A Nonhuman Primate Model for Urinary Bladder Regeneration Using Autologous Sources of Bone Marrow-Derived Mesenchymal Stem Cells§


  • Arun K. Sharma,

    Corresponding author
    1. Division of Pediatric Urology; Children's Memorial Hospital of Chicago, Chicago, Illinois, USA
    2. Department of Urology, Northwestern University Feinberg School of Medicine, Chicago, Illinois, USA
    3. Institute for BioNanotechnology in Medicine (IBNAM), Northwestern University, Chicago, Illinois, USA
    • The Feinberg School of Medicine at Northwestern University, Institute for BioNanotechnology in Medicine, 303 East Superior Street, IBNAM 11-113, Chicago, Illinois 60611, USA
    Search for more papers by this author
    • Telephone: 312-503-1101; Fax: 312-503-1222

  • Matthew I. Bury,

    1. Division of Pediatric Urology; Children's Memorial Hospital of Chicago, Chicago, Illinois, USA
    Search for more papers by this author
    • Author contributions: A.K.S.: conception and design, collection and/or assembly of data, data analysis and interpretation, manuscript writing, final approval of manuscript; M.I.B.: collection and/or assembly of data, data analysis and interpretation; A.J.M.: collection and/or assembly of data, manuscript writing; N.J.F.: collection and/or assembly of data, data analysis and interpretation, manuscript writing; J.W.M.: collection and/or assembly of data; N.T.: collection and/or assembly of data; L.C.H.: provision of study material or patients, manuscript writing; D.J.M.: collection and/or assembly of data; E.Y.C.: financial support, collection and/or assembly of data, data analysis and interpretation.

  • Andrew J. Marks,

    1. Division of Pediatric Urology; Children's Memorial Hospital of Chicago, Chicago, Illinois, USA
    Search for more papers by this author
  • Natalie J. Fuller,

    1. Division of Pediatric Urology; Children's Memorial Hospital of Chicago, Chicago, Illinois, USA
    Search for more papers by this author
  • Jay W. Meisner,

    1. Division of Pediatric Urology; Children's Memorial Hospital of Chicago, Chicago, Illinois, USA
    Search for more papers by this author
  • Natalie Tapaskar,

    1. Division of Pediatric Urology; Children's Memorial Hospital of Chicago, Chicago, Illinois, USA
    Search for more papers by this author
  • Lisa C. Halliday,

    1. Biologic Resources Laboratory, University of Illinois at Chicago, Chicago, Illinois, USA
    Search for more papers by this author
  • Derek J. Matoka,

    1. Division of Pediatric Urology; Children's Memorial Hospital of Chicago, Chicago, Illinois, USA
    Search for more papers by this author
  • Earl Y. Cheng

    1. Division of Pediatric Urology; Children's Memorial Hospital of Chicago, Chicago, Illinois, USA
    2. Department of Urology, Northwestern University Feinberg School of Medicine, Chicago, Illinois, USA
    3. Institute for BioNanotechnology in Medicine (IBNAM), Northwestern University, Chicago, Illinois, USA
    Search for more papers by this author

  • Disclosure of potential conflicts of interest is found at the end of this article.

  • §

    First published online in STEM CELLSEXPRESS November 23, 2010.


Animal models that have been used to examine the regenerative capacity of cell-seeded scaffolds in a urinary bladder augmentation model have ultimately translated poorly in the clinical setting. This may be due to a number of factors including cell types used for regeneration and anatomical/physiological differences between lower primate species and their human counterparts. We postulated that mesenchymal stem cells (MSCs) could provide a cell source for partial bladder regeneration in a newly described nonhuman primate bladder (baboon) augmentation model. Cell-sorted CD105+/CD73+/CD34/CD45 baboon MSCs transduced with green fluorescent protein (GFP) were seeded onto small intestinal submucosa (SIS) scaffolds. Baboons underwent an approximate 40%–50% cystectomy followed by augmentation cystoplasty with the aforementioned scaffolds or controls and finally enveloped with omentum. Bladders from sham, unseeded SIS, and MSC/SIS scaffolds were subjected to trichrome, H&E, and immunofluorescent staining 10 weeks postaugmentation. Immunofluorescence staining for muscle markers combined with an anti-GFP antibody revealed that >90% of the cells were GFP+/muscle marker+ and >70% were GFP+/Ki-67+ demonstrating grafted cells were present and actively proliferating within the grafted region. Trichrome staining of MSC/SIS-augmented bladders exhibited typical bladder architecture and quantitative morphometry analyses revealed an approximate 32% and 52% muscle to collagen ratio in unseeded versus seeded animals, respectively. H&E staining revealed a lack of infiltration of inflammatory cells in grafted animals and in corresponding kidneys and ureters. Simple cystometry indicated recovery between 28% and 40% of native bladder capacity. Data demonstrate MSC/SIS composites support regeneration of bladder tissue and validate this new bladder augmentation model. STEM CELLS 2011;29:241–250


