|Home | About | Journals | Submit | Contact Us | Français|
Adipose tissue is often a key structure to restore in reconstructive and augmentative surgeries. Current materials for soft tissue reconstruction or augmentation suffer from shortcomings such as suboptimal volume retention, donor site morbidity, and poor biocompatibility. A series of experiments are presented here to describe our stem cell–based approach to engineering human adipose tissue with predefined shape and dimensions. These findings indicate the real possibility that biologically viable adipose tissue can be engineered by taking a teaspoon full of tissue fluid containing the patient's adult stem cells, expanding them, differentiating them into adipogenic cells, and encapsulating them into appropriate biocompatible polymer materials. The end result is anticipated to be minimal donor site trauma related to needle size, immune compatibility because the patient's own stem cells are used, and long-term volume maintenance because stem cells are capable of replenishing adipogenic cells to retain the predefined shape and dimensions of the engineered soft tissue. Upcoming challenges include long-term volume maintenance, tissue maturation, angiogenesis, scaling up, and host tissue integration. Conceptually, stem cell–derived soft tissue grafts are realizable in plastic and reconstructive surgical procedures.
Soft tissue defects represent substantial challenges for contemporary medical practice. Many types of breast cancer and facial cancer are life threatening and, once resected, leave patients with soft tissue defects and disfiguration. Mastectomy and tumor resection surgeries are examples of nonelective surgical procedures that mandate the replacement of lost soft tissue to restore physical shape and physiological function. Facial reconstructive surgeries are necessary after cancer resections such as basal cell carcinoma, squamous cell carcinoma, melanoma, and other head and neck cancers.1,2,3 Improved medical managements of cancer have prolonged survival rate, with the result that cancer survivors wish to have the morphology and function of their soft tissue defects restored.1,4,5 Injuries take place during war, leading to soft tissue trauma in addition to skeletal injuries.6,7 Work-related accidents and traffic accidents are additional causes of soft tissue defects. Soft tissue is missing after burns, necessitating the reconstruction of not only the skin but also subcutaneous soft tissue.8 In congenital anomalies such as hemifacial microsomia (one half of the face underdeveloped relative to the other half), a considerable amount of soft tissue is missing and needs to be reconstructed.9,10 In 2003, about 6 million individuals in the United States received reconstructive surgical procedures performed by plastic surgeons.11
For decades, surgeons have utilized their creative wisdom to improve soft tissue reconstruction procedures. Autologous soft tissue grafts and synthetic materials are predominant in soft tissue reconstruction procedures. Each approach has its own pros and cons and is suitable for specific types of patients and soft tissue defects based on clinical judgment. The current procedures for soft tissue reconstruction and augmentation fall short of the ideal, with adverse side effects such as donor site morbidity, leakage, extrusion, unnatural texturing, resorption, and immune rejection. Although autologous mature adipose tissue has been used as the preferable filler for soft tissue defects, the results are unpredictable.12 Volume reduction appears to be a primary concern, with autologous grafts resorbing up to 60%.13,14 Volume reduction of autologous soft tissue grafts may be due to apoptosis of the transplanted mature adipocytes because of their low tolerance for ischemia and slow revascularization rate.13,14 Lipectomy or liposuction procedures provide another route of fat transplantation. Single cell suspensions of mature adipocytes isolated from liposuction aspirate have been incorporated in scaffolds and transplanted to fill soft tissue defects. However, volume reduction remains an issue probably because 85% of the cytoplasm of mature adipocytes is lipid,14 leading to a great need for angiogenesis. In the most severe cases, fat grafts from liposuction aspirates are associated with suboptimal blood supply and necrosis of the graft. Mechanical stresses resulting from liposuction are projected to damage up to 90% of the adipocytes upon implantation.13 Another complicating factor is that mature adipocytes are fully differentiated and do not proliferate,15 leading to a shortage of adipogenic cells in fat grafts.
