- Email: [email protected]
Nanostructure of hyaluronan acyl-derivatives in the solid state
Carbohydrate Polymers 195 (2018) 468–475
Contents lists available at ScienceDirect
Carbohydrate Polymers journal homepage: www.elsevier.com/locate/carbpol
Nanostructure of hyaluronan acyl-derivatives in the solid state a,⁎
T
Josef Chmelař , Petr Bělský , Jiří Mrázek , Daniel Švadlák , Martina Hermannová , Miroslav Šloufc, Ivan Krakovskýd, Daniela Šmejkalováa, Vladimír Velebnýa b
a
a
a
a
Contipro a.s., Dolní Dobrouč 401, 561 02, Czech Republic New Technologies Research Centre, University of West Bohemia, Univerzitní 8, 306 14 Pilsen, Czech Republic Institute of Macromolecular Chemistry AS CR, v.v.i., Heyrovského náměstí 2, 162 06 Prague 6, Czech Republic d Department of Macromolecular Physics, Faculty of Mathematics and Physics, Charles University, V Holešovičkách 2, 180 00 Prague 8, Czech Republic b c
A R T I C LE I N FO
A B S T R A C T
Keywords: Hyaluronan Hydrophobization Nanostructure Solid-state Small-angle X-ray scattering Electron microscopy
Acyl derivatives of hyaluronan (acyl-HA) are promising materials for biomedical applications. Depending on the acyl length and the degree of substitution, these derivatives range from self-assembling water-soluble polymers to materials insoluble in aqueous environments. The behaviour of acyl-HA was studied in solution, but little attention was paid to the solid state, despite its importance for applications such as medical device fabrication. We thus used X-ray scattering and electron microscopy to explore the solid-state nano-structure of acyl-HA. The set of samples included various substituents, substitution degrees and molecular weights. The obtained data showed that all studied acyl-HA materials contained structures with dimensions on the order of nanometres that were not present in unmodified HA. The size of the nanostructures increased with the acyl length, while the degree of substitution and molecular weight had negligible effects. We suggest that the observed nanostructure corresponds to a distribution of hydrophobic domains in a hydrophilic matrix of unmodified HA segments.
1. Introduction Hyaluronan (HA) is a naturally occurring anionic linear polysaccharide with the repeating unit consisting of D-glucuronic acid and N-acetyl-D-glucosamine linked by alternating β(1 → 4) and β(1 → 3) glycosidic bonds. HA is present, for example, in the extracellular matrix, skin, synovial fluid, vitreous humour, or various connective tissues. It plays an important role in processes such as tissue hydration, lubrication, or wound healing (Dicker et al., 2014). As an important constituent of our body, HA is biocompatible, biodegradable and nontoxic (Garg & Hales, 2004), making it a promising material for applications in medicine. Due to its hydrophilic character, native HA is soluble in water. However, insolubility or limited solubility is required for applications such as medical device fabrication or drug delivery. A common approach for overcoming HA solubility is covalent cross-linking. The published methods include processes based on enzymatic reactions (Dvorakova et al., 2014), photochemistry (Bobula et al., 2015), or chemical cross-linking agents such as polyvalent hydrazides (Vercruysse, Marecak, Marecek, & Prestwich, 1997), glutaraldehyde, divinyl sulfone and carbodiimides (Collins & Birkinshaw, 2007, 2008). Another option is to chemically modify HA with hydrophobic side groups such as long acyl chains (Creuzet, Kadi, Rinaudo, & Auzély-
⁎
Corresponding author. E-mail address: [email protected] (J. Chmelař).
https://doi.org/10.1016/j.carbpol.2018.04.111 Received 1 March 2018; Received in revised form 23 April 2018; Accepted 27 April 2018 Available online 30 April 2018 0144-8617/ © 2018 Elsevier Ltd. All rights reserved.
Velty, 2006; Finelli et al., 2014; Huerta-Angeles, Bobek, Příkopová, Šmejkalová, & Velebný, 2014; Šmejkalová et al., 2012) or octenyl succinic anhydride (Eenschooten et al., 2012). These hydrophobized derivatives can be used in drug delivery applications (Eenschooten et al., 2012; Choi et al., 2010; Šmejkalová et al., 2017) or for the preparation of polymeric films (Foglarová et al., 2016) and fibres (Zápotocký et al., 2016). The nanostructure of biomaterials can have a significant impact on their performance in applications (Tang et al., 2016). In the case of native HA, there is a number of studies concerned with the nanostructure and chain conformations both in the solid state (Cowman & Matsuoka, 2005; Haxaire, Braccini, Milas, Rinaudo, & Perez, 2000) and in aqueous solutions (Buhler & Boué, 2004; Cowman & Matsuoka, 2005; Matteini et al., 2009). On the contrary, the nano-scale behaviour of hydrophobized HA was only studied in solution with emphasis on drug delivery applications (Eenschooten et al., 2012; Šmejkalová et al., 2014). The solid-state structure of these derivatives was not yet explored, despite the intention to use them also in solid products. Apart from its practical importance, this topic is also interesting in primary research, since the nanostructure of polyelectrolytes such as HA is a complex problem (Svergun & Koch, 2003) that is still not completely understood. Small-angle X-ray scattering (SAXS) is commonly used for studying
Carbohydrate Polymers 195 (2018) 468–475
J. Chmelař et al.
receptor interactions. Furthermore, the molecular weight of HA is not expected to change significantly during the reaction (Huerta-Angeles et al., 2014). The exact reaction procedure for each derivative was chosen based on our previous experience. The palmitoyl (HA-C16) and lauroyl (HA-C12) derivatives were prepared by the acylation of HA with a symmetric fatty acid anhydride (Šmejkalová et al., 2012). Briefly, HA was first dissolved in demineralized water with a subsequent addition of triethylamine and the catalyst N,N-dimethylaminopyridine. After one hour of mixing, the solution was diluted with tetrahydrofuran and further homogenized. Next, the symmetric anhydride was added. At the end of the reaction, the mixture was diluted with 50% 2-propanol, followed by the addition of a saturated sodium chloride solution. The product was precipitated with absolute 2-propanol, washed, decanted and dried for 48 h at 40 °C. The hexyl (HA-C6), decanoyl (HA-C10) and oleyl (HA-C18:1) derivatives were prepared by a similar method that differs in the use of a mixed fatty acid anhydride instead of the symmetric one (Huerta-Angeles et al., 2014). The degree of HA substitution (DS) was determined by gas chromatography, as described elsewhere (Chmelař et al., 2017). Briefly, the method is based on the alkaline hydrolysis of the samples, extraction of the hydrolyzed fatty acids and their quantification by gas chromatography. Note that we express DS as the number of acyl chains per 100 HA repeating disaccharide units. For example, a DS of 20% means that there are in average 20 acyl chains on each 100 disaccharide units.
the nanostructure of polymers (Chu & Hsiao, 2001). However, papers describing the use of SAXS for HA based systems are relatively scarce. As examples, we would mention studies concerned with highly viscous solutions of cross-linked HA (Gamini et al., 2002) and HA-chitosan polyelectrolyte complexes (Lalevée et al., 2016). In the case of hydrophobically modified polysaccharides, SAXS was used, for example, to study the nanostructure of palmitoyl-chitosan hydrogels (Chiu et al., 2009) or the assembly of hydrophobically modified alginate in aqueous solution, sol and hydrogel (Choudhary & Bhatia, 2012). However, to the best of our knowledge, SAXS was not yet used to study hydrophobized HA in the solid state. Other methods used for studying the nanostructure of polysaccharides are small-angle neutron scattering and various light scattering techniques (Buhler & Boué, 2004; Maki, Furusawa, Dobashi, Sugimoto, & Wakabayashi, 2017). For visualization, atomic force microscopy (Moffat, Morris, Al-Assaf, & Gunning, 2016), transmission electron microscopy (Chiu et al., 2009) or highresolution scanning electron microscopy (Hussain & Jaisankar, 2017) were used. This paper aims to elucidate the solid-state nanostructure of HA modified with acyl side chains (acyl-HA). For this purpose, a large set of acyl-HA derivatives was prepared that covered a broad range of acyl chain lengths, degrees of substitution and HA molecular weights. Most experiments were carried out with powder samples. Furthermore, selected derivatives were also measured in the form of a film to show whether the nanostructure is influenced by the sample form and the method of its preparation (powders by precipitation, films by slow drying). For characterization, wide- and small-angle X-ray scattering, transmission electron microscopy and extra-high resolution scanning electron microscopy were used.
2.2. Preparation of acyl-HA films The acyl-HA films were prepared by solution casting in a custom built apparatus (Foglarová et al., 2016), where the solution is evaporated between two plates with controlled temperature. This design enables precise process control, ensuring good reproducibility. First, the acyl-HA solution was prepared by dissolving the polymer in 50 wt.% aqueous 2-propanol to a concentration of 10 mg/ml. To ensure complete dissolution, the solution was vigorously stirred at 25 °C for 18 h (Collins & Birkinshaw, 2013). The acyl-HA solution (20 ml) was applied into the drying apparatus on a hydrophobized glass substrate. The apparatus was then closed, the lower and upper plate temperature set to 50 and 20 °C, respectively, and the solution was left to dry for 6 h. Subsequently, the apparatus was opened and the dry film removed from the substrate. The used film preparation conditions were chosen based on previous experiments (Foglarová et al., 2016).
