SEARCH

SEARCH BY CITATION

Keywords:

  • tendon healing;
  • tenocytes;
  • microcarrier;
  • bioreactor

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. Literature Cited

Tendon healing is a time consuming process leading to the formation of a functionally altered reparative tissue. Tissue engineering-based tendon reconstruction is attracting more and more interest. The aim of this study was to establish tenocyte expansion on microcarriers in continuous bioreactor cultures and to study tenocyte behavior during this new approach. Human hamstring tendon-derived tenocytes were expanded in monolayer culture before being seeded at two different seeding densities (2.00 and 4.00 × 106 cells/1000 cm2 surface) on Cytodex™ type 3 microcarriers. Tenocytes' vitality, growth kinetics and glucose/lactic acid metabolism were determined dependent on the seeding densities and stirring velocities (20 or 40 rpm) in a spinner flask bioreactor over a period of 2 weeks. Gene expression profiles of tendon extracellular matrix (ECM) markers (type I/III collagen, decorin, cartilage oligomeric protein [COMP], aggrecan) and the tendon marker scleraxis were analyzed using real time detection polymerase chain reaction (RTD-PCR). Type I collagen and decorin deposition was demonstrated applying immunolabeling. Tenocytes adhered on the carriers, remained vital, proliferated and revealed an increasing glucose consumption and lactic acid formation under all culture conditions. “Bead-to-bead” transfer of cells from one microcarrier to another, a prerequisite for continuous tenocyte expansion, was demonstrated by scanning electron microscopy. Type I and type III collagen gene expression was mainly unaffected, whereas aggrecan and partly also decorin and COMP expression was significantly downregulated compared to monolayer cultures. Scleraxis gene expression revealed no significant regulation on the carriers. In conclusion, tenocytes could be successfully expanded on microcarriers. Therefore, bioreactors are promising tools for continuous tenocyte expansion. © 2013 American Institute of Chemical Engineers Biotechnol. Prog., 30:142–151, 2014


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. Literature Cited

Tendon ruptures and tears remain major orthopaedic challenges.[21, 22] Healing of tendon injury is time consuming and leads to scar formation as a functionally altered reparative tissue.[1] Tendon is a hypovascular and hypocellular tissue with a low metabolite exchange rate.[2] The resident cells, the tenocytes, are embedded within their abundant ECM.[3] This ECM consists mainly of type I collagen arranged in dense parallel fibril bundles.[4] In healing tendons, type III collagen expression is increased.[2] The main proteoglycan in tendon is decorin.[23, 24] However, aggrecan and the glycoprotein COMP are also present in lower amounts in the tendon ECM.[24, 25] Tendon tissue engineering could be an attractive approach to improve tendon repair, e.g., by using biodegradable biomaterials seeded with tenocytes,[5] but it requires substantial cell expansion in vitro.

During monolayer expansion, tenocytes display an altered phenotype.[6-8] With increasing passage number, human Achilles tendon-derived tenocytes cultured in monolayer became more rounded and were more widely spaced at confluence. The ratio of type III to type I collagen increased from passage 1 to 8. Decorin expression significantly decreased with passage number.[6] Protein levels of type I and III collagen and decorin were downregulated after four monolayer passages in human biceps tendon fibroblasts,[8] and the number of transcripts of type I collagen and decorin decreased after the first monolayer passage in rabbit Achilles tenocytes.[7] However, Almarza et al. investigated rat tendon fibroblasts until passage 5 and detected no major differences in the expression profile.[9] Previously, we found that tenocyte gene expression profile strongly depends on 2D or 3D culture conditions and clearly differs from that in native tendon.[10] Scleraxis is a typical tendon marker[11] with significantly higher expression in tendon compared with tenocytes expanded in monolayer.[10]

In recent years, bioreactor cultivation procedures using microcarriers have been intensively studied for chondrocyte expansion.[26, 27] They provide a suitable basis for continuous cell expansion without passaging using trypsin/ethylenediaminetetraaceticacid (EDTA) or other agents required for cell detachment that might interfere with tenocytes' metabolism and lead to the loss of freshly synthesized ECM. Moreover, human chondrocytes can be cultured on microcarriers under retention of their chondrocytic phenotype.[12] Furthermore, bioreactor cultures are less time-, cost-, and labor-intensive than monolayer cultures, since cell expansion can be achieved by simply adding empty microcarriers.[13] Hence, no loss of freshly synthesized cell-associated ECM occurs in microcarrier culture. In addition, bioreactor cultures have the advantage of allowing regulation and automation, e.g., by using advanced bioreactor systems and monitoring of the physical, chemical and biological milieu. The stirring, necessary for permanent floating of the microcarriers, provides mechanostimulation of the cultures.

