Regenerative Medicine: Past and Present
- First Online:
- Cite this article as:
- Atala, A. Medicine Studies (2009) 1: 11. doi:10.1007/s12376-009-0008-6
- 77 Views
Novel therapies resulting from regenerative medicine and tissue engineering technology may offer new hope for patients with injuries, end-stage organ failure, degenerative disorders and many other clinical issues. Currently, patients suffering from diseased and injured organs are treated with transplanted organs. However, there is a shortage of donor organs that is worsening yearly as the population ages and new cases of organ failure increase. Scientists in the field of regenerative medicine and tissue engineering are now applying the principles of cell transplantation, material science, and bioengineering to construct biological substitutes that can restore and maintain normal function in diseased and injured tissues. In addition, the stem cell field is a rapidly advancing part of regenerative medicine, and new discoveries in this field create new options for this type of therapy. For example, new types of stem cells, such as amniotic fluid and placental stem cells, that can circumvent the ethical issues associated with embryonic stem cells have been discovered. The process of somatic cell nuclear transfer and the creation of induced pluripotent cells provide still other potential sources of stem cells for cell-based tissue engineering applications. While stem cells are still in the research phase, some therapies arising from tissue engineering endeavors that make use of autologous, adult cells have already entered the clinical setting, indicating that regenerative medicine holds much promise for the future.
KeywordsRegenerative medicineStem cellTissue engineeringCell transplantationBiomaterials
In a generous interpretation, we find four medical firsts described in this passage: the use of anesthesia, surgery, cloning, and tissue engineering. Just thirty years ago, two of these, tissue engineering and cloning, were not possible. However, the use of one body part for another, or the exchange of parts from one person to another was mentioned in the medical literature even in antiquity.
So the Lord God caused the man to fall into a deep sleep, and while he was asleep, He took part of the man’s rib, and closed up the place with flesh. Then the Lord God made a woman from the part He had taken out of the man, and He brought her to the man.
The field of urology was the earliest to gain from the advent of transplantation, with the kidney being the first entire organ to be replaced in a human in 1955 (Murray et al. 1955). In the early 1960s, Murray performed an allogenic kidney transplantation from a nongenetically identical patient into another. This transplant, which overcame the immunologic barrier, marked a new era in medical therapy and opened the door for use of transplantation as a means of therapy for different organ systems. However, lack of good methods for immunosuppression, the inability to monitor and control rejection, and a severe organ donor shortage opened the door for other alternatives.
As time passed, synthetic materials were introduced to replace or to rebuild diseased tissues or parts in the human body. The manufacture of new materials, such as tetrafluoroethylene (Teflon) and silicone, opened a new field that included a wide array of devices that could be applied for human use. Although these devices could provide for structural replacement, the functional component of the original tissue was not restored. Simultaneously, increased knowledge of the biologic sciences, including cell biology, molecular biology, and biochemistry, led to new techniques for cell harvesting, culture, and expansion. Studies of the extracellular matrix and its interaction with cells, and with growth factors and their ligands, led to better understanding of cell and tissue growth and differentiation.
In the 1960s, researchers began to combine the new devices and materials sciences with cell biology, and a new field that is now termed tissue engineering was born. As more scientists from different fields came together with the common goal of tissue replacement, the field of tissue engineering became more formally established. Tissue engineering is now defined as “an interdisciplinary field which applies the principles of engineering and life sciences towards the development of biological substitutes that aim to maintain, restore or improve tissue function”. The first use of the term tissue engineering in the literature can be traced to a reference dealing with corneal tissue in 1985. In the last two decades, scientists have attempted to engineer virtually every tissue of the human body. This article will review some of the progress that has been achieved.
The Basic Components of Tissue Engineering and Regenerative Medicine
The field of regenerative medicine encompasses various areas of technology, such as tissue engineering, stem cells, and cloning. Tissue engineering, one of the major components of regenerative medicine, follows the principles of cell transplantation, materials science, and engineering towards the development of biological substitutes that can restore and maintain normal function. Tissue engineering strategies generally fall into two categories: a) the use of acellular matrices, which depend on the body’s natural ability to regenerate for proper orientation and direction of new tissue growth, and b) the use of matrices with cells. Acellular tissue matrices are usually prepared by manufacturing artificial scaffolds, or by removing cellular components from tissues via mechanical and chemical manipulation to produce collagen-rich matrices (Dahms et al. 1998; Piechota et al. 1998; Yoo et al. 1998; Chen et al. 1999). Cells can be used for therapy via injection, either with carriers such as hydrogels, or alone.
When cells are used for tissue engineering, a small piece of donor tissue is dissociated into individual cells. These cells are either implanted directly into the host, or are expanded in culture, attached to a support matrix, and then reimplanted into the host after expansion. The source of donor tissue can be heterologous (such as bovine), allogeneic (same species, different individual), or autologous. Ideally, both structural and functional tissue replacement will occur with minimal complications. The preferred cells to use are autologous cells, where a biopsy of tissue is obtained from the host, the cells are dissociated and expanded in culture, and the expanded cells are implanted into the same host (Cilento et al. 1994; Atala 1998, 1999, 2001; Yoo et al. 1998; Amiel and Atala 1999; Oberpenning et al. 1999; Yoo et al. 1999; Amiel et al. 2006). The use of autologous cells, although it may cause an inflammatory response, avoids rejection and thus the deleterious side effects of immunosuppressive medications can be avoided.
Most current strategies for tissue engineering depend upon a sample of autologous cells from the diseased organ of the host. However, for many patients with extensive end-stage organ failure, a tissue biopsy may not yield enough normal cells for expansion and transplantation. In other instances, primary autologous human cells cannot be expanded from a particular organ, such as the pancreas. In these situations, stem cells are envisioned as being an alternative source of cells from which the desired tissue can be derived. Stem cells can be derived from discarded human embryos (human embryonic stem cells), from fetal tissue, or from adult sources (bone marrow, fat, skin).
Cloning has also played a role in the development of the field of regenerative medicine. So-called “nuclear transplantation” or “nuclear transfer” involves the introduction of a nucleus from a donor cell into an enucleated oocyte to generate an embryo with a genetic makeup identical to that of the donor. Stem cells can be derived from this source, which may have the future potential to be used therapeutically. Therefore, this type of cloning has also been called “therapeutic cloning”.
