Glycopolypeptide conformations in bioactive block copolymer assemblies influence their nanoscale morphology

We describe the preparation and assembly of glycosylated amphiphilic diblock copolypeptides, where the hydrophilic glycosylated segments adopt either α-helical or disordered conformations. In this study, glycosylated amphiphilic diblock copolypeptides were prepared using poly(L-leucine), poly(L), as the hydrophobic segment, and poly(α-D-galactopyranosyl-L-lysine), poly(α-gal-K), or poly(α-D-galactopyranosyl-L-cysteine sulfone), poly(α-gal-CO2), as the hydrophilic segment. The poly(α-gal-K) and poly(α-gal-CO2) segments are known to be fully α-helical (>90% at 20 °C) and fully disordered in water, respectively. We found that block copolypeptides containing galactosylated hydrophilic segments of either α-helical or disordered conformation give different assembly morphologies, where the disordered glycopolypeptide segments favor vesicle formation and also present sugar residues that can bind to biological targets.


Introduction
2][3] To precisely control the nanostructured morphology of these delivery vehicles, including shape (e.g.spherical micelles, cylindrical micelles, discs, or vesicles) and internal structure (e.g.spotted, segmented, or core-shell), the assembly of amphiphilic block copolymers in aqueous solution has been highly useful. 4,5owever, the incorporation of multifunctionality (e.g. for cellular targeting, uptake, or intracellular release of cargos), [6][7][8] can oen perturb the self-assembly process, leading to different morphologies with altered stabilities.0][11] Here, we describe the preparation and assembly of glycosylated amphiphilic diblock copolypeptides, where the hydrophilic glycosylated segments adopt either a-helical or disordered chain conformations.These distinct glycopolypeptide conformations were found to signicantly impact block copolymer self-assembly.These results show how careful choice of polypeptide components can be used to direct assembly of nanocarriers into desired morphologies and simultaneously introduce bioactive functionality.
Glycosylation of polymeric drug and gene carriers has been shown to lower cytotoxicity, enhance aqueous solubility, and provide targeting to specic cells and organs. 12,13It is also well known in biology that the way in which sugar functional groups are presented greatly affects their ability to bind targets and signal cells. 14We recently reported the preparation of fully glycosylated, high molar mass synthetic polypeptides, which are water soluble and mimic the structures of naturally occurring glycoproteins. 15,16 key feature of these glycopolymers is that their chain conformations are readily controlled, either by choice of peptide backbone or by selective oxidation of side-chain functional groups, such that fully a-helical or fully disordered chains can be obtained.We have now incorporated these glycopolypeptides as hydrophilic segments in amphiphilic diblock copolypeptides to study their aqueous self-assembly and evaluate the properties of the resulting nanostructures.8][19][20] This has been difficult to study since hydrophilic polypeptide segments presenting similar functionality but differing only in conformation are rare.Here, we have found that block copolypeptides containing galactosylated hydrophilic segments of either a-helical or disordered conformation give different assembly morphologies, where the disordered glycopolypeptide segments favor vesicle formation and present sugar residues that can bind to biological targets (Fig. 1).
2][23][24][25] The rod-like conformations of these hydrophobic segments were found to favor side-by-side packing resulting in lamellar vesicle membranes, while samples with disordered hydrophobic segments were found to pack into spherical micelles, 26,27 similar to other, conformationally disordered, synthetic block copolymers. 52][23][24][25] However, it has been difficult to unequivocally determine the role of the hydrophilic chain conformation in directing nanostructure, since there have been signicant differences between the disordered and a-helical hydrophilic segments in these materials (e.g.ionic vs. nonionic).For example, both nonionic poly(N 3 -2-(2-(2-methoxyethoxy)ethoxy)acetyl-L-lysine) 100 -b-poly(L-leucine) 20  (K P 100 L 20 ) with an a-helical hydrophilic segment and ionic poly(L-homoarginine$HCl) 60 -block-poly(L-leucine) 20 (R H 60 L 20 ) with a disordered hydrophilic segment can be assembled into micron sized vesicles in water. 18,19Some ionic vesicles can be neutralized by adjustment of pH, which results in a transition from disordered to a-helical conformation, but the uncharged, a-helical polypeptide segments are sparingly soluble in water and precipitate above micromolar concentrations. 28,29Likewise, K P 100 L 20 samples with nonionic, disordered hydrophilic segments have been prepared using racemic K P residues, yet inter-chain Hbonding interactions in the resulting disordered segments limit their aqueous solubility as well. 18Consequently, there remains a need for fully a-helical and disordered hydrophilic polypeptide segments that display similar functionality to determine how hydrophilic chain conformation can be used to direct nanoscale assembly and present polypeptide functionality.