The clinical needs of pediatric patients suffering from developmentally abnormal urinary bladders have traditionally relied heavily on surgical intervention in the form of bladder augmentation cystoplasty to create a reservoir for urine collection and expulsion [1, 2]. The structural components of these artificial bladders have been derived from bowel tissue to increase compliance and decrease leak point pressure while simultaneously increasing overall bladder capacity [3, 4]. The risks involved with using bowel tissue include perforation, electrolyte imbalances, excessive mucus production, and malignant transformation [4, 5]. Although bladder augmentation cystoplasty is considered a state of the art procedure, it is still a stopgap measure. Recent clinical studies have attempted to create functional bladder tissue to overcome the problems associated with traditional bladder augmentation but have been met by numerous obstacles. Data demonstrate the use of myelomeningocele patient-derived urothelial and bladder smooth muscle cells (bSMCs) used to reconstruct the urothelium and smooth muscle components of the bladder [6]. Although highly innovative in approach, these studies demonstrated lackluster physiological effects and did not address the possibility that the use of pathologic bladder cells may eventually result in the reformation of a diseased bladder state and the decline in urodynamic function of patients undergoing this treatment. Hence, the sources of cells to be used in a bladder regenerative setting should be disease-free and ideally come from autologous sources as to avoid unwanted immune responses.

Experimental data has demonstrated the utility of bone marrow (BM) cells in a bladder regenerative setting with data suggesting the phenotypic and physiological similarities between mesenchymal stem cells (MSCs) and bSMCs [7–9]. The characterization of a battery of contractile proteins shows that MSCs have a contractile protein profile similar to bSMCs [8]. Further data presented in this study demonstrate both cell types can also respond to agonist stimulation with statistically indistinguishable contractile responses in vitro. A second study describes the use of MSCs in a nude rat bladder augmentation model [9]. Data from this study indicate that bladder reconstitution occurs more robustly using epitope-defined MSCs compared with normal human bSMCs seeded on an elastomeric scaffold compared with unseeded control animals. MSC-seeded scaffolds maintained high levels of smooth muscle markers and displayed smooth muscle bundle formation by trichrome staining with statistically superior muscle to collagen ratios at 10 weeks postimplantation compared with bSMCs-seeded scaffold constructs [9]. This data implies that MSCs can serve as an alternative cell source for potentially diseased bSMCs.

Although a majority of the bladder augmentation studies have been performed in canine, rodent, or rabbit models, [10–13] there exists the conundrum as to whether these models are good representations of the lower urinary tract in humans as anatomical and physiological differences often result in poorly translatable data. The high regenerative capacity of urothelial cells in various models may possibly confound overall analyses of regenerative process in cystectomized animals. Studies describing this phenomenon in nude rats demonstrate the rapid invasion of native urothelium onto the luminal side of cystectomized bladders that have been seeded with cells. This rapid invasion of urothelium may prevent the luminally seeded cell populations from proliferating and contributing to the grafts [9, 12]. Alternatively, the cells on the luminally seeded side of the graft may undergo cell death as a consequence of being in contact with urine. Although this regrowth of urothelium is also seen in humans, mechanisms governing the growth of cells in humans and lower species are poorly understood and may not be comparable. Second, there are several factors to consider when attempting to extrapolate data from animal models to humans. The limited graft size in rodent models provides an inadequate model of regeneration due to the rate at which the regenerative process occurs. Small augmented areas tend to regenerate within a matter of weeks, whereas larger grafts, as seen in canine models, can take up to several months to regenerate completely [9, 14, 15]. More importantly, there exist anatomical differences between the lower nonhuman primates (NHPs) species and humans that include the musculature component of the lower urinary tract further questioning the translational validity of lower NHP models [16–18]. NHP are phylogenetically more similar to humans than are rodents or canines, thus displaying strong anatomical and physiological resemblances to humans making the NHP an ideal model to examine bladder tissue engineering strategies [16–18]. Within the context of this study, we describe the use of a baboon bladder augmentation model in conjunction with an autologous source of BM MSCs for partial bladder regeneration.


Baboon MSC Isolation and Characterization

Female baboon (Papio anubis; 18–20 kg; Tulane National Primate Research Center, New Orleans, LA) BM aspirates (∼20 ml) were obtained from the humeri and iliac crests under ketamine (10 mg/kg, IM) and xylazine (1–2 mg/kg, IM) anesthesia. The BM was diluted 1:8 in Dulbecco's phosphate-buffered saline (DPBS; Invitrogen Corporation, San Diego, CA) and the mononuclear cell (MNC) fraction was isolated by centrifugation over a 60% Percoll (Pharmacia, Uppsala, Sweden) density gradient at 500g for 30 minutes at room temperature (RT) [19]. The MNC fraction was collected and cultured in MSC growth media (MSCGM, Lonza, Inc., Walkersville, MD) at a density of 4,000 cells/cm2 and allowed to proliferate. The adherent cell population was isolated and subjected to antibody staining and fluorescent-activated cell sorting with antibodies against CD105, CD73, CD34, and CD45 (Becton Dickinson Biosciences, San Diego, CA) to isolate CD105+/CD73+/CD34/CD45 baboon MSCs (bMSCs) [20]. Antibodies were used at dilutions of 1:100 to 1:150. Cells were cultured at 37°C with 5% CO2 in air and all bMSCs used throughout this study were passage 3 or earlier. bMSCs were subjected to coerced differentiation into an adipocyte lineage using a protocol for human MSC differentiation [21]. bMSCs were plated at 2,500 cells/cm2 and grew for approximately 10 days in MSCGM until reaching 100% confluency. bMSCs underwent three cycles of an induction/maintenance regimen in which MSCs were cultured in adipogenic induction media (Lonza, Inc.) for 3 days followed by 3 days of maintenance in adipogenic maintenance media. Induced bMSCs were stained with Oil Red O (Milipore Corporation, Temecula, CA) to visualize adipocytes by an established protocol [21]. Guidelines for the animal research protocols were established and approved by the Animal Care Committee at the University of Illinois at Chicago.

