Biochemical profiling of rat embryonic stem cells grown on electrospun polyester fibers using synchrotron infrared microspectroscopy
Therapeutic options for spinal cord injuries are severely limited; current treatments only offer symptomatic relief and rehabilitation focused on educating the individual on how to adapt to their new situation to make best possible use of their remaining function. Thus, new approaches are needed, and interest in the development of effective strategies to promote the repair of neural tracts in the central nervous system inspired us to prepare functional and highly anisotropic polymer scaffolds. In this work, an initial assessment of the behavior of rat neural progenitor cells (NPCs) seeded on poly(3-hydroxybutyrate-co-3-hydroxyhexanoate) fiber scaffolds using synchrotron-based infrared microspectroscopy (SIRMS) is described. Combined with a modified touch imprint cytology sample preparation method, this application of SIRMS enabled the biochemical profiles of NPCs on the coated polymer fibers to be determined. The results showed that changes in the lipid and amide I–II spectral regions are modulated by the type and coating of the substrate used and the culture time. SIRMS studies can provide valuable insight into the early-stage response of NPCs to the morphology and surface chemistry of a biomaterial, and could therefore be a useful tool in the preparation and optimization of cellular scaffolds.
KeywordsPoly(3-hydroxybutyrate-co-3-hydroxyhexanoate) Electrospinning Neural progenitor cells FTIR spectroscopy
The intrinsic characteristics of the central nervous system (CNS) are a major impediment to its spontaneous recovery in response to injury and, as a consequence, lesions cause permanent functional deficits that depend on their location and extent. In fact, complete functional repair of a spinal cord injury (SCI) cannot generally be achieved without precise and significant aid . Thus, SCI is a global problem that not only affects the physical and psychological well-being of patients and their families, but also places an enormous burden on the economic resources of developed countries and increases the mortality rates in developing nations . To give an example, the incidence of traumatic SCI in Western Europe was recently reported to be between 218 and 316 cases (around half of which are due to traffic accidents) per million habitants, whereas for North America (the US and Canada), the mean is over three times that . In Canada alone, the annual cost of SCI in 2012 was reported to be over 2 billion €, at least 32% of which was ascribed to attendant care. In 2017, the National Spinal Cord Injury Statistical Center in the USA estimated that, depending on the severity of the SCI, the average lifetime cost of treatment and care for a person who suffers a SCI at the age of 25 was 1.4–4.3 M€ . Hence, there is a great deal of interest in improving this situation.
The multifactorial nature of this type of injury severely limits therapeutic options, and the regeneration of injured axons demands a timely, structured strategy that is able to present multiple signals in a favorable environment [5, 6]. Among the diverse options that are being explored to repair neural tissue are a variety of tissue engineering approaches that are designed to facilitate the regeneration of axons using anisotropic scaffolds with biochemical cues [7, 8, 9]. These include fibers of natural and synthetic materials that mimic the spinal cord extracellular matrix and could be combined with growth factors and different types of exogenous uncommitted cells  or homing factors to stimulate endogenous stem cell migration towards the injury site .
The development of fibrous substrates resembling the extracellular matrix to support cell adhesion, differentiation, and proliferation has drawn great interest from researchers involved in tissue engineering [12, 13, 14] since the renaissance of electrospinning , a technique that allows the preparation of nonwoven fibers with diameters ranging from tens of nanometers to a few microns. While the tissue engineering of electrospun scaffolds has shown some limitations (i.e., it is difficult to fabricate complex three-dimensional structures and cell infiltration is poor due to small pore sizes) , these substrates offer the possibility of exploring cell behavior on exposure to a variety of chemical and physical cues in a reproducible manner, and they facilitate the translation of new and innovative strategies to in vivo preclinical research.
Scaffolds based on PHB [poly(3-hydroxybutyrate)] and its copolymers with other β-hydroxy acids have been employed in experimental models of CNS lesions [17, 18]. Such scaffolds offer several advantages in relation to their intended applications: they possess a reproducible and well-defined polymer microstructure, amenable to processing by a wide variety of methods, and they are both biodegradable and biocompatible. However, as is often the case for many synthetic polymers, the intrinsic surface properties of these materials do not facilitate interaction between the scaffold and the cells. Thus, the functionality of the scaffolds must be modified by the physical adsorption or covalent binding of specific chemical moieties and macromolecular fragments to its surface.
