Human amniotic fluid stem cell injection therapy for urethral sphincter regeneration in an animal model

This study was approved by the Ethics Committee of Kyungpook National University School of Medicine. All participants provided informed consent. Amniotic fluids (10 mL each) were obtained from four women undergoing routine amniocentesis at a gestational age of 15 to 19 weeks. Amniotic fluids were centrifuged and supernatants discarded. Cell pellets were resuspended with Chang Medium (α-MEM, 15% embryonic stem cell-fetal bovine serum (Gibco-Invitrogen, Grand Island, NY, USA) with 18% Chang B and 2% Chang C (Irvine Scientific, Irvine, CA, USA)} in a petri dish. Non-adherent cells were discarded after one week. Adherent cells were passaged for expansion on reaching 80% confluence, and culture medium was replaced every three days.

hAFSCs (Passage 3) were evaluated by flow cytometry with phycoerythrin (PE)- or fluorescein isothiocyanate-conjugated mouse monoclonal antibodies specific for embryonic stem cell marker SSEA4, MSC markers CD44, CD73, CD90 and CD105, hematopoietic stem cell marker CD45, and immunologic markers HLA-ABC and HLA-DR (BD Biosciences, San Jose, CA, USA), according to the manufacturer's instructions. Approximately 10,000 cells were measured using a fluorescence activated cell sorter (FACS; BD Biosciences) system equipped with the CellQuest program. A homogenous stem cell population was obtained by sorting the cells twice using C-KIT antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA) with a magnetic activated cell sorting system (MACS, Miltenyi Biotec, Bergischu Gladbach, Germany). In case #1, a cell population with high expression for C-KIT and SSEA4 as well as low expression for HLA-DR was obtained and used for subsequent studies.

The optimum myogenic condition for hAFSC induction was determined by testing three different induction media: (i) myogenic medium (DMEM, 0.5% chick embryo extract, 10% horse serum, Gibco-Invitrogen) containing 3 μM of 5-aza-20-deoxycytidine (5-azaC; Sigma-Aldrich, St. Louis, MO, USA), (ii) myogenic medium containing 5 ng/mL of transforming growth factor-β (TGF-β; Peprotech, Rocky Hill, NJ, USA), and (iii) conditioned medium (CM; collected from cultured human skeletal muscle cells). After 24-hour treatment with 5-azaC or TGF-β, cells were cultured up to a further 14 days. Cell viabilities at 1, 3, 5 and 14 days of differentiation were measured using a CCK-8 assay kit (Dojindo, Kumamoto, Japan,) according to the manufacturer's instructions. The genotypic and morphologic conversion of hAFSC into myocytes was analyzed by real-time polymerase chain reaction (PCR) and immunocytochemical (ICC) staining using routine methods. The primer sequences and antibody information are listed in Tables 1 and 2, respectively. The C2C12 cell line and human fibroblasts served as positive and negative controls for ICC staining respectively. The same experiment was repeated three times independently.

All experimental protocols were approved by the Animal Ethics Committee, Kyungpook National University School of Medicine. Female imprinting control region (ICR) mice, weighing 20 to 25 g, were obtained from Hyochang Science (Daegu, Korea). Animals were prepared for aseptic surgery under general anesthesia (isoflurane). A lower midline abdominal incision was made and the bladder and urethra were exposed. The SUI model was created using a bilateral pudendal nerve transection technique instead of crush injury of the pudendal nerve. This allowed us to confirm the functional recovery of incontinence by regeneration of the urethral sphincter and to eliminate the effect of pudendal nerve regeneration. The pudendal nerve on each side was identified and transected with microsurgical scissors under microscopic magnification (n = 30). Laparotomy was closed in layers with absorbable 4 to 0 vicryl sutures. An additional 15 mice underwent a sham operation (lower midline incision and closure) and served as normal controls.