bMSC Viral Transduction

Sorted bMSCs were transduced with copGFP self-inactivating lentiviral particles (Santa Cruz Biotechnology Inc., Santa Cruz, CA) as per manufacturer's instructions [22]. Briefly, bMSCs were plated at 25,000 cells/cm2 and 100 μl of a 5,000 infectious viral units per microliter of viral stock was added to the bMSCs in the presence of 2 μg/ml polybrene, and cells were incubated for 48 hours at 37°C with 5% CO2 in air. Cells were visualized under fluorescence to determine transduction efficiency.

bMSC Scaffold Seeding

Single layer surgical grade acellular small intestinal submucosa (SIS; Biodesign Surgisis; Cook Biotech Inc., West Lafayette, IN) were placed within a sterilized tissue culture frame consisting of beveled stainless steel brackets with an outer area of 46 cm2 and inner area of 29 cm2. The SIS containing bracket was submerged in complete MSCGM and underwent three media changes in a 24-hour span and was placed in fresh media and left to incubate overnight at 37°C and 5% CO2 in air. The following day, the SIS was washed once with fresh MSCGM media and seeded with green fluorescent protein positive (GFP+) bMSCs at 80,000 cells/cm2 onto one side of the SIS with a minimal amount of media to keep the cells in a hospitable environment. The cells remained in this configuration for 5 hours at 37°C and 5% CO2 until the cells attached. The frame was then flipped over and the other side was seeded in a similar manner except that the seeded side was submerged in fresh media while the seeding process was underway. bMSC/SIS composites were allowed to grow in culture for approximately 1 week prior to bladder augmentation cystoplasty with media changes every 2–3 days. Unseeded SIS scaffold was incubated in MSCGM prior to implantation.

Contrast Radiographs and Cystometry Studies

Baboons underwent radiographs and cystometry studies to determine bladder capacities. Animals were identified as PA1 (sham control), PA2 (unseeded SIS), PA3, and PA4 (bMSC/SIS constructs). Survey and contrast radiographs were taken of the caudal abdomen to visualize the bladder prior to the surgical procedure. All radiographic films were taken using a Summit Nova 325 x-ray machine and processed using an Optimax 2010 film processor. Lateral and anteroposterior (AP) survey radiographic views were taken first, followed by lateral and AP views with the urinary bladder filled with contrast medium. Contrast medium was injected into the bladder via the Foley catheter until maximum distension was achieved. After adequate radiographs were taken, the contrast medium was removed via the Foley catheter and the baboons were transferred to the surgical suite. Cystometry was performed on baboons prior to bladder augmentation and sacrifice. Under intramuscular ketamine sedation, intravenous (i.v.) propofol was administered and endotracheal intubation obtained. A 10-French Foley catheter was placed per urethra in the bladder using aseptic technique and the balloon inflated with 3 ml of sterile water. Slight traction was applied and the catheter taped in place to prevent leakage around the catheter. With the baboon in a lateral decubitus position, a gravity manometer was attached to the Foley catheter using i.v. tubing. Using a syringe, the bladder was slowly filled with normal saline solution through the IV tubing in 20-ml interval. At each interval, a pressure measurement was taken using the manometer. This process was continued until the intravesical pressure reached 20 cm H2O. The volume required to produce a pressure of 20 cm H2O was considered the bladder capacity.

Baboon Bladder Augmentation Cystoplasty

After sedation, the bladder was filled with 100 ml of normal saline through a 10-French Foley catheter and clamped. A 5-cm lower midline incision was made in the abdomen using a scalpel, and subcutaneous tissues dissected using electrocautery developing the extravesical space of Retzius. The peritoneum was entered adjacent to the urachus and the bladder dissected further. Once the dome of the bladder was exposed, approximately 40%–50% cystectomy was performed by removing a square of tissue at the dome of the bladder using electrocautery. After cystectomy, animal PA1 underwent closure of the bladder in two layers using running vicryl 4-0 sutures. For animal PA2, a piece of SIS was cut to size to fill in the gap left by the cystectomy. For animals PA3 and PA4, a dual-sided MSC SIS composite was used in a similar fashion. Once cut to size, the SIS was sutured to the bladder using four running absorbable monofilament sutures. Integrity of the augmentation was verified by irrigating the bladder through the Foley catheter with 100 ml of normal saline solution and injection of saline into the bladder from the nonaugmented region of the bladder. The closure was covered with a pedicle flap of greater omentum, which was tacked in place using 4-0 vicryl sutures. A 1/4-inch latex penrose drain was placed through a separate stab wound in the left lower quadrant and sewn to the skin with nonabsorbable monofilament sutures. The fascia, subcutaneous tissue, and skin were closed with absorbable sutures. Intravenous antibiotics were administered postoperatively as per facility protocol. Urinary catheters were removed after 3 days and drains after 5 days by facility veterinary staff. At 10 weeks postsurgery, animals were administered ketamine and xylazine sedation and underwent repeat cystometry as described earlier. Animals were sacrificed using pentobarbital overdose as per facility protocol. Whole specimens of bladder were obtained as well as the right kidney and ureter to inspect for any evidence of upper urinary tract changes secondary to surgery.

