Optimizing structural and mechanical properties of ...

Materials Science and Engineering C 71 (2017) 465–472

Contents lists available at ScienceDirect

Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec

Optimizing structural and mechanical properties of cryogel scaffolds for use in prostate cancer cell culturing A. Cecilia a,⁎, A. Baecker b, E. Hamann a, A. Rack c, T. van de Kamp d, F.J. Gruhl b, R. Hofmann a, J. Moosmann e, S. Hahn a, J. Kashef a, S. Bauer a, T. Farago a, L. Helfen a,c, T. Baumbach a,d a

Institute for Photon Science and Synchrotron Radiation (IPS), Karlsruhe Institute of Technology, Hermann-von-Helmholtz-Platz 1, D-76344 Eggenstein-Leopoldshafen, Germany Institute of Microstructure Technology (IMT), Karlsruhe Institute of Technology (KIT), Hermann-von-Helmholtz-Platz 1 Bldg 329, Eggenstein-Leopoldshafen, Karlsruhe D-76344, Germany c European Synchrotron Radiation Facility (ESRF), 6 rue Jules Horowitz, 38000 Grenoble, France d Laboratory for Applications of Synchrotron Radiation (LAS), Karlsruhe Institute of Technology, 6980, D-76128 Karlsruhe, Germany e Institute of Materials Research, Helmholtz-Zentrum Geesthacht (HZG), Max-Planck-Str. 1, D-21502 Geesthacht, Germany b

a r t i c l e

i n f o

Article history: Received 18 May 2016 Received in revised form 13 September 2016 Accepted 18 October 2016 Available online 21 October 2016 Keywords: Cryogel scaffolds Prostate cancer cell Synchrotron X-ray computed micro tomography

a b s t r a c t Prostate cancer (PCa) currently is the second most diagnosed cancer in men and the second most cause of cancer death after lung cancer in Western societies. This sets the necessity of modelling prostatic disorders to optimize a therapy against them. The conventional approach to investigating prostatic diseases is based on two-dimensional (2D) cell culturing. This method, however, does not provide a three-dimensional (3D) environment, therefore impeding a satisfying simulation of the prostate gland in which the PCa cells proliferate. Cryogel scaffolds represent a valid alternative to 2D culturing systems for studying the normal and pathological behavior of the prostate cells thanks to their 3D pore architecture that reflects more closely the physiological environment in which PCa cells develop. In this work the 3D morphology of three potential scaffolds for PCa cell culturing was investigated by means of synchrotron X-ray computed micro tomography (SXCμT) fitting the according requirements of high spatial resolution, 3D imaging capability and low dose requirements very well. In combination with mechanical tests, the results allowed identifying an optimal cryogel architecture, meeting the needs for a well-suited scaffold to be used for 3D PCa cell culture applications. The selected cryogel was then used for culturing prostatic lymph node metastasis (LNCaP) cells and subsequently, the presence of multi-cellular tumor spheroids inside the matrix was demonstrated again by using SXCμT. © 2016 Elsevier B.V. All rights reserved.

1. Introduction The proliferation mechanisms of cancer cells and their responses to drugs, compounds or signaling molecules have been studied by 2D cell culturing for many decades. Limitations of this method have been increasingly recognized over the years [1], the main disadvantage being the inability of the rigid plastic dishes or plastic flasks in which the cells are cultured to reflect the realistic 3D structuring of the cell microenvironment and thus to provide an optimal surrounding for the culturing of established and primary cell lines [2]. An important progress in cell culturing has been the introduction of 3D culture systems, i.e. cryogel scaffolds that are synthetized by using natural and synthetic polymers [3]. Cryogels are macroporous systems, which are formed at temperatures below 0 °C. The monomer or polymer precursors are dissolved in water which forms ice crystals at such low temperatures. During this process, the growing water ice crystals build a network surrounded by a non-frozen microphase, in which the precursors are concentrated and undergo chemical reactions. After the

⁎ Corresponding author.

http://dx.doi.org/10.1016/j.msec.2016.10.038 0928-4931/© 2016 Elsevier B.V. All rights reserved.

complete polymerization and thawing of the gels at room temperature, the molten ice crystals leave behind a highly interconnected porous system [4]. This interconnected system of pores provides mechanical stability and mimics simultaneously the 3D cellular environment where the cells, which are to be investigated, can proliferate and differentiate. Moreover, it guarantees the nutrient and oxygen supply to the cells as well as the removal of metabolic waste products. Thanks to all these properties, cryogel scaffolds represent a useful system to analyze functional and biomechanical features of cell-matrix interactions and pathogenesis [5,6]. To be successfully applied to cell culturing, a cryogel scaffold should optimally mimic the natural environment and the extracellular matrix of the tissue in which the cells grow. This implicates that the cryogel has to exhibit specific mechanical and architectural properties that affect the biological functionality of the system. The thickness of the pore walls, for example, influences the stiffness and mechanical strength of the cryogel material and thus the cell adhesion to the scaffold. Both the mechanical and the morphological parameters of a cryogel can be controlled by altering its composition (monomer/polymer and cross-linker concentration) and/or the synthesis process of

