- Email: [email protected]
Evaluation of cell-surface interaction using a 3D spheroid cell culture model on artificial extracellular matrices
Materials Science and Engineering C 73 (2017) 310–318
Contents lists available at ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Evaluation of cell-surface interaction using a 3D spheroid cell culture model on artificial extracellular matrices Wolfgang Metzger a,⁎, Sandra Rother b, Tim Pohlemann a, Stephanie Möller c, Matthias Schnabelrauch c, Vera Hintze b, Dieter Scharnweber b a b c
Department of Trauma, Hand and Reconstructive Surgery, Saarland University, Building 57, 66421 Homburg, Germany Institute of Materials Science, Max Bergmann Center of Biomaterials, TU Dresden, Budapester Straße 27, 01069 Dresden, Germany Biomaterials Department, INNOVENT e. V., Prüssingstraße 27 B, 07745 Jena, Germany
a r t i c l e
i n f o
Article history: Received 13 July 2016 Received in revised form 25 October 2016 Accepted 17 December 2016 Available online 20 December 2016 Keywords: Spheroid Cell-surface interaction Sulfated hyaluronan Glycosaminoglycans
a b s t r a c t Since decades, cell-surface interactions are studied in 2D cell culture approaches, but cells organized in 3D (spheroids) reflect the normal situation of cells in tissues much better due to intense cell-cell-contacts. Accordingly, this study aimed to prove, if spheroids could be used to study cell-surface interaction. Spheroids consisting of fibroblasts and/or osteoblasts were seeded on artificial extracellular matrices. Here, nonsulfated hyaluronan as a biological relevant compound of the extracellular matrix was chemically sulfated to different extents and co-fibrillised with collagen. The changes of the spheroid diameters and the migration distance of outgrown cells after seeding on the matrices were used as parameters to evaluate cell-surface interaction quantitatively. Fibroblast-based spheroids reacted in the initial phase of adhesion with different spheroid sizes on the contact with the matrices. In contrast, the reaction of osteoblasts was more pronounced at later time points exhibiting a decrease of the size of the spheroids with increasing sulfation degree of the matrix. The migration of the cells was impaired by increasing sulfation degree, which might be caused by an increased expression of focal adhesion relevant proteins. In summary, spheroids can be used in cell-surface interaction studies and additional analytical tools could be implemented. © 2016 Elsevier B.V. All rights reserved.
1. Introduction The interaction of cells with an artificial surface of different chemical composition or different topographical features is the basis for the medical outcome of the implantation of any biomaterial. Accordingly, cellsurface interactions are relevant for a wide range of materials and applications e.g. osseointegration of orthopedic implants [1], restenosis of vascular stents [2] or development of novel scaffold materials for tissue engineering approaches [3] to name a few which have been studied with traditional 2D cell culture approaches in the past. Spheroids are spherical 3D cell aggregates. The cells adhere to each other in a self-assembly process without the need for using any stabilizing scaffold [4]. Depending on the cell type and the chosen experimental setup, a variety of methods can be used to generate spheroids from single cell suspensions, e.g. spinner flasks techniques [5], gyratory rotation systems [6], the carboxymethyl cellulose technique [7], hanging drop [8] or liquid overlay technique (LOT) [9]. In addition, novel single-use products are available on the market facilitating the high throughput ⁎ Corresponding author. E-mail address: [email protected] (W. Metzger).
http://dx.doi.org/10.1016/j.msec.2016.12.087 0928-4931/© 2016 Elsevier B.V. All rights reserved.
generation of spheroids [10]. In contrast to traditional 2D cell culture approaches, the cells do not adhere to an artificial surface e.g. polystyrene or glass, but they adhere to each other. Due to the intensive cell-cell contacts in three dimensions, the cells can interact with each other by paracrine mechanisms and by direct cell-cell contacts [11,12]. Furthermore, the cells can interact with proteins of the extracellular matrix (ECM) secreted by the spheroid forming cells their selves [13]. Hence, spheroids are able to reflect the physiological situation of cells in a tissue much better compared to conventional 2D cell cultures [14–16]. It is well known from literature, that a direct comparison of cells in 2D cultures with 3D cultures reveals clear differences in terms of cellular differentiation [17], gene expression profiles or secretion of growth factors [18]. Furthermore, cells organized in spheroids exhibited an increased resistance against hypoxia [18], a lower sensitivity against therapeutics [15], alterations in signaling pathways [19] and a decreased rate in apoptotic cell death [18]. Traditionally, spheroids were used in tumor research because of their more physiological sensitivity against cytotherapeutics [20]. However, during the last years, spheroids were also used to study angiogenesis, both in vivo [21] and in vitro [7], and in tissue engineering approaches where they might serve as building blocks [22]. This study was conducted to test the hypothesis, that
W. Metzger et al. / Materials Science and Engineering C 73 (2017) 310–318
spheroids might also be a useful tool to study the interaction of cells with artificial surfaces, e.g. artificial ECMs (aECMs). The native ECM synthesized by the local cells of a tissue plays a crucial role as 3D microenvironment for a variety of biological process like cell adhesion, proliferation, migration and differentiation [23]. Glycosaminoglycans (GAGs) are functional ECM components, able to interact with numerous biological mediators (e.g. growth factors) and cells [24]. Hence, engineering ECM signals into biomaterials is a promising approach for applications in regenerative medicine [25]. GAGs such as native non-sulfated hyaluronan (HA) can be chemically sulfated leading to sulfated derivatives (sHA) with different interaction profiles [26]. Biophysical studies revealed that the interaction between mediator proteins and GAGs are mainly controlled by the sulfation degree and pattern as well as the carbohydrate backbone [27–30]. Thus, studying these HA derivatives contributes to an improved understanding of the structure-function relationships of GAG derivatives in their interaction with biological mediators. Accordingly, aECMs composed of fibrillar collagen (Coll) and GAGs can be used as functional biomaterials to mimic the cellular microenvironment [31]. In vitro studies with cells relevant for wound healing showed prominent effects for solute GAGs as well as those incorporated in Coll-based aECMs such as anti-inflammatory properties towards macrophages and dendritic cells, enhanced osteogenic differentiation of human mesenchymal stromal cells, improved proliferation of human dermal fibroblasts, decreased osteoclast resorption and improved osseointegration of dental implants in minipigs [32– 36]. Thus, aECMs are promising candidates for bioinspired biomaterials able to control and promote healing processes in bone and skin in particular in health-compromised patients. This study aimed to prove, that spheroids could also be used to study cell-ECM interactions quantitatively by seeding spheroids on aECMsbased coatings. Spheroids (10,000 cells) consisting of normal human osteoblasts (NHost), normal human dermal fibroblasts (NHDF) or co-cultures of both cell types in a 1:1 ratio were seeded on three different types of Coll-based aECMs containing low- or high-sulfated HA as well as a combination of both. Coll type I served as control. The changes in the size of the spheroids as well as the migration of outgrown cells on the aECMs were used as quantitative parameters to describe the interaction of the cells organized as spheroids with the aECMs and the influence of the cellular composition of the spheroids. 2. Materials and methods 2.1. Cell culture NHost (Lonza, Basel, Switzerland) were used in passage six and NHDF (PromoCell, Heidelberg, Germany) were used in passage five. Both cell types were cultivated under standard incubation conditions (95% humidity, 5% CO2) at 37 °C in NHost specific osteoblast growth medium (OGM) or in fibroblast specific Quantum 333 medium (Q333, PAA laboratories, Pasching, Austria), respectively. All cell culture media were supplemented with Penicillin-Streptomycin (1%, PAA). Cells were expanded in 75 cm2 or 25 cm2 tissue flasks (Greiner BioOne, Frickenhausen, Germany) until they reached near-confluence. For subcultivation of the cells, standard trypsination procedures were applied and the cells were seeded at a seeding density of 4000–5000 cells/cm2. The cell numbers were quantified with a hemocytometer.
311
solid agarose coating provides a non-adherent surface preventing cellular adhesion. Depending on the chosen cellular composition, a total number of 10,000 cells was transferred to each well. The cell culture medium was composed according to the cellular composition of the spheroids and standard cell culture conditions were applied. The cells sedimented in the wells but they were not able to adhere to the agarose. Consequently, they adhered to each other in a self-assembly process and after an overnight incubation one well-defined spheroid was generated per well (Fig. 1). Microscopic images of spheroids prior to nuclear staining were made with a Nikon Eclipse TS 100 phase contrast microscope (Nokia, Chiyoda, Japan) equipped with a Sony Cybershot DSC H10 digital camera (Sony, Minato, Japan). 2.3. Preparation of aECMs Low- and high-sulfated HA derivatives (sHA1, sHA4) were produced and characterized as previously described [27,37]. The sulfation degree (average number of sulfate groups per repeating disaccharide unit) is 1.4 for sHA1 and 3.6 for sHA4. The weight-average molecular weight analyzed by gel permeation chromatography (GPC) using laser light scattering detection is 20,255 Da for sHA1 (polydispersity index (PD) detected by GPC = 2.81) and 28,610 Da for sHA4 (PD = 1.74). Rat tail Coll type I was purchased from Corning (Kaiserslautern, Germany). Chamber slides (Sarstedt, Nümbrecht, Germany) were coated with Coll-based aECMs via in vitro-fibrillogenesis of 1 mg/ml Coll type I monomers dissolved in acetic acid in the presence or absence of 2.5 mM disaccharide units of sHA1 or sHA4 as well as a combination of both dissolved in 60 mM phosphate buffer (pH 7.4) for 16–18 h at 37 °C. After drying of the aECM suspensions and two washing steps with dH2O, the matrices were analyzed for their Coll and sHA contents as described previously [28,37,38]. Before use in cell culture experiments, the coatings were incubated with phosphate buffered saline (PBS) at 37 °C for 1 h. The analysis of aECM composition reveals a Coll-to-GAG mass ratio of 13:1 for sHA1, 12:1 for sHA4 and 11:1 for sHA1/sHA4. 2.4. Seeding of spheroids on aECMs Spheroids were picked from the 96 well plates and were collected in a petri-dish (Greiner BioOne) filled with the appropriate cell culture medium using pipette tips with a large orifice (VWR, Radnor, PA, USA) to prevent spheroids from damage. Single spheroids were collected out of the petri-dish in a total volume of 30–40 μl and placed in the middle of one cavity of the chamber slides coated with the prepared aECMs. The spheroids were allowed to adhere to the aECMs under standard cell culture conditions for 1 h. Subsequently, 300 μl of the appropriate cell culture medium was added and the spheroids were further incubated until day one, day three or day six, respectively. For fixation of the spheroids, the medium was removed from the cavities and the spheroids were washed twice with PBS. Fixation was done with paraformaldehyde (Carl Roth, Karlsruhe, Germany, 4% in PBS) for 10 min at room temperature, followed by an additional rinsing with PBS. In order to visualize cell nuclei, spheroids as well as out grown cells were stained with Hoechst 33,258 nuclear dye (Sigma-Aldrich, c = 50 μg/ml in PBS) for 15 min at room temperature followed by washing with PBS, rinsing with dH2O and air drying. Finally, the chambers were removed from the chamber slides and the slides were stored at 4 °C in the dark until microscopic analysis.