For cell-based tissue engineering, the expanded cells are seeded onto a scaffold synthesized with the appropriate biomaterial. In tissue engineering, biomaterials replicate the biologic and mechanical function of the native extracellular matrix (ECM) found in tissues in the body by serving as an artificial ECM. Biomaterials provide a three-dimensional space for the cells to form into new tissues with appropriate structure and function, and also can allow for the delivery of cells and appropriate bioactive factors (e.g. cell adhesion peptides, growth factors), to desired sites in the body (Kim and Mooney 1998). As the majority of mammalian cell types are anchorage-dependent and will die if no cell-adhesion substrate is available, biomaterials provide a cell-adhesion substrate that can deliver cells to specific sites in the body with high loading efficiency. Biomaterials can also provide mechanical support against in vivo forces such that the predefined three-dimensional structure is maintained during tissue development. Furthermore, bioactive signals, such as cell-adhesion peptides and growth factors, can be loaded along with cells to help regulate cellular function.
The ideal biomaterial should be biodegradable and bioresorbable to support the replacement of normal tissue without inflammation. Incompatible materials are destined for an inflammatory or foreign-body response that eventually leads to rejection and/or necrosis. Degradation products, if produced, should be removed from the body via metabolic pathways at an adequate rate that keeps the concentration of these degradation products in the tissues at a tolerable level (Bergsma et al. 1995). The biomaterial should also provide an environment in which appropriate regulation of cell behavior (adhesion, proliferation, migration, and differentiation) can occur such that functional tissue can form. Cell behavior in the newly formed tissue has been shown to be regulated by multiple interactions of the cells with their microenvironment, including interactions with cell-adhesion ligands (Hynes 1992) and with soluble growth factors. Since biomaterials provide temporary mechanical support while the cells undergo spatial reorganization into tissue, the properly chosen biomaterial should allow the engineered tissue to maintain sufficient mechanical integrity to support itself in early development, while in late development, it should have begun degradation such that it does not hinder further tissue growth (Kim and Mooney 1998).
Generally, three classes of biomaterials have been utilized for engineering tissues: naturally derived materials (e.g., collagen and alginate), acellular tissue matrices (e.g., bladder submucosa and small intestinal submucosa), and synthetic polymers such as polyglycolic acid (PGA), polylactic acid (PLA), and poly(lactic-co-glycolic acid) (PLGA). These classes of biomaterials have been tested in respect to their biocompatibility (Pariente et al. 2001, 2002). Naturally derived materials and acellular tissue matrices have the potential advantage of biological recognition. However, synthetic polymers can be generated reproducibly on a large scale with controlled properties such as strength, degradation rate, and microstructure.
Naturally Derived Materials
Collagen is the most abundant and ubiquitous structural protein in the body, and may be readily purified from both animal and human tissues with an enzyme treatment and salt/acid extraction (Li 1995). Collagen implants, under normal conditions, degrade through a process involving phagocytosis of collagen fibrils by fibroblasts (Arora et al. 2000). This is followed by sequential attack by lysosomal enzymes including cathepsins B1 and D. Under inflammatory conditions, the implants can be rapidly degraded largely by matrix metalloproteins (MMPs) and collagenases (Arora et al. 2000). However, the in vivo resorption rate of a collagen implant can be regulated by controlling the density of the implant and the extent of intermolecular crosslinking. The lower the density, the greater the space between collagen fibers and the larger the pores for cell infiltration, leading to a higher rate of implant degradation. Collagen contains cell adhesion domain sequences (e.g., RGD) that may assist to retain the phenotype and activity of many types of cells, including fibroblasts (Silver and Pins 1992) and chondrocytes (Sams and Nixon 1995).
Alginate, a polysaccharide isolated from seaweed, has been used as an injectable cell delivery vehicle (Smidsrod and Skjak-Braek 1990) and a cell immobilization matrix (Lim and Sun 1980) owing to its gentle gelling properties in the presence of divalent ions such as calcium. Alginate is relatively biocompatible and approved by the Food and Drug Administration (FDA) for human use as wound dressing material. Alginate is a family of copolymers of d-mannuronate and l-guluronate. The physical and mechanical properties of alginate gel are strongly correlated with the proportion and length of polygluronic block in the alginate chains (Smidsrod and Skjak-Braek 1990).
Acellular Tissue Matrices
Acellular tissue matrices are collagen-rich matrices prepared by removing cellular components from tissues. The matrices are often prepared by mechanical and chemical manipulation of a segment of tissue (Dahms et al. 1998; Piechota et al. 1998; Chen et al. 1999; Yoo et al. 1998). The matrices slowly degrade upon implantation, and are replaced and remodeled by ECM proteins synthesized and secreted by transplanted or in growing cells.
Polyesters of naturally occurring a-hydroxy acids, including PGA, PLA, and PLGA, are widely used in tissue engineering. These polymers have gained FDA approval for human use in a variety of applications, including sutures (Gilding 1981). The ester bonds in these polymers are hydrolytically labile, and these polymers degrade by nonenzymatic hydrolysis. The degradation products of PGA, PLA, and PLGA are nontoxic natural metabolites and are eventually eliminated from the body in the form of carbon dioxide and water (Gilding 1981). The degradation rate of these polymers can be tailored from several weeks to several years by altering crystallinity, initial molecular weight, and the copolymer ratio of lactic to glycolic acid. Since these polymers are thermoplastics, they can be easily formed into a three dimensional scaffold with a desired microstructure, gross shape, and dimension by various techniques, including molding, extrusion, solvent casting (Mikos et al. 1994), phase separation techniques, and gas foaming techniques (Harris et al. 1998). Many applications in tissue engineering often require a scaffold with high porosity and ratio of surface area to volume. Other biodegradable synthetic polymers, including poly(anhydrides) and poly(ortho-esters), can also be used to fabricate scaffolds for tissue engineering with controlled properties (Peppas and Langer 1994).