One week after generation of the incompetent urethral sphincter model, animals were anesthetized with isoflurane and the bladder and urethra were exposed by a lower abdominal incision. The urethra was slightly retracted and hAFSCs were injected at the three and nine o'clock area of the external sphincter using a 26G Hamilton microsyringe (Hamilton Company, Reno, NV, USA). The depth of injection was determined by an experienced operator. Each injection consisted of 0.5 × 10⁶ undifferentiated AFSCs (Passage 5) in 5 μL of Plasma-Lyte. Three experimental groups were established: a sham-operation group (Ctrl), a pudendal neurectomy without cell injection group (Cell (-)) and a pudendal neurectomy with cell injection group (Cell (+)) (n = 15 for each group). The fate of injected cells in vivo was traced by immunohistochemical (IHC) staining with human nuclear-specific antibody at 3, 5, 7 and 14 days after injection.

Silica-coated magnetic nanoparticles incorporating rhodamine B isothiocyanate (MNPs@SiO2 (RITC)) were kindly provided by Dr. Jae-sung Bae (Kyungpook National University, Daegu, Korea). MNPs@SiO2 (RITC) is a novel cell-tracking agent with multimodal fluorescence and magnetic properties. MNPs@SiO2 (RITC) nanoparticles become integrated into the cells by endocytosis. The nanoparticles are not shed by cells and the signal cannot be detected after cell death. When cells are proliferating or differentiating, MNPs@SiO2 (RITC) transfers into daughter cells, where each cell particle number becomes gradually reduced. MNPs@SiO2 (RITC) is biocompatible and has been used for various stem cell tracking studies [8‐10]. In the present study, the optimal concentration and exposure time of MNPs@SiO₂ (RITC) for efficient uptake of nanoparticles into AFSCs was established by incubating cells at 37°C with various concentrations (0.01, 0.05, 0.1 or 0.2 mg/mL) of MNPs@SiO₂ (RITC) for different exposure times (24 to 72 hours). Briefly, the AFSCs (2 × 10⁴ per chamber well) were cultured in Lab-Tek™ Chamber Slides (four-chamber mounted Permanox slides; Nunc, Rochester, NY, USA). On reaching about 60 to 70% confluence, cells were incubated with MNPs@SiO₂ (RITC) in a 37°C/5% CO₂ incubator. Labeling was stopped by washing the cells three times with PBS. Cells were then fixed by incubation with 4% paraformaldehyde for 20 minutes at 4°C and washed with PBS. Individual coverslips were mounted onto VECTASHIELD mounting medium with DAPI (Vector Laboratories, Burlingame, CA, USA). Cells were examined with a fluorescence microscope (Olympus BX51, Tokyo, Japan) and confocal microscope (Olympus FluoViewTM FV1000) to determine the intracellular localization of nanoparticles.

For in vivo cell tracking, unsorted MNPs@SiO2 (NIR-797) labeled cells (0.5 × 10⁶) were injected into the urethral sphincter region of mice (n = 9), and optical images were obtained at 3, 7, 10 and 14 days using Optix exPlore (ART, Montreal, QC, Canada) with the filter set for NIR797. Images were corrected for background and autofluorescence using non-labeled cells as controls. Acquired images were analyzed with eXplore Optix OptiView software. For optical imaging, the animal was anesthetized by intravenous injection of a Rompun (Bayer, Leverkusen, Germany):Zoletil (Virbac Australia, Milperra NSW, Australia):saline mixture (1:5.7:10; 100 μL/mouse) according to the manufacturers' guidelines.