2.2. Generation of spheroids 2.5. Quantification of spheroid size and cellular migration As described previously, spheroids were generated using the LOT. All spheroids consisted of 10,000 cells and were either NHDF- and NHostmono-culture-spheroids or co-culture-spheroids of both cell types (1:1). In brief, a 1% agarose stock solution (Sigma-Aldrich, Taufkirchen, Germany) was prepared in distilled water (dH2O) and autoclaved. The bottom of 96 well flat-bottomed plates (Greiner BioOne) were coated with 50 μl hot liquid agarose per well. After cooling down, the semi-
The changes in spheroid size and cellular migration was evaluated and quantified microscopically. We used an Axioskop fluorescence microscope equipped with the Axiovision software for analyses (Zeiss, Oberkochen, Germany). Sizing of the spheroids was done on the basis of images taken at 50× magnification showing complete spheroids on all three days of analysis. The Axiovision software enables for direct
312
W. Metzger et al. / Materials Science and Engineering C 73 (2017) 310–318
Fig. 1. Microphotographs (phase contrast) showing the self-assembly of NHDF over time (0−22 h) applying the LOT. Scale bar 200 μm.
determination of the diameter of adhered spheroids. The diameter of each spheroid resulted from the mean value of the vertical and the horizontal measurement. The quantification of cellular migration was done on the basis of images taken at 100× magnification applying the measuring module of the Axiovision software. We defined cellular migration as the distance between the edge of the spheroid and the migration front of the outgrown cells. On day one, outgrown cells and the edge of the spheroids could be documented in one visual field. On day three and day six, this was not possible anymore due to the increased migration distance. Hence, several visual fields were observed starting from the edge of the spheroids and ending with the migration front. Since the dimensions of one visual field are known, the migration distance could be quantified easily. If possible, the cellular migration was quantified in all four cardinal directions for each spheroid. The number of outgrown cells from the spheroids was not determined in a quantitative way but was estimated in a semi-quantitative way. 2.6. Statistical analyses All numerical results presented in this study are shown as mean values with standard deviations. For each time point and each aECM, six spheroids of each cell type were seeded. Four to six spheroids were analyzed for the quantification of their size. A one way ANOVA followed by a Student-Newman-Keuls test for all pairwise multiple comparison was run. In case of not normally distributed data, a Kruskal-Wallis One Way ANOVA on Ranks followed by a Dunn's test for all pairwise multiple comparisons was run. As for the quantification of the spheroid size, the quantification of cellular migration based on the analysis of four to six spheroids, too. We always tried to measure the migration front in all four cardinal directions, but this was not possible for all samples. Statistical differences between the groups were identified as already described for the spheroid size. All statistical tests (p b 0.05) were done using the SigmaPlotSoftware Version 11.0 (Systat Software, Chicago, IL, USA). 3. Results 3.1. Generation of spheroids As previously reported [4,12], the LOT is suitable to generate uniform spheroids with a constant size and with a high yield. In this study, only
7.3 to 10.4% of the initially generated spheroids could not be used. In some cases, we found fibres of unknown origin in the cavities. In such cases, cells adhered to the fibres and did not form regular spherical 3D aggregates but elongated 3D aggregates. Furthermore, sometimes small spheroids could be seen in near vicinity of the main central spheroids. But the majority of spheroids in all three groups exhibited a well-defined morphology and a very regular shape. This can still be seen after seeding of the spheroids to the aECMs on day one (Figs. 2–4 and 6).
3.2. Size of the spheroids on aECMs Depending on the cellular composition of the spheroids, their size varied independently from the aECMs. This can be clearly seen, if the size of the adhered spheroids on day one is compared directly (Figs. 2–4a, d, g, j and 6): NHDF-spheroids exhibited the smallest size, NHost-spheroids were bigger and the co-culture-spheroids were in between. This impression is confirmed by the quantification of the diameter of the spheroids presented in Fig. 5. Especially on day three and day six (Fig. 5b, c), these cell specific differences in the size of the spheroids can be clearly seen for spheroids adhered to Coll (day three NHDF 467 ± 138 μm, NHost 664 ± 73 μm, co-culture 626 ± 62 μm; day six NHDF 454 ± 32 μm, NHost 805 ± 53 μm, co-culture 618 ± 88 μm) or to Coll/sHA1 (day three NHDF 374 ± 42 μm, NHost 564 ± 20 μm, co-culture 553 ± 30 μm; day six NHDF 443 ± 83 μm, NHost 697 ± 83 μm, coculture 529 ± 32 μm). The comparison of the spheroid size on the aECM revealed significant differences for NHDF on day one: NHDF-spheroids exhibited a maximum diameter on Coll (402 ± 24 μm) and on Coll/ sHA1/sHA4 (454 ± 23 μm). These differences could not be found on day three and on day six. No differences of spheroid size after adhesion to the aECMs could be seen for NHost on day one, but these spheroids covered a bigger area on Coll and on Coll/sHA1 on day three and day six. On all aECMs, the size of NHost-spheroids further increased until day six indicating a progressive transition from the spherical 3D morphology to a more 2D morphology, which is impressively shown in Fig. 4d–f. The size of the co-culture spheroids consisting of 50% NHDF and 50% NHost correlated with the cellular composition of these spheroids. Hence, the spheroids size on the four aECMs under investigation were approximately in between the size determined for the mono-culture experiments with a maximum size on Coll and Coll/sHA1 (Fig. 5a–c).
W. Metzger et al. / Materials Science and Engineering C 73 (2017) 310–318
313
Fig. 2. Microphotographs (phase contrast) of spheroids consisting of 10,000 NHDF on day one, day three and day six after seeding on different aECMs. Coll (a–c), Coll/sHA1 (d–f), Coll/sHA4 (g–i), Coll/sHA1/sHA4 (j–l). Scale bar 200 μm.
3.3. Outgrowth of cells from spheroids seeded on aECMs On day one after seeding on the aECMs, only few cells grew out from the spheroids on all aECMs under investigation, independent from the cellular composition of the spheroids (Figs. 2–4a, d, g, j). In some cases, almost no outgrown cells could be detected, especially on the aECM with a higher sulfation degree of the associated GAG derivative (Coll/sHA4 and Coll/sHA1/sHA4; Figs. 2g, 3g and j and 6h and j–l). As expected, more outgrown cells could be found on day three. Based on a semi-quantitative estimation, more cells grew out from NHDF-spheroids (Fig. 2) and co-culture-spheroids (Fig. 4) compared to NHostspheroids (Fig. 3). As already described for day one, outgrowth of cells from all three types of spheroids was enhanced on Coll and Coll/sHA1 (Figs. 2–4b, e). On day six, the number of outgrown cells was further increased. In general, NHDF-spheroids impressed by a high number of outgrown cells on Coll and Coll/sHA1 resulting in a confluent layer (Fig. 2c and f) surrounding the initial spheroid. On Coll/sHA1, the spheroid almost disappeared and could hardly be recognized in the resulting confluent 2D layer (Fig. 2f). In contrast, a clearly reduced number of
outgrown NHDF could be found on Coll/sHA4 and Coll/sHA1/sHA4 and the initial spheroid could still be seen (Fig. 2i and l). Comparable findings could be made for NHost on the different aECMs (Fig. 3) but with an overall reduced number of cells compared to NHDF. The reaction of the co-culture-spheroids on the contact with the aECMs was somehow in the middle between the reactions of the mono-culture-spheroid correlating to their cellular composition (Fig. 4). Of note, a higher number of outgrown cells from the co-culture-spheroids could be seen on Coll/ sHA4 and Coll/sHA1/sHA4 (Fig. 4i and l) compared to NHDF-spheroids (Fig. 2i and l) or NHost-spheroids (Fig. 3i and l). 3.4. Cellular migration on aECMs The second quantitative analytical parameter to assess the cell-surface interaction in our proposed 3D model was the determination of the migration front (Fig. 7). This parameter describes the migration capability of cells, which were outgrown from the adhered spheroids. Even on day one, clear differences in cellular migration on the aECMs could be observed (Fig. 7a). Cells from all three types of spheroids
314
W. Metzger et al. / Materials Science and Engineering C 73 (2017) 310–318
Fig. 3. Microphotographs (phase contrast) of spheroids consisting of 10,000 NHost on day one, day three and day six after seeding on different aECMs. Coll (a–c), Coll/sHA1 (d–f), Coll/sHA4 (g–i), Coll/sHA1/sHA4 (j–l). Scale bar 200 μm.
exhibited the highest migration on Coll followed by Coll/sHA1. NHDF did not differ in their migration on aECMs with different sulfation degree, but NHost as well as the co-culture showed a decreased cellular migration on Coll/sHA4 and Coll/sHA1/sHA4. On day three and day six, the migration of cells from all three types of spheroids exhibited a comparable graduated trend in cellular migration: The highest cellular migration was found on Coll followed by Coll/sHA1, Coll/sHA4 and Coll/sHA1/sHA4. Many of these results differed significantly (Fig. 7b and c). Differences between cellular migration on Coll/sHA4 and Coll/ sHA1/sHA4 could not be found for all spheroid types on all aECM especially not on day six (Fig. 7c). It is noteworthy, that the quantification of cell migration revealed a possible synergistic effect of the co-cultivation of NHDF with NHost, which is most pronounced on day six on the aECMs with a high sulfation degree. On Coll/sHA4, the mean cell migration of NHDF was 1280 ± 507 μm and of NHost 385 ± 44 μm but for the co-culture-spheroids, we determined a mean cell migration of 1800 ± 242 μm. The results for NHost and the co-culture differed significantly (p = 0.001) as revealed by an additional statistic test. A similar effect could be seen on Coll/sHA1/sHA4. On this aECM, the mean cell migration of NHDF was 887 ± 315 μm and of NHost was 368 ± 72 μm but
for the co-culture-spheroids we determined a mean cell migration of 1813 ± 257 μm. Here again, the results differed significantly (NHost and NHDF versus co-culture: p b 0.001).