An interest in developing effective strategies to experimentally promote the repair of neural tracts in the CNS led us to prepare functional and highly anisotropic scaffolds of poly(3-hydroxybutyrate-co-3-hydroxyhexanoate) [P(HB-co-HHx)] via electrospinning. Since the physical and chemical characteristics of polymer scaffolds play an important role in the modeling of cellular behavior during growth and proliferation [19, 20, 21], the response of neural precursor cells (NPCs) to stimuli induced by P(HB-co-HHx) scaffolds must be characterized in order to guide and optimize approaches to SCI functional repair. In particular, studying the biochemical profile of cells and their local environment using vibrational microspectroscopy could highlight distinctive spectral markers associated with a specific cell response.
Fourier-transform infrared (FTIR) spectroscopy and (particularly) high-resolution synchrotron-based infrared microspectroscopy (SIRMS) have been successfully employed to characterize many biological tissues [22, 23, 24, 25, 26] and to explore several aspects of cell biology, such as cell differentiation [27, 28, 29, 30, 31], changes in cell physiological state [32, 33], and the effects of exogenous agents on cell biochemical profiles [34, 35]. The brightness of the synchrotron IR source provides diffraction-limited spatial resolution, allowing high-quality spectra to be obtained through small apertures that can be closely matched to the size of the cells or features of interest in the sample. The study described in the present paper applied SIRMS to search for spectral markers that allow the roles of the scaffold morphology and surface modification in the tuning and control of neural progenitor cell response during cell differentiation and subsequent proliferation to be evaluated. Spectra were obtained from NPCs cultured for up to 48 h on electrospun P(HB-co-HHx) scaffolds impregnated with poly-L-lysine (PLL) and laminin (L). Poly-L-lysine is a polypeptide commonly used in CNS-originated cell cultures to facilitate cell adhesion to culture plates, and laminin is a protein present in the extracellular matrix of the CNS. Samples obtained at three time points were analyzed to assess the evolution of the cell profile. We monitored variations in the extracellular matrix composition using IR spectroscopic biomarkers, particularly those associated with lipid and protein contents, during the adhesion, differentiation, and proliferation of NPCs on the scaffolds . The spectroscopic findings were correlated with the results of immunochemical studies of cell morphology and membrane markers of differentiating and/or differentiated cells in culture.
Materials and methods
Poly(3-hydroxybutyrate-co-3-hydroxyhexanoate) [P(HB-co-HHx)] with a molar composition of 14.3% in 3-hydroxyhexanoate, molecular weight Mw = 1.9 × 105 g mol−1 and polydispersity index Mw/Mn = 1.47, was supplied by Professor Guo-Qiang Chen (Tsinghua University, China). Prior to use, the copolymer was washed with ethanol under constant stirring overnight, filtered, and vacuum dried to a constant weight. Dichloromethane supplied by Carlo Erba Réactifs-SdS (Val de Reuil, France) and methanol, ethanol, and 1,1,1,3,3,3-hexafluoro-2-propanol purchased from Sigma–Aldrich (Steinheim, Germany) were used as supplied.
Neurobasal medium and B27 supplement were purchased from GIBCO (Paisley, UK); human bFGF and EGF were from Peprotech (Rocky Hill, NJ, USA); L-glutamine, L-glutamate, penicillin, streptomycin, bisBenzimide H 33258, and a fragment of laminin containing the IKVAV sequence were from Sigma–Aldrich; and fungizone was from Invitrogen (Madrid, Spain). Rabbit polyclonal anti-GFAP was provided by Acris (Herford, Germany) and monoclonal antibody for nestin was from Santa Cruz Biotechnology (Dallas, TX, USA); antibody goat anti-rabbit IgG conjugated to Alexa Fluor 488 and antibody goat anti-rabbit IgG conjugated to Alexa Fluor 594 were provided by Molecular Probes (Eugene, OR, USA).
