Chapter 30: Regenerative Medicine: Multipotential Cell Therapy for Tissue Repair

Regenerative medicine is a concept that evolved from knowledge in genome regulation and modification, from understanding of embryonic development and “stemness” of cells, and from 50 years of experience in human transplant biology. Therefore, a narrow view of any of these disciplines is not sufficient for illuminating the mechanisms of action underlying the already accomplished successes and for guiding the potential of novel basic biology discoveries into clinically meaningful regenerative medicine (Fig. 30–1).

Accordingly, this chapter spans major organ systems (marrow, liver, pancreas, brain, and spinal cord) to demonstrate their connectivity and shared biologic responses deployed at the times of acute and chronic injury. Furthermore, the goals of regenerative therapies are different than those of commonly used drugs. Medications are typically aimed at amelioration of symptoms, while regenerative medicine seeks to either recruit the patient’s reparative cells or to replace the malfunctioning tissue altogether to restore the deficient organ to adequate function.

Regenerative medicine harnesses the body’s own repair mechanisms to replace, restore, or regenerate damaged or malfunctioning cells and tissues in conditions as diverse as diabetes, heart disease, spinal cord injury, and types of blindness. Some regenerative medicine therapies are already in use, for example using unrelated hematopoietic cell transplant to regenerate a patient’s immune system after their marrow has been destroyed by chemotherapy or radiation. There are some therapies that are in the early stages of clinical trials, for example using a patient’s cells seeded onto a biomesh scaffolding to grow a new trachea, ear, or nose. Some therapies are on the cusp of progressing into clinical trials, such as differentiating human embryonic stem cells into beta cells that could produce insulin in diabetic patients. And some therapies, such as growing new lungs from patient cells and repairing a spinal cord injury with a cellular bridge, remain tantalizingly out of reach.

The zygote has the ability to give rise to a complete organism. Any cellular genome in the organism has the ability to code for any protein in the body. Although we know this, the mechanisms underlying the ability of a progenitor cell to differentiate have been challenging to elucidate. For the earliest critical steps in this long and complex process, we must look at developmental biology. Nuclear transfers in amphibians done by Briggs, King, and Gurdon^1,2,3 established that bidirectionality of cellular fate determination is possible. It was, established by McGrath and Solter that this process is driven by a multitude of factors of such temporal and spatial complexity that it would make reprogramming of mammalian cells by nuclear transfer impossible.^4,5 Recent experimentation has focused on editing the genome itself, finding success in both mouse and human DNA models. Much work remains to bring this technology into human therapies, but in the foreseeable future, cells and organisms will no longer be seen as being given sealed orders at birth, but rather the instructions contained in their developmental program can be thought of as “software” that can be rewritten and used to reprogram the genomic “hardware” of a cell.

It has been even more difficult to imagine and later define the possibility that a differentiated cell could be instructed to revert to an immature state and undergo a respecification to another differentiated cellular phenotype or an asymmetrical division to generate more immature cells. Yamanaka’s experimental proof reduced this perceived complexity to a four-factor recipe sufficient to restore skin fibroblasts to induced pluripotent stem cells (iPSCs).^6,7,8

This chapter gives a broad overview of the state of regenerative medicine as it stands today, a complex and rapidly advancing field that holds tremendous promise in treating, and even curing, many of the disorders that cause pain and suffering.

EMBRYONIC STEM CELLS

Four to 5 days after fertilization, an egg becomes a blastocyst, a ball of approximately 100 to 150 cells. A small group of inner cells within the blastocyst are pluripotent and have the potential to replicate indefinitely and to become any of the differentiated types of tissue in the body. These pluripotent cells are called embryonic stem cells (ESCs). ESCs have the dual ability to self-renew (copy themselves) and differentiate (produce more specialized types of cells of the body).

ESCs were first isolated from mice when embryonic carcinoma (EC) cells⁹ were shown to proliferate indefinitely like stem cells and were used to generate a chimeric mouse. Further development of the culture conditions for EC cells that used feeder (supporting) cells, along with discovery of cell-surface antigens, like SSEA-1 and F-9 antigen on EC cells, led to the isolation of the first ESCs from a mouse embryo in 1981.¹⁰ In 1995, Thomson isolated ESC lines from a nonhuman primate,¹¹ followed by the first successful isolation of human embryonic stem cell (hESC) lines in 1998.¹² The use of human ESCs in research, however, has been severely limited because of the social and religious concerns that the blastocyst is destroyed when the ESC lines are generated.

hESC lines can be cultured on feeder cells where they divide infinitely. They can also be grown without feeder cells, where they develop into clusters known as embryoid bodies. Using cells from a human blastocyst in clinical therapy has been difficult, so the 2006 discovery by Yamanaka and Takahashi that pluripotent stem cells could be created from skin cells⁶ revolutionized the stem cell research. Their pioneering work showed that inducing skin cells with four genes (Oct4, SOX2, Klf4, and c-Myc) would generate pluripotent embryonic stem-like cells in vitro. These cells, known as iPSCs, have the basic properties of hESCs while being derived from somatic cells rather than blastocysts.

An alternative approach for obtaining pluripotent stem cells is somatic cell nuclear transfer (SCNT).¹³ In this process, the nucleus of an adult human cell is placed in an egg cell that has had its nucleus removed. As this cell divides it can be a source of pluripotent stem cells. A recent study verified that the SCNT pluripotent stem cells are more similar to ESCs than are iPSCs.¹⁴ Although much research remains to be done, SCNT pluripotent stem cells appear to have potential in regenerative medicine.

