Background. Investigating the adhesion activity of different scaffold materials and topologies is important not just for bone, but for all other kinds of, tissue engineering. We studied how two different fiber orientations (random and aligned) electrospun into 2D scaffolds influence the attachment of cells of the MG-63 line.
Materials and methods. MG-63 human osteosarcoma derived cells were cultured in an incubator on nanofibrous scaffolds with random and aligned fiber orientations for 1h. Cell adhesion was evaluated by optical microscopy and based on the amount of DNA measured by PicoGreen analysis.
Results. Optical microscopy images revealed a higher number of attached cells and greater elongation on aligned fibers than on random ones. Controversially the PicoGreen assay showed no significant difference in cell numbers with a mean of 2.75Â±0.338 Âµg/ml and 2.93Â±0.227 Âµg/ml DNA concentration on random and aligned scaffolds respectively. The Student's T test between the samples produced a p-number of 0.5533.
Get your grade
or your money back
using our Essay Writing Service!
Conclusions. Thought the two assays produced seemingly opposite results, our findings suggest that fiber alignment has no significant effect on cellular adhesion in the case of MG-63 cells. Nonetheless a more thorough investigation is required before a finite conclusion can be drawn.
Keywords: Adhesion activity, MG-63, osteoblast-like, electrospinning, scaffolds, random, aligned, fiber orientation
A very important part of today's Tissue Engineering (TE) investigations is realizing the possibilities that the research of the past decades yielded, by optimizing the TE techniques through understanding the different factors that influence the outcome of tissue cultures. By mapping out different preferences of different cell types, which scaffold materials and properties, culture environments and stimuli, or what combinations of these give the best result, will we be able to create the desired Tissue Engineered product, which will be a great and useful alternative to "traditional" medical treatments.
To contribute to this greater process we chose to investigate whether osteoblast-like cells prefer random or aligned fibrous scaffolds. Understanding substrate preferences of osteoblasts is not just important in designing scaffolds for tissue engineering purposes, but for the many biomedical (non-TE) implants (orthopaedic, dental etc.), that are currently being developed.
1.1. Cell attachment and adhesion
Cell attachment and adhesion is critical, as those cells, that do not attach will not proliferate, secrete ECM and will die; as such these factors influence cellular growth, morphology, proliferation and differentiation (Goto et al, 2007). This explains why a lot of research is being done in this area on many of the different cell types, including osteoblasts. However there are, many variables (chemical, mechanical, geometric) that need to be considered in cell attachment and adhesion.
On the example of osteoblasts it was shown that surface topology (microgrooves) can influence cell alignment and ECM secretion (Matsuzaka et al, 2003). It was also shown that chemical patterning can influence this, though it's outcome greatly depends not just on their relative position, but different geometry (Calvanti et al, 2007), different concentrations of the same chemical (Charest et al, 2006) and even keeping the cells in different media (Teixeira et al, 2005) can give completely different result. Surface roughness is a two-edged blade as some roughness, compared to a smooth surface, will produce better adhesion, but worse proliferation (Zinger et al, 2004). It was also shown that, if human fetal osteoblasts (hFOB 1.19) attach to a surface and survive, even if just barely they will eventually proliferate (Liu et al, 2007).
Experiments were conducted on nano-fibrous material, and it was shown that their use for bone tissue engineering may be preferable compared to solid-walled architecture as they promote osteoblast differentiation and biomineralization. (Woo et al, 2007)
How all this information, as it was gathered using 2D cultures, can be transferred onto and used for 3D constructs is a difficult, and probably not straightforward, question.
Electrospinning, invented in 1934, is a fiber fabrication technique with which fibers, from micrometer to 100 nm diameter (even as low as 5 nm have been reported ) can be produced (Dalton et al, 2007; Frenot et al, 2003; Tan et al, 2005).