One of the limitations of applying cell-based regenerative medicine techniques to organ replacement has been the inherent difficulty of growing specific cell types in large quantities. Even when some organs, such as the liver, have a high regenerative capacity in vivo, cell growth and expansion in vitro may be difficult. By studying the privileged sites for committed precursor cells in specific organs, as well as exploring the conditions that promote differentiation, one may be able to overcome the obstacles that limit cell expansion in vitro. For example, urothelial cells could be grown in the laboratory setting in the past, but only with limited expansion. Several protocols were developed over the past two decades that identified the undifferentiated cells, and kept them undifferentiated during their growth phase (Cilento et al. 1994; Scriven et al. 1997; Liebert et al. 1991, 1997; Puthenveettil et al. 1999). Using these methods of cell culture, it is now possible to expand a urothelial strain from a single specimen that initially covered a surface area of 1 cm2 to one covering a surface area of 4202 m2 (the equivalent of one football field) within 8 weeks (Cilento et al. 1994). These studies indicated that it should be possible to collect autologous bladder cells from human patients, expand them in culture, and return them to the donor in sufficient quantities for reconstructive purposes (Cilento et al. 1994; Freeman et al. 1997; Liebert et al. 1991, 1997; Puthenveettil et al. 1999; Harriss 1995; Nguyen et al. 1999). Major advances have been achieved within the past decade on the possible expansion of a variety of primary human cells, with specific techniques that make the use of autologous cells possible for clinical application.
Alternate Cell Sources: Stem Cells, Nuclear Transfer, and Induced Pluripotent (iPS) Cells
Human embryonic stem cells exhibit two remarkable properties: the ability to proliferate in an undifferentiated but pluripotent state (self-renewal), and the ability to differentiate into many specialized cell types (Brivanlou et al. 2003). They can be isolated by aspirating the inner cell mass from the embryo during the blastocyst stage (5 days post-fertilization), and are usually grown on feeder layers consisting of mouse embryonic fibroblasts or human feeder cells (Richards et al. 2002). More recent reports have shown that these cells can be grown without the use of a feeder layer (Amit et al. 2004) and thus avoid the exposure of these human cells to mouse viruses and proteins. These cells have demonstrated longevity in culture by maintaining their undifferentiated state for at least 80 passages when grown using current published protocols (Thomson et al. 1998; Reubinoff et al. 2000).
Human embryonic stem cells have been shown to differentiate into cells from all three embryonic germ layers in vitro. Skin and neurons have been formed, indicating ectodermal differentiation (Schuldiner et al. 2000; Reubinoff et al. 2001; Schuldiner et al. 2001; Zhang et al. 2001). Blood, cardiac cells, cartilage, endothelial cells, and muscle have been formed, indicating mesodermal differentiation (Kaufman et al. 2001; Kehat et al. 2001; Levenberg et al. 2002). Pancreatic cells have been formed, indicating endodermal differentiation (Assady et al. 2001). In addition, as further evidence of their pluripotency, embryonic stem cells can form embryoid bodies, which are cell aggregations that contain all three embryonic germ layers while in culture, and can form teratomas in vivo (Itskovitz-Eldor et al. 2000).
An alternate source of stem cells is the amniotic fluid and placenta. Amniotic fluid and the placenta are known to contain multiple partially differentiated cell types derived from the developing fetus. We isolated stem cell populations from these sources, called amniotic fluid and placental stem cells (AFPSC) that express embryonic and adult stem cell markers (De Coppi et al. 2007). The undifferentiated stem cells expand extensively without feeders and double every 36 hours. Unlike human embryonic stem cells, the AFPSC appear not to form tumors in vivo. Lines maintained for over 250 population doublings retained long telomeres and a normal karyotype. AFS cells are broadly multipotent. Clonal human lines verified by retroviral marking can be induced to differentiate into cell types representing each embryonic germ layer, including cells of adipogenic, osteogenic, myogenic, endothelial, neuronal and hepatic lineages. In this respect, they meet a commonly accepted criterion for pluripotent stem cells, without implying that they can generate every adult tissue. Examples of differentiated cells derived from AFS cells and displaying specialized functions include neuronal lineage secreting the neurotransmitter l-glutamate or expressing G-protein-gated inwardly rectifying potassium (GIRK) channels, hepatic lineage cells producing urea, and osteogenic lineage cells forming tissue engineered bone. The cells could be obtained either from amniocentesis or chorionic villous sampling in the developing fetus, or from the placenta at the time of birth. The cells could be preserved for self use, and used without rejection, or they could be banked. A bank of 100,000 specimens could potentially supply 99 percent of the US population with a perfect genetic match for transplantation. Such a bank may be easier to create than with other cell sources, since there are approximately 4.5 million births per year in the USA (De Coppi et al. 2007).
In addition, stem cells for tissue engineering could be generated through cloning procedures. Frogs were the first successfully cloned vertebrates derived from nuclear transfer (Gurdon et al. 1958; Gurdon 1962a, b, c) but the nuclei were derived from non-adult sources. In the past fifteen years, tremendous advances in nuclear cloning technology have been reported, indicating the relative immaturity of the field. There has been remarkable interest in the field of nuclear cloning since the birth of Dolly in 1997. However, Dolly was not even the first cloned mammal to be produced via nuclear transfer; in fact, live lambs were produced in 1996 using nuclear transfer and differentiated epithelial cells derived from embryonic discs (Campbell et al. 1996). The significance of Dolly was that she was the first mammal to be derived from an adult somatic cell using nuclear transfer (Wilmut et al. 1997). Since then, animals from several species have been grown using nuclear transfer technology, including cattle (Cibelli et al. 1998), goats (Baguisi et al. 1999), mice (Wakayama et al. 1998), and pigs (Betthauser et al. 2000; De Sousa et al. 2002).
Two types of nuclear cloning, reproductive cloning and somatic cell nuclear transfer (so-called “therapeutic cloning”), have been described, and a better understanding of the differences between the two types may help to alleviate some of the controversy that surrounds these technologies (Colman and Kind 2000; Vogelstein et al. 2002). Banned in most countries for human applications, reproductive cloning is used to generate an embryo that has the identical genetic material as its cell source. This embryo can then be implanted into the uterus of a female to give rise to an infant that is a clone of the donor. On the other hand, somatic cell nuclear transfer technology is used to generate early stage embryos that are explanted in culture to produce embryonic stem cell lines whose genetic material is identical to that of its source. These autologous stem cells have the potential to become almost any type of cell in the adult body, and thus would be useful in tissue and organ replacement applications (Hochedlinger et al. 2004). Therefore, somatic cell nuclear transfer may provide an alternative source of transplantable cells. According to data from the Centers for Disease Control (CDC), an estimated 3000 Americans die every day of diseases that could have been treated with stem cell derived tissues (Lanza et al. 1999; Lanza et al. 2001) With current allogeneic tissue transplantation protocols, rejection is a frequent complication because of immunologic incompatibility, and immunosuppressive drugs are usually administered (Hochedlinger et al. 2004, pp. 114–117). The use of transplantable tissue and organs derived from somatic cell nuclear transfer may potentially lead to the avoidance of immune responses that typically are associated with transplantation of non-autologous tissues (Lanza et al. 1999).