4. Discussion GAGs represent important compounds of the ECM. A well-balanced interaction between the ECM and soluble growth factors is crucial for a physiological course of important biological events [26,39]. Consequently, Coll-based aECMs are interesting as scaffold materials for tissue engineering approaches [25] and especially oversulfated GAGs have the potential for the biofunctionalization of implants [40]. In previous studies, we characterized the biological effects of solute as well as matrixbound chemically oversulfated GAGs on several cell types in 2D in vitro experiments [30,32,37,40]. Spheroids as 3D aggregates of cells are able to reflect the normal situation of cells in tissues better than 2D cell culture approaches [16]. Consequently, this study aimed to prove, if spheroids could be used to study the interactions of cells with Coll/GAG-coated surfaces.
W. Metzger et al. / Materials Science and Engineering C 73 (2017) 310–318
315
Fig. 4. Microphotographs (phase contrast) of spheroids consisting of 5000 NHDF and 5000 NHost on day one, day three and day six after seeding on different aECMs. Coll (a-c), Coll/sHA1 (d–f), Coll/sHA4 (g–i), Coll/sHA1/sHA4 (j–l). Scale bar 200 μm.
It is widely accepted, that spheroids reflect the physiological situation of cells better than traditional 2D cultures [14–16]. Depending on the size and the age of a spheroid, gradients of nutrients, oxygen and metabolites are formed which could not be found in a flat 2D cell layer. As already reported in previous studies [4,12], the LOT allows for the highly reproducible generation of mono- and co-culture-spheroids of defined size. Spheroids incubated on cell-repellent agarose exhibit a continuous decrease in size over time, probably due to reorganization processes within the spheroids or a continuous decrease in the diameter of single cells due to their 3D arrangement in the spheroid. If spheroids are seeded on an adhesive surface, cells will start to grow out of the spheroid after its sedimentation. Over time, the spheroid will lose its 3D architecture and will increasingly transform into a 2D layer. The changes in the size of the spheroids and the migration of outgrown cells were used in this study as quantitative parameters for the evaluation of the cell-ECM interaction. Kramer et al. describe in their review the usage of a spheroid model to study basics and principles of cell migration. Spheroids were placed on plastic surfaces and the concentric area of outgrown cells were quantified, but they do not
evaluate the influence of different surfaces on cellular outgrowth [41]. To our knowledge, our study is the first attempt to use spheroids for studying the direct cell-surface interaction. Harvesting spheroids consisting of 10,000 cells was quite easy, because they can be seen with naked eye. After washing for removing cell debris, we seeded them in a small volume onto aECM-coatings and allowed them to adhere for 1 h before adding additional medium. In the future, the pre-incubation time should be prolonged, because not all spheroids were stationary after 1 h and detached again. In some cases, they adhered at the edge of the aECM-coating and therefore the migration could not be quantified in all four cardinal directions for every spheroid. The initial size of the spheroids is influenced by their cellular composition. As already described by our group, NHDF-spheroids are basically smaller compared to spheroids composed of osteoblasts even without adherence to any surface. However, spheroids of the same cellular composition could be generated reproducibly with the LOT resulting in spheroids of a constant size with a very low standard deviation [12]. It is likely, that differences in the size of single cells are responsible for
316
W. Metzger et al. / Materials Science and Engineering C 73 (2017) 310–318
Fig. 5. Diameter of spheroids (10,000 cells) consisting of NHDF, NHost and NHDF-NHost on day one (a), day three (b) and day six (c) after seeding on Coll (white bars), Coll/ sHA1 (black bars), Coll/sHA4 (light grey) and Coll/sHA1/sHA4 (dark grey). Significant differences (p b 0.05) are indicated by lines above the bars, n = 4–6 spheroids.
the differences in spheroid's size. In order to test this hypothesis, NHDF and NHost-spheroids could be dissociated into single cells [4] and the diameter of the suspended cells could be quantified by Casy®Technology. On day one NHDF-spheroids clearly reacted to the different kinds of aECMs and exhibited significant differences in their size. On day three and day six, no more differences could be seen indicating, that NHDF mainly react on the contact with the aECM in terms of early adhesion, whereas NHost-spheroids show no reaction on the early contact with the aECM. In a previous study, the influence of sulfated GAG derivatives on the cellular behavior was evaluated in 2D experiments indicating a positive influence of a higher sulfation degree of HA on initial cell adhesion and cell proliferation of human dermal fibroblasts [32]. Such a promoting effect could not be shown in this study for most of the spheroids and GAG-containing coatings under investigation. Of interest, on day one which can be considered as an early time point of spheroid adhesion the diameter of NHDF-spheroids was higher on Coll/sHA1/sHA4 compared to all other GAGs. Furthermore, in this study, which did not focus on cellular proliferation directly, NHDF from just one donor were used and not fibroblasts derived from biopsies from different
donors. Accordingly, donor specific variations must be considered and of course, fibroblasts organized in a spheroid might react different on the contact with aECMs compared to single cells. The diameter of NHost-spheroids clearly differed on day three and day six exhibiting the greatest diameter on Coll without GAGs. These results indicate that NHDF and NHost react differently to the contact with the aECM and that NHost prefer attachment to aECMs with no or low sulfated HA. The co-culture-spheroids respond according to cellular composition and no synergetic effects could be observed for this parameter. It is noteworthy, that the quantification of the diameter of the spheroids is sometimes challenging. Especially for later time points, the dimensions of the remaining spheroids are difficult to define (see Fig. 2f). Due to the high migration of NHDF (see Fig. 7), this problem mainly occurred for NHDF-spheroids leading to higher standard deviations for this type of spheroid (see Fig. 5b and c). However, in most cases, the dimensions of the remaining spheroids were still clearly to see until day six (see Figs. 3c, f, i, l and 4c, f, i, l). The diameter of the spheroids as a quantitative analysis describes, how fast the transition of the 3D organization of the cells to a 2D cell layer takes place. Accordingly, this parameter is able to describe, what is more attractive for the cells; remaining in the cell complex or growing on the surface. For the analysis of the cellular migration, spheroids in all three groups exhibited the same tendency: the highest cellular migration was determined on Coll, and the presence of sHA4 within the matrices impaired cellular migration. The observed differences were more pronounced on day three and on day six. But in contrast to the analysis of the spheroid's size, a possible synergistic effect of the co cultivation of NHDF and NHost in terms of an increased migration can be seen. Our results indicate a decreasing cellular migration with increasing sulfation degree of HA which might be also influenced by the different surface charges and fibril morphologies of the aECMs [37]. In previous studies, we demonstrated that the presence of GAGs during in vitro fibrillogenesis of Coll type I leads to a concentration- and sulfation dependent decrease of the fibril diameter [37,42]. However, because the cellular response to solute GAGs and aECM-bound GAGs are reported to be comparable we strongly believe that the differences between the fibril morphologies are of secondary importance in this work [37,40]. It is known, that the sulfation degree of GAGs can directly influences integrin signaling and related proteins responsible for cell adhesion via focal adhesions [43]. The authors were able to show, that integrin α5 and integrin β1, which are responsible for cell-matrix interactions and vinculin, which is an essential part of the focal adhesions were up-regulated on sHA compared to HA. Focal adhesions are specific protein clusters which are connected to the actin-fibres within in the cells and which are in contact with the underlying surface via transmembranous integrins [44]. Binding of integrins to ECM proteins is a major mechanism of cell-surface interaction [45]. Integrin signaling enables cells to sense their environment and to translate extracellular signals into intracellular signals [46]. Focal adhesions are responsible for an intensive adhesion of cells to a surface and an increased number of focal adhesions indicated by a high integrin expression upon contact with sulfated GAG derivatives might correlate with the lower cell migration on these aECMs. Cellular migration of the co-culture-spheroids does not reflect their cellular composition but is in the range of the NHDF-spheroids. Of interest, synergistic effects of co-cultures in spheroids could also be demonstrated for the rate of apoptotic cells: previously we reported a decreased rate of apoptotic endothelial cells after cocultivation with osteoblasts or NHDF, respectively [47]. In this study, we were able to show, that spheroids can be successfully used in cell-surface interaction studies. These studies using spheroids are not restricted to the influence of aECMs on cellular behavior. They can be extended to other related topics e.g. influence on nano- and microtopographies on cell proliferation and cell differentiation, which we studied in 2D in vitro studies before [1,2,48]. Furthermore, spheroids might be a useful tool for the evaluation of biofunctionalized surfaces, like surfaces with covalently immobilized growth factors [49]. For
W. Metzger et al. / Materials Science and Engineering C 73 (2017) 310–318
317
Fig. 6. Microphotographs (fluorescence) of spheroids (10,000 cells) consisting of NHDF (a, d, g, j), NHost (b, e, h, k) and NHDF-NHost (c, f, i, l) on day one after seeding on Coll (a–c), Coll/ sHA1 (d–f), Coll/sHA4 (g–i), Coll/sHA1/sHA4 (j–l). The nuclei of the outgrown cells can be easily detected after bisbenzimide staining and clear differences in the number of outgrown cells and the cellular migration between the aECM under investigation can be seen. Scale bar 200 μm.
further studies, the different cell populations could be stained with fluorescence dyes for living cells prior to the generation of the spheroids. This approach offers the possibility to evaluate a cell selective effect of the surface: Which cell prefers growing on the surface? Which cell prefers the 3D aggregate? Can we see any cell selective properties of the surface? Is it more attractive for a cell to stay within the spheroid or does the cell prefer to leave the 3D cell complex and to adhere to an artificial surface? Furthermore, cells could be detached from the surface, sorted and further examined by e.g. quantitative RT-PCR for the analysis of the differentiation status.
deal with the influence of different surface chemistries but also with the influence of different surface topographies on cellular physiology. Since spheroids reflect the physiological situation in normal tissues much better than traditional 2D-cultures, they might be a valuable addition to 2D studies. Labeling of different cell types prior to the formation of the spheroids can be done to enable further analytical tools e.g. determination of cell differentiation, cell proliferation or apoptosis by means of cell sorting and quantitative RT-PCR.