Although there are other techniques in existence capable of creating fibers of micrometer diameters (e.g. wet spinning, dry spinning, melt spinning, gel spinning), they are unable or unsuitable to be used on a nanometer scale (Frenot et al, 2003). With electrospinning we can manufacture fibers of different polymers (e.g. PLA, PGA, PLGA and PCI (Huang et al, 2004)) with many different diameters, mechanical properties and complex shapes, thus making it ideal to produce non-woven fabric for many nanotechnology applications such as air and water filtration, drug delivery and, most important for us, tissue engineering (He et al, 2005) .
1.2.1. Benefits of elecrospinning
Always on Time
Marked to Standard
Fibers of sub-micron or nanoscale have great advantages over other materials, by having much higher surface to volume or surface to mass ratio, small pore size and superior mechanical properties, such as stiffness and tensile strength, which make them very useful in a multitude of applications. (Dalton et al, 2007; Frenot et al, 2003; Ramakrishna et al, 2005; Huang et al, 2003; Thompson et al, 2007)
1.2.2. The process of electrospinning
This process uses a high voltage electric field to force a solution (or melt) of polymer out of a pipette or a small diameter needle. Due to the build up of charge in the fluid a so called "Taylor cone" forms from a droplet at the end of the pipette. When the charge reaches the critical amount a jet of polymer erupts forming an elongated fiber. The solvent evaporates in flight as the fiber travels to a grounded metal sheet, a place of lower potential, where it is collected. (Frenot et al, 2003; Ramakrishna et al, 2005; Tan et al, 2005; Huang et al, 2003)
With this arrangement only non-woven products of random fiber orientation can be produced. As the need to create aligned fiber topologies has risen several methods have been suggested. These are (Huang et al, 2003):
A fast rotating cylinder collector
Auxiliary electrodes and electric fields
A thin wheel with sharp edge
A frame collector
1.2.3. The parameters
As electrospinning is a complex process, a multitude of parameters influence it's outcome, which have been examined both theoretically (Thompson el at, 2007) and in practice (Tan et al, 2005).
These variables, classified into three groups, are:
Solution properties: Viscosity, polymer concentration, molecular weight, electrical conductivity, elasticity, surface tension
Processing conditions: Applied voltage, distance from needle to collector, volume feed rate, needle diameter
Ambient conditions: Temperature, humidity, atmospheric pressure (Tan et al, 2005)
Thompson's theoretical model showed that volumetric charge density, distance from nozzle to collector, needle diameter, relaxation time, and viscosity influence the outcome the most, while the other parameters have only moderate or low effect. (Thompson el at, 2007)
In comparison, Tan et al's results in some instances confirmed and at other occasions contradicted Thompson's findings. For example it was shown that decreased polymer concentration or increased electrical conductivity resulted in the greatest drop in diameter size and that low molecular weight produces non-uniform fibers more easily. (Tan et al, 2005) As such Tan et al classifies polymer concentration as a factor of great effect, while Thompson considers it only a moderately important parameter.
The cause of the difference between the two findings can be, that a mathematical model can never be perfectly accurate, especially in the case of such a complex process, or because some factors may be unintentionally left out and not taken into consideration, or simply just because the two experiments may have been based on different kinds of polymers.
1.3. Electrospinning in bone tissue engineering
Electrospinning has been widely studied and applied in bone tissue engineering research, mainly because it can be used to fabricate fibers into multiple forms and as such scaffolds with biomimetic properties can be easily produced (Jang et al, 2009). The recent work in using electrospun scaffolds concerns itself with coating the fibers (e.g. with calcium phosphate) and adding growth factors (e.g. bone morphogenetic protein 2 (BMP-2)) or nanoparticles to improve bioactivity (Li et al, 2006; Yang et al, 2008; Nandakumar et al, 2009; Sahoo et al, before print). One of the inherent feature of electrospinning, that only thin scaffolds can be manufactured using it, has somewhat limited its use in bone tissue engineering, where usually bone defects of relatively large size are present. One of the methods, currently under investigation, is trying to address this matter by bonding multiple electrospun scaffolds using biocompatible gel, thus increasing the thickness of the product. (McCullen et al, 2010)
2. Methods and materials
2.1. Cell line and culture
For this experiment, though some argue it's usability in such experiments because of its origin, and whether it's phenotype is closer to osteoblasts or osteoprogenitors (Sautier et al, 2002), we have chosen the human osteosarcoma derived (Billiau et al, 1975) MG-63 ostoblast-like cell line.