Materials and methods
Unless stated otherwise, reactions were conducted in ovendried glassware under an atmosphere of dinitrogen using anhydrous solvents.Hexanes, THF, DCM, and DMF were puried by rst purging with dry nitrogen, followed by passage through columns of activated alumina.Deionized water (18 MU cm) was obtaining by passing in-house deionized water through a Millipore Milli-Q Biocel A10 purication unit.All commercially obtained reagents were used as received without further purication unless otherwise stated.Reaction temperatures were controlled using an IKA temperature modulator, and unless stated otherwise, reactions were performed at room temperature (RT, approximately 20 C).Thin-layer chromatography (TLC) was conducted with EMD gel 60 F254 precoated plates (0.25 mm) and visualized using a combination of UV, anisaldehyde, and phosphomolybdic acid staining.Selecto silica gel 60 (particle size 0.032-0.063mm) was used for ash column chromatography. 1 H NMR spectra were recorded on Bruker spectrometers (at 500 MHz) and are reported relative to residual protons in the deuterated solvents.Data for 1 H NMR spectra are reported as follows: chemical shi (d ppm), multiplicity, coupling constant (Hz) and integration.Splitting patterns are designated as follows: s, singlet; d, doublet; t, triplet; q, quartet; m, multiplet and br, broad. 13C NMR spectra were recorded on Bruker Spectrometers (at 125 MHz).Data for 13 C NMR spectra are reported in terms of chemical shi.Highresolution mass spectrometry (HRMS) was performed on a Micromass Quatro-LC Electrospray spectrometer with a pump rate of 20 mL min À1 using electrospray ionization (ESI).All Fourier Transform Infrared (FTIR) samples were prepared as thin lms on NaCl plates, and spectra were recorded on a Perkin Elmer RX1 FTIR spectrometer and are reported in terms of frequency of absorption (cm À1 ).Tandem gel permeation chromatography/light scattering (GPC/LS) was performed on an SSI Accuow Series III liquid chromatograph pump equipped with a Wyatt DAWN EOS light scattering (LS) and Optilab rEX refractive index (RI) detectors.Separations were achieved using 10 5 , 10 4 , and 10 3 Å Phenomenex Phenogel 5 mm columns using 0.10 M LiBr in DMF as the eluent at 60 C. All GPC/LS samples were prepared at concentrations of 5 mg mL À1 .The preparation of 2,3,4,6-tetra-O-acetyl-a-D-galactopyranosyl-L-lysine-N-carboxyanhydride (a-gal-K NCA), 15 2,3,4,6-tetra-O-acetyl-a-D-galactopyranosyl-L-cysteine-N-carboxyanhydride (a-gal-C NCA), 16 L-leucine N-carboxyanhydride (Leu NCA), 30 and (PMe 3 ) 4 Co 31 have been previously reported.

Preparation of glycosylated diblock copolypeptides
All polymerization reactions were performed in a dinitrogen lled glove box.To a solution of a-gal-K NCA or a-gal-C NCA (1 equiv.) in THF (50 mg mL À1 ) was rapidly added, via syringe, a solution of (PMe 3 ) 4 Co in THF (0.05 equiv., 30 mg mL À1 ).The reaction was stirred at RT and polymerization progress was monitored by FTIR.Polymerization reactions were generally complete within 3 hours.Immediately upon polymerization completion, aliquots were removed for GPC/LS and endgroup analysis using isocyanate terminated PEG (M w ¼ 1000 Da). 16,32A solution of Leu NCA in THF (0.33 equiv., 50 mg mL À1 ) was added and the polymerization was monitored by FTIR.Polymerization reactions were generally complete within 3 hours.Aer complete consumption of NCA, reactions were removed from the drybox and precipitated into hexanes.Solids were collected by centrifugation and washed with 2 portions of water at pH 2 (HCl), followed by DI water.The polymers were lyophilized to yield white solids (95-99% yield).