5. Conclusions
The authors would like to thank Daniela Sossong, Andreas Brauer and Aline Katzschner for excellent technical assistance. They further gratefully acknowledge financial support by the DFG (TRR 67, A2, A3, Z3).
In this study, we were able to show, that spheroids can be successfully used for the analysis of cell-surface interactions. Such studies might
Acknowledgments
318
W. Metzger et al. / Materials Science and Engineering C 73 (2017) 310–318
Fig. 7. Mean migration distance of cells outgrown from spheroids (10,000 cells) consisting of NHDF, NHost and NHDF-NHost on day one (a), day three (b) and day six (c) after seeding on Coll (white bars), Coll/sHA1 (black bars), Coll/sHA4 (light grey) and Coll/ sHA1/sHA4 (dark grey). Significant differences (p b 0.05) are indicated by lines above the bars, n = 4–6 spheroids.
References [1] W. Metzger, B. Schwab, M.M. Miro, S. Grad, A. Simpson, M. Veith, G. Wennemuth, V. Zaporojtchenko, S. Verrier, J.S. Hayes, M. Bubel, T. Pohlemann, C. Aktas, J. Biomed. Nanotechnol. 10 (2014) 831–845. [2] M. Oberringer, E. Akman, J. Lee, W. Metzger, C. Akkan, E. Kacar, A. Demir, H. AbdulKhaliq, N. Pütz, G. Wennemuth, T. Pohlemann, M. Veith, C. Aktas, Mater. Sci. Eng. C 33 (2013) 901–908. [3] Y. Li, Y. Wang, J. Ye, J. Yuan, Y. Xiao, Mater. Sci. Eng. C 68 (2016) 177–183. [4] N. Dorst, M. Oberringer, U. Grässer, T. Pohlemann, W. Metzger, Ann. Anat. 196 (2014) 303–311. [5] S. Nath, G.R. Devi, Pharmacol. Ther. 163 (2016) 94–108. [6] A. Moscona, Exp. Cell Res. 22 (1961) 455–475. [7] M. Heiss, M. Hellstrom, M. Kalen, T. May, H. Weber, M. Hecker, H.G. Augsutin, T. Korff, FASEB J. 29 (2015) 3076–3084. [8] L. Pfisterer, T. Korff, Methods Biol. 1430 (2016) 167–177. [9] J.M. Yuhas, A.P. Li, A.O. Martinez, A.J. Ladman, Cancer Res. 37 (1977) 3639–3643. [10] S. Messner, I. Agarkova, W. Moritz, J.M. Kelm, Arch. Toxicol. 87 (2013) 209–213.
[11] A. Ivascu, M. Kubbies, J. Biomol. Screen. 11 (2006) 922–932. [12] W. Metzger, D. Sossong, A. Bächle, N. Pütz, G. Wennemuth, T. Pohlemann, M. Oberringer, Cytotherapy 13 (2011) 1000–1012. [13] T.M. Achilli, J. Meyer, J.R. Morgan, Expert. Opin. Biol. Ther. 12 (2012) 1347–1360. [14] N.T. Elliott, F. Yuan, J. Pharm. Sci. 100 (2011) 59–74. [15] E. Fennema, N. Rivron, J. Rouwkema, C. van Blitterswijk, J. de Boer, Trends Biotechnol. 31 (2013) 108–115. [16] K.A. Fitzgerald, M. Malhotra, C.M. Curtin, F.J. O'Brien, C.M. O'Driscoll, J. Control, Release 215 (2015) 39–54. [17] M.C. Arufe, A. De la Fuente, I. Fuentes-Boquete, F.J. De Toro, F.J. Blanco, J. Cell. Biochem. 108 (2009) 145–155. [18] S.H. Bhang, S.W. Cho, W.G. La, T.J. Lee, H.S. Yang, A.Y. Sun, S.H. Baek, J.W. Rhie, B.S. Kim, Biomaterials 32 (2011) 2734–2747. [19] J.A. Green, K.M. Yamada, Adv. Drug Deliv. Rev. 59 (2007) 1293–1298. [20] C. Unger, N. Kramer, A. Walzl, M. Scherzer, M. Hengstschlager, H. Dolznig, Adv. Drug Deliv. Rev. 79–80 (2014) 50–67. [21] M.W. Laschke, T.E. Schank, C. Scheuer, S. Kleer, S. Schuler, W. Metzger, D. Eglin, M. Alini, M.D. Menger, Acta Biomater. 9 (2013) 6876–6884. [22] M.W. Laschke, M.D. Menger, Trends Biotechnol. (2016)http://dx.doi.org/10.1016/j. tibtech.2016.08.004 (Epub ahead of print). [23] L.A. Hidalgo-Bastida, S.H. Cartmell, Tissue Eng. Part B Rev. 16 (2010) 405–412. [24] A. Imberty, H. Lortat-Jacob, S. Perez, Carbohydr. Res. 342 (2007) 430–439. [25] S.F. Badylak, D.O. Freytes, T.W. Gilbert, Acta Biomater. 5 (2009) 1–13. [26] D. Scharnweber, L. Hubner, S. Rother, U. Hempel, U. Anderegg, S.A. Samsonov, M.T. Pisabarro, L. Hofbauer, M. Schnabelrauch, S. Franz, J. Simon, V. Hintze, J. Mater. Sci. Mater. Med. 26 (2015) 232. [27] V. Hintze, S. Moeller, M. Schnabelrauch, S. Bierbaum, M. Viola, H. Worch, D. Scharnweber, Biomacromolecules 10 (2009) 3290–3297. [28] V. Hintze, A. Miron, S. Moeller, M. Schnabelrauch, H.P. Wiesmann, H. Worch, D. Scharnweber, Acta Biomater. 8 (2012) 2144–2152. [29] V. Hintze, S.A. Samsonov, M. Anselmi, S. Moeller, J. Becher, M. Schnabelrauch, D. Scharnweber, M.T. Pisabarro, Biomacromolecules 15 (2014) 3083–3092. [30] J. Salbach-Hirsch, S.A. Samsonov, V. Hintze, C. Hofbauer, A.K. Picke, M. Rauner, J.P. Gehrcke, S. Moeller, M. Schnabelrauch, D. Scharnweber, M.T. Pisabarro, L.C. Hofbauer, Biomaterials 67 (2015) 335–345. [31] S. Bierbaum, D. Scharnweber, in: P. Ducheyne, K.E. Healy, D.W. Hutmacher, D.W. Grainger, C.J. Kirkpatrick (Eds.), Comprehensive Biomaterials, 2, Elsevier, Amsterdam 2011, pp. 127–153. [32] A. van der Smissen, V. Hintze, D. Scharnweber, S. Moeller, M. Schnabelrauch, A. Majok, J.C. Simon, U. Anderegg, Biomaterials 32 (2011) 8938–8946. [33] U. Hempel, V. Hintze, S. Moeller, M. Schnabelrauch, D. Scharnweber, P. Dieter, Acta Biomater. 8 (2012) 659–666. [34] J. Salbach, S. Kliemt, M. Rauner, T.D. Rachner, C. Goettsch, S. Kalkhof, M. von Bergen, S. Moeller, M. Schnabelrauch, V. Hintze, D. Scharnweber, L.C. Hofbauer, Biomaterials 33 (2012) 8418–8429. [35] S. Franz, F. Allenstein, J. Kajahn, I. Forstreuter, V. Hintze, S. Moeller, J.C. Simon, Acta Biomater. 9 (2013) 5621–5629. [36] P. Korn, M.C. Schulz, V. Hintze, U. Range, R. Mai, U. Eckelt, M. Schnabelrauch, S. Moeller, J. Becher, D. Scharnweber, B. Stadlinger, J. Biomed. Mater. Res. A. 102 (2014) 2334–2344. [37] S. Rother, J. Salbach-Hirsch, S. Moeller, T. Seemann, M. Schnabelrauch, L.C. Hofbauer, V. Hintze, D. Scharnweber, ACS Appl. Mater. Interfaces 7 (2015) 23787–23797. [38] V. Hintze, A. Miron, S. Moeller, M. Schnabelrauch, S. Heinemann, H. Worch, D. Scharnweber, J. Tissue Eng. Regen. Med. 8 (2014) 314–324. [39] I.T. Swinehart, S.F. Badylak, Dev. Dyn. 245 (2016) 351–360. [40] J. Salbach-Hirsch, N. Ziegler, S. Thiele, S. Moeller, M. Schnabelrauch, V. Hintze, D. Scharnweber, M. Rauner, L.C. Hofbauer, J. Cell. Biochem. 115 (2014) 1101–1111. [41] N. Kramer, A. Walzl, C. Unger, M. Rosner, G. Krupitza, M. Hengstschlaeger, H. Dolznig, Mutat. Res. 752 (2013) 10–24. [42] A. Miron, S. Rother, L. Huebner, U. Hempel, I. Käppler, S. Moeller, M. Schnabelrauch, D. Scharnweber, V. Hintze, Macromol. Biosci. 14 (2014) 1783–1794. [43] S. Kliemt, C. Lange, W. Otto, V. Hintze, S. Moeller, M. von Bergen, U. Hempel, S. Kalkhof, J. Proteome Res. 12 (2013) 378–389. [44] G.R. Owen, D.O. Meredith, I. Ap Gwynn, R.G. Richards, Eur. Cell Mater. 9 (2005) 85–96. [45] A. Chopra, M.E. Murray, F.J. Byfield, M.G. Mendez, R. Halleluyan, D.J. Restle, D. RazBen Aroush, P.A. Galie, K. Pogoda, R. Bucki, C. Marcinkiewicz, G.D. Prestwich, T.I. Zarembinski, C.S. Chen, E. Puré, J.Y. Kresh, P.A. Janmey, Biomaterials 35 (2014) 71–82. [46] J.S. Hayes, E.M. Czekanska, R.G. Richards, Adv. Biochem. Eng. Biotechnol. 126 (2012) 1–31. [47] R. Walser, W. Metzger, A. Görg, T. Pohlemann, M.D. Menger, M.W. Laschke, Eur. Cell Mater. 26 (2013) 222–233. [48] M. Veith, O.C. Aktas, W. Metzger, D. Sossong, H. Ullah Wazir, I. Grobelsek, N. Pütz, G. Wennemuth, T. Pohlemann, M. Oberringer, Biofabrication 2 (2012) 035001. [49] W. Metzger, N. Grenner, S.E. Motsch, R. Strehlow, T. Pohlemann, M. Oberringer, Tissue Eng. 13 (2007) 2751–2760.