There are many disorders and defects that arise from errors in the complex process of embryonic development. Research over the past decade advanced our understanding of the critical steps in the embryonic development of mice; however, information about the embryonic development of humans remains limited. Although there is overlap with what has been learned from studying mouse embryos, human embryonic growth is different and unique. However, by growing hESCs in the laboratory the multitude of regulatory factors that control the different stages of cell, tissue, and organ system development can be studied and can also provide insights into how our adult tissues are maintained and repaired and allows identification of the causes of birth defects by discovering what interferes with the normal path of cell fate acquisition. The hESCs are also used to produce laboratory disease models in specialized cells like nerve, heart, or beta cells and can also be used for development of new drug therapies.

Despite the much-anticipated potential of hESCs to differentiate and replace malfunctioning cells in the body, progress toward clinical use has been hindered by the possibility of teratoma formation or the immune rejection of the allogeneic transplanted cells and production issues.

INDUCED PLURIPOTENT STEM CELLS

Generation of iPSCs has connected several previous observations into a coherent outline. For example, the ability of transcription factor MyoD to change fibroblasts to myoblasts¹⁵ and of transcription factor Antennapedia to change development of antennae into legs in Drosophilla,¹⁶ uncovered potential of a differentiated cell to assume an alternative cell fate as a result of defined, externally provided signals.

The understanding of induced pluripotency has become more refined as additional reprogramming factors are identified,¹⁷ the critical role of epigenetic regulation is uncovered,¹⁸ and with the dynamics of iPSC generation (from initially random event to deterministic process) more fully developed.¹⁹

The reprogramming technology applied to human cells iPSCs allows for modeling various, typically genetic, disorders.^20,21 Furthermore, organoid cultures derived from the patients themselves allow for high throughput drug testing that would be impossible without the supply of differentiated cells from patient-specific iPSCs.

The first preclinical example of iPSC technology conceptually applied to human disease has been amelioration of the sickle cell anemia phenotype in a murine model.²² At the time of this writing, the first iPSC-based clinical trial opened in Japan for individuals with exudative age-related macular degeneration.²³

New knowledge derived from the rapidly expanding iPSC field has also reenergized the technology of direct reprogramming, whereby one differentiated cellular phenotype (such as a dermal fibroblast) can be induced to convert into another somatic cell (such as a neuron) without the intermediate iPSC stage. In contrast to the expandable iPSC-based generation of differentiated cells, the process of direct reprogramming makes it more challenging to produce the large numbers of cells needed for therapeutic intervention.

An example of this strategy has been in vivo trans-differentiation of exocrine pancreatic cells or biliary epithelial cells into insulin-producing endocrine cells in rodent models.^24,25 A conceptually different concept to solve the same clinical challenge has been blastocyst complementation whereby rat iPSCs were injected into blastocysts that had been derived from mice deficient in pancreatic organogenesis, which resulted in the development of a functional rat pancreas in mice.²⁶ In addition to reducing the cell numbers needed to create a physiologically meaningful effect, the efficacy of both strategies may be enhanced by targeting them into a permissive cellular niche.

HEMATOPOIETIC STEM CELLS

The earliest advances with clinical potential will most likely arise from understanding reprogramming in hematopoietic stem cells. Not only was hematopoietic cell transplantation the first stem cell therapy, developed close to half a century ago, but reports of using defined factors to turn committed blood progenitor cells into transplantable hematopoietic cells^27,28 suggest that robust generation of clinical-grade, patient-specific autologous grafts for transplantation is possible.

Equally important has been combination of pluripotentiality of ESCs and iPSCs with their commitment to specific lineages, such as hemogenic differentiation program. Derivation of hematopoietic stem cells from murine ESC and their genetic correction has been used in murine severe combined immune deficiency²⁹; similarly in a model of sickle cell anemia, hematopoietic stem cells derived from gene-corrected murine iPSCs²² have established preclinical proof-of-concept for combined gene correction and stem cell engineering. Furthermore, insights from murine embryogenesis were applied to in vitro induction of mesoderm and ESC differentiation to blood cells via coculture with feeder cells or generation of embryoid bodies. These seemingly straightforward concepts, however, have proven challenging to mimic in human ESCs and iPSCs. Despite many attempts, current technology appears to lead only to low hematopoietic chimerism after transplantation of hematopoietic stem cells derived from pluripotent human cells.^30,31 An alternative to generation of transplantable human hematopoietic stem cells is direct conversion of fibroblasts to hematopoietic stem cells without the iPSC intermediate. This is done by using forced expression of OCT4³² and differentiation of human pluripotent progenitor cells by forced expression of GATA-1, ETV2, and TAL-1 into hemoendothelial cells.³³

The replacement of hematopoiesis by marrow transplantation is the prototype of regenerative medicine. While the initial experimentation with marrow transfers on both sides of the Atlantic was almost immediately recognized as a pioneering effort in hematology, it was only later understood as a turning point in the larger field of regenerative medicine. The critical evidence was the ability of a relatively small number of donor cells to repopulate the host and reconstitute its full lymphohematopoietic system. Although initially applied to leukemia and lymphoma therapy in an effort to replace the malignant lymphohematopoiesis with a healthy wild-type system, it later became clear that the immune elimination of the tumor (graft-versus-leukemia, graft-versus-lymphoma) is the dominant mechanism behind successful therapy in many cases.

This remarkable regenerative capacity of hematopoietic stem cells established marrow, and later cord blood, transplantation as the blueprint for other stem cell therapies.

