To build microcavities, we used the soft lithography technique. Cost-efficiency and small detail production are some advantages this method offers. The microfluidic pattern was created and printed in high resolution. The pattern consisted of microchannels and microcavities. The generated transparency mask was used in the photolithography of the SU-8 photoresist (Microchem, USA). A layer of negative photoresist was spin-coated onto the 100-mm glass wafer (Shin-Etsu Group, Japan). The master was developed by immersion of the SU-8 wafer into developer propylene glycol methyl ether acetate (PGMEA) (Sigma, USA) under agitation for 10 min to remove the non-cross-linked photoresist. The polydimethylsiloxane (PDMS) (15:1 elastomer: curing agent; Sylgard 184 Silicone elastomer kit; Dow Corning, USA) was poured over the master and incubated overnight in an oven at 60o C. The PDMS replica was oxidized by plasma treatment to make the PDMS mold surface hydrophilic. The negatively charged PDMS surface was exposed to an ethylene dichloride/N-hydroxysuccinimide (EDC/NHS) solution (10 mg/ml) for 24 h and then rinsed thoroughly with distilled water. Diluted fibronectin solution was added to the PDMS surface and incubated for 24 h at 4 °C. The microcavities and microchannels system was brought into a cell culture cabinet and then sterilized by rinsing with 70% ethanol and rinsed thoroughly with sterile phosphate-buffered saline (PBS). Pneumatic micropumps were used to transport the cells and media in the microchannels.

Healthy child-bearing mothers gave informed consent, approved by the local Ethical Committee of the Tarbiat Modares University, for cord blood donation. Cord blood samples were collected at the time of birth. Briefly, after the baby had been delivered, the cord was clamped at both distal ends and cut. A needle was injected into the umbilical vein at a prepared site to harvest the UCB into a sterile 250 ml blood bag (BesaatQom Iran) containing 21 ml of anticoagulant citrate phosphate dextrose (CPD) solution by gravity flow. The cord blood bag was gently rotated to mix the anticoagulant with the blood. The collected UCB was immediately transferred to our laboratory. The UCB unit was processed within 2 h. Isolation of UCB-mononuclear cells (UCB-MNCs) was performed. Briefly, red blood cells were depleted by hydroxyethyl starch (Fresenius Kabi) within 1 h. Subsequently, the UCB-MNCs were isolated from the whole blood by density gradient separation medium (1.077 g/ml, Ficoll-Hypaque). In this procedure, red blood cells and plasma were removed and an MNC fraction containing hematopoietic stem cells was isolated. Positive selection of CD133 antigen-carrying cells was performed by the MidiMACS (MiltenyiBiotech, Germany) magnetic separation columns according to the protocols provided by the manufacturer. To further purify CD133+ cells, isolated cells were subjected to a second column separation. The purity of CD133+ cells in the selected population was above 95%. The CD133+ cell viability and number were determined by trypan blue in a hemacytometer.

To determine the purity of CD133 cells that were isolated with the magnetic-activated cell sorting (MACS) and cells expanded with different methods, cells were harvested and washed in a PBS containing 1% of Stain Buffer (FBS) and were each labeled with 10µl of phycoerythrin (PE)-conjugated anti-human CD133 (Miltenyi Biotech, Germany), following manufacturer instructions. Isotype controls were included to confirm specificity. After a wash, the Becton Dickinson FACSCalibur flow cytometer (Becton Dickinson, San Jose, CA, USA) was used to distinguish the fluorochrome antibody-tagged cells. Six thousand to 10,000 events were counted. The FACSCalibur data were analyzed using the Cyflogic (CyFlo Ltd, FINLAND).

The CD133+ cells were seeded at 104 cells/ml in 24-well tissue culture plates by the conventional method and 104 cells/ml in the microfluidic bioreactor system in stem span serum-free medium (Stem Cell Technologies Inc., Vancouver, Canada) and were cultured for 12 days. Both mediums were supplemented with the proper mixture of growth factors and cytokines (50 ng/ml recombinant human stem cell factor, 20 ng/ml recombinant human thrombopoietin and 50 ng/ml recombinant human FLT3-L). All the cytokines were purchased from R&D Systems (Minneapolis, USA). The conventional culture was performed, as the control, in 24-well tissue culture plates (Nunc, Denmark) with 2 ml of culture medium per well. Both cultures were maintained at 37ºC in a humidified incubator with 5% CO2. The medium was replaced with the cytokine-containing fresh medium every 3 days. After culturing for 12 days, all expanded cells under conventional conditions were harvested and counted with a hemocytometer. In the microfluidic device, the covering layer of PDMS was removed aund the content of desired micro-cavities was harvested with a micropipette by flushing with sterile phosphate-buffered solution with ethylenediaminetetraacetic acid (PBS-EDTA).

