Project

Poly(2-oxazoline) drug conjugates

Duration
01 January 2014 → 31 December 2017
Funding
Regional and community funding: IWT/VLAIO
Research disciplines
  • Natural sciences
    • Organic chemistry
Keywords
Poly(2-oxazoline) clinic imino ethers
 
Project description
During this PhD we were able to set important steps forward towards a better understanding of the cationic ring-opening polymerization (CROP) of cyclic imino-ethers (Figure 7). While we made significant progress in this area, there is still a long road ahead to fully understand the CROP of cyclic imino-ethers.
Figure 119: General simplified overview of the CROP of 2-oxazolines.
This PhD research was done in two stages:
In the first stage, we took a closer look at monomers with an unexpected (co)polymerization behavior. We were able to find a correlation between the functionality present on the monomer
side-chains and the observed propagation rate constants (kp). All of the faster propagating monomers exhibited functional groups that have the ability to interact at the reactive polymer
chain end (the 2-oxazolinium) and/or on monomer stage resulting in an increased reactivity for the living polymerization. We were able to explain the reactivity of existing monomers but also rationally developed new monomers with control over their reactivity. This ultimately led to the successful design and synthesis of different linear copolymers with different monomer
distributions (random, gradient and (quasi) diblock).
In the second stage of this PhD, we focused on the influence of monomer ring-size on (co)polymerization between 2-oxazolines (five-membered ring), 2-oxazines (six-membered
ring) and also 2-oxazepines (seven- membered ring). The different ring-sizes are expected to result in differences in kp, originating from differences in both nucleophilicity of the monomer and of conformation-induced sterical hindrance (or stereo-electronic effects).
Chapter 6
In Figure 120 we provide a complete overview of all monomers that were investigated during this PhD thesis.
Figure 120: Overview of all utilized monomers in this PhD thesis. 2-methyl-2-oxazoline (MeOx) 1, 2-ethyl-2-oxazoline
(EtOx)2, 2-n-butyl-2-oxazoline (n-ButOx) 3, 2-(but-3-enyl)-2-oxazoline (ButenOx) 4, 2-(but-3-ynyl)-2-oxazoline (ButynOx) 5,
2-(penta-4-ynyl)-2-oxazoline (PentynOx) 6, 2-(dec-9-enyl)-2-oxazoline (DecenOx) 7, 2-decanyl-2-oxazoline (DecanOx) 8, 2-
(pent-4-en-2-yl)-2-oxazoline (α-Me-ButenOx) 9, 2-(2-methylpent-4-en-2-yl)-2-oxazoline (α,α-diMe-ButenOx) 10, 2-methyl-2-
oxazine (MeOzi) 11, 2-n-propyl-2-oxazozine (n-PropOzi) 12, 2-phenyl-2-oxazoline (PhOx)13, 2-phenyl-2-oxazine (PhOzi) 14
and 2-phenyl-2-oxazepine (PhOpi) 15.
Stage 1: functional side-chains and their effect on (co)polymerization rates By comparing the CROP of 2-n-butyl-2-oxazoline (n-ButOx) 3, 2-(but-3-enyl)-2-oxazoline
(ButenOx) 4, 2-(buta-4-ynyl)-2-oxazoline (ButynOx) 5 and 2-(penta-4-ynyl)-2-oxazoline (PentynOx) 6 under standard CROP conditions, we could study the effect of unsaturations in
the hydrocarbon side-chains on the kp. It was hypothesized that the unsaturated side-chain can have a cation-π interaction with cationic intermediates and transition states, resulting in a
significant reduction of the activation energy for incorporating a neutral monomer, by favorable interactions coming from the side-chain of the incoming monomer (pre-reactive complex) or
the side-chain of the 2-oxazolinium polymer chain end. This hypothesis was confirmed by preliminary DFT modeling data provided by the center of molecular modeling (CMM). The
highest rate acceleration in this set of monomers was observed for the alkynes. This can be explained by a larger available surface of π-electron density, which makes interacting
conformations with the cation of the 2-oxazolinium more likely. This rationale further fits with the observed clear chain length dependency, as ButynOx 5 has a higher kp then PentynOx 6.