MESENCHYMAL STROMAL CELLS

Originally defined by how they were identified—they adhered to the surface of a culture dish³⁴—marrow-derived mesenchymal stromal/stem cells (MSCs) were then identified as key support cells in the cellular niche. Evidence from different sources (e.g., marrow, umbilical cord blood, and adipose tissue) suggests that MSCs have different functions in various organs (e.g., as pericytes in adventitia of blood vessels, or as supporting cells in marrow periosteal and endovascular hematopoietic niches).^35,36,37 In addition to this developmental heterogeneity, cultured MSCs display various levels of “stemness,” and the cellular products used in therapeutic applications may be more a cell culture artifact than a counterpart to physiologic functionally integrated MSCs. This is not necessarily a disadvantage, as the culture process enables both amplification of cell numbers and defined release criteria for clinical use.

The most striking application of MSCs in medicine to date relies not on the regenerative capacity of MSCs alone but on the immunosuppressive potential of MSC cultures in the setting of severe graft-versus-host disease (GVHD),^38,39,40 a serious complication of allogeneic hematopoietic cell transplantation (HCT).⁴¹ The treatment options for individuals with glucocorticoid-resistant severe GVHD have been inadequate, and mortality in this subgroup remains high. In a paradigm-changing study, it was demonstrated that culture-expanded MSCs can ameliorate severe GVHD.^38,42 The induction and maintenance of MSC-driven regulation of immune and inflammatory reactions⁴³ has made it possible to assess their role in autoimmune and inflammatory disorders such as Crohn disease, arthritis, diabetes, organ rejection, and bridge therapy before solid-organ transplantation.^44,45,46

The regenerative potential of MSCs has long been sought as a tool to rebuild and replace tissues damaged by acute or chronic injury whereby injected cultured MSCs activate endogenous repair mechanisms and disappear in the process.⁴⁷ In this capacity, MSCs in preclinical models have been shown to alleviate ischemic injury in the heart, brain, and kidneys; toxic insults to lung and liver; and degenerative damage to joints; and it is possible that in some settings allogeneic MSCs may be even more effective than autologous MSCs.⁴⁸

Lastly, MSCs are relatively easy to gene-modify and thus can be used as cellular vectors for delivering gene therapeutic agents in both inborn genetic disorders and acquired conditions, such as in anticancer therapy or for trophic factor support after surgery.

CARDIAC REPAIR

Cardiovascular disease is a leading cause of death in the world, leading to an estimated 17 million deaths per year.⁴⁹ As life expectancy in the developed world rises, so too do risk factors associated with chronic heart disease. It is estimated that there are approximately 800,000 new cases of acute myocardial infarction (AMI)⁵⁰ annually. Heart failure occurs when there is significant deprivation of oxygen to cardiac tissue, which results in decreased cardiac output and function as a result of loss of cardiomyocytes, scar formation, and tissue remodeling.⁵⁰ Identifying ways to regenerate or repair heart tissue will be key to developing effective treatment options for heart failure.

Since the mid-1990s, scientists have been investigating the potential of adult progenitor cells for use in heart regeneration. These early studies were triggered by the discovery that certain adult tissue-specific stem cells could be differentiated in vitro to become cardiac-like cells.⁵¹ This discovery led to many preclinical studies that assessed the ability of adult stem cells to repair or enhance cardiac function after various types of injury.

The cells reported to differentiate in vitro to cardiac-like cells in vivo are satellite cells, which are undifferentiated skeletal muscle myoblasts,⁵² which led to studies using autologous skeletal myoblasts surgically implanted into the heart muscle.⁵³ Although these cells survived for short periods of time, they retained their intrinsic contractile properties and did not fully integrate into the cardiac tissue,⁵⁴ which led to arrhythmias and gave little long-term significant benefit in overall heart function.

Some marrow-derived cell populations (lin−; c-kit+) were capable of differentiating to myocytes expressing cardiomyocyte markers such as Nkx2.5, Gata4, and MEF.⁵⁵ These marrow-derived cells were shown to survive in infarcted hearts and were capable of differentiating into smooth muscle and endothelial cells but not cardiomyocytes in vivo.⁵⁶ Additional studies demonstrated that other marrow-derived progenitor cell populations (endothelial progenitors, angioblasts, or CD34+ cells) were able to contribute to angiogenesis and neovascularization of the infarcted myocardium.⁵⁷ This differed from more immature marrow progenitor populations, called side-population cells (Lin− c-kit+ Sca-1+), that not only contribute to neovascularization but also regenerate myocardium.⁵⁸ These marrow-derived side-population cells homed to the border zone of infarction and resulted in improved left ventricular function.⁵⁸ The cardiac regeneration capacity of the marrow-derived progenitor cells facilitated large-scale clinical studies using heterogeneous populations of bone mononuclear cells (MNCs), also called epithelial progenitor cells (EPCs), for cardiac repair in patients with AMI^59,60 or ischemic cardiomyopathy.⁶¹ These studies showed only moderate improvements, and therefore led to further refinement of the selection criteria for marrow-derived cells (CD34+/CD133+) and changes in the route of administration (intracoronary injection) in the Regeneration by Intracoronary Infusion of Selected Population of Stem Cells in Acute Myocardial Infarction (REGENT) study, which still only showed modest success.⁶²

The moderate success of marrow-derived MNCs spurred the investigation of other populations of adult progenitor cells, as well as new routes of administration. The investigations led to the discovery of MSCs, as well as endogenous cardiac-derived stem cells, that could differentiate into cardiomyocytes and endothelial cells in animal models.⁶³ These findings led to clinical studies comparing marrow-derived MNCs versus the new MSCs used for intracoronary injection into patients with ischemic cardiomyopathy.⁶⁴ These tests showed significant improvement in left ventricle ejection fraction in response to MSC treatment.

To date, the clinical success of cell therapy approaches for cardiac regeneration has been mixed. This is in contrast to the promising early preclinical studies that showed significant improvement in many different measures of cardiac function. This difference has been attributed to the differences between rodent cardiac injury models and human clinical pathology, the cell population administration route, the origin of the cell populations, and the limited number of cells injected.