466

A. Cecilia et al. / Materials Science and Engineering C 71 (2017) 465–472

the gels, which can be optimized depending on the application they are devoted to [7,8,9,10]. Cryogel scaffolds also represent a promising basis for the in vitro mimics of prostatic disorders such as benign prostatic hyperplasia and various stages of prostate cancer including castration resistant and metastatic disease [11]. Research efforts have been oriented towards finding the most suitable cryogel composition to model prostatic disorders and thus to improve the therapeutic approaches against prostate cancers, with PCa being the second most diagnosed cancer in men and the second most cause of cancer death after lung cancer in Western societies [12,13]. In order to be used for PCa cell culturing, a cryogel needs to possess mechanical and architectural properties suitable to mimic the prostate gland tissue inside which the cells normally proliferate. Its elastic modulus should be in the range of the prostatic tissue elastic modulus (between 29 kPa and 71.3 kPa [14]). Besides that, the pores of the scaffold need to be interconnected to a certain extent to guarantee both the transport of nutrients to the cells and the removal of the metabolic waste products during the cell culturing. In this work three different cryogel matrices were studied, as potential scaffolds for prostatic cancer cell culturing (see below for cryogels composition). To investigate the inner structure of the cryogel scaffolds and thus to judge whether their 3D architecture is suitable for the desired application, the samples were inspected with synchrotron X-ray computed micro tomography (SXCμT). SXCμT is a standard technique that allows us to investigate nondestructively the 3D internal structure of samples at sub-micrometer spatial resolution. Other imaging techniques are available, namely magnetic resonance tomography (NMR) and electron microscopy (TEM and SEM). However, the former exhibits a much lower spatial resolution (typically 25 μm) and throughput (typically 1 h per tomogram) and the latter, despite having a high quasi 2D resolution, has very limited depth information. In addition, both TEM and SEM require the physical sectioning of the sample [15,16]. In this work, we complement SXCμT with results achieved from mechanical tests for determining the elastic modulus and water uptake capacity of the investigated samples to designate an optimal cryogel composition, which combines mechanical stability and good mass transport, therefore meeting the requirements for realistic prostatic cell culturing. In a next step, we culture PCa cells in the chosen cryogel scaffold and evaluate the presence of multi-cellular tumor spheroids inside the matrix by means of SXCμT.

Table 1 Sample composition of the three different cryogel samples used in this work. Sample name

Cryogel description

Sample1 PEGda-fibrinogen-gelatin

Sample2 Chitosan-agarose-gelatin (CAG)

Sample3 Poly(2-hydroxyethyl methacrylate)-alginate-gelatin (pHAG)

Cryogel components Poly(ethylene glycol) diacrylate Fibrinogen from human plasma (50–70% protein) Gelatin (from cold water fish skin) Chitosan from shrimp shells Agarose (low gelling temperature) Gelatin (from cold water fish skin) 2-Hydroxyethyl methacrylate (HEMA) Gelatin (from cold water fish skin) Sodium alginate (from brown algae)

extracted from cold water fish skin (Teleostean) by acid extraction, its molecular weight is 60,000 [21]. All cryogel scaffolds were synthesized at the same temperature of − 21 °C, which leads to an optimal interconnected ice crystal system and thus a higher porosity as would be attained at different synthesis temperatures, see exemplarily Fig. 1 for the composition of sample3. In this case, the temperature of −21 °C allows to achieve an elastic modulus of 31.94 kPa, which lies well in the range of the prostatic tissue elastic moduli (between 29 kPa and 71.3 kPa [14]). For the production process of each cryogel, glutaraldehyde was used as chemical cross-linker to improve mechanical properties and the water sensitivity of gelatin [23]. At low concentration of b 0.6% v/v applied during synthesis the cytotoxicity of glutaraldehyde can be neglected [24]. In the following sections, the details of the production processes for each sample investigated are provided. 2.1.1. Sample1 The cryogel from which sample1 has been derived is normally used for wound repair and culturing of dermal fibroblasts [17] and has an elastic modulus of ~ 129 kPa. The latter was tuned to 38.75 kPa by adding polyethylene glycol diacrylate (PEGda) to the reaction mixture.

