Biomacromolecules 2005, 6, 1987-1991
1987
Synthesis and Characterization of Random and Block Copolypeptides Derived from γ-Methylglutamate and Leucine N-Carboxyanhydrides Vishal Goury,† Dhanjay Jhurry,*,† Archana Bhaw-Luximon,† Bruce M. Novak,‡ and Joe¨l Belleney§ Department of Chemistry, Faculty of Science, University of Mauritius, Re´ duit, Mauritius, Department of Chemistry, North Carolina State University, Raleigh, North Carolina 27695-8204, and Laboratoire de Chimie des Polyme` res, UMR 7610, Universite´ Pierre et Marie Curie, 4 Place Jussieu, 75252 Paris Cedex 05, France Received December 10, 2004; Revised Manuscript Received March 9, 2005
The synthesis of random and block copolypolyeptides derived from γ-methylglutamate and leucine N-carboxyanhydrides using Al-Schiff’s base complexes and allylamine as initiators is here reported. The copolymer structures were confirmed by 1H and 13C NMR. The calculation of the statistical average block lengths reveals the presence of longer methylglutamate units in the copolymer. The determination of the reactivity ratios indicated a slightly higher reactivity of γ-methylglutamateNCA as compared to leucineNCA. Block copolypeptides containing glutamate and leucine units were obtained by sequential polymerization of the two NCAs using Al-Schiff’s base complexes or allylamine in dioxane as solvent. Based on 13C NMR spectra of copolymers exhibiting two signals corresponding to peptide linkages, we confirmed the block structure and concluded that the copolymerization proceeds by attack of an amino group present on a glutamate chain end onto a LeuNCA. The copolymerization with allylamine was also shown, from calculation of the average block lengths of sequences, to exhibit living behavior. Viscometry analysis further showed that molar masses of the copolypeptides obtained with Al-Schiff’s base were quite close to those derived from allylamine, supporting the proposed mechanism of copolymerization. Introduction The synthesis and properties of copolypeptides derived from N-carboxyanhydrides has been reviewed by Kricheldorf.1,2 They represent an interesting class of materials which find various applications, in particular as biocompatible materials. Of these synthetic copolypeptides, those with a random primary structure have received most attention by several research groups,1,3-4 who have focused mainly on the kinetic aspects, reactivity ratio determination, chain structure, and conformation. The initiators most commonly used to obtain random copolymers were either primary amines such as benzylamine and hexylamine or tertiary amines such as triethylamine. From the reactivity ratios based on conventional copolymerization of a range of NCAs, a reactivity order was established as follows: GlyNCA > AlaNCA > γ-MeNCA ∼ γ-BzGlu > LeuNCA > ValNCA. The chemical heterogeneity of the copolypeptides was found to be closely related to the type of initiator used. Anderson and co-workers5-7 have shown that triethylamineinitiated copolymerization of NCAs with differing reactivities, such as γ-BzGluNCA and ValNCA, ended up with a significant amount of interchain heterogeneity due to initia* To whom correspondence should be addressed. E-mail: djhurry@ uom.ac.mu. † University of Mauritius. ‡ North Carolina State University. § Universite ´ Pierre et Marie Curie.
tion occurring throughout the polymerization. With primary amines which give rise to fast initiation, interchain heterogeneity is negligible, whereas intrachain heterogeneity is much more pronounced. The synthesis of block copolypeptides and mostly triblocks has also been reported by a number of groups.1,8 NCAs were sequentially copolymerized mostly using primary amines. The mechanism of copolymerization has been well established in that an amino end group present on the first polypeptide chain initiates polymerization of a second NCA. Deming9-11 has more recently shown that zero valent nickel and cobalt initiators were very efficient for polymerization of NCAs. The living character of the polymerizations enabled the synthesis of various diblock copolypeptides. Hadjichristidis et al.12 also obtained block copolypeptides with narrow molecular weight distributions using primary amine initiators. They showed that high purity conditions are essential for a living process and used high vacuum techniques to create and maintain impurity-free environments during NCA polymerization. The structural characterization of block copolypeptides was mainly achieved by IR and circular dichroism spectroscopies. Only a few reports of characterization of copolypeptides using 13C NMR are found in the literature. 1,13 We reported recently that Al-Schiff’s base complexes of the HAPENAlOR family are effective initiators for the polymerization of γ-methylglutamate N-carboxyanhydride
10.1021/bm049219m CCC: $30.25 © 2005 American Chemical Society Published on Web 05/19/2005
1988
Biomacromolecules, Vol. 6, No. 4, 2005
Goury et al.
