Adhesion and morphology of mammalian cells on nanoporous and nonporous spherical carbon substrates

Three spherical activated carbons (SACs) were used as substrates for mammalian cell proliferation. SACs were obtained by carbonizing styrene-co-divinylbenzene ion exchangers 35WET, XAD4, or 1200H. The new materials (XAD_C, WET_C, and H_C) were characterized by adsorption–desorption nitrogen isotherms and mercury intrusion porosimetry. XAD_C and WET_C exhibited well-developed BET surface areas, similar total pore volumes, and highly different pore size distributions. H_C was nonporous spherical material—reference material. The XAD_C was meso-macroporous, but the WET_C was micro-mesoporous. All SACs were not cytotoxic toward Leydig TM3 cells. The differences in porous structure and morphology of the carbon scaffolds led to morphological differences in adhered cells. The monolayer of cells was distributed flat over the entire WET_C and H_C surfaces. Leydig TM3 cells adhered to nonporous SAC but were easily washed out due to weak adhesion. The cells adhered in clusters to XAD_C and proliferated in clusters. As microscopic techniques and viability tests demonstrated, only nanoporous carbons provided a good surface for the attachment and proliferation of eukaryotic cells.


Introduction
The porous carbons have unique physicochemical properties, making them useful for industrial, technical, and medical applications [1].These materials are efficient in energy storage (supercapacitors, batteries, or hydrogen tanks), energy conversion (fuel and solar cells) [2], sensors, and environmental engineering (adsorption of sulfur oxide and nitrogen oxides, or water purification) [3][4][5].Significantly, they can be eventually used in clinics as biocompatible scaffolds [6] or as an efficient support for ECM synthesis [7].However, according to Fuhrer and colleagues [8], carbon structures have not yet been insightfully explored in biomedicine.
Activated carbons are produced from various precursors by carbonization and activation.These precursors can be of natural origins such as hard coal, brown coal, peat, wood, and fruit stones [12] or be polymeric substrates such as phenolformaldehyde resins [13,14], furfural [15], poly(ethylene terephthalate) [16], and ion exchange resins [17].The polymeric precursors of activated carbon microparticles have several advantages, but the crucial one is the high purity of the final product [1].
The highly porous activated carbons with their large surface areas were efficiently applied for extracorporeal blood purification by hemoperfusion [18].These materials were noncytotoxic, as it was demonstrated on the V79 cell line (the cell line from Chinese hamster lungs).
Activated carbons served as scaffolds for successful human fetal osteoblast cells (hFOB 1.19) [6] proliferation.In 2013, de Araújo Farias and colleagues used activated carbon fibers for an umbilical cord stromal stem cell culture [6,19].They observed the spontaneous differentiation of stem cells into osteocytes and chondrocytes without any specific medium supplements.Crowder et al [20] also reported that threedimensional graphene foams were suitable substrates for the adhesion and differentiation of human mesenchymal stem cells into osteocytes.Activated carbons were also applied as scaffolds for mineralization, enhancing bone tissue growth, and enabling easy and permanent deposition of the extracellular matrix (ECM) [21].
Efficient mammalian cell adhesion to the scaffold surface is a prerequisite for tissue engineering.Interactions between cells and the adhesion substrate depend on surface energy, morphology, functionality, and stiffness [22,23].The presence of extracellular matrix proteins preloaded before cell adhesion or synthesized by the adhering cells is also favorable [24].Many mammalian cells prefer substrate surfaces with relatively disordered morphology [25].Cell contractility and adhesion strength govern cell interactions with a curved surface.However, significant cell contraction or excessive curvature of the material surface may lead to a lack of cell adhesion [25].
One crucial factor in the cell adhesion process can be surface nanoporosity.Bello et al used nanoporous and polished titanium discs for Chinese hamster ovary (CHO-K1) cell culture [26].They observed the stimulation of cell spreading through the formation of numerous focal adhesions and filopodia on the nanoporous titanium surface.Thick parallel actin filaments in the cell cytoskeleton were also formed, suggesting the proper biomechanical adaptation of the cell on the nanostructured surface.
A well-controlled nanoporosity also characterizes mesoporous-activated carbon microparticles.Their nanoporosity can be easily controlled by chemical [27] or physical [28] activation processes or precursor modification upon synthesis [6,29].Therefore, our research aimed to synthesize activated carbon microparticles of various porosities as scaffolds for the growth of mammalian cells and for the storage of their secretions.
Leydig TM3 cells were chosen as mammalian cells for our study's culture on the scaffold surface.Leydig TM3 cells are glandular cells of the interstitial tissue and produce testosterone [30,31].We presumed the additional function of spherical carbon microparticles as substance storage, i.e. testosterone, apart from being a biocompatible scaffold [32].Retaining hormones inside a porous carbon structure holds promise for future infertility treatments.
In this research, we synthesized activated carbons from polymers using a novel method.The spherical activated carbons (SACs) were prepared from raw ion exchangers (Amberlite XAD4, Amberlyst 35 WET, and Amberjet 1200H) via a simple heat treatment followed by washing in a less expensive production process than the soft or hard templating methods.It allowed for a reduction in the cost of their synthesis and the number of potentially toxic substances introduced into the cell culture.Therefore, it was in line with the green chemistry approach in biomaterial development.Interestingly, the heat treatment of ion exchangers afforded the formation of changes in pores in the resulting carbons.Nevertheless, the synthesis procedure afforded SAC samples with a high specific surface area of 828 m 2 •g −1 .The shape, structure (morphology), and nanoporosity depended on the precursors' original character despite using the same copolymer styrene-divinylbenzene.Potentially, the availability of different types of commercial ion exchangers will allow us to obtain a variety of porous or/and nanoporous carbon structures in the future.As far as we know, this study is the first time to evaluate the different effects of nanoporosity versus porosity on cell adhesion.