2. Materials and methods 2.1. Samples preparation The samples investigated in this work (cf. Table 1) were synthesized at the Institute of Microstructure Technology (IMT) at the Karlsruhe Institute for Technology (KIT, Karlsruhe) by modifying the chemical composition of cryogels which are already used in tissue engineering applications, namely wound repair and dermal fibroblast culturing as well as cartilage repair and lung muscle regeneration [17,18,19]. The composition of each cryogel was adjusted either by adding new constituents to the reaction mixture (polyethylene glycol diacrylate in sample1) or by changing the proportions of the constituents already in use (chitosan, agarose and gelatin for sample2 and poly(2hydroxyethyl methacrylate)-alginate-gelatin for sample3) in order to tune their elastic moduli to that of the prostatic cell tissue. The common feature of the cryogels is the inclusion of gelatin in their compositions. Gelatin is in fact a denatured collagen possessing the amino acid sequence Arg-Gly-Asp (RGD), which is known to enhance cell attachment and proliferation [20]. Besides that, gelatin is biocompatible, non-immunogenic, and inexpensive. Gelatin used was

Fig. 1. Total porosity and elasticity (E-Modulus) vs. different synthesis temperature for sample3.1 1 All quantitative measurements were performed on at least triplicates. All values are expressed as means ± standard deviations (SD). One-way ANOVA with post hoc Tukey tests [22] were used to compare the quantitative measurements (elastic modulus and total porosity) relatively to the room temperature results and p b 0.05 was used to assess statistical significance (p N 0.05 not significant; p b 0.0001 ***; p b 0.001 **; p b 0.05 *).


467

Fig. 2. Schematic of the experimental set-up used for the white beam synchrotron tomography.

In detail, a solution of fibrinogen from human plasma (Sigma) was mixed, by stirring, with a solution of cold water fish skin gelatin (Sigma) and PEGda (Sigma) as cross-linker. After addition of ammonium persulfate/petramethylethylenediamine (APS/TEMED, 0.5%/0.1%

w/v) the filled syringes were frozen overnight at −21 °C to start the polymerization of PEGda and glutaraldehyde (25% v/v) as cross-linker between gelatin and fibrinogen. The polymerized cryogels were cut into 3 mm thick disks with a diameter of 8 mm (by using a home-made

Fig. 3. Flow chart detailing the methodologies used to investigate the cryogel scaffolds.

468


Fig. 4. (a) X-ray tomography reconstructed slice and inset: detail of the highly absorbing regions (b) 3D volume rendering of sample1.

cutting tool), thawed at room temperature and stored in doubledistilled water at 4 °C prior to use. 2.1.2. Sample2 The cryogel on which sample2 is based finds application in cartilage tissue engineering [18] and has a high mechanical strength (~950 kPa). Its elastic modulus was adjusted to 43.16 kPa by changing the proportions of the cryogel constituents, i.e. chitosan-agarose-gelatin (CAG). Chitosan used for the synthesis of sample2 was extracted from shrimp shells. Its deacetylation degree is ≥75% and its molecular weight varies between 190 kDa and 375 kDa [25]. Starting from the self-gelation property of agarose at low temperatures, gelatin was cross-linked with chitosan via peptide bond formation by glutaraldehyde. Chitosan (Sigma) was dissolved in 1% aqueous acetic acid solution (pH 2.4) with the help of mechanical stirring at room temperature. After complete dissolution of chitosan and addition of gelatin (Sigma) the stirring was continued. The final concentration of chitosan–gelatin in the polymer reaction mixture depended on the chitosan concentration used. Separately, agarose (low gelling temperature, Sigma) was dissolved in deionized water by boiling until the solution became clear (indicating that agarose has completely dissolved). After cooling down to 50– 55 °C the chitosan-gelatin solution was added and mixed followed by the addition of glutaraldehyde solution (0.2%, v/v) from the stock solution (25%, v/v). The reaction mixture was poured into 3 ml syringes and incubated overnight at −21 °C. The ready-to-use cryogel samples finally were stored at 4 °C. 2.1.3. Sample3 This cryogel was generated by modifying the composition of a scaffold usually used in lung tissue engineering [19], initially showing a low elasticity of ~ 5–6 kPa [26]. The tuning of the elastic modulus to 31.94 kPa was achieved by modifying the proportion of all the cryogel components. Different concentrations of 2-hydroxyethyl methacrylate (HEMA, Merck), gelatin (Sigma) and alginate (from brown algae, Sigma), containing water as solvent were used to synthesize the cryogel matrix. The gels were prepared by mixing gelatin and alginate in double-distilled water. The monomer HEMA and the cross linker PEGda (Sigma) were added to the reaction mixture after complete dissolution of the previous polymers. The gel formation was initiated by adding APS/TEMED (0.5%/0.1% w/v) and glutaraldehyde (25% v/v) as cross-linker between gelatin and alginate. The mixture was poured immediately into 3 ml syringes and frozen overnight at −21 °C. The polymerized cryogels were cut into 3 mm thick disks with a diameter of 8 mm, thawed at room temperature and stored in double-distilled water at 4 °C prior to use. After synthesis of each cryogel, the scaffolds were washed three times in phosphate buffered saline (PBS) and stored in distilled water