(γ-MeGluNCA).14 In the present paper, we describe the synthesis of statistically random and block copolymers based on γ-MeGluNCA and leucine NCA (LeuNCA) using the Al complexes. We make use of 13C NMR spectroscopy in deuterated trifluoroacetic acid to compare the primary structures of the latter copolymers with those derived from primary and tertiary amine initiators such as allylamine and triethylamine, respectively. Experimental Section Materials. Anhydrous dioxane sealed under nitrogen was purchased from Aldrich and used without further purification. Dichloromethane was refluxed over calcium hydride, distilled under nitrogen, and stored under argon. Toluene was refluxed under nitrogen, then distilled over sodium, and stored under argon. γ-MeGluNCA and LeuNCA were offered as gifts by Flamel Technologies Co. (Lyon, France) and were used without further purification. These NCAs were synthesized from L-glutamic acid 5-methyl ester and L-leucine using triphosgene. Initiators. HAPENAlOiPr (I) and HAPENAlOMe (II) were synthesized according to procedures described previously.15,16 See the Supporting Information for 1H NMR characterization of the Al complexes. Allylamine (III) and triethylamine (IV) were purchased from Aldrich and used without further purification Homopolymerization of γ-MeGluNCA and LeuNCA. All polymerizations were carried out in Schlenk tubes under argon. The NCA (2.5 mmol) was first dissolved in dichloromethane, and the initiator (I or II) was subsequently added. The reaction was allowed to proceed at room temperature until all of the monomer was consumed, as evidenced by 1 H NMR. The solvent was then removed under vacuum. The polymer was isolated after several washings with water and acetone and finally dried under vacuum at room temperature. Nonsequential Copolymerization of γ-MeGluNCA and LeuNCA. Stoichiometric amounts of both monomers, γ-MeGluNCA (1.25 mmol) and LeuNCA (1.25 mmol), were first dissolved in DCM (5 mL), and the calculated amount of initiator (I, II, III or IV) was then added to achieve [M]/[I] ) 50, 75, and 100. Upon completion of copolymerization at ambient temperature, the solvent was removed under vacuum. The polymer was isolated after several washings with water followed by acetone and dried under vacuum. The formation of a copolymer was confirmed by 1H NMR, and the proportions of the respective units were calculated. Determination of Reactivity Ratios. A set of copolymerization experiments were carried out with varying γ-MeGluNCA and LeuNCA initial feed ratio, in the presence of complex II ([M]/[I] ) 50), as described in the previous paragraph. The polymerizations were conducted to about 1012% conversion (duration approx 15 min), as calculated by gravimetry. The copolymers were isolated by precipitation in methanol and thoroughly dried before analysis. The copolymer composition was determined by 1H NMR (TFAd). Treatment of the data was done by either Finemann and Ross method or Kelen-Tudo¨s method at low polymer conversion.
Figure 1. 13C NMR showing the carbonyl region of a typical random copolymer obtained with initiator (I) in DCM (monomer feed ratio: γ-MeGluNCA/LeuNCA ) 1; [M]/[I] ) 75; [monomer] ) 0.5 M).
Block Copolymerization of γ-MeGluNCA and LeuNCA. γ-MeGluNCA (2.5 mmol) was polymerized in 5 mL of dichloromethane with the required amount of initiator. After the homopolymerization was complete, a solution of LeuNCA (2.5 mmol) in dichloromethane (5 mL) was added to the reaction mixture. Polymerization was allowed to proceed at ambient temperature for the required time. The solvent was removed under vacuum, and the polymer was isolated after several washings with water followed by acetone and dried under vacuum. The copolymer was characterized by 1H and 13C NMR. Characterization. 1H and 13C NMR spectra were recorded in 5 mm tubes in CDCl3, CD2Cl2 or deuterated trifluoroacetic acid at 25 °C on a FT Bruker 250 MHz spectrometer. Viscometric analysis was conducted in Ubbelohde tubes (type A). Results and Discussion Nonsequential copolymerization of γ-MeGluNCA and LeuNCA. A series of copolypeptides were synthesized in DCM at room temperature using (I) and (II) at various total monomer to initiator concentrations ([M]/[I] ) 50, 75, and 100). For comparison with previous works, copolymerizations were also conducted under the same experimental conditions in the presence of a primary amine (III) and a tertiary amine (IV) as initiators. The mechanism of homopolymerization in the presence of the Al alkoxides has been established previously,14 and it proceeds by nucleophilic attack of the C(5)dO by an amino group generated during the initiation step. 1 H NMR analyses of purified samples show a slight enrichment in γ-methylglutamate units, independent of the type of initiator used. Typical 1H and 13C spectra of copolymers obtained with the various initiators are presented in the Supporting Information. A finer analysis of the copolymers was made by focusing on their carbonyl region in the 13C NMR spectra. Thus, it was found that copolymers obtained with initiators I, II, or III are very similar in that four main chain carbonyl resonances are observed. On the basis of the 13C NMR spectra of the homopolymers, the signals at 173.7 and 174.7 ppm (Figure 1) have been assigned
Biomacromolecules, Vol. 6, No. 4, 2005 1989
Synthesis and Characterization of Copolypeptides Table 1. Average Block Lengths of γ-Meglutamate (L h Glu) and Leucine (L h Leu) Residuesa
a
copolymers
initiator
Lh Glu
Lh Leu
CP 1 CP 2 CP 3 CP 4
I II III IV
2.86 3.12 2.84 4.83
2.43 2.53 2.53 4.56
Monomer feed: 50/50; [M]/[I] ) 50.