Our overall aim is to develop an efficient, controlled and automated bioreactor system for a continuous culture of tenocytes. Here, we demonstrate for the first time that hamstring tenocytes are able to colonize Cytodex™ type 3 microcarriers. Expansion in a spinner flask bioreactor system led to “bead-to-bead” cell transition of the tenocytes, underlining their migratory potency. Cells remained vital over 4 weeks in the spinner flask system. Substantial expression of genes coding for the main ECM proteins type I and type III collagen as well as scleraxis was maintained by the tenocytes on microcarriers. Therefore, expansion in spinner flasks using microcarriers could serve as a tool to continuously expand primary tendon-derived cells, and provides the basis for the development of a more advanced automated and controlled bioreactor culture system.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. Literature Cited

Human tenocyte isolation and expansion

Human primary hamstring tendon-derived tenocytes were isolated as described previously[10, 28] from hamstring tendons (midsubstance of Musculus (M.) semitendinosus, M. semimembranosus, M. gracilis tendons) of three healthy middle-aged donors (average age 43 (±17.6) years, two male and one female donor). Hamstring tendons are often used as autotransplants for anterior cruciate ligament reconstruction. This study was approved by the Charité-Universitätsmedizin Berlin review board for experiments with human-derived tissues. After careful removal of the peritendineum, pieces of human tendon were cultured in growth medium at 37°C and 5% CO2 for several days. Growth medium consisted of Ham's F-12/DMEM 1:1 containing 10% fetal calf serum (FCS) (Biochrom, Berlin, Germany), 25 μg mL−1 ascorbic acid (Sigma–Aldrich, Munich, Germany), 50 IU mL−1 streptomycin, 50 IU mL−1 penicillin, 0.5 μg mL−1 partricin, essential amino acids, 2 mM l-glutamine (all: Biochrom). After 1–2 weeks, tenocytes continuously emigrated from these explants and adhered to Petri dishes. When cell density reached confluence, the cells were removed using 0.05% trypsin/1.0 mM EDTA and multiplied up to passage 6 in monolayer culture to obtain enough cells for the microcarrier experiments.

Preparation of microcarriers

For cultivation of tenocytes, Cytodex™ type 3 microcarriers (GE Healthcare, Munich, Germany) were reconstituted using phosphate buffered saline (PBS) (Biochrom). After autoclaving and threefold washing of the microcarriers with cell culture medium, cells were seeded to determine the optimal adherence time as a basis for further cultivation.

Determination of tenocyte adherence time

To determine the optimal cell adherence time, 1.00 × 106 cells and 0.037 g microcarrier (100 cm2 surface area) were put in a Sigmacote- (Sigma–Aldrich) pretreated 100-mL glass bottle (Schott, Mainz, Germany) with a volume of 20-mL culture medium.[13] Sigmacote prevents the attachment of cells on the surface of the bottle. Cells attached on 60 microcarriers were counted after 0.5, 1, 3, and 6 h using a phase-contrast microscope (Leica Microsystems, Wetzlar, Germany). Applying SigmaPlot 8.0 software (Systat Software, Erkrath, Germany), regression analysis was performed, whereby an average cell count per microcarrier was calculated, after comparing the measured frequency of cell distribution with a Poisson distribution.[14] In brief, the frequencies of microcarriers with distinct cell numbers (w) were plotted against carriers with another distinct cell number (c). According to the measuring points the Poisson distribution curve was adjusted to determine the average cell count per microcarrier (up).

  • display math(1)

Microcarrier culture of tenocytes in spinner flask bioreactor systems

For cultivation of tenocytes in bioreactors, 0.37 g of Cytodex™ type 3 microcarriers (1000 cm2 surface area) and 2.00 or 4.00 × 106 cells were used with 2 × 106 cells being the minimal recommended seeding cell number and 4 × 106 cells the doubled amount. After adherence over the determined time of 3 h in siliconized glass bottles (100 mL size) in a volume of 20 mL growth medium (Ham's F-12/DMEM 1:1 containing 10% FCS, 50 IU mL−1 streptomycin, 50 IU mL−1 penicillin, 2 mM l-glutamine) without stirring, the colonized carriers were transferred to spinner flasks with two glass impellers (Integra Biosciences, Fernwald, Germany) and cultivated for 2 weeks with 20 or 40 rpm in a volume of 50 mL with 20 rpm being the minimal stirring velocity recommended by the microcarrier manufacturer and 40 rpm the doubled stirring velocity. For demonstration of “bead-to-bead” transfer, cells in five combinations of seeding densities (2.00–6.50 × 106 cells) and stirring velocities (60–75 rpm) were cultivated in spinner flasks. After 2 weeks, half of the microcarriers were replaced by empty carriers and further cultivated for 2 weeks. All vessels and spinner flasks were siliconized using Sigmacote. All cells on microcarriers were cultured in growth medium at 37°C and 5% CO2. Half of the medium was changed every second day.

Scanning electron microscopy (SEM) of cell-seeded microcarriers

For the visualization of cell growth on the microcarriers after 2 weeks of culture, SEM was performed. The carriers were washed twice with PBS and incubated with 2.5% glutaraldehyde (Merck) for 10 min. To dehydrate the microcarriers washing steps with increasing ethanol (Merck) concentrations (30%-absolute) were performed. For drying, samples were incubated one time with hexamethyldisilazane (Sigma–Aldrich) for 5 min and two times for 10 min. The final drying was performed overnight under an exhaust hood. After sputtering the microcarriers and cells (Emitech K550, Röntgenanalytik Messtechnik, Taunusstein, Germany) in an argon atmosphere (10−2 mbar, 240 s, 30 μA), observation with a scanning electron microscope (Philips XL20, FEI, Eindhoven, Netherlands) was performed.