at room temperature for at least 1 h to remove residues to a degree where their impact on cell proliferation could safely be neglected [27]. After the optimal cryogel composition was determined, this scaffold system was used for culturing PCa cells derived from lymph node metastasis (LNCaP) cells [28]. A solution containing 2.5 ∗ 105 cells in 20 μl cell culture medium was pipetted onto the cryogel scaffolds. The cells were then allowed to attach to the scaffold by keeping them for 2 h in an incubator at 37 °C. After 2 h fresh medium was added to the sample. During the cell culturing, the proliferation of the PCa cells was observed over a time period of three weeks, and the metabolic activity of the cells was monitored on day 0, 7, 14 and 21 using CellTiter-Glo®-3D (Promega). This viability assay is a homogenous method to determine the number of living cells in 3D cell culture model systems. After 21 days of culturing, cryogel scaffolds were prepared for investigation by optical microscopy. These scaffolds were paraffin embedded and cut into 7 μm thick slices. The cell spheroids were stained with hematoxylin/eosin (HE), hematoxylin staining acidic structures purpleblue (e.g. the cell nucleus) and eosin basic structures reddish or pink (e.g. the cytoplasm, cell membranes and extracellular fibers) [29]. For details on sample preparation, see [30]. Another uncut cryogel scaffold was used to nondestructively investigate the cell proliferation by means of SXCμT. After 21 days of culturing, the PCa cells were fixed with glutharaldehyde solution (2.5% v/v) and dried prior to investigation by X-rays. 2.2. Experimental methods The tomography experiments on the scaffolds without cells were performed at the TOPO-TOMO beamline [31] of the ANKA synchrotron located at the Karlsruhe Institute of Technology (KIT). The available white beam energy spectrum ranges between 1.5 keV and 40 keV with a flux at sample position (30 m from the source) of 1016 ph/s over an area of 10 mm × 10 mm. The white-beam spectrum (used for the tomography experiments) was filtered with 200 μm Al. The experimental set-up is depicted in Fig. 2. The samples were positioned on an air-bearing rotary stage positioned on a linear stage, which was used to move the sample into the

Table 2 Morphological and mechanicals properties of the cryogels. Sample

Pore volume (%)

Elastic modulus (kPa)

Water uptake capacity (%)

Pore area range (μm2)

Mean pore area (μm2)

Sample1 Sample2 Sample3

11.8 84.7 43.5

38.75 43.16 31.94

38.87 88.40 65.58

– 149–1.75e5 9.3–3.8e4

9541 1759


469

Fig. 5. (a) X-ray tomography reconstructed slice and (b) 3D volume rendering of sample2.

beam shortly after reference radiographs (flatfields) were acquired in order to correct for beam inhomogeneities. The X-ray projections were recorded with an indirectly converting X-ray area detector composed of a 200-μm-thick LuAG:Ce scintillator, an Optique Peter white beam microscope providing a magnification of 3.6 × [32] and a PCO.DIMAX camera with 2016 × 2016 pixels with a physical pixel size of 11 μm. The effective pixel size was 3.05 μm and the effective field of view 6 mm × 6 mm. For each tomography experiment 3000 projections were recorded within 40 s, and the distance between the sample and the detector (propagation distance) was set to 200 mm. Before reconstruction, the frames were processed with the phase retrieval ImageJ plugin ANKAphase [33]. The tomographic reconstruction was done with the PyHST software developed at the European Synchrotron Radiation Facility (ESRF, Grenoble, France). The segmentation of the reconstructed slices was done with the modules “Image Labeling” and “Labels Analysis” of the AVIZO Fire 8.1 3D Software. 3D volume rendering (Figs. 4b, 5b and 6b) was performed with Drishti 2.5.1. The tomography experiment on the scaffold used for cell culturing was performed at the ID19 beamline of the ESRF by using the socalled “pink beam” radiation. The spectrum was filtered with 1.4 mm of Aluminum that cuts off the low-energy part of the beam. The resulting energy spectrum was centered around 26.3 keV with an effective bandwidth of ΔE/E ~ 3%. The distance between the sample and the detector was set to 7100 mm. The detector system used is composed of a pco.edge camera and an optical system that provides a magnification of 4 × and that was designed and custom-made by optique peter [34] to work in combination with 4k × 4k chips (with larger pixels around 14–15 μm like FReLoN etc.). The effective pixel size is 1.8 μm. The mechanical integration of