Contents lists available at ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Evaluation of cell-surface interaction using a 3D spheroid cell culture model on artificial extracellular matrices Wolfgang Metzger a,⁎, Sandra Rother b, Tim Pohlemann a, Stephanie Möller c, Matthias Schnabelrauch c, Vera Hintze b, Dieter Scharnweber b a b c
Department of Trauma, Hand and Reconstructive Surgery, Saarland University, Building 57, 66421 Homburg, Germany Institute of Materials Science, Max Bergmann Center of Biomaterials, TU Dresden, Budapester Straße 27, 01069 Dresden, Germany Biomaterials Department, INNOVENT e. V., Prüssingstraße 27 B, 07745 Jena, Germany
a r t i c l e
i n f o
Article history: Received 13 July 2016 Received in revised form 25 October 2016 Accepted 17 December 2016 Available online 20 December 2016 Keywords: Spheroid Cell-surface interaction Sulfated hyaluronan Glycosaminoglycans
a b s t r a c t Since decades, cell-surface interactions are studied in 2D cell culture approaches, but cells organized in 3D (spheroids) reflect the normal situation of cells in tissues much better due to intense cell-cell-contacts. Accordingly, this study aimed to prove, if spheroids could be used to study cell-surface interaction. Spheroids consisting of fibroblasts and/or osteoblasts were seeded on artificial extracellular matrices. Here, nonsulfated hyaluronan as a biological relevant compound of the extracellular matrix was chemically sulfated to different extents and co-fibrillised with collagen. The changes of the spheroid diameters and the migration distance of outgrown cells after seeding on the matrices were used as parameters to evaluate cell-surface interaction quantitatively. Fibroblast-based spheroids reacted in the initial phase of adhesion with different spheroid sizes on the contact with the matrices. In contrast, the reaction of osteoblasts was more pronounced at later time points exhibiting a decrease of the size of the spheroids with increasing sulfation degree of the matrix. The migration of the cells was impaired by increasing sulfation degree, which might be caused by an increased expression of focal adhesion relevant proteins. In summary, spheroids can be used in cell-surface interaction studies and additional analytical tools could be implemented. © 2016 Elsevier B.V. All rights reserved.
1. Introduction The interaction of cells with an artificial surface of different chemical composition or different topographical features is the basis for the medical outcome of the implantation of any biomaterial. Accordingly, cellsurface interactions are relevant for a wide range of materials and applications e.g. osseointegration of orthopedic implants [1], restenosis of vascular stents [2] or development of novel scaffold materials for tissue engineering approaches [3] to name a few which have been studied with traditional 2D cell culture approaches in the past. Spheroids are spherical 3D cell aggregates. The cells adhere to each other in a self-assembly process without the need for using any stabilizing scaffold [4]. Depending on the cell type and the chosen experimental setup, a variety of methods can be used to generate spheroids from single cell suspensions, e.g. spinner flasks techniques [5], gyratory rotation systems [6], the carboxymethyl cellulose technique [7], hanging drop [8] or liquid overlay technique (LOT) [9]. In addition, novel single-use products are available on the market facilitating the high throughput ⁎ Corresponding author. E-mail address: [email protected] (W. Metzger).
http://dx.doi.org/10.1016/j.msec.2016.12.087 0928-4931/© 2016 Elsevier B.V. All rights reserved.
generation of spheroids [10]. In contrast to traditional 2D cell culture approaches, the cells do not adhere to an artificial surface e.g. polystyrene or glass, but they adhere to each other. Due to the intensive cell-cell contacts in three dimensions, the cells can interact with each other by paracrine mechanisms and by direct cell-cell contacts [11,12]. Furthermore, the cells can interact with proteins of the extracellular matrix (ECM) secreted by the spheroid forming cells their selves [13]. Hence, spheroids are able to reflect the physiological situation of cells in a tissue much better compared to conventional 2D cell cultures [14–16]. It is well known from literature, that a direct comparison of cells in 2D cultures with 3D cultures reveals clear differences in terms of cellular differentiation [17], gene expression profiles or secretion of growth factors [18]. Furthermore, cells organized in spheroids exhibited an increased resistance against hypoxia [18], a lower sensitivity against therapeutics [15], alterations in signaling pathways [19] and a decreased rate in apoptotic cell death [18]. Traditionally, spheroids were used in tumor research because of their more physiological sensitivity against cytotherapeutics [20]. However, during the last years, spheroids were also used to study angiogenesis, both in vivo [21] and in vitro [7], and in tissue engineering approaches where they might serve as building blocks [22]. This study was conducted to test the hypothesis, that
W. Metzger et al. / Materials Science and Engineering C 73 (2017) 310–318
spheroids might also be a useful tool to study the interaction of cells with artificial surfaces, e.g. artificial ECMs (aECMs). The native ECM synthesized by the local cells of a tissue plays a crucial role as 3D microenvironment for a variety of biological process like cell adhesion, proliferation, migration and differentiation [23]. Glycosaminoglycans (GAGs) are functional ECM components, able to interact with numerous biological mediators (e.g. growth factors) and cells [24]. Hence, engineering ECM signals into biomaterials is a promising approach for applications in regenerative medicine [25]. GAGs such as native non-sulfated hyaluronan (HA) can be chemically sulfated leading to sulfated derivatives (sHA) with different interaction profiles [26]. Biophysical studies revealed that the interaction between mediator proteins and GAGs are mainly controlled by the sulfation degree and pattern as well as the carbohydrate backbone [27–30]. Thus, studying these HA derivatives contributes to an improved understanding of the structure-function relationships of GAG derivatives in their interaction with biological mediators. Accordingly, aECMs composed of fibrillar collagen (Coll) and GAGs can be used as functional biomaterials to mimic the cellular microenvironment [31]. In vitro studies with cells relevant for wound healing showed prominent effects for solute GAGs as well as those incorporated in Coll-based aECMs such as anti-inflammatory properties towards macrophages and dendritic cells, enhanced osteogenic differentiation of human mesenchymal stromal cells, improved proliferation of human dermal fibroblasts, decreased osteoclast resorption and improved osseointegration of dental implants in minipigs [32– 36]. Thus, aECMs are promising candidates for bioinspired biomaterials able to control and promote healing processes in bone and skin in particular in health-compromised patients. This study aimed to prove, that spheroids could also be used to study cell-ECM interactions quantitatively by seeding spheroids on aECMsbased coatings. Spheroids (10,000 cells) consisting of normal human osteoblasts (NHost), normal human dermal fibroblasts (NHDF) or co-cultures of both cell types in a 1:1 ratio were seeded on three different types of Coll-based aECMs containing low- or high-sulfated HA as well as a combination of both. Coll type I served as control. The changes in the size of the spheroids as well as the migration of outgrown cells on the aECMs were used as quantitative parameters to describe the interaction of the cells organized as spheroids with the aECMs and the influence of the cellular composition of the spheroids. 2. Materials and methods 2.1. Cell culture NHost (Lonza, Basel, Switzerland) were used in passage six and NHDF (PromoCell, Heidelberg, Germany) were used in passage five. Both cell types were cultivated under standard incubation conditions (95% humidity, 5% CO2) at 37 °C in NHost specific osteoblast growth medium (OGM) or in fibroblast specific Quantum 333 medium (Q333, PAA laboratories, Pasching, Austria), respectively. All cell culture media were supplemented with Penicillin-Streptomycin (1%, PAA). Cells were expanded in 75 cm2 or 25 cm2 tissue flasks (Greiner BioOne, Frickenhausen, Germany) until they reached near-confluence. For subcultivation of the cells, standard trypsination procedures were applied and the cells were seeded at a seeding density of 4000–5000 cells/cm2. The cell numbers were quantified with a hemocytometer.
311
solid agarose coating provides a non-adherent surface preventing cellular adhesion. Depending on the chosen cellular composition, a total number of 10,000 cells was transferred to each well. The cell culture medium was composed according to the cellular composition of the spheroids and standard cell culture conditions were applied. The cells sedimented in the wells but they were not able to adhere to the agarose. Consequently, they adhered to each other in a self-assembly process and after an overnight incubation one well-defined spheroid was generated per well (Fig. 1). Microscopic images of spheroids prior to nuclear staining were made with a Nikon Eclipse TS 100 phase contrast microscope (Nokia, Chiyoda, Japan) equipped with a Sony Cybershot DSC H10 digital camera (Sony, Minato, Japan). 2.3. Preparation of aECMs Low- and high-sulfated HA derivatives (sHA1, sHA4) were produced and characterized as previously described [27,37]. The sulfation degree (average number of sulfate groups per repeating disaccharide unit) is 1.4 for sHA1 and 3.6 for sHA4. The weight-average molecular weight analyzed by gel permeation chromatography (GPC) using laser light scattering detection is 20,255 Da for sHA1 (polydispersity index (PD) detected by GPC = 2.81) and 28,610 Da for sHA4 (PD = 1.74). Rat tail Coll type I was purchased from Corning (Kaiserslautern, Germany). Chamber slides (Sarstedt, Nümbrecht, Germany) were coated with Coll-based aECMs via in vitro-fibrillogenesis of 1 mg/ml Coll type I monomers dissolved in acetic acid in the presence or absence of 2.5 mM disaccharide units of sHA1 or sHA4 as well as a combination of both dissolved in 60 mM phosphate buffer (pH 7.4) for 16–18 h at 37 °C. After drying of the aECM suspensions and two washing steps with dH2O, the matrices were analyzed for their Coll and sHA contents as described previously [28,37,38]. Before use in cell culture experiments, the coatings were incubated with phosphate buffered saline (PBS) at 37 °C for 1 h. The analysis of aECM composition reveals a Coll-to-GAG mass ratio of 13:1 for sHA1, 12:1 for sHA4 and 11:1 for sHA1/sHA4. 2.4. Seeding of spheroids on aECMs Spheroids were picked from the 96 well plates and were collected in a petri-dish (Greiner BioOne) filled with the appropriate cell culture medium using pipette tips with a large orifice (VWR, Radnor, PA, USA) to prevent spheroids from damage. Single spheroids were collected out of the petri-dish in a total volume of 30–40 μl and placed in the middle of one cavity of the chamber slides coated with the prepared aECMs. The spheroids were allowed to adhere to the aECMs under standard cell culture conditions for 1 h. Subsequently, 300 μl of the appropriate cell culture medium was added and the spheroids were further incubated until day one, day three or day six, respectively. For fixation of the spheroids, the medium was removed from the cavities and the spheroids were washed twice with PBS. Fixation was done with paraformaldehyde (Carl Roth, Karlsruhe, Germany, 4% in PBS) for 10 min at room temperature, followed by an additional rinsing with PBS. In order to visualize cell nuclei, spheroids as well as out grown cells were stained with Hoechst 33,258 nuclear dye (Sigma-Aldrich, c = 50 μg/ml in PBS) for 15 min at room temperature followed by washing with PBS, rinsing with dH2O and air drying. Finally, the chambers were removed from the chamber slides and the slides were stored at 4 °C in the dark until microscopic analysis.