Table 2. Reactivity Ratios Determined by Finemann-Ross and Kelen-Tudos Methods method
γ-MeGluNCA (r1)
LeuNCA (r2)
Finemann-Ross Kelen-Tudos
2.55 ( 0.1 2.56 ( 0.2
2.22 ( 0.09 2.22 ( 0.09
to Glu-Glu and Leu-Leu sequences of the copolymer, respectively. The other two signals at 173.3 and 175.1 ppm have been attributed to Glu-Leu and Leu-Glu sequences, and their presence indicates a statistical distribution of comonomers. With triethylamine, broader carbonyl peaks are observed and sequence effects are less pronounced, due probably to increased interchain heterogeneity. From the signal intensities (Ii) of the carbonyl peaks of the corresponding sequence and by using eqs 1 and 2, we have calculated the statistical average block lengths1 (L h Glu and L h Leu) of the sequences for various copolymers, as listed in Table 1. The block lengths of random copolymers resulting from initiators III and IV were also computed. On average, triethylamine leads to longer blocks (CP 4, Table 1) than allylamine (CP 3, Table 1). This enhanced intrachain heterogeneity observed for copolymerization in the presence of triethylamine was also reported by Anderson et al.3-5 for γ-OBz-GluNCA and ValNCA. For the Al-complex-initiated copolymerizations (CP 1 and 2), the glutamate sequences were found to be longer than the leucine residues. This indicates a preference for the propagating glutamate residue to react on a γ-MeGluNCA and a tendency to form longer blocks. L h Glu ) {IGluGlu/IGluLeu} + 1
(1)
L h Leu ) {ILeuLeu/ILeuGlu} + 1
(2)
where Li is the average block length and Ii is the intensity of peak i for the corresponding sequence To support the previous findings, the reactivity ratios were further determined using both Finemann-Ross and KelenTudos methods at low polymer conversions. The copolymer ratio was analyzed by 1H NMR by comparing intensities of GluCOOCH3 units with CH3 Leu units. The reactivity ratio values are summarized in Table 2. The results obtained by both methods are in good agreement and indicate a slightly higher reactivity of γ-MeGluNCA as compared to LeuNCA, which is in accordance with previously reported results.1 This trend was explained on the basis of increased bulkiness of the side chain which reduces the nucleophilicity of the nitrogen and electrophilicity of the carbonyl C-5. The values indicate a tendency for the monomers to form blocks, as suggested by the statistical block lengths calculated previously.