the optical system was done at ESRF, the design is white-beam like, i.e. the lens looks on the mirror and there is no objective downstream of the scintillator. Here, the recorded frames were processed with the so-called “quasi-particle” phase-retrieval algorithm developed from R. Hofmann et al. [35,36]. The tomographic reconstruction was performed with an inhouse code developed in the UFO framework project [37]. As outlined in the flow chart shown in Fig. 3, the morphology study of synthetized cryogels was performed by SXCμT and complemented with mechanical tests (stress and water uptake capacity measurements) to identify the sample meeting the requirements for realistic prostatic cell culturing, namely sample3 (see below). The stress tests were conducted at IMT on uniform disk-like samples of cryogel saturated in RPMI (Roswell Park Memorial Institute) medium (cell culture medium) for 24 h. Uniaxial stress was exerted onto these disks using a Zwick/Roell Z0.5 machine with a 50 N load cell and a constant linear compression rate of 1.0 mm/min [30]. This allowed to extract the stress-strain curve and the elastic modulus, the latter being defined a as the slope in the elastic region. The water uptake capacity measurement was carried out at room temperature by a conventional gravimetric procedure [5]: before and after swelling the samples in RPMI culture medium for 10 min they were weighed to determine the relative fluid uptake capacity. 3. Experimental results and discussion As illustrated in Fig. 4, the tomographic reconstructed slice and 3D volume of sample1 indicate that this cryogel is characterized by a dense structure with only few and not interconnected pores that constitute 11.8% of the total sample volume.

Fig. 6. (a) X-ray tomography reconstructed slice and (b) 3D volume rendering of sample3.

470


Fig. 7. Comparison of segmented slices selected from the central part of the 3D volumes of (a) sample2 and (b) sample3. The individual pores were differently colored to better illustrate their geometry.

Due to such a low porosity, the water uptake capacity (cf. Table 2) and in this context the nutrient and oxygen supply would not be guaranteed, the latter being the most important feature which ensures the survival and growth of the cells in the scaffold. The absence of pore networking detected with SXCμT is an indication that sample1 cannot be used for PCa culturing. Further, the reconstructed tomographic slice (Fig. 4a) evidences the presence of bright aggregates of spherical structures with diameter b 12 μm. These regions can be ascribed to higher density regions inside the scaffold and thus to inhomogeneities in the cryogel composition. A possible explanation to the observed inhomogeneity is that the crosslinker concentration of PEGda in the cryogel composition was so high that the polymer chains already started networking in the solution before the ice crystals started forming. The tomography study performed on sample2 showed that the synthesis process of this sample was successful. The pores of sample2 constitute a large portion (84.7%) of the sample volume and were found to be highly interconnected (Fig. 5), demonstrating the suitability of this cryogel for cell culturing. As it can be seen in Fig. 5a, the architecture of sample2 has a honeycomb like structure that is typical of cryogels containing chitosan agarose and gelatine (CAG) in their initial composition [38]. The thickness of the pore walls, which could not be determined in sample1 due to the unsuccessful synthesis process, ranges between 24 μm and 33 μm. The production process was also successful for sample3 and similarly to sample2, the pores turned out to be interconnected (Fig. 6). However, sample3 is characterized by a more compact structure with respect to sample2 with the pores occupying 43.5% of the total volume. As illustrated in Fig. 6a, the pores of this cryogel have a characteristic elongated geometry disrupted by transversal thinner pore walls. The thickness of the pore walls in sample3 ranges between 44 μm and 85 μm and the thickness of the transversal interconnections vary between 22 μm and 51 μm. The results of the mechanical tests performed at IMT are listed in Table 2. Both sample2 and sample3 show elastic moduli values compatible with the elastic modulus of the prostatic cell tissue (Table 2). In addition, their water uptake capacity is sufficiently high to guarantee the nutrients transport to the cells and the removal of waste products. In other words, from the viewpoint of the mechanical properties these two cryogels are suitable for PCa cell culturing. To quantify and compare the pore sizes in sample2 and sample3, reconstructed slices selected from the central part of the corresponding 3D volume were segmented with AVIZO (Fig. 7). The individual pore areas in each segmented slice were marked with different colors to illustrate more clearly the structure of the two cryogels. The comparison of the segmented slices demonstrates that the architectures of sample2 and sample3 are quite dissimilar, which unavoidably affects the biological functionality of the system and thus the use of the cryogel for the desired application. The pores in sample2 are bigger than the voids in sample3, with the mean pore area in

