Synthesis of Carbohydrate-Grafted Glycopolymers Using a Catalyst-Free, Perfluoroarylazide-Mediated Fast Staudinger Reaction

Glycopolymers have gained increasing importance in investigating glycan-lectin interactions, as drug delivery vehicles and in modulating interactions with proteins. The synthesis of these glycopolymers is still a challenging and rigorous exercise. In this regard, the highly efficient click reaction, copper (I)-catalyzed alkyne-azide cycloaddition, has been widely applied not only for its efficiency but also for its tolerance of the appended carbohydrate groups. However, a significant drawback of this method is the use of the heavy metal catalyst which is difficult to remove completely, and ultimately toxic to biological systems. In this work, we present the synthesis of carbohydrate-grafted glycopolymers utilizing a mild and catalyst-free perfluorophenyl azide (PFPA)-mediated Staudinger reaction. Using this strategy, mannose (Man) and maltoheptaose (MH) were grafted onto the biodegradable poly(lactic acid) (PLA) by stirring a PFAA-functionalized PLA with a phosphine-derivatized Man or MH in DMSO at room temperature within an hour. The glycopolymers were characterized by 1H-NMR, 19F-NMR, 31P-NMR and FTIR.


Introduction
Carbohydrates are not only the core of metabolism in many biological systems, but are also integral in many cells as structural [1], communication [2], and recognition [3] elements. Synthetic carbohydrate-functionalized polymers, i.e., glycopolymers, has become an important tool in fundamental glycobiology research, and in biomedical applications such as sensing [4,5], drug delivery [6], and cryopreservation [7,8]. Synthetic glycopolymers can be categorized broadly into two types based on the mode of synthesis: (1) polymerization of carbohydrate-derivatized monomers (such as allyl [9], acrylamide [10] and acrylate [11]), and (2) grafting of carbohydrates onto a polymer backbone [12][13][14][15][16][17]. The latter technique, typically referred to as post-polymerization modification, allows facile synthesis of the polymer backbone and control over carbohydrate grafting density. To successfully graft carbohydrates onto the polymer, the conjugation reaction must be highly efficient, of high yield, and tolerant of the functional groups on the polymer. The popular 'click' type of reactions, particularly the copper(I)-catalyzed alkyne-azide cycloaddition (CuAAC), meet these requirements and have been applied in the synthesis of glycopolymers by post-polymerization and in glycosylation of surfaces [18][19][20]. This reaction tolerates a wide variety of carbohydrate side groups. Analytically, it offers the advantage in the form of the distinct chemical shift of the triazole proton in 1 H-NMR that facilitates straightforward characterization [21,22]. However, a drawback of this reaction is the necessity of Cu (I) catalyst that has proved difficult to be removed completely from the