2.2. Generation of spheroids 2.5. Quantification of spheroid size and cellular migration As described previously, spheroids were generated using the LOT. All spheroids consisted of 10,000 cells and were either NHDF- and NHostmono-culture-spheroids or co-culture-spheroids of both cell types (1:1). In brief, a 1% agarose stock solution (Sigma-Aldrich, Taufkirchen, Germany) was prepared in distilled water (dH2O) and autoclaved. The bottom of 96 well flat-bottomed plates (Greiner BioOne) were coated with 50 μl hot liquid agarose per well. After cooling down, the semi-
The changes in spheroid size and cellular migration was evaluated and quantified microscopically. We used an Axioskop fluorescence microscope equipped with the Axiovision software for analyses (Zeiss, Oberkochen, Germany). Sizing of the spheroids was done on the basis of images taken at 50× magnification showing complete spheroids on all three days of analysis. The Axiovision software enables for direct
312
W. Metzger et al. / Materials Science and Engineering C 73 (2017) 310–318
Fig. 1. Microphotographs (phase contrast) showing the self-assembly of NHDF over time (0−22 h) applying the LOT. Scale bar 200 μm.
determination of the diameter of adhered spheroids. The diameter of each spheroid resulted from the mean value of the vertical and the horizontal measurement. The quantification of cellular migration was done on the basis of images taken at 100× magnification applying the measuring module of the Axiovision software. We defined cellular migration as the distance between the edge of the spheroid and the migration front of the outgrown cells. On day one, outgrown cells and the edge of the spheroids could be documented in one visual field. On day three and day six, this was not possible anymore due to the increased migration distance. Hence, several visual fields were observed starting from the edge of the spheroids and ending with the migration front. Since the dimensions of one visual field are known, the migration distance could be quantified easily. If possible, the cellular migration was quantified in all four cardinal directions for each spheroid. The number of outgrown cells from the spheroids was not determined in a quantitative way but was estimated in a semi-quantitative way. 2.6. Statistical analyses All numerical results presented in this study are shown as mean values with standard deviations. For each time point and each aECM, six spheroids of each cell type were seeded. Four to six spheroids were analyzed for the quantification of their size. A one way ANOVA followed by a Student-Newman-Keuls test for all pairwise multiple comparison was run. In case of not normally distributed data, a Kruskal-Wallis One Way ANOVA on Ranks followed by a Dunn's test for all pairwise multiple comparisons was run. As for the quantification of the spheroid size, the quantification of cellular migration based on the analysis of four to six spheroids, too. We always tried to measure the migration front in all four cardinal directions, but this was not possible for all samples. Statistical differences between the groups were identified as already described for the spheroid size. All statistical tests (p b 0.05) were done using the SigmaPlotSoftware Version 11.0 (Systat Software, Chicago, IL, USA). 3. Results 3.1. Generation of spheroids As previously reported [4,12], the LOT is suitable to generate uniform spheroids with a constant size and with a high yield. In this study, only
7.3 to 10.4% of the initially generated spheroids could not be used. In some cases, we found fibres of unknown origin in the cavities. In such cases, cells adhered to the fibres and did not form regular spherical 3D aggregates but elongated 3D aggregates. Furthermore, sometimes small spheroids could be seen in near vicinity of the main central spheroids. But the majority of spheroids in all three groups exhibited a well-defined morphology and a very regular shape. This can still be seen after seeding of the spheroids to the aECMs on day one (Figs. 2–4 and 6).
3.2. Size of the spheroids on aECMs Depending on the cellular composition of the spheroids, their size varied independently from the aECMs. This can be clearly seen, if the size of the adhered spheroids on day one is compared directly (Figs. 2–4a, d, g, j and 6): NHDF-spheroids exhibited the smallest size, NHost-spheroids were bigger and the co-culture-spheroids were in between. This impression is confirmed by the quantification of the diameter of the spheroids presented in Fig. 5. Especially on day three and day six (Fig. 5b, c), these cell specific differences in the size of the spheroids can be clearly seen for spheroids adhered to Coll (day three NHDF 467 ± 138 μm, NHost 664 ± 73 μm, co-culture 626 ± 62 μm; day six NHDF 454 ± 32 μm, NHost 805 ± 53 μm, co-culture 618 ± 88 μm) or to Coll/sHA1 (day three NHDF 374 ± 42 μm, NHost 564 ± 20 μm, co-culture 553 ± 30 μm; day six NHDF 443 ± 83 μm, NHost 697 ± 83 μm, coculture 529 ± 32 μm). The comparison of the spheroid size on the aECM revealed significant differences for NHDF on day one: NHDF-spheroids exhibited a maximum diameter on Coll (402 ± 24 μm) and on Coll/ sHA1/sHA4 (454 ± 23 μm). These differences could not be found on day three and on day six. No differences of spheroid size after adhesion to the aECMs could be seen for NHost on day one, but these spheroids covered a bigger area on Coll and on Coll/sHA1 on day three and day six. On all aECMs, the size of NHost-spheroids further increased until day six indicating a progressive transition from the spherical 3D morphology to a more 2D morphology, which is impressively shown in Fig. 4d–f. The size of the co-culture spheroids consisting of 50% NHDF and 50% NHost correlated with the cellular composition of these spheroids. Hence, the spheroids size on the four aECMs under investigation were approximately in between the size determined for the mono-culture experiments with a maximum size on Coll and Coll/sHA1 (Fig. 5a–c).
W. Metzger et al. / Materials Science and Engineering C 73 (2017) 310–318
313
Fig. 2. Microphotographs (phase contrast) of spheroids consisting of 10,000 NHDF on day one, day three and day six after seeding on different aECMs. Coll (a–c), Coll/sHA1 (d–f), Coll/sHA4 (g–i), Coll/sHA1/sHA4 (j–l). Scale bar 200 μm.
3.3. Outgrowth of cells from spheroids seeded on aECMs On day one after seeding on the aECMs, only few cells grew out from the spheroids on all aECMs under investigation, independent from the cellular composition of the spheroids (Figs. 2–4a, d, g, j). In some cases, almost no outgrown cells could be detected, especially on the aECM with a higher sulfation degree of the associated GAG derivative (Coll/sHA4 and Coll/sHA1/sHA4; Figs. 2g, 3g and j and 6h and j–l). As expected, more outgrown cells could be found on day three. Based on a semi-quantitative estimation, more cells grew out from NHDF-spheroids (Fig. 2) and co-culture-spheroids (Fig. 4) compared to NHostspheroids (Fig. 3). As already described for day one, outgrowth of cells from all three types of spheroids was enhanced on Coll and Coll/sHA1 (Figs. 2–4b, e). On day six, the number of outgrown cells was further increased. In general, NHDF-spheroids impressed by a high number of outgrown cells on Coll and Coll/sHA1 resulting in a confluent layer (Fig. 2c and f) surrounding the initial spheroid. On Coll/sHA1, the spheroid almost disappeared and could hardly be recognized in the resulting confluent 2D layer (Fig. 2f). In contrast, a clearly reduced number of
outgrown NHDF could be found on Coll/sHA4 and Coll/sHA1/sHA4 and the initial spheroid could still be seen (Fig. 2i and l). Comparable findings could be made for NHost on the different aECMs (Fig. 3) but with an overall reduced number of cells compared to NHDF. The reaction of the co-culture-spheroids on the contact with the aECMs was somehow in the middle between the reactions of the mono-culture-spheroid correlating to their cellular composition (Fig. 4). Of note, a higher number of outgrown cells from the co-culture-spheroids could be seen on Coll/ sHA4 and Coll/sHA1/sHA4 (Fig. 4i and l) compared to NHDF-spheroids (Fig. 2i and l) or NHost-spheroids (Fig. 3i and l). 3.4. Cellular migration on aECMs The second quantitative analytical parameter to assess the cell-surface interaction in our proposed 3D model was the determination of the migration front (Fig. 7). This parameter describes the migration capability of cells, which were outgrown from the adhered spheroids. Even on day one, clear differences in cellular migration on the aECMs could be observed (Fig. 7a). Cells from all three types of spheroids
314
W. Metzger et al. / Materials Science and Engineering C 73 (2017) 310–318
Fig. 3. Microphotographs (phase contrast) of spheroids consisting of 10,000 NHost on day one, day three and day six after seeding on different aECMs. Coll (a–c), Coll/sHA1 (d–f), Coll/sHA4 (g–i), Coll/sHA1/sHA4 (j–l). Scale bar 200 μm.
exhibited the highest migration on Coll followed by Coll/sHA1. NHDF did not differ in their migration on aECMs with different sulfation degree, but NHost as well as the co-culture showed a decreased cellular migration on Coll/sHA4 and Coll/sHA1/sHA4. On day three and day six, the migration of cells from all three types of spheroids exhibited a comparable graduated trend in cellular migration: The highest cellular migration was found on Coll followed by Coll/sHA1, Coll/sHA4 and Coll/sHA1/sHA4. Many of these results differed significantly (Fig. 7b and c). Differences between cellular migration on Coll/sHA4 and Coll/ sHA1/sHA4 could not be found for all spheroid types on all aECM especially not on day six (Fig. 7c). It is noteworthy, that the quantification of cell migration revealed a possible synergistic effect of the co-cultivation of NHDF with NHost, which is most pronounced on day six on the aECMs with a high sulfation degree. On Coll/sHA4, the mean cell migration of NHDF was 1280 ± 507 μm and of NHost 385 ± 44 μm but for the co-culture-spheroids, we determined a mean cell migration of 1800 ± 242 μm. The results for NHost and the co-culture differed significantly (p = 0.001) as revealed by an additional statistic test. A similar effect could be seen on Coll/sHA1/sHA4. On this aECM, the mean cell migration of NHDF was 887 ± 315 μm and of NHost was 368 ± 72 μm but
for the co-culture-spheroids we determined a mean cell migration of 1813 ± 257 μm. Here again, the results differed significantly (NHost and NHDF versus co-culture: p b 0.001).