The cells were of passage 98 and were cultured in a T-75 flask in a media of DMEM (Dulbecco's Modified Eagle Medium) with 10% of FCS (Fetal Calf Serum), 1% of L-Glutamine and 1% of antibiotics and antimyotics.
This Essay is
a Student's Work
This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.Examples of our work
(This media composition was used throughout the experiment)
The culturing took place in an incubator at 37Â°C and 5% CO2 content. After culturing the media was removed and the cells were washed with 5 ml of PBS (Phosphate Buffered Saline), and de-attached from flask surface with 3ml Trypsin-EDTA. (Please note that after the adding of the trypsin the cells should be regarded as from passage 99.) To enhance de-attachment the cells were placed back into the incubator for 5 minutes after trypsinization. The Trypsin was deactivated using 5 ml of quenching media and the cells were centrifuged for 5 minutes on 10000 rpm. The cell pellet was resuspended in 5ml media. From the mixture 10 Âµl was removed for cell counting with haemocytometer.
2.2. The material of the fibers
We have chosen poly(Îµ-caprolactone) (PCL), a synthetic polymer already frequently and widely used in medical applications, as such in orthopaedic tissue engineering (Jeong and Hollister, 2010; Oh et al, 2007) , as our fiber material. Its positive properties, e.g. that it's biocompatible, biodegradable, easy to fabricate and available off-the-shelf (Chen et al, 2008; Jeong and Hollister, 2010; Oh et al, 2007; Lin et al, 1999) outweigh its shortcomings, the great hydrophobicity, slow biodegradation rate and low, but still present, toxicity (Huang et al, 2010).
2.3. Preparing the scaffolds
The PCL polymer was suspended in chloroform and dimethylformamide solution. The ratio of the chemicals in the solution was 7:3 respectively. The polymer was then electrospun in the two arrangements (random and aligned) onto 24mm by 24mm coverslips to create 2D scaffolds. 3 scaffolds with random and 3 with aligned arrangement were created. The parameters used for the different orientations were: Volume of 0.012 ml with 0.01 ml/min flow rate, 22G blunt needle, working distance of 15 cm and 4 kV external voltage for random, and volume of 0.018 ml with 0.01 ml/min flow rate, 22G blunt needle, working distance of 20 cm and 4.5 kV external voltage for aligned scaffolds.
The scaffolds then were sterilized in 70% IMS for 1 hour, dryed for 1 hour, and placed in a 6-well plate using sterile forceps.
2.5. Cell seeding protocol
1.6 million cells were separated and re-suspended in 8 ml media. From this suspension 100.000 cells in 500 Âµl media were seeded onto each scaffold. After the seeding process the scaffolds were placed into the incubator at 37Â°C and 5% CO2. After approx. 1 hour the scaffolds were removed from the incubator and the morphology of the cells was assessed by optical microscopy. The media then was removed from the wells and the scaffolds were washed with PBS.
2.6. DNA extraction
To extract the DNA from the cells for the PicoGreen analysis the cell membranes had to be broken down. This was done three fold:
Firstly (and mainly) chemically, by adding 400 Âµl lysis buffer of 1:20 diluted TE buffer and 0.1% Triton-X to each well. The cells were lysed for 5 minutes and the solutions from each well then were transferred to 1.5 ml eppendorf tubes. Each sample was additionally vortexed to further brake down the cell membranes physically. The samples then were stored at -20Â°C overnight (which also propagates the down breaking by freezing).
