³C-NMR spectroscopy

'H and ¹³C-NMR spectroscopy was performed on a VARIAN Mercury 400 MHz NMR in CDCI₃. Data was acquired using VNMR software. Chemical shifts are reported in ppm relative to tetramethylsilane. Differential Scanning Calorimetry (DSC)

Differential Scanning Calorimetry (DSC) was performed on a TA Q100 DSC.

Approximately 5 mg of dried polymer was weighed into aluminum hermetic pans.

Temperature profiles from -50 °C up to 130 °C with a heating and cooling rate of 10

"C/min were applied. TA Universal Analysis software was used for data acquisition. Melting and crystallization temperatures were determined from the second heating run and the first cooling from the melt respectively. IR measurements

IR measurements were performed on a PerkinElmer SpectrumOne FT-IR Spectrometer, using a universal ATR sampling accessory. Kinetic attenuated total reflection Fourier Transform InfraRed (ATR-FTIR) spectroscop measurements were performed on a Bio- Rad Excalibur FTS3000MX infrared spectrometer using the golden gate setup. 32 scans were recorded per spectrum with a resolution of 4 cm^'1. Data was acquired using Varian Resolutions Pro. UV-curing measurements were performed under nitrogen atmosphere using a Driel Spectral Luminator with a wavelength of 365 nm. UV curing was performed using a Philips HPR-125 mercury discharge lamp with an output of 125 W/ 1.15 A.

Gel content

The gel content of the cross-linked polymers was determined with soxhlet extraction in chloroform.

Scanning electron microscopy

Scanning electron microscopy was performed on a FEI Quanta 3D FEG in the high vacuum mode. Samples were coated with gold before measurements. Examples 1-4 : enzymatic ring opening copolymerizaHon of Globalide and 1,5- dioxepan-2-one

1.08 g of a stock solution of Globalide (Gl) and 1 ,5-dioxepan-2-one (DXO) in mass feed ratios (GI:DXO) of 91:09; 73:27; 50:50 and 28:72, respectively and dried toluene (0.90 g) was prepared and dried overnight at 40 "C in the presence of 4 A molecular sieves.

Novozym 435 (40 mg) was dried in the reaction flask over molsieves at 40 °C overnight in a vacuum oven. After drying, the oven was opened under nitrogen flow. The flask containing Novozym 435 was placed in an oil bath at 60 °C and the stock solution was added.

The amount of water in the reaction mixture was measured at the beginning of the reaction using Karl-Fischer Coulometry. The water content of all reactions was below 0.07 mg water per gram of reaction mixture. Water present in the enzyme was used as initiator After several hours the viscosity of the mixture prevented proper stirring. After 24 hrs dichloromethane was added to the reaction mixture to redissoive the product. After filtering off the enzyme, the filtrate was precipitated in cold methanol. The polymer was dried at room temperature in vacuum. Typical yields were between 80 - 90 % after precipitation.

The obtained polymer was a poly(Globalide-co-1,5-dioxepan-2-one), (i.e. P(GI-co-DXO)) with different built-in GI:DXO ratios. See Table 1.

Examples 5-8 : enzymatic ring opening copolymerizations of Gfobalide and 4- ethyl caprolactone

The procedure of examples 1-4 was repeated for stock solutions comprising Globalide (Gl) and 4-methyl caprolactone (4MeCL) in mass feed ratios (GI:4 eCI) of 89:11; 73:27; 50:50 and 25:75, respectively. The remainder of the procedure was identical as to the procedure described in examples 1-4.

Examples 9-11: enzymatic ring opening opolymerizations of Ambrettolide and 2- oxo-crown-4-ether.

The procedure of examples 1-4 was repeated for stock solutions comprising Ambrettolide (Am) and 2-oxo-12-crown-4-ether (OC) in mass feed ratios (Am:Oc) of 77:23; 44:56 and 27:73, respectively. The remainder of the procedure was identical as to the procedure described in examples 1-4, except for the precipitation step after filtering off the enzyme. Precipitation of the polymers containing 2-oxo-12-crown-4-ether (OC) was troublesome in methanol and therefore these copolymers were precipitated in hexane.