For the colony-forming cell (CFC) assay, 12 days after culture, cells were assayed for normal hematopoietic progenitor cell differentiation using an in vitro methylcellulose-based CFC assay (MethoCult® H4435 Enriched, StemCell Technologies) as per manufacturer instructions. Cells were seeded at 1 × 104 cells/ml in 24-well plates containing Iscove Modified Dulbecco Media (IMDM), 20% of FBS, 0.9% of methylcellulose and required cytokines. The plates were incubated at 37°C for 14 days in a humidified atmosphere containing 5% of CO2. After 14 days, colonies were counted using an inverted microscope. Colonies containing more than 50 cells were scored as colony-forming units (CFUs).

The long-term culture-initiating cell (LTC-IC) assay (Stem Cell Technologies) was performed according to manufacturer instructions. In brief, mitomycin C-treated feeder cells of murine mesenchymal stem cells were established in 35 mm tissue culture dishes at a seeding density of 3 × 105 cells/ml. Three hours after treatment with mitomycin C, the superficial layer was exchanged with a suspension of 5 × 105 cells/ml in the human long-term culture medium (HLTM) and incubated at 37°C, 5% of CO2 for 35 days. Once a week, half of the medium was extracted and replaced by new HLTM and fresh hydrocortisone.

Following 5 weeks of culture at 37°C, the contents of each dish were harvested, and 5×104 cells were plated into methylcellulose-containing dishes (MethoCult, H4230; StemCell Technologies). The dishes were cultured at 37°C, 5% of CO2, for 18 days until cells were scored.

Total cellular ribonucleic acid (RNA) of freshly isolated cells (day 0), cells expanded in 24-well tissue culture plates (day 12) and cells expanded in the microfluidic system (day 12) was extracted using the RNeasy Mini Kit (Qiagen, Valencia, CA) according to the protocols provided by the manufacturer. The quantity and purity of the RNA were determined by spectrophotometric analysis at 260 nm and 280 nm and the quality of all RNA samples was assessed by electrophoresis in 1.2% agarose gel. Subsequently, the complementary deoxyribonucleic acid (cDNA) was synthesized using the RevertAid™ First Strand cDNA Synthesis Kit (Fermentas, USA).

The quantitative reverse transcription polymerase chain reaction (RT-qPCR) was performed using cDNA pools from freshly isolated and expanded CD133+ cells. RT-qPCR assays were developed to determine CXCR4 expression levels using the Sybr Green method. All RT-qPCR reactions were performed on the STEP ONE ABI platform (Applied Biosystems, USA). Gene-specific primers were designed according to Primer-Express software. Forward and reverse primers were designed in different exons or exon-exon junction to eliminate DNA contamination (Table 1). The Sybr Green Master Mix was purchased from ABI (SYBR green PCR kit; Applied Biosystems, Foster City, CA). The relative initial amount of mRNA of a particular gene was extrapolated from the standard curve. To plot the standard curve, we used a pool of all the samples serially diluted in four log2 steps and run in parallel to the samples. The total volume of each reaction was 10 μl, containing 300 nM of each forward and reverse primer and 125ng of cDNA. Appropriate negative controls were run for each reaction. The PCR amplification included the first step of 10 min at 95°C of denaturation, followed by 45 cycles of amplification at 95°C for 15 s and 60°C for 60 s. All of the reactions were performed in triplicate. The relative obtained amounts were normalized to the housekeeping gene beta-actin. The relative change in gene expression was calculated by dividing each gene expression signal by that of the internal standard (beta-actin) in each sample.

Specimens were prepared for SEM analysis on day 4 using the following protocol. The superficial medium was gently removed and cell-microfluidic constructs were fixed two times. First, in the PBS with 2.5% of glutaraldehyde for 90 min at 4°C and then a second fixation in 1% osmium tetroxide for 2 h at 4 °C. Samples were dehydrated by soaking in a graded ethanol series of 30%, 50%, 70%, 90% and 100%. Samples were finally vacuum dried. The construct was mounted on an SEM specimen stub, sputter-coated with gold-palladium and imaged by SEM (Electronic Engineering College, Tehran University).

The percentage of CXCR4 positive cells was analyzed using the quadrant statistics of the dot plot (co-expression analysis) and expressed as the mean ± standard error in the mean of at least three experiments. Differences between populations of cells were analyzed using a two-tailed Student's t-test.