2. Material and methods 2.1. Material Commercial sodium hyaluronate (HA) with various molecular weights was provided by Contipro a.s. The molecular weight data were obtained using size exclusion chromatography with multiangle laser light scattering detection (SEC-MALLS), as described elsewhere (Podzimek, Hermannova, Bilerova, Bezakova, & Velebny, 2010). Sodium chloride, sodium hydroxide, trichloromethane, 2-propanol, hexane, triethylamine and tetrahydrofuran were obtained from Lachner. Hydrochloric acid, formic acid, undecanoic acid, and HPLC grade 2-propanol were obtained from Sigma-Aldrich. Fatty acids (hexanoic, decanoic, lauric, palmitic and cis-oleic), fatty acid anhydrides and N,Ndimethylaminopyridine were obtained from TCI Europe. Benzoyl chloride was obtained from Merck. The acyl-HA derivatives were prepared by the esterification of OH groups on the HA backbone (Fig. 1). Advantageously, the used modification does not involve the carboxyl groups that remain available for
2.3. X-ray scattering The small- and wide-angle X-ray scattering (SAXS/WAXS) experiments were performed on a SAXSess mc2 instrument (Anton Paar). All samples (powders and films) were measured in point collimation geometry using a GeniX Microfocus X-ray point source with a Cu anode (50 kV and 1 mA) and single-bounce focusing X-ray optics. Image plates together with a CyclonePlus® Reader (PerkinElmer, Inc.) were used to record 2D X-ray scattering patterns. During measurement, the whole chamber of the instrument was evacuated to a pressure below 1 mbar to minimize undesirable parasitic absorption and scattering by air. The range of the scattering vector magnitude q was 0.2 − 28 nm−1. Both the SAXS and WAXS range were thus covered. Note that q is defined as
q=
4π sin θ λ
(1)
where λ is the X-ray wavelength (0.15418 nm for CuKα) and θ is the half-scattering angle. The film samples were fixed directly in the sample holder and the powder samples were glued in between two pieces of scotch tape. Since all 2D patterns were circularly symmetric, the 1D radial scattering-intensity profiles were obtained from them by azimuthal averaging using the SAXSquant software (Anton Paar). The scattering intensities were corrected with respect to background scattering. Finally, incoherent-
Fig. 1. Scheme of acyl-HA structure denoting a substituted disaccharide unit, where R is the aliphatic chain of the acyl group and n is the number of repeating disaccharide units. 469
Carbohydrate Polymers 195 (2018) 468–475
J. Chmelař et al.
scattering background (assumed to be constant) was subtracted from the SAXS profiles. The profiles of powder samples show a typical I ∼ q−4 dependency (Porod scattering) in the low-q region originating from the scattering on the surfaces of powder particles. The scattering of interest, i.e., the component originating from nanostructures, is superimposed on this power-law scattering. Thus, the profiles of powder samples had to be further processed to extract the SAXS profile component owing to internal nanostructures from the overall profile. This procedure is described in the Supplementary material. The SAXS profiles of both the film and powder samples showed a distinct scattering peak. The position of the peak maximum qmax was used to determine the Bragg distance dB according to
dB =
2π qmax
Table 1 List of native HA and acyl-HA used in this study. Mw is the weight average molecular weight of HA used for the modification (SEC-MALLS data) and DS is the degree of substitution determined by gas chromatography (Chmelař et al., 2017).
(2)
The precise position of the peak maximum was determined by fitting the Lorentz peak function in a narrow interval around an estimated maximum. 2.4. Electron microscopy Nanostructure of bulk materials was observed using transmission electron microscopy (TEM) and extra-high resolution scanning electron microscopy (XHRSEM). The samples for both TEM and XHRSEM in the form of thin films were embedded in an epoxy resin (LRW resin; Polysciences, Inc.) and prepared by dry ultramicrotomy at room temperature on an Ultracut UCT ultramicrotome (Leica). Dry ultramicrotomy (dry cutting without water trough, with direct collection of the ultrathin sections on TEM grids) had to be used because of the very high sample hydrophilicity and water swelling. The thin sections (thickness 60 nm) were observed in a Tecnai G2 Spirit Twin 12 TEM microscope (FEI) at accelerating voltage 120 kV and bright field imaging. The cut sections were fixed on a conductive support, transferred to a Magellan 400 XHRSEM microscope (FEI), plasma-etched inside the microscope chamber (using internal plasma cleaning device; preetching 30 s, final etching 180 s), and observed at a low energy of landing electrons (2 keV) using in-lens secondary electrons detector and high magnifications (up to 200kx).
Side chain
Mw/kDa
DS/%
dB,powder/nm
dB,film/nm
HA-1 HA-2 HA-C6-1 HA-C10-1 HA-C12-1 HA-C12-2 HA-C12-3 HA-C12-4 HA-C12-5 HA-C12-6 HA-C16-1 HA-C16-2 HA-C16-3 HA-C16-4 HA-C16-5 HA-C16-6 HA-C16-7 HA-C18:1-1 HA-C18:1-2
None None Hexyl Decanoyl Lauroyl Lauroyl Lauroyl Lauroyl Lauroyl Lauroyl Palmitoyl Palmitoyl Palmitoyl Palmitoyl Palmitoyl Palmitoyl Palmitoyl Oleyl Oleyl
330 1000 40 250 330 330 330 330 330 330 330 1000 330 330 80 330 330 15 15
0 0 65 33 6 7 21 32 38 46 8 13 20 25 29 38 48 10 14
None None 3.0 3.9 5.0 5.0 5.1 4.6 4.8 4.5 6.1 5.6 5.5 5.6 5.7 5.7 5.4 6.2 6.4
None None 3.0 4.2 – – – – 5.0 – 6.7 – 7.1 6.7 – 6.8 – – 7.0
samples only the SAXS profiles originating from the nanostructures, which are obtained by subtracting the Porod scattering from the acylHA profiles, as described in Section 2.3. and the Supplementary material. The SAXS profiles of a native HA film (HA-2) and a palmitoyl-HA film (HA-C16-6) are depicted in Fig. 2b. Since the films are non-porous bulk materials, they do not exhibit surface scattering (contrary to the powders). The profile of the native HA film did not show any q-dependence, only background signal was observed. This suggests that the native HA film did not contain any nanostructures. On the other hand, the SAXS profile of the palmitoyl-HA film showed a distinct scattering peak, which is clear evidence for the presence of nanostructures. It thus seems that the form of the sample does not have any influence on the presence of nanostructures. The effect of DS on the SAXS profiles of acyl-HA powders (Fig. 3a–c) was studied using lauroyl-HA, palmitoyl-HA and oleyl-HA powders with various DS. The effect of the acyl chain length (Fig. 3d) was studied using hexanoyl, decanoyl, lauroyl, palmitoyl, and oleyl derivatives. Importantly, a distinct scattering peak was observed in all profiles. The scattering intensity increased with increasing DS and increasing acyl chain length. In the range of DS investigated, the position of the scattering peak did not depend on DS for a given acyl chain length. However, with increasing acyl chain length, the scattering peak clearly shifted to lower q values. The effects of DS and acyl chain length on the SAXS profiles of acylHA films are demonstrated in Fig. 4. The influence of DS was studied on a set of palmitoyl-HA films, while the effect of the acyl chain length was studied using hexanoyl, decanoyl, palmitoyl, and oleyl-HA films. As in the case of powders, a distinct scattering peak was observed in all profiles. The scattering intensity increased with increasing DS and increasing acyl chain length. The scattering peak shifted to lower q values with increasing acyl chain length, but its position did not change substantially with DS. As for the powder samples, the SAXS profiles of all acyl-HA films showed a distinct scattering peak. It can be noticed that the scattering peaks of the films are slightly shifted towards lower q with respect to the powders, indicating that the nanostructures in powders might be somewhat smaller. However, it is also possible that scattering on the powder particles could contribute to the observed shift. The second difference between powders and films is that the dependence of the intensity on DS is not monotonous in the case of powders. This is
3. Results The native HA and acyl-HA listed in Table 1 were analysed using Xray scattering. All samples were measured in the form of a powder and, additionally, selected samples were also measured in the form of a film. Note that the scattering vector magnitude range of the used instrument enables to study structures with dimensions ranging approximately from 1 to 25 nm (depending also on the contrast). The SAXS profiles of a native HA powder (HA-1) and a palmitoyl-HA powder (HA-C16-6) are shown in Fig. 2a. The native HA profile could be well fitted by the sum of the Porod’s law and a constant (representing background signal) as follows:
I = k q− 4 + B
Sample
(3)
where I is the scattering intensity, k is the proportionality constant, q is the scattering vector magnitude and B is the background constant. Porod’s law describes the surface (Porod) scattering of compact particles with a sharp interface (e.g., powder particles). Since the SAXS profile of the native HA powder contained only Porod scattering (i.e., scattering by the powder particles themselves), we can suggest that nanostructures (long range order) are not present in the native HA powder. On the other hand, the palmitoyl-HA sample showed an additional scattering peak superimposed on the Porod scattering of the powder (Fig. 2a), indicating the presence of nanostructures. Since we aim to study these nanostructures, we will further present for the powder 470
Carbohydrate Polymers 195 (2018) 468–475
J. Chmelař et al.
Fig. 2. Comparison of native HA (HA-1 powder, HA-2 film) and acyl-HA (HA-C16-6) SAXS profiles for powders (a) and films (b). Since only background signal was observed for native HA, the data are presented without the subtraction of a constant background.
scattering peak. This enabled to calculate the Bragg distance dB from the position of the scattering peak maximum (Eq. (2)). The obtained dB are listed in Table 1. The Bragg distance provides information about the characteristic scale in electron density variations (heterogeneity). However, the Bragg distance depends on the molecular structure of the system in a complex manner and its further direct interpretation is impossible without a priori assuming a structural model (Guinier & Fournet, 1955). Since we do not have sufficient information to develop such a model, the Bragg distance will be used only qualitatively.
because the thickness and void fraction of the irradiated powder layer was not fixed due to the applied experimental technique. Therefore, the intensities cannot be compared among each other. In the high-q region (Porod’s region of nanostructures), the SAXS profiles of both acyl-HA powders (Fig. 3) and films (Fig. 4) showed a q−n decay with n ≈ 4. This decay exponent is characteristic for systems consisting of multiple domains or phases separated by relatively sharp boundaries (Glatter & Kratky, 1982). The SAXS profiles of acyl-HA derivatives showed a distinct
Fig. 3. SAXS profiles of acyl-HA powder samples after the subtraction of Porod scattering: lauroyl-HA (a), palmitoyl-HA (b), and oleyl-HA (c) with various DS; samples with varying acyl chain length (d). 471
Carbohydrate Polymers 195 (2018) 468–475
J. Chmelař et al.
Fig. 4. SAXS profiles of acyl-HA film samples: palmitoyl-HA with various DS (a) and samples with varying acyl chain length (b).
Fig. 5. Dependence of the Bragg distance dB on the length of the acyl side chain.