Estimation of proliferation and vitality testing of tenocytes on the microcarriers

During the cultivation of tenocytes on microcarriers, the cell proliferation was determined daily using 3-(4,5-Dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) assay as described by Frauenschuh et al.[13] Nearly 200 μL of the microcarrier suspension were supplemented with the MTT (Sigma–Aldrich) solution. After 2 h of incubation the MTT assay was stopped using lysis buffer consisting of 81% isopropanol (Merck, Darmstadt, Germany), 4% 1 M HCl (Merck) and 15% sodium dodecyl sulfate (SDS) solution (20% SDS in H2O) (Sigma–Aldrich). Absorption at a wavelength of 570 nm was measured with a Mithras LB 940 plate reader (Berthold Technologies, Bad Wildbad, Germany) and compared with standards to determine the cell number. The standards were prepared with the same cells as used for the cultivation when starting the bioreactor cultivation at day 0 by using defined cell numbers in the range of 0–1 × 105 cells per 200 μL. Growth rates for exponential growth and population doublings (PD) were calculated according to the following formulas with growth rate μ, time points tfinal and texp, corresponding cell numbers Xfinal and Xexp, and the PD. The time point texp was chosen for each cultivation procedure as the day when the exponential growth starts. The time point tfinal is the final day of culture.

  • display math(2)
  • display math(3)

After 2 weeks in culture on microcarriers tenocyte vitality was measured using fluorescein diacetate (FDA) (Sigma–Aldrich) for labeling of living cells and propidium iodide (PI) (Sigma–Aldrich) for staining of dead cells. A concentration of 3 μg mL−1 for FDA and 0.1 mg mL−1 for PI diluted in PBS was used. One mL of the microcarrier solution was used for the staining. Carriers were washed with PBS and incubated with 200 μL of the PI/FDA solution for 5 min at room temperature in darkness. Cell vitality was examined by fluorescence microscopy (CKX41, Olympus Soft Imaging Solutions, Hamburg, Germany). Images were taken using an Olympus camera Color View II (Olympus Soft Imaging Solutions).

Glucose-/lactic acid assay of tenocyte culture supernatants

Estimation of glucose and lactic acid concentrations was performed daily by an enzymatic-amperometric method during 2 weeks of tenocyte spinner flask culture. Therefore, supernatants of microcarrier cultures were taken each day and 10 μL were measured using the Sensostar GL one analyzer (DiaSys Diagnostic Systems, Holzheim, Germany) for the determination of glucose and lactic acid concentration.

Ribonucleic acid (RNA)-isolation

After 14 days of cell culture, 15 mL of the microcarrier suspension was collected and centrifuged at 310g. Cells in the pellet were lysed using the RLT-lysis buffer of the Qiagen RNeasy miniKit (Qiagen, Hilden, Germany). Subsequently, RNA was isolated using the Qiagen RNeasy miniKit according to manufacturer's protocol.

Messenger RNA analysis by real time detection polymerase chain reaction

Real time detection polymerase chain reaction (RTD-PCR) analyses were performed to obtain quantitative gene expression data for aggrecan (ACAN), cartilage oligomeric matrix protein (COMP), type I and III collagen (COL1A1, COL3A1), decorin (DEC) and scleraxis (SCX). Ribonucleic acid quantity and quality was measured using the RNA 6000 Nano assay (Agilent Technologies, Böblingen, Germany). Reverse transcription was performed with the Quanti Tect Reverse Transcription Kit (Qiagen). In brief, 1-μL aliquots of the complementary deoxyribonucleic acid (cDNA) (16.7 ng) were amplified using RTD-PCR in a 20-μL reaction mixture using TaqMan Gene Expression Assays (Applied Biosystems, Foster City, USA) and specific primer pairs for the above mentioned genes and the house-keeping gene β-actin (ACTB) (Table 1). All assays were performed in an Opticon 1 - Real-Time-Cycler (Biorad, Munich, Germany). The conditions of amplification were as follows: 10 min at 95°C, and then for 40 cycles 15 s at 95°C, 30 s at 60°C, followed by 6°C cooling. A lack of primer dimers and specificity of amplification were confirmed by efficacy testing and agarose gel electrophoresis of PCR products; all showed a single band of expected size. Relative gene expression levels were normalized with β-actin and calculated with the 2-deltaCT method.[15]

Table 1. Oligonucleotides used for RTD-PCR Analysis
Gene/AliasGene SymbolForward/Reverse PrimerProbe/Referencebp
β-actinACTBTGGGACGACATGGAGAA/GCCCCCCTGAACCCTAA147
GAAGGTCTCAAACATGATCTGG  
Type I collagenCOL1A1GGCAACGATGGTGCTAA/AATGCCTGGTGAACGTG138
GACCAGCATCACCTCTGTC  
  NCBI Gene ReferenceCat. no. 
ScleraxisSCXNM_001080514.1Hs03054634_g163
Type III collagenCOL3A1NM_000090.3Hs 00164103_m166
DecorinDECNM_133503.2Hs00370384_m177
COMPCOMPNM_000095.2Hs00164359_m1101
AggrecanACANNM_0132272.2Hs00202971_m193