This chain length dependency is indeed most readily explained by a favorable interaction between the separate functions, in the form of a complex between the unsaturation and cation
of the 2-oxazolinium. By including 2-(dec-9-enyl)-2-oxazoline (DecenOx) 7 and 2-decanyl-2-oxazoline (DecanOx) 8 in this study, we hoped to gain a better insight into the chain length
dependency for longer alkenes and more general for the cation-π interaction. Unfortunately, due to (in all probability) solubility issues of DecenOx 7 and especially DecanOx 8 no real
conclusions could be drawn from these experiments, even after lowering the monomer concentration to 2 mol/L. However it was noted in these experiments that the kp of ButenOx 4
is concentration dependent, which can be explained by the previously mentioned pre-reactive complex or that the alkenes in the solution can undergo the cation-π interactions acting as
accelerating ‘cosolvent’. Next to intramolecular π-cation separation, also intermolecular πcation separation has a significant kinetic effect.
The above kinetic studies resulted in a better understanding of the properties of side-chain interactions towards the CROP, leading to a general proposal to explain this type of interactions, with also some predictive value for the homopolymerization.
Besides these kinetic studies, we also optimized the synthesis of ButenOx 4, as significant quantities of ButenOx 4 were required. The literature described a multistep synthesis route by Schlaad et al. with an overall yield of 41% utilizing rather expansive chemicals and exhibit low atom efficiency of ButenOx.1 This multistep literature synthesis route of ButenOx 4 was replaced with an one step synthesis (α-deprotonation route) with 67% yield. In addition a significant higher atom economy and utilizing less expansive chemicals was also achieved.
In the future it would be interesting to study 2-oxazolines with different alkene side chain lengths, to determine the chain length dependency of the cation-π interaction for these type of
unsaturated monomers. While the first steps are already undertaken, computational modeling would be a highly interesting addition to gain an even further understanding of the cation-π interaction. By combining these data, new type of monomers that possibly could undergo cation-π interaction could be identified. In addition copolymers between the unsaturated monomers and for instance 2-methyl-2-oxazoline (MeOx) 1 and 2-ethyl-2-oxazoline (EtOx) 2 could be highly interesting for kinetic studies or for materials.
Unsaturated side-chains with α-substitutions As a next step in the first stage, ButenOx 4 was compared to mono and double α-methylated ButenOx, 2-(pent-4-en-2-yl)-2-oxazoline (α-Me-ButenOx) 9 and 2-(2-methylpent-4-en-2-yl)-2-oxazoline (α,α–diMe-ButenOx) 10, respectively. This α-substitution was found to be a highly efficient way of lowering the kp of ButenOx 4. The kp of α-Me-ButenOx 9 dropped roughly to the same kp as n-ButOx 3, indicating that for α-Me-ButenOx 9 the unfavorable steric repulsion effect and the favorable electronic (cation-π) effect balanced each other. By further increasing the substitution, α,α–diMe-ButenOx 10, the decrease was even more pronounced, as the kp drops with a factor of almost 200, when compared to ButenOx 4. The resulting new polymers,
poly(α-Me-ButenOx) and poly(α,α–diMe-ButenOx) exhibit an increased glass transition temperature (Tg) comparable to poly(ButenOx). While poly(ButenOx) and poly(α-MeButenOx) did not exhibit a melting temperature (Tm) poly(α,α–diMe-ButenOx) did exhibit a slow recurring one after annealing at elevated temperatures.
The ButenOx type monomers (ButenOx 4, α-Me-ButenOx 9 and α,α–diMe-ButenOx 10) were copolymerized with MeOx 1 and EtOx 2. Similar to 2-methoxycarbonylethyl-2-oxazoline (C2-MestOx) and 2-methoxycarbonylpropyl-2-oxazoline (C3-MestOx) reported in literature,2 an inversion of the incorporation rate in the copolymerization with MeOx 1 and EtOx 2, compared to their homopolymerization, was observed for ButenOx 4. In general a slower incorporation of the cation-π undergoing monomer was observed in the copolymerization with MeOx 1 and EtOx 2 when compared to their relative ratio during the homopolymerization. This could also be explained in similar fashion as for C2-MestOx, C3-MestOx, ButenOx 4, α-Me-ButenOx 9 and α,α–diMe-ButenOx 10 the nucleophilicity of the imine is lowered, due to the electron withdrawing side-chain, when compared to MeOx 1 and EtOx 2.