While promising, somatic cell nuclear transfer technology has certain limitations that require further improvements before it can be applied widely in replacement therapy under the label of “therapeutic cloning”. Currently, the efficiency of the overall cloning process is low. The majority of embryos derived from animal cloning do not survive after implantation (Solter 2000; Rideout et al. 2001; Hochedlinger and Jaenisch 2002). To improve cloning efficiency, further improvements are required in the multiple complex steps of nuclear transfer, such as enucleation and reconstruction, activation of oocytes, and cell cycle synchronization between donor cells and recipient oocytes (Dinnyes et al. 2002).
Recently, exciting reports of the successful transformation of adult cells into pluripotent stem cells through a type of genetic “reprogramming” have been published. Reprogramming is a technique that involves de-differentiation of adult somatic cells to produce patient-specific pluripotent stem cells, without the use of embryos. Cells generated by reprogramming would be genetically identical to the somatic cells (and thus, the patient who donated these cells) and would not be rejected. Yamanaka was the first to discover that mouse embryonic fibroblasts (MEFs) and adult mouse fibroblasts could be reprogrammed into an “induced pluripotent state (iPS)” (Takahashi and Yamanaka 2006). This group used mouse embryonic fibroblasts (MEFs) engineered to express a neomycin resistance gene from the Fbx15 locus, a gene expressed only in ES cells. They examined 24 genes that were thought to be important for embryonic stem cells and identified four key genes that, when introduced into the reporter fibroblasts, resulted in drug-resistant cells. These were Oct3/4, Sox2, c-Myc, and Klf4. This experiment indicated that expression of the four genes in these transgenic MEFs led to expression of a gene specific for ES cells. Mouse embryonic fibroblasts and adult fibroblasts were co-transduced with retroviral vectors, each carrying one of the four genes, and transduced cells were selected via drug resistance. The resultant iPS cells possessed the immortal growth characteristics of self-renewing ES cells, expressed genes specific for ES cells, and generated embryoid bodies in vitro and teratomas in vivo. When iPS cells were injected into mouse blastocysts, they contributed to a variety of cell types. However, although iPS cells selected in this way were pluripotent, they were not identical to ES cells. Unlike ES cells, chimeras made from iPS cells did not result in full-term pregnancies. Gene expression profiles of the iPS cells showed that they possessed a distinct gene expression signature that was different from that of ES cells. In addition, the epigenetic state of the iPS cells was somewhere between that found in somatic cells and that found in ES cells, suggesting that the reprogramming was incomplete.
These results were improved significantly by Wernig and Jaenisch in July 2007 (Wernig et al. 2007). Fibroblasts were infected with retroviral vectors and selected for the activation of endogenous Oct4 or Nanog genes. Results from this study showed that DNA methylation, gene expression profiles, and the chromatin state of the reprogrammed cells were similar to those of ES cells. Teratomas induced by these cells contained differentiated cell types representing all three embryonic germ layers. Most importantly, the reprogrammed cells from this experiment were able to form viable chimeras and contribute to the germ line like ES cells, suggesting that these iPS cells were completely reprogrammed. Wernig et al. observed that the number of reprogrammed colonies increased when drug selection was initiated later (day 20 rather than day 3 post-transduction). This suggests that reprogramming is a slow and gradual process and may explain why previous attempts resulted in incomplete reprogramming.
It has recently been shown that reprogramming of human cells is possible (Takahashi et al. 2007; Yu et al. 2007). Yamanaka showed that retrovirus-mediated transfection of OCT3/4, SOX2, KLF4, and c-MYC generates human iPS cells that are similar to hES cells in terms of morphology, proliferation, gene expression, surface markers, and teratoma formation. Thompson’s group showed that retroviral transduction of OCT4, SOX2, NANOG, and LIN28 could generate pluripotent stem cells without introducing any oncogenes (c-MYC). Both studies showed that human iPS were similar but not identical to hES cells.
Another concern is that these iPS cells contain three to six retroviral integrations (one for each factor) which may increase the risk of tumorigenesis. Yamanaka et al. studied the tumor formation in chimeric mice generated from Nanog-iPS cells and found 20% of the offspring developed tumors due to the retroviral expression of c-myc (Okita et al. 2007). An alternative approach would be to use a transient expression method, such as adenovirus-mediated system, since both Jaenisch and Yamanaka showed strong silencing of the viral-controlled transcripts in iPS cells (Meissner et al. 2007; Okita et al. 2007). This indicates that these viral genes are only required for the induction, not the maintenance, of pluripotency.
A further concern is the use of transgenic donor cells for reprogrammed cells in the mouse studies. Both studies used donor cells from transgenic mice harboring a drug resistance gene driven by Fbx15, Oct3/4, or Nanog promoters so that, if these ES cell-specific genes were activated, the resulting cells could be easily selected using neomycin. However, the use of genetically modified donors hinders clinical applicability for humans. To assess whether iPS cells can be derived from nontransgenic donor cells, wild type MEF and adult skin cells were retrovirally transduced with Oct3/4, Sox2, c-Myc, and Klf4 and ES-like colonies were isolated by morphology alone, without the use of drug selection for Oct4 or Nanog (Meissner et al. 2007). IPS cells from wild type donor cells formed teratomas and generated live chimeras. This study suggests that transgenic donor cells are not necessary to generate iPS cells.
Although this is an exciting phenomenon, it is unclear why reprogramming adult fibroblasts and mesenchymal stromal cells have similar efficiencies (Takahashi and Yamanaka 2006). It would seem that cells that are already multipotent could be reprogrammed with greater efficiency, since the more undifferentiated donor nucleus the better SCNT performs (Blelloch et al. 2006). This further emphasizes our limited understanding of the mechanism of reprogramming, yet the potential for this area of study is exciting.
Tissue Engineering of Specific Structures
Investigators around the world, including our laboratory, have been working towards the development of several cell types and tissues and organs for clinical application.