Histological Tissue Staining

Tissue specimens consisting of full thickness baboon bladder, kidney, and ureters were isolated immediately following euthanasia and were fixed in 10% buffered formalin phosphate (Fisher Scientific, Pittsburgh, PA). Specimens were dehydrated through a series of graded ethanol exchanges followed by paraffin embedding according to established protocols [23]. Samples were sectioned at a 5 μm thickness using a RM2125 RT Microtome (Leica, Nannockburn, IL) onto glass slides and subjected to Masson's Trichrome (Sigma-Aldrich, St. Louis, MO) staining as previously described [9, 24]. Tissue containing slides were deparaffinized by heating followed by treatment with xylenes, graded ethanol washes, and deionized water. Slides were placed in Bouin's solution, hematoxylin, and Scarlet-acid fuchsin for 15, 5, and 5 minutes, respectively, with in between water washes. Slides were subjected to a mixture of phosphotungstic acid (PTA)/phosphomolybdic acid hydrate (PMA), followed by aniline blue solution and a 1% acetic acid wash. Slides were placed in 95%–100% ethanol and rinsed in xylene. After drying, a coverslip was placed over the tissue and secured with Permaslip (Alban Scientific Inc, St Louis, MO). Alternatively, samples were subjected to H&E staining as previously described [25].

Quantitative Morphometry of Trichrome-Stained Bladder Tissue

Explanted bladder specimens that underwent Trichrome staining were evaluated for muscle and collagen content by an established protocol [9]. Muscle to collagen ratios from Trichrome-stained samples were digitally quantified using a Nikon Eclipse Microscope (Nikon Inc., Melville, NY) and Spot Advanced Imaging Software (Diagnostic Instruments, Sterling Heights, MI). Images were opened with Adobe Photoshop CS3 (Adobe Systems Inc., San Jose, CA). Selected pixels were quantified using the image histogram tool and a muscle to collagen ratio was calculated from these values. In instances where red blood cells or debris were present, images were edited to remove these structures to preserve a more accurate extrapolation of the muscle to collagen from the red to blue.

Immunofluorescent Staining of Augmented Tissues

Following the embedding and sectioning processes, slides were further prepared and visualized for GFP expression using a fluorescent microscope (Nikon Eclipse 2000-U; Nikon Inc.) using standard settings for 488-nm emission and excitation. Tissue samples were subjected to staining with an anti-GFP antibody (Santa Cruz Biotechnology, Santa Cruz, CA). Antigen retrieval was also performed on specimen slides that were boiled for 15 minutes in citrate buffer (0.01 M citrate, pH 6.0, 0.05% Tween-20), followed by a 40 minutes cool down. Slides were blocked for 15 minutes in bovine serum albumin (5 mg/ml) followed by a 30 minutes incubation at RT with the anti-GFP antibody at a 1:100 dilution. After washing with DPBS, slides were incubated for 30 minutes with an Alexa Red 568 secondary antibody at a 1:400 dilution. Slides were rinsed with DPBS, air dried, and mounted with Vectashield (Vector Laboratories, Burlingame, CA). Primary antibodies used in this study were directed against epitopes for markers of bSMCs including caldesmon, calponin, transgelin, pan-cytokeratin, alpha smooth muscle actin (α-SMA), smooth muscle myosin heavy chain (SMMHC), elastin, or Ki-67 (Santa Cruz Biotech, Santa Cruz, CA or Millipore, Billerica, MA) in conjunction with either an Alexa Red 568 or fluorescein isothiocyanate (FITC)–conjugated secondary antibody (Molecular Probes, Carlsbad, CA) [9]. Primary antibodies were used at dilutions from 1:100 to 1:250, whereas secondary antibodies were used at a 1:400 dilution. Immunofluorescent quantification was carried out by using a Nikon Eclipse 50i Microscope (Nikon Inc.) and Spot Advanced Imaging Software (Diagnostic Instruments). Sample images (1,600 × 2,000 pixels, bit depth 24) were opened with Adobe Photoshop CS3 (Adobe Systems Inc.). The number of GFP+ (red in color) and GFP+/marker+ (muscle or Ki-67 [marker]; orange in color) were manually counted using the eraser tool within Adobe Photoshop to mark exclusively colored cells.

Statistical Analyses

Statistical data of quantified trichrome images were derived by using a t test (two sample assuming unequal variances) expressed as mean ± with 95% confidence intervals using Microsoft Excel. p < .05 is statistically significant.