Methods
2.2.1.Samples preparation and synthesis of SACs.The microparticles were the SACs.SACs were prepared according to Choma et al [27].First, 10 g of raw ion exchangers (Amberlite XAD4, Amberlyst 35 WET, and Amberjet 1200H) were treated with 50% orthophosphoric acid to increase flame retardancy [28].Polymeric spheres were dried at 100 °C.The polymer was pyrolyzed with a nitrogen atmosphere to the spherical carbon material using the temperature ramp: 100 °C for 15 min, 300 °C for 15 min, and 500 °C for 45 min.The carbon spheres were washed using two different mixtures, one by one, in a Soxhlet apparatus.The first solution was ∼35% hydrochloric acid/ ethanol (0.5 V/V), and the second was a mixture of water and ethanol (1:1 V/V).The SACs suspensions were adjusted to neutral pH and dried at 90 °C.Purified SACs were denoted as XAD_C when produced from XAD4, WET_C-from 35WET, and X_C-from 1200H.All SACs were sieved through 300-micrometer pore sieves.Finally, SACs were autoclaved, as described below.
2.2.2.Visualization of the SACs' morphology.The micrographs of SACs were taken with a scanning electron microscope STEM (FEI, Quanta FEG250).The SACs' surfaces were observed using a high vacuum mode with the Everhart-Thornley detector (ETD).Furthermore, the SACs' shape and diameter dimensions were imaged under an inverted optical microscope (Primo Vert, ZEISS).The diameters of carbon materials were measured using ZEISS ZEN software (Carl Zeiss Ltd).