ESCs, because of their pluripotency and unlimited ability to proliferate, have been the subject of extensive preclinical investigation for many tissues, particularly cardiac tissue repair.^65,66 However, there has been less enthusiasm for hESC-derived cardiomyocytes for human cell therapy because their allogeneic nature requires concomitant immunosuppressive therapy and because ethical issues surround their derivation. Despite these challenges, clinical studies have begun to collect ESCs for cardiac differentiation with the intent of being used in a trial for AMI patients. However, as about any somatic cell can be used to generate embryonic stem-like iPSCs,⁶⁷ with the capability to differentiate into cardiomyocytes, endothelial cells, and smooth muscle cells.⁶⁸ Initial preclinical studies of murine iPSC-derived cardiomyocytes, injected into ischemic myocardium, led to rejection of transplanted cells by immune reaction as well as continuous proliferation that led to teratoma formation.⁶⁹ Although iPSCs are attractive for their allogeneic potential they have potential disadvantages for human cell therapy based on their oncogenic nature, epigenetic memory, and maintenance of potency for other cells types.

Growing tissues in vitro for use in regenerative therapies has been investigated as another delivery method of cells for heart repair. This tissue engineering approach involves seeding cells onto scaffolds and growing them for later engraftment or for the generation of whole organs.⁷⁰ In these systems, cells are transplanted with the scaffolding to the cardiac wall, which provides structural support and a better microenvironment for the migration of cells into the damaged myocardium.⁷¹

A common theme in most preclinical and clinical cell therapy-based studies is the demonstration of improvements in cardiac function that is not correlated to number of cells injected or their longevity after administration. This observation has led investigators to speculate that transplanted cells improve cardiac function through paracrine rather than structural effects. This model suggests that observed improvements in myocardial regeneration or vasculogenesis are a result of transplanted cells secreting molecules that are known to improve cardiovascular function after injury.⁷² The effects of paracrine factors include decreased inflammation, increased angiogenesis, induction of proliferation of cardiomyocytes or protection of existing ones, and activation of endogenous stem cells.⁷³ Ischemic hearts subjected to secreted factors showed an increase in the expression of genes involved in cardiogenesis and a downregulation of cell-death markers, effects that contribute to survival of ischemic cardiomyocytes.⁷⁴ The advantage of the paracrine model is the potential for the commercial development of paracrine factors that have proven potential for cardiac repair.

LUNG REPAIR

The prevalence of lung diseases, like the chronic obstructive pulmonary diseases (COPDs) that include asthma and emphysema, has increased dramatically over the last 50 years. Lung disease is expected to become the third leading cause of disease-related death in the world by 2020. New therapeutic approaches from regenerative medicine are being developed ranging from stem cell therapies to bioengineering of entire tissues of the respiratory system for transplantation. These approaches are based on initial observations that endothelial progenitor cells and mesenchymal stem cells can differentiate in vitro to cells expressing lung epithelial markers and contribute to mature functional bioengineered tissues.

Throughout the pulmonary tract there exist many different niche environments containing distinct epithelial cell types that contribute to the complexity of the lung. Identification of a true endogenous stem cell population that is responsible for maintaining lung tissue under steady state and injury has been challenging and has been a source of controversy.⁷⁵ Evidence from rodent models and human lungs suggests that the adult endogenous airway, alveolar epithelial cells, lung stroma, and pulmonary vasculature all contain putative stem cell populations that can repair damaged tissue.^76,77 These studies suggest the lung has a regional hierarchy of stem and progenitor cells that are specific for proximal versus distal airways as well as alveoli.

Identifying endogenous lung stem cells is complex because many different subpopulations of basal epithelial cells exhibit restricted patterns or roles in self-renewal for steady-state maintenance or after injury.^78,79 In the distal airway, putative progenitor cells have been identified in the neuroepithelial body,⁸⁰ bronchoalveolar duct junction,⁸¹ by specific markers of self-renewing lung epithelial cells,^82,83 and by function as bronchiolar alveolar stem cells (BASCs). This is in contrast to alveolar epithelial repair thought to be regulated by type 2 alveolar epithelial cells (ATII) because they have been shown to be precursors to type 1 (ATI) cells.^84,85 Identification of regionally specific stem cell populations is further complicated by the demonstration that isolated distal airway progenitors (BASCs, CK5+/p63+) can differentiate into ATII and ATI cells.^86,87 Regardless, all of these cells show unique functions in repair after injury, reside in different locations in the distal airway and alveolar epithelium, and play different roles as endogenous lung epithelial progenitors.

Many preclinical studies have shown that EPCs can increase function in pulmonary lung injury models.^88,89,90 This improvement in function could be because of contributions to structure, paracrine effects, modulation of immune responses, or a combination of these.⁹⁰ EPCs have also been demonstrated to preferentially home to sites of injury in the lung after systemic administration⁹¹; consequently, autologous EPCs have been used clinically in pulmonary hypertension patients and showed improved cardiopulmonary outcomes.^92,93

Marrow MSCs are known for their immunomodulatory effects in a wide range of diseases.^94,95 The beneficial effect of MSCs results from secretion of soluble mediators and microsomal particles that influence lung progenitor cells directly or indirectly through mediation of inflammatory cells that subsequently promote repair.^96,97 Both preclinical and clinical studies have shown efficacy in either systemic or intratracheal administration of MSCs in acute lung injury models, asthma, COPD, and a host of other inflammation-related lung injuries or diseases.^75,98 Although different studies have shown varying degrees of efficacy, there are still significant gaps in our understanding of the mechanisms of MSC action on ameliorating disease symptoms and of the specific subtype of MSCs used. This is important, as studies have demonstrated that certain MSCs can have negative effects in some lung disease models, such as pulmonary fibrosis.^99,100

Although there has been significant preclinical investigation into using EPC and MSC therapy approaches to lung repair, clinical studies have been slow to develop. However, there are a growing number of clinical trials in development focused on using MSCs for chronic lung diseases where preclinical data show the most promise. The most recent is the PROCHYMAL phase II trial looking at systemic administration of marrow MSCs for moderate to severe COPD,¹⁰¹ which showed the safety of using MSCs and also preliminary evidence for decrease in markers of inflammation.