Fig. 6. Dependence of the Bragg distance dB on the DS for HA-C12 and HA-C16 derivatives measured in the powder form.
The dependence of the Bragg distance dB on the acyl chain length is depicted in Fig. 5. One can clearly see that dB increases with increasing acyl chain length and that this increase is approximately linear. Note that the acyl chain length is expressed as the number of carbon atoms of the given fatty acid. The dB determined for the powders and films did not differ significantly, which is not surprising given the similarity of their SAXS profiles. The dB values obtained for the films were slightly larger than those obtained for the powders, in agreement with the already discussed shift of the acyl-HA film scattering peaks to lower q. The dependence of the Bragg distance dB on the degree of HA substitution is depicted in Fig. 6 for the HA-C12 and HA-C16 powder samples. One can clearly see that dB does not depend on the degree of HA substitution (in the studied range of properties), with average values of 4.8 and 5.7 nm for HA-C12 and HA-C16, respectively. The situation is analogous for the film samples, as demonstrated for HA-C16 films in the Supplementary material (Fig. S2). The used instrument is capable of measuring X-ray scattering also in the WAXS range, which enables to assess crystallinity. The SAXS/WAXS profiles of a native HA film (HA-1) and a palmitoyl-HA film (HA-C16-6) are depicted in Fig. 7. In the WAXS region, only the so-called amorphous halo was present for both native HA and acyl-HA. No crystalline reflections were observed in the investigated q-range (corresponding to a maximum 2θ of 41°). These data clearly show that the studied materials are not crystalline. Based on the SAXS data, two film samples were selected for microscopic observations: native HA (HA-1) as a control without nanostructure and palmitoyl-HA (HA-C16-6) as a sample with nanostructure.
Fig. 7. SAXS/WAXS profiles of a native HA film and a palmitoyl-HA film.
In a TEM microscope, both samples appeared morphologically homogeneous (results not shown). In a XRHSEM microscope after plasma etching, a certain difference between the two samples was found (Fig. 8). The native HA film again appeared morphologically homogeneous (Fig. 8a), while the palmitoyl-HA film displayed some faint
472
Carbohydrate Polymers 195 (2018) 468–475
J. Chmelař et al.
Fig. 8. XHRSEM micrographs showing plasma-etched cut surfaces of native HA (a) and palmitoyl-HA (b) films. The palmitoyl-HA film showed structures that could be enhanced using contrast enhancement (c) in combination with Wiener filter (d) and binarized using local threshold (e). The binary image was transformed to reciprocal space using 2D-DFFT (f) to evaluate periodic distances (Slouf & Vackova, 2014; Slouf et al., 2008).
morphology (Fig. 8b). After contrast adjustment using ImageJ (Schneider, Rasband, & Eliceiri, 2012), the morphological features were somewhat enhanced but the adjusted micrograph suffered from high noise (Fig. 8c). Better results were obtained by combining noise/shift reduction by means of Wiener filter (using Python/SciPy (Oliphant, 2007)) and contrast enhancement like in ImageJ (Fig. 8d). The image was binarized using automated local threshold in ImageJ (Fig. 8e) and transformed to reciprocal space using two-dimensional discrete fast Fourier transform (2D-DFFT; using Python/Scipy (Oliphant, 2007; Slouf & Vackova, 2014)). The final 2D-DFFT image (Fig. 8f) could be used to calculate the periodicity in the sample. The 2D-DFFT image was radially averaged (using Python/SciPy (Oliphant, 2007; Slouf & Vackova, 2014)) to obtain a 1D-DFFT radial profile (Fig. 9). The radial profile corresponds to a calculated 1D diffraction pattern of the observed structures and, consequently, the peaks in the 1D-DFFT profile represent periodic distances, d, which can be evaluated using a few simple formulas analogous to those used in the field of X-ray diffraction (Slouf et al., 2008). The calculation for the investigated palmitoyl-HA film is shown directly in Fig. 9. Although the morphology in XHRSEM micrographs was very weak and the micrographs had to be heavily processed to enhance the signal, it is noteworthy that the calculated periodic distance d ≈ 5.5 nm was in satisfactory agreement with the SAXS results. We can conclude that careful Fourier-transform-based analysis of XHRSEM micrographs showing plasma-etched ultramicrotomed samples confirmed the periodicity observed in the parallel SAXS experiments.
Fig. 9. Radial profile calculated from the Fourier-transformed image of the palmitoyl-HA film (Fig. 8f) showing a strong maximum corresponding to a real space periodic distance d ≈ 5.5 nm. The principle of the periodic distance calculation is given directly in the image, more details can be found elsewhere (Slouf & Vackova, 2014; Slouf et al., 2008).
particles. This confirms that native HA is not a nanostructured material, i.e., it forms one homogeneous phase on the nanoscale. On the other hand, all studied acyl-HA derivatives exhibited distinct scattering peaks with various intensities. Analogous results were obtained by XHRSEM, which showed periodic morphological features for the palmitoyl-HA film, while the native HA film appeared morphologically homogeneous. This clearly shows that the nanostructure (long range order) observed for acyl-HA is related to the presence of the hydrophobic acyl side chains. The SAXS profiles and the evaluated dB showed only a minor
4. Discussion The data presented in the Results revealed several interesting features. Importantly, the SAXS profiles of native HA did not show any scattering peaks or shoulders or Porod scattering, except for the powder sample, where the Porod scattering originates from the powder 473
Carbohydrate Polymers 195 (2018) 468–475
J. Chmelař et al.
5. Conclusions
influence of sample form, despite the significantly different ways of solidification: the powder samples were precipitated, while the films were produced by slow solvent evaporation. Therefore, the observed solid-state nanostructure seems to be inherent to the materials. The WAXS profiles only contained a weak broad halo (Fig. 7) representing short-range order (length scale below 1 nm) that is characteristic for liquids and glasses, showing that the acyl-HA derivatives are amorphous. When these findings are combined, we can conclude that the nanostructure of acyl-HA is related to local organization of amorphous polymer chains that is induced by the presence of the acyl side chains. Considering the chemical structure of acyl-HA, we suggest that the observed solid-state nanostructure corresponds to a distribution of hydrophobic domains in a hydrophilic matrix (microphase-separated structure). The Bragg distance dB obtained from the SAXS profiles could be related to the average distance of the hydrophobic domains. The creation of this structure is driven by unfavorable interactions between the hydrophobic acyl chains and the hydrophilic HA backbone. Such interactions are well known especially in solution, where they can lead to the formation of polymeric micelles (Šmejkalová et al., 2017). In the solid-state, the hydrophobic domains act as physical cross-links and, at sufficient DS and acyl chain length, can provide insolubility or limited solubility to forms such as films (Foglarová et al., 2016) or fibres (Zápotocký et al., 2016). A single hydrophobic domain (as seen by SAXS, i.e., a domain differing in electronic density) should certainly contain the interacting acyl chains. It is not possible to reliably determine whether the modified HA backbone segments are also part of this domain. However, some information can be obtained from the Porod region of the SAXS profile of nanostructures. In this region, the SAXS profiles showed a q−n decay with n ≈ 4, indicating a nanostructure with multiple domains or phases that have a relatively sharp boundary. This means that the hydrophobic domains in acyl-HA are rather well defined. The persistence length of HA was reported to be around 10 nm (Buhler & Boué, 2004; Carn et al., 2012). This value is higher than the Bragg distance, suggesting that HA chains will behave as rigid at the length scale of the nanostructures. Consequently, the HA chains will be expanded and the acyl groups of a single acyl-HA chain will contribute to a larger number of hydrophobic domains. In the literature, helical conformations were suggested for native HA in the solid state (Cowman & Matsuoka, 2005; Haxaire et al., 2000). Our results are not in conflict with these findings, since the backbone chain itself could be in a helical conformation, although some defects caused by the presence of the acyl side chains and their interactions can be expected. It is also interesting to compare dB to the length of the acyl side chain lacyl, which can be calculated from the number of carbon atoms of the given acyl chain NC as (Tanford, 1980): lacyl (nm) = 0.15 + 0.1265 NC
We studied the nanostructure of acyl-HA in the solid state using Xray scattering and electron microscopy. The SAXS profiles of unmodified HA did not show any scattering peaks, while all studied acylHA samples exhibited distinct peaks corresponding to nanometre-sized structures. Analogous results were obtained by XHRSEM. The SAXS data showed only minor influence of sample form (powder or film), so the observed nanostructure is not a result of sample preparation. The size of the nanostructures increased with the acyl chain length, while the degree of substitution and molecular weight had negligible effects. WAXS profiles showed that the samples were non-crystalline. Based on these results, we suggest that the solid-state nanostructure of acyl-HA corresponds to a distribution of hydrophobic domains in a hydrophilic matrix of unmodified HA segments. The hydrophobic domains act as physical cross-links and can provide insolubility or limited solubility to the material similarly to covalent cross-linking. These findings contribute to our understanding of the macroscopically observed behaviour of acyl-HA materials and should improve our ability to design final product properties (e.g., the swelling and degradation of resorbable implants). Declarations of interest None. Acknowledgements On the part of New Technologies - Research Centre, the result was developed within the CENTEM project, reg. no. CZ.1.05/2.1.00/ 03.0088, cofunded by the ERDF as part of the Ministry of Education, Youth and Sports OP RDI programme and, in the follow-up sustainability stage, supported through CENTEM PLUS (LO1402) by financial means from the Ministry of Education, Youth and Sports under the National Sustainability Programme I. Electron microscopy at the Institute of Macromolecular Chemistry (IMC) and at the collaborating Institute of Scientific Instruments (ISI) was supported by project TE01020118 (Technology Agency of the CR); the final XHRSEM micrographs come from ISI and were obtained by Aleš Paták. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.carbpol.2018.04.111. References
(4)
Šmejkalová, D., Hermannová, M., Šuláková, R., Průšová, A., Kučerík, J., & Velebný, V. (2012). Structural and conformational differences of acylated hyaluronan modified in protic and aprotic solvent system. Carbohydrate Polymers, 87(2), 1460–1466. Šmejkalová, D., Nešporová, K., Hermannová, M., Huerta-Angeles, G., Čožíková, D., Vištejnová, L., et al. (2014). Paclitaxel isomerisation in polymeric micelles based on hydrophobized hyaluronic acid. International Journal of Pharmaceutics, 466(1–2), 147–155. Šmejkalová, D., Muthný, T., Nešporová, K., Hermannová, M., Achbergerová, E., HuertaAngeles, G., et al. (2017). Hyaluronan polymeric micelles for topical drug delivery. Carbohydrate Polymers, 156, 86–96. Bobula, T., Běťák, J., Buffa, R., Moravcová, M., Klein, P., Žídek, O., et al. (2015). Solidstate photocrosslinking of hyaluronan microfibres. Carbohydrate Polymers, 125, 153–160. Buhler, E., & Boué, F. (2004). Chain persistence length and structure in hyaluronan solutions: Ionic strength dependence for a model semirigid polyelectrolyte. Macromolecules, 37(4), 1600–1610. Carn, F., Guyot, S., Baron, A., Perez, J., Buhler, E., & Zanchi, D. (2012). Structural properties of colloidal complexes between condensed tannins and polysaccharide hyaluronan. Biomacromolecules, 13(3), 751–759. Chiu, Y. L., Chen, S. C., Su, C. J., Hsiao, C. W., Chen, Y. M., Chen, H. L., et al. (2009). pHtriggered injectable hydrogels prepared from aqueous N-palmitoyl chitosan: In vitro characteristics and in vivo biocompatibility. Biomaterials, 30(28), 4877–4888. Chmelař, J., Kotzianová, A., Hermannová, M., Šuláková, R., Šmejkalová, D., Kulhánek, J., et al. (2017). Evaluating the degree of substitution of water-insoluble acyl derivatives
Since more data were measured for powders, further analysis will be limited to these samples. The dependence of dB on lacyl is linear (see Fig. S3 in the Supplementary material) and can thus be easily fitted, yielding: dB (nm) = 2.07 lacyl + 1.23
(5)
The slope is close to 2, suggesting that dB is somehow related to twice the length of the acyl chain. The intersection of 1.23 is relatively close to the value of 1 nm, which was reported as the length of the HA repeating unit (Finelli et al., 2014). Unfortunately, it is not possible to directly relate these observations to the nanostructure based on the measured data, but they might be valuable input into further studies utilizing, for example, molecular simulations. In general, simulations or thermodynamic calculations (Lundy et al., 2017) should be the next step in elucidating the nanostructure of acyl-HA materials.