Block Copolymerization of γ-MeGluNCA and LeuNCA. Synthesis of (PMeGlu)-b-PLeu) using Allylamine. The polymerization of NCAs and the synthesis of block copolymers using primary amines were reported by a number of groups in the past. It has been generally argued that the formation of block copolymers was successful because the reaction proceeds via the “amine” mechanism in which case Ri . Rp. However, some reports17-19 have shown that the molar mass distribution of some polypeptide blocks were broad due to possible side reactions such as the occurrence of the “activated monomer mechanism”. The extent of such reactions will depend on the nucleophilic versus basic character of the primary amine. In a more recent work, Dimitrov and Schlaad20 have used an amino-terminated polystyrene to polymerize Z-L-lysine NCA, and they show by SEC, after precipitation and extraction, that a block copolymer is indeed obtained free of homopolymer contaminants at 70-80% yield. To the best of our knowledge, allylamine has not been used previously for the synthesis of well-defined glutamate and leucine blocks, and we have therefore tested its efficiency. Copolymerization was carried out sequentially by first polymerizing γ-MeGluNCA in DCM or dioxane under experimental conditions as described in the previous section. After total consumption of this monomer as evidenced by 1 H NMR, an equivalent amount of LeuNCA was added to the solution and allowed to polymerize at room temperature. 13 C NMR analysis of the purified polymer sample revealed the presence of two distinct carbonyl signals centered at 173.75 and 174.7 ppm, characteristic of methylglutamate and leucine peptide units (see the Supporting Information for the 13 C NMR spectrum of copolymer obtained with initiator III). The presence of these two signals supports a block copolymer structure. It is to be noted that the splitting of the methylglutamate peaks is due to the presence of both R-helix and random coil conformations of this segment. The block structure was further confirmed by calculating the length of the two blocks from the integration of the allyl group (δCH2 ) 5.1 ppm) anchored at one end of the chain. The values thus obtained were found to be in relatively good agreement with the [M]/[I] ratio in the case of dioxane, as one would expect for a living process. For instance, 1H NMR calculations gave DPn values of 45 and 56 for glutamate and leucine sequences respectively as compared to an expected DPn of 50, which is within acceptable experimental errors. (see the Supporting Information for 1H NMR spectrum of poly(Glub-Leu) in dioxane with integration of signals). In summary, we can conclude that the copolymerization proceeds according to Scheme 1 whereby an amino group present on a glutamate chain end acts as a macroinitiator for the polymerization of LeuNCA. The results obtained with DCM were not conclusive. Indeed, the lengths of both sequences were found to be much higher than the calculated length, probably because of the partial precipitation of the polypeptide chains in that solvent during the polymerization. Moreover, the presence of homopolymers cannot be ruled out. Synthesis of PMeGlu-b-PLeu using Al-Schiff’s Base Complexes. Kinetics of γ-MeGluNCA Polymerization. Be-
1990
Biomacromolecules, Vol. 6, No. 4, 2005
Goury et al.
Scheme 1. Sequential Copolymerization Pathway
fore attempting the sequential block copolymerization experiments using the Al complexes, the kinetics of γ-MeGluNCA homopolymerization was initially studied in both DCM and dioxane at room temperature. The polymer conversion was monitored at various time intervals using 1 H NMR by comparing the intensites of either the OCH3 signals (for DCM polymerization) or the CH2 signals (for dioxane polymerization) that are due to residual monomer and polymer. Linear semilogarithmic plots of monomer conversion against time are obtained (Figure 2), implying that the rate of polymerization is first order with respect to γ-MeGluNCA concentration. The polymerization is found to be faster in DCM as compared to dioxane, as depicted by the rate constants listed in Table 3. Synthesis of Copolymers. On the basis of the better results obtained with allylamine in dioxane, we then attempted the synthesis of glutamate and leucine block copolypeptides using Al complexes I and II in that solvent. The 13C spectra (Figure 3) of the purified copolymers were
Figure 2. Semilogarithmic plot of γ-MeGluNCA conversion against time (initiators: I and II; [M]/[I] ) 50; temp. ) 25 °C).
Table 3. Apparent Rate Constants for γ-MeGluNCA in DCM and Dioxane at 25 °C Using Initiators I and IIa initiator I II I II a
solvent DCM dioxane
kapp (h-1) 0.37 0.75 0.11 0.10
[M]/[I] ) 50.
identical to those obtained with allylamine; that is, they present two single signals corresponding to the methylglutamate and leucine peptide units. The copolymerization proceeds by attack of an amino group present on a glutamate chain end onto a LeuNCA, as illustrated in Scheme 1. The DPn of copolypeptide chains resulting from HAPENAlOiPr-initiated reactions could not be determined by 1H NMR as was the case for the allylamine-initiated polymerizations, because the protons of the isopropoxy end-group are overlapped with main chain proton signals. For further comparison of the copolypeptides synthesized, the polymer samples were subjected to viscometry analysis in TFA. The reduced viscosity results are given in Table 4. As can be seen, the reduced viscosities of copolymers obtained in DCM are much higher than those synthesized in dioxane, which reflect a higher molar mass of the corresponding polypeptides and a divergence from living conditions. This result is in agreement with the NMR findings. Moreover, the viscosity
Figure 3. 13C spectrum showing the carbonyl region (TFA-d) of a block copolymer obtained with initiator (I) in dioxane.