Results and Discussion
While the availability of pendant functional groups like the hydroxyl on the polymer would be suited for condensation with a carboxyl-modified sugar, this strategy is less desirable as it requires the use of coupling agents that can be difficult to remove like DCC/DMAP [39] or strong acid [40] that would hydrolyze the polymer.
Our design for grafting carbohydrates to PLA was to functionalize PLA with PFPA followed by reaction with a phosphine-derivatized carbohydrate. A PLA-co-PLA-PFPA copolymer is synthesized so that the density of PFPA and the carbohydrate can be varied and controlled. A hydroxy-functionalized polylactide copolymer 4 (Scheme 1A) was synthesized to conjugate PFPA and subsequently graft the carbohydrate. The benzyl-derivatized lactide monomer 1 was synthesized according to Scheme 1B. Ring-opening copolymerization of lactide 1 and lactide 2 in toluene using stannous octoate as the catalyst gave polylactide copolymer 4 in 74% yield [5]. The ratio of the two monomers can be varied so as to control the grafting density of carbohydrate on the PLA polymer. In this work, the polymer obtained had a monomer ratio (m:n) of 1:22 for 1 and 2 after copolymerization. Deprotection of the benzyl group on 3 gave the hydroxy-functionalized polylactide copolymer 4 in 63% yield [5]. The m:n ratio was calculated from the 1 H-NMR spectrum of 4 ( Figure 1A) by taking the ratio of the methine protons (H-a and H-c) that overlap at 5.2 ppm. The integral value of the H-a is equivalent to methylene protons H-b/2. Subtraction of this value from the overlapped (H-a + H-c) gives the integration of H-c and therefore H-a/H-c can be calculated.     PFPA-functionalized PLA copolymer 5 was prepared by esterification of 4 with carboxy-derivatized PFPA using DCC and a catalytic amount of DMAP giving 5 in 86% yield (Scheme 2).
PFPA-functionalized PLA copolymer 5 was prepared by esterification of 4 with carboxyderivatized PFPA using DCC and a catalytic amount of DMAP giving 5 in 86% yield (Scheme 2). The 1 H-NMR spectrum of 5 showed a near complete reaction as demonstrated by a marked shift of the methylene proton H-b from 4.05 ppm to 4.83 ppm ( Figure 1). No degradation of the copolymers to the free lactic acid monomer was observed during the reaction, as evidenced by the absence of peaks at ~1.2 ppm which belongs to the free lactic acid. In addition, the characteristic azide peak at 2133 cm −1 appeared in the Fourier transform-infrared spectroscopy (FTIR) spectrum of copolymer 5 ( Figure 2). The phosphine-derivatized carbohydrates were prepared from an amine-derivatized carbohydrate and the NHS-functionalized phosphine (Scheme 3). A monosaccharide, D-mannose (Man), and an oligosaccharide, D-maltoheptaose (MH), were used as model carbohydrates in this study. The amine-Man 6 was synthesized according to previously reported procedure (see Scheme S2 and detailed procedures in SI) [41]. The amine-MH 7 was synthesized following the procedure in Scheme S1 (see detailed procedures in SI). Reaction of NHS-functionalized triphenylphosphine with excess of 6 or 7 in DMSO at room temperature gave the triphenylphosphine-derivatized Man (8) or MH (9). The 1 H-NMR spectrum of 5 showed a near complete reaction as demonstrated by a marked shift of the methylene proton H-b from 4.05 ppm to 4.83 ppm ( Figure 1). No degradation of the copolymers to the free lactic acid monomer was observed during the reaction, as evidenced by the absence of peaks at~1.2 ppm which belongs to the free lactic acid. In addition, the characteristic azide peak at 2133 cm −1 appeared in the Fourier transform-infrared spectroscopy (FTIR) spectrum of copolymer 5 ( Figure 2). PFPA-functionalized PLA copolymer 5 was prepared by esterification of 4 with carboxyderivatized PFPA using DCC and a catalytic amount of DMAP giving 5 in 86% yield (Scheme 2). The 1 H-NMR spectrum of 5 showed a near complete reaction as demonstrated by a marked shift of the methylene proton H-b from 4.05 ppm to 4.83 ppm ( Figure 1). No degradation of the copolymers to the free