Molecular weight determination
The degree of polymerization (DP) of the rst segment, poly(agal-C) or poly(a-gal-K), was determined by 1 H NMR end-group analysis of the aliquot end-capped with PEG. 32Integrations were calibrated using the polyethylene glycol chemical shi found at d 3.64, and the polypeptide DPs were found to be (a-gal-C) 65 and (a-gal-K) 67 .Polydispersity indices were determined using GPC/ LS: M w /M n for (a-gal-C) 65 ¼ 1.09 and M w /M n for (a-gal-K) 67 ¼ 1.07.The DP of the oligoleucine segments were determined by 1 H NMR integrations relative to the glycosylated segment.Final

Removal of protecting groups from glycosylated diblock copolypeptides
To a solution of the protected diblock copolymer in DCM : methanol 1 : 2 (10 mg mL À1 ) was added hydrazine monohydrate (4 equiv./OAcgroup).The reactions were stirred overnight at room temperature and the product was then observed as a white precipitate.Reactions were quenched by addition of drops of acetone.Et 2 O was added and the solids collected by centrifugation (99% yield).The solids were taken up with water and transferred to 2000 molecular weight cutoff (MWCO) dialysis tubing and dialyzed against Millipore water for 3 days, with water changes twice per day.Dialyzed copolymers were lyophilized to dryness to yield white uffy solids (80% yield aer dialysis). Poly

Circular dichroism spectra of diblock glycopolypeptides
Circular dichroism spectra were recorded on an OLIS RSM CD spectrophotometer running in conventional scanning mode.Spectra (190-250 nm) were recorded in a quartz cuvette of 0.1 cm path length with samples prepared using Millipore deionized water.All spectra were recorded as an average of 3 scans.The spectra are reported in units of molar ellipticity [q] (deg cm 2 dmol À1 ).The formula used for calculating molar ellipticity, [q], was [q] ¼ (q Â 100 Â M w )/(c Â l) where q is the experimental ellipticity in millidegrees, M w is the average molecular weight of a residue in g mol À1 , c is the polypeptide concentration in mg mL À1 , and l is the cuvette pathlength in cm.The percent ahelical content of the glycopolypeptides was estimated using the formula % a-helix ¼ 100 Â (À[q] 222 + 3000)/39000), where [q] 222 is the measured molar ellipticity at 222 nm. 33The measured a-helicity of (a-gal-K) 65 L 20 was 94%.

Self assembly of diblock glycopolypeptides in water
Solid (a-gal-C O2 ) 65 L 20 or (a-gal-K) 65 L 20 was dispersed in THF to produce a 1% (w/v) suspension.The suspension was placed in a bath sonicator for 30 minutes to evenly disperse the polypeptide and reduce large particulates.An equivalent amount of Millipore water was then added to give a 0.5% (w/v) suspension.The suspension became clear as the solution was mixed by vortexing.The mixture was then dialyzed (2000 MWCO membrane) against Millipore water overnight with 3 water changes.Vesicular assemblies were also obtained via slow evaporation of the THF from the mixed solvent suspension.

Differential interference contrast (DIC) microscopy
Self assembled copolypeptide suspensions of (a-gal-C O2 ) 65 L 20 or (a-gal-K) 65 L 20 (0.5% (w/v)) in water were visualized on glass slides with a spacer between the slide and the coverslip (doublesided tape or Secure Seal Imaging Spacer, Grace Bio-labs) allowing the structures to be minimally disturbed during focusing.The samples were imaged using a Zeiss Axiovert 200 DIC/Fluorescence Inverted Optical Microscope.

Extrusion of vesicle assemblies
A 0.2% (w/v) aqueous (a-gal-C O2 ) 65 L 20 vesicle suspension was extruded using an Avanti Mini-Extruder.Serial extrusion of vesicle suspensions were performed through Whatman Nuceleopore Track-Etched polycarbonate (PC) membranes with decreasing lter pore sizes: 3 times through a 1.0 mm lter, 3 times through a 0.4 mm lter, and 3 times through a 0.2 mm lter.The PC membranes and lter supports were soaked in Millipore water for 10 minutes prior to extrusion.

Dynamic light scattering (DLS) on extruded vesicles
A 0.2% (w/v) aqueous solution of extruded (a-gal-C O2 ) 65 L 20 vesicles was placed in a disposable cuvette and analyzed with a Malvern Zetasizer Nano ZS model Zen 3600 (Malvern Instruments Inc, Westborough, MA).A total scattering intensity of approximately 1 Â 10 5 cps was targeted.

Laser scanning confocal microscopy (LSCM) of uorescently labeled vesicles
LSCM images of (a-gal-C O2 ) 65 L 20 aqueous vesicle suspensions were taken on a Leica Inverted TCS-SP1 MP-Inverted Confocal and Multiphoton Microscope equipped with an argon laser (476 and 488 nm blue lines), a diode (DPSS) laser (561 nm yellowgreen line), and a helium-neon laser (633 nm far red line).The copolypeptide suspensions (20 mL of 0.2% (w/v) in Millipore water) were incubated with lipophilic DiOC 18 uorescent dye (1 mL of 0.008% (w/v) in DMSO, Molecular Probes) for 2 hours before imaging.Suspensions of the uorescently labeled copolypeptides (0.2% (w/v)) were visualized on glass slides with a spacer between the slide and the cover slip (Secure Seal Imaging Spacer, Grace Bio-labs) allowing the self-assembled structures to be minimally disturbed during focusing.Imaging of an xy plane with an optical z-slice (700 nm) showed that the assemblies were water lled, unilamellar vesicles.

