Injectable hydrogel delivery plus preconditioning of mesenchymal stem cells: exploitation of SDF-1/CXCR4 axis toward enhancing the efficacy of stem cells' homing

Cell isolation and culture

Discarded adipose tissues were obtained from healthy women, who underwent aesthetic liposuction. The washed lipoaspirates were treated with freshly prepared collagenase solution (1 mg collagenase type I [Invitrogen, Carlsbad, CA]), 10 mg of BSA (Biowest, Nuaillé, France) as carrier protein and 2 mM CaCl₂ were dissolved in 1 mL PBS (for each 3 mL lipoaspirates) and incubated for 45 min at 37°C with gentle agitation in a shaker water bath. The mixture was centrifuged at 800 g for 10 min and the infranatant pellet (stromal vascular fraction cells; SVF) was separated and washed with PBS and re-centrifuged at 400 g for 6 min. The SVF pellet was resuspended and cultivated in low glucose Dulbecco's modified Eagle (DME)-medium (Biowest) containing 2 ng/mL basic fibroblast growth factor (bFGF, Royan Institute, Iran), 10% fetal bovine serum (FBS, Invitrogen), 100 µg/mL streptomycin and 100 U/mL penicillin (Invitrogen). Resuspended pellets from each of the 10 mL lipoaspirates were seeded in one 75 cm² tissue culture treated flask. Medium was exchanged every 2–3 days and the cells became ∼90% confluent after 3–5 days. Then, the cells were trypsinized and replated into new flasks at a density of 5000–10000 cells/cm². ASCs were characterized by morphological analysis, FACS pattern and standard differentiation assays. The cells at passage 3 were used for all experimental studies.

Study design and delivery routes of ex vivo expanded ASCs into animal model

After the general anesthetization of Wistar rats with ketamine and xylazine, contractually, in all experimental groups (five rats per each group) as shown in Figure 1 (groups A–G), 1 mL of CH-GP-HEC hydrogel containing SDF-1α (100 ng/mL) was injected subcutaneously on the right side of the rat's back and 1 mL of empty CH-GP-HEC hydrogel on the left side. In the meantime, DiI-labeled human ASCs (pretreated or untreated) were delivered from three different routes (Figure 1). The animals were treated and maintained according to the standard guidelines of Animal Care and Use Committee of Ferdowsi University.

Adipogenic and osteogenic differentiation

ASCs at passage 3 were induced to differentiate toward the adipogenic and osteogenic lineages according to the method of previous work (da Silva Meirelles et al., 2008). Briefly, the cells were seeded in six-well plates and grown to 80–90% confluency for adipogenic differentiation study. Adipogenesis was induced by culturing ASCs for 14 days in the adipogenic differentiation medium containing DMEM supplemented with 10% FBS, 200 µM indomethacin, 1 µM dexamethasone, and 10 mM β-glycerophosphate (all reagents were from Sigma–Aldrich, Munich, Germany). The differentiation medium was changed every third day and assessed using an Oil Red O stain (Sigma–Aldrich) as an indicator of intracellular lipid droplets and counterstained with Hematoxylin (Merck, White House Station, NJ). Osteogenesis was performed by inducing ASCs at 100% confluency for 21 days in an osteogenic medium containing DMEM supplemented with 10% FBS, 50 µM ascorbate-2-phosphate (Sigma–Aldrich), 0.1 µM dexamethasone, and 10 mM β-glycerophosphate. Osteogenesis was assessed by staining the calcium mineralization of extracellular matrix and alkaline phosphatase (AP) activity of osteoblast with Alizarin Red S (pH 4.1–4.3) stain (Sigma–Aldrich) and BCIP/NBT substrate (BD Bioscience, San Jose, CA), respectively.