lactic acid monomer was observed during the reaction, as evidenced by the absence of peaks at ~1.2 ppm which belongs to the free lactic acid. In addition, the characteristic azide peak at 2133 cm −1 appeared in the Fourier transform-infrared spectroscopy (FTIR) spectrum of copolymer 5 ( Figure 2). The phosphine-derivatized carbohydrates were prepared from an amine-derivatized carbohydrate and the NHS-functionalized phosphine (Scheme 3). A monosaccharide, D-mannose (Man), and an oligosaccharide, D-maltoheptaose (MH), were used as model carbohydrates in this study. The amine-Man 6 was synthesized according to previously reported procedure (see Scheme S2 and detailed procedures in SI) [41]. The amine-MH 7 was synthesized following the procedure in Scheme S1 (see detailed procedures in SI). Reaction of NHS-functionalized triphenylphosphine with excess of 6 or 7 in DMSO at room temperature gave the triphenylphosphine-derivatized Man (8) or MH (9). The phosphine-derivatized carbohydrates were prepared from an amine-derivatized carbohydrate and the NHS-functionalized phosphine (Scheme 3). A monosaccharide, D-mannose (Man), and an oligosaccharide, D-maltoheptaose (MH), were used as model carbohydrates in this study. The amine-Man 6 was synthesized according to previously reported procedure (see Scheme S2 and detailed procedures in SI) [41]. The amine-MH 7 was synthesized following the procedure in Scheme S1 (see detailed procedures in SI). Reaction of NHS-functionalized triphenylphosphine with excess of 6 or 7 in DMSO at room temperature gave the triphenylphosphine-derivatized Man (8) or MH (9). To prepare Man-or MH-grafted PLA 10 and 11, copolymer 5 was added directly to the reaction mixture of 8 or 9 in DMSO and stirred at room temperature for 1 hour. The products were purified by dialysis for 48 h and dried by lyophilization. After the carbohydrate was grafted, the aromatic and the carbohydrate peaks appeared in the 1 H-NMR spectra of 10 and 11 at 7.5-7.7 ppm and 3.0-6.0 ppm, respectively ( Figure 3). In the 31 P-NMR spectra of the products (Figure 4), a new peak was observed at 13 ppm after conjugation of the phosphine onto the copolymer. The absence of any peaks higher than 13 ppm confirmed the absence of byproducts resulting from the oxidation of phosphine.
The yield of conjugation was obtained from the 1 H-NMR spectrum by taking the ratio of peak integration of the phenyl protons (Ph) at 7.5-7.7 ppm and the methyl protons H-b at 1.48 ppm, together with the previously calculated monomer ratio of m:n = 1:22 to give the formula: (1) Using this equation, the yields were calculated to be 25% and 35% for the Man-grafted PLA copolymer 10 and MH-grafted PLA copolymer 11, respectively (see Figures  Scheme 3. Synthesis of phosphine-derivatized mannose 8 and maltoheptaose 9, and subsequent grafting to PFPA-PLA copolymer 5 to yield mannose-polymer 10 and maltoheptaose-polymer 11, respectively. To prepare Man-or MH-grafted PLA 10 and 11, copolymer 5 was added directly to the reaction mixture of 8 or 9 in DMSO and stirred at room temperature for 1 h. The products were purified by dialysis for 48 h and dried by lyophilization. After the carbohydrate was grafted, the aromatic and the carbohydrate peaks appeared in the 1 H-NMR spectra of 10 and 11 at 7.5-7.7 ppm and 3.0-6.0 ppm, respectively ( Figure 3). In the 31 P-NMR spectra of the products (Figure 4), a new peak was observed at 13 ppm after conjugation of the phosphine onto the copolymer. The absence of any peaks higher than 13 ppm confirmed the absence of byproducts resulting from the oxidation of phosphine.
The yield of conjugation was obtained from the 1 H-NMR spectrum by taking the ratio of peak integration of the phenyl protons (Ph) at 7.5-7.7 ppm and the methyl protons H-b at 1.48 ppm, together with the previously calculated monomer ratio of m:n = 1:22 to give the formula: Using this equation, the yields were calculated to be 25% and 35% for the Man-grafted PLA copolymer 10 and MH-grafted PLA copolymer 11, respectively (see Figures S5 and S6 for peak integrations).