2.7. PicoGreen assay
The number of the cells, their adhesion, on each scaffold was determined using the PicoGreen assay. The samples were taken out of the freezer and were kept on ice prior to use, before being transferred to a 96-well plate. 100 Âµl of the samples and the Î» DNA standards (0-2 Âµg/ml) were put into different wells and were all done in double. 100 Âµl of 1:200 diluted PicoGreen (2 mg/ml) was added to each used well. The plate was incubated in foil before being excited at 485 nm and read at 528 nm in a Biotek Synergy 2 fluorescence plate reader (Gem 5 software).
2.8. Statistical analysis
Student's T-test was used to analyse the data. Data are represented as meanÂ±SD. (SD = standard deviation)
3.1. Evaluation of cell morphology
Cell morphology and alignment was studied with optical microscopy after 1h of culture on the scaffolds. Though this evaluation is not quantitative, as it is based on the subjective estimation of the experimenter, rather than quantitative data and the images might not even be representative of the whole scaffold, the effects of parallel fiber topology can be easily seen on the image in Fig. 1 (A). A high ratio of the cells (56 out of 73, based on a quick numerical counting) shows spread morphology, a sign of attachment, and seem to be aligned with the fibers, especially well displayed in the upper right corner of the image. Some cells (23%) have retained spherical morphology and have not attached, but their amount is small compared to those in Fig. 1 (B) where little to no cell attachment or alignment can be seen. The presence of spherical un-attached cells in Fig. 1 (A) suggest that there is place for improvement in the case of parallel orientation aswell, nonetheless, based on these findings parallel fiber topology seems to be the more favourable substrate from the perspective of cellular attachment.
3.2. PicoGreen analysis
A standard DNA curve, shown in Fig. 2, was fitted to four points of absorbance measurements. Fitting this curve was necessary to determine the relationship between the amount of light absorbed in the plate reader and the DNA found in the samples.
In Fig. 3 we can see the concentration of DNA in Âµg/ml found in each sample after 1h of culture: three with random (R1, R2, and R3) and two, due to one of the samples being lost, with aligned (A1 and A3) topology. There is some variation visible between samples of the same kind.
A more representative comparison between the two topologies is found in Fig. 4, where the mean concentration found in the random and aligned samples, 2.75Â±0.338 Âµg/ml and 2.93Â±0.227 Âµg/ml respectively, are displayed, showing no significant difference between cell numbers on the two arrangements (Student's T-test, p=0.5533).
Our aim with this study was to determine and compare the effects of random and aligned topologies on the MG-63 osteoblast like cells. Based on cell morphology observations, it would be easy to conclude that aligned fibers promote cell attachment, alignment and elongation considerably better, with only 23% of the cells seemingly unaffected, compared to those with random orientation, as cells cultured upon these remained spherical in shape and showed no sign of attachment. Interestingly, there is a quite puzzling controversy between the results of the two assays used, as based on the concentrations of DNA found in our samples, the two scaffold types show no significant difference in cell numbers. The simplest explanation is human error, as mistakes during the experiment could easily have altered its outcome. The images may not be representative of the whole of or all of the scaffolds or there may be biocompability issues, that were previously not taken into account, with the poly(Îµ-caprolactone) fibers, that may block the adhesion of the cells in some way.
4.1. Comparison of results to literature
In literature the effects of fiber orientation on tissue types other than bone have been thoroughly investigated. Skeletal muscle cells have been cultured on electrospun PLC scaffolds with both aligned and random orientation. Thought the initial results showed greater cell alignment and myotube formation on unidirectional fibers, compared to random arrangements and cells cultured on flat surfaces as controls, no significant difference between cell adhesion and proliferation was found (Choi et al, 2008). Similarly, results have shown in the case of NIH 3T3 fibroblast cultured on poly(D,L-lactic-co-glycolic acid) (PLGA) fiber meshes, that with increased fiber diameter and orientation cells displayed greater cell-area and aspect ratio, cell proliferation seemed not to be sensitive to these parameters. (Bashur et al, 2006). As an example for stem cell behaviour, human MSCs exhibited elongated morphology on aligned collagen and collagen/heparin meshes, while no statistical difference in cell numbers was found between the aligned and random matrices (Lanfer et al, 2009).