Transmission electron microscopy (TEM) of extruded vesicles
Extruded (a-gal-C O2 ) 65 L 20 vesicle suspensions were diluted to 0.1% (w/v) with Millipore water.Samples (4 mL) were placed on a 300 mesh Formvar/carbon coated copper grid (Ted Pella) and allowed to remain on the grid for 60 seconds.Filter paper was used to remove the residual sample.One drop of 2% (w/v) uranyl acetate (negative stain) was then placed on the grid for 90 seconds, and subsequently removed by washing with drops of millipore water and removing the excess liquid with lter paper.The grids were allowed to dry before imaging using a JEM 1200-EX (JEOL) transmission electron microscope at 80 kV.

Encapsulation of Texas Red labeled dextran in vesicles
Vesicles composed of (a-gal-C O2 ) 65 L 20 were prepared as described above, except the aqueous phase contained 0.125 mg mL À1 Texas Red labeled dextran (M n ¼ 3000 Da).Vesicle solutions were dialyzed in 8000 MWCO tubing overnight to remove unencapsulated Texas Red labeled dextran.As a control to test for dye adsorption to vesicles, pre-formed (a-gal-C O2 ) 65 L 20 vesicles were incubated for 16 hours with a 0.125 mg mL À1 Texas Red labeled dextran solution and then dialyzed in 8000 MWCO tubing overnight.The samples were then imaged using DIC and LSCM microscopy as described above.No uorescence was observed in the pre-formed vesicles incubated with Texas Red labeled dextran solution.

Evaluation of carbohydrate-lectin binding by turbidity
Ricinus communis Agglutinin I (RCA 120 ) was purchased from Vector labs, and Conconavalin A (Con A) was purchased from Sigma-Aldrich.Lectin solutions were prepared at a concentration of 2 mg mL À1 in 10 mM phosphate, 0.15 M NaCl, pH 7.8.Lectin solutions (600 mL) were transferred to cuvettes and baseline measurements were taken.A solution of (a-gal-C O2 ) 65 L 20 vesicles (140 nm average diameter) was prepared at a concentration of 1 mg mL À1 in DI water, and 60 mL of the solution was added to cuvettes containing either RCA 120 or Con A. Final glycopolypeptide concentrations were 3.3 mM.The solutions were gently mixed and absorbance spectra were recorded over time.All measurements were performed in triplicate.

Cell culture
The HeLa cell line was maintained in Minimum Essential Medium (MEM) supplemented with 10% FBS, 1 mM sodium pyruvate, 100 units per mL penicillin, and 100 mg mL À1 streptomycin at a pH of 7.4 in a 37 C humidied atmosphere with 5% CO 2 using standard tissue culture protocols.