Copolymers containing more than 50 % ambrettolide or globalide were obtained as white powders, copolymers containing less than 50 % of one of these monomers were obtained as oils. Due to the different reactivity ratios of the monomers all polymerizations were carried out over at least 24 hours to ensure incorporation of both monomers and complete randomization of the polymer.

Comparative examples A-E : enzymatic ring opening homopolymerizations of Globalide.Ambrettolide, 1,5-dioxepan-2-one, 4-methyl caprolactone and 2-oxo- crown-4-ether, respectively.

The procedure of examples 1-4 was repeated for stock solutions comprising Globalide, Ambrettolide, 1 ,5-dioxepan-2-one, 4-methyl caprolactone or 2-oxo-crown-4-ether as a single monomer, respectively. The remainder of the procedure was identical as to the procedure described in examples 1-4, All polymers were characterized by N R, DSC and SEC, and the results are summarized in Table 1. Relatively consistent molecular weights were obtained, with the exception of 44 kg/mol for P(GI-co-DXO) (50/50). The variation in the obtained molecular weights is probably due to varying amounts of water in the polymerization mixture which acts as a polymerization initiator. Polydispersities are all around 2, which is expected for enzymatic ring opening polymerizations (eROP) because lipases are transesterification catalysts.

P(GI-co-4MeCL) (Examples 5-8) was prepared using the same ratios as the P(GI-co- DXO) set (Examples 1-4). Here, the incorporation ratio was also in agreement with the feed ratio. All molecular weights obtained were around 17 kg/mol, with the exception of P(GI-co-4MeCL), for which only a mass of 6 kg/mol could be obtained.

Despite the same polymerization procedure, the molecular weights of OC-containing copolymers were significantly lower than those of the other copolymers. One probable cause can be the hydrophilicity of OC. It is more difficult to dry the reactants which causes a higher water amount in the reaction mixture and consequently a lower molecular weight.

All copolymers were obtained as random copolymers. NMR results:

Polyglobalide: H-NMR: δ (ppm) = 5.53 - 5.29 (m, 2H, CH=CH), 4.06 - 4.03 (m, 2H,

C=OOCH₂), 2.30 - 2.26 (m, 2H, CH₂C=00), 2.12 - 1.92, 1.71 - 1.59, 1.25 (m, 22H. CH₂).

"C-NMR: δ (ppm) = 173.8 (C=0), 133.5, 132.9, 131.5, 131.0, 128.6, 128.1, 125.0

(CH=CH), 63.9. 63.7 (C=OOCH₂), 34.4 (CH₂C=00), 32.6, 32.5 (CH₂-CH=CH-CH₂), 32.0

(C=OOCH₂-CH₂), 29.6 - 25.0 (CH₂).

Polyambrettolide: ¹H-NMR: δ (ppm) = 5.38 - 5.36 (m, 2H, CH=CH), 4.06 - 4,03 (t, 2H,

C=OOCH₂), 2.30 - 2.26 (m, 2H, CH₂C=00), 1.96, 1.63 - 1.57, 1.34 - 1.29 (m, 24H, CH₂).

¹³C-NMR: δ (ppm) = 173.8 (C=0), 131.0, 130.5 (CH=CH), 64.2 (C^OCHz), 34.8

(CH₂C=00), 31.7 (CH₂-CH=CH-CH₂), 31.3 (C=OOCH₂-CH₂), 29.5 - 24.9 (CH₂).