Type I collagen and decorin immunolabelling

Tenocytes were cultured for 2 weeks on the microcarriers in spinner flask bioreactors. Then, the tenocytes colonizing the microcarriers were fixed for 15 min with 4% paraformaldehyde, subsequently rinsed with tris-buffered saline (TBS) and blocked with protease-free donkey serum (5% diluted in TBS) for 30 min at room temperature (RT). Then, the carriers colonized with the tenocytes were incubated with polyclonal rabbit anti-human type I collagen (27.5 μg mL−1, Acris Antibodies, Hiddenhausen, Germany), anti-human decorin antibody (1 μg mL−1, Acris) or anti-human aggrecan antibody (29.5 μg mL−1, R&D systems, Minneapolis, MN) in a humidifier chamber overnight at 4°C. Microcarriers were subsequently washed with TBS before incubation with donkey-anti-rabbit-Alexa-Fluor®488 (10 mg mL−1, Invitrogen, Carlsbad, USA) secondary antibody for 30 min at RT. Negative controls included omitting the primary antibody during the staining procedure. Cell nuclei were counterstained using 4′,6-diamidino-2-phenylindole (DAPI) (0.1 μg mL−1, Roche, Mannheim, Germany). Labeled microcarriers were rinsed several times with TBS, embedded with Fluoromount G (Southern Biotech, Biozol Diagnostica, Birmingham, USA) and examined using fluorescence microscopy (Axioskop 40, Carl Zeiss, Jena, Germany). Images were taken using an Olympus camera XC30 (Olympus Soft Imaging Solutions GmbH).

Statistical analysis

Normalized PCR data were expressed as the mean and standard error of the mean (mean ± SEM). Differences between experimental groups were considered significant at P < 0.05 as determined by Student's paired two-tailed t test combined with one way ANOVA and Bonferoni post hoc test (GraphPad Prism 5, GraphPad software, San Diego, USA).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. Literature Cited

Determination of an optimal cell adherence time

To determine the optimal adherence time, cells and microcarriers were put in a coated glass bottle (n = 3 donors) and 60 carriers were counted after 0.5, 1, 3, and 6 h, respectively. We found that 45.8% (±5.3%) of the cells attached to microcarriers after 0.5 h, 56.4% (±8.8%) after 1 h, 59.8% (±14.5%) after 3 h and 63.2% (±12.5%) after 6 h (Figure 1A). The cell adherence followed a Poisson distribution curve, the cell count per microcarrier revealed a mean number of adhered cells of 4.1 (±0.5) after 0.5 h, 5.1 (±0.8) after 1 h, 5.4 (±1.3) after 3 h and 5.7 (±1.1) after 6 h. Clearly, human hamstring tenocytes had colonized the Cytodex™ type 3 microcarriers. An adherence time of 3 h was chosen for further experiments.

image

Figure 1. Adherence kinetics and culture expansion of tenocytes in monolayer culture and on microcarriers at different cell densities. Determination of the adherence kinetics of human hamstring tenocytes (n = 3 donors) on Cytodex™ type 3 microcarriers (seeding density 1.00 × 106 cells/100 cm2 microcarrier surface) by cell counting revealed a mean increase of adhered cells (in percent of seeded cells, Poisson-normalized mean values) of cells per microcarrier with progressing time, with only a marginal change between 3 and 6 h. Hence, a period of 3 h was chosen for further experiments (A). Cultivation of 2.00 × 106 cells/1000 cm2 (B) and 4.00 × 106 cells/1000 cm2 (C) on Cytodex™ type 3 microcarriers for 14 days revealed an increased cell proliferation over time. Cell morphology in monolayer culture in proliferating (D) and confluent states (E) are shown, after photographing on an inverted microscope. Representative cell proliferation up to day 14 on Cytodex™ type 3 microcarriers is demonstrated for one donor (2 × 106 cells/1000 cm2 seeded cells and 40 rpm stirring velocity) at day 5 (F), day 10 (G), and day 14 (H) (all ×100, scale bars: 200 µm).