4. Discussion GAGs represent important compounds of the ECM. A well-balanced interaction between the ECM and soluble growth factors is crucial for a physiological course of important biological events [26,39]. Consequently, Coll-based aECMs are interesting as scaffold materials for tissue engineering approaches [25] and especially oversulfated GAGs have the potential for the biofunctionalization of implants [40]. In previous studies, we characterized the biological effects of solute as well as matrixbound chemically oversulfated GAGs on several cell types in 2D in vitro experiments [30,32,37,40]. Spheroids as 3D aggregates of cells are able to reflect the normal situation of cells in tissues better than 2D cell culture approaches [16]. Consequently, this study aimed to prove, if spheroids could be used to study the interactions of cells with Coll/GAG-coated surfaces.
W. Metzger et al. / Materials Science and Engineering C 73 (2017) 310–318
315
Fig. 4. Microphotographs (phase contrast) of spheroids consisting of 5000 NHDF and 5000 NHost on day one, day three and day six after seeding on different aECMs. Coll (a-c), Coll/sHA1 (d–f), Coll/sHA4 (g–i), Coll/sHA1/sHA4 (j–l). Scale bar 200 μm.
It is widely accepted, that spheroids reflect the physiological situation of cells better than traditional 2D cultures [14–16]. Depending on the size and the age of a spheroid, gradients of nutrients, oxygen and metabolites are formed which could not be found in a flat 2D cell layer. As already reported in previous studies [4,12], the LOT allows for the highly reproducible generation of mono- and co-culture-spheroids of defined size. Spheroids incubated on cell-repellent agarose exhibit a continuous decrease in size over time, probably due to reorganization processes within the spheroids or a continuous decrease in the diameter of single cells due to their 3D arrangement in the spheroid. If spheroids are seeded on an adhesive surface, cells will start to grow out of the spheroid after its sedimentation. Over time, the spheroid will lose its 3D architecture and will increasingly transform into a 2D layer. The changes in the size of the spheroids and the migration of outgrown cells were used in this study as quantitative parameters for the evaluation of the cell-ECM interaction. Kramer et al. describe in their review the usage of a spheroid model to study basics and principles of cell migration. Spheroids were placed on plastic surfaces and the concentric area of outgrown cells were quantified, but they do not
evaluate the influence of different surfaces on cellular outgrowth [41]. To our knowledge, our study is the first attempt to use spheroids for studying the direct cell-surface interaction. Harvesting spheroids consisting of 10,000 cells was quite easy, because they can be seen with naked eye. After washing for removing cell debris, we seeded them in a small volume onto aECM-coatings and allowed them to adhere for 1 h before adding additional medium. In the future, the pre-incubation time should be prolonged, because not all spheroids were stationary after 1 h and detached again. In some cases, they adhered at the edge of the aECM-coating and therefore the migration could not be quantified in all four cardinal directions for every spheroid. The initial size of the spheroids is influenced by their cellular composition. As already described by our group, NHDF-spheroids are basically smaller compared to spheroids composed of osteoblasts even without adherence to any surface. However, spheroids of the same cellular composition could be generated reproducibly with the LOT resulting in spheroids of a constant size with a very low standard deviation [12]. It is likely, that differences in the size of single cells are responsible for
316
W. Metzger et al. / Materials Science and Engineering C 73 (2017) 310–318
Fig. 5. Diameter of spheroids (10,000 cells) consisting of NHDF, NHost and NHDF-NHost on day one (a), day three (b) and day six (c) after seeding on Coll (white bars), Coll/ sHA1 (black bars), Coll/sHA4 (light grey) and Coll/sHA1/sHA4 (dark grey). Significant differences (p b 0.05) are indicated by lines above the bars, n = 4–6 spheroids.
the differences in spheroid's size. In order to test this hypothesis, NHDF and NHost-spheroids could be dissociated into single cells [4] and the diameter of the suspended cells could be quantified by Casy®Technology. On day one NHDF-spheroids clearly reacted to the different kinds of aECMs and exhibited significant differences in their size. On day three and day six, no more differences could be seen indicating, that NHDF mainly react on the contact with the aECM in terms of early adhesion, whereas NHost-spheroids show no reaction on the early contact with the aECM. In a previous study, the influence of sulfated GAG derivatives on the cellular behavior was evaluated in 2D experiments indicating a positive influence of a higher sulfation degree of HA on initial cell adhesion and cell proliferation of human dermal fibroblasts [32]. Such a promoting effect could not be shown in this study for most of the spheroids and GAG-containing coatings under investigation. Of interest, on day one which can be considered as an early time point of spheroid adhesion the diameter of NHDF-spheroids was higher on Coll/sHA1/sHA4 compared to all other GAGs. Furthermore, in this study, which did not focus on cellular proliferation directly, NHDF from just one donor were used and not fibroblasts derived from biopsies from different
donors. Accordingly, donor specific variations must be considered and of course, fibroblasts organized in a spheroid might react different on the contact with aECMs compared to single cells. The diameter of NHost-spheroids clearly differed on day three and day six exhibiting the greatest diameter on Coll without GAGs. These results indicate that NHDF and NHost react differently to the contact with the aECM and that NHost prefer attachment to aECMs with no or low sulfated HA. The co-culture-spheroids respond according to cellular composition and no synergetic effects could be observed for this parameter. It is noteworthy, that the quantification of the diameter of the spheroids is sometimes challenging. Especially for later time points, the dimensions of the remaining spheroids are difficult to define (see Fig. 2f). Due to the high migration of NHDF (see Fig. 7), this problem mainly occurred for NHDF-spheroids leading to higher standard deviations for this type of spheroid (see Fig. 5b and c). However, in most cases, the dimensions of the remaining spheroids were still clearly to see until day six (see Figs. 3c, f, i, l and 4c, f, i, l). The diameter of the spheroids as a quantitative analysis describes, how fast the transition of the 3D organization of the cells to a 2D cell layer takes place. Accordingly, this parameter is able to describe, what is more attractive for the cells; remaining in the cell complex or growing on the surface. For the analysis of the cellular migration, spheroids in all three groups exhibited the same tendency: the highest cellular migration was determined on Coll, and the presence of sHA4 within the matrices impaired cellular migration. The observed differences were more pronounced on day three and on day six. But in contrast to the analysis of the spheroid's size, a possible synergistic effect of the co cultivation of NHDF and NHost in terms of an increased migration can be seen. Our results indicate a decreasing cellular migration with increasing sulfation degree of HA which might be also influenced by the different surface charges and fibril morphologies of the aECMs [37]. In previous studies, we demonstrated that the presence of GAGs during in vitro fibrillogenesis of Coll type I leads to a concentration- and sulfation dependent decrease of the fibril diameter [37,42]. However, because the cellular response to solute GAGs and aECM-bound GAGs are reported to be comparable we strongly believe that the differences between the fibril morphologies are of secondary importance in this work [37,40]. It is known, that the sulfation degree of GAGs can directly influences integrin signaling and related proteins responsible for cell adhesion via focal adhesions [43]. The authors were able to show, that integrin α5 and integrin β1, which are responsible for cell-matrix interactions and vinculin, which is an essential part of the focal adhesions were up-regulated on sHA compared to HA. Focal adhesions are specific protein clusters which are connected to the actin-fibres within in the cells and which are in contact with the underlying surface via transmembranous integrins [44]. Binding of integrins to ECM proteins is a major mechanism of cell-surface interaction [45]. Integrin signaling enables cells to sense their environment and to translate extracellular signals into intracellular signals [46]. Focal adhesions are responsible for an intensive adhesion of cells to a surface and an increased number of focal adhesions indicated by a high integrin expression upon contact with sulfated GAG derivatives might correlate with the lower cell migration on these aECMs. Cellular migration of the co-culture-spheroids does not reflect their cellular composition but is in the range of the NHDF-spheroids. Of interest, synergistic effects of co-cultures in spheroids could also be demonstrated for the rate of apoptotic cells: previously we reported a decreased rate of apoptotic endothelial cells after cocultivation with osteoblasts or NHDF, respectively [47]. In this study, we were able to show, that spheroids can be successfully used in cell-surface interaction studies. These studies using spheroids are not restricted to the influence of aECMs on cellular behavior. They can be extended to other related topics e.g. influence on nano- and microtopographies on cell proliferation and cell differentiation, which we studied in 2D in vitro studies before [1,2,48]. Furthermore, spheroids might be a useful tool for the evaluation of biofunctionalized surfaces, like surfaces with covalently immobilized growth factors [49]. For
W. Metzger et al. / Materials Science and Engineering C 73 (2017) 310–318
317
Fig. 6. Microphotographs (fluorescence) of spheroids (10,000 cells) consisting of NHDF (a, d, g, j), NHost (b, e, h, k) and NHDF-NHost (c, f, i, l) on day one after seeding on Coll (a–c), Coll/ sHA1 (d–f), Coll/sHA4 (g–i), Coll/sHA1/sHA4 (j–l). The nuclei of the outgrown cells can be easily detected after bisbenzimide staining and clear differences in the number of outgrown cells and the cellular migration between the aECM under investigation can be seen. Scale bar 200 μm.
further studies, the different cell populations could be stained with fluorescence dyes for living cells prior to the generation of the spheroids. This approach offers the possibility to evaluate a cell selective effect of the surface: Which cell prefers growing on the surface? Which cell prefers the 3D aggregate? Can we see any cell selective properties of the surface? Is it more attractive for a cell to stay within the spheroid or does the cell prefer to leave the 3D cell complex and to adhere to an artificial surface? Furthermore, cells could be detached from the surface, sorted and further examined by e.g. quantitative RT-PCR for the analysis of the differentiation status.
deal with the influence of different surface chemistries but also with the influence of different surface topographies on cellular physiology. Since spheroids reflect the physiological situation in normal tissues much better than traditional 2D-cultures, they might be a valuable addition to 2D studies. Labeling of different cell types prior to the formation of the spheroids can be done to enable further analytical tools e.g. determination of cell differentiation, cell proliferation or apoptosis by means of cell sorting and quantitative RT-PCR.