Various biomaterials without cells, such as PGA and acellular collagen-based matrices from small intestine and bladder, have been used experimentally (in animal models) for the regeneration of urethral tissue (Atala et al. 1992; Chen et al. 1999, 2000; Olsen et al. 1992; Kropp et al. 1998; Sievert et al. 2000). Some of these biomaterials, like acellular collagen matrices derived from bladder submucosa, have also been seeded with autologous cells for urethral reconstruction. De Filippo and colleagues were able to replace tubularized urethral segments using tube-shaped collagen matrices seeded with cells (Atala 2002; De Filippo et al. 2002).
Unfortunately, the above techniques are not applicable for tubularized urethral repairs. The collagen matrices are able to replace urethral segments only when used in an onlay fashion. However, if a tubularized repair is needed, the collagen matrices should be seeded with autologous cells to avoid the risk of stricture formation and poor tissue development (De Filippo et al. 2002). Therefore, tubularized collagen matrices seeded with autologous cells can be used successfully for total penile urethra replacement.
Currently, gastrointestinal segments are commonly used as tissues for bladder replacement or repair. However, gastrointestinal tissues are designed to absorb specific solutes, whereas bladder tissue is designed for the excretion of solutes. Due to the problems encountered with the use of gastrointestinal segments, numerous investigators have attempted alternative materials and tissues for bladder replacement or repair.
The success of cell transplantation strategies for bladder reconstruction depends on the ability to use donor tissue efficiently and to provide the right conditions for long term survival, differentiation, and growth. Urothelial and muscle cells can be expanded in vitro, seeded onto polymer scaffolds, and allowed to attach and form sheets of cells (Atala et al. 1993a). These principles were applied in the creation of tissue engineered bladders in an animal model that required a subtotal cystectomy with subsequent replacement with a tissue engineered organ in beagle dogs (Oberpenning et al. 1999). Urothelial and muscle cells were separately expanded from an autologous bladder biopsy, and seeded onto a bladder-shaped biodegradable polymer scaffold. The results from this study showed that it is possible to tissue engineer bladders that are anatomically and functionally normal. Clinical trials for the application of this technology are currently being conducted.
Male and Female Reproductive Organs
Reconstructive surgery is required for a wide variety of pathologic penile conditions, such as penile carcinoma, trauma, severe erectile dysfunction, and congenital conditions such as ambiguous genitalia, hypospadias, and epispadias. One of the major limitations of phallic reconstructive surgery is the scarcity of sufficient autologous tissue.
The major components of the phallus are corporal smooth muscle and endothelial cells. The creation of autologous functional and structural corporal tissue de novo would be beneficial. Autologous cavernosal smooth muscle and endothelial cells were harvested, expanded, and seeded on acellular collagen matrices and implanted in a rabbit model (Kershen et al. 2002; Kwon et al. 2002). Histologic examination confirmed the appropriate organization of penile tissue phenotypes, and structural and functional studies, including cavernosography, cavernosometry, and mating studies, demonstrated that it is possible to engineer autologous functional penile tissue. Our laboratory is currently working on increasing the size of the engineered constructs.
Congenital malformations of the uterus may have profound implications clinically. Patients with cloacal exstrophy and intersex disorders may not have sufficient uterine tissue present for future reproduction. We investigated the possibility of engineering functional uterine tissue using autologous cells (Wang et al. 2003). Autologous rabbit uterine smooth muscle and epithelial cells were harvested, then grown and expanded in culture. These cells were seeded onto preconfigured uterine-shaped biodegradable polymer scaffolds, which were then used for subtotal uterine tissue replacement in the corresponding autologous animals. Upon retrieval 6 months after implantation, histological, immunocytochemical, and Western blot analyses confirmed the presence of normal uterine tissue components. Biomechanical analyses and organ bath studies showed that the functional characteristics of these tissues were similar to those of normal uterine tissue. Breeding studies using these engineered uteri are currently being performed.
Similarly, several pathologic conditions, including congenital malformations and malignancy, can adversely affect normal vaginal development or anatomy. Vaginal reconstruction has traditionally been challenging due to the paucity of available native tissue. The feasibility of engineering vaginal tissue in vivo was investigated (De Filippo et al. 2003). Vaginal epithelial and smooth muscle cells of female rabbits were harvested, grown, and expanded in culture. These cells were seeded onto biodegradable polymer scaffolds, and the cell-seeded constructs were then implanted into nude mice for up to 6 weeks. Immunocytochemical, histological, and Western blot analyses confirmed the presence of vaginal tissue phenotypes. Electrical field stimulation studies in the tissue-engineered constructs showed similar functional properties to those of normal vaginal tissue. When these constructs were used for autologous total vaginal replacement, patent vaginal structures were noted in the tissue-engineered specimens, while the non-cell-seeded structures were noted to be stenotic (De Filippo et al. 2003).
The kidney is a complex organ with multiple cell types and a complex functional anatomy that renders it one of the most difficult to reconstruct (Amiel and Atala 1999; Auchincloss and Bonventre 2002). Previous efforts in tissue engineering of the kidney have been directed toward the development of extracorporeal renal support systems made of biological and synthetic components (Aebischer et al. 1987a, b; lp et al. 1988; Lanza et al. 1996; MacKay et al. 1998; Humes et al. 1999a, b; Amiel et al. 2000; Joki et al. 2001) and ex vivo renal replacement devices are known to be life-sustaining. However, there would be obvious benefits for patients with end-stage kidney disease if these devices could be implanted long term without the need for an extracorporeal perfusion circuit or immunosuppressive drugs.
We applied the principles of both somatic cell nuclear transfer and tissue engineering in an effort to produce genetically identical renal tissue in a large animal model, the cow (Bos taurus) (Lanza et al. 2002). Bovine skin fibroblasts from adult Holstein steers were obtained by ear notch, and single donor cells were isolated and microinjected into the perivitelline space of donor enucleated oocytes. The resulting blastocysts were implanted into progestin-synchronized recipients to allow for further in vivo growth. After 12 weeks, cloned renal cells were harvested, expanded in vitro, then seeded onto biodegradable scaffolds. The constructs, which consisted of the cells and the scaffolds, were then implanted into the subcutaneous space of the same steer from which the cells were cloned to allow for tissue growth
Cloned renal cells were seeded on scaffolds consisting of three collagen-coated cylindrical silastic catheters. The ends of the three membranes of each scaffold were connected to catheters that terminated into a collecting reservoir. This created a renal neoorgan with a mechanism for collecting the excreted urinary fluid. These scaffolds with the collecting devices were transplanted subcutaneously into the same steer from which the genetic material originated, and then retrieved 12 weeks after implantation.