MSC Characterization and Scaffold Seeding

bMSCs initially isolated by fluorescence activated cell sorting (FACS) displayed typical fibroblast morphology when plated in tissue culture flasks and was analogous to their human counterparts (Fig. 1A, 1B). To ascertain whether or not the cells that were isolated were indeed bMSCs, the cells were subjected to coerced differentiation into adipocytes, a hallmark cell type of MSC terminal differentiation [20]. During the second induction phase of the protocol, it was observed that the presence of adipocytes appeared at a greater frequency than human MSC control samples that were seeded in the same manner and continued to increase in number until the conclusion of the assay. Figure 1C demonstrates the conversion of bMSCs into adipocytes. Ex vivo expansion of bMSCs demonstrated rapid and robust growth that was similar to human MSC samples. Transduced populations of bMSCs were approximately 85% GFP+ prior to scaffold seeding and subfractions of cells continued to grow in culture for several weeks. Figure 1D depicts GFP+ bMSCs that were seeded on SIS. Cells appeared to grow normally on the SIS and maintained their fibroblastic morphology up until implantation.

Figure 1.

Baboon mesenchymal stem cell (bMSC) characterization. (A): Fibroblast morphology of bMSCs plated in vitro, (B) cells are similar in appearance as compared with human MSCs, (C) adipocytes derived from bMSCs, and (D) green fluorescent protein positive (GFP+) bMSCs grown on small intestinal submucosa in vitro prior to augmentation procedures (×100).

Contrast Radiographs and Bladder Cystometry Studies

Volumetric capacity was determined by the volume instilled into the bladder to produce an intravesical pressure of 20 cm H2O. The pressure of 20 cm of H2O was chosen as a highly conservative pressure in which the potential bladder volume could be reasonably obtained while not compromising the bladder-scaffold anastomoses. This was based on many years of clinical expertise dealing with augmentation cystoplasty procedures in the urologic pediatric patient population. Animal PA1 had a presurgical volumetric capacity of 180 ml and a postsurgical capacity of 70 ml (Fig. 2B). Animal PA2 had a large capacity bladder presurgery of 460 ml (Fig. 2A), whereas postoperative values obtained indicated the capacity decreased to 130 ml. (Fig. 2C) Animal PA3 had a presurgical capacity of approximately 180 ml, although this may be an overestimation due to some leakage of urine around the urinary catheter at capacity and the postoperative capacity value was 60 ml (Fig. 2D). Animal PA4 had an initial 380 ml capacity preoperatively and 150 ml capacity postoperatively (Fig. 2E). Cystometry indicated bladder capacity recovery at 39%, 28%, 33%, and 40% for animals PA1, PA2, PA3, and PA4, respectively, of native bladder capacity at 10 weeks. Preoperative pressures increased slowly during bladder filling until nearing volumetric capacity, at which point intravesical pressure increased more sharply. Postoperatively, however, bladder pressures began increasing quickly during filling in all animals, suggesting that the indicated reduced bladder compliance.

Figure 2.

Contrast radiographs demonstrating animal bladder capacity. (A): Animal PA2 prior to augmentation cystoplasty procedures. Capacity for PA2 was 460cc and was reduced to 28% (C). Postoperative images of animals PA1–PA4 reveal the smooth contour of bladder with even filling. No gross defects were observed ([B–E], 10 weeks post-surgery). Abbreviation: R, right side.

Bladder Surgical Studies

Bladder cystectomy and augmentation procedures were conducted similarly to those performed in the clinical environment. Figure 3 depicts images throughout the course of the augmentation procedure of animal PA4. Figure 3A depicts the initial 5-cm incision through the subcutaneous layer with exposure of the abdominal wall and musculature. After physical separation of the musculature, the urinary bladder was exposed (Fig. 3B). The high level of vascularization is represented by the black arrows. Electrocautery of the bladder musculature removed approximately 40%–50% of the bladder (Fig. 3C), which was splayed and held open by sutures. (Fig. 3D) MSC-seeded SIS was placed in the defect region (Fig. 3E–3G) and sewn together with sutures. This created a liquidtight seal between the native bladder and the MSC-seeded SIS patch. (Fig. 3H). There was no evidence of leakage once the MSC/SIS composite was sewn into place and was confirmed by the injection of sterile saline into the bladder underside. As a final step in the augmentation procedure, omentum was sutured over the augmented bladder to provide additional support to the bladder and a potential source of nutrition due to the highly vascularized nature of the omentum (Fig. 3I). Figure 3J depicts the same bladder 10 weeks postsurgery immediately after euthanasia. The bladder is viewed in its distended form and appears healthy and devoid of necrotic tissue in the region of regeneration on gross visual inspection. Figure 3K is the bladder removed from the host and splayed open with regions of apparent regeneration depicted by black arrows. The ureters and kidneys were also examined grossly and displayed no adverse effects of the augmentation procedure. Figure 3L depicts a hemisection of the right kidney with no apparent signs of distress. Animals urinated normally and demonstrated no adverse effects from the surgery while also exhibiting normal social behaviors. The anatomical structures of the lower urinary tract were found to be similar to that of humans without any note of deviation on gross visual examination.

Figure 3.