The porosity of SACs evaluated by mercury intrusion.
The SACs' porosity was evaluated with the Micromeritics' AutoPore IV 9500 apparatus.Before adsorption measurements, the samples were preliminarily degassed at 110 °C for 1 h and then at 25 °C for 0.5 h to a pressure of 50 μm Hg.The porosity of materials was characterized by the increased pressure in the range of 0.05-414 MPa posed to the sample immersed in mercury (the non-wetting liquid).The method forced mercury to enter the material's pores.This phenomenon, described by the Washburn equation (1), allowed the calculation of the critical diameter of the capillary in which mercury could penetrate under a certain pressure, according to the equation: where r is an effective pore radius (capillaries with an assumed cylindrical shape) filled with mercury under a pressure p, σ is a surface tension at the mercury-adsorbent interface (usually a value of 0.480 N•m −1 is assumed for carbon materials), and θ-the contact angle of the solid surface for mercury (the most common value for carbon adsorbents is 2,478 rad), Δp is the pressure needed for mercury to penetrate into the pores.The Thermo Fisher Scientific UV-vis Evolution 220 spectrophotometer was used to measure the phenylaniline concentration's changes before and after exposure (adsorption) to SACs.The SACs at concentrations of 2.0, 3.75, and 15 mg •ml −1 were incubated in the solution of phenylalanine dissolved at a concentration of 1 mg ml −1 in water.The samples were incubated at room temperature for 24 h in the dark.After centrifugation, the supernatants at a volume of 4 ml were used for spectral measurements of phenylalanine content in a quartz cuvette at a wavelength of 260 nm.The phenylalanine concentration was read from the calibration curve, previously obtained by the dilution method.This curve was characterized by a coefficient of determination R 2 equal to 0.9999.The adsorbed phenylalanine was calculated by comparing (the difference) the amino acid concentration in the stock solution and the supernatant after 24 h of incubation and expressed as a percentage.
Then, the medium was removed from the wells.TM3 cell suspensions in DMEM/F12 were added to these wells (containing XAD_C, WET_C, or H_C carbon spheres) at a density of 5•10 5 ml −1 per well and incubated for 24 h at 37 °C, in an atmosphere of 5% CO 2 , at 90% humidity.Then, SACs with the TM3 cells attached to the surface were visualized by SEM and optical microscopy and used in metabolic and viability assays.
2.2.8.Visualization of the cell adhesion to SACs.SACs with adherent Leydig TM3 cells were processed according to Osuchowska et al [35].The samples were fixed in 4% paraformaldehyde and 0.4% glutaraldehyde solution for 30 min.Then, the samples were treated with 1% osmium tetroxide and incubated at 4 °C for 16 h on a plate sealed with parafilm.After this time, the samples underwent the dehydration treatment with successive water/ethanol solutions of increasing ethanol concentrations: 30%, 50%, 70%, 90%, 96%, and 99%, followed by increasing acetone concentrations from 30% to 100%.Afterward, the SACs with the adherent cells were dried in a critical point dryer (Leica EM CPD300, Germany) and covered by platinum in a sputter coater (Leica EM ACE200, Germany).The SACs' surfaces and TM3 cell morphology were observed using scanning electron microscopy with high vacuum mode and the ETD detector.
2.2.9.The viability of adherent Leydig TM3 cells.The viability of TM3 cells on the individual SAC's surface was evaluated with the LIVE/DEAD ® Viability/Cytotoxicity Kit (Molecular Probes, Life Technologies Corp., Oregon, USA).
The assay was performed according to the manufacturer's instructions.Carbon spheres covered with cells were gently transferred to a new dish to show only the cells attached to them and not to the bottom of the dish.After photographs, they were rinsed three times with warm PBS to remove poorly attached cells and re-imaged.The extensive lighting system Laser Scanning Confocal System LSM 700 microscope Zeiss Axio Observer.Z1 (CSLM) was used for cell visualization.
The evaluation was conducted at two excitation wavelengths: 488 nm and 555 nm, respectively, for calcein-AM and ethidium homodimer-1.
2.2.10.MTT assay.The metabolic activity of TM3 cells was determined using methyl-thiazolyl diphenyl-tetrazolium bromide In this study, cellular metabolic activity was used to indicate the TM3 cells' density and viability.The assay was performed according to the manufacturer's instructions.After 24 h incubations, all types of SACs were transferred to the 2 ml Eppendorf tubes.MTT assays were carried out in the wells after removing the SACs to the 2 ml Eppendorf tubes and, simultaneously, in the tubes.Therefore, the metabolic activity of the cells in the wells and the cells' adhesion to the SAC's surface in the Eppendorf tubes were measured.Absorbance was measured with a microplate reader CLARIOstar (BMG LABTECH) at 550 nm.The control samples constituted TM3 cells seeded on the surface of the well at density 5•10 5 ml −1 .The results were normalized and expressed as a percentage of the average OD value recorded for the TM3 cells adherent to the well's surface and SAC's surface.The assays were performed in four repetitions.

Imaging of SACs
The XAD_C spheres were corrugated, and their diameters were reduced (on average) by 138 μm compared to the manufacturer's data (https://sigmaaldrich.com/PL/pl/-product/sigma/xad4#), accessed on November 11th, 2022), and finally, the mean diameter of XAD_C was equal to 407 ± 69.4 μm.In turn, the carbonization process of the WET_C precursor did not destroy the spherical shape of the carbon spheres despite many production and purification stages.The surface of WET_C was smooth after the complete process of synthesis.The precursor sphere's mean diameter was 487 ± 69.6 μm (https://lenntech.com/Data-sheets/Dow-Amberlyst- Figure 1(A) shows the histogram of all SACs sphere diameters and the optical microscope images.The sphere size distribution was similar for all SACs; the widest curve had XAD_C (figure 1(B)).Second was WET_C (figure 1(B)).The most stable dimension had H_C (figure 1(D)).
Figure 2 presents the surface morphology of XAD_C, WET_C, and H_C under a scanning electron microscope.Carbonization resulted in the shrinkage and creasing of spheres, especially in XAD_C (figure 2(C)).The WET_C surface was smoother than XAD_C; only at higher magnifications, depressions, dislocations, and pores were observed due to its well-developed nanosurface (figures 2(E) and (F)).The H_C surfaces were smooth, even at high magnifications (figures 2(H) and (I)).The few deficiencies in the continuity of the carbon structure could have been caused by the cleaning process, pouring, etc.