Functional FN-coated microcavities were fabricated and characterized for their usage in the investigation of UCB-HSC expansion. The microstructure (LPDMS) molds were replicated from lithographically designed silicon master structures. The human fibronectin fibrils were covalently attached via EDC/NHS onto the PDMS replicas containing channels and micro-cavities of approximately 150µm in diameter. The homogeneity of the FN layers was analyzed using scanning electron microscopy, showing a smooth surface on microcavities (Figure 1). Additionally, results from the SEM imaging showed the microfluidic bioreactor has separate, symmetrical cavity-like structures (Figure 2).

Figure 1.

Homogeneity of FN layer on PDMS before FN immobilization (A) and after FN (B) with × 5000 magnification in SEM.

(0.1MB).

Figure 2.

SEM image of microcavity structure with × 50 (A) and × 150 (B) magnification. Cavities are sized at 50µm.

(0.34MB).

On average, the number of CD133+ cells separated by the MACS was 106. Purity assessments of enriched CD133+ cell fractions were analyzed by a flow cytometer. Flow cytometer analysis after isolation showed that the mean purity of separated cells among samples was 97% for CD133+ cells (Figure 3).

Figure 3.

Purity assessment of CD133+ cells from UCB-MNCs after MACS separation.

(0.08MB).

The first goal of this study was to determine the expansion rate of CD133+ cells in microfluidic bioreactors. The CD133+ cells extracted from cord blood were cultured in duplicate in microfluidic devices. At the same time, CD133+ cells were also cultured in 24-well plates for comparison purposes. The expansion of the total nucleated cells (TNCs) and CD133+ cells (n = 5) was calculated. On day 12, the microfluidic system expansion results showed significant improvement, compared to those of the tissue culture plate (TCP), with p < 0.05. The TNCs cultured in the TCP generated a 45-fold expansion, while it expanded 65-fold in the microfluidic system. On day 12, the CD133+ cell expansion was 4.5-fold in the TCP and 8.5-fold in the microfluidic system (Figure 4).

Figure 4.

The proliferation of total cells and CD133+ cells. a) TNC cultured in TCP generated 45-fold expansion, while it expanded 65-fold in the microfluidic system (p < 0.05). b) CD133+ cells yielded hematopoietic colonies after 12 days of culture in methylcellulose containing CFU-E, CFU-GEMM and CFU-GM in a microfluidic system (* p < 0.05)

(0.26MB).

On day 12, expanded cells in the TCP and microfluidic bioreactor were cultured in methylcellulose to compare their colony-forming potential. After 14 days of culturing in methylcellulose, the number of burst-forming unit-erythroid (BFU-E), colony-forming unit-erythroid (CFU-E), granulocyte-macrophage progenitor (CFU-GM) and colony-forming unit – granulocyte, erythrocyte, monocyte and megakaryocyte (CFU-GEMM) colonies were counted by their morphology. It was shown that microcavities of the bioreactor were more efficient in expanding CD133+ cells, but these results were statistically insignificant (p >0.05). This test was repeated three times and, on average, the tissue culture produced 123 colonies and the microfluidic bioreactor produced 149 colonies. The appearance of the colonies is shown in Figure 4.

A critical point of in vitro HSC expansion is the homing ability of these cells and thereby their potential in reviving hematopoiesis. The CXCR4 is a vital factor in such a process. To better understand the role the microfluidic system plays in CXCR4 expression of the expanded hematopoietic cell, we analyzed cell surface expression levels of CXCR4 molecules on CD133+ cells and CXCR4 messenger RNA (mRNA) expression in CD133+ cells cultured in different systems. We first examined the cell-membrane expression of the CXCR4 in the expanded cells by flow cytometry using a phycoerythrin (PE)-conjugated anti-CXCR4. On average, the percentage of freshly purified CD133+ cells (day 0) that were positive for CXCR4 was 77.6% ± 15.2%. Therefore, it was considered to be used as a baseline to study changes in CXCR4 expression during the culture of CD133+ cells. The median percentage of CXCR4 expression in the expanded cells after 12 days was lower in the conventional system, compared to the microfluidic system, with a median expression of 8.9% in the microfluidic system, compared to 7.7% in the conventional system (p < 0.05).

In the second step, in order to assess CXCR4 expression in hematopoietic stem cells, real-time PCR was conducted on RNA extracted from fresh HSC of cord blood and HSC expanded under the conventional and microfluidic system. Threshold cycle values were normalized to the glyceraldehyde 3-phosphate dehydrogenase (GAPDH) expression and the level of CXCR4 expression was calculated relative to the freshly enriched CD133+, as the untreated control. According to our experiment, the CXCR4 gene expression in the microfluidic system remained the same, while expression in the traditional system was reduced. However, the results were not statistically significant (data are not shown). This illustrated the fact that besides surface expression, mRNA levels of CXCR4 were also increased. Thus, the CXCR4 gene is up-regulated in the microfluidic system, compared to the traditional type.