BRAIN AND SPINAL CORD REPAIR

Stem cell-based therapy is rapidly developing as a way to improve outcomes following brain and spinal cord injury or disease. The human CNS is composed of more than 100 billion nerve cells connected in a complex network that must work seamlessly throughout our lives. Conditions affecting the CNS—such as stroke, brain, spinal cord injury, and neurodegenerative diseases—affect millions of people worldwide. Challenges for therapeutic intervention include the complex pathology of these conditions, as well as the specialized anatomical structures of the CNS that prevent easy access from systemic administration of therapies (e.g., the blood–brain barrier).

As with other areas of regenerative medicine, much effort has been spent on the identification of cell types with the best potential for CNS repair. Although it was initially thought that the adult CNS did not contain progenitor cells for repair, it is now recognized that the human CNS does retain an endogenous neural stem cell (NSC) population that retains some capacity for repair, although only in select regions and of a limited nature. Isolated adult and fetal NSCs can be expanded and differentiated into neurons, astrocytes, and oligodendrocytes, the three main CNS cell types. Adult NSCs are retained throughout life, and are found in the striatal subventricular zone and the dentate gyrus of the hippocampus. In preclinical studies, endogenous NSCs have shown to provide the most significant improvement in functional recovery in rodent stroke models.¹⁰² Isolated human fetal NSCs (CD133+),¹⁰³ are currently being investigated in a number of clinical studies.^104,105 In addition, multiple NSC-based cell therapy trials are being conducted to determine the safety and efficacy in patients with amyotrophic lateral sclerosis (ALS, sometimes called Lou Gehrig disease).^106,107 Although all trials to date have confirmed the safety of using these cells, any benefit from them has yet to be reported.

Although other adult stem cell types (endothelial progenitor cells, umbilical cord blood cells [UBCs], dental pulp stem cells [DPSCs]) have been investigated preclinically, only MSCs have been shown to have the same level of efficacy as NSCs. Marrow-derived MSCs have been the primary focus of preclinical and clinical studies because of their relative abundance and potential for autologous cell transplantation.^108,109 Administration of MSCs, regardless of route, have been shown to improve outcome measures in rodent models of injury and disease.^106,110 Based on these findings, there have been a number of early phase clinical trials initiated to study the effects of MSC transplantation following CNS injury^111,112 or disease.^113,114 In both cases, MSCs have been shown to be both safe and a feasible approach. Although not designed to test efficacy, many trials have observed improvements in functional outcomes.¹¹⁵

From studies of the efficacy of adult stem cells for therapy in preclinical models of CNS injury or disease over the past 20 years, there is strong evidence that transplanted adult stem cells can migrate to the site of injury and promote functional improvement. The mechanism of action, however, remains controversial.¹¹⁶ There is speculation that the benefits of cell-based therapy arise from multiple factors. From the host of preclinical studies on MSCs for treatment of stroke, improvements in function have been attributed to increased angiogenesis, neurogenesis, prosurvival signals, and mitigation of immune responses.^117,118 These mechanisms are potentially mediated by various soluble factors that act through a paracrine mechanism secreted by transplanted stem cells that benefit the local environment. Many of these secreted factors have begun to be identified and are well-known mediators of neurogenesis (bone-derived neurotrophic factor [BDNF]), angiogenesis (vascular endothelial growth factor [VEGF]), and immune regulation (transforming growth factor-β₁[TGF-β₁]).^117,119

The use of pluripotent cell types (ESCs, iPSCs) holds significant potential as a therapeutic approach for CNS repair. Clinical application of ESCs/iPSCs is limited because of the inability to isolate pure differentiated populations of neuronal cell types.^120,121 Despite this limitation, considerable progress has been made in preclinical studies for remyelination following spinal cord injury.^122,123 These studies and others have demonstrated that ESCs/iPSCs can be differentiated to oligodendrocyte progenitor cells (OPCs), migrate within the spinal cord, and produce myelin. Again, the efficacy and mechanism of recovery remain controversial.^{124,125,126,127} In 2010, the Geron Company began recruitment of patients for a phase I clinical trial for treatment of spinal cord injury with ESC-derived OPCs. Despite significant enthusiasm from the patient population, and report of no adverse events on a small cohort of treated patients (n = 4), significant methodologic¹²⁸ and economic obstacles forced the early discontinuation of the trial.