Characterization of SACs by mercury porosimetry
Mercury porosimetry supplemented the gaseous adsorption method presented below.Mercury (the non-moisturizing liquid) is injected into the material's pores at a specific pressure.Therefore, the mercury was forced into the smallest pores at different pressures, and the Washburn equation was used to calculate the relationships.
The analysis showed a wide range of pores, even up to the size of ten microns.WET_C had a particularly welldeveloped microporosity (in the range of up to 2 nm) and narrow mesoporosity (in the range of 2-50 nm), which correlated well with the results determined with nitrogen  adsorption-desorption isotherms (figure 3(D)).In turn, XAD_C has a reasonably evenly developed porosity throughout the entire range, including the pores sized one to ten micrometers.The H_C material could not be evaluated by this method due to the lack of nanoporosity.Based on the data obtained (table 1), one could hypothesize that the SACs' pore structure differences might influence the cells' adhesion and proliferation.

Characterization of the SAC's porosity by nitrogen adsorption isotherms
The porous structure of XAD4, 35WET was shrunken after exposure to high temperatures upon the production of SACs.The surface of 1200H after the same high temperature) treatment was smooth.Two techniques were used to define the nanoporosity of SACs in various dimensions (ranges), i.e. nitrogen adsorption-desorption isotherms and mercury porosimetry.Figure 4(A) presents the low-temperature nitrogen adsorption-desorption isotherms measured for XAD_C, WET_C, and H_C.The porous structure of XAD_C and WET_C were of type IVA, according to the recommended IUPAC classification [36].
The synthesized carbons acquire significantly different porosity; the highest specific surface area (SSA) of 828 m 2 •g −1 exhibited WET_C carbon derived from 35WET resin, even though the SSA of the starting polymer was as low as 40 m 2 •g −1 .Carbon denoted as XAD_C has about twice lower SSA (428 m 2 •g −1 ) than its polymeric precursor (990 m 2 •g −1 ), whereas both samples, the starting polymer 1200H and the derived carbon H_C possessed rather low porosities; their SSAs were below 100 m 2 •g −1 .The low porosity of H_C (the carbon can be considered non-porous) resulted from cracks and discontinuities formed under exposure to high temperatures.Thus, it was characterized by minimal nitrogen adsorption.In this case, the porosity of the polymer precursor was practically not preserved.The precursor pores were partly preserved in XAD_C carbon.Interestingly, the heat treatment of the 35WET ion exchanger afforded the formation of abundant additional pores in the resulting carbon.The pore number, size, and shape depended on the precursors' original character and the synthesis and purification method.
Figures 4(B)-(D) present the pore size distributions (PSDs) of three activated carbon microparticles in three range scales (from nanometers to micrometers).The PSD curve of WET_C represented two distinct bands with various pore dimensions (figure 4(B)).This material had better-developed micropores (ultramicropores and small supermicropores, mainly), as depicted by the first peak position.The second band was the broad peak of large mesopores and small macropores in the 5-90 nm range.XAD_C exhibited a negligible peak in the 0-2 nm range (figure 4(D)), representing micropores and narrow mesopores.XAD_C was further characterized by a broad band of mesopores in the 2-60 nm range, formed after the original polymeric precursor nanostructure collapsed.XAD_C was rather mesoporous with a minimal proportion of micropores.The PSD curve of H_C is a flat line in the nanometer dimension, proving the lack of a developed porous structure.In turn, the nanoporosity of both materials (XAD_C and WET_C) was quite developed in quite similar dimensions.
Table 2 collects structural parameters based on the nitrogen adsorption-desorption data determined for carbon materials and their precursors.As can be seen, WET_C carbon had a better-developed microporosity than XAD_C, H_C, and all precursors.The specific surface area (S BET ) was 427.6 m 2 •g −1 for XAD_C, 827.7 m 2 •g −1 for WET_C, and 57.5 m 2 •g −1 for H_C.These materials' total pore volumes (V T ) were equal to 0.59 cm 3 •g −1 for XAD_C, 0.76 cm 3 •g −1 for WET_C, and 0,04 cm 3 •g −1 for H_C.WET_C had a better-developed microporosity than XAD_C, H_C, and all precursors.Whereas XAD_C had a better-developed mesoporosity than WET_C; the respective values were equal to 61% and 38%.It can be summarized that XAD_C was a mesoporous material while WET_C was a micro-mesoporous.The lack of porosity in H_C enabled its use as a reference material for the adhesion of eukaryotic cells.