Flowcytometric analysis

Flow cytometry was performed to characterize the ASCs and also to determine the expression of CXCR4 on the surface of both untreated and pretreated ASCs (passage 3). Cell cultures were prepared by trypsin/EDTA digest and washed twice with cold D-PBS, resuspended in D-PBS buffer containing 0.5% FBS at a concentration of 1 × 10⁶ cells/mL. One hundred microliter (1 × 10⁵ cells) of the suspension was transferred to a 1.5 mL microtube and was incubated with either respective antibody or isotype-matched control antibody. The cells were vortexed gently and incubated at RT for 45 min in the dark. Then, 400 μL of D-PBS buffer was added to each tube and the data was acquired by a flow cytometer (FACScalibur, BD Bioscience) and analyzed with Flowing Software 2.5 program.

Antibodies used in flow cytometry analysis to characterize ASCs were rabbit anti-CD44 polyclonal antibody, rabbit anti-CD34 polyclonal antibody (all from antibodies-online, Aachen Germany), mouse anti-CD90 monoclonal antibody, rabbit anti-CD11b polyclonal antibody (all from Novus Biologicals, Littleton, CO), rabbit anti-CD105 polyclonal antibody, rabbit anti-CD45 polyclonal antibody (all from Bioss, Inc., Woburn, MA) and for analysis of CXCR4 expression (mouse anti-CXCR4 monoclonal antibody; from Neuromics, Edina, MN) on the surface of ASCs. For each antibody, respective isotype control was used.

Migration assay

ASCs chemotaxis assay was performed in 24-well plates based on the Boyden chamber principle using a polycarbonate membrane (8 µm pores size, Costar, corning, Tewksbury MA). ASCs were suspended in serum-free DMEM containing 0.5% BSA at a concentration of 25 × 10⁴/mL. ASCs suspension at a concentration of 25 × 10³ cells/mL in 100 μL was loaded onto the top wells, whereas 0.6 mL SDF-1α (100 ng/mL) was added to the bottom chamber. After 18-h of incubation at 37°C, 5% CO₂, the non-migrating cells were completely wiped off from the top surface of the filters by cotton swab, and the migrating cells adhering to the undersurface of the filter membrane were stained with DAPI and counted. To demonstrate that ASCs migration toward SDF-1α is correlated with the upregulation of CXCR4, the cells were preincubated with AMD3100, a CXCR4-selective antagonist (De Clercq, 2003), at a concentration of 10 μg/mL for 30 min at RT. All experiments were performed three times in triplicate wells. At least five fields were counted from each well and the average number of migrating cells was determined.

Conventional and real-time PCR

Total RNA was extracted using Tripure reagent (Roche, Basel, Switzerland) according to the manufacturer's instructions. For all RNA samples, the quantity and purity were determined using the Nanodrop ND-1000 spectrophotometer (Nanodrop, BIO-TEK, Winooski, VT), integrity by gel eleterophoresis for 28S and 18S rRNA. Reverse transcriptase–polymerase chain reaction (RT-PCR) of CXCR4 and β-actin (housekeeping gene) was performed using 1 µg of total RNA for first strand cDNA synthesis. False positives and DNA contamination were controlled by omitting reverse transcriptase in control reactions. Conventional RT-PCR was performed to 35 cycles, 94°C for 30 seconds, 57°C for 40 s and 72°C for 30 s. Primers for RT-PCR were: β-actin (forward) GCTCAGGAGGAGCAAT (3′) and (reverse) GGCATCCACGAAACTAC (3′) with 187 bp product length; CXCR4 (forward) TCTGGAGAACCAGCGGTTAC (3′) and (reverse) GACGCCAACATAGACCACCT (3′) with 501 bp product length.

Quantitative real-time PCR analysis was performed using SYBR Green PCR Master Mix (Takara, Shiga, Japan) according to the manufacturer's instructions. Briefly, for relative quantification, we used a method that compares the amount of target normalized to a housekeeping reference gene (β-actin). Melting curve analysis was performed to check for a single amplicon. The formula used was 2^−ΔΔCt, representing the n-fold differential expression of a specific gene in a treated sample compared with the control sample, where C_t is the mean of threshold cycle at which the amplification of the PCR product is first detected. Before using this method, we performed a validation experiment comparing the standard curve of the reference and the target to demonstrate that the efficiencies were approximately equal. The identity of the amplified products was checked by electrophoresis. The primer specific for real-time PCR was: CXCR4 (forward) ATCCCTGCCCTCCTGCTGACTATTC (3′) and (reverse) GAGGGCCTTGCGCTTCTGGTG (3′) with 231 bp product length. The primer sequence for β-actin was the same as for RT-PCR.