Baboon bladder augmentation cystoplasty. Baboon that underwent a bladder cystectomy and augmented with a baboon mesenchymal stem cell-seeded small intestinal submucosa scaffold. (A–J): The surgical procedure from initial incision to the layering of omentum at the finale. (K): Bladder directly post euthanasia. (L): Right kidney sample showing no gross signs of damage after the augmentation procedure.

Quantitative Morphometry of Regenerated Bladder Tissue

Muscle to collagen ratios are indicative of overall bladder tissue regeneration as previously described [9, 26]. Masson's trichrome staining revealed distinct differences between unseeded and SIS-seeded samples in areas of regeneration at 10 weeks postbladder augmentation. Figure 4A represents native bladder tissue (“native” being defined as bladder tissue that was not part of the augmented region but part of the endogenous animal bladder), whereas Figure 4B depicts after it underwent cystectomy from animal PA1. The tissue in Figure 4B is nearly identical to that found in Figure 4A. Figure 4E depicts native bladder tissue from animal PA2 that was augmented with unseeded SIS, whereas Figure 4F shows the region in which the SIS patch was grafted demonstrating skewed muscle to collagen ratio expression. Figure 4I and 4M shows the native bladder tissue samples from animals PA3 and PA4 prior to augmentation, whereas Figure 4J and 4N shows the samples demonstrating the robust growth of bladder tissue that was derived from the seeded MSCs that formed smooth muscle-like bundles in these animals. Figure 4C, 4G, 4K, 4O and 4D, 4H, 4L, 4P represent H&E staining of bladder tissue of animals PA1–4 postsurgery and right kidney samples, respectively. Staining reveals a paucity of inflammatory cells in these tissue samples with the exception of Figure 4G which depicts the unseeded SIS graft. The sections of proximal and distal (labeled P and D, respectively) ureters (insets in Figs. 4D, 4H, 4L, 4P) also appear normal. Ureter tissue samples were taken approximately 1-inch away from the kidney and the bladder. Staining further revealed a lack of microscopic disruption in augmented tissue. Yellow arrows indicate smooth muscle bundles (Fig. 4A, 4B, 4E, 4I, 4J, 4M, 4N), red arrows indicate areas of collagen accumulation (Fig. 4F), and black arrows indicate histologically normal glomeruli (Fig. 4D, 4H, 4L, 4P) from kidney samples of animals PA1–4. Finally, laboratory values obtained from blood samples of animals PA2 and PA3 presurgery and postsurgery were normal. Levels of sodium, potassium, chloride, blood urea nitrogen, albumin, total protein, creatinine, calcium, and bilirubin were within acceptable levels. These data taken together indicate that kidney function was not affected by the bladder augmentation procedure. Quantitative morphometry studies corroborated microscopic trichrome results. Native bladder tissue was found to be approximately 60% muscle in content at 10th week time point. Specifically, PA1 was 65.8% ± 1.5%, PA2 62.0% ± 1.8%, PA3 62.2% ± 1.1%, and PA4 was 63.9% ± 2.5%. Animals PA3 and PA4 augmented with MSC/SIS-seeded scaffolds demonstrated muscle content approaching that of the native tissue in new areas of regenerated tissue. PA3 was 50.9% ± 2.2% regenerated tissue versus 62.2% ± 1.1% native, whereas PA4 52.9% ± 2.0% regenerated tissue versus 63.9% ± 2.5% native. Regenerated tissue in PA2 showed approximately half the muscle content of native tissue (31.9% ± 3.2% regenerated vs. 62.0% ± 1.8% native). Statistically, muscle content for regenerated, but not native, tissue was considered significantly different for MSC/SIS scaffold composites (n = 2 baboons, n = 14 regenerated tissue, and n = 15 native tissue images per animal) versus the unseeded-SIS (n = 1 baboons, n = 12 regenerated tissue, and n = 15 native tissue images) grafts (regenerated p < .0001, native p > .05). Data presented in Figure 5. Data demonstrate that bMSC-seeded SIS scaffolds were able to recapitulate the native bladder environment to levels approaching native muscle/collagen levels.

Figure 4.

Histological analyses. Masson's trichrome and H&E staining of native, grafted/surgical tissue areas, and kidney/ureter samples of PA1 (A–D), PA2 (E–H), PA3 (I–L), and PA4 (M–P). Trichrome images and kidney/ureter images (×100); H&E-grafted tissue images (×200); and insets (D) indicates distal and (P) indicates proximal ureters (×40).

Figure 5.

Trichrome quantification. Seeded grafts showed more muscle content than unseeded graft in regenerated, but not native, tissue. Muscle content in regenerated tissue was similar for the two seeded grafts; comparison with the unseeded graft showed statistically significant differences whether the two animals were considered separately or grouped (*, p < .0001, error bars, 95% confidence intervals).