LIVER AND PANCREAS REPAIR

The liver is an essential organ that coordinates glycogen storage, drug detoxification, production of various serum proteins, and secretion of bile, which plays a critical role in food digestion and metabolism. It is interspersed with small microscopic canals known as canaliculi through which the bile drains to the gall bladder. The numerous bile canaliculi join together into many larger bile ducts, which join to become a branched structure that forms the common hepatic duct. The part of the common hepatic duct that is outside the liver is called the extra hepatic bile duct, which joins the cystic duct from the gall bladder to form the common bile duct and connect to the exocrine pancreas. This whole branched structure forms the biliary tree.¹²⁹ Liver, biliary tree, and pancreas originate from the anterior definitive endoderm and have been found to share stem cell populations during the early stages of development.¹³⁰

The normal liver has extraordinary potential to regenerate following partial removal of the liver or liver injury. The hepatocytes and cholangiocytes (biliary epithelial cells) are normally quiescent, but in response to liver injury, these cells proliferate and contribute to regeneration.¹³¹ Average liver turnover is maintained by differentiation and proliferation of parenchymal or nonparenchymal cells. However, in chronic liver disease, when the liver cannot be repaired by self-duplication of existing hepatocytes, small bipotential progenitor cells are activated that have the ability to differentiate into either hepatocytes or cholangiocytes.¹³² Severe injury in alcoholic liver disease has been found to stimulate increased proliferation of hepatic stem/progenitor cells in humans¹³³ and oval cells in rodents.¹³⁴ Some experiments have suggested that stem progenitor cells are not only activated in the injured liver, but that the normal adult human liver contains a large number of liver progenitor cells that may contribute to liver homeostasis.

Liver transplantation is currently the only option in acute liver failure. Two small clinical trials conducted with hepatocyte transplantation resulted in limited restored enzyme function in one patient, but it was not enough for survival and the patient eventually needed a liver transplant.¹³⁵ It remains unknown if hepatocytes can contribute to long-term rescue in patients. Additionally, hepatocytes are difficult to produce in clinically sufficient numbers, as they lose their viability and function when cultured in vitro.

Transplanted human ESC/iPSC-derived mature hepatocytes can express liver-specific enzymes such as albumin, antitrypsin, and cytochrome P450, which can improve liver function.^136,137 Furthermore, a functional liver organ bud generated from iPSCs was found to be engrafted and integrated within the host organism, even including development of blood vessels.¹³⁸ When iPSCs are injected into the blastocysts of fumarylacetoacetate hydrolase (FAH)-deficient mice, iPSC-derived hepatocytes can repopulate the damaged liver efficiently and restore liver function.¹³⁹

The liver and pancreas both harbor a niche of stem cells, collectively known as the hepatic stem/progenitor cells.^140,141 The fetal biliary tree, arising from the ductal plate cells during development and consisting of the intrahepatic and extrahepatic ducts, has been shown to harbor a rich source of stem/progenitor cells.¹⁴² These stem/progenitor cells are quite distinctive from hepatoblasts, which contribute toward the hepatocytes or cholangiocytes during development.¹⁴³ These stem/progenitor cells present in the biliary tree differ from hepatoblasts by their ability to self-renew and differentiate into hepatocytes, cholangiocytes, or pancreatic islets, depending on the microenvironment. Sox9 expression is detected throughout the pancreatic ducts, intra- and extrahepatic ducts, and in the intestinal crypt connected through the major duodenal papilla, forming a contiguous Sox9+ zone.^144,145 Remarkably, when an adenovirus contacting three pancreatic transcription factors (Pdx1, Ngn3, and MafA) was delivered into the liver, it was able to reprogram the Sox9+ population of cells within the bile ducts into functional insulin-secreting, beta-like cells.²⁵ When both Ad-PNM and a peroxisome proliferator-activated receptor (PPAR) agonist, WY14643, were given in animals, it resulted in injury to the liver, led to an increase in the cell division rate of Sox9+ cells lining bile ducts, and contributed toward making more insulin-positive beta cells.¹⁴⁶ In an injured liver, the Wnt signaling pathway is activated, which stimulates discreet subsets of progenitor cell populations, which, in turn, engage in liver regeneration.¹⁴⁷ Isolation of stem/progenitor cell populations based on their surface markers like Lgr5 or EpCAM has identified the possibility of differentiating them toward a hepatocyte- or beta-like cell fate.^145,148

The pancreas plays an important role in digestion, as well as in maintaining blood glucose homeostasis. It is composed of ducts, with the main duct (pancreatic duct) running the length of the pancreas. The pancreatic duct merges with the bile duct to form the major duodenal ampulla, which drains the pancreatic fluid into the first portion of the small intestine, the duodenum. The pancreas is composed of exocrine and endocrine parts. The exocrine component plays an integral part in digestion. The endocrine component contains the islets of Langerhans that produce and secrete hormones into the bloodstream. The pancreatic hormones, insulin and glucagon, work together to maintain proper sugar levels in the blood.

Diabetes mellitus is a metabolic syndrome caused by having an insufficient number of insulin-producing beta cells. In type 1 diabetes, beta cells are destroyed by the body’s own immune system. In type 2 diabetes, although the pancreas has functioning beta cells, insulin resistance causes the liver to release too much sugar into the bloodstream, and the beta cells cannot secrete enough insulin to maintain normal glucose homeostasis. The American Diabetes Association estimated in 2012 that approximately 9.3 percent of the United States population is living with diabetes.¹⁴⁹ Beta-cell therapy holds promise for treating type 1 diabetes by replenishing beta cells in the body; however, donor pancreases are in short supply and the demand for transplantable beta cells cannot be met.

In addition, beta-like cells have been successfully generated from hESCs and iPSCs using overexpression of transcription factors, chemicals, or growth factors.¹⁵⁰ Current protocols for differentiation of pluripotent stem cells to beta cells follow a five-stage procedure that recapitulates the embryonic stages of development.¹⁵¹ Only the first four stages have been carried out successfully in vitro. The fifth stage—which involves maturation to glucose-responsive, insulin-secreting beta cells and other islet cells—until recently could only be carried out by implantation in vivo.¹⁵² A long-awaited directed differentiation of insulin-producing cells from hESCs has been accomplished; this fully defined ex vivo technology is immediately relevant.¹⁵³

Also nonendocrine cells within the pancreas have been found to transdifferentiate or reprogram to a beta cell fate²⁴ upon their being induced with the three-pancreatic-gene cocktail (Pdx1, Ngn3, and MafA). Another source for adult cell reprogramming to beta cells has been described from glucagon-producing alpha cells. A study showed that overexpression of Pax4 (a gene responsible for specifying endocrine fate) in the alpha cells was able to force them to become beta-like cells.¹⁵⁴ In another study, it was observed that near-complete ablation of beta cells forced regeneration of beta cells from former alpha cells.¹⁵⁵ However such in vivo studies have not been established in humans or other primates.