Poly(1,5-dioxepan~2-one): ¹H-NMR: δ (ppm) = 4.30 - 4.28 (t, 2H, C=OOCH₂), 3.66 - 3.64 (m, 2H, CH₂0), 2.56 - 2.54 (m, 2H, CH₂C=0). ¹³C-NMR: δ (ppm) = 171.3 (C=0). 68.7

(C=OOCH₂), 66.4 (OCH₂-CH₂C=0), 63.4 (C=OOCH₂-CH₂), 34.8 (CH₂C=00).

Poly(4-methylcaprolactone): ¹H-NMR: δ (ppm) = 4.13 - 4.06 (m, 2H, C=OOCH₂), 2.34 -

2.27 (m, 2H, CH₂C=0). 1.73 - 1.41 (m, 6H, CH₂, CH, CH₃).

Poly(2-oxo-12-crown-4-ether): H-NMR: δ (ppm) = 4.24 (m, 2H, C=OOCH₂), 4.12 (s, 2H, CH₂C=0), 3.70 - 3.55 (m, 10H, CH₂0). Poly(globalide-co-1,5-dioxepan-2-one) (55/45): 'H-N R: δ (ppm) = 5.51 - 5.28 (m, CH=CH), 4.28 - 4.00 (m, C=OOCH₂), 3.68 - 3.63 (m, CH₂0), 2.55 - 2.23 (m, CH₂C=00), 2.11 - 1.20 (m, CH₂). ¹³C-NMR: δ (ppm) = 174.0 (C=0, Gl), 171.4 (C=0, DXO), 133.7, 133.5, 131.6, 131.5, 128.6, 128.1 , 125.0, 124.8 (CH=CH), 68.9 (C=OOCH₂, DXO), 66.6 (OCH₂-CH₂C=0) 64.3 - 63.2 (C=OOCH₂, Gl), 35.0, 34.8, 34.4. 34.1 (CH₂C=00, DXO- DXO, DXO-GI, GI-DXO. Gl-GI), 32.6, 32.5 (CH₂-CH=CH-CH₂), 32.0 (C=OOCH₂-CH₂). 29.5 - 24.9 (CH₂).

Poly(globalide-co-4-methylcaprolactone) (53/47): 'H-N R: δ (ppm) = 5.55 - 5.30 (m, CH=CH), 4.15 - 4.02 (m, C=OOCH₂). 2.35 - 2.25 (m, CH₂C=00), 2.13 - 1.22 (m, CH₂, CH, CH₃). ¹³C-NMR: δ (ppm) = 173.9, 173.7 (C=0), 133.6, 133.5, 131.5, 131.1 , 129.0. 128.1. 125.0 (CH=CH), 63.9. 63.7. 62.6. 62.4 (C=OOCH₂. 4MeCI-4MeCI. 4MeCI-GI. Gl- 4 eCI, Gl-GI), 35.2. 34.3

CH₂). 29.6 - 23.5 (CH₂, CH), 19.0 (CH₃).

Poly(ambrettolide-co-2-oxo-12-crown-4-ether)) (37/63): H-NMR: δ (ppm) = 5.39 - 5.36 (m, CH=CH), 4.25 - 4.01 (m, C=OOCH₂. CH₂C=00 from OC), 3.71 - 3.54 (m, CH₂0), 2.32 - 2.25 (m, CH₂C=00 Am), 1.93 - 1.32 (m. CH₂). ¹³C-NMR: δ (ppm) = 177.0, 174.0 (C=0), 130.5, 130.1 (CH=CH), 75.6 (CH₂C=00). 71.1 , 69.9, 68.6 (COCC) 65.7 - 63.3 (C=OOCH₂, COCC), 34.4. 34.2, 32 5. 32.4 (CH₂C=00, OC-OC, OC-Am, Am-OC, Am- Am), 29.6 - 24.8 (CH₂).

Table 1 Properties of polyesters obtoined via eROP of macrolactones.