Chemical analysis of the collected urine-like fluid, including urea nitrogen and creatinine levels, electrolyte levels, specific gravity, and glucose concentration, revealed that the implanted renal cells possessed filtration, reabsorption, and secretory capabilities.
Histological examination of the retrieved implants revealed extensive vascularization and self-organization of the cells into glomeruli and tubule-like structures. A clear continuity between the glomeruli, the tubules, and the silastic catheter was noted that allowed the passage of urine into the collecting reservoir. Immunohistochemical analysis with renal-specific antibodies revealed the presence of renal proteins, RT-PCR analysis confirmed the transcription of renal specific RNA in the cloned specimens, and Western blot analysis confirmed the presence of elevated renal-specific protein levels.
Since previous studies have shown that bovine clones harbor the oocyte mitochondrial DNA (Evans et al. 1999; Hiendleder et al. 1999; Steinborn et al. 2000), the donor egg’s mitochondrial DNA (mtDNA) was thought to be a potential source of immunologic incompatibility. Differences in mtDNA-encoded proteins expressed by cloned cells could stimulate a T-cell response specific for mtDNA-encoded minor histocompatibility antigens when the cloned cells are implanted back into the original nuclear donor (Fischer Lindahl et al. 1991). We used nucleotide sequencing of the mtDNA genomes of the clone and fibroblast nuclear donor to identify potential antigens in the muscle constructs. Only two amino acid substitutions were noted to distinguish the clone and the nuclear donor and, as a result, a maximum of two minor histocompatibility antigens could be defined. Given the lack of knowledge regarding peptide-binding motifs for bovine MHC class I molecules, there is no reliable method to predict the impact of these amino acid substitutions on bovine histocompatibility.
Oocyte-derived mtDNA was also thought to be a potential source of immunologic incompatibility in the cloned renal cells. Maternally transmitted minor histocompatibility antigens in mice have been shown to stimulate both skin allograft rejection in vivo and cytotoxic T lymphocytes expansion in vitro (Fischer Lindahl et al. 1991) that could prevent the use of these cloned constructs in patients with chronic rejection of major histocompatibility matched human renal transplants (Hadley et al. 1992; Yard et al. 1993). We tested for a possible T-cell response to the cloned renal devices using delayed-type hypersensitivity testing in vivo and Elispot analysis of interferon-gamma secreting T-cells in vitro. Both analyses revealed that the cloned renal cells showed no evidence of a T-cell response, suggesting that rejection will not necessarily occur in the presence of oocyte-derived mtDNA. This finding may represent a step forward in overcoming the histocompatibility problem of stem cell therapy (Yard et al. 1993).
These studies were the first demonstration of the use of somatic cell nuclear transfer for regeneration of tissues in vivo. Cells derived from nuclear transfer could be successfully harvested, expanded in culture, and transplanted in vivo with the use of biodegradable scaffolds on which the single suspended cells could organize into tissue structures that are genetically identical to that of the host.
In the United States, over 5 million people currently live with some form of heart disease, and many more are diagnosed each year. While many medications have been developed to assist the ailing heart, the treatment for end-stage heart failure still remains transplantation. Unfortunately, as with other organs, donor hearts are in short supply, and even when a transplant can be performed, the patient must endure the side effects created by lifelong immunosuppression. Thus, alternatives are desperately needed, and the development of novel methods to regenerate or replace damaged heart muscle using tissue engineering and regenerative medicine techniques is an attractive option.
Cell therapy for infracted areas of the heart is attractive, as these methods involve a rather simple injection into the damaged area of a patient’s heart, rather than a rigorous surgical procedure, to complete. Various types of stem cells have been investigated for their potential to regenerate damaged or dead heart tissue in this manner. Skeletal muscle cells, bone marrow stem cells (both mesenchymal and hematopoietic), amniotic fluid stem cells, and embryonic stem cells have been used for this purpose. In this technique, cells are suspended in a biocompatible matrix that can range from simple normal saline to complex yet biocompatible hydrogels depending on the type of injection to be performed. The cells are either injected into the damaged area of the heart itself, or they are injected into the coronary circulation with the hope that they will home to the damaged area, take up residence there, and begin to repair the tissue. However, injectable therapies have been shown to be relatively inefficient, and cell loss is quite substantial. Newer methods of tissue engineering include the development of engineered “patches,” which are comprised of cells adhered to a biomaterial, that can theoretically be used to replace the damaged area of the heart. These techniques have promise, but require further research into the optimal cell types and biomaterials for this purpose before they can be used extensively in the clinic (see Jawad et al. 2007 for an excellent review of these methods).
However, the methods described above could only be used in cases where a relatively small section of heart muscle was damaged. In cases where a large area or even the whole heart has become nonfunctional, a more radical approach may be required. In these situations, the use of a bioartificial heart would be ideal, as rejection would be avoided and the problems associated with a mechanical heart (such as thromboembolus formation) would be abolished. To this end, Ott et. al., recently developed a novel heart construct in vitro using decellularized cadaveric hearts. By reseeding the tissue scaffold that remained after a specialized decellularization process with various types of cells that make up a heart (cardiomyocytes, smooth muscle cells, endothelial cells, and fibrocytes) and culturing the resulting construct in a bioreactor system designed to mimic physiologic conditions, this group was able to produce a construct that could generate pump function on its own (Ott et al. 2008). This study suggests that production of bioartificial hearts may one day be possible.
The liver can sustain a variety of insults, including viral infection, alcohol abuse, surgical resection of tumors, and acute drug-induced hepatic failure. The current therapy for liver failure is liver transplantation. However, this therapy is limited by the shortage of donors and the need for lifelong immunosuppressive therapy. Cell transplantation has been proposed as a potential solution for liver failure. This is based on the fact that the liver has enormous regenerative potential in vivo suggests that in the right environment, it may be possible to expand liver cells in vitro in sufficient quantities for tissue engineering (Bhandari et al. 2001). Many approaches have been tried, including development of specialized media, co-culture with other cell types, identification of growth factors that have proliferative effects on these cells, and culture on three-dimensional scaffolds within bioreactors (Bhandari et al. 2001).