The ability to correct defective cells or give them enhanced properties (e.g., antitumor effects) represents a novel approach to transplantation medicine and sets the stage for individualized therapies. Two options exist for this strategy: provision of functional copies of a gene delivered by a viral or nonviral gene transfer system or in situ correction of the disease-causing sequence. Several major studies have used clinically employed viral transgenesis of hematopoietic stem cells (HSCs). In 2010, a γ-retroviral vector was used to deliver the complementary DNA for the IL2RG gene to CD34+ progenitors from patients with X-linked severe combined immunodeficiency (SCID-X1). Normalization of the immune system occurred in most patients; however, four patients developed acute T-cell leukemia from promiscuous LMO2 oncogene activation by the viral vector.¹⁵⁶ To mitigate the potential for viral elements to dysregulate endogenous gene expression, investigators have used self-inactivating lentiviral vectors for the gene therapy of X-linked adrenoleukodystrophy,¹⁵⁷ metachromatic leukodystrophy,¹⁵⁸ and Wiskott-Aldrich syndrome,¹⁵⁹ in gene therapy trials using hematopoietic progenitors.

Despite this, the integrating nature of viral vectors, with a preference for transcriptionally active areas, makes more precise gene targeting a highly desirable goal. Such precision can be achieved with genome editing nucleases that are rationally designed and with engineered proteins that have the unifying characteristic of recognizing and contacting a unique sequence of DNA. Most studies have tethered these proteins to nuclease domains or used their inherent ability to cut DNA. Once the DNA is broken, two predominant repair pathways have been used for therapeutic genome engineering: nonhomologous end-joining (NHEJ) and homologous recombination (HR). NHEJ is an error-prone pathway that, in the absence of a donor template, repairs the DNA break in a way that can cause small insertions or deletions (“indels”) that can permanently disrupt coding DNA sequences. Gene repair relies on the error-free HR pathway. In gene repair, the inclusion of an exogenous single- or double-stranded DNA donor template that contains homologous sequences to the target site avoids disruption. In response to a double-strand break (DSB), the donor template acts as the template for repair and allows for the precise and permanent insertion of user-defined sequences at the target locus. Both repair pathways can be used therapeutically. The major candidates employed for DNA cleavage are the meganucleases (MNs), zinc finger nucleases (ZFNs), transcription activator-like effector nucleases (TALENs), and clustered regularly interspaced short palindromic repeats (CRISPRs)/Cas9 system.

Each class of reagent has been used for stem and progenitor cell genome modification with ZFNS, which is, to date, the first to enter clinical application. Human ESC engineering with ZFNs was first used to target the HUES-3 and HUES-1 cell lines with ZFNs designed for inactivation of the CCR5 gene, a coreceptor for HIV entry to a cell.¹⁶⁰ ZFNs and a donor sequence containing green fluorescent protein (GFP) were introduced into exon 3 of the CCR5 gene and approximately 5 percent rates of targeted integration were observed. Importantly, the cells maintained their pluripotency and ability to self-renew.¹⁶⁰ This study established a precedent for inserting genes of interest into a specified spot in the ESC genome via HR. Others extended this to allow for gene addition, a placement of an inducible expression cassette at the so-called safe harbor locus AAVS1. Using ZFNs for the first exon of the PPP1R12C gene on chromosome 19, a “standalone” expression cassette was introduced containing a promoter, puromycin gene, and a polyadenylation signal (or gene trap targeting vector) containing a splice acceptor-2A-puromycin gene that relied on proper targeting and splicing with the first exon of the PPP1R12C gene.¹⁶¹ As such, gene targeting at the AAV locus allows for placement of a gene with a promoter that drives the desired level of expression or is controlled by the native PPP1R12C promoter that is constitutively active.¹⁶² In another study, employing the safe harbor strategy did not appear to alter the pluripotent nature of ESCs.^161,162,163 The ability to modify genes in pluripotent target cells is important to disease modeling in vitro, which has become a new foundation for the acceleration of translational research. Although these studies established the ability to modify ESCs at a site-specific and “safe harbor” locus, widespread use is limited by the relatively small number of approved ESC lines and the even smaller number of disease-specific ones.

As an elegant solution to address the potential paucity of disease-specific stem cells and to remove the potential for variability between stem cell lines, ZFNs have been used to generate isogenic control and Parkinson disease (PD) cell lines.¹⁶⁴ This work centered on engineering the A53T or E46K PD mutations into disease-free ESCs or repairing the mutation in PD patient-derived iPSCs.¹⁶⁴ In this way they mitigated the effects of the numerous genetic differences and modifiers that exist between individuals and ESC and iPSC clones.