As anticipated, and as required for our goal of controlling copolymer architectures, the different substitution grades of the α-position resulted in a broad range of different copolymers derived from the ButenOx type monomers at the one hand, and MeOx 1 or EtOx 2 at the other hand, ranging from amphiphilic copolymers with random, gradient with different gradient slopes, to quasi diblock linear architectures. By increasing the concentration of alkene containing monomer in the copolymerization an increase in the kp of both monomers is achieved, while retaining their apparent reactivity ratio (rapp).3 While the homopolymers of ButenOx 4, α-MeButenOx 9 and α,α–diMe-ButenOx 10 all exhibited a lower Tg then MeOx 1 and EtOx 2 the influence was minimal on the copolymers as the mol% were deliberately kept low. The water solubility of the copolymers was also minimal influenced in these copolymers. This means that we were successful in synthesizing linear polymer with different monomer distributions, but with the same alkene functionality without alternating the final copolymer properties too much.
This alkene functionality can be used to introduce a wide range of other functionalities (e.g.through thiol-ene), all of which will be placed along the polymer chain with a predictable
gradient.
The first steps towards applications of our newly developed controlled synthesis of functional poly(2-oxazoline) copolymers have been taken in collaboration with other groups (this work is
not included in this thesis).4,5 However, these applications are, at the time of writing, limited to ButenOx 4 containing copolymers. In future work, it would be highly interesting to further explore these copolymers to gain a better understanding of the influence of the obtained gradients copolymers on different applications, as it is known that gradient copolymers exhibit different behavior then block and random copolymers.6
Stage 2: monomer ring size its effect on (co)polymerization rates
In the second stage of this PhD we synthesized gradient and random copolymers based on 2-oxazoline (MeOx 1 and EtOx 2) and 2-oxazine monomers. While the initiation rates of 2-
oxazines, due to their higher nucleophilicity, is significantly higher when compared to 2-oxazolines, the conformationally induced steric hindrance (sofa conformation) of 2-oxazines
results in an at least four times slower homopolymerization compared to 2-oxazolines.7 This lower monomer electrophilicity (a steric effect in its 2-oxazinium state towards incoming monomer) outweighs the increased nucleophilicity in the homopolymerization of 2-oxazines (i.e. the CROP of 2-methyl-2-oxazine (MeOzi) 11 is slower than the CROP of both MeOx 1 or EtOx 2). However, we were able to determine that in the copolymerization between 2-oxazines and 2-oxazolines, the 2-oxazines exhibit a faster incorporation rate than 2-oxazolines. Such copolymerizations yield gradient copolymers, indicating that the higher nucleophilicity of 2-
oxazines is dominant, and both ring size monomers suffer from a less electrophilic chain end.
Indeed, the extra steric hindrance of the 2-oxazines has only a secondary effect on the kp’s in the copolymerization, which can be seen in the copolymerizations with EtOx 2. For this reason these unexpected but useful findings were patented.8 By also exploiting the α-substitution effect, going from MeOzi 11 to 2-iso-propyl-2-oxazine (i-PropOzi 12), the incorporation rate of the 2-oxazine could be reduced in the copolymerization with MeOx 1 or EtOx 2, resulting in (near) random copolymers containing 2-oxazoline and 2-oxazine. It has to be noted that for the polymerization of i-PropOzi 12 and MeOx 1 the reactivity even switches, MeOx 1 is faster incorporated then i-PropOzi 12, when compared to the rest.
The first steps towards an application for the 2-oxazoline - 2-oxazine copolymers have been made in collaboration with the University of Twente. After the hydrolysis of the side-chain
amide bonds, the expected polyamine polymers were obtained, which are at this time being
evaluated for their use as gene delivery vectors. In future work it would be interesting to see if the 2-oxazoline that is functionalized by a cyclopropyl would increase its incorporation rate of the 2-oxazoline relative to 2-oxazines. Due to time constrains, higher cost (as for the application
we have in mind the side-chains are hydrolyzed off) and the inherent instability of 2-cyclopropyl-2-oxazoline this was not yet achieved. In theory this would also yield (near) random
copolymers. However, an extra benefit would be that the overall polymerization time to full conversion could be reduced.