Preparation of P(HB-co-HHx) fiber scaffolds
The fiber scaffolds were prepared by electrospinning P(HB-co-HHx) solutions in a mixture of solvents. Briefly, 1 g of copolymer was dissolved in 3.5 mL of a mixture of dichloromethane (3.3 mL) and 1,1,1,3,3,3-hexafluoro-2-propanol (0.2 mL) at room temperature for 12 h with constant stirring.
The copolymer solutions were immediately electrospun in a home-built apparatus consisting of a high-voltage power supply (30 kV 600 W, SL Series, Spellman, Hauppauge, NY, USA), a blunt tip needle (0.584 mm i.d.) connected to the positive pole of the power supply, and a grounded 7 cm diameter rotatable drum as collector. In addition, the flow (0.2 mL/h) of the solution through the needle and into the electric field was controlled with an infusion pump (100 Series, KD Scientific, Holliston, MA, USA). Usually, the distance between the needle tip and the rotating drum was 16 cm, the rotation speed was ~1200 rpm, and the applied voltage was 11.5 kV. The temperature in the chamber during electrospinning was maintained at around 22 °C and the relative humidity varied between 20 and 24%. Fibers were collected for 3.5 h on aluminum foil affixed to the rotating drum to obtain mechanically stable scaffolds with a mean thickness of approximately 35 μm. After vacuum drying them overnight, the deposited fibers were cut into disks (∅ = 14 mm) and stored in a dry environment at 4 °C until used.
The experiments followed the European Parliament and Council Directive (2013/63/EU) and the Spanish regulation (RD 53/2013) on the protection of animals for experimental use. This study was approved by our institutional animal use and care committee for animal welfare (Hospital Nacional de Parapléjicos, registered as SAPA001). Wistar rat embryos (E15) were obtained from previously anesthetized pregnant rats by cesarean section. The rats were bred and maintained at the animal house of the Hospital Nacional de Parapléjicos in Toledo.
Striata from E15 rats were dissected and mechanically dissociated into individual cells . This cell suspension was incubated in neurobasal medium and B27 supplement containing human bFGF (10 ng/mL) and EGF (20 ng/mL) as well as L-glutamine (0.5 mM), L-glutamate (25 mM), penicillin (100 U/mL), streptomycin (0.1 mg/mL), and fungizone (2.5 lg/mL). After about 7 days in this “NB27” medium, floating neurospheres were formed and collected by low-speed centrifugation, and were washed free of glutamate by suspension in PBS and further centrifugation. Neurospheres were dissociated by mild trypsinization, passed through a 25-gauge needle, and expanded every 3–4 days.
Neural precursor cell differentiation
After four expansion passages, NPCs grown as neurospheres for cell expansion were suspended in a 1:1 mixture of NB27 and DMEM incorporating 10% bovine serum. The neurosphere suspension was plated onto a 24-well cell culture plate that contained 14 mm ∅ disks of P(HB-co-HHx) fiber substrates treated with 50 μg/mL poly-L-lysine and/or 20 μg/mL of a laminin fragment containing the IKVAV sequence. ZnS IR windows coated with laminin were used as a control. The cells were incubated for 3 h to allow cell attachment and 24 and 48 h for neurosphere differentiation. After treatment, the cells were fixed for immunocytochemistry and characterized by SR-FTIR microspectroscopy.
The treated cells on PLL- and/or laminin-coated substrates were fixed in 2% paraformaldehyde/sucrose in PBS (12 min, 25 °C), washed with PBS, and immunostained as follows. They were incubated for 30 min, at 25 °C, in PBS containing 1% normal goat serum with 0.1% Triton X100. The cells were then incubated (16 h at 4 °C) in the same mixture containing the primary antibody. After repeated washing with PBS, the cells on the substrates were treated with the secondary antibody (45 min, 25 °C) H 33258, 10 g/mL for 10 min at 25 °C, washed three times with PBS, mounted on a new 24-well plate with glycerol/PBS (1/1), and examined on a fluorescence microscope (DMI 6000B, Leica, Wetzlar, Germany). Mouse monoclonal anti-nestin (2Q178, Santa Cruz Biotechnology) and rabbit polyclonal anti-GFAP at 1/500 dilution were used for immunocytochemistry and revealed by secondary antibodies: goat anti-rabbit IgG conjugated to Alexa Fluor 594 or goat anti-mouse IgG conjugated to Alexa Fluor 488, respectively, at a dilution of 1/1000.