The effect of surface topology on bone cells has been extensively studied. Primary osteoblasts have been shown not just to elongate and proliferate around surface patterns (e.g. nanogrooves), but upregulated ECM production (also aligned with the surface topology) has been reported. (Lamers et al, 2010; Lenhert et al, 2005; Matsuzaka et al, 2003). There is evidence of the effects of nanopatterning on cellular behaviour in the case of osteoblast-like cancer derived cells aswell. MG-63 cells have been shown, after 4 and 24 h spent in a bioreactor, to align along grooves and proliferate faster compared to smooth surfaces. (Yang et al, 2009). Interestingly the opposite can be observed when culturing cells of the MCT3T-E1 line: Cell numbers on microgrooved substrates after 7 days of culture were shown not to be significantly bigger than those on flat surfaces. (Wang et al, 2000)
When considering these above mentioned findings, we must take care, as cell numbers in these studies were usually counted after several days in culture, giving time to the cells to proliferate. In our investigation, under only one hour spent in the incubator, there was not time for them to do so, as such our cell number is only representative of cellular adhesion and not proliferation, and should not be compared directly to the above mentioned ones. Still based on the results of the other scientific investigators: Cells, primary or derived, stem or differentiated, those with bone origins or from other tissue types, seem to prefer aligned topologies, whether those are fibrous or grooved, and display greater cellular alignment and ECM production when cultured upon them. But, if we are looking at the issue from the viewpoint of proliferation, an additional significant increase in cell number, besides the greater adhesion, has only been witnessed when using MG-63 osteoblast-like cells on nanogrooved surfaces (Yang et al, 2009). The lack of difference in most of the investigations can be due to that in most cases the cells on both arrangements had time to react properly. It was shown that cells do attach and adhere even to completely unfavourable substrates, if they manage to survive, it only takes much longer (Liu et al, 2007). This may be the case here aswell: Random orientation may not be completely foreign to the cells, they are able to attach to the fibers, they only require more time to do so, than with aligned fibers, and will eventually reach the same number of cells.
The study by Yang et al is particularly interesting, because we used the same cell line in our experiment, and our results based on our PicoGreen analysis, suggest the opposite. The difference could be explained by numerous reasons: The geometrical differences, as our scaffold was built up by fibers and theirs was a nanotextured substrate. We used PCL, while their choice of material was silicone. Differences in cell passage number, the culture media, time and environment may also have played a role. Thought it can be easily understood, how the above mentioned factors can be the cause of significant differences, as it was shown that subtle changes in parameters can greatly influence the outcome (Matsuzaka et al, 2003; Calvanti et al, 2007; Teixeira et al, 2005; Zinger et al, 2004; Charest et al, 2006), the answer may be as simple as that the effects of grooved surfaces and fibrous topologies cannot be compared.
The finite conclusion in this study is controversial, as based on current literature and our findings in Fig. 1, it seems more likely that aligned fibers influence cellular attachment in a positive way, while the outcome of the PicoGreen assay suggests the opposite. As the later is quantative and the one, that represents the whole of the scaffolds, we find our self obliged to rely on its results as the basis of our conclusion.
Nonetheless, we feel that such verdicts should not be drawn hastily, based on only one analysis carried out on the topic. Many more experiments should follow, some with the same, some with different parameters (e.g. fiber diameter, porosity, culture time, media composition), to get a statistically more stable base. Even if we have all these at our disposal, it must be kept in mind that, the conclusion then will still be only definitive for the MG-63 cell line and can only imply how primary osteoblasts or other bone cells would behave under similar circumstances.