Preadipocytes are preferable to mature adipocytes because preadipocytes are believed to be capable of proliferating to a certain extent ex vivo and tolerate hypoxic environments to a greater extent than mature adipocytes.16,17,18 Several preadipocyte cell lines such as 3T3-L1 and 3T3-F442A have been used for engineering adipose tissues in biomaterials such as PGA (polyglycolic acid) meshes.19,20 These adipogenic cell lines, despite their considerable value for in vitro studies, are immortalized and thus not as valuable as primary preadipocytes for in vivo engineering of adipose tissue. Primary preadipocytes can be isolated from animal or human fat tissue by excision or liposuction procedures and have been successfully used for in vivo engineering of adipose tissue. For example, preadipocytes seeded in porous PLA (polylactic acid) scaffolds generate adipose tissue in vivo.17 Peptide-linked alginate implants supported the adhesion and proliferation of seeded sheep preadipocytes and adipose tissue formation.21 Basic fibroblast growth factor exogenously delivered to human and murine preadipocytes promoted continuous adipogenic differentiation and adipose tissue formation in collagen or Matrigel scaffolds.22,23 However, some preadipocytes are probably damaged during excision or liposuction procedures.16 At this time, it is uncertain whether preadipocytes can be expanded ex vivo to sufficient numbers for healing large soft tissue defects. Whether preadipocytes can proliferate for sufficient cell cycles to maintain the long-term viability of soft tissue grafts is also unknown. Thus, more studies using preadipocytes are warranted to determine their ex vivo expandability and in vivo renewability.
Mesenchymal stem cells (MSCs) derive from embryonic stem cells and differentiate into cell lineages that form all connective tissues such as adipose tissue, cartilage, bone, skeletal muscles, and interstitial fibrous tissue.24,25,26,27,28,29 MSCs can be isolated from bone marrow, adipose tissue, deciduous teeth, and skeletal muscle. In addition to their multipotent nature, MSCs can self-renew and replicate for many passages without losing their stemness.24 MSCs were first isolated and identified as fibroblast-like cells that are adherent to cell culture Petri dishes.30 In humans, MSCs are typically aspirated from the bone marrow of the superior iliac crest, whereas murine MSCs are commonly aspirated from the mid-diaphysis of the tibia and femur.24 Upon exposure to various established induction media such as adipogenic, chondrogenic, osteogenic, and myogenic media, MSCs differentiate into distinct pathways and begin to express genes and protein markers characteristic of corresponding cell lineages.25,26,31
MSCs are progenitors of adipogenic cells and adipocytes. Although preadipocytes give rise to adipocytes, it is not clear whether preadipocytes can replicate as readily as MSCs. It is known that end-stage adipocytes neither replicate nor differentiate further. Thus, a shortage of adipose tissue–forming cells contributes, among other factors, to volume reduction after current soft tissue grafting procedures. It is not surprising that the surgical literature is replete with continuing aspiration to “grow” soft tissue grafts from the patient's own stem cells for reconstructive needs.4,32,33,34,35 Tissue engineering of adipose tissue implants from the patient's own adult stem cells is believed to be capable of overcoming most of these associated deficiencies.35 Pittenger et al published a seminal study demonstrating that human mesenchymal stem cells (hMSCs) can be differentiated into adipogenic cells in monolayer culture in Petri dishes.25 Subsequently, a large number of genes expressed during adipogenic differentiation of human MSCs in vitro have been delineated.36 MSC-derived adipocytes are anticipated to proliferate longer and yield a larger cell population in ways that are exactly what is needed for adipose tissue engineering. Our recent study demonstrates that hMSC-derived adipogenic cells generate adipose tissue grafts in vivo in a biocompatible hydrogel.27 The shape and dimensions of the hMSC-derived adipose tissue grafts are maintained virtually 100% after 4 weeks of in vivo implantation subcutaneously in the dorsum of immunodeficient mice.27 Our additional data discussed here further demonstrate this effect.