Scanning electron microscopy
The polymer scaffolds were analyzed prior to cell culture using scanning electron microscopy (SEM). Briefly, the scaffolds were coated with approx. 5 nm Au/Pd and observed on a Philips (Eindhoven, Netherlands) XL30 scanning electron microscope at ambient temperature using the parameters indicated in each micrograph.
Preparation of imprints for FTIR microspectroscopy measurements
To acquire IR spectra of cell cultures on polymer fibers, imprints of the substrates with cells were prepared . Briefly, stored fixed samples were moistened with two drops of distilled water, placed between two 1-mm-thick ZnS IR windows, and lightly compressed for 30 min to allow the transfer of cellular material from the polymer scaffold onto the window in contact with the cell-seeded surface. After this time, the press was carefully opened and the window with the imprint was exposed to air and dried overnight at room temperature (see Fig. S1 in the “Electronic supplementary material,” ESM, for the methodology).
Cell cultures on ZnS windows were observed as prepared; no imprints were made.
ZnS windows were chosen for imprinting purposes due to their wide transparency domain in the mid-infrared and their relatively good mechanical properties, and because they are biocompatible, allowing cell growth (they show a cell viability of >90%, and cell morphologies are similar to those obtained in standard polystyrene culture flasks ).
Synchrotron radiation FTIR microspectroscopy (SIRMS)
Spectra of imprints and cell clusters were recorded at the biological endstation of the SMIS beamline at Synchrotron SOLEIL, employing both bending and edge radiation from a bending magnet at a constant current (top-up mode) of around 400 mA. The spectra were obtained on a Continuum XL (Nicolet (Thermo Scientific), Waltham, MA, USA) microscope equipped with a liquid-nitrogen-cooled MCT/A detector, a 32×/NA0.65 Schwarzschild objective, a motorized knife-edge aperture, and an xyz motorized stage, which was coupled to a Nicolet 5700 FTIR spectrometer (Thermo Fisher Scientific, Villebon-sur-Yvette, France). The microscope was operated in dual aperture mode with a 15 × 15 μm2 spatial aperture. A spectral resolution of 4 cm−1 was achieved and 128 scans were accumulated at each data point to obtain a high signal to noise ratio.
The acquired spectra were first visualized using OMNIC 9.2 (Thermo Scientific, Madison, WI, USA) in order to predefine the spectral regions of interest and to eliminate spectra exhibiting strong Mie scattering effects.
The spectra of cells and imprints were analyzed by multivariate pattern recognition techniques using The Unscrambler X, version 10.3 (Camo Software AS, Oslo, Norway). Two spectral regions, 3050–2800 cm−1 and 1800–1400 cm−1, corresponding mainly to lipid and amide I–II bands, respectively, were selected. These regions were preprocessed before analysis [third-degree polynomial, 7/7 point smoothing prior to calculation of the second derivative (third-degree polynomial, 7/7 points) and unit vector normalization]. Principal component analysis (PCA) was carried out using preprocessed, mean-centered data on the two spectral regions combined, as well as on the amide I–II bands. Four to seven principal components (PCs) were calculated using the singular-value decomposition (SVD) algorithm and leverage correction. 2D score plots and 1D loading plots were examined, respectively, to identify any spectral clustering and to obtain information on the spectral bands leading to those clusters. Unsupervised cluster analysis was performed by considering the k-means and Euclidean distance.