Example Polymer Feed Ratio M * PDI* T "

ratio (NMR) [g mol] ro I' )

Im/m-%] Im/m-%]

1 P(GI-co-DXO) 91/09 91/09 11 000 2.5 42.4 26.8

2 P(Gi-co-DXO) 73/27 71/29 13 000 1.8 35.3 17.9

3 P(GI-co-DXO) 50/50 55/45 44 000 1.8 26.6 5.6

4 P(GI-co-DXO) 28/72 29/71 22 000 2.0 0.3 -25.0

5 P(GI-co-4 eCL) 89/11 89/11 17 000 2.5 40.9 23.8

6 P(GI-co-4 eCL) 73/27 75/25 17 000 2.0 36.5 18.3

7 P(GI-co-4 eCL) 50/50 53/47 18 000 2.2 22.0 0.1

8 P(GI-co-4 eCL) 25/75 33/67 6 000 2.0 - -

9 P(Am-co-OC) 77/23 71/29 14 000 2.0 53 34

10 P{Am-co-OC) 44/56 37/63 8000 1.6 37 17 11 P(Am-co-OC) 27/73 27/73 8 000 1.4 35 16

A PGI - - 24 000 1.9 46.2 29.8

B PAm - - 24 000 1.9 54.9 37.7

C PDXO - - 6000 2.1 - -

D P4MeCL - - 4 000 1.5 - -

E OC - - 1 000 1.57 - -

"Determined by SEC in THF at 40 °C (polystyrene standards), determined from the second DSC heating run. 'Determined from the first cooling run.

Thermal properties

In Table 1 all melting points (T_m) and crystallization temperatures (T_c), as determined with Differential Scanning Calorimetry (DSC), of the copolymers with different compositions are listed.

Homopolymers of DXO, 4MeCL and OC are amorphous and no melting point could be observed. All copolymers that are not completely amorphous show a single crystallization and melting process.

The difference between T_m and T_c is very small for all polymers (ca. 20 °C), which indicates that crystallization in these materials is fast.

All glass transistion temperatures ( are far bejow 0 °C. The T_g of PCL, a less flexible material, is known to be -60 °C, all copolymers have a T_g < -60 "C. However, these could not be measured due to the limit of the DSC (-60 ^eC).

A trend can be observed in the melting (ranging from 42.4 to 0.3 °C) and crystallization temperatures of P(GI-co-DXO) with different compositions (examples 1-4), namely the higher the amount of DXO incorporated, the lower the melting and crystallization temperature. Explanation may be found in the fact that the homopolymer of DXO is completely amorphous, and incorporation of this monomer will lower the melting and crystallization temperature.

The copolymer of example 8 (copolymer with mass feed ratio(GI:4MeCI) of 25:75) is the only in this set which is obtained as a completely amorphous copolymer. No melting and crystallization temperatures could be observed. Moreover, no glass transition could be observed either, which is probably due to the limit of the DSC (-60 °C).

All other GI-4MeCL copolymers show melting transitions and here the same trend is observed as for the DXO copolymers, namely the incorporation of 4MeCL results in lowering of the melting and crystallization temperature. The observed influence of incorporation of OC on the melting temperature is the same as for the other comonomers, DXO and 4MeCL. Examples 12-16 thermal cross/inking of poly(globalide-co-e-caprotactone) with dicumylperoxide.

In order to determine the lowest amount of cross-linkable comonomer needed to obtain a fully cross-linked material, a series of copolymers was made according to the procedure given in Examples 1-4, with globalide and ε- caprblactone (see label 2). These copolymers show the same polymerization behavior and trends as the other copolymer sets reported earlier in this chapter. All polymers were obtained as random copolymer. Moreover, the molecular weights are representative for the ones obtained with the other copolymer systems (compare Tables 1 and 2). The amount of globalide was varied between 10 and 50 m/m-%.

For the cross-linking, all materials were melted, mixed with dicumyl-peroxide in the melt and cured at 150 °C for 30 min. Afterwards the cured materials were cooled to room temperature.