Download figure to PowerPoint

Tenocyte microcarrier culture

Human tenocytes and Cytodex™ type 3 microcarriers were combined in a spinner flask bioreactor system and cultivated for 2 weeks. All tested parameter combinations (cell seeding density of 2.00 or 4.00 × 106 cells/1000 cm2 microcarrier surface and 20 or 40 rpm microcarrier stirring velocity) led to a substantial cell proliferation (Table 2). Using MTT assays, in mean (n = 3 donors), for 2.00 × 106 seeded cells, 1.45 (±0.17) × 106 (20 rpm) or 1.15 (±0.07) × 106 (40 rpm) cells adhered to the carriers after 1 day of cultivation, reaching a total cell number of 5.83 (±3.39) × 106 cells (20 rpm) with 2.0 PD and 7.09 (±3.39) × 106 cells (40 rpm) with 2.6 PD. The cells colonizing the microcarriers barely changed in cell number for the first days. For a stirrer rotation velocity of 20 rpm, the cells began a period of increased proliferation at day 9 with an exponential growth rate of 0.206 PD/day while using 40 rpm; the cell growth started at day 6 with an exponential growth rate of 0.198 PD/day (Figure 1B). For 4.00 × 106 seeded cells 2.64 (±0.79) × 106 (20 rpm) or 1.83 (±1.42) × 106 (40 rpm) cells populated the microcarriers at day 1. After 14 days of culturing, 9.68 (±1.00) × 106 cells using 20 rpm (1.9 PD) and 9.01 (±3.07) × 106 cells using 40 rpm (2.3 PD) were detected with the MTT assay. With 4.00 × 106 seeded cells and 20 rpm an increased cell proliferation was measured starting with day 1, with an exponential growth rate of 0.103 PD/day, whereas an enhanced cell proliferation using 40 rpm was measured beginning at day 5 with an exponential growth rate of 0.174 PD/day (Figure 1C). The proliferation was also observed in phase contrast microscopy in monolayer (Figure 1D,E) and microcarrier culture (Figure 1F–H). Taken together, cell proliferation was detected in all chosen culture settings. After an initial stationary phase, exponential cell proliferation started in all spinner flasks with relatively high growth rates and PD compared to the mean cell proliferation in monolayer cultures for the two passages (5–7 days in cell culture flasks) before microcarrier cultivation (mean growth rate μ = 0.093 (±0.036) PD/day and 0.9 (±0.3) PD).

Table 2. Proliferation of Tenocytes on Cytodex™ Type 3 Microcarriers
Seeded Cells [×106]rpmMean Adhered Cells at day 1 [×106]Mean Adhered Cells at day 14 [×106]Population DoublingsStart of Exponential Cell Growth (day)Growth Rate µ (Start of Exponential Cell Growth to Day 14) (/day)
2201.45 (±0.17)5.83 (±3.39)2.090.206
2401.15 (±0.07)7.09 (±3.39)2.660.198
4202.64 (±0.79)9.68 (±1.00)1.910.103
4401.83 (±1.42)9.01 (±3.07)2.350.174

SEM of cell-seeded microcarriers

To characterize and monitor the populated microcarriers, SEM investigations were performed. At day 14, SEM observations revealed at a 200-fold magnification that nearly all microcarriers were populated with tenocytes in response to all different parameters tested for cultivation (Figure 2A). Closer examination (1000-fold) demonstrated mostly the stretched fibroblastoid morphology of the human hamstring tenocytes on the Cytodex™ type 3 microcarrier surfaces (Figure 2B).

image

Figure 2. Morphology, cell vitality and glucose/lactic acid metabolism of tenocytes in microcarrier culture. Scanning electron microscopy analysis demonstrated tenocytes populating the Cytodex™ type 3 microcarriers after 14 days of culture (A, ×200, scale bar: 100 µm) in different cell densities. Tenocytes showed mostly a stretched cell morphology on microcarriers (B, ×1000, scale bar: 20 µm) (A and B: 4.00 × 106 cells/1000 cm2 and 40 rpm stirring velocity). PI/FDA staining for cell vitality revealed the survival of cells after 14 days of culture (C, ×40, scale bar: 500 µm) and also showed the stretched cell morphology (D, ×100, scale bar: 200 µm) (C and D: 2.00 × 106 cells/1000 cm2 and 20 rpm stirring velocity). During microcarrier cultivation the mean glucose concentration of all cultures under different culture conditions decreased, and the lactic acid production increased, indicating cell proliferation (E, 2.00 × 106 cells/1000 cm2 and 20 rpm stirring velocity; F, 2.00 × 106 cells/1000 cm2 and 40 rpm stirring velocity; G, 4.00 × 106 cells/1000 cm2 and 20 rpm stirring velocity; H, 4.00 × 106 cells/1000 cm2 and 40 rpm stirring velocity).

Download figure to PowerPoint

Cell viability

Most of the cells colonizing the microcarriers were viable. PI/FDA staining confirmed the presence of only a few dead cells (Figure 2C). The cell viability was verifiable in all used samples for all tested culture conditions. None of the chosen parameters of different seeding densities and stirring velocities impaired the vitality of the cells. At 200-fold magnification the stretched fibroblastoid morphology of the adherent living cells was visible (Figure 2D).

Glucose/lactic acid metabolism of the tenocytes on the carrier

Analysis of the glucose/lactic acid concentration of all samples revealed a continuous decrease of the glucose concentration over the 14 days of tenocyte culture on Cytodex™ type 3 microcarriers in the spinner flask bioreactor system starting at 6.86 (±0.01) mmol L−1 down to 5.84 (±0.89) mmol L−1 for 2.00 × 106 cells/1000 cm2 and 20 rpm stirring velocity, 6.94 (±0.20) mmol L−1 down to 5.77 (±0.55) mmol L−1 for 2.00 × 106 cells/1000 cm2 and 40 rpm stirring velocity, 7.17 (±0.34) mmol L−1 down to 5.19 (±0.28) mmol L−1 for 4.00 × 106 cells/1000 cm2 and 20 rpm stirring velocity, and 7.00 (±0.22) mmol L−1 down to 5.16 (±0.83) mmol L−1 for 4.00 × 106 cells/1000 cm2 and 40 rpm stirring velocity, whereas the lactic acid concentration was continuously increasing from 3.05(±0.74) mmol L−1 up to 4.54 (±1.09) mmol L−1 for 2.00 × 106 cells/1000 cm2 and 20 rpm stirring velocity, 2.55 (±0.27) mmol L−1 up to 5.35 (±0.90) mmol L−1 for 2.00 × 106 cells/1000 cm2 and 40 rpm stirring velocity, 2.91 (±0.37) mmol L−1 up to 4.85 (±0.34) mmol L−1 for 4.00 × 106 cells/1000 cm2 and 20 rpm stirring velocity, and 2.58 (±0.20) mmol L−1 up to 5.81 (±1.19) mmol L−1 for 4.00 × 106 cells/1000 cm2 and 40 rpm stirring velocity (Figure 2E–H). Taken together, changes in glucose/lactic acid concentration indicated an increasing glucose consumption and metabolite formation suggesting cell proliferation independent from stirring velocity and seeding density.