sample2 being five times larger than the one of sample3. The higher pore dimensions of sample2 can be a limit for the PCa cell culturing. In fact, when the cells are cultured in bigger pores, the probability is higher that hypoxic cores form inside the tumor cell aggregates, where the cells do not receive nutrients and subsequently die [39]. The composition of sample3 leads to a characteristic compartment structure of the polymer chains system, which results from the presence of transversal thin walls inside the elongated pores. The smaller average pore size in sample3 allows for a controlled and increased uptake kinetic of moisture (e.g. cell culture medium), which facilitates a more reliable cell attachment and proliferation. Besides that, the geometry of the pores in sample3 resembles very closely the glandular prostatic structures, in which the cancer cells proliferate when the prostate is affected by prostate adenocarcinoma [40]. In conclusion, the combination of the architectural and mechanical properties of sample3 makes this cryogel the optimal in vitro 3D scaffold for PCa cell culturing. After having found that sample3 is optimal for PCa cell culturing, prostatic lymph node metastatic cells (LNCaP) were cultivated up to 21 days in cryogel scaffolds identical to sample3, to investigate the effect of chemical and physical scaffold characteristics on 3D cell culturing [30]. The metabolic activity of the cells was monitored on day 0, 7, 14 and 21. Starting from the seeded cells (2.5*105 cells in 20 μl cell culture medium) at t = 0 days, the fold change of proliferation was calculated at each time step by using the following equation: (Nt − N0)/N0, where Nt is the number of cells measured at the time t and N0 is the initial

Fig. 8. Fold change of proliferation of LNCaP cells vs culturing time. The fold change of proliferation was calculated relatively to the initial number of cells at day 0 (2.5 ∗ 105 cells in 20 μl cell culture medium).2 2 All quantitative measurements were performed on at least triplicates. All values are expressed as means ± standard deviations (SD). One-way ANOVA with post hoc Tukey tests [22] were used to compare the quantitative measurements at the different culture time points relatively to the results at t = 7 days and p b 0.05 was used to assess statistical significance (p N 0.05 not significant; p b 0.0001 ***; p b 0.001 **; p b 0.05 *).


471

The analysis of the reconstructed tomographic slices demonstrated the development of PCa cell aggregates inside the cryogel compartments, with the cell assemblies deriving their characteristic granular appearance from the multi-cellular tumor spheroids that compose each ensemble (Fig. 10a). The 3D visualization of cells is particularly relevant for the development of in vitro cryogel scaffolds applicable to PCa culturing. In fact, the tomography volumes allow to quantitatively assess the 3D structure of PCa cell aggregates at the different culturing times and consequently to access the cell proliferation rate. In addition, SXCμT can be applied to investigate the 3D development of the PCa tumor agglomerates in particular applications, e.g. in presence of hormonal signaling molecules, thus contributing to model prostatic disorders to optimize a therapy against them. 4. Conclusions and outlook Fig. 9. Spheroids distribution in sample3.

number of cells [30]. The study of the PCa cell proliferation is depicted in Fig. 8 and indicates an increase of the cell population by a factor of 9 after culturing for 21 days. Such a long term cultivation time is another advantage of cryogel scaffolds with respect to conventional 2D cell culturing, where the flask (/dish) surface is almost completely covered by the cells after 4–5 culturing days [41]. To investigate the presence of hypoxic cores inside the proliferated PCa cells, an optical microscopy study was performed on selected slices extracted from the top, middle and bottom of a cryogel undergoing culturing for 21 days. The total number of spheroids with and without hypoxic cores was quantified by post-processing the images with the module “MaterialStatistics” of the AMIRA 5.6 3D Software [27]. As shown in Fig. 9, the total number of cell spheroids decreases from the top over the middle to the bottom position of the cryogel, due to the cell seeding process from the top of the scaffold. In the same way, also the number of spheroids with hypoxic cores is higher on the top position, where the higher nutrient and oxygen supply favors the spheroids growth, thus enhancing the probability that hypoxic cores are formed. However, the number of hypoxic cores represents only a small fraction of the total number of spheroids (~8%) [27]. Further, one of the cryogel was also used to nondestructively investigate the cell proliferation by means of SXCμT. After 21 days of culturing, the PCa cells were fixed with glutharaldehyde solution (2.5% v/v) and dried prior to tomographic scans. The presence of PCa cell groups in the tomographic slices (Fig. 10a) was confirmed by the comparison with high spatial resolution optical microscopy images, acquired on a cryogel scaffold identical to sample3 and equally used for culturing the PCa cells (Fig. 10b).