The amount of cross-links in the material was estimated with IR-spectroscopy. The signal of the trans unsaturated carbon-carbon bond around 970 cm ¹ decreased upon curing and was used to monitor the consumption of double bonds in the process. A clear trend can be seen in the percentage of cross-linked monomer residues in the polymer. The lower the amount of residual cross-linkabie comonomer, the higher the percentage of reacted double bonds (58 % for P(GI-co-CL) (10/90) to 13 % for P(GI-co-CL) (40/60)). This is probably due to the formation of the network itself. As soon as there are some bonds cross-linked, the peroxide is no longer able to penetrate the material and cross-linking stops. When the amount of Gl decreases, the network becomes less dense and this facilitates a continuous and more complete cross-linking of the unsaturated bonds in the network.

The gel content of the cross-linked polymers was determined with soxhlet extraction in chloroform. In Table 2 it can be seen that 10 m/m-% of cross-linkable monomer is sufficient to obtain a fully cross-linked material. This means that at least 90 % of an monomer can be incorporated without losing the advantages of a cross-linked system, the advantages being the possibility to make completely amorphous and easily degradable materials with adjustable modulus by changing the amount of cross-linked double bonds. In literature it was described that monomers like DXO could only be incorporated up to 20 m/m-% in non-cured systems, due to the unfavorable influence on the melting point. However, incorporation of a higher amount of comonomer is desirable as it can enhance the hydrophilicity and thus degradation properties of materials significantly. By cross- linking the material can be tuned to the desired properties without taking the melting point and crystallinity into account. This opens various possibilities for degradable biocompatible polymers.

Table 2 Cross-linking properties of copolymers containing glabalide and caprolactone.

Feed Ratio Mn ^b PDI Reacted Gel Reacted Gel

Entry GI/CL Gi/CL* C=O^d content content*

[g/mol] H [m/m- [w/w- [m/m- [w/w-

[m/m-%] [m/m-%] %] %] %] %]

12 50/50 47/53 20 000 3.3 n.d. 99.2 23 89

13 40/60 40/60 21 000 3.0 13 96.8 37 88

14 30/70 32/68 23 000 2.4 16 97.5 43 82

15 20/80 22/78 18000 2.8 33 96.2 25 79

16 10/90 10/90 16000 2.4 58 94.5 11 33

"Determined with ^!H-NMR, ^Determined bij SEC in tetrahydrofuran, ^Determined by IR, "^Cross-linked using thermal methods, 'Cross-linked using thJol-enc chemistry. examples 17-21 crosslinking of polyfglobalide-co-e-caprolactonej using thiol-ene chemistry

Examples 12-16 were repeated (examples 17-21, respectively) utilizing ethylene glycol bis(3-mercaptopropionate) as a dithiol and 4-hydroxybenzophenone as a UV initiator. Different results were obtained in comparison with the thermal cross-linker (dicumyl peroxide). Cured films containing≥ 20 m/m-% cross-linkable monomer were obtained as transparent films, only experiment 21 (equivalent to experiment 16, see table 2) was obtained as an opaque film.

The amount of reacted C=C bonds was determined by IR spectroscopy.

Using this curing system, no trend could be observed in the relation of cross-link density and amount of cross-linkable monomer. In contrast to the thermal curing, where the double bonds are connected directly, using the thiol-ene reaction a flexible linker is placed between the double bonds. This creates a more open network where it is still possible for the dithiols to reach the unreacted double bonds. Noticeable was also that all reactions were finished within twenty minutes. The gel content of all copolymers was lower than when using the thermal cross-link system. However, the polymer was mixed with dithiol in a 2/1 mole ratio. Since only a percentage of the added dithiol is used in the cross-linking (with a maximum of 43 m/m-% in experiment 19, equivalent to experiment 14 in Table 2), the unreacted dithiol was still present as a small molecule in the cured film. These small molecules are washed out easily upon extraction. Taking this into account during calculation of the gel content (weight after extraction devided by the weight before extraction), a much higher gel content has been calculated. Only curing of experiment 21 (equivalent to experiment 16 in Table 2) did not result in complete network formation. Upon addition of low amounts of cross-linkable monomer, the crystallinity of the cured material can be tuned to a certain extend.