Demonstration of “bead-to-bead” transition of tenocytes

The migration of tenocytes from densely to less densely populated microcarriers is one prerequisite for continuous cell culture. To confirm this “bead-to-bead” transfer in the continuously stirred spinner flask bioreactor system, SEM pictures were analyzed. Here, it could clearly be demonstrated that cells migrated from one carrier to another (Figure 3A). For further analysis, tenocytes seeded in five combinations of cell densities (2.00–6.50 × 106 cells) and stirring velocities (60–75 rpm) were cultivated on microcarriers for 14 days. After that time half of the colonized microcarriers were substituted with empty ones. We found that the freshly added microcarriers were immediately colonized by “bead-to-bead” transfer via direct contact between confluent and empty microcarriers. After further 14 days of culture, 82.3% (±8.4%) of the microcarriers were populated with cells (Figure 3B).

image

Figure 3. “Bead-to-bead” cell transfer after addition of empty microcarriers. Scanning electron microscopy analysis revealed the migration of tenocytes from one microcarrier to another during continuously stirred cultivation (A, exemplarily 2.00 × 106/1000 cm2 cells with 20 rpm, ×400, scale bar: 50 µm). After 14 days of cultivation, one half of the microcarriers were removed and substituted by empty carriers (B, arrow). During further cultivation for 14 days, 82.3% (±8.4%) of the unpopulated carriers were colonized with tenocytes (mean of five combinations of cell densities 2.00–6.50 × 106 cells and stirring velocities 60–75 rpm).

Download figure to PowerPoint

Gene expression profile of the tenocytes on the microcarriers

Expression of genes coding for tendon ECM proteins and the tendon marker scleraxis was determined using RTD-PCR analysis (Figure 4A–F). The gene expression profile for type I collagen (COL1A1) and type III collagen (COL3A1) did not significantly differ between monolayer and spinner flask cultures, except for COL1A1 which was significantly downregulated at 40 rpm and 4 × 106 cells (P = 0.015). Decorin (DEC) gene expression was significantly downregulated on the microcarriers when compared with the monolayer culture at 2.00 × 106 seeded tenocytes under both stirring conditions (P = 0.038, P = 0.0189). In a similar manner, the gene expression of COMP (2 × 106 cells at 20 rpm: P = 0.0166; 2 × 106 cells at 40 rpm: P = 0.0026; 4 × 106 cells at 40 rpm: P = 0.0022) and also of the cartilage proteoglycan aggrecan (ACAN) (2 × 106 cells at 20 rpm: P = 0.0023; 2 × 106 cells at 40 rpm: P = 0.0009; 4 × 106 cells at 20 rpm: P = 0.0038; 4 × 106 cells at 40 rpm: P = 0.0001) significantly decreased on the microcarriers when compared with the monolayer culture.

image

Figure 4. Gene expression analysis for ECM markers and scleraxis. Type I (A, COL1A1) and type III collagen (B, COL3A1), decorin (C, DEC), aggrecan (D, ACAN), cartilage oligomeric matrix protein (E, COMP) and scleraxis (F, SCX) gene expression was determined in human tenocytes cultured for 14 days on Cytodex™ type 3 microcarriers at a seeding density of 2.00 or 4.00 × 106 cells/1000 cm2 and stirring velocities of 20 or 40 rpm in comparison to monolayer (ML) culture by RTD-PCR. The gene expression of the tenocytes in monolayer culture was normalized (set to 1.00). Data of three independent experiments using human tenocytes derived from three different donors are shown. * P < 0.05; P** < 0.01, P*** < 0.005.

Download figure to PowerPoint

A divergent influence of the stirring velocity could only be observed for the COMP gene expression, whereby the higher stirring speed leads to a stronger suppression of its gene expression. This effect was significant for 4.00 × 106 seeded cells. When comparing the tested seeding densities no significant difference became evident.