Immunofluorescence Staining

GFP+ bMSCs/SIS composites used for bladder augmentation failed to display GFP expression on visualization using standard fluorescence microscopy. It has been previously demonstrated that the paraffin-embedding process can have deleterious effects on GFP expression [27]. To access GFP-expressing cells, antigen retrieval was performed, followed by staining with an anti-GFP antibody. Figure 6A represents animal PA4-expressing GFP epitopes (red cells). Figure 6B, 6C; 6D, 6E; 6F, 6G; 6H, 6I; and 6J, 6K represents caldesmon; calponin; α-SMA; SMMHC; and transgelin (in vivo and in vitro samples). In vivo samples are dual stained with the anti-GFP antibody. There is ample expression of each smooth muscle marker that was derived from BM-derived MSCs used in the augmentation demonstrating that the MSC-seeded grafts contributed to regeneration of the muscle component of the bladder. Similarly, Figure 6L is Ki-67 stained tissue with the anti-GFP antibody showing clear evidence of proliferating bMSCs, whereas Figure 6M shows the cells staining positive for Ki-67 in vitro. Figure 6N is pan-cytokeratin staining of augmented tissue which is positive only for baboon urothelium indicated by white arrows. Cells were negative for pan-cytokeratin staining for in vitro conditions. Staining for elastin and osteocalcin were negative in both in vitro and in vivo samples. Results were similar for animal PA3. Quantification GFP+/marker+ cells found within the grafted area of animals PA3 and PA4 revealed >90% of the initial transplanted GFP+ cells displaying expression of smooth muscle-associated proteins or >70% with Ki-67+ (Table 1). There was also no evidence that the labeled bMSCs integrated into the luminal side of the graft. This data described clearly demonstrates that bMSCs can remain within the graft for extended periods with active proliferation and contribute to the regeneration of the muscular component of the bladder wall.

Figure 6.

Immunofluorescence imaging. (A): Anti–green fluorescent protein (GFP)/Alexa 568 (red). (B, C): Caldesmon. (D, E): Calponin. (F, G): Alpha smooth muscle actin. (H, I): Smooth muscle myosin heavy chain. (J, K): Transgelin. (L, M): Ki-67. (N): Pan-cytokeratin. (A, B, D, F, H, J, L, N): Tissue samples also with anti-GFP/Alexa 568 (×400; pan-cytokeratin ×200). (C, E, G, I, K, M): In vitro (×200). All images from PA4 and stained with 4',6-diamidino-2-phenylindole.

Table 1. Profile of transplanted baboon mesenchymal stem cells, 10 weeks postgraft: Expression of smooth muscle-associated proteins and proliferation marker Ki-67
  1. For both animals, >90% of transplanted GFP+ cells showed expression of smooth muscle-associated proteins, with a proliferation fraction of >70%. Data, shown as % positive, represent mean ± 95% confidence interval for 12–15 fields (totaling 1,915–3,910 GFP+ cells) surveyed per marker per animal.

  2. Abbreviations: α-SMA, alpha smooth muscle actin; SMMHC, smooth muscle myosin heavy chain.

inline image


The tightly orchestrated events involved with micturition can become greatly altered by bladder dysfunction. Regenerative medicine-based schemes that have been implemented to replace bladder tissue largely gone unfulfilled. Although bladder augmentation is standard procedure in instances of pediatric neuropathic bladder, there are many short- and long-term complications [28, 29]. A number of attempts to regenerate functional bladder tissue have used small animal rodent models and the canine species [9–12, 15, 30]. Although these models provide a vast array of information about the possible mechanisms involved with the bladder regenerative process, the question still remains as to whether or not these results are applicable to human subjects undergoing similar procedures. Hence, within the context of this study, we evaluated the use of a NHP baboon model of bladder regeneration.

The justification to use an animal model that is highly analogous to humans is warranted based on documented bladder anatomical and physiological differences between human and lower level primates [17, 18, 31, 32]. It has been assumed that data from small animal models could be directly extrapolated to the human condition. The cellular makeup of the bladder also varies quite dramatically when comparing different species. Fascicles within the muscle bundles of the human bladder detrusor muscle exist in specific orientations and complex arrangements which eventually dictate the physiological attributes of the bladder wall [33]. However, similar regions within the rabbit bladder, for example, appear much less structurally complex and simplistic in nature when compared with human bladder [34, 35]. Aside from the gross/ultrastructural anatomical differences, molecular differences are also apparent. Structural contractile proteins that are found within bSMCs, such as the family of actin molecules, are arranged in specific ratios to maximize contraction and relaxatory cycles of the bladder during micturition and bladder filling. In normal bSMCs, the α, β, and γ isoforms of actin are expressed in relative proportions of 33:25:42% (human), 41:19:40% (rat), and 44:10:46% (mouse) [35–37]. The disproportionateness between the rodent and human species raises the question of the translatability of bladder augmentation data from species to species. The skewing of the these ratios as in the case of a xenogenic tissue-based bladder augmentation model, may be more reflective of disease states or trauma rather than normal physiological conditions. Finally, the use of immunocompromised rodents typically used for bladder studies eliminates the investigation of any unknown or yet to be determined immunoregulatory components involved in the regenerative process.

Baboons, however, are phylogenetically very similar to humans and thus share many salient anatomical and physiological features. During the initial phases of baboon fetal development, the urinary tract grows and develops comparably to humans accentuated by a late burst of growth during in late gestation [38]. A variety of studies in numerous fields comprising decades of research activity with NHPs as preclinical models have been used in the elucidation of physiological mechanisms [39–41]. As there is a paucity of information with regards to baboon bladder regeneration, our study is the first report on a NHP bladder augmentation model using an autologous source of BM MSCs. This provides an exceptional proof of principle model that is clinically relevant, which is used to study the complexities involved with human bladder regeneration.