Toward realizing the therapeutic potential of stem cells, investigators derived a fibroblast cell line from a humanized mouse model of sickle cell anemia; reprogrammed these cells into iPSCs; performed gene correction using a plasmid donor; differentiated the cells into hematopoietic progenitor cells; and transplanted them into sickle cell mice to reconstitute normal erythropoiesis.²² Proof of principle for a similar strategy using human cells was demonstrated using ZFNs to correct the sickle cell mutation in iPSCs that were subsequently differentiated into cells of the erythroid lineage.¹⁶⁵ Numerous studies using ZFNs, TALENs, and CRISPR/Cas9 have shown the ability to correct disease-causing mutations in iPSCs or in primary cells that are subsequently differentiated into iPSCs. A major limitation for these strategies for hematologic disorders is the poor and/or absent ability of iPSCs to form from true blood progenitors ex vivo that are capable of reconstituting a functional circulatory system. However, the most streamlined path to translational use is likely to be direct modification of a patient’s own HSCs. To date only two reports document the ability to mediate HR in HSCs. In 2007, maximal rates of 0.11 percent gene targeting at the CCR5 locus utilizing ZFNs and a donor containing GFP or a puromycin resistance gene were reached.¹⁶⁰ Subsequently, optimized conditions involving ZFNs delivered as mRNA and the donor construct delivered on an integrase-deficient lentiviral (IDLV) cassette were used to correct the IL2RG gene from an individual with SCID-X1 and observed multilineage repopulation in transplanted mice.¹⁶⁶ This specialized delivery methodology in conjunction with the use of the StemRegenin 1 aryl hydrocarbon receptor antagonist¹⁶⁷ and/or 16,16-dimethyl-prostaglandin E₂ (dmPGE₂)^168,169 allowed for HR in HSCs that normally preferentially employ NHEJ. These data provide a strong platform for first-in-human studies.

The NHEJ arm of DNA repair also holds potential for therapeutic use. A promising avenue of investigation to use NHEJ to permanently disrupt genes has been employed clinically in T cells from HIV patients.¹⁷⁰ However, because of the ability of HIV to infect non–T-cell subsets,¹⁷¹ studies have also investigated CCR5 disruption in a preclinical humanized mouse model and showed that the modified cells are resistant to HIV-1 infection.^172,173 These data are especially relevant due to the recent treatment of patients using HCT of grafts with homozygous CCR5Δ32 mutations that disrupt cellular entry of the HIV particles.¹⁷⁴ This treatment protocol was initiated for an individual with HIV/AIDS who developed acute myeloid leukemia (the “Berlin patient”) in an attempt to cure both the malignancy and the HIV infection.^174,175 Because of the paucity of CCR5Δ32/CCR5Δ32 donors, ZFN-modified HSCs are thought to be an ideal strategy for widespread implementation of this regimen. However, a 2013 evaluation of this individual showed reinfection, possibly with an HIV strain different from the initial one, indicating that CCR5 loss alone may not result in full HIV resistance. The CXCR4 receptor is a coreceptor for HIV cellular entry, and a significant number of HIV patients harbor the CXCR4-using HIV strain.¹⁷⁶ A recent combinatorial ZFN approach has been investigated in the laboratory using CXCR4 and CCR5 adenoviral-borne ZFNs, with a demonstrated ability to remove both HIV coreceptors simultaneously in human T cells.¹⁷⁷ A potential clinical limitation of this approach is the fact that CXCR4 is a critical homing molecule for HSCs¹⁷⁷ and its disruption may perturb normal HSC homeostasis.

A joint approach using nuclease-induced gene disruption with chimeric antigen receptor (CAR) expression has mitigated the potential for cell-mediated alloreactivity and maximized the antitumor cellular effects. When the T-cell receptor (TCR) α and β chains were disrupted and paired with a Wilms tumor CAR, a potent tumoricidal activity without GVHD resulted.¹⁷⁸ A therapeutic success employing T cells transduced with a CD19-specific CAR have been achieved in patients with chronic lymphoid leukemia.¹⁷⁹ Scientists have extended these findings to include CAR expression with TCR-α disruption via ZFNs.¹⁸⁰ Future improvements to this technology will include more tumor-specific antigen recognition and/or temporizing CAR expression in order to minimize B-cell aplasia and tumor lysis syndrome.¹⁷⁹

Their blood lineage plasticity and their expansive clinical application makes HSCs a desirable cell type for genome engineering with designer nucleases; however, their limited ability to form extrahematopoietic tissue limits their wide use in comprehensive disease modeling and in regenerative medicine outside the lymphohematopoietic system. ESCs and iPSCs are powerful tools for filling this void and performing nuclease genome modification of multilineage stem cells. Collectively, the convergence of stem cell technology and precision genome engineering holds tremendous potential to increase the therapeutic benefit of cell-based therapies while minimizing allogeneic transplant-associated risks. Crucial to the realization of their clinical potential will be rigorous safety assessments for each platform.

Both nucleases and pluripotent stem cells have potentially deleterious aspects that could limit their effectiveness. For pluripotent stem cells this relates to the presence or accumulation of genetic and epigenetic modifications prior to or during reprogramming. Both ESCs and iPSCs are subject to these modifications in vitro, which may manifest in the same line or even within the same culture vessel during propagation.¹⁸¹ Aneuploidy has been reported in iPSCs, their parental cellular precursors, and ESCs. Studies by the International Stem Cell Initiative suggest that karyotypic abnormalities may occur in as many as one of every three cell lines.¹⁸² Trisomy 12 is the most common abnormality in human ESCs and iPSCs, and chromosome 17 trisomy occurs frequently in murine ESCs.^182,183

By definition, engineered nucleases are designed to recognize a specific DNA sequence; however, they may also exhibit off-target (OT) effects due to overlapping or low-complexity sequence recognition between the primary target and the OT site. Unbiased genome-level screens are a powerful way to assess putative OT sites, and ZFN, TALEN, and CRISPR/Cas9 have shown excellent safety profiles to date^184,185; however, this high-resolution methodology will need to be performed for each gene target candidate. In summary, engineered nucleases allow for unparalleled specificity and flexibility that complement the attributes of progenitor cells. The ability to precisely manipulate the genome will support individualized ex vivo therapies and will allow for more uniform disease modeling in vitro.