Type I collagen and decorin expression by the tenocytes on the microcarriers

Cell associated type I collagen (Figure 5A) and decorin (Figure 5B) formation could be detected in tenocyte spinner flask bioreactor cultures. The formation was independent of the tested culture conditions. Aggrecan formation could not be demonstrated in tenocytes cultured on the microcarriers using immunofluorescence labeling.

image

Figure 5. Type I collagen and decorin immunofluorescence labeling of tenocytes cultured on microcarriers. At day 14, type I collagen (A) and decorin (B) immunolabeling demonstrated the formation of both ECM proteins by tenocytes colonizing the Cytodex™ type 3 microcarriers. This was independent of the chosen culture conditions and is shown here in representative pictures (2.00 × 106 cells/1000 cm2, 40 rpm stirring velocity, ×100, scale bars: 200 µm). Tenocyte nuclei were counterstained using DAPI.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. Literature Cited

Tissue engineering techniques require high cell numbers and hence, time-consuming and cost-intensive cell expansion procedures. Cell amplification for tissue engineering is usually performed using monolayer cultivation. At confluence and between each passage, tenocytes have to be detached using trypsin/EDTA, whereby the cells lose most of their freshly synthesized ECM. Additionally, it is well known that extensive monolayer culturing can lead to a phenotypic drift in tenocytes.[6-8] Microcarrier culture provides a high culturing surface in a relatively small volume, and thus presents many advantages over monolayer culturing. Therefore, microcarrier cultures have been applied in different bioreactor systems for the expansion of several anchorage-dependent cell types, such as chondrocytes and mesenchymal stem cells (MSC). These cell types survived for 4 weeks on the carriers. Published data for human tenocyte expansion on microcarriers in a bioreactor system are so far not yet available. The results of the present study demonstrate that human hamstring tenocytes can be successfully expanded on Cytodex™ type 3 microcarriers. In contrast to monolayer culturing, bioreactors provide a more continuous cell expansion. We show that tenocytes fulfill the prerequisite for such continuous culturing, namely “bead-to-bead” migration from one carrier to another. Furthermore, this study reveals some aspects of the behavior of tenocytes cultured on microcarriers for at least 2 weeks regarding glucose/lactic acid metabolism, expression of typical tendon-derived genes, and synthesis of type I collagen and decorin.

Tenocytes rapidly adhered on the microcarriers, retained a typical fibroblastic morphology, survived over the whole culturing period of 2 weeks, and easily colonized newly added empty beads by “bead-to-bead” transition, which indicates their high migratory capacity. The estimated adherence time of 3 h is comparable to observations with other mesenchymal cell types. For the cultivation of bovine chondrocytes on Cytodex™ type 3 and other dextran-based microcarriers, a period of 3 h was suggested as an optimal seeding time.[12] Also for porcine bone marrow-derived MSC, 3 h was estimated as the optimal seeding time leading to the highest number of adhered cells, compared with other adhesion times.[13] For human MSC, a minimum adherence time of 2 h was suggested with 4 h achieving a maximum seeding density.[16]

Different stirring velocities and seeding parameters can be chosen for optimization of microcarrier culture. In the present study, two cell densities and two stirring velocities were selected. Varying the numbers of microcarriers in order to change the surface area for cell adherence was not analyzed here. Also, changes in the stirring profile considering discontinuous stirring with resting times, or changes in direction of stirring rotation, were not evaluated. Stirring is necessary for homogenous cell distribution and provides some mechanical stimulation, which is of high advantage for culturing mechanosensitive cells such as tenocytes. On the contrary, high stirring velocities exhibit increasing shear stress and impair “bead-to-bead” transition of the cells, since the contact time of the beads is reduced. Also, the Cytodex™ type 3 microcarrier can become damaged when using high stirring velocities. For 20 and 40 rpm, the number of damaged microcarriers did not increase over the whole time of cultivation and the higher stirring speed, 40 rpm, seemed not to affect tenocyte vitality. Even with higher velocities a “bead-to-bead” transfer of tenocytes was detected. Besides the stirring velocity, the impeller geometry influences the cell proliferation and cell survival during cultivation.[16] In this study, we used two glass pendula that cause mild shear forces while maintaining optimum mixing compared to a paddle impeller, which develops considerable shear forces.[17]

For cell growth over 2 weeks, cell proliferation could be measured for all chosen conditions (2.00 or 4.00 × 106/1000 cm2 cells and 20 or 40 rpm). Furthermore, tenocytes revealed continuous metabolic activity, metabolizing glucose into lactate, and maintained many of their differentiated functions such as scleraxis (SCX) gene expression. After an adherence time of 3 h and cultivation on microcarriers for 1 day, all cell numbers decreased compared to the initially seeded cell numbers. This has also been observed for other cell types, such as porcine MSC.[13] In general, the resulting total number of cells after the microcarrier culture was higher for 4.00 × 106/1000 cm2 cells seeded, due to the higher initial seeding density. Comparing the PD, a higher mean value was achieved using the higher stirring velocity of 40 rpm compared to the lower velocity of 20 rpm when using the same cell seeding density. When both initial cell seeding densities were compared under the same stirring conditions, a lower initial cell seeding density resulted in a higher PD. Therefore, the highest PD was achieved using 2.00 × 106/1000 cm2 cells with 40 rpm stirring velocity.