The regenerative features of human MSCs in a rodent bladder augmentation model have previously been described and demonstrate the great plasticity of MSCs [9]. Within the context of this model, bMSCs played a similar role in the regenerative process of the bladder. Epitope-defined bMSCs were able to assimilate into the augmented area of the bladder and function in what appears to be a normal manner as the augmented animals urinated normally following the augmentation procedure and displayed no evidence of ureteral, lithogenesis, or kidney disturbances. Animals PA3 and PA4 displayed robust muscle layer bladder architecture as seen in Figure 4 with no evidence of cell death or grossly necrotic tissue with muscle content approaching that of native tissue, whereas displaying muscle markers and bMSCs were also highly proliferative as indicated by Ki-67 staining. The bladder of the sham control animal, although reduced dramatically in capacity, appeared to also function normally and display normal bladder architecture. Finally, the approximate 50% reduction in muscle content based on quantitative morphometry data and high collagen levels in the unseeded SIS graft are suggestive of scarification as a result of an inflammatory reaction by the host against the SIS. It has been reported that SIS can invoke a localized inflammatory response within a rodent augmentation model [42]. The lack of scarification in the MSC-seeded scaffolds and the subsequent robust muscle growth may be attributed to the fact that MSCs can act as immune modulators. In a variety of different systems, MSCs have been shown to possess immunomodulatory characteristics including an abrogation in proinflammatory cytokines [43–45]. The grafts in our model may have benefited from the protective mechanism brought forth by the bMSCs. Fibrosis was not encountered on either side of the bladder as both sides of the graft were seeded with bMSCs. It was also noted that native baboon urothelium had invaded the grafted area to provide a watertight barrier, and seeded bMSCs were not present as determined by antibody staining. This phenomenon has been reported in other species as well [9, 12].

Urodynamic testing provides a means to assess the physiological aspects of the micturition process by gathering data to potentially diagnose and subsequently treat disorders of the lower urinary tract. A key urodynamic parameter is the volumetric capacity of the bladder which is directly dependent on the anatomy and structural integrity of the bladder. The level of heterogeneity with regards to bladder anatomy and subsequent capacity was evident within the animals in this study and is more representative of differing patient populations seen in clinical settings. Typical bladder capacity of adult humans is approximately 500 ml [46], whereas we observed preoperative bladder capacities for the baboons were approximately within 200–500 ml range. The removal of a significant portion of the bladder and ultimate restoration provided by the augment returned bladder capacities to acceptable levels at 10 weeks. Both animals grafted with seeded scaffolds had a mean 37% (± 4% SD, n = 2) recovered bladder capacity, which was indistinguishable from the sham control (39%, n = 1). The unseeded graft contained the largest disparity with regards to bladder capacity recovery at 28%. The high levels of collagen present within the augmentation patch may have contributed to a decrease in elastic modulus of the bladder as a whole, thus decreasing overall bladder capacity through x-rays (Fig. 2). This suggests that postaugmented animals had no overall, gross structural abnormalities of the bladder. We speculate that the bladder capacity may have reached normal levels in animals that were augmented with cell seeded scaffolds had the length of the study been extended, as other animal studies have demonstrated. Previously described canine models for bladder regeneration were carried out for approximately 1-year yielding data that demonstrated trends of increasing capacity over time with the implementation of a cell-seeded graft [30]. At the structural and functional levels, the seeded scaffolds performed similarly to the sham control and better than the unseeded control with regards to cystometry measurements. Although the animals were fairly homologous in physical makeup (sex and weight), there was quite a difference in measured bladder capacity. The anatomical differences with regards to female baboons have not been previously characterized and may explain the variation in bladder capacity as there are degrees of anatomical variation in humans as well. Alternatively, slight variations in body temperature, reaction to ventilation conditions, and anesthetics may have affected bladder capacity readings even though every possible measure was taken to create identical conditions to make these measurements for each animal. This study is the first to describe the use of a defined population of BM-derived MSCs in an autologous setting using a baboon urinary bladder augmentation model. Data obtained from this work provides insight into the bladder regenerative process at the gross, microscopic, and physiological levels as well as providing sufficient evidence that MSCs can be used to regenerate the smooth muscle portion of the bladder wall.


This newly described bladder augmentation model incorporating autologous sources of BM MSCs represents a unique insight into the bladder regenerative process. Analyses of tissue-engineered bMSC/SIS composites along with cystometry data provide strong evidence that MSCs can be exploited for tissue engineering purposes also illustrating that cell-seeded scaffolds function markedly better than the unseeded control and similar to the sham. This study demonstrates the feasibility of MSCs in a partial bladder regenerative setting using a model that may be directly extrapolated to the human condition.


We thank Amber Hoggat, DVM, Heather Charles, CVT, Susana Stacha, CVT, and Sarah Smith, CVT for their assistance. This study was supported by a generous gift from Sara C. Star. We also thank Drs. Asgi Fazleabas (Michigan State University) and Nadim Mahmud (University of Illinois at Chicago) for thoughtful discussions.