Extracorporeal bioartificial liver devices that use porcine hepatocytes have been designed and applied. These devices are designed to filter and purify the patient’s blood as would the patient’s own liver, and the blood is returned to the patient in a manner similar to kidney dialysis. Another cell-based approach is the injection of liver cell suspensions. This has been performed in animal models. Intraportal hepatocyte injection has also been used in patients with Crigler-Najjar Syndrome Type 1 (Fox et al. 1998); however, complications such as portal vein thrombosis and pulmonary embolism are major concerns, especially when large cell numbers are used (Nieto et al. 1989). Finally, cells including stem cells, oval progenitor cells, and mature hepatocytes have been seeded onto liver shaped biocompatible matrices to engineer artificial, implantable livers. These have been tested in various animal models (Gilbert et al. 1993; Kaufmann et al. 1999), however the transplantation efficiency as well as the functionality of these constructs must be improved substantially before the technology can be moved into the clinic.
Articular Cartilage and Trachea
Full-thickness articular cartilage lesions have limited healing capacity and thus represent a difficult management issue for the clinicians who treat adult patients with damaged articular cartilage (Hunter 1995; O’Driscoll 1998). Large defects can be associated with mechanical instability and may lead to degenerative joint disease if left untreated (Buckwalter and Lohmander 1994; Buckwalter and Mankin 1998). Chondrocytes were expanded and cultured onto biodegradable scaffolds to create engineered cartilage for use in large osteochondral defects in rabbits (Schaefer et al. 2002). When sutured to a subchondral support, the engineered cartilage was able to withstand physiologic loading and underwent orderly remodeling of the large osteochondral defects in adult rabbits, providing a biomechanically functional template that was able to undergo orderly remodeling when subjected to quantitative structural and functional analyses.
Few treatment options are currently available for patients who suffer from severe congenital tracheal pathology, such as stenosis, atresia, and agenesis, due to the limited availability of autologous transplantable tissue in the neonatal period. Tissue engineering in the fetal period may be a viable alternative for the surgical treatment of these prenatally diagnosed congential anomalies, as cells could be harvested and grown into transplantable tissue in parallel with the remainder of gestation. Chondrocytes from both elastic and hyaline cartilage specimens have been harvested from fetal lambs, expanded in vitro, then dynamically seeded onto biodegradable scaffolds (Fuchs et al. 2002). The constructs were then implanted as replacement tracheal tissue in fetal lambs. The resultant tissue-engineered cartilage was noted to undergo engraftment and epithelialization, while maintaining its structural support and patency.
Recently, Martin Birchall’s group moved this technology into a human patient with end-stage airway disease (Macchiarini et al. 2008). This group was able to remove the cellular material and MHC antigens from a human donor trachea, and, using a specialized bioreactor, seed this acellular matrix with chondrocytes and epithelial cells derived from the patient to receive the graft. This construct was then used to replace the patient’s left main bronchus. There were no perioperative complications, and the left lung ventilated normally as soon as the graft was placed. At three months post-surgery, the patient’s lung function was in the normal range for her age and sex, and she was able to function normally. While longer follow-up and larger study populations are needed, this report indicates that tissue engineering may be a new option for patients with airway disease.
Injectable bulking agents can be endoscopically used in the treatment of both urinary incontinence and vesicoureteral reflux. The advantages in treating urinary incontinence and vesicoureteral reflux with this minimally invasive approach include the simplicity of this quick outpatient procedure and the low morbidity associated with it. Several investigators are seeking alternative implant materials that would be safe for human use (Kershen and Atala 1999).
The ideal substance for the endoscopic treatment of reflux and incontinence should be injectable, nonantigenic, nonmigratory, volume stable, and safe for human use. Toward this goal long term studies were conducted to determine the effect of injectable chondrocytes in vivo (Atala et al. 1993b). It was initially determined that alginate, a liquid solution of gluronic and mannuronic acid, embedded with chondrocytes, could serve as a synthetic substrate for the injectable delivery and maintenance of cartilage architecture in vivo. Alginate undergoes hydrolytic biodegradation and its degradation time can be varied depending on the concentration of each of the polysaccharides. The use of autologous cartilage for the treatment of vesicoureteral reflux in humans would satisfy all the requirements for an ideal injectable substance.
Chondrocytes derived from an ear biopsy can be readily grown and expanded in culture. Neocartilage formation can be achieved in vitro and in vivo using chondrocytes cultured on synthetic biodegradable polymers. In these experiments, the cartilage matrix replaced the alginate as the polysaccharide polymer underwent biodegradation. This system was adapted for the treatment of vesicoureteral reflux in a porcine model (Atala et al. 1994). These studies showed that chondrocytes can be easily harvested and combined with alginate in vitro, the suspension can be easily injected cystoscopically, and the elastic cartilage tissue formed is able to correct vesicoureteral reflux without any evidence of obstruction.
Two multicenter clinical trials were conducted using this engineered chondrocyte technology. Patients with vesicoureteral reflux were treated at ten centers throughout the US. The patients had a similar success rate as with other injectable substances in terms of cure. Chondrocyte formation was not noted in patients who had treatment failure. It is supposed that the patients who were cured have a biocompatible region of engineered autologous tissue present, rather than a foreign material (Diamond and Caldamone 1999). Patients with urinary incontinence were also treated endoscopically with injected chondrocytes at three different medical centers. Phase 1 trials showed an approximate success rate of 80% at follow-up 3 and 12 months postoperatively (Bent et al. 2001). Several of the clinical trials involving bioengineered products have been placed on hold because of the costs involved with the specific technology. With a bioengineered product, costs are usually high because of the biological nature of the therapies involved. As with any therapy, the cost that the medical health care system can allow for a specific technology is limited. Therefore, the costs of bioengineered products have to be lowered for them to have an impact clinically. This is currently being addressed for multiple tissue-engineered technologies. As the technologies advance over time, and the volume of the application is considered, costs will naturally decrease.
Injectable Muscle Cells
The potential use of injectable cultured myoblasts for the treatment of stress urinary incontinence has been investigated (Chancellor et al. 2000; Yokoyama et al. 2000). Myoblasts were labeled with fluorescent latex microspheres (FLM) in order to track them after injection. Labeled myoblasts were directly injected into the proximal urethra and lateral bladder walls of nude mice with a micro-syringe in an open surgical procedure. Tissue harvested up to 35 days post-injection contained the labeled myoblasts, as well as evidence of differentiation of the labeled myoblasts into regenerative myofibers. The authors reported that a significant portion of the injected myoblast population persisted in vivo. Similar techniques of sphincteric derived muscle cells have been used for the treatment of urinary incontinence in a pig model (Strasser et al. 2004). The fact that myoblasts can be labeled and survive after injection and begin the process of myogenic differentiation further supports the feasibility of using cultured cells of muscular origin as an injectable bioimplant.