Manufacturing of porous materials

Materials

P(Am-co-DXO) was synthesized via the same route as described for all other copolymers in Examples 1-4.

Polyethyleneglycol (PEG) of different masses were purchased from Aldrich.

Sugar was purchased from Van Gilze and grinded to < 125 pm particles using a microgrinder purchased from IKA and a sieve with a threshold of 125 pm.

THF was purchased from Biosolve and used without further purification.

4-hydroxybenzophenone was purchased from Fluka and ethylene glycol bis(3- mercaptopropionate) was purchased from Wako. Example 22 Synthesis and characterization of a porous structure of polyambrettolide

Sugar, poly(ethylene glycol) (PEG) and different Pluronics® were used as porogen to make porous structures. Sugar was grinded to a crystal size of < 125 pm. Various concentrations of sugar were added to polyambrettolide, ranging from 1 to 10 grams of sugar per gram of polymer. The porogen was added to a concentrated unsaturated polyester solution in tetrahydrofuran. Adding more than 10 grams of sugar per gram of polymer resulted in disintegration of the scaffold.

All samples were UV-cured in the presence of ethylene glycol bis(3-mercaptopropionate) as a dithiol and 4-hydroxybenzophenone as an initiator. Due to the opaque character of the mixture, curing was performed using longer exposure times. The pore size and pore size distribution were analyzed with scanning electron microscopy (SEM). When water was added, swelling of the scaffolds was observed. The best results were obtained by adding a tenfold of sugar with respect to Pam. This provided the most open structure. However, interconnection of the pores is not easy to determine. Moreover, the pore size of these pores is difficult to determine due to the irregular shape of the pores.

Particulate leaching using PEG of different molecular wejghts (35 kg/mol and 600 kg/mol) using different ratios of porogen/polymer (ranging from 0.5 - 8) resulted in molecular mixing of the porogen with the polymer. Micropores instead of the desired macropores were formed.

Example 23 Synthesis and characterization of a porous structure of polyambrettolide-co-DXO.

Degradable scaffolds were prepared with P(Am-co-DXO) (50/50) and using sugar as porogen. These scaffolds were prepared in the same way as for the homopolymer (see example 22). Analyzing the scaffold with SEM revealed the open structure of the scaffold. It could also be observed that the to part of the scaffold has a more dense structure than the inside, what would be ideal for nerve guidance channels. A continuous porous network is formed in the interior of the material.

Biocompatibility

During construction of porous scaffolds a dithiol and a UV initiator were used to cross-link the material (see examples 22 and 23). After curing the structures were put in THF for 4 hours to leach out the unreacted dithiol and residual initiator. Afterwards the materials were put in water during several days to leach out the porogen. However, it cannot be excluded that some unreacted agents remain in the material, which could have an influence on the biocompatibility. Therefore cytotoxicity tests were performed on the porous materials and on the cross-linked homopolymer.

The porous scaffold was found to be less toxic than the cured homopolymer. The latter was cured via the thermal method while the porous material via UV-curing. Keeping in mind that the porous polymer is a copolymer and the solid material a homopolymer, UV- curing seems to result in lower toxicity than the thermal method. Due to the low toxicity of the uncured polymer, the curing step must be responsible for the toxicity. One possibility is that despite the leaching in THF, there are still rests of curing agents in the scaffolds. Another explanation might be the toxicity of the scaffold itself. By introducing thiol ethers in the polymer, toxicity may be introduced.

