Tissue Engineering Part A: Research Advances

The purpose of this study was to examine the growth and morphology of a salivary epithelial cell line (HSG) in vitro on several biodegradable substrata as an important step toward developing an artificial salivary gland. The substrates examined were poly-L-lactic acid (PLLA), polyglycolic acid (PGA), and two co-polymers, 85% and 50% PLGA, respectively. The substrates were formed into 20- to 25-mm disks, and the cells were seeded directly onto the polymers or onto polymers coated with specific extracellular matrix proteins. The two copolymer substrates became friable over time in aqueous media and proved not useful for these experiments. The purified matrix proteins examined included fibronectin (FN), laminin (LN), collagen I, collagen IV, and gelatin. In the absence of preadsorbed proteins, HSG cells did not attach to the polymer disks. The cells, in general, behaved similarly on both PLLA and PGA, although optimal results were obtained consistently in PLLA. On FN-coated PLLA disks, HSG cells were able to form a uniform monolayer, which was dependent on time and FN concentration. Coating of disks with LN, collagen I, and gelatin also promoted monolayer growth. This study defines the conditions necessary for establishing a monolayer organization of salivary epithelial cells with rapid proliferation on a biodegradable substrate useful for tissue engineering.

Clinical outcomes of bone allograft procedures may be improved by modifying the surface of the graft with an osteoconductive biopolymeric coating. In this comparative in vitro study, we evaluated the dimensional stability, mechanical strength, hydrophilicity, and water uptake of biodegradable foams of poly(propylene fumarate) (PPF) and poly(d,l-lactic-co glycolic acid) (PLGA) when applied as surface coatings to cortical bone. Cortical bone samples were divided into four groups: Type I, untreated bone; Type II, laser-perforated bone; Type III, partially demineralized bone; and Type IV, laser-perforated and partially demineralized bone. Results show that PPF wets easily, achieving 12.5% wt/wt in 30 min. Compressive tests on the PPF foam material showed that the compressive strength was 6.8 MPa prior to in vitro incubation but then gradually reduced to 1.9 MPa at 8 weeks. Push-out and pulloff strength tests showed that initially both PPF and PLGA foam coatings had comparable adherence strengths to the cortical bone samples (100-150 N). When additional geometrical surface alteration by perforation and demineralization of the bony substrate was employed, in vitro adherence of the PPF foam coating was further increased to 120 N, demonstrating a statistically significant improvement of push-out strength throughout the entire 8-week observation period (p < 0.0002 for all four data points). The pore geometry of PPF-foam coatings changed little over the 2-month evaluation period. In comparison, PLGA foam coating around the cortical bone samples rapidly lost structure with a decrease of 67% in strength seen after 1-week in vitro incubation. These new types of bone allografts may be particularly useful where the use of other replacement materials is not feasible or practical.

For in vivo cell implantation techniques to be successful, the energy and metabolic substrate requirement of the cells being grown must be met. Certain cells with high-energy requirements (e.g., hepatocytes, pancreatic island cells) experience a high degree of cell death after implantation due to a limited supply of oxygen. We proposed that the pleural cavity might be an oxygen-rich environment and hence an excellent site for cell implantation. To test the hypothesis that the delivery of oxygen to the pleural cavity is directly proportional to the inspired oxygen concentration we measured the pO₂ of saline instilled in the pleural cavity as compared to that of the peritoneal cavity. We postulated that the physiologic basis for any difference was the result of direct diffusion of oxygen into the pleural space across the alveoli. The study was conducted on sheep (n = 6), after induction of general anesthesia, in two phases, control and experimental. Saline was instilled into the peritoneal and pleural cavities via catheters, after equilibration at given FiO₂, the pO₂ of the paline aspirated from the two cavities was compared. In the experimental group, animals were sacrificed (no circulation) and ventilated. The same sequence of steps as in the control phase were repeated. In the control group, the pO₂ of saline aspirated from the pleural cavity approached the arterial pO₂ at all FiO₂ levels. The pO₂ of the peritoneal saline aspirate fell over time. In the experimental phase (no circulation), the pO₂ of the pleural cavity saline rose to >400 mmHg. We conclude that this is a result of direct diffusion and is a potential source of unlimited oxygen supply not dependent on vascular supply.

Small intestinal submucosa (SIS) is a naturally occurring, acellular biomaterial derived from porcine jejunum, which promotes constructive tissue remodeling when applied as a xenogeneic graft material. Galactosyl-α(1,3)galactose (Gal) is a cell-associated epitope responsible for hyperacute rejection of porcine whole-organ xenografts in primates. Because SIS is harvested from porcine tissue, it may contain the Gal epitope. The goals of this study were to determine if Gal is present in SIS and, if it is present, to determine if human serum complement can be activated in vitro following exposure to porcine-derived SIS. SIS was probed for Gal by immunohistochemical methods and by lectin-peroxidase staining. SIS stained strongly positive with human serum, which contains naturally occurring antibodies to Gal, followed by anti-immunoglobulin G (IgG) or anti-IgM peroxidase conjugate. Blocking with the lectin I-B₄, which is specific for the Gal epitope, decreased the intensity of staining. Exposure of SIS to α-galactosidase reduced staining to negligible amounts. The Gal epitope is distributed transmurally throughout the SIS material. Subtyping of the immunoglobulins that bind to SIS showed that IgG₂ is the major immunoglobulin of human plasma that binds to SIS. SIS did not activate complement in vitro as measured by radioimmunoassay for C3a.