A scaffold material is usually required to provide initial adhesion and support for tissue-forming cells in engineered grafts. Selecting the appropriate scaffold for a particular tissue engineering application is crucial to the success of the endeavor. There is no one scaffold for every application, and several scaffolds should be investigated regarding their biomaterial properties and cytocompatibility. Tissue development is dependent on the structural environment, cell-biomaterial interaction, and potentially biological signals incorporated in the scaffold.34
It is notable that previous work on cell-based adipose tissue engineering has primarily utilized porous scaffolds.19,21,22,37 The advantage of porous scaffolds is their potential to facilitate vascular ingrowth. However, a common drawback of porous scaffolds in adipose tissue engineering is premature resorption or degradation, or both, of the porous scaffold, potentially associated with enzymes produced by invading host cells. Hydrogels have been used extensively as tissue engineering scaffolds. Hydrogels encompass a broad arena of scaffolds including collagen, elastin, hyaluronic acid, agarose, alginate, chitosan, and polyethylene glycol (PEG) hydrogels. PEG hydrogels have been used to encapsulate various cell types, including chondrocytes and osteoblasts.26,27,38,39 The rheological and recovery properties of PEGDA hydrogel have been studied with several parameters found to be comparable to human abdominal adipose tissue.40 PEG hydrogel is advantageous as a scaffold for adipose tissue engineering applications. PEG is hydrophilic, biocompatible, and undergoes slow degradation, which is useful for volume retention of the engineered adipose tissue.41,42 Indeed, we have found that hMSC-derived adipogenic tissue grafts in PEG hydrogel maintained the predefined shape and dimensions virtually 100% after 4 weeks of in vivo implantation subcutaneously in the dorsum of immunodeficient mice.27 Our data demonstrate that PEG hydrogel encapsulating hMSC-derived adipogenic cells maintains is far superior in shape and dimension maintenance to collagen sponges.
Reconstructive and plastic surgeries are about shape and dimensions, in addition to the scientific basis of surgical practice.18 Hypothetically, a successfully tissue-engineered kidney from stem cells, when realized, does not need to have the precise shape of the patient's normal kidney as long as the engineered kidney functions well in vivo. In contrast, soft tissue defects must be restored to have the original shape and dimensions in addition to having physiological function. Current soft tissue reconstruction procedures suffer from postoperative volume reduction or shrinkage.14,43,44 Volume reduction can be as severe as up to 60% over time.43,44,45 In fact, “postoperative volume reduction” of soft tissue grafts is identified as a key issue among other complications of soft tissue reconstruction procedures in the surgical literature.14,43,44,45
Several previous reports have shown that the dimensions of the engineered adipose tissue are retained to various degrees.16,21,37 For example, porous alginate hydrogel seeded with fibroblasts appears to be more successful in retaining the volume of engineered soft tissue than cell-free hydrogel.37 Maintenance of volume of the alginate constructs seeded with preadipocytes is between 19% to 88% after 8 weeks of subcutaneous implantation in rats, with the highest percentage of volume maintenance upon seeding fibroblasts in alginate and in situ solidification after subcutaneous injection.37 However, in all previous reports some changes in the shape and dimensions of the engineered adipose tissue have occurred.16,19,21,37 Our data demonstrate that the shape and dimensions of the adipose tissue engineered by encapsulating adipogenic cells in a PEG hydrogel with pore sizes that are sufficiently small to disallow host cell invasion have been maintained virtually 100% following 4 weeks of in vivo implantation.
Although volume reduction is probably caused by a multitude of factors such as scarring or suboptimal tissue integration, the key issue is hypothesized to be a shortage of incorporated cells that are capable of (1) sustained synthesis of adipose and fibrous matrices to maintain the volume and (2) self-replication to replenish the supply of adipose and fibrous matrix-forming cells. These two needs are not only what MSCs are designed for but also what end-stage cells such as adipocytes cannot offer. Thus, tissue engineering approaches to improve soft tissue reconstruction procedures unavoidably must deal with volume retention.