MTS cell proliferation assay
The MTS cell proliferation assay (CellTiter 96 AQ ueous Non-Radioactive Cell Proliferation Assay) was used to quantify any cytotoxic effects of the (a-gal-C O2 ) 65 L 20 and R H 60 L 20 vesicle suspensions.HeLa cells were seeded at a density of 4 Â 10 4 cells per cm 2 on a 96-well plate prior to the experiment.At the start of the experiment, the cell culture medium was aspirated, and the cells were incubated in medium containing the copolypeptide vesicles for 5 h in a 37 C humidied atmosphere with 5% CO 2 .The incubation medium was the same as the cell culture medium except for the absence of FBS, penicillin, and streptomycin.Following the 5 h incubation period, the medium containing copolypeptide vesicles was aspirated.Fresh medium containing 20% MTS was then added to the cells.The cells were placed back into a CO 2 incubator for 1 h and then the absorbance at 490 nm (A 490 ) was measured with an Innite F200 plate reader (Tecan Systems Inc., San Jose, CA, USA).The background absorbance was read at 700 nm (A 700 ) and subtracted from A 490 .The relative survival of the cells at each polypeptide concentration was quantied by taking the ratio of the (A 490 -A 700 ) values and comparing between the experimental and control cells.
Attempts were made to assemble vesicles from the galactose containing copolypeptides using mixed solvent annealing, which has been found to assist formation of ordered nanostructures in many other block copolypeptide systems. 18,21The hydrophilic chain conformations of the galactosylated block copolypeptides were found to strongly inuence their self assembly in water.The sample with a disordered hydrophilic segment (a-gal-C O2 ) 65 L 20 , gave exclusively vesicles with diameters ranging from hundreds of nanometers to a few microns in diameter (Fig. 3A).The vesicular morphology of the (a-gal-C O2 ) 65 L 20 assemblies was conrmed by labeling their hydrophobic domains with DiOC 18 dye and imaging thin slices through suspensions of the samples using laser scanning confocal microscopy (LSCM), which revealed their membrane structure and hydrophilic interior (Fig. 3C).The ability of these vesicles to encapsulate hydrophilic cargos was also shown by their retention of Texas Red labeled dextran (M n ¼ 3000 Da) aer removal of unencapsulated cargo by dialysis (Fig. 3D).][8] In contrast to the results above, the highly a-helical sample, (a-gal-K) 65 L 20 gave nearly no vesicles, and instead an abundance of micron sized irregular aggregates and some platelike objects were observed (Fig. 3B).Comparable aggregation behavior has been observed previously in block copolypeptides containing similarly rigid hydrophilic segments. 18The inability of (a-gal-K) 65 L 20 to form vesicles is likely due to rigidity of these hydrophilic domains, which compacts these segments and limits their ability to effectively solubilize and stabilize the assemblies against further aggregation.The rod-like nature of the a-helical hydrophilic segments in (a-gal-K) 65 L 20 also likely acts to stiffen any membranes formed, leading to rigid sheetlike membranes that lack the exibility needed to accommodate vesicle curvature. 18Related block copolypeptides containing ahelical hydrophilic domains, 18 were also found to only form rigid, inexible membranes regardless of length of the hydrophilic segments used, indicating the helical chain conformation strongly inuences assembled morphology. 18The presence of different hydrophilic chain conformations in (a-gal-K) 65 L 20 and (a-gal-C O2 ) 65 L 20 thus signicantly altered their self-assembled structures, where the desired vesicles were favored by the disordered segments in (a-gal-C O2 ) 65 L 20 .The exibility in the poly(a-gal-C O2 ) segments also allows better mixing with water, essentially increasing their hydrophilicity, compared to the conformationally rigid poly(a-gal-K) chains.The more open structure of solvated disordered poly(a-gal-C O2 ) segments should also partially frustrate packing of the rigid hydrophobic oligoleucine segments, making the vesicle membranes themselves more dynamic, exible and able to accommodate curvature. 21o determine if (a-gal-C O2 ) 65 L 20 segments retain the bioactivity of their pendant galactose units, we incubated (a-gal-C O2 ) 65 L 20 vesicles with lectins in a precipitation assay.We chose Ricinus communis agglutinin (RCA 120 ) for the polymer binding lectin since it is known to specically and selectively bind to galactosyl groups, and concanavalin A (Con A) as a control lectin  that binds mannosyl and glucosyl, but not galactosyl, groups. 14hen the galactosyl-vesicles were incubated with RCA 120 , turbidity of the solution was found to increase rapidly from aggregation due to lectin binding (Fig. 4). 14,34Furthermore, the vesicle suspensions showed no increase in turbidity when incubated with Con A, indicating the interaction with RCA 120 is a specic binding interaction between this lectin and the galactosyl groups of the polymers.

Conclusions
Overall, we have found that the chain conformation of hydrophilic polypeptide segments can play a signicant role in dictating structure in self assembled nanostructures.The use of hydrophilic segments with disordered conformations in amphiphilic diblock copolypeptides was found to be particularly effective both for formation of vesicular assemblies as well as presentation of functionality in an accessible, active form.The (a-gal-C O2 ) 65 L 20 vesicles were also found to be minimally cytotoxic (Fig. 5), as compared to cationic polypeptide vesicles such as R H 60 L 20 , 19 making them attractive for development as biofunctional drug carriers with controlled nanostructure.

Fig. 5
Fig. 5 Relative survival of HeLa cells incubated for 5 hours with copolypeptide vesicles determined using the MTS assay.Samples are (a-gal-C O2 ) 65 L 20 (solid line) and R H 60 L 20 (dashed line).Error bars represent the standard deviation from an average of three measurements.

Table 1
Characterization and properties of (a-gal-C) 65 L 20 and (a-gal-K) 65 L 20 diblock copolypeptides a Hydrophilic segment lengths (number average molecular weight, M n , for a-gal-K and a-gal-C segments) and polydispersity index (M w /M n ) determined using gel permeation chromatography and 1 H NMR. b Calculated using 1 H NMR. c Total isolated yield of diblock glycopolypeptide.d Structures observed visually using optical microscopy (V ¼ vesicle, A ¼ irregular aggregate, P ¼ plate).