Finally, to gain a better understanding in the ring size influence on the kp we were interested to expand the known polymerized cylic imino-ethers (2-oxazolines and 2-oxazines) with 2-oxazepines. In collaboration with the organic synthesis group, we were able to synthesize for the first time a 4, 5, 6 and 7 unsubstituted 2-oxazepine (2-phenyl-2-oxazepine (PhOpi) 15).
Although, these PhOpi 15 compounds were already reported by multiple literature reports,9–14 we found unambiguous proof, utilizing NMR and single crystal X-ray diffraction, that none of these authors succeeded in synthesizing an 2-oxazepine ring, but rather misidentified a simple acylated pyrrolidine ring, which is considerably less strained, and is also completely unreactive in CROP. Using a new approach, such novel seven-membered ring monomers were prepared
on gram scale, allowing for the first time to investigate their polymerization behavior in CROP.
Due to the small scale, suitable conditions for the CROP of this PhOpi 15 (and two references:
2-phenyl-2-oxazoline (PhOx) 13 and 2-phenyl-2-oxazine (PhOzi) 14) were found to be in chorobenzene as solvent at 120°C, allowing online NMR monitoring of reaction progress. In
these polymerizations, PhOx 13 and PhOzi 14 polymerized as expected. While PhOzi 14 is already 7.5 times slower than PhOx 13 in these conditions, a much more pronounced effect was found for PhOpi 15. PhOpi 15 polymerized almost 550 times slower than PhOx 13, indicating a significant increase in reaction barriers. This extremely slow polymerization of PhOpi 15 could be explained by severe conformational effects in the monomer, also observed in the structure determined by X-ray diffraction. Compared to 2-oxazolines and 2-oxazines, the lone
pairs on the imino-ether oxygen atom are forced in an out-of-plane conformation, preventing oxygen-to-imine conjugation in the seven-membered ground state. In order to arrive at a
reactive conformation, significant ring strain needs to be introduced, strongly affecting the required energy to reach a reasonable transition state wherein a significant ‘amide bonding’ character can be present (requiring a planarization around the oxygen atom). This planarization of the imino-ether moiety is a given in the ground states of 2-oxazines and 2-oxazolines, but is completely absent in the strained 2-oxazepines.
It would be very interesting to study, in the future, the material properties of poly(2-oxazepine)s and compare these to equivalent poly(2-oxazoline)s and poly(2-oxazine)s. In addition
copolymerizations between 2-oxazoline or 2-oxazine with 2-oxazepines would also be very interesting. This could shed further light on the importance of the conformation induced sterical hindrance.
References
(1) Gress, A.; Völkel, A.; Schlaad, H. Thio-Click Modification of poly[2-(3-Butenyl)-2-Oxazoline].
Macromolecules 2007, 40 (22), 7928–7933 DOI: 10.1021/ma071357r.
(2) Bouten, P. J. M.; Hertsen, D.; Vergaelen, M.; Monnery, B. D.; Catak, S.; Van Hest, J. C. M.; Van
Speybroeck, V.; Hoogenboom, R. Synthesis of poly(2-Oxazoline)s with Side-chain Methyl Ester
Functionalities: Detailed Understanding of Living Copolymerization Behavior of Methyl Ester Containing
Monomers with 2-Alkyl-2-Oxazolines. J. Polym. Sci. Part A Polym. Chem. 2015, 53 (22), 2649–2661
DOI: 10.1002/pola.27733.
(3) Puskas, J. E.; McAuley, K. B.; Chan, S. W. P. Fundamental Aspects of Measuring Copolymerization
Reactivity Ratios Using Real-Time FTIR Monitoring. Macromol. Symp. 2006, 243 (1), 46–52 DOI:
10.1002/masy.200651105.