Phenylalanine accumulation on SACs
Phenylalanine was used as a testosterone surrogate to determine the level of bioactive molecule adsorption on SACs.This amino acid was chosen because it has good solubility in the water phase.It is a nonpolar and hydrophobic amino acid, like the testosterone molecule.Moreover, it has only one heteroatom (nitrogen) and no substituents in the aromatic ring.It is also a DMEM/F12 medium component, which could adsorb on SACs. Figure 5   tested (2.0 mg•ml −1 , 3.75 mg•ml −1 , 15.0 mg•ml −1 ), the adsorbed phenylalanine concentration was several times higher on WET_C than on XAD_C and H_C.WET_C adsorbed at least six times more phenylalanine than XAD_C or H_C (max.20 times more phenylalanine adsorbed in WET_C in 2.0 mg •ml −1 carbon concentration).The adsorption data of phenylalanine on XAD_C, WET_C, and H_C is shown in table 3.
Additionally, adsorption capacities for all SAC materials were calculated relative to 1 g of carbon adsorbent.The qualitative results presented are the average values of all concentrations for the given SAC used in the measurements.Table 1.Structural parameters calculated by analysis of mercury adsorption for XAD_C and WET_C.For each SAC, the results were measured in triplicate at every out of the three concentrations.Average values of SACs' adsorption capacities (C a ) relative to phenylalanine were as follows: XAD_C C a = 6,1 ± 1,70 mg•g −1 (C) , WET_C C a = 72,3 ± 19,63 mg•g −1 (C) , and H_C C a = 6,4 ± 0,87 mg•g −1 (C) .

Leydig TM3 cell viability by confocal microscopy
Leydig TM3 cells attached to SACs and the bottom of culture wells were stained with the LIVE/DEAD™ Viability/Cytotoxicity Kit and observed under a confocal microscope 24 h after.No differences in the viability between the cells that adhered to the carbon surface and those grown in the culture wells were observed under a confocal microscope (data not shown).The vast majority of Leydig TM3 cells on the SACs' surface were alive, as shown in the Z-stack photographs (figure 6).Most cells showed no membrane damage and green fluorescence, and only a few showed red fluorescence suggesting membrane discontinuity.TM3 cells adhered in a similar number to the surfaces of all SACs.One also might pay attention to the dense packing of cells adjacent to the surface.It could be due to the efficient cell adhesion or their divisions, as Leydig TM3 cells divide every 12 h.After washing, we observed a slight decrease in the number of cells on XAD_C and WET_C, while almost all cells were washed off from the H_C carbon (figures 6(D)-(F)).

Morphology of adherent Leydig TM3 cells
The morphology of TM3 cells attached to SACs was visualized several times through repeated experiments.The randomly selected XAD_C, WET_C, and H_C surfaces with attached TM3 cells are shown in figure 7.No damage to the Leydig TM3 cells was observed.The cellular membrane was not perturbed, and numerous microvilli on the cell surface were retained and visible.However, the morphology, size, and three-dimensional arrangement of TM3 cells on XAD_C compared to WET_C and H_C were different.These differences could be observed at higher magnifications (figures 7(A3), (B3), and (C3)).TM3 cells attached to XAD_C spheres (figures 7(A1)-(A3)) were elongated with numerous cell-to-cell connections.The cells also formed several agglomerates (clusters) (figures 7(A2) and (A3)).In turn, TM3 cells on the smooth surface of WET_C V micro -the volume of micropores calculated by integration of PSD up to 2.0 nm; e V meso -the volume of mesopores-the difference between V t and V micro ; f Mesoporosity-the ratio of V meso to V t expressed in percent.surface of the WET_C spheres formed cell monolayers but not clusters, as could be observed for XAD_C and H_C.In turn, TM3 cells mostly adhered to the craters, holes, and dislocations, which interrupted the continuity of the H_C surface.Smooth surfaces of H_C without the developed nanopores prevented the anchoring of Leydig TM3 cells.

Leydig TM3 cell metabolic activity on SACs
The MTT assay is a colorimetric method for measuring cell metabolic activity and viability.The metabolic activity of cells adhered to SACs (figure 8(B)) and the bottom of the wells (figure 8(A)), in which the incubation was carried out, was measured with the MTT assay.The results in figure 8(B) presented that TM3 cells were active metabolically on the SACs' surface.This finding confirmed the observations made under the confocal microscope after the cell staining with the LIVE/DEAD™ Viability/Cytotoxicity kit.The highest cellular metabolic activity of the cells was recorded for XAD_C and WET_C at concentrations equal to 2.00 and 3.75 mg •ml −1 .Along with the increase in SAC's concentration above 7.50 mg• ml −1 , the metabolic activity of Leydig TM3 cells decreases both on the SACs' and well surface.The decrease in cell metabolic activity is evident for H_C microparticles (figure 8(A)).Furthermore, significant differences in the metabolic activity of TM3 cells adhered to XAD_C and WET_C unlike H_Cs' surface were observed (figure 8(B)).For TM3 cells that adhered to XAD_C and WET_C, OD550 values ranged from 40% to 90%, and for H_C 5%-9% compared to control samples.