474
Carbohydrate Polymers 195 (2018) 468–475
J. Chmelař et al.
behavior of hyaluronan in relation to its physical properties as probed by molecular modeling. Glycobiology, 10(6), 587–594. Huerta-Angeles, G., Bobek, M., Příkopová, E., Šmejkalová, D., & Velebný, V. (2014). Novel synthetic method for the preparation of amphiphilic hyaluronan by means of aliphatic aromatic anhydrides. Carbohydrate Polymers, 111, 883–891. Hussain, I. S. A., & Jaisankar, V. (2017). An eco-friendly synthesis: Characterization and antibacterial applications of novel almond gum −poly(acrylamide) based hydrogel silver nanocomposite. Polymer Testing, 62, 154–161. Lalevée, G., Sudre, G., Montembault, A., Meadows, J., Malaise, S., Crépet, A., et al. (2016). Polyelectrolyte complexes via desalting mixtures of hyaluronic acid and chitosan—Physicochemical study and structural analysis. Carbohydrate Polymers, 154, 86–95. Lundy, R., Flynn, S. P., Cummins, C., Kelleher, S. M., Collins, M. N., Dalton, E., et al. (2017). Controlled solvent vapor annealing of a high χ block copolymer thin film. Physical Chemistry Chemical Physics, 19, 2805–2815. Maki, Y., Furusawa, K., Dobashi, T., Sugimoto, Y., & Wakabayashi, K. (2017). Small-angle X-ray and light scattering analysis of multi-layered Curdlan gels prepared by a diffusion method. Carbohydrate Polymers, 155, 136–145. Matteini, P., Dei, L., Carretti, E., Volpi, N., Goti, A., & Pini, R. (2009). Structural behavior of highly concentrated hyaluronan. Biomacromolecules, 10(6), 1516–1522. Moffat, J., Morris, V. J., Al-Assaf, S., & Gunning, A. P. (2016). Visualisation of xanthan conformation by atomic force microscopy. Carbohydrate Polymers, 148, 380–389. Oliphant, T. E. (2007). Python for scientific computing. Computing in Science & Engineering, 9, 10–20. Podzimek, S., Hermannova, M., Bilerova, H., Bezakova, Z., & Velebny, V. (2010). Solution properties of hyaluronic acid and comparison of SEC-MALS-VIS datawith off-line capillary viscometry. Journal of Applied Polymer Science, 116(5), 3013–3020. Schneider, C. A., Rasband, W. S., & Eliceiri, K. W. (2012). NIH Image to ImageJ: 25 years of image analysis. Nature Methods, 9, 671–675. Slouf, M., & Vackova, T. (2014). Application of fourier transformation in image analysis of micrographs. Materials Structure, 22, 91–94. Slouf, M., Synkova, H., Baldrian, J., Marek, A., Kovarova, J., Schmidt, P., et al. (2008). Structural changes of UHMWPE after e-Beam irradiation and thermal treatment. Journal of Biomedical Materials Research Part B: Applied Biomaterials, 85B, 240–251. Svergun, D. I., & Koch, M. H. J. (2003). Small-angle scattering studies of biological macromolecules in solution. Reports on Progress in Physics, 66, 1735–1782. Tanford, C. (1980). The hydrophobic effect. New York: John Wiley & Sons. Tang, Z., He, C., Tian, H., Ding, J., Hsiao, B. S., Chu, B., et al. (2016). Polymeric nanostructured materials for biomedical applications. Progress in Polymer Science, 60, 86–128. Vercruysse, K. P., Marecak, D. M., Marecek, J. F., & Prestwich, G. D. (1997). Synthesis and in vitro degradation of new polyvalent hydrazide cross-linked hydrogels of hyaluronic acid. Bioconjugate Chemistry, 8(5), 686–694. Zápotocký, V., Pospíšilová, M., Janouchová, K., Švadlák, D., Baťová, J., Šogorková, J., et al. (2016). Fabrication of biodegradable textile scaffold based on hydrophobized hyaluronic acid. International Journal of Biological Macromolecules, 95, 903–909.
of hyaluronan using Raman spectroscopy: Method development and comparison with gas chromatography and 1H NMR. Analytical Methods, 9, 232–239. Choi, K. Y., Chung, H., Min, K. H., Yoon, H. Y., Kim, K., Park, J. H., et al. (2010). Selfassembled hyaluronic acid nanoparticles for active tumor targeting. Biomaterials, 31(1), 106–114. Choudhary, S., & Bhatia, S. R. (2012). Rheology and nanostructure of hydrophobically modified alginate (HMA) gels and solutions. Carbohydrate Polymers, 87(1), 524–530. Chu, B., & Hsiao, B. S. (2001). Small-angle X-ray scattering of polymers. Chemical Reviews, 101(6), 1727–1761. Collins, M. N., & Birkinshaw, C. (2007). Comparison of the effectiveness of four different crosslinking agents with hyaluronic acid hydrogel films for tissue-culture applications. Journal of Applied Polymer Science, 104(5), 3183–3191. Collins, M. N., & Birkinshaw, C. (2008). Investigation of the swelling behavior of crosslinked hyaluronic acid films and hydrogels produced using homogeneous reactions. Journal of Applied Polymer Science, 109(2), 923–931. Collins, M. N., & Birkinshaw, C. (2013). Hyaluronic acid solutions—A processing method for efficient chemical modification. Journal of Applied Polymer Science, 130(1), 145–152. Cowman, M. K., & Matsuoka, S. (2005). Experimental approaches to hyaluronan structure. Carbohydrate Research, 340(5), 791–809. Creuzet, C., Kadi, S., Rinaudo, M., & Auzély-Velty, R. (2006). New associative systems based on alkylated hyaluronic acid. Synthesis and aqueous solution properties. Polymer, 47(8), 2706–2713. Dicker, K. T., Gurski, L. A., Pradhan-Bhatt, S., Witt, R. L., Farach-Carson, M. C., & Jia, X. (2014). Hyaluronan: A simple polysaccharide with diverse biological functions. Acta Biomaterialia, 10(4), 1558–1570. Dvorakova, J., Kucera, L., Kucera, J., Svik, K., Foglarova, M., Muthny, T., et al. (2014). Chondrogenic differentiation of mesenchymal stem cells in a hydrogel system based on an enzymatically crosslinked tyramine derivative of hyaluronan. Journal of Biomedical Materials Research Part A, 102(10), 3523–3530. Eenschooten, C., Vaccaro, A., Delie, F., Guillaumie, F., Tømmeraas, K., Kontogeorgis, G. M., et al. (2012). Novel self-associative and multiphasic nanostructured soft carriers based on amphiphilic hyaluronic acid derivatives. Carbohydrate Polymers, 87(1), 444–451. Finelli, I., Chiessi, E., Oddo, L., Galesso, D., Renier, D., & Paradossi, G. (2014). Collective dynamics and transient behavior of partially hydrophobic hyaluronic acid chains. Macromolecular Chemistry and Physics, 215(2), 140–147. Foglarová, M., Chmelař, J., Huerta-Angeles, G., Vágnerová, H., Kulhánek, J., Tománková, K. B., et al. (2016). Water-insoluble thin films from palmitoyl hyaluronan with tunable properties. Carbohydrate Polymers, 144, 68–75. Gamini, A., Paoletti, S., Toffanin, R., Micali, F., Michielin, L., & Bevilacqua, C. (2002). Structural investigations of cross-linked hyaluronan. Biomaterials, 23(4), 1161–1167. Garg, H., & Hales, C. (2004). Chemistry and biology of hyaluronan. Oxford: Elsevier. Glatter, O., & Kratky, O. (1982). Small-angle X-ray scattering. London: Academic Press Inc. Guinier, A., & Fournet, G. (1955). Small angle scattering of X-rays. New York: Wiley. Haxaire, K., Braccini, I., Milas, M., Rinaudo, M., & Perez, S. (2000). Conformational
475
Contents lists available at ScienceDirect
Carbohydrate Polymers journal homepage: www.elsevier.com/locate/carbpol
Nanostructure of hyaluronan acyl-derivatives in the solid state a,⁎
T
Josef Chmelař , Petr Bělský , Jiří Mrázek , Daniel Švadlák , Martina Hermannová , Miroslav Šloufc, Ivan Krakovskýd, Daniela Šmejkalováa, Vladimír Velebnýa b
a
a
a
a
Contipro a.s., Dolní Dobrouč 401, 561 02, Czech Republic New Technologies Research Centre, University of West Bohemia, Univerzitní 8, 306 14 Pilsen, Czech Republic Institute of Macromolecular Chemistry AS CR, v.v.i., Heyrovského náměstí 2, 162 06 Prague 6, Czech Republic d Department of Macromolecular Physics, Faculty of Mathematics and Physics, Charles University, V Holešovičkách 2, 180 00 Prague 8, Czech Republic b c
A R T I C LE I N FO
A B S T R A C T
Keywords: Hyaluronan Hydrophobization Nanostructure Solid-state Small-angle X-ray scattering Electron microscopy
Acyl derivatives of hyaluronan (acyl-HA) are promising materials for biomedical applications. Depending on the acyl length and the degree of substitution, these derivatives range from self-assembling water-soluble polymers to materials insoluble in aqueous environments. The behaviour of acyl-HA was studied in solution, but little attention was paid to the solid state, despite its importance for applications such as medical device fabrication. We thus used X-ray scattering and electron microscopy to explore the solid-state nano-structure of acyl-HA. The set of samples included various substituents, substitution degrees and molecular weights. The obtained data showed that all studied acyl-HA materials contained structures with dimensions on the order of nanometres that were not present in unmodified HA. The size of the nanostructures increased with the acyl length, while the degree of substitution and molecular weight had negligible effects. We suggest that the observed nanostructure corresponds to a distribution of hydrophobic domains in a hydrophilic matrix of unmodified HA segments.