The use of injectable muscle precursor cells has also been investigated for use in the treatment of urinary incontinence due to irreversible urethral sphincter injury or maldevelopment. Muscle precursor cells are the quiescent satellite cells found in each myofiber that proliferate to form myoblasts and eventually myotubes and new muscle tissue. Intrinsic muscle precursor cells have previously been shown to play an active role in the regeneration of injured striated urethral sphincter (Yiou et al. 2003a). In a subsequent study, autologous muscle precursor cells were injected into a rat model of urethral sphincter injury, and both replacement of mature myotubes as well as restoration of functional motor units was noted in the regenerating sphincteric muscle tissue (Yiou et al. 2003b). This is the first demonstration of the replacement of both sphincter muscle tissue and its innervation by the injection of muscle precursor cells. As a result, muscle precursor cells may be a minimally invasive solution for urinary incontinence in patients with irreversible urinary sphincter muscle insufficiency.
Patients with testicular dysfunction and hypogonadal disorders are dependent on androgen replacement therapy to restore and maintain physiological levels of serum testosterone and its metabolites, dihydrotestosterone and estradiol (Machluf et al. 2000, 2003). Currently available androgen replacement modalities, such as testosterone tablets and capsules, Depo-Provera injections, and skin patches may be associated with fluctuating serum levels and complications such as fluid and nitrogen retention, erythropoiesis, hypertension, and bone density changes (Santen and Swerdloff 1990). Since Leydig cells of the testes are the major source of testosterone in men, implantation of heterologous Leydig cells or gonadal tissue fragments have previously been proposed as a method for chronic testosterone replacement (Tai et al. 1989; van Dam et al. 1989). These approaches, however, were limited by the failure of the tissues and cells to produce testosterone.
Encapsulation of cells in biocompatible and semipermeable polymeric membranes has been an effective method to protect against a host immune response as well as to maintain viability of the cells while allowing the secretion of desired therapeutic agents (Tai and Sun 1993; De Vos et al. 1997). Alginate poly-l-lysine-encapsulated Leydig cell microspheres were used as a novel method for testosterone delivery in vivo (Machluf et al. 2003). Elevated stable serum testosterone levels were noted in castrated adult rats over the course of the study, suggesting that microencapsulated Leydig cells may be a potential therapeutic modality for testosterone supplementation.
In addition, Raya-Rivera et al. reported the creation of a novel, bioengineered testicular prosthesis created from cartilage in the laboratory (Raya-Rivera et al. 2008). To produce this prosthesis, testis-shaped polymer scaffolds were seeded with chondrocytes and maintained in a bioreactor for four weeks. At this time, milky white cartilage covered the entire scaffold, producing a solid testicular prosthesis with a hollow center. Testosterone enanthate was injected into this hollow center, and the filled prostheses were implanted into castrated athymic mice. Interestingly, the mice showed a burst in systemic testosterone levels on the first day post-implantation, and then maintained physiologic testosterone levels for the entire study period (40 weeks). These results indicate that it may be possible to engineer a biocompatible prosthesis in the laboratory that can also provide testosterone replacement.
Another area of intense study in regenerative medicine is the pancreas, since the ability to regenerate the insulin-producing cells this organ contains could lead to novel treatments or a cure for diabetes. In a series of exciting experiments, Zhou et al. recently demonstrated that regeneration of the insulin-producing cells of the pancreas, the β-cells, may be possible using cellular reprogramming techniques (Zhou et al. 2008). Using a mouse model, they showed that in vivo activation of a specific combination of three transcription factors (Ngn3, Pdx1, and Mafa) using adenoviral vectors led to the reprogramming of adult differentiated pancreatic exocrine cells into cells that closely resemble β-cells. These cells were similar to native β-cells in size, shape, and ultrastructure, and they expressed genes that are specific to β-cells as well. Interestingly, these cells secreted insulin and expressed vascular endothelial growth factor (VEGF), which allowed them to remodel local vasculature in a manner similar to native β-cells. In fact, these reprogrammed cells were able to partially amelioriate hyperglycemia in diabetic mice, suggesting that reprogramming techniques for treating disease may one day become a reality.
The engineering of large organs will require a vascular network of arteries, veins, and capillaries to deliver nutrients to each cell. One possible method of vascularization is through the use of gene delivery of angiogenic agents such as vascular endothelial growth factor (VEGF) with the implantation of vascular endothelial cells (EC) in order to enhance neovascularization of engineered tissues. Skeletal myoblasts from adult mice were cultured and transfected with an adenovirus encoding VEGF and combined with human vascular endothelial cells (Nomi et al. 2002). The mixtures of cells were injected subcutaneously in nude mice, and the engineered tissues were retrieved up to 8 weeks after implantation. The transfected cells were noted to form muscle with neovascularization by histology and immunohistochemical probing with maintenance of their muscle volume, while engineered muscle of nontransfected cells had a significantly smaller mass of cells with loss of muscle volume over time, less neovascularization, and no surviving endothelial cells. These results indicate that a combination of VEGF and endothelial cells may be useful for inducing neovascularization and volume preservation in engineered tissue.
The delivery of anti-angiogenic agents may help to slow tumor growth for a variety of neoplasms. Encapsulated hamster kidney cells transfected with the angiogenesis inhibitor endostatin were used for local delivery on human glioma cell line xenografts (Joki et al. 2001). The release of biologically active endostatin led to significant inhibition of endothelial cell proliferation and substantial reduction in tumor weight. Continuous local delivery of endostatin via encapsulated endostatin-secreting cells may be an effective therapeutic option for a variety of tumor types.
Summary and Conclusion
Regenerative medicine efforts are currently underway experimentally for virtually every type of tissue and organ within the human body. As regenerative medicine incorporates the fields of tissue engineering, cell biology, nuclear transfer, and materials science, personnel who have mastered the techniques of cell harvest, culture, expansion, transplantation, as well as polymer design are essential for the successful application of these technologies to extend human life. Various tissues are at different stages of development, with some already being used clinically, a few in preclinical trials, and some in the discovery stage. Recent progress suggests that engineered tissues may have an expanded clinical applicability in the future and may represent a viable therapeutic option for those who would benefit from the life-extending benefits of tissue replacement or repair.
The author wishes to thank Jennifer L. Olson, Ph.D. for editorial assistance with this manuscript.