PCR was also performed, for tracking of hASCs from three different delivery routes, on DNA extracted from implanted hydrogels, lungs and spleen by human specific primer for CXCR4 gene. PCR of rRNA 18S was also used to prove infiltration of cells into the scaffold. Primers for PCR were: rRNA 18S (forward) ACTCAACACGGGAAACCTCA (3′) and (reverse) AACCAGACAAATCGCTCCAC (3′) with 123 bp product length; human specific CXCR4 (forward) GAGGAGTTAGCCAAGATGTGAC (3′) and (reverse) CCCATAGTGACTTCATTATATCCTTC (3′) with 116 bp product length.

Preparation of CH-GP-HEC thermosensitive injectable hydrogel

CH-GP-HEC injectable hydrogel was prepared according to our previous study (Naderi-Meshkin et al., 2014a). Briefly, the chitosan solution was obtained by dissolving 150 mg of chitosan (medium viscosity, 84% degree of deacetylated; Polysciences, Warrington, PA) in 9 mL of 0.1 M acetic acid solution, prepared in PBS and subjected to autoclave sterilization. To prepare the experimental mixtures, 1 mL of 1.3 M GP solution (filter-sterilized) was added dropwise to 9 mL of chitosan while stirred on ice for 15 min under aseptic conditions. Separately, HEC (Sigma–Aldrich) was dissolved at a concentration of 20 mg/mL in DMEM and sterilized by microfiltration. The SDF-1α (100 ng/mL) or cell pellet of ASCs (1 × 10⁷ cells/mL) was homogeneously resuspended in 1 mL HEC, and then mixed with 4 mL chitosan-GP at 4°C. Immediately, the thermo-gelling cell suspension or/and SDF-1α-containing solution was injected into dorsum of rat (1 mL/injection site).

Histological assessment of transplanted animals

Fifteen days after transplantation, the rats were anesthetized (ketamine 50 mg/mL and xylazine 5 mg/mL, i.p.), perfused with 4% buffered paraformaldehyde (Haddad-Mashadrizeh et al., 2013) and the remaining implanted scaffolds were individually dissected and removed from the dorsum of rats. The tissues from lung and spleen were also removed. Immediately, half of all samples from each implanted scaffold, lung and spleen were processed for RNA analysis and the other half was fixed in 4% PFA, embedded in paraffin and used for histological analysis. Then, the embedded specimens were sectioned (4 μm) along the longitudinal axis of the implant, and the sections were sought for presence of DiI-labeled ASCs under an Olympus IX71 fluorescent microscope (Olympus, Tokyo, Japan) equipped with digital camera.

Statistical analysis

The SPSS v.21 statistical program was used to analyze the data. The values are reported as mean ± SD. Samples were obtained from at least three independent experiments and used to calculate the mean value. Significance of the data was examined in the level of P ≤ 0.05 using the one-way ANOVA and the Tukey test.

ASCs characterization

Ex vivo expanded ASCs derived from lipoaspirates reached to passage 3 (P3) after 9–12 days of culture. They exhibited spindle-shaped morphology as shown in Figures 2A and 2B. Characterization of ASCs at P3 was performed by differentiating into mesenchymal tissues such as adipose (Figures 2C and 2D) and bone (Figures 2E–2H). MSC-derived adipocytes showed red lipid droplet by Oil Red O staining as compared with undifferentiated control. Undifferentiated MSCs (without extracellular calcium deposits) were slightly reddish by Alizarin Red S (Figure 2E) and showed weak alkaline phosphatase activity (Figure 2G), whereas MSC-derived osteoblasts (with extracellular calcium deposits) were bright orange-red (Figure 2F) and featured very high phosphatase activity (Figure 2H). In addition, ASCs were characterized by positive expression of CD44, CD105, CD90 and negative expression of CD34, CD45, and CD11b surface markers (Figure 2I).