5. Conclusions
The authors would like to thank Daniela Sossong, Andreas Brauer and Aline Katzschner for excellent technical assistance. They further gratefully acknowledge financial support by the DFG (TRR 67, A2, A3, Z3).
In this study, we were able to show, that spheroids can be successfully used for the analysis of cell-surface interactions. Such studies might
Acknowledgments
318
W. Metzger et al. / Materials Science and Engineering C 73 (2017) 310–318
Fig. 7. Mean migration distance of cells outgrown from spheroids (10,000 cells) consisting of NHDF, NHost and NHDF-NHost on day one (a), day three (b) and day six (c) after seeding on Coll (white bars), Coll/sHA1 (black bars), Coll/sHA4 (light grey) and Coll/ sHA1/sHA4 (dark grey). Significant differences (p b 0.05) are indicated by lines above the bars, n = 4–6 spheroids.
References [1] W. Metzger, B. Schwab, M.M. Miro, S. Grad, A. Simpson, M. Veith, G. Wennemuth, V. Zaporojtchenko, S. Verrier, J.S. Hayes, M. Bubel, T. Pohlemann, C. Aktas, J. Biomed. Nanotechnol. 10 (2014) 831–845. [2] M. Oberringer, E. Akman, J. Lee, W. Metzger, C. Akkan, E. Kacar, A. Demir, H. AbdulKhaliq, N. Pütz, G. Wennemuth, T. Pohlemann, M. Veith, C. Aktas, Mater. Sci. Eng. C 33 (2013) 901–908. [3] Y. Li, Y. Wang, J. Ye, J. Yuan, Y. Xiao, Mater. Sci. Eng. C 68 (2016) 177–183. [4] N. Dorst, M. Oberringer, U. Grässer, T. Pohlemann, W. Metzger, Ann. Anat. 196 (2014) 303–311. [5] S. Nath, G.R. Devi, Pharmacol. Ther. 163 (2016) 94–108. [6] A. Moscona, Exp. Cell Res. 22 (1961) 455–475. [7] M. Heiss, M. Hellstrom, M. Kalen, T. May, H. Weber, M. Hecker, H.G. Augsutin, T. Korff, FASEB J. 29 (2015) 3076–3084. [8] L. Pfisterer, T. Korff, Methods Biol. 1430 (2016) 167–177. [9] J.M. Yuhas, A.P. Li, A.O. Martinez, A.J. Ladman, Cancer Res. 37 (1977) 3639–3643. [10] S. Messner, I. Agarkova, W. Moritz, J.M. Kelm, Arch. Toxicol. 87 (2013) 209–213.
[11] A. Ivascu, M. Kubbies, J. Biomol. Screen. 11 (2006) 922–932. [12] W. Metzger, D. Sossong, A. Bächle, N. Pütz, G. Wennemuth, T. Pohlemann, M. Oberringer, Cytotherapy 13 (2011) 1000–1012. [13] T.M. Achilli, J. Meyer, J.R. Morgan, Expert. Opin. Biol. Ther. 12 (2012) 1347–1360. [14] N.T. Elliott, F. Yuan, J. Pharm. Sci. 100 (2011) 59–74. [15] E. Fennema, N. Rivron, J. Rouwkema, C. van Blitterswijk, J. de Boer, Trends Biotechnol. 31 (2013) 108–115. [16] K.A. Fitzgerald, M. Malhotra, C.M. Curtin, F.J. O'Brien, C.M. O'Driscoll, J. Control, Release 215 (2015) 39–54. [17] M.C. Arufe, A. De la Fuente, I. Fuentes-Boquete, F.J. De Toro, F.J. Blanco, J. Cell. Biochem. 108 (2009) 145–155. [18] S.H. Bhang, S.W. Cho, W.G. La, T.J. Lee, H.S. Yang, A.Y. Sun, S.H. Baek, J.W. Rhie, B.S. Kim, Biomaterials 32 (2011) 2734–2747. [19] J.A. Green, K.M. Yamada, Adv. Drug Deliv. Rev. 59 (2007) 1293–1298. [20] C. Unger, N. Kramer, A. Walzl, M. Scherzer, M. Hengstschlager, H. Dolznig, Adv. Drug Deliv. Rev. 79–80 (2014) 50–67. [21] M.W. Laschke, T.E. Schank, C. Scheuer, S. Kleer, S. Schuler, W. Metzger, D. Eglin, M. Alini, M.D. Menger, Acta Biomater. 9 (2013) 6876–6884. [22] M.W. Laschke, M.D. Menger, Trends Biotechnol. (2016)http://dx.doi.org/10.1016/j. tibtech.2016.08.004 (Epub ahead of print). [23] L.A. Hidalgo-Bastida, S.H. Cartmell, Tissue Eng. Part B Rev. 16 (2010) 405–412. [24] A. Imberty, H. Lortat-Jacob, S. Perez, Carbohydr. Res. 342 (2007) 430–439. [25] S.F. Badylak, D.O. Freytes, T.W. Gilbert, Acta Biomater. 5 (2009) 1–13. [26] D. Scharnweber, L. Hubner, S. Rother, U. Hempel, U. Anderegg, S.A. Samsonov, M.T. Pisabarro, L. Hofbauer, M. Schnabelrauch, S. Franz, J. Simon, V. Hintze, J. Mater. Sci. Mater. Med. 26 (2015) 232. [27] V. Hintze, S. Moeller, M. Schnabelrauch, S. Bierbaum, M. Viola, H. Worch, D. Scharnweber, Biomacromolecules 10 (2009) 3290–3297. [28] V. Hintze, A. Miron, S. Moeller, M. Schnabelrauch, H.P. Wiesmann, H. Worch, D. Scharnweber, Acta Biomater. 8 (2012) 2144–2152. [29] V. Hintze, S.A. Samsonov, M. Anselmi, S. Moeller, J. Becher, M. Schnabelrauch, D. Scharnweber, M.T. Pisabarro, Biomacromolecules 15 (2014) 3083–3092. [30] J. Salbach-Hirsch, S.A. Samsonov, V. Hintze, C. Hofbauer, A.K. Picke, M. Rauner, J.P. Gehrcke, S. Moeller, M. Schnabelrauch, D. Scharnweber, M.T. Pisabarro, L.C. Hofbauer, Biomaterials 67 (2015) 335–345. [31] S. Bierbaum, D. Scharnweber, in: P. Ducheyne, K.E. Healy, D.W. Hutmacher, D.W. Grainger, C.J. Kirkpatrick (Eds.), Comprehensive Biomaterials, 2, Elsevier, Amsterdam 2011, pp. 127–153. [32] A. van der Smissen, V. Hintze, D. Scharnweber, S. Moeller, M. Schnabelrauch, A. Majok, J.C. Simon, U. Anderegg, Biomaterials 32 (2011) 8938–8946. [33] U. Hempel, V. Hintze, S. Moeller, M. Schnabelrauch, D. Scharnweber, P. Dieter, Acta Biomater. 8 (2012) 659–666. [34] J. Salbach, S. Kliemt, M. Rauner, T.D. Rachner, C. Goettsch, S. Kalkhof, M. von Bergen, S. Moeller, M. Schnabelrauch, V. Hintze, D. Scharnweber, L.C. Hofbauer, Biomaterials 33 (2012) 8418–8429. [35] S. Franz, F. Allenstein, J. Kajahn, I. Forstreuter, V. Hintze, S. Moeller, J.C. Simon, Acta Biomater. 9 (2013) 5621–5629. [36] P. Korn, M.C. Schulz, V. Hintze, U. Range, R. Mai, U. Eckelt, M. Schnabelrauch, S. Moeller, J. Becher, D. Scharnweber, B. Stadlinger, J. Biomed. Mater. Res. A. 102 (2014) 2334–2344. [37] S. Rother, J. Salbach-Hirsch, S. Moeller, T. Seemann, M. Schnabelrauch, L.C. Hofbauer, V. Hintze, D. Scharnweber, ACS Appl. Mater. Interfaces 7 (2015) 23787–23797. [38] V. Hintze, A. Miron, S. Moeller, M. Schnabelrauch, S. Heinemann, H. Worch, D. Scharnweber, J. Tissue Eng. Regen. Med. 8 (2014) 314–324. [39] I.T. Swinehart, S.F. Badylak, Dev. Dyn. 245 (2016) 351–360. [40] J. Salbach-Hirsch, N. Ziegler, S. Thiele, S. Moeller, M. Schnabelrauch, V. Hintze, D. Scharnweber, M. Rauner, L.C. Hofbauer, J. Cell. Biochem. 115 (2014) 1101–1111. [41] N. Kramer, A. Walzl, C. Unger, M. Rosner, G. Krupitza, M. Hengstschlaeger, H. Dolznig, Mutat. Res. 752 (2013) 10–24. [42] A. Miron, S. Rother, L. Huebner, U. Hempel, I. Käppler, S. Moeller, M. Schnabelrauch, D. Scharnweber, V. Hintze, Macromol. Biosci. 14 (2014) 1783–1794. [43] S. Kliemt, C. Lange, W. Otto, V. Hintze, S. Moeller, M. von Bergen, U. Hempel, S. Kalkhof, J. Proteome Res. 12 (2013) 378–389. [44] G.R. Owen, D.O. Meredith, I. Ap Gwynn, R.G. Richards, Eur. Cell Mater. 9 (2005) 85–96. [45] A. Chopra, M.E. Murray, F.J. Byfield, M.G. Mendez, R. Halleluyan, D.J. Restle, D. RazBen Aroush, P.A. Galie, K. Pogoda, R. Bucki, C. Marcinkiewicz, G.D. Prestwich, T.I. Zarembinski, C.S. Chen, E. Puré, J.Y. Kresh, P.A. Janmey, Biomaterials 35 (2014) 71–82. [46] J.S. Hayes, E.M. Czekanska, R.G. Richards, Adv. Biochem. Eng. Biotechnol. 126 (2012) 1–31. [47] R. Walser, W. Metzger, A. Görg, T. Pohlemann, M.D. Menger, M.W. Laschke, Eur. Cell Mater. 26 (2013) 222–233. [48] M. Veith, O.C. Aktas, W. Metzger, D. Sossong, H. Ullah Wazir, I. Grobelsek, N. Pütz, G. Wennemuth, T. Pohlemann, M. Oberringer, Biofabrication 2 (2012) 035001. [49] W. Metzger, N. Grenner, S.E. Motsch, R. Strehlow, T. Pohlemann, M. Oberringer, Tissue Eng. 13 (2007) 2751–2760.