1. Introduction Hyaluronan (HA) is a naturally occurring anionic linear polysaccharide with the repeating unit consisting of D-glucuronic acid and N-acetyl-D-glucosamine linked by alternating β(1 → 4) and β(1 → 3) glycosidic bonds. HA is present, for example, in the extracellular matrix, skin, synovial fluid, vitreous humour, or various connective tissues. It plays an important role in processes such as tissue hydration, lubrication, or wound healing (Dicker et al., 2014). As an important constituent of our body, HA is biocompatible, biodegradable and nontoxic (Garg & Hales, 2004), making it a promising material for applications in medicine. Due to its hydrophilic character, native HA is soluble in water. However, insolubility or limited solubility is required for applications such as medical device fabrication or drug delivery. A common approach for overcoming HA solubility is covalent cross-linking. The published methods include processes based on enzymatic reactions (Dvorakova et al., 2014), photochemistry (Bobula et al., 2015), or chemical cross-linking agents such as polyvalent hydrazides (Vercruysse, Marecak, Marecek, & Prestwich, 1997), glutaraldehyde, divinyl sulfone and carbodiimides (Collins & Birkinshaw, 2007, 2008). Another option is to chemically modify HA with hydrophobic side groups such as long acyl chains (Creuzet, Kadi, Rinaudo, & Auzély-
⁎
Corresponding author. E-mail address: [email protected] (J. Chmelař).
https://doi.org/10.1016/j.carbpol.2018.04.111 Received 1 March 2018; Received in revised form 23 April 2018; Accepted 27 April 2018 Available online 30 April 2018 0144-8617/ © 2018 Elsevier Ltd. All rights reserved.
Velty, 2006; Finelli et al., 2014; Huerta-Angeles, Bobek, Příkopová, Šmejkalová, & Velebný, 2014; Šmejkalová et al., 2012) or octenyl succinic anhydride (Eenschooten et al., 2012). These hydrophobized derivatives can be used in drug delivery applications (Eenschooten et al., 2012; Choi et al., 2010; Šmejkalová et al., 2017) or for the preparation of polymeric films (Foglarová et al., 2016) and fibres (Zápotocký et al., 2016). The nanostructure of biomaterials can have a significant impact on their performance in applications (Tang et al., 2016). In the case of native HA, there is a number of studies concerned with the nanostructure and chain conformations both in the solid state (Cowman & Matsuoka, 2005; Haxaire, Braccini, Milas, Rinaudo, & Perez, 2000) and in aqueous solutions (Buhler & Boué, 2004; Cowman & Matsuoka, 2005; Matteini et al., 2009). On the contrary, the nano-scale behaviour of hydrophobized HA was only studied in solution with emphasis on drug delivery applications (Eenschooten et al., 2012; Šmejkalová et al., 2014). The solid-state structure of these derivatives was not yet explored, despite the intention to use them also in solid products. Apart from its practical importance, this topic is also interesting in primary research, since the nanostructure of polyelectrolytes such as HA is a complex problem (Svergun & Koch, 2003) that is still not completely understood. Small-angle X-ray scattering (SAXS) is commonly used for studying
Carbohydrate Polymers 195 (2018) 468–475
J. Chmelař et al.
receptor interactions. Furthermore, the molecular weight of HA is not expected to change significantly during the reaction (Huerta-Angeles et al., 2014). The exact reaction procedure for each derivative was chosen based on our previous experience. The palmitoyl (HA-C16) and lauroyl (HA-C12) derivatives were prepared by the acylation of HA with a symmetric fatty acid anhydride (Šmejkalová et al., 2012). Briefly, HA was first dissolved in demineralized water with a subsequent addition of triethylamine and the catalyst N,N-dimethylaminopyridine. After one hour of mixing, the solution was diluted with tetrahydrofuran and further homogenized. Next, the symmetric anhydride was added. At the end of the reaction, the mixture was diluted with 50% 2-propanol, followed by the addition of a saturated sodium chloride solution. The product was precipitated with absolute 2-propanol, washed, decanted and dried for 48 h at 40 °C. The hexyl (HA-C6), decanoyl (HA-C10) and oleyl (HA-C18:1) derivatives were prepared by a similar method that differs in the use of a mixed fatty acid anhydride instead of the symmetric one (Huerta-Angeles et al., 2014). The degree of HA substitution (DS) was determined by gas chromatography, as described elsewhere (Chmelař et al., 2017). Briefly, the method is based on the alkaline hydrolysis of the samples, extraction of the hydrolyzed fatty acids and their quantification by gas chromatography. Note that we express DS as the number of acyl chains per 100 HA repeating disaccharide units. For example, a DS of 20% means that there are in average 20 acyl chains on each 100 disaccharide units.
the nanostructure of polymers (Chu & Hsiao, 2001). However, papers describing the use of SAXS for HA based systems are relatively scarce. As examples, we would mention studies concerned with highly viscous solutions of cross-linked HA (Gamini et al., 2002) and HA-chitosan polyelectrolyte complexes (Lalevée et al., 2016). In the case of hydrophobically modified polysaccharides, SAXS was used, for example, to study the nanostructure of palmitoyl-chitosan hydrogels (Chiu et al., 2009) or the assembly of hydrophobically modified alginate in aqueous solution, sol and hydrogel (Choudhary & Bhatia, 2012). However, to the best of our knowledge, SAXS was not yet used to study hydrophobized HA in the solid state. Other methods used for studying the nanostructure of polysaccharides are small-angle neutron scattering and various light scattering techniques (Buhler & Boué, 2004; Maki, Furusawa, Dobashi, Sugimoto, & Wakabayashi, 2017). For visualization, atomic force microscopy (Moffat, Morris, Al-Assaf, & Gunning, 2016), transmission electron microscopy (Chiu et al., 2009) or highresolution scanning electron microscopy (Hussain & Jaisankar, 2017) were used. This paper aims to elucidate the solid-state nanostructure of HA modified with acyl side chains (acyl-HA). For this purpose, a large set of acyl-HA derivatives was prepared that covered a broad range of acyl chain lengths, degrees of substitution and HA molecular weights. Most experiments were carried out with powder samples. Furthermore, selected derivatives were also measured in the form of a film to show whether the nanostructure is influenced by the sample form and the method of its preparation (powders by precipitation, films by slow drying). For characterization, wide- and small-angle X-ray scattering, transmission electron microscopy and extra-high resolution scanning electron microscopy were used.
2.2. Preparation of acyl-HA films The acyl-HA films were prepared by solution casting in a custom built apparatus (Foglarová et al., 2016), where the solution is evaporated between two plates with controlled temperature. This design enables precise process control, ensuring good reproducibility. First, the acyl-HA solution was prepared by dissolving the polymer in 50 wt.% aqueous 2-propanol to a concentration of 10 mg/ml. To ensure complete dissolution, the solution was vigorously stirred at 25 °C for 18 h (Collins & Birkinshaw, 2013). The acyl-HA solution (20 ml) was applied into the drying apparatus on a hydrophobized glass substrate. The apparatus was then closed, the lower and upper plate temperature set to 50 and 20 °C, respectively, and the solution was left to dry for 6 h. Subsequently, the apparatus was opened and the dry film removed from the substrate. The used film preparation conditions were chosen based on previous experiments (Foglarová et al., 2016).