Mice were placed under general anesthesia with ether to avoid muscle relaxation. LPP and CP were measured one, two and four weeks after cell injection using the vertical tilt/intravesical pressure clamp model of SUI as previously described [11], and the Power Lab^® system (AD Instruments Pty Ltd, Bella Vista NSW, Australia). Before measuring, the spinal cord was transected at the T9 to T10 level to eliminate reflex bladder activity in response to increasing intravesical pressure. This suprasacral spinal cord transection does not interfere with the spinal continence reflexes of the bladder neck and urethra [12]. Under general anesthesia, the bladder was exposed by a midline incision. A transvesical catheter with a fire-flared tip (PE-25) was inserted into the dome of the bladder and the abdominal wall and overlying skin were closed with sutures. The mice were then mounted on a tilt table and placed in the vertical position. The intravesical pressure was clamped by connecting a large 50 mL syringe reservoir to the bladder catheter and the pressure transducer using PE-50 tubing and three-way stopcocks. The intravesical pressure was increased in 1- to 3-cm H₂O steps from 0 cm H₂O upward until visual identification of the leak point height. The pressure at this leak point was referred to as the LPP. The intravesical pressure was then decreased in 1- to 3-cm H₂O steps downward until the leak ceased. The pressure at this leak cessation point was taken as the CP. The averages of three consecutive LPP and CP measurements were taken as data points for each animal.

Animals were sacrificed after LPP and CP measurement and the bladder-urethra complex was removed en bloc. Tissue specimens were fixed in 10% buffered formalin, processed and cut into 4- to 6-μM-thick sections for routine staining with hematoxylin and eosin (H&E), and for IHC staining. The presence and migration of injected cells in situ were analyzed by IHC using human nuclear specific antibody (HuNu; Cell Signaling Technology, Danvers, MA, USA). Myogenic conversion of injected hAFSCs in vivo was confirmed with real-time PCR using human primers for myogenic lineage markers. To investigate whether injected hAFSCs induced a host myogenic response, real-time PCR analysis was performed using mouse primers for myogenic lineage markers. The neuromuscular junction formation was analyzed at Week 4 by IHC staining for the acetylcholine receptor with α-Bungarotoxin. The possibility that hAFSC injection induced host neuroregeneration was investigated by real-time PCR with mouse neurogenic-related primers.

The immunogenicity of hAFSCs was evaluated by flow cytometry to determine expression of HLA-DR on the cell surface. The in vivo immunosuppression effect of hAFSCs was examined by retrieving tissue one week after injection and performing IHC staining with cytotoxic T cell marker (CD8, BD Pharmigen, San Jose, CA, USA). As a negative control, human fibroblasts were injected (n = 3). Tumorigenicity was analyzed by injecting 1 × 10⁶ hAFSCs into the subcapsular space of the kidney (n = 9). Animals were sacrificed eight weeks later and kidneys were harvested for histologic confirmation.

Our main outcome measure was to compare differences between LPP and CP between the hAFSCs injection and control groups in the SUI animal model. We collected data from five animals in each group at one, two and four weeks after injection. The results were also compared with those of the normal control (n = 5, at each time point). Data were presented as means ± SD. Statistical analysis was undertaken with Student's t-test or one-way analysis of variance (ANOVA). A P-value of less than 0.05 was considered statistically significant. When the value was found to be significant after assessment using the ANOVA statistical test, the Tukey post-hoc comparison was used.

We obtained adherent cells from all four amniotic fluid samples. FACS analysis on culture-expanded cells (Passage 3) showed the cells were strongly positive for the mesenchymal markers, CD44, CD73, CD90 and CD105; weakly positive for the embryonic marker, SSEA-4; and negative for the hematopoietic lineage marker, CD45 and MHC Class II antigen, HLA-DR (Figure 1A). Double MACS analysis showed 98.40% of the cell population in case #1 was C-KIT (+) (Figure 1B). The C-KIT (+) populations in other cases were less than 90% (data not shown). Therefore, we used the C-KIT (+) cells from case #1 for subsequent studies.

When hAFSCs were cultured in myogenic induction media, the cells become elongated and with spindle shapes at Day 7. These findings were similar in all three groups. CCK-8 assays, performed at Day 14 showed a 2.7- (P < 0.0001) and 1.58- (P < 0.0001) fold higher cell viability for cells cultured in CM compared with cells cultured in medium containing 5-azaC or TGF-β, respectively (Figure 2A). Real-time PCR showed that expression of early myogenic differentiation markers (PAX7 and MYOD) was dominant at Day 3, whereas the mid to late myogenic marker (DYSTROPHIN) became dominant at Day 7. Each gene expression level varied depending on the medium type (Figure 2B). These results were confirmed by ICC staining (Figure 2C).