4.2. Suggestions for other materials that could be used with electrospinning
As the alignment of PCL fibers didn't improve cell adhesion, modifying, augmenting the current process or even the use of other materials may be viable.
Current literature contains many possible ways with which we could improve the process even if we continue using PCL as our "base" material. This could be done by simply changing some of the parameters of the electrospinning process, as bone cells may prefer scaffolds with different diameters, porosity or mechanical properties. (Badami et al, 2006)
The surface chemistry is just as important as the architecture of the scaffold, as such coating the fibers with bone-like apatite or calcium phosphate (CaP) ceramic (Yang et al, 2008), something which the cells will recognize as similar to their natural environment, may stimulate and enhance cell adhesion. Growth factors could also be incorporated. For example BMP-2, a growth factor with well known osteogenic effects, has been electrospun into a scaffold and remained bioactive. (Li et al, 2007)
Poly(Îµ-caprolactone) (PCL) have been first suggested and used for bone tissue engineering, by Yoshimotoa et al (2003), because of its low cost, low toxicity and slow degradation, but there are a lot of other organic and non-organic materials, already tried in tissue engineering research, which could be more, or at least as, suitable as PCL. Other synthetic polymers of the group of poly(Î±-hydroxyl acid) (e.g. poly(lactic acid) (PLA), poly(glycolic acid) (PGA)) can be just as easily electrospun and all have their advantages and disadvantages compared to PCL, when used in bone tissue engineering. As the inherent hydrophobic chemistry of the synthetic polymers limit cell adhesion, the use of natural polymers, those that can be electrospun into different diameters and are already found in the body e.g. collagen type I (a major component of bone), may be preferable. The use of bone-bioactive inorganic materials, such as calcium phosphate and bioactive glass, has also been suggested. (Jang et al, 2009)
We find most interesting and suggest the approach of combining the above mentioned material types to create polymer-inorganic composites, as these mimic the hydroxyapatite/collagen structure of the bone in 3D (Jang et al, 2009). Electrospinning together the brittle, but compressively resistant, inorganic ceramics, with a matrix of flexible fibre, that is more resistant to tensile stress, gives the scaffold the mechanical properties desirable for bone tissue engineering, and also may prove to be a more suitable environment for the cells. Though the polymers most used to fabricate these composites are poly(L-lactic acid) (PLLA), poly(lactic-glycolic acid) (PLGA) and poly(Îµ-caprolactone) (PCL) (Patlolla et al, 2010), other combinations, composed mostly of natural polymers, like poly-L-lactide/ hydroxyapatite/collagen (Mohamma et al, 2009) and hydroxyapatite/chitosan (Zhang et al, 2008) have been investigated aswell and may be promising scaffolds for tissue engineering.
In this study we used MG-63 osteosarcoma derived cells on 2D scaffolds composed of electrospun poly(Îµ-caprolactone) (PCL) fibers in either aligned or random orientation to determine the effect of these different topologies on cellular behaviour, with the emphasis mainly on attachment and alignment.
We have concluded that aligned fiber orientation does not influence cellular attachment, as after 1 h of culture, there was no significant difference in cell number between the two scaffold types, though cells on scaffolds with quasi-parallel fibers aligned more and showed elongated morphology, compared to those on random fibers, on optical microscopy images.
As our results have controversy among themselves and seem to be the opposite of those found in current literature, this topic needs a more thorough investigation, before a more finite conclusion can be drawn, one that can be the foundation of future studies in the field of tissue engineering and the realization of the possibilities it holds.
The authors would like to thank Dr. Sarah H. Cartmell for this great opportunity to experience the work of cell and tissue engineers in practice, Deepak Kumar for his help and friendly guidance during the practical and the Institute of Science and Technology in Medicine.