Neural precursor cell differentiation: immunostaining
Characteristic FTIR bands observed from neural progenitor cells during adhesion and differentiation on P(HB-co-HHx) fibers and ZnS IR windows
Band max., 2nd derivative, cm−1
PC-1, 3 h vs 24 h, cm−1 (loading)
PC-1, 24 h vs 48 h, cm−1 (loading)
νas of methylene groups, mainly in lipids
νs of methylene groups, mainly in lipids
νs of carbonyl ester lipids
β-Turn (secondary structure of protein) and antiparallel β-sheets
α-Helical protein structure
Unordered protein structures
β-Sheet (secondary structure of protein)
Overall protein absorbance
C=C stretching in aromatic amino acids and β-turn (secondary structure of protein)
A detailed review of the spectral profiles from cells on each type of substrate (P(HB-co-HHx)/PLL/L, P(HB-co-HHx)/L, and ZnS/L) at given culture times showed only minor fluctuations in the lipid band intensities (data not shown).
When IR spectra were grouped according to culture time, in the region between 1700 and 1600 cm−1 (amide I band, Fig. 4b), a decrease in the intensity of the band attributed to α-helix structure (1657 cm−1) was observed at 48 h . A shift of more than 15 cm−1 in the band associated with β-sheet structure , from 1638 to 1622 cm−1, and an increase in the band intensity at 1693 cm−1 were also observed; the latter is generally assigned to β-turns and antiparallel β-sheet structures . Also, these changes appear to be correlated with those observed for the amide II  peak at 1516 cm−1.
An assessment of the amide I band for cells on each type of substrate at the studied culture times showed that the presence of α-helix structures is favored on ZnS substrates, while β-sheet structures are more likely to be formed on P(HB-co-HHx)/L fibers (see Fig. S3a and b in the ESM). The FTIR cell profiles on P(HB-co-HHx)/PLL/L initially exhibit an intermediate behavior, whereas all cells grown on the polymer fibers show similar spectra at 48 h (see Fig. S3c in the ESM).
The FTIR spectrum of NPCs on P(HB-co-HHx) fibers showed a marked decrease in lipid band intensity at the longest culture time studied, 48 h. This observation is consistent with previous observations which showed a reduction in lipid content that was associated with a loss of stem cell pluri- or multipotency . However, there is still no clear consensus about this, as the opposite behavior—an increase in lipid production with cell differentiation—has also been reported . Several outcomes may be anticipated, depending on the differentiation path of the NPCs. NPC differentiation to an oligodendrocyte-like cell could increase the intensity of the lipid bands due to an increase in myelin. On the other hand, if NPCs differentiate to neuron or astrocyte-like cells, a reduction in the lipid signal intensity might be observed. In our case, previous results have shown that on standard culture plates with the culture medium used in this study, neurospheres differentiate to an intermediate stage, maintaining their differentiation potential to some extent .
The changes observed in the amide I band, corresponding to an increase in the amount of β-sheets and β-turns present, support the above statement. The immunochemistry results show that the GFAP signal increased progressively with increasing culture time, particularly on ZnS/L IR windows and P(HB-co-HHx)/PLL/L fibers. The presence of the protein nestin, a biomarker for NPCs, was more prominent during the initial stages of cell culture.
An increase in the heterogeneity of the biochemical profiles of cells with increasing culture time can be anticipated because, for the electrospun polyester fiber substrates, the likelihood of finding cells at different stages of differentiation would increase over time. This would lead to a broadening of the spectral cluster determined by PCA. Although the outcome of this report is influenced by the limited set of variables studied and the duration of cell culture, it is worth pointing out that the cells on fibers coated only with laminin exhibited the greatest overlap between the three distinct populations identified by FTIR microspectroscopy. On the other hand, the NPCs on ZnS IR/L windows showed the narrowest distributions. Thus, although the culture time is a major influence on the cell IR biochemical profile, these observations suggest that the type of substrate/coating present affects the cell differentiation pathway, and could be adjusted to modulate the cellular response appropriately. In this regard, given the cell profile heterogeneity and the short exposure times to the biomaterials, the observation that spectral clustering was influenced by the type and coating of the cell scaffolds used is remarkable. These findings are in agreement with results from other studies showing that the surface chemistry and topography substantially influence the outcome of NPC differentiation , and highlight the potential of synchrotron FTIR microspectroscopy for studying the interactions between cells and biomaterials. Nevertheless, further experiments incorporating different surface chemistries and extended culture times are required to attain a more robust response.