2. Material and methods 2.1. Material Commercial sodium hyaluronate (HA) with various molecular weights was provided by Contipro a.s. The molecular weight data were obtained using size exclusion chromatography with multiangle laser light scattering detection (SEC-MALLS), as described elsewhere (Podzimek, Hermannova, Bilerova, Bezakova, & Velebny, 2010). Sodium chloride, sodium hydroxide, trichloromethane, 2-propanol, hexane, triethylamine and tetrahydrofuran were obtained from Lachner. Hydrochloric acid, formic acid, undecanoic acid, and HPLC grade 2-propanol were obtained from Sigma-Aldrich. Fatty acids (hexanoic, decanoic, lauric, palmitic and cis-oleic), fatty acid anhydrides and N,Ndimethylaminopyridine were obtained from TCI Europe. Benzoyl chloride was obtained from Merck. The acyl-HA derivatives were prepared by the esterification of OH groups on the HA backbone (Fig. 1). Advantageously, the used modification does not involve the carboxyl groups that remain available for
2.3. X-ray scattering The small- and wide-angle X-ray scattering (SAXS/WAXS) experiments were performed on a SAXSess mc2 instrument (Anton Paar). All samples (powders and films) were measured in point collimation geometry using a GeniX Microfocus X-ray point source with a Cu anode (50 kV and 1 mA) and single-bounce focusing X-ray optics. Image plates together with a CyclonePlus® Reader (PerkinElmer, Inc.) were used to record 2D X-ray scattering patterns. During measurement, the whole chamber of the instrument was evacuated to a pressure below 1 mbar to minimize undesirable parasitic absorption and scattering by air. The range of the scattering vector magnitude q was 0.2 − 28 nm−1. Both the SAXS and WAXS range were thus covered. Note that q is defined as
q=
4π sin θ λ
(1)
where λ is the X-ray wavelength (0.15418 nm for CuKα) and θ is the half-scattering angle. The film samples were fixed directly in the sample holder and the powder samples were glued in between two pieces of scotch tape. Since all 2D patterns were circularly symmetric, the 1D radial scattering-intensity profiles were obtained from them by azimuthal averaging using the SAXSquant software (Anton Paar). The scattering intensities were corrected with respect to background scattering. Finally, incoherent-
Fig. 1. Scheme of acyl-HA structure denoting a substituted disaccharide unit, where R is the aliphatic chain of the acyl group and n is the number of repeating disaccharide units. 469
Carbohydrate Polymers 195 (2018) 468–475
J. Chmelař et al.
scattering background (assumed to be constant) was subtracted from the SAXS profiles. The profiles of powder samples show a typical I ∼ q−4 dependency (Porod scattering) in the low-q region originating from the scattering on the surfaces of powder particles. The scattering of interest, i.e., the component originating from nanostructures, is superimposed on this power-law scattering. Thus, the profiles of powder samples had to be further processed to extract the SAXS profile component owing to internal nanostructures from the overall profile. This procedure is described in the Supplementary material. The SAXS profiles of both the film and powder samples showed a distinct scattering peak. The position of the peak maximum qmax was used to determine the Bragg distance dB according to
dB =
2π qmax
Table 1 List of native HA and acyl-HA used in this study. Mw is the weight average molecular weight of HA used for the modification (SEC-MALLS data) and DS is the degree of substitution determined by gas chromatography (Chmelař et al., 2017).
(2)
The precise position of the peak maximum was determined by fitting the Lorentz peak function in a narrow interval around an estimated maximum. 2.4. Electron microscopy Nanostructure of bulk materials was observed using transmission electron microscopy (TEM) and extra-high resolution scanning electron microscopy (XHRSEM). The samples for both TEM and XHRSEM in the form of thin films were embedded in an epoxy resin (LRW resin; Polysciences, Inc.) and prepared by dry ultramicrotomy at room temperature on an Ultracut UCT ultramicrotome (Leica). Dry ultramicrotomy (dry cutting without water trough, with direct collection of the ultrathin sections on TEM grids) had to be used because of the very high sample hydrophilicity and water swelling. The thin sections (thickness 60 nm) were observed in a Tecnai G2 Spirit Twin 12 TEM microscope (FEI) at accelerating voltage 120 kV and bright field imaging. The cut sections were fixed on a conductive support, transferred to a Magellan 400 XHRSEM microscope (FEI), plasma-etched inside the microscope chamber (using internal plasma cleaning device; preetching 30 s, final etching 180 s), and observed at a low energy of landing electrons (2 keV) using in-lens secondary electrons detector and high magnifications (up to 200kx).
Side chain
Mw/kDa
DS/%
dB,powder/nm
dB,film/nm
HA-1 HA-2 HA-C6-1 HA-C10-1 HA-C12-1 HA-C12-2 HA-C12-3 HA-C12-4 HA-C12-5 HA-C12-6 HA-C16-1 HA-C16-2 HA-C16-3 HA-C16-4 HA-C16-5 HA-C16-6 HA-C16-7 HA-C18:1-1 HA-C18:1-2
None None Hexyl Decanoyl Lauroyl Lauroyl Lauroyl Lauroyl Lauroyl Lauroyl Palmitoyl Palmitoyl Palmitoyl Palmitoyl Palmitoyl Palmitoyl Palmitoyl Oleyl Oleyl
330 1000 40 250 330 330 330 330 330 330 330 1000 330 330 80 330 330 15 15
0 0 65 33 6 7 21 32 38 46 8 13 20 25 29 38 48 10 14
None None 3.0 3.9 5.0 5.0 5.1 4.6 4.8 4.5 6.1 5.6 5.5 5.6 5.7 5.7 5.4 6.2 6.4
None None 3.0 4.2 – – – – 5.0 – 6.7 – 7.1 6.7 – 6.8 – – 7.0
samples only the SAXS profiles originating from the nanostructures, which are obtained by subtracting the Porod scattering from the acylHA profiles, as described in Section 2.3. and the Supplementary material. The SAXS profiles of a native HA film (HA-2) and a palmitoyl-HA film (HA-C16-6) are depicted in Fig. 2b. Since the films are non-porous bulk materials, they do not exhibit surface scattering (contrary to the powders). The profile of the native HA film did not show any q-dependence, only background signal was observed. This suggests that the native HA film did not contain any nanostructures. On the other hand, the SAXS profile of the palmitoyl-HA film showed a distinct scattering peak, which is clear evidence for the presence of nanostructures. It thus seems that the form of the sample does not have any influence on the presence of nanostructures. The effect of DS on the SAXS profiles of acyl-HA powders (Fig. 3a–c) was studied using lauroyl-HA, palmitoyl-HA and oleyl-HA powders with various DS. The effect of the acyl chain length (Fig. 3d) was studied using hexanoyl, decanoyl, lauroyl, palmitoyl, and oleyl derivatives. Importantly, a distinct scattering peak was observed in all profiles. The scattering intensity increased with increasing DS and increasing acyl chain length. In the range of DS investigated, the position of the scattering peak did not depend on DS for a given acyl chain length. However, with increasing acyl chain length, the scattering peak clearly shifted to lower q values. The effects of DS and acyl chain length on the SAXS profiles of acylHA films are demonstrated in Fig. 4. The influence of DS was studied on a set of palmitoyl-HA films, while the effect of the acyl chain length was studied using hexanoyl, decanoyl, palmitoyl, and oleyl-HA films. As in the case of powders, a distinct scattering peak was observed in all profiles. The scattering intensity increased with increasing DS and increasing acyl chain length. The scattering peak shifted to lower q values with increasing acyl chain length, but its position did not change substantially with DS. As for the powder samples, the SAXS profiles of all acyl-HA films showed a distinct scattering peak. It can be noticed that the scattering peaks of the films are slightly shifted towards lower q with respect to the powders, indicating that the nanostructures in powders might be somewhat smaller. However, it is also possible that scattering on the powder particles could contribute to the observed shift. The second difference between powders and films is that the dependence of the intensity on DS is not monotonous in the case of powders. This is
3. Results The native HA and acyl-HA listed in Table 1 were analysed using Xray scattering. All samples were measured in the form of a powder and, additionally, selected samples were also measured in the form of a film. Note that the scattering vector magnitude range of the used instrument enables to study structures with dimensions ranging approximately from 1 to 25 nm (depending also on the contrast). The SAXS profiles of a native HA powder (HA-1) and a palmitoyl-HA powder (HA-C16-6) are shown in Fig. 2a. The native HA profile could be well fitted by the sum of the Porod’s law and a constant (representing background signal) as follows:
I = k q− 4 + B
Sample
(3)
where I is the scattering intensity, k is the proportionality constant, q is the scattering vector magnitude and B is the background constant. Porod’s law describes the surface (Porod) scattering of compact particles with a sharp interface (e.g., powder particles). Since the SAXS profile of the native HA powder contained only Porod scattering (i.e., scattering by the powder particles themselves), we can suggest that nanostructures (long range order) are not present in the native HA powder. On the other hand, the palmitoyl-HA sample showed an additional scattering peak superimposed on the Porod scattering of the powder (Fig. 2a), indicating the presence of nanostructures. Since we aim to study these nanostructures, we will further present for the powder 470
Carbohydrate Polymers 195 (2018) 468–475
J. Chmelař et al.
Fig. 2. Comparison of native HA (HA-1 powder, HA-2 film) and acyl-HA (HA-C16-6) SAXS profiles for powders (a) and films (b). Since only background signal was observed for native HA, the data are presented without the subtraction of a constant background.
scattering peak. This enabled to calculate the Bragg distance dB from the position of the scattering peak maximum (Eq. (2)). The obtained dB are listed in Table 1. The Bragg distance provides information about the characteristic scale in electron density variations (heterogeneity). However, the Bragg distance depends on the molecular structure of the system in a complex manner and its further direct interpretation is impossible without a priori assuming a structural model (Guinier & Fournet, 1955). Since we do not have sufficient information to develop such a model, the Bragg distance will be used only qualitatively.
because the thickness and void fraction of the irradiated powder layer was not fixed due to the applied experimental technique. Therefore, the intensities cannot be compared among each other. In the high-q region (Porod’s region of nanostructures), the SAXS profiles of both acyl-HA powders (Fig. 3) and films (Fig. 4) showed a q−n decay with n ≈ 4. This decay exponent is characteristic for systems consisting of multiple domains or phases separated by relatively sharp boundaries (Glatter & Kratky, 1982). The SAXS profiles of acyl-HA derivatives showed a distinct
Fig. 3. SAXS profiles of acyl-HA powder samples after the subtraction of Porod scattering: lauroyl-HA (a), palmitoyl-HA (b), and oleyl-HA (c) with various DS; samples with varying acyl chain length (d). 471
Carbohydrate Polymers 195 (2018) 468–475
J. Chmelař et al.
Fig. 4. SAXS profiles of acyl-HA film samples: palmitoyl-HA with various DS (a) and samples with varying acyl chain length (b).
Fig. 5. Dependence of the Bragg distance dB on the length of the acyl side chain.
Fig. 6. Dependence of the Bragg distance dB on the DS for HA-C12 and HA-C16 derivatives measured in the powder form.