Discussion
Several publications have recently pointed out the functionality of nanoporous structures and their potential use in biomedical applications [26,[37][38][39][40].However, the efficiency of mammalian cell adhesion to activated carbons has yet to be sufficiently elaborated.Similarly, some doubts exist about whether these materials are not toxic to mammalian cells.These problems partly stem from microplate assays commonly used in biomedical laboratories to determine the cytotoxicity of chemical compounds.The mesoporous carbon materials readily adsorb the dyes used in such assays, thus distorting the interpretation of the results [18].To gain knowledge of these phenomena, this work conducted the synthesis and characteristics of SACs with different nanoporous structures for biomedical applications, i.e. mammalian cell in vitro culture.The idea was that carbonaceous porous 3D materials could be suitable as scaffolds due to their porosity and the fact that the porosity can be easily managed and monitored during production [6,29].Three different SACs were investigated, and Leydig TM3 cells were chosen as a biological research subject.
All SACs analyzed in this study had significantly different morphology and porosity.The differences in nanoporosity parameters of the three materials were depicted from nitrogen adsorption-desorption isotherms and mercury porosimetry data.These types of analysis are a standard approach for characterizing different materials with a developed specific surface in the nano and micro-range [7,41,42].The applied measurement methods can estimate pore sizes [1,9,27,28,33] and provide information on the potential adsorption capability of high molecular mass compounds from the liquid phase [1,3,4,43,44].The analysis of nitrogen adsorption-desorption isotherms in this study demonstrated the nanoporosity of WET_C, while mercury porosimetry showed that the porosity of XAD_C is of a broader range-above 100 nm.However, the drawback of mercury porosimetry is that the fractures could be defined as single pores with dimensions corresponding only to the fracture slit (a crevice).It means that only the total volumes of consecutive pore fractions could represent a source of reliable information [45], and the dimensions may be understated compared to microscopic analyses (morphology).Hence, that is why H_C was not evaluated by mercury porosimetry.For the characterization of the H_C porosity, the low-temperature nitrogen analysis was adopted.According to the IUPAC classification of nanoporosity, WET_C had narrow micropores and small mesopores located directly on the surface, whereas XAD_C had micropores, mesopores, and macropores on micro-scale and up to 10 μm.Carbon H_C as a nonporous material was used to compare the efficiency of cell adhesion on nanoporous and nonporous substrates.
Cai et al [46] demonstrated that nanosized pores are the closest to natural tissue morphology and positively affect cell adhesion by influencing collagen and ECM synthesis.In our study, WET_C and XAD_C were suitable substrates for TM3 Leydig cell adhesion.The detailed SEM observations showed that adherent TM3 cells had normal morphology with numerous microvilli on the surface.The adherent TM3 cells on the smooth surface of WET_C were flattened and grown as a monolayer.TM3 cells on XAD_C were more rounded and grown in clusters.It seemed that TM3 cells displayed stronger cell-to-cell interactions than cell-to-substrate interactions on this carbonaceous material.The glossy H_C surface displayed low cell-to-substrate interactions with TM3 cells.This type of surface topography caused cell culture degradation during SEM preparations.The cells were removed from the carbon surface due to washings.It proves the poor strength of adhesion, probably due to the need for more close contact (interaction) of microvilli with the carbonaceous surface.
This assumption was also confirmed by washing the carbon spheres three times with PBS and comparing the microscopic image before and after washing.Cell adhesion to XAD_C and WET_C was strong, and washing did not detach the cells from the substrate.Triple washing removed most of the cells from the surface of H_C.Poor adhesion of TM3 cells to the surface of this activated carbon makes the correct interpretation of images and viability tests difficult.
The cells were predominantly viable before and after washing; the few red dots marked cell nuclei with damaged cell membranes.However, it should be remembered that besides the natural cell death associated with the cell cycle, there is also physical contact of the spheres during preparation and imaging and the possibility of mechanical damage to the cells under the influence of friction.
Differences in the SAC's surface topography caused differences in the morphology of adherent cells on the surface of all materials; however, the adhesion process could also be influenced by other factors, such as the presence of adhesive proteins on the surface.It was shown by Chavez et al [47], that the increased presence of mesopores on the carbonaceous material surface significantly decreased MC3T3 cell (murine osteoblast cells) adhesion.This phenomenon was probably due to the intensive adsorption of adhesion promoters or nutrients in porous materials during cell growth.As a result, the cells could not proliferate efficiently on the mesoporous film substrate.
However, our results contradict Chavez and colleagues [47], because micro-and mesoporous SACs were suitable substrates for cell adhesion in this work.In our opinion, the possible explanation for these differences was the lack of a pre-saturation step in Chavez and colleagues' experiments.These carbon materials adsorbed the necessary nutrient components from the growth media.It led to lower metabolic activity and viability of the cells.In our experiments, the presaturation step lasted for 24 h.Therefore, all SACs were saturated with the media compounds and did not adsorb them from the environment during cells' experiments.
The efficiency of nutrient accumulation is vital for the growth and viability of the cells.The cells grown on WET_C demonstrate an increased cell-substrate surface contact due to the material morphology and more than twice the volume of micropores, which might provide more nutrients to the cells and promote their flattening.In contrast, XAD_C had broader and deeper pores, which was correlated with a lower concentration of medium compounds; thus, the nutrient availability was lower, and the cells grew rounded and in clusters.Moreover, the bridges between cells were observed, creating connections.Clusters of cells preferred to settle in macropores close to the sphere's surface instead of penetrating deep into the craters.The cells had a healthy morphology on the nonporous H_C material.It can be assumed that due to the topography of the H_C surface, which is very smooth and nonporous, the cells could grow in clusters and only in the cracks offering the possibility of the cell's anchoring.