CXCR4 has dynamic surface and intracellular expression

Flow cytometry was performed to detect the expression of SDF-1α receptor (CXCR4) on the surface of ASCs at passage 3 (Supplementary Figure S1). The results showed that CXCR4 could not be detected from extracellular staining of ASCs, even those pretreated with DFX, CoCl₂, VPA, LiCl, and physical hypoxia; while 95.2% of a white blood cells population as positive control, expressed CXCR4 on their surface (Supplementary Figure S1). On the contrary, analysis of CXCR4 by in vitro migration assay showed that CXCR4 can be functionally represented on the cell membrane of DFX-pretreated ASCs to mediate SDF-1α-dependent homing within 24 h of reculturing after trypsinization (Supplementary Figures S2E and G). Furthermore, introduction of CXCR4 selective antagonist (AMD3100) has shown to significantly inhibit the chemotaxis of the ASCs toward SDF-1, confirming that SDF-1/CXCR4 interactions are responsible for the migratory response in DFX-pretreated ASCs (Supplementary Figures S2F and S2G). In addition, conventional and real time RT-PCR confirmed that ASCs express intracellular CXCR4 (Figures 3A and B).

Gelation time and degradation of CH-GP-HEC injectable hydrogel

The CH-GP-HEC with final concentration of chitosan, GP, and HEC; 12 mg/mL, 120 mM, and 4 mg/mL, respectively, gelled after 20 min at 37°C, which is optimum for clinical purpose. The gelation temperature and time of CH-GP-HEC hydrogel are adjustable and depend on the concentration of GP. CH-GP-HEC-based hydrogel was nearly degraded after one month in vivo (Figure 4A). Cells or/and therapeutic factors like SDF-1α can simply mix with the CH-GP-HEC hydrogel and injected to desired sites.

CXCR4 expression and SDF-1α release increase homing and retention

Although a few recent studies reported the in vivo SCs recruitment using SDF-1α-containing scaffold in a subcutaneous implantation model (Bladergroen et al., 2009; Kimura and Tabata, 2010; Thevenot et al., 2010; Chen et al., 2013; Zhang et al., 2013), no available studies investigated the recruitment and retention of pretreated ASCs in vivo. In this study, we designed a delivery platform for the controlled release of SDF-1α, an important chemokine for stem cell recruitment and homing, to investigate effect of pretreatment with DFX on ASCs and the efficiency of the three delivery routes. Results showed that pretreatment of hASCs with DFX can improve the efficacy and retention of hASCs mixed or migrated toward SDF-1α-releasing hydrogel injected subcutaneously in dorsum of Wistar rat in all three delivery routes (Figures 5A–5F) as compared with untreated experimental groups E, F, and G (Figures 5G–5L). On the other hand, the enriched population of DiI-labled cells was observed in the hydrogel containing SDF-1α as shown in Figures 5B, 5D, 5F, 5H, 5J, and 5L compared to the hydrogel without SDF-1α in Figures 5A, 5C, 5E, 5G, 5I, and 5K, respectively. For example, in delivery route 1 (mixed), incorporated SDF-1α effectively retained DFX-pretreated ASCs within the CH-GP-HEC hydrogel (Figure 5B). Additionally, since human specific genes were also detected by PCR only in experimental group B where DFX-pretreated ASCs were mixed with SDF-1α-containing hydrogel (Figure 4C), but not all other groups, it was concluded that SDF-1α operates in both homing and retention of CXCR4-positive cells. Furthermore, intensive cell infiltrations were observed within the CH-GP-HEC in this group accompanied with vascularization nearby (Figure 5O). This result was confirmed by the fact that the CH-GP-HEC hydrogels provided optimally permissive structural support for migration of cells (Figure 4B), indicating the hydrogels to be compatible with homing of endogenous cells.