We differentiated commercially available hMSCs, from an anonymous adult donor, into adipogenic cells by our previous method.27 Then hMSC-derived adipocytes were encapsulated in PEG hydrogel and implanted into surgically created subcutaneous pockets in the dorsum of immunodeficient mice. After 4 weeks of in vivo implantation, stem cell–derived adipose tissue grafts were harvested. Figure Figure11 demonstrates the retrieval process for PEG hydrogel scaffolds. Of particular interest, the PEG hydrogel encapsulating hMSC-derived adipogenic cells, such as those in Figure Figure1D,1D, appeared darker in color and well attached to the host subcutaneous tissue (Fig. 1). Figure Figure22 demonstrates the shape, dimensions, and different photo-opaqueness of PEG adipose tissue grafts after in vivo implantation. Notice in particular that the PEG hydrogel construct encapsulating hMSC-derived adipogenic cells not only maintained the original shape and dimensions but also demonstrated considerable photo-opaqueness (Fig. 2D), suggesting that a substantial amount of biological structures had been formed in the hMSC-derived adipogenic PEG construct. In contrast to the PEG hydrogel constructs, collagen sponges harvested after the same 4-week in vivo implantation lost their original shape and dimensions (Fig. 3). As seen in Figure Figure4,4, all collagen sponges, whether cell free or seeded with hMSCs or hMSC-derived adipogenic cells, lost the original shape and shrank a substantial amount (Fig. 4). Oil Red O staining, a histological marker for lipid accumulation, of hMSC-derived adipogenic PEG and collagen constructs demonstrated positive staining (Fig. 5). However, H&E staining showed no host cell invasion into the PEG constructs, suggesting that all the generated adipose tissue was from the implanted human MSC-derived adipogenic cells (Fig. 5A). In comparison, there are likely to be host mouse cells invading collagen sponges, leaving the possibility of a host cell contribution to adipogenesis within the collagen constructs (Fig. 6).
Human MSCs can readily differentiate into adipogenic cells not only in monolayer culture in Petri dishes but also in three-dimensional hydrogel and in vivo in animal models, as evidenced in the present study. Human MSCs can be readily isolated from patients whose soft tissue defects need surgical repair so that autologous cell therapy is available to circumvent issues of immunorejection.46 Our present work has demonstrated that hMSC-derived adipose tissue grafts in PEG hydrogel are capable of maintaining their predefined shape and dimensions virtually 100% following 4 weeks of subcutaneous implantation.27 Further long-term studies are necessary to characterize shape and dimension maintenance after extended in vivo implantation periods.
Angiogenesis remains a challenge for engineered tissues including adipose tissue. Despite isolated meritorious attempts, controlled angiogenesis needs to be induced in engineered soft tissue grafts. Recent work has shown that angiogenesis in fat pads derived from preadipocyte cell lines is from the host instead of preadipocytes,20 indicating the need to stimulate host angiogenesis. Although adipose tissue is the primary component of soft tissue, soft tissue consists of more than adipose tissue. For example, host fibrous tissue may be needed in engineered soft tissue grafts. Several meritorious approaches for inducing angiogenesis in bone tissue engineering can be instructive for enabling angiogenesis in engineered adipose tissue.
Scale-up is a common tissue engineering challenge including the engineering of adipose tissue. The engineered adipose tissue graft in the present work is up to 9 mm in diameter and is sufficient for certain facial soft tissue defects and other small soft tissue defects. However, soft tissue defects after mastectomy and large trauma demand multiple fold scale-ups. Issues such as mass transport, nutrient diffusion, angiogenesis, and nerve innervation may become increasingly challenging upon scaling up of engineered soft tissue.
Despite all the challenges, we are at an exciting time in the history of innovation in reconstructive and augmentative surgical procedures. Several research and development teams have devised well-designed strategies to deal with many of the issues identified here. With influx of bioengineering talents into the field of cell-based soft tissue reconstruction and regeneration, we are optimistic that innovative therapies with long-lasting results are likely to be realized sooner than many of us are willing to predict.
We are grateful to Aurora Lopez for general technical assistance. The research work described here has been supported by grants R01DE13964 and R01DE15391 from the National Institute of Dental and Craniofacial Research and R01EB02332 from the National Institute of Biomedical Imaging and Bioengineering, National Institutes of Health.