Conclusions
In this work, we utilized an electron-deficient perfluorophenyl azide-mediated fast Staudinger reaction to efficiently synthesize carbohydrate grafted-polylactide glycopolymers by postpolymerization modification. A polylactide copolymer was synthesized by stannous octoatecatalyzed cationic ring-opening copolymerization, which was subsequently modified with PFPA. Conjugation with a phosphine-derivatized carbohydrate yield mannose-and maltoheptaose-grafted polylactide glycopolymers in 35% and 25% yields. These yields are comparable to those obtained by other groups in post-polymerization synthesis of glycopolymers using other techniques [42,43]. The main limiting factor of efficient grafting being steric hindrance between ligands and the polymer backbone restricting access to the reactive sites. In our work, this occurs between carbohydratephosphines reaction with the pendant PFPA groups, which is further complicated by the difference in polarity between the PLA and D-mannose/maltoheptaose. However, the grafting reaction is fast, carried out under mild conditions without the use of any catalyst. This metal catalyst-free approach to glycopolymer synthesis is significant as it eliminates the concerns over the potential toxicity of heavy metals, making these glycopolymers attractive for biomedical applications.

Conclusions
In this work, we utilized an electron-deficient perfluorophenyl azide-mediated fast Staudinger reaction to efficiently synthesize carbohydrate grafted-polylactide glycopolymers by postpolymerization modification. A polylactide copolymer was synthesized by stannous octoatecatalyzed cationic ring-opening copolymerization, which was subsequently modified with PFPA. Conjugation with a phosphine-derivatized carbohydrate yield mannose-and maltoheptaose-grafted polylactide glycopolymers in 35% and 25% yields. These yields are comparable to those obtained by other groups in post-polymerization synthesis of glycopolymers using other techniques [42,43]. The main limiting factor of efficient grafting being steric hindrance between ligands and the polymer backbone restricting access to the reactive sites. In our work, this occurs between carbohydratephosphines reaction with the pendant PFPA groups, which is further complicated by the difference in polarity between the PLA and D-mannose/maltoheptaose. However, the grafting reaction is fast, carried out under mild conditions without the use of any catalyst. This metal catalyst-free approach to glycopolymer synthesis is significant as it eliminates the concerns over the potential toxicity of heavy metals, making these glycopolymers attractive for biomedical applications.

Conclusions
In this work, we utilized an electron-deficient perfluorophenyl azide-mediated fast Staudinger reaction to efficiently synthesize carbohydrate grafted-polylactide glycopolymers by post-polymerization modification. A polylactide copolymer was synthesized by stannous octoate-catalyzed cationic ring-opening copolymerization, which was subsequently modified with PFPA. Conjugation with a phosphine-derivatized carbohydrate yield mannose-and maltoheptaose-grafted polylactide glycopolymers in 35% and 25% yields. These yields are comparable to those obtained by other groups in post-polymerization synthesis of glycopolymers using other techniques [42,43]. The main limiting factor of efficient grafting being steric hindrance between ligands and the polymer backbone restricting access to the reactive sites. In our work, this occurs between carbohydrate-phosphines reaction with the pendant PFPA groups, which is further complicated by the difference in polarity between the PLA and D-mannose/maltoheptaose. However, the grafting reaction is fast, carried out under mild conditions without the use of any catalyst. This metal catalyst-free approach to glycopolymer synthesis is significant as it eliminates the concerns over the potential toxicity of heavy metals, making these glycopolymers attractive for biomedical applications.

Materials and Instruments
All chemicals and solvents were purchased from Sigma-Aldrich (St. Louis, MO, USA), TCI America (Portland, Oregon, USA) or Fisher Scientific (Hampton, NH, USA), and were used without further purification unless otherwise noted. Dichloromethane (DCM), dimethylformamide (DMF), ethyl acetate (EtOAc) and dimethyl sulfoxide (DMSO) were purified by distillation over CaH 2 . Deuterated solvents were acquired from Cambridge Isotope Lab., Inc. (Tewksbury, MA, USA). Amberlite IRC-120H + resin was activated by washing with NaOH and HCl, followed by water, ethanol and toluene.
Nuclear magnetic resonance spectroscopy data were collected on either a Bruker 500 MHz spectrometer ( 1 H-NMR) or a Bruker 200 MHz spectrometer ( 19 F-and 31 P-NMR) (Bruker Corporation, Billerica, MA, USA). FT-IR spectra were recorded on a Nicolet 6700 FT-IR spectrometer (Thermo Fisher Scientific, Waltham, MA, USA).