Degradation

Enzymatic degradation of aliphatic polyesters is determined by the rate of surface erosion. Therefore the surface per unit volume has a large influence on the degradation rate. By creating porous materials this surface area per unit volume is increased significantly Degradation experiments were conducted by putting 100 mg of P(Am-co-DXO) (50/50) scaffold in Phosphate Buffered Saline (PBS) with 0.1 mg/mL Lipase PS and a pH of 7.4. Swelling of the material was observed and degradation was faster compared to the solid polymer. Within 100 days the solid material showed only 30 % degradation, whereas 30 % of the porous material degraded in 55 days. The degradation curves both seem linear. Although only 40 % degradation is obtained within 70 days, it is expected that full degradation can be reached.

Claims

1. A copolymer from at least an unsaturated macrolactone A and a lactone B,

wherein the lacton B are chosen from the group consisting of ε-caprolactone (CL), substituted lactones and lactones that contain one or more etherlinkages in the lactone ring.

2. The copolymer according to claim 1, wherein the lactone B is chosen from the group consisting of substituted lactones and lactones that contain one or more etherlinkages in the lactone ring.

3. The copolymer according to claim 1 or 2, wherein the macrolactone A is optionally substituted and has a ring size ranging from 10 to 50 carbon atoms.

4. The copolymer according to claim 1 or 2, wherein the macrolactone A is chosen from the group consisting of 5-tetradecen-14-olide, 11-pentadecen-15-olide, 12-pentadecen-15-olide (also known as globalide), 7-hexadecen-16-olide (also known as ambretto!ide), and 9-hexadecen-16-olide.

5. The copolymer according to anyone of claims 2-4, wherein the lactone B is chosen from the group consisting of, 4-methyl caprolactone (4MeCL), 4-ethylcaprolactone, 1 ,5-dioxepan-2-one (DXO), 1 ,4-dioxan-2-one, 10-oxahexadecanolide,

11-oxahexadecanolide, 12-oxahexadecanolide, 12-oxahexadecen-16-oIide and 2- oxo-12-crown-4-ether (OC).

6. The copolymer according to anyone of claims 2-4, wherein the lactone B are

selected from the group consisting of 1 ,5-dioxepan-2-one (DXO) and 2-oxo-12- crown-4-ether (OC).

7. The copolymer according to any one of claims 2-3, wherein the macrolactone is globalide (Gl) or ambrettolide (Am) and the lacton B is selected from the group consisting of 1 ,5-dioxepan-2-one (DXO) and 2-oxo-12-crown-4-ether (OC).

8. A copolymer according to any one of claim 1-7 for use in medical applications.

9. A process for preparing a copolymer comprising the steps of:

a. providing an unsaturated macrolactone monomer A and a lactone B

optionally in a solvent;

b. subjecting said monomers to an enzyme catalyzed ring opening

polymerization using a lipase type enzyme as a catalyst.

10. The process according to claim 8, wherein the lipase type enzyme is selected from the group consisting of Candida Antarctica Lipase B (CALB), Pseudomonas cepacia lipase, preferably Candida Antarctica Lipase B (CALB).

11. The process according to any one of claims 9-10, wherein said ring opening polymerization is performed in the presence of a solvent, wherein the solvent is selected from toluene, THF, xylene, 1 ,2-dichloroethane, benzene, cyc!ohexane, chloroform, dioxane, heptanes and acetonitrile, preferably the solvents are chosen from toluene. THF, xylene and cyclohexane.

12. A process for preparing a porous copolymer structure comprising the steps of a. Providing the copolymer according to anyone of claims 1-7

b. Mixing the copolymer with a porogen and crossiinking agent

c. Crossiinking the copolymer

d. Leaching the porogen with a leaching solvent

3. The process according to claim 12, wherein the porogen is a medicall accepted compound, preferably sugar or a salt.

14. The process according to claim 12 or 13, wherein the crossiinking is performed using a dithiol, preferably ethylene glycol bis(3-mercaptopropionate) as a crosslinker and a UV initiator, and wherein the crossiinking reaction is initiated by exposure to UV radiation.

15. Nerve guide tubes comprising a crosslinked porous structure of a copolymer according to any one of the claims 1-7.