In conclusion, three cryogels scaffolds, namely Gelatin-Fibrinogen, Chitosan-Agarose-Gelatin and Poly(2-hydroxyethyl methacrylate)-Alginate-Gelatin, developed for PCa cell culturing have been investigated by means of SXCμT. These measurements were complemented with mechanical tests to evaluate the elastic modulus and the water uptake capacity of each sample. It turned out that the cryogel based on Gelatin-Fibrinogen lacks pore interconnections, which are demanded to guarantee nutrient proliferation and thus cell survival. As such, the system would just represent a low porous surface for culturing cells as conventional 2D monolayer. The architectural properties of the other two samples demonstrate that the cryogel based on Poly(2-hydroxyethyl methacrylate)-Alginate-Gelatin is the most suitable candidate to be used as scaffold for cell culturing thanks to its smaller, but sufficiently large pore sizes and their high interconnectivity. Both the properties enable a more reliable cell attachment and thus cell survival. SXCμT was also used to investigate the presence of multi-cellular tumor spheroids that were cultured inside the optimal cryogel composition. Our results show that several tumor aggregates developed in the scaffold, demonstrating the scaffold usefulness in studying the progression and behavior of prostate cancer cells. We believe that our investigation is particularly relevant for the development of in vitro cryogel scaffolds: it permits a nondestructive monitoring of the 3D development of PCa cells in presence of certain signaling molecules to mimic the endocrine environment in which the cells proliferate. Based on the present study, further investigations focusing on long-term cultivation of LNCaP cells as well as other PCa cell lines (e.g. PC3, VCaP) can be envisaged. Beside the analysis of spheroid formation and distribution within the scaffold, a treatment of the cells with dihydrotestosterone (DHT) is feasible additionally. The Androgen

Fig. 10. (a) X-ray tomography reconstructed slice of a cryogel scaffold (sample3) used for culturing PCa cells for 21 days and (b) high resolution optical microscope image of a cryogel containing PCa cells (purple aggregates) cultured under the same experimental conditions.

472


signaling via androgen receptor (AR) plays an important role and leads to an increase of cellular proliferation of androgen sensitive PCa cell lines (e.g. LNCaP) [42,43]. In this manner, it can mimic more closely the physiological environment. Furthermore, the dispersion of the DHT across the characteristic pore structure and its effect on the spheroid growth can be monitored non-destructively by SXCμT. All this provides for an optimal toolbox for prostate cancer research, and in turn, implies an environment suitable to drug testing and screening. References [1] R. Edmondson, J. Jenkins Broglie, A.F. Adcock, L. Yang, Assay and Drug Dev. Technologies 12 (4) (2014) 207–218. [2] F. Pampaloni, E.G. Reynaud, E.H.K. Stelzer, Mol. Cell. Biol. 8 (2007) 839–845. [3] M. Ravi, V. Paramesh, S.R. Kaviya, E. Anuradha, F.D.P. Solomon, J. Cell. Physiol. 230 (2014) 16–26. [4] V.I. Lozinsky, I.Y. Galaev, Trends Biotechnol. 21 (10) (2003) 445–451. [5] A. Tripathi, A. Kumar, Macromol. Biosci. 11 (2011) 22–35. [6] S. Rhee, Exp. Mol. Med. 41 (2009) 858–865. [7] H. Kirsebom, D. Topgaard, I.Y. Galaev, B. Mattiasson, Langmuir 26 (2010) 16129–16133. [8] T. Yeung, P.C. Georges, L.A. Flanagan, B. Marg, M. Ortiz, M. Funaki, N. Zahir, W. Ming, V. Weaver, P.A. Janmey, Cell Motil. Cytoskeleton 60 (2005) 24–34. [9] A. Barbetta, R. Bedini, R. Pecci, M. Dentini, Ann. Ist. Super. Sanita 48 (2012) 10–18. [10] I.N. Savina, V. Cnudde, S. D'Hollander, L. Van Hoorebeke, B. Mattiasson, I.Y. Galaev, F.D. Prez, Soft Matter 3 (2007) 1176–1184. [11] N. Sampson, H. Neuwirt, M. Puhr, H. Klocker, I.E. Eder, Endocr. Related Cancer 20 (2013) R49–R64. [12] J.M. Martínez-Jabaloyas, J.A. March-Villalba, M.M. Navarro-García, F. Dasi, Expert. Opin. Biol. Ther. 13 (2013) 1–5. [13] R. Siegel, J. Ma, Z. Zou, A. Jemal, CA Cancer J. Clin. 64 (2014) 9–29. [14] K. Boehm, G. Salomon, B. Beyer, J. Schiffmann, K. Somonis, M. Graefen, L. Budaeus, J. Urol. 193 (2015) 784–800. [15] J.M. Tyszka, A.J. Ewald, J.B. Wallingford, S.E. Fraser, Developmental Dynamics 235(2): spc1, Cover Image, and Published Online 4, May 2006. [16] L.F. Drummya, J. Yangb, D.C. Martin, Ultramicroscopy 99 (2004) 247–256. [17] M.B. Dainiak, I.U. Allan, I.N. Savina, L. Cornelio, E.S. James, S.L. James, S.V.J. SL, S.V. Mikhalovsky, H. Jungvid, I.Y. Galaev, Biomaterials 31 (2010) 67–76. [18] S. Bhat, A. Tripathi, A. Kumar, J.R. Soc, Interface 8 (2011) 540–554.