Synthesis of 3-benzyloxy-2-hydroxypropionic acid (II)
Synthesized according to literature procedure [47]. To a 200 mL of 0.7 M aqueous solution of trifluoroacetic acid, H-Ser(benzyl)-OH (I, 10.0 g, 52 mmol) was added, and the mixture was stirred at room temperature until all solids were dissolved. Then, 50 mL of aqueous NaNO 2 (5.3 g, 77 mmol) was added dropwise with a syringe pump under Ar protection and the reaction was stirred for another 3 h. After confirming of the consumption of starting material by TLC, NaCl (10 g) was added and the mixture was extracted with ethyl acetate three times followed by washing with brine and dried over MgSO 4 . After passing through a flash column using CH 2 Cl 2 /MeOH/AcOH (v/v/v 100:8:1), compound II was obtained (7.2 g, 72%). 3

Synthesis of 3-(benzyloxy)-2-(2-bromopropanoyloxy)propanoic acid (III)
Synthesized per literature [47]. Compound II (10.0 g, 51 mmol) and 2-bromopropionyl chloride (6.3 mL, 61.2 mmol) were mixed in a 50-mL round-bottom flask backfilled with Ar. The reaction mixture was heated to 70 • C and stirred for 6 h. Upon the completion of the reaction, the crude product was heated at 60 • C under reduced pressure to remove unreacted 2-bromopropionyl chloride and 2-bromopropionyl acid. After cooling to room temperature, the residue was washed with water and extracted with ethyl acetate 3 times. The combined organic phase was washed with brine and dried over MgSO 4 . Compound III was obtained as a brown oil after further purification by flash column CH 2 Cl 2 /MeOH/AcOH (v/v/v 100:2:0.5) (12.6 g, 75%). 1
4.5. Synthesis of 3-(benzyloxymethyl)-6-methyl-1,4-dioxane-2,5-dione (Monomer 1) Synthesized as described in literature [7]. A solution of compound IV (10.0 g, 26.8 mmol) in dry CH 2 Cl 2 (100 mL) was added dropwise to refluxing dry acetone (1 L) containing DIEA (8.8 mL, 53.6 mmol) under Ar. It took 10 h to finish the addition and the reaction was refluxed for another hour. The solvents were removed under reduced pressure and ether was added to dissolve the crude product. Insoluble ammonium iodide was filtered and the filtrate was concentrated. After purification by flash column chromatography using hexanes/ethyl acetate (v/v 4:1), the title compound 1 was obtained as a yellow oil (2.1 g, 31%). The diastereomers were used directly without separation.

Synthesis of PLA copolymer 3
Monomer 1 (1.0 g, 4.0 mmol), recrystallized L-lactide (2, 1.0 g, 6.9 mmol) and Sn(Oct) 2 (10 mg in 1 mL anhydrous toluene) were added into a 5-mL round-bottom flask. The mixture was heated to 70 • C under vacuum for 1 h. Ar was filled in the flask and the temperature was increased to 140 • C. The mixture was stirred until the stir bar stopped moving. After cooling to room temperature, the solid was dissolved in CH 2 Cl 2 and hexanes was added. The precipitate was re-dissolved in CH 2 Cl 2 , and this dissolution/precipitation was repeated three times, and the precipitate was finally dried under vacuum to give copolymer 1c as a dark brown solid (1.48 g, 74%). 1

Synthesis of PLA copolymer 4
Following literature synthesis [47]. Copolymer 3 (1.0 g) was dissolved in 50 mL ethyl acetate/methanol (3:1), and catalytic amount of Pd/C was added. The mixture was purged with Ar for 20 min and filled with H 2 under vigorous stirring. After 12 h, the solution was passed through a pile of Celite to remove Pd/C and the filtrate was dried under vacuum to give copolymer 4 as a light brown solid (630 mg, 63%). 1   Copolymer 4 (100 mg) and PFPA-COOH (20 mg, 0.09 mmol) were added together with DCC (41 mg, 0.2 mmol) and DMAP (2.4 mg, 0.02 mmol) to 10 mL anhydrous dichloromethane under Ar, and the mixture was stirred at room temperature overnight. The solution was then concentrated to 3 mL and was poured into 50 mL of hexane/methanol (v/v 9:1) to precipitate the crude product. The yellow precipitate was dissolved in dichloromethane and was precipitated in hexane/methanol. This dissolution/ precipitation was repeated for a total of 3 times. Finally, the precipitate was