(4) Dargaville, T. R.; Lava, K.; Verbraeken, B.; Hoogenboom, R. Unexpected Switching of the Photogelation
Chemistry When Cross-Linking Poly(2-Oxazoline) Copolymers. Macromolecules 2016, 49 (13), 4774–
4783 DOI: 10.1021/acs.macromol.6b00167.
(5) Kalaoglu-Altan, O. I.; Verbraeken, B.; Lava, K.; Gevrek, T. N.; Sanyal, R.; Dargaville, T.; De Clerck, K.;
Hoogenboom, R.; Sanyal, A. Multireactive Poly(2-Oxazoline) Nanofibers through Electrospinning with
Crosslinking on the Fly. ACS Macro Lett. 2016, 5 (6), 676–681 DOI: 10.1021/acsmacrolett.6b00188.
(6) Filippov, S. K.; Verbraeken, B.; Konarev, P. V.; Svergun, D. I.; Angelov, B.; Vishnevetskaya, N. S.;
Papadakis, C. M.; Rogers, S.; Radulescu, A.; Courtin, T.; Martins, J. C.; Starovoytova, L.; Hruby, M.;
Stepanek, P.; Kravchenko, V. S.; Potemkin, I. I.; Hoogenboom, R. Block and Gradient Copoly(2-
Oxazoline) Micelles: Strikingly Different on the Inside. J. Phys. Chem. Lett. 2017, 8 (16), 3800–3804
DOI: 10.1021/acs.jpclett.7b01588.
(7) Saegusa, T.; Kobayashi, S.; Nagura, Y. Isomerization Polymerization of 1,3-Oxazine. II. Kinetic Studies
of the Ring-Opening Isomerization Polymerization of Unsubstituted 5,6-Dihydro-4H-1,3-Oxazine.
Macromolecules 1974, 7 (3), 265–272 DOI: 10.1021/ma60039a003.
Chapter 6
246
(8) Hoogenboom, R.; Verbraeken, B. Poly(cyclic Imino Ether)s. WO2017EP66489, 2018.
(9) Petersson, M. J.; Jenkins, I. D.; Loughlin, W. A. The Use of Phosphonium Anhydrides for the Synthesis
of 2-Oxazolines{,} 2-Thiazolines and 2-Dihydrooxazine under Mild Conditions. Org. Biomol. Chem.
2009, 7 (4), 739–746 DOI: 10.1039/B818310D.
(10) Pouliot, M.-F.; Angers, L.; Hamel, J.-D.; Paquin, J.-F. Synthesis of 2-Oxazolines and Related NContaining Heterocycles Using [Et2NSF2]BF4 as a Cyclodehydration Agent. Tetrahedron Lett. 2012, 53
(32), 4121–4123 DOI: https://doi.org/10.1016/j.tetlet.2012.05.130.
(11) Reddy, D. N.; Thirupathi, R.; Tumminakatti, S.; Prabhakaran, E. N. A Method for Stabilizing the Cis
Prolyl Peptide Bond: Influence of an Unusual N→π∗ Interaction in 1,3-Oxazine and 1,3-Thiazine
Containing Peptidomimetics. Tetrahedron Lett. 2012, 53 (33), 4413–4417 DOI:
https://doi.org/10.1016/j.tetlet.2012.06.031.
(12) Schweiker, S. S.; Loughlin, W. A.; Lohning, A. S.; Petersson, M. J.; Jenkins, I. D. Synthesis, Screening
and Docking of Small Heterocycles as Glycogen Phosphorylase Inhibitors. Eur. J. Med. Chem. 2014, 84,
584–594 DOI: https://doi.org/10.1016/j.ejmech.2014.07.063.
(13) Alhalib, A.; Kamouka, S.; Moran, W. J. Iodoarene-Catalyzed Cyclizations of Unsaturated Amides. Org.
Lett. 2015, 17 (6), 1453–1456 DOI: 10.1021/acs.orglett.5b00333.
(14) Mollo, M. C.; Orelli, L. R. Microwave-Assisted Synthesis of 2-Aryl-2-Oxazolines, 5,6-Dihydro-4H-1,3-
Oxazines, and 4,5,6,7-Tetrahydro-1,3-Oxazepines. Org. Lett. 2016, 18 (23), 6116–6119 DOI:
10.1021/acs.orglett.6b03122.