[19] D. Singh, S.M. Zo, A. Kumar, S.S. Han, J. Biomater. Sci. Polym. Ed. 24 (2013) 1343–1359. [20] N. Goodstone, L. Muller, A. Cartwright, Tissue Eng. 10 (3–4) (2004) 621–631. [21] A.A. Karim, R. Bath, Food Hydrocoll. 23 (2009) 563–576. [22] L.S. Kao, C.E. Green, J. Surg. Res. 144 (1) (2008) 158–170 (January). [23] S. Farris, J. Song, Q. Huang, J. Agric. Food Chem. 58 (2010) 998–1003. [24] A. Bigi, G. Cojazzi, S. Panzavolta, K. Rubini, N. Roveri, Biomaterials 22 (2001) 763–768. [25] www.sigmaaldrich.com, last visit September 2016. [26] I. Levental, P. Georges, P. Janmey, Soft Matter 3 (2007) 299–306. [27] A. Baecker, PhD thesis (in preparation). [28] J.S. Horoszewicz, S.S. Leong, E. Kawinski, J.P. Karr, H. Rosenthal, T.M. Chu, E.A. Mirand, G.P. Murphy, Cancer Res. 43 (4) (1983) 1809–1818. [29] C. C. Fischbach, R. Chen, T. Matsumoto, T. Schmelzle, J.S. Brugge, P.J. Polverini, J.D. Mooney, Nat. Methods 4 (2007) 855–860. [30] A. Bäcker, B. Göppert, S. Sturm, P. Abaffy, T. Sollich, F.J. Gruhl, SpringerPlus 5 (902) (2016) 1–12. [31] A. Rack, T. Weitkamp, S. Bauer, P. Modregger, A. Cecilia, T. Santos Rolo dos, T. Rack, D. Haas, R. Simon, R. Heldele, M. Schulz, B. Mayzel, A.N. Danilewsky, T. Waterstradt, W. Diete, H. Riesemeier, B.R. Müller, NIM B 267 (2009) (1978-1988). [32] A. Douissard, A. Cecilia, X. Rochet, X. Chapel, T. Martin, T. van de Kamp, L. Helfen, T. Baumbach, L. Luquot, X. Xiao, J. Meinhardt, A. Rack, J. Instrum. 7 (2012) 1–26. [33] T. Weitkamp, D. Haas, D. Wegrzynek, A. Rack, J. Synchrotron Radiat. 18 (2011) 617–629. [34] http://www.optiquepeter.com/fr/index.php, last visit February 2016. [35] J. Moosmann, V. Altapova, L. Helfen, D. Hänschke, R. Hofmann, T. Baumbach, J. Phys. Conf. Ser. 425 (2013) 1–5. [36] R. Hofmann, A. Schober, S. Hahn, J. Moosmann, J. Kashef, M. Hertel, V. Weinhardt, D. Hänschke, L. Helfen, I.A. Sánchez Salazar, J.-P. Guigay, X. Xiao, T. Baumbach, Opt. Express 24 (2016) 4331–4348. [37] M. Vogelgesang, T. Farago, T. F. Morgeneyer, L. Helfen, T. dos Santos Rolo, A. Myagotin and T. Baumbach, J. Synchrotron Radiat. [38] E.A. Naumenko, I.D. Guryanov, R. Yendluri, Y.M. Lvova, R.F. Fakhrullin, Nanoscale 8 (2016) 7257–7271. [39] M. Zanoni, F. Piccinini, C. Arienti, A. Zamagni, S. Santi, R. Polico, A. Bevilacqua, A. Tesei, Nature -Scientific Reports 6 (2016) 1–11. [40] Y. Peng, Y. Jiang, L. Eisengart, M.A. Healy, F.H. Straus, X.J. Yang, J. Pathol. Inform. 2 (2016) 33. [41] http://docs.abcam.com/pdf/protocols/Cell_Culture.pdf. [42] E. Tan, J. Li, H.E. Xu, K. Melcher, E. Yong, Acta Pharmacol. Sin. 36 (2015) 3–23. [43] C. Lee, D.M. Sutkowski, J.A. Sensibar, D. Zelner, I. Kim, I. Amsel, N. Shaw, G.S. Prins, J.M. Kozlowski, Endocrinology 136 (2) (1995) 796–803.