Another crucial property of nanoporous carbons that enables medical application is their inertness, which provides these structures with good biocompatibility [25].All three carbonaceous SACs were non-cytotoxic.Leydig TM3 cells were alive on the entire surface of SACs after 24 h incubation, as it was visualized by confocal microscopy.It was confirmed by the quantitative MTT assay, where the metabolic activity of the adhered cells was shown.The maximum cell adhesion and their metabolic activity were observed at a certain cell-tocarbon proportion (at an SACs concentration of 3.75 mg ml −1 and the density of TM3 cells-5 × 10 5 ml).At higher SAC concentrations, the recorded OD values began to decline insignificantly.The SACs can efficiently adsorb dyes based on viability and metabolic assays.This effect becomes apparent at higher carbon concentrations because the cells occupy only a tiny fraction of the total area for adsorption.
Differences in structure and properties of nanoporous materials produce varying adsorption levels of the biologically active molecules on their surface [48][49][50].Nagy et al [51] used nanoporous carbon materials for hemoperfusion or as a medium for protein storage.Liu et al [52] investigated mesoporous carbon prepared by the inverse replica process to introduce sparingly soluble drugs (carvedilol), simultaneously monitoring their concentration and distribution pathways in vivo.The most resembling materials to those presented in this study were mesoporous carbon spheres synthesized by Wang et al [53] for the innovative blood-cleansing application [54].They successfully adsorbed various compounds of a broad spectrum of molecular weights (MW) and dimensions, such as creatinine (small MW), vitamin B 12 (medium MW), and α-chymotrypsin (large MW).The reports presented above encouraged us to consider the hypothesis on the SACs' adsorption of testosterone, which could be secreted by Leydig cells adjacent to the surface of spherical carbons.
To verify this hypothesis, we investigated the adsorption of phenylalanine from an aqueous solution by SACs.The WET_C adsorbed phenylalanine much more efficiently than XAD_C or H_C (20 times more at the carbon concentration of 2 mg ml −1 , 12 times-at 3.75 mg ml −1 , and six times-at 15 mg ml −1 ).The WET_C micropores are located directly on the surface of spheres; therefore, they can be better wetted with water.According to the adsorption theory, molecule deposition occurs first in the smallest pores [1,34].In this study, the SACs exhibited an adsorption capacity for phenylalanine in the range of about 6.1-72 mg g −1 .Compared to other carbon adsorbents described in the literature, such as activated meal sunflower [55] or activated defective coffee beans [56], the SACs' phenylalanine storage values were comparable.SACs fared worse than materials from activated corn cobs [57] or activated date stones [58].However, the materials described in [57,58] were also characterized by higher pore structure parameters, and, more importantly, they were activated with substances that could impair the in cell culture.The importance of adsorption and storage of biomolecules was presented in [59].As reported in the paper, spherical materials are promising candidates for selective and even separation of biomolecules.According to [59], the size of the pores and their total volume determine the rate of adsorption and capacity concerning biomolecules.Spherical carbon aerogels (SCAs) obtained by Long et al [59] were used in the adsorption of the following biomolecules: l-phenylalanine, vitamin B12, α-chymotrypsin, and bovine serum albumin (BSA).However, the latest paper in which phenylalanine adsorption by porous carbon material is described in [60] used activated coconut carbon to confirm the legitimacy of using this material as an alternative in amino acid therapy.This material can help to remove excess aromatic and heterocyclic amino acids in the treatment of i.e. phenylketonuria, kidney, and liver diseases.
Therefore, we can conclude that SACs adsorb efficiently small organic molecules at micropores, and their meso-and macroporosity augment the adhesion of mammalian cells.The synergy of these features and others described/demonstrated earlier in this paper suggests that SACs are noteworthy for use as scaffolds for cell transfer.Further, this work indicated how important each of the dimensions of porosity and nanoporosity is in cell culture.Concerning the work [60], our research, taking into account the proposed materials, convinces the development of SACs in new promising medical applications.
Currently, mesoporous or micro-mesoporous materials are obtained using hard or soft mapping methods [1,2,4,9].They require the usage of expensive and potentially toxic reagents for eukaryotic cells.The material synthesis presented in this work recycles polymer [1] and uses only a simple reagent, orthophosphoric acid, which can be easily washed out to neutralize the final material.Orthophosphoric acid is also a component of phosphate-buffered saline (PBS) used in cell culture.This is undoubtedly another argument confirming the legitimacy of using the described synthesis methodology of SACs.It is an economically reasonable [1,9] path to produce mesoporous carbon materials, which their synthesis still needs to be.
In this manuscript, we used two different SACs and their non-porous substitute of different porosity to study differences in mammalian cell morphology and adhesion.It has not been done yet.We also verified the hypothesis that the spherical surface porosity may affect cell physiology.The differences in the SACs' nanoporosity were shown with lowtemperature nitrogen adsorption-desorption isotherms, and mercury intrusion, and visualized under electron and confocal microscopes.Moreover, we used two methods to characterize the nanoporosity and describe carbon sphere dimensions since mistakes often appear in the literature, and roughness is confused with porosity.This was why we decided to use the word nanoporosity and not porosity in our work.
We also demonstrated for the first time that SACs appeared noncytotoxic toward mammalian cells but differed significantly as substrates for efficient mammalian cell adhesion.WET_C and XAD_C absorbed to a high extent amino acids and other molecules.We indicated that this phenomenon might introduce some perturbances to the commonly used assays in cell biology, i.e. cell proliferation assay, causing bias to the previously reported results.We discussed how to avoid these perturbances and provide the most reliable results on cell adhesion to porous biocompatible carbon materials, as mentioned earlier by Fuhrer et al [8].Despite the Bello et al publication [26], which we referred to in the discussion, we believe the nanoporous structure is essential for proper and efficient cell adhesion.In our opinion, the adsorption properties (the possibility of storage) are the indisputable advantage of this type of structure.However, it requires appropriate preparation (saturation).We do not support the claim of Bello et al [26] that better adhesion is achieved on flat smooth carbon surfaces.Our results, however, correlate well and confirm Peñalver et al [41] results.The growth of adherent cells is possible due to the concentration of nutrients and oxygen on the developed surface of the activated carbon.