Biomacromolecules, Vol. 6, No. 4, 2005 1991
Synthesis and Characterization of Copolypeptides Table 4. Reduced Viscosity of Homo- and Copolypeptidesa polymer
initiator
solvent
poly(MeGlu) polyLeu poly(MeGlu-b-Leu) poly(MeGlu-b-Leu) poly(MeGlu-b-Leu) poly(MeGlu-b-Leu)
III
dioxane
I III I
dioxane DCM
ηred (g-1 cm3)b
c DPexp n , Glu/Leu
66 33 116 81 277 254
51 46 45/56 d 72/80 d
a DP b -3 in trifluoron,theoretical ) [M]/[I])50, 100% conversion. 0.5 g dm acetic acid at 25 °C (Ubbelohde tube type A). c Determined from 1H NMR spectra. d Undetermined.
values of copolymers obtained with either allylamine or HAPENAlOiPr in dioxane are quite close. Conclusions We have shown in this paper that Al-Schiff’s base complexes of the HAPENAlOR family can be successfully used as initiators to synthesize well-defined blocky-type and true block copolypeptides containing glutamate and leucine units. Reactivity ratios of the NCAs were determined for the nonsequential copolymerizations. The values which are greater than unity indicate that the monomers have a tendency to form blocky sequences and the statistical average block lengths of the sequences were also computed based on the NMR intensities of the carbonyl peptide signals. Moreover, we showed that copolypetides obtained with Al complexes and allylamine exhibited a higher percentage of intrachain heterogeneity in comparison with those obtained from triethylamine. In addition, it was found that the sequential polymerization of the NCAs in the presence of both allylamine and HAPENAlOR complexes yield block copolymers, as confirmed by NMR and viscometry measurements. Acknowledgment. We are thankful to the Tertiary Education Commission (Mauritius) for providing a scholar-
ship to V.G. for his PhD Thesis. We also extend our warmest thanks to Prof R. G. Gilbert for hosting V.G. at the Key Centre for Polymer Colloids (Sydney) for 3 months. We are most grateful to Flamel Technologies Co., Lyon, and Dr. G. Soula (G.M.) for sending us free samples of NCAs. Supporting Information Available. NMR spectra of random and block copolypeptides. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Kricheldorf, H. R. In Models of Biopolymers by Ring-Opening Polymerization; Penczek, S., Ed.; CRC: Boca Raton, FL, 1990; pp 98-132 and references therein. (2) Kricheldorf, H. R. R-Amino acid N-Carboxyanhydrides and Related Heterocycles; Springer-Verlag: New York, 1987. (3) Hiltner, A.; Anderson, J. M.; Baer, E. J. Macromol Sci., Phys Ed. 1973, 8, 431. (4) Hiltner, A.; Anderson, J. M.; Borkovski, E. Macromolecules 1972, 5, 446. (5) Sederel, W.; Deshmane, S.; Hayashi, T.; Anderson, J. M. Biopolymers 1978, 17, 2835. (6) Deshmane, S.; Hayashi, T.; Sederel, W.; Anderson, J. M. Biopolymers 1978, 17, 2851. (7) Mitra, S. B.; Patel, N. K.; Anderson. J. M. Int. J. Biol. Macromol. 1979, 1, 55. (8) Hayashi, T.; Walton, A. G.; Anderson, J. M. Macromolecules 1977, 10, 346. (9) Deming, T. J. Nature 1997, 390, 386. (10) Deming, T. J. Macromolecules 1999, 32, 4500. (11) Deming, T. J. AdV. Drug DeliVery ReV. 2002, 54, 1145. (12) Aliferis, T.; Iatrou, H.; Hadjichristidis, N. Biomacromolecules 2004, 5, 1653. (13) Kricheldorf, H. R.; Schilling, G. Makromol. Chem. 1978, 179, 1175. (14) Bhaw-Luximon, A.; Jhurry, D.; Belleney J.; Goury V. Macromolecules 2003, 36 (4), 977. (15) Bhaw-Luximon, A.; Jhurry, D.; Spassky, N. Polym. Bull. (Berlin) 2000, 44, 31-38. (16) Jhurry, D.; Bhaw-Luximon, A.; Spassky, N. Macromol Symp. 2001, 175, 67. (17) Gallot, B. Prog. Polym. Sci. 1996, 21, 1035-1088. (18) Nakajima, A.; Hayashi, T.; Kugo, K.; Shinoda, K. Macromolecules 1979, 12, 840-843. (19) Deming, T. J. AdV. Mater. 1997, 9, 299-311. (20) Dimitrov, I.; Schlaad, H. Chem Comm. 2003, 2944.
BM049219M