The dependence of the Bragg distance dB on the acyl chain length is depicted in Fig. 5. One can clearly see that dB increases with increasing acyl chain length and that this increase is approximately linear. Note that the acyl chain length is expressed as the number of carbon atoms of the given fatty acid. The dB determined for the powders and films did not differ significantly, which is not surprising given the similarity of their SAXS profiles. The dB values obtained for the films were slightly larger than those obtained for the powders, in agreement with the already discussed shift of the acyl-HA film scattering peaks to lower q. The dependence of the Bragg distance dB on the degree of HA substitution is depicted in Fig. 6 for the HA-C12 and HA-C16 powder samples. One can clearly see that dB does not depend on the degree of HA substitution (in the studied range of properties), with average values of 4.8 and 5.7 nm for HA-C12 and HA-C16, respectively. The situation is analogous for the film samples, as demonstrated for HA-C16 films in the Supplementary material (Fig. S2). The used instrument is capable of measuring X-ray scattering also in the WAXS range, which enables to assess crystallinity. The SAXS/WAXS profiles of a native HA film (HA-1) and a palmitoyl-HA film (HA-C16-6) are depicted in Fig. 7. In the WAXS region, only the so-called amorphous halo was present for both native HA and acyl-HA. No crystalline reflections were observed in the investigated q-range (corresponding to a maximum 2θ of 41°). These data clearly show that the studied materials are not crystalline. Based on the SAXS data, two film samples were selected for microscopic observations: native HA (HA-1) as a control without nanostructure and palmitoyl-HA (HA-C16-6) as a sample with nanostructure.
Fig. 7. SAXS/WAXS profiles of a native HA film and a palmitoyl-HA film.
In a TEM microscope, both samples appeared morphologically homogeneous (results not shown). In a XRHSEM microscope after plasma etching, a certain difference between the two samples was found (Fig. 8). The native HA film again appeared morphologically homogeneous (Fig. 8a), while the palmitoyl-HA film displayed some faint
472
Carbohydrate Polymers 195 (2018) 468–475
J. Chmelař et al.
Fig. 8. XHRSEM micrographs showing plasma-etched cut surfaces of native HA (a) and palmitoyl-HA (b) films. The palmitoyl-HA film showed structures that could be enhanced using contrast enhancement (c) in combination with Wiener filter (d) and binarized using local threshold (e). The binary image was transformed to reciprocal space using 2D-DFFT (f) to evaluate periodic distances (Slouf & Vackova, 2014; Slouf et al., 2008).
morphology (Fig. 8b). After contrast adjustment using ImageJ (Schneider, Rasband, & Eliceiri, 2012), the morphological features were somewhat enhanced but the adjusted micrograph suffered from high noise (Fig. 8c). Better results were obtained by combining noise/shift reduction by means of Wiener filter (using Python/SciPy (Oliphant, 2007)) and contrast enhancement like in ImageJ (Fig. 8d). The image was binarized using automated local threshold in ImageJ (Fig. 8e) and transformed to reciprocal space using two-dimensional discrete fast Fourier transform (2D-DFFT; using Python/Scipy (Oliphant, 2007; Slouf & Vackova, 2014)). The final 2D-DFFT image (Fig. 8f) could be used to calculate the periodicity in the sample. The 2D-DFFT image was radially averaged (using Python/SciPy (Oliphant, 2007; Slouf & Vackova, 2014)) to obtain a 1D-DFFT radial profile (Fig. 9). The radial profile corresponds to a calculated 1D diffraction pattern of the observed structures and, consequently, the peaks in the 1D-DFFT profile represent periodic distances, d, which can be evaluated using a few simple formulas analogous to those used in the field of X-ray diffraction (Slouf et al., 2008). The calculation for the investigated palmitoyl-HA film is shown directly in Fig. 9. Although the morphology in XHRSEM micrographs was very weak and the micrographs had to be heavily processed to enhance the signal, it is noteworthy that the calculated periodic distance d ≈ 5.5 nm was in satisfactory agreement with the SAXS results. We can conclude that careful Fourier-transform-based analysis of XHRSEM micrographs showing plasma-etched ultramicrotomed samples confirmed the periodicity observed in the parallel SAXS experiments.
Fig. 9. Radial profile calculated from the Fourier-transformed image of the palmitoyl-HA film (Fig. 8f) showing a strong maximum corresponding to a real space periodic distance d ≈ 5.5 nm. The principle of the periodic distance calculation is given directly in the image, more details can be found elsewhere (Slouf & Vackova, 2014; Slouf et al., 2008).
particles. This confirms that native HA is not a nanostructured material, i.e., it forms one homogeneous phase on the nanoscale. On the other hand, all studied acyl-HA derivatives exhibited distinct scattering peaks with various intensities. Analogous results were obtained by XHRSEM, which showed periodic morphological features for the palmitoyl-HA film, while the native HA film appeared morphologically homogeneous. This clearly shows that the nanostructure (long range order) observed for acyl-HA is related to the presence of the hydrophobic acyl side chains. The SAXS profiles and the evaluated dB showed only a minor
4. Discussion The data presented in the Results revealed several interesting features. Importantly, the SAXS profiles of native HA did not show any scattering peaks or shoulders or Porod scattering, except for the powder sample, where the Porod scattering originates from the powder 473
Carbohydrate Polymers 195 (2018) 468–475
J. Chmelař et al.
5. Conclusions
influence of sample form, despite the significantly different ways of solidification: the powder samples were precipitated, while the films were produced by slow solvent evaporation. Therefore, the observed solid-state nanostructure seems to be inherent to the materials. The WAXS profiles only contained a weak broad halo (Fig. 7) representing short-range order (length scale below 1 nm) that is characteristic for liquids and glasses, showing that the acyl-HA derivatives are amorphous. When these findings are combined, we can conclude that the nanostructure of acyl-HA is related to local organization of amorphous polymer chains that is induced by the presence of the acyl side chains. Considering the chemical structure of acyl-HA, we suggest that the observed solid-state nanostructure corresponds to a distribution of hydrophobic domains in a hydrophilic matrix (microphase-separated structure). The Bragg distance dB obtained from the SAXS profiles could be related to the average distance of the hydrophobic domains. The creation of this structure is driven by unfavorable interactions between the hydrophobic acyl chains and the hydrophilic HA backbone. Such interactions are well known especially in solution, where they can lead to the formation of polymeric micelles (Šmejkalová et al., 2017). In the solid-state, the hydrophobic domains act as physical cross-links and, at sufficient DS and acyl chain length, can provide insolubility or limited solubility to forms such as films (Foglarová et al., 2016) or fibres (Zápotocký et al., 2016). A single hydrophobic domain (as seen by SAXS, i.e., a domain differing in electronic density) should certainly contain the interacting acyl chains. It is not possible to reliably determine whether the modified HA backbone segments are also part of this domain. However, some information can be obtained from the Porod region of the SAXS profile of nanostructures. In this region, the SAXS profiles showed a q−n decay with n ≈ 4, indicating a nanostructure with multiple domains or phases that have a relatively sharp boundary. This means that the hydrophobic domains in acyl-HA are rather well defined. The persistence length of HA was reported to be around 10 nm (Buhler & Boué, 2004; Carn et al., 2012). This value is higher than the Bragg distance, suggesting that HA chains will behave as rigid at the length scale of the nanostructures. Consequently, the HA chains will be expanded and the acyl groups of a single acyl-HA chain will contribute to a larger number of hydrophobic domains. In the literature, helical conformations were suggested for native HA in the solid state (Cowman & Matsuoka, 2005; Haxaire et al., 2000). Our results are not in conflict with these findings, since the backbone chain itself could be in a helical conformation, although some defects caused by the presence of the acyl side chains and their interactions can be expected. It is also interesting to compare dB to the length of the acyl side chain lacyl, which can be calculated from the number of carbon atoms of the given acyl chain NC as (Tanford, 1980): lacyl (nm) = 0.15 + 0.1265 NC
We studied the nanostructure of acyl-HA in the solid state using Xray scattering and electron microscopy. The SAXS profiles of unmodified HA did not show any scattering peaks, while all studied acylHA samples exhibited distinct peaks corresponding to nanometre-sized structures. Analogous results were obtained by XHRSEM. The SAXS data showed only minor influence of sample form (powder or film), so the observed nanostructure is not a result of sample preparation. The size of the nanostructures increased with the acyl chain length, while the degree of substitution and molecular weight had negligible effects. WAXS profiles showed that the samples were non-crystalline. Based on these results, we suggest that the solid-state nanostructure of acyl-HA corresponds to a distribution of hydrophobic domains in a hydrophilic matrix of unmodified HA segments. The hydrophobic domains act as physical cross-links and can provide insolubility or limited solubility to the material similarly to covalent cross-linking. These findings contribute to our understanding of the macroscopically observed behaviour of acyl-HA materials and should improve our ability to design final product properties (e.g., the swelling and degradation of resorbable implants). Declarations of interest None. Acknowledgements On the part of New Technologies - Research Centre, the result was developed within the CENTEM project, reg. no. CZ.1.05/2.1.00/ 03.0088, cofunded by the ERDF as part of the Ministry of Education, Youth and Sports OP RDI programme and, in the follow-up sustainability stage, supported through CENTEM PLUS (LO1402) by financial means from the Ministry of Education, Youth and Sports under the National Sustainability Programme I. Electron microscopy at the Institute of Macromolecular Chemistry (IMC) and at the collaborating Institute of Scientific Instruments (ISI) was supported by project TE01020118 (Technology Agency of the CR); the final XHRSEM micrographs come from ISI and were obtained by Aleš Paták. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.carbpol.2018.04.111. References
(4)
Šmejkalová, D., Hermannová, M., Šuláková, R., Průšová, A., Kučerík, J., & Velebný, V. (2012). Structural and conformational differences of acylated hyaluronan modified in protic and aprotic solvent system. Carbohydrate Polymers, 87(2), 1460–1466. Šmejkalová, D., Nešporová, K., Hermannová, M., Huerta-Angeles, G., Čožíková, D., Vištejnová, L., et al. (2014). Paclitaxel isomerisation in polymeric micelles based on hydrophobized hyaluronic acid. International Journal of Pharmaceutics, 466(1–2), 147–155. Šmejkalová, D., Muthný, T., Nešporová, K., Hermannová, M., Achbergerová, E., HuertaAngeles, G., et al. (2017). Hyaluronan polymeric micelles for topical drug delivery. Carbohydrate Polymers, 156, 86–96. Bobula, T., Běťák, J., Buffa, R., Moravcová, M., Klein, P., Žídek, O., et al. (2015). Solidstate photocrosslinking of hyaluronan microfibres. Carbohydrate Polymers, 125, 153–160. Buhler, E., & Boué, F. (2004). Chain persistence length and structure in hyaluronan solutions: Ionic strength dependence for a model semirigid polyelectrolyte. Macromolecules, 37(4), 1600–1610. Carn, F., Guyot, S., Baron, A., Perez, J., Buhler, E., & Zanchi, D. (2012). Structural properties of colloidal complexes between condensed tannins and polysaccharide hyaluronan. Biomacromolecules, 13(3), 751–759. Chiu, Y. L., Chen, S. C., Su, C. J., Hsiao, C. W., Chen, Y. M., Chen, H. L., et al. (2009). pHtriggered injectable hydrogels prepared from aqueous N-palmitoyl chitosan: In vitro characteristics and in vivo biocompatibility. Biomaterials, 30(28), 4877–4888. Chmelař, J., Kotzianová, A., Hermannová, M., Šuláková, R., Šmejkalová, D., Kulhánek, J., et al. (2017). Evaluating the degree of substitution of water-insoluble acyl derivatives
Since more data were measured for powders, further analysis will be limited to these samples. The dependence of dB on lacyl is linear (see Fig. S3 in the Supplementary material) and can thus be easily fitted, yielding: dB (nm) = 2.07 lacyl + 1.23
(5)
The slope is close to 2, suggesting that dB is somehow related to twice the length of the acyl chain. The intersection of 1.23 is relatively close to the value of 1 nm, which was reported as the length of the HA repeating unit (Finelli et al., 2014). Unfortunately, it is not possible to directly relate these observations to the nanostructure based on the measured data, but they might be valuable input into further studies utilizing, for example, molecular simulations. In general, simulations or thermodynamic calculations (Lundy et al., 2017) should be the next step in elucidating the nanostructure of acyl-HA materials.