Endothelial cell seeding of synthetic small diameter vascular grafts (SSDVG) has been shown to diminish thrombosis and intimal hyperplasia, resulting in improved graft patency. However, endothelial cell retention on seeded grafts when exposed to physiological shearing conditions remains poor. We report that the genetic engineering of endothelial cells to overexpress endothelial nitric oxide synthase (eNOS), may create improved anti-thrombotic and anti-hyperplastic endothelial cell phenotypes for SSDVG seeding. eNOS-overexpressing endothelial cells may potentially overcome the biochemical loss due to shear induced reduction in endothelial cell coverage on SSDVG. Bovine aortic endothelial cells (BAEC) were transfected with the human eNOS gene, and co-incubated with either human whole blood or bovine aortic smooth muscle cells (BASMC) in vitro. eNOS-transfected BAEC significantly overexpressed eNOS compared to control β-Gal-transfected and untransfected BAEC up to 120 h post transfection. In co-incubation and co-culture assays, human platelet aggregation decreased by 46% and BASMC proliferation decreased by 67.2% when compared to incubation with untransfected BAEC.

As cell-based therapies receive approval for clinical evaluation and use, the development of reliable methods to quantify cell number and control the dose of therapy delivered is becoming increasingly important. An example is the determination of the number and volume of primary porcine hepatocytes used in an extracorporeal treatment for patients with liver disease. Conventional cell counting using optical microscopy was compared against two alternate methods to quantify isolated porcine hepatocytes: (1) automated cell counting using a commercially available particle characterization instrument, and (2) quantitation by cell mass. Methods were compared based on accuracy, precision, specificity, linear range, and ruggedness. The automated method delivered substantially improved accuracy, precision, and ruggedness when compared to the conventional optical method. It also provided valuable information about the size distribution of cell preparations, which often contained clumps of cells, and showed that processing steps such as cryopreservation can alter the size characteristics of a cell population. The automated method was also faster, and was well suited to use in a commercial manufacturing process. The mass-based method was simple and inexpensive, but suffered from nonlinearity at low cell concentrations. Automated cell quantitation using a commercially available particle characterization instrument proved to be the preferred method for obtaining accurate and consistent porcine hepatocyte counts in a timely manner.

In vivo magnetic resonance imaging (MRI) and relaxometry were performed to assess noninvasively the tissue reaction and the biological integration of hydrogels made of poly[N-(2-hydroxypropyl) methacrylamide] (PHPMA) after implantation in the trapezius muscle of rabbits. The benefits of incorporating RGD peptide sequences in the polymer backbone were also investigated. The histological status of each implant was probed by the trend of their transversal relaxation times, T₂, while their biocompatibility was evaluated by analyzing the host tissue response through the evolution of the relaxation times of the adjacent muscle tissue. MR results showed the good acceptability of both hydrogels by the host tissue. The transversal relaxation curves of each implant exhibited two distinct phases as a function of implantation time: (1) a monoexponential phase, dominated by the influx of fluids inside the implants; and (2) a biexponential phase related to the infiltration of cells and the granulation tissue formation within the porous structure of each polymer. These MR findings were correlated with the results of conventional histological analyses. The present study demonstrates the effectiveness of MR methods in noninvasively monitoring the biocompatibility and histological status of implanted porous biomaterials.

De novo adipogenesis at the implanted site of a basement membrane extract (Matrigel) was induced through controlled release of basic fibroblast growth factor (bFGF). bFGF was incorporated into biodegradable gelatin microspheres for its controlled release. When the mixture of Matrigel and bFGF-incorporated gelatin microspheres was implanted subcutaneously into the back of mice, a clearly visible fat pad was formed at the implanted site 6 weeks later. Histologic examination revealed that the de novo formation of adipose tissue accompanied with angiogenesis was observed in the implanted Matrigel at bFGF doses of 0.01, 0.1, and 1 μg/site, the lower and higher doses being less effective. The de novo formation induced by the bFGF-incorporated microspheres was significantly higher than that induced by free bFGF of the same dose. The mRNA of a lipogenesis marker protein, glycerophosphate dehydrogenase, was detected in the formed adipose tissues, biochemically indicating de novo adipogenesis. Free bFGF, the bFGF-incorporated gelatin microspheres, or Marigel alone and bFGF-free gelatin microspheres with or without Matrigel did not induce formation of adipose tissue. This de novo adipogenesis by mixture of Matrigel and the bFGF-incorporated gelatin microspheres will provide a new idea for tissue engineering of adipose tissue.