A slight influence of microcarrier culture on tenocyte gene expression profile could be detected when compared with monolayer culture. The expression of COL1A1 and COL3A1, both coding for the main tendon ECM proteins type I and type III collagen, remained mainly unaffected on the microcarriers except for a downregulation of the COL1A1 expression observed at 40 rpm stirring combined with 4.00 × 106 cells as an initial seeding density, when compared with monolayer culture. Previously, we found that COL1A1 expression is upregulated in monolayer culture when compared with native tendon.[10] In contrast to collagen, decorin (DEC), aggrecan (ACAN), and COMP expressions were significantly impaired when 2.00 × 106/1000 cm2 cells were seeded onto the carriers. In a previous study, we compared the expression of these genes in native tendon and monolayer culture, whereby their suppression could be detected in cultured tenocytes.[10] Looking at 4.00 × 106/1000 cm2 cells at 20 rpm, the downregulation of DEC and COMP expression was less pronounced on the microcarriers, recommending these parameters for future culturing; this is supported also by the observation that SCX expression was highest under these conditions. The reason for inhibition of proteoglycan and COMP expression by bioreactor culture remains unclear.

One limitation of this study is that a monolayer expansion phase preceded the spinner flask culture. It is probable that monolayer culturing might induce irreversible alterations in the gene expression profile. Hence, future studies should follow, based on tenocytes directly cultured on microcarriers in a bioreactor system immediately after their isolation. This experimental setting would show whether the cells maintain their tenocytic expression profile more closely when directly transferred into microcarrier culture.

Cytodex™ type 3 carriers served as a model system in the present study. They consist of a non-porous cross-linked dextran matrix coated with a thin layer of denatured porcine type I collagen. The size of each carrier is between 141 and 211 μm with an average of 175 μm and in a dry weight of 1 g, about 3 × 106 carriers are contained. The use of other materials for microcarrier preparation might stimulate tenocyte proliferation and modify their expression profile. Other well-characterized, dextran-based commercial available microcarriers for mesenchymal cells are Cytodex™ type 1 and −2. Cytodex™ type 1 consists of a dextran matrix with substituted N,N-dimethylaminoethyl groups. Cytodex™ type 2 is composed of a dextran matrix with a surface layer of N,N,N-trimethyl-2-hydroxyaminopropyl groups. Other commercially available microcarriers are based on materials similar to the surface coating of Cytodex™ type 3. Other prominent commercially available microcarriers are the porcine gelatin-based porous carriers Cultispher-G and -S.

Degradable microcarriers consisting of biomaterials favoring tenocyte specific gene expression are also of high interest. For tissue engineering applications, poly(lactic-co-glycolic acid) (PLGA) scaffold cultures under static conditions showed promising results.[10] Moreover, PLGA-based carriers have already been used for chondrocyte cultivation.[18] The use of PLGA-based microcarriers is not only considered for tenocyte expansion but also for subsequent tenocyte differentiation in tissue engineering applications. That would enable an application without preceding enzymatic cell harvesting. Besides non-porous PLGA microspheres,[18] porous PLGA microcarriers could also be of interest. They exhibit an increased surface area to achieve a higher cell density, and they have already been used as an injectable cell delivery system for adipose tissue engineering in a rabbit model.[19] Furthermore, microcarriers based on polylactic acid (PLA) are also already established. Collagen-coated PLA microcarriers have been successfully used for chondrocyte cultivation and as injectable cell carriers.[20] In particular, microcarriers consisting of collagen might be of interest, because collagen is the main component of natural tendon ECM. Furthermore, the coating of the carriers with particular growth factors as a slow release system might provide a basis to improve/stabilize tenogenic differentiation. For clinical applications, tenocytes seeded on microcarriers could also be combined with scaffolds, to render them immobilized when injected, to support tendon defect healing.

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. Literature Cited

In summary, we report for the first time that hamstring-derived human tenocytes can be cultivated on microcarriers. Seeded on Cytodex™ type 3 microcarriers, the cell number increased over 2 weeks of cultivation, while tenocytes still synthesized type I collagen and decorin. According to the gene expression analysis, type I (COL1A1) and type III collagen (COL3A1) expression remained nearly unchanged; decorin (DEC), aggrecan (ACAN) and COMP expression decreased. Looking at the different seeding and stirring conditions, a tenocyte adhesion time of 3 h, seeding of 4.00 × 106 cells/1000 cm2 Cytodex™ type 3 microcarrier surface, and a stirring velocity of 20 rpm led to the most promising results, whereby a higher expression rate of genes coding for tendon typical ECM proteins could be observed. However, based on the number of PD, the application of the parameter set of 3-h adhesion, 2.00 × 106 cells/1000 cm2 microcarrier surface, and 40 rpm resulted in a slightly more effective proliferation. Direct “bead-to-bead” transfer, the prerequisite for continuous cultivation, was verified by colonization of freshly added microcarriers and by SEM. Taken together, this study provides the scientific and technical basis for the development of an automated and controlled bioreactor system for continuous cultivation of adherent growing human tenocytes, beginning with the biopsy and ending with the cell product for clinical application.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. Literature Cited

The authors thank Marion Lemke, Barbara Walewska, and Johanna Golla for excellent technical assistance. This study was supported by the Berlin-Brandenburg Center for Regenerative Therapies-BCRT (Bundesministerium für Bildung und Forschung-1315848A). The authors also acknowledge the support of the Sonnenfeld Foundation. They see no possible conflict of interests, financially or otherwise.

Literature Cited

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. Literature Cited