Neural progenitor cells were cultured on electrospun P(HB-co-HHx) fiber substrates coated with either laminin or poly-L-lysine/laminin. At different times after cell seeding, the samples were fixed and the cellular material on the fibers was partly transferred onto IR windows using a modified touch imprint cytology method. This strategy proved to be very effective and enabled the biochemical profiles of the NPCs that evolved on the coated polymer fibers to be determined with SIRMS. The intensities of bands in the lipid and amide I–II regions were observed to be influenced by the substrate type and coating and the culture time. These results demonstrate the important insights that SIRMS studies can provide into the responses of NPCs to biomaterials (in this case to the morphology and surface chemistry of polymeric scaffolds), suggesting that such studies could be a fundamental tool in the preparation and optimization of cellular scaffolds for CNS tissue engineering and regenerative medicine.
Financial support was provided by the Ministerio de Economía y Competitividad (MINECO), projects FIS/ISCIII PI11/01436 and PI11/00592, MAT2015-65184-C2-2-R, and the CSIC. The research leading to these results was funded by the European Community’s Seventh Framework Programme (FP7/2007-2013) under grant agreement no. 312284 (CALIPSO). Synchrotron measurements were performed at the SMIS beamline of Synchrotron SOLEIL. PS is grateful to MINECO for Ramón y Cajal postdoctoral fellowships (RYC-2014-16759). We would like to thank all the beamline staff for their helpful support and advice.
Compliance with ethical standards
Conflicts of interest
The authors declare that they have no competing interests.
- 1.Schwab JM, Brechtel K, Mueller CA, Failli V, Kaps HP, Tuli SK, et al. Experimental strategies to promote spinal cord regeneration—an integrative perspective. Prog Neurobiol. 2006;78:91–116.Google Scholar
- 2.Singh A, Tetreault L, Kalsi-Ryan S, Nouri A, Fehlings MG. Global prevalence and incidence of traumatic spinal cord injury. Clin Epidem. 2014;6:309–31.Google Scholar
- 3.Lee BB, Cripps RA, Fitzharris M, Wing PC. The global map for traumatic spinal cord injury epidemiology: update 2011, global incidence rate. Spinal Cord. 2014;52:110–6.Google Scholar
- 4.NSCISC. Spinal cord injury: facts and figures at a glance. Birmingham, AL: National Spinal Cord Injury Statistical Center; 2017. https://www.nscisc.uab.edu/public_pages/FactsFiguresArchives/SCI%20Facts%20and%20Figures%20at%20a%20Glance%202017.pdf. Accessed 12/02/2018.
- 10.Prestwich GD, Erickson IE, Zarembinski I, West M, Tew WP. The translational imperative: making cell therapy simple and effective. Acta Biomater. 2012;8:4200–7.Google Scholar
- 33.Whelan DR, Bambery KR, Heraud P, Tobin MJ, Diem M, McNaughton D, et al. Monitoring the reversible B to A-like transition of DNA in eukaryotic cell using Fourier transform infrared spectroscopy. Nucl Acids Res. 2011;39:5439–48.Google Scholar
- 34.Yousef I, Bréard J, SidAhmed-Adrar N, Maâmer-Azzabi A, Marchal C, Dumas P, Le Naour F. Infrared spectral signatures of CDCP1-induced effects in colon carcinoma cells. Analyst. 2011;136:5162–8.Google Scholar
- 35.Gianoncelli A, Vaccari L, Kourousias G, Cassese D, Bedolla DE, Kenig S, et al. Soft X-ray microscopy radiation damage on fixed cells investigated with synchrotron radiation FTIR microscopy. Sci Rep. 2015;5:10250. https://doi.org/10.1038/srep10250.
- 43.Murphy BM, D’Antonio J, Manning MC, Al-Azzam W. Use of the amide II infrared band of proteins for secondary structure determination and comparability of higher order structure. Curr Pharm Biotechnol. 2014;15:880–9.Google Scholar
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.