Conclusions
This work demonstrated that SACs with high specific surface areas of up to 828 m 2 •g −1 were successfully synthesized from commercial ion exchangers via a simplified synthesis procedure based on heat treatment.The as-obtained porous spherical carbons could be used as scaffolds for in vitro culture of mammalian cells.However, some restrictions regarding these carbon forms have also been demonstrated upon studying their properties.These are materials with enormous sorption capabilities, which can be used, for example, to store products of cell metabolism, a phenomenon beneficial for biomedical applications.On the other hand, porous carbon binds the components of the medium, thus making it difficult to conclude the efficiency of cell proliferation.As this study shows, only the pre-saturation of activated carbons with the medium allows for controlled cultivation of cells on their surface.
Significant differences in the metabolic activity of TM3 cells adhered to highly porous carbons XAD_C and WET_C were observed compared to nonporous H_C carbon, which was used as a control sample.The highest cellular metabolic activity of the cells was recorded for XAD_C and WET_C at concentrations equal to 2.00, 3.75, and 7.50 mg •ml −1 .For TM3 cells that adhered to XAD_C and WET_C, optical density (OD; λ = 550 nm) values ranged from 40 to 90%, whereas for the H_C carbon sample, the value was as low as 5%-9%.The binding efficiency of low molecular weight organic components depended on the morphology and porosity of the carbon materials in our research, and such a phenomenon was demonstrated for the first time.Our results indicate the need for more extensive research on the adhesion and growth of mammalian cells on spherical porous materials while observing the variable parameters of these unique materials employed as scaffolds for cell culture.In addition, the greater the adsorption of the amino acid, the larger the surface of the carbon sphere is occupied by the cells and, at the same time, they are more flattened.

Figure 3 .
Figure 3. Pore size distribution measured by consecutive intrusion/ extrusion cycles of mercury on SACs.

Figure 5 .
Figure 5. Phenylalanine content after the adsorption on SACs as evaluated by spectrometry.
deviation.(figure 7(B1)-(B3)) showed a more flattened morphology.These cells produced multiple protrusions; wide lamellipodia or narrower filopodia are shown in figure 7(B3).Therefore, TM3 cells may have adhered more efficiently to the WET_C surface than to the XAD_C surface due to a larger contact surface with the substrate and smaller curvature of the spheres.The surface of WET_C spheres was more uniformly covered with the cells than XAD_C.The cells attached to the

Figure 8 .
Figure 8.The metabolic activity of TM3 cells: (A) the cells remaining in the microplate wells after incubation with SACs for 24 h; (B)-TM3 cells adherent to the SACs' surface.
a S BET -specific surface area calculated by the Brunauer-Emmett-Teller method; b S SP -single point surface area at p/p °= 0.3; c V t -total pore volume at p/p °= 0.99; d