In vitro ICC staining with HuNu confirmed the presence of hAFSCs (Figure 3A). IHC staining confirmed focal detection of injected cells at the injection site at Day 3. Cells then migrated into the surrounding urethral sphincter region (Figure 3B). The cell signal decreased from Day 7 and had disappeared by Day 14.

The optimum concentration and treatment time for MNPs@SiO2 (RITC) labeling of hAFSCs were 0.1 mg/mL and 24 hours, respectively (Figure 4A). The maximum labeling efficacy was 94.31% (Figure 4B). MNPs@SiO2 (RITC) uptake was uniform for each cell preparation and density. The cell viability assay confirmed that the nanoparticles did not induce cytotoxicity at a wide range of concentrations (0.05 to 0.2 mg/mL) and exposure times (up to 72 hours) (Figure 4C). Injection of MNPs@SiO2 (RITC) labeled hAFSCs into the periurethral region confirmed that optical imaging could identify these cell clusters at the injection site. The signal intensity decreased gradually until Day 10 after injection and disappeared thereafter (Figure 4D).

LPP and CP were measured one, two and four weeks after injection (Figure 5). The mean LPP and CP for the Cell (+) group were similar to those of the Cell (-) group at Week 1 (17.9 ± 0.5 vs. 16.6 ± 2.1 and 9.9 ± 1.3 vs. 9.1 ± 0.9 cmH₂O). However, the mean LPP was significantly higher for the Cell (+) group than for the Cell (-) group at Week 2 (18.1 ± 2.8 vs. 11.6 ± 1.2 cmH₂O, P = 0.0014) and Week 4 (20.2 ± 3.3 vs. 15.2 ± 2.1 cmH₂O, P = 0.0202). The mean CP for the Cell (+) and Cell (-) groups showed similar differences to those found for LPP at the same time points (Week 2: 12.4 ± 1.7 vs. 6.9 ± 1.1 cmH₂O, P = 0.0001 and Week 4: 14.4 ± 3.4 vs. 8.4 ± 1.1 cmH₂O, P = 0.0051).

After measurement of LPP and CP, the mice were sacrificed and the whole urethra was excised. The entire length of the urethra was about 7 mm and the rhabdosphincter was located about 6 mm distant from external urethral orifice (Figure 6A). H&E staining identified normal-appearing circular muscle mass regeneration over time at the urethral sphincter region in the Cell (+) group. In contrast, the Cell (-) group showed only scant muscle regeneration and an atrophic sphincter. These results were confirmed with IHC staining using MyoD antibody (Figure 6B). Real-time PCR analysis showed that the expression of genes related to early (PAX7, MYF5 and MYOD) and mid to late (MYOGENIN, MEF2 and MLP) myogenic differentiations corresponded correctly with time. Human gene expression was highest in the first week then gradually reduced (Figure 7A), whereas mouse gene expression gradually increased with time (Figure 7B). The IHC for neuromuscular junction formation showed that the Cell (+) group had a similar expression level of α-bungarotoxin for acetylcholine receptor as the normal control (Figure 7C). Real-time PCR showed neurogenic gene expression (Nestin, Vimentin, Neurofilament, Microtubule-associated Protein 2, β-Tubulin III, Glial Fibrillary Acidic Protein) was significantly higher (P < 0.05) in the Cell (+) group compared with the Cell (-) group (Figure 7D).

FACS analysis showed that HLA-DR expression in hAFSCs was lower (0.26% total) than that of the isotype control (1.48% total) (Figure 8A). IHC staining of the urethral sphincter tissue of hAFSC injected animals revealed scant CD8 lymphocyte aggregation at one week, whereas the Cell (-) and human fibroblast injected animals had significant CD8 lymphocyte accumulation (Figure 8B). Histologic analysis revealed no teratoma formation in tissues retrieved eight weeks after renal subcapsular injection of hAFSCs (Figure 8C).