474
Carbohydrate Polymers 195 (2018) 468–475
J. Chmelař et al.
behavior of hyaluronan in relation to its physical properties as probed by molecular modeling. Glycobiology, 10(6), 587–594. Huerta-Angeles, G., Bobek, M., Příkopová, E., Šmejkalová, D., & Velebný, V. (2014). Novel synthetic method for the preparation of amphiphilic hyaluronan by means of aliphatic aromatic anhydrides. Carbohydrate Polymers, 111, 883–891. Hussain, I. S. A., & Jaisankar, V. (2017). An eco-friendly synthesis: Characterization and antibacterial applications of novel almond gum −poly(acrylamide) based hydrogel silver nanocomposite. Polymer Testing, 62, 154–161. Lalevée, G., Sudre, G., Montembault, A., Meadows, J., Malaise, S., Crépet, A., et al. (2016). Polyelectrolyte complexes via desalting mixtures of hyaluronic acid and chitosan—Physicochemical study and structural analysis. Carbohydrate Polymers, 154, 86–95. Lundy, R., Flynn, S. P., Cummins, C., Kelleher, S. M., Collins, M. N., Dalton, E., et al. (2017). Controlled solvent vapor annealing of a high χ block copolymer thin film. Physical Chemistry Chemical Physics, 19, 2805–2815. Maki, Y., Furusawa, K., Dobashi, T., Sugimoto, Y., & Wakabayashi, K. (2017). Small-angle X-ray and light scattering analysis of multi-layered Curdlan gels prepared by a diffusion method. Carbohydrate Polymers, 155, 136–145. Matteini, P., Dei, L., Carretti, E., Volpi, N., Goti, A., & Pini, R. (2009). Structural behavior of highly concentrated hyaluronan. Biomacromolecules, 10(6), 1516–1522. Moffat, J., Morris, V. J., Al-Assaf, S., & Gunning, A. P. (2016). Visualisation of xanthan conformation by atomic force microscopy. Carbohydrate Polymers, 148, 380–389. Oliphant, T. E. (2007). Python for scientific computing. Computing in Science & Engineering, 9, 10–20. Podzimek, S., Hermannova, M., Bilerova, H., Bezakova, Z., & Velebny, V. (2010). Solution properties of hyaluronic acid and comparison of SEC-MALS-VIS datawith off-line capillary viscometry. Journal of Applied Polymer Science, 116(5), 3013–3020. Schneider, C. A., Rasband, W. S., & Eliceiri, K. W. (2012). NIH Image to ImageJ: 25 years of image analysis. Nature Methods, 9, 671–675. Slouf, M., & Vackova, T. (2014). Application of fourier transformation in image analysis of micrographs. Materials Structure, 22, 91–94. Slouf, M., Synkova, H., Baldrian, J., Marek, A., Kovarova, J., Schmidt, P., et al. (2008). Structural changes of UHMWPE after e-Beam irradiation and thermal treatment. Journal of Biomedical Materials Research Part B: Applied Biomaterials, 85B, 240–251. Svergun, D. I., & Koch, M. H. J. (2003). Small-angle scattering studies of biological macromolecules in solution. Reports on Progress in Physics, 66, 1735–1782. Tanford, C. (1980). The hydrophobic effect. New York: John Wiley & Sons. Tang, Z., He, C., Tian, H., Ding, J., Hsiao, B. S., Chu, B., et al. (2016). Polymeric nanostructured materials for biomedical applications. Progress in Polymer Science, 60, 86–128. Vercruysse, K. P., Marecak, D. M., Marecek, J. F., & Prestwich, G. D. (1997). Synthesis and in vitro degradation of new polyvalent hydrazide cross-linked hydrogels of hyaluronic acid. Bioconjugate Chemistry, 8(5), 686–694. Zápotocký, V., Pospíšilová, M., Janouchová, K., Švadlák, D., Baťová, J., Šogorková, J., et al. (2016). Fabrication of biodegradable textile scaffold based on hydrophobized hyaluronic acid. International Journal of Biological Macromolecules, 95, 903–909.
of hyaluronan using Raman spectroscopy: Method development and comparison with gas chromatography and 1H NMR. Analytical Methods, 9, 232–239. Choi, K. Y., Chung, H., Min, K. H., Yoon, H. Y., Kim, K., Park, J. H., et al. (2010). Selfassembled hyaluronic acid nanoparticles for active tumor targeting. Biomaterials, 31(1), 106–114. Choudhary, S., & Bhatia, S. R. (2012). Rheology and nanostructure of hydrophobically modified alginate (HMA) gels and solutions. Carbohydrate Polymers, 87(1), 524–530. Chu, B., & Hsiao, B. S. (2001). Small-angle X-ray scattering of polymers. Chemical Reviews, 101(6), 1727–1761. Collins, M. N., & Birkinshaw, C. (2007). Comparison of the effectiveness of four different crosslinking agents with hyaluronic acid hydrogel films for tissue-culture applications. Journal of Applied Polymer Science, 104(5), 3183–3191. Collins, M. N., & Birkinshaw, C. (2008). Investigation of the swelling behavior of crosslinked hyaluronic acid films and hydrogels produced using homogeneous reactions. Journal of Applied Polymer Science, 109(2), 923–931. Collins, M. N., & Birkinshaw, C. (2013). Hyaluronic acid solutions—A processing method for efficient chemical modification. Journal of Applied Polymer Science, 130(1), 145–152. Cowman, M. K., & Matsuoka, S. (2005). Experimental approaches to hyaluronan structure. Carbohydrate Research, 340(5), 791–809. Creuzet, C., Kadi, S., Rinaudo, M., & Auzély-Velty, R. (2006). New associative systems based on alkylated hyaluronic acid. Synthesis and aqueous solution properties. Polymer, 47(8), 2706–2713. Dicker, K. T., Gurski, L. A., Pradhan-Bhatt, S., Witt, R. L., Farach-Carson, M. C., & Jia, X. (2014). Hyaluronan: A simple polysaccharide with diverse biological functions. Acta Biomaterialia, 10(4), 1558–1570. Dvorakova, J., Kucera, L., Kucera, J., Svik, K., Foglarova, M., Muthny, T., et al. (2014). Chondrogenic differentiation of mesenchymal stem cells in a hydrogel system based on an enzymatically crosslinked tyramine derivative of hyaluronan. Journal of Biomedical Materials Research Part A, 102(10), 3523–3530. Eenschooten, C., Vaccaro, A., Delie, F., Guillaumie, F., Tømmeraas, K., Kontogeorgis, G. M., et al. (2012). Novel self-associative and multiphasic nanostructured soft carriers based on amphiphilic hyaluronic acid derivatives. Carbohydrate Polymers, 87(1), 444–451. Finelli, I., Chiessi, E., Oddo, L., Galesso, D., Renier, D., & Paradossi, G. (2014). Collective dynamics and transient behavior of partially hydrophobic hyaluronic acid chains. Macromolecular Chemistry and Physics, 215(2), 140–147. Foglarová, M., Chmelař, J., Huerta-Angeles, G., Vágnerová, H., Kulhánek, J., Tománková, K. B., et al. (2016). Water-insoluble thin films from palmitoyl hyaluronan with tunable properties. Carbohydrate Polymers, 144, 68–75. Gamini, A., Paoletti, S., Toffanin, R., Micali, F., Michielin, L., & Bevilacqua, C. (2002). Structural investigations of cross-linked hyaluronan. Biomaterials, 23(4), 1161–1167. Garg, H., & Hales, C. (2004). Chemistry and biology of hyaluronan. Oxford: Elsevier. Glatter, O., & Kratky, O. (1982). Small-angle X-ray scattering. London: Academic Press Inc. Guinier, A., & Fournet, G. (1955). Small angle scattering of X-rays. New York: Wiley. Haxaire, K., Braccini, I., Milas, M., Rinaudo, M., & Perez, S. (2000). Conformational
475