Staff Working Paper

Applications of Human Stem Cells in Research and Medicine

Introduction

Research using human and animal stem cell preparations continues to be an extremely active area.  It is developing new research tools, new knowledge about pathways of cell differentiation and opening new vistas of cell transplantation therapy for human diseases.  As part of an approach to fulfilling its mandate to "monitor stem cell research", The President's Council on Bioethics asked several experts to contribute review articles on various areas of stem cell research (see articles by Drs. Gearhart¹, Itescu^2,3, Jaenisch⁴ Ludwig and Thomson⁵, Prentice⁶ and Verfaillie⁷ in the Appendices).  These reviews and the present paper emphasize peer-reviewed, published work with stem cell preparations through July 2003.  Interested readers should also consult the wide variety of review articles on these types of stem cells that have appeared previously⁸.   

Heterogeneous human stem cell preparations derived from bone marrow have been in clinical use as treatment (bone marrow transplants) for various forms of cancer for many years.  More recently, hematopoietic stem cells have been isolated and purified from bone marrow and are being studied⁹.  Because this is a clinically well-established part of current medical treatment, we will not emphasize it further in this report. Instead we will focus on new types of stem cell preparations, and ways in which they might be used in research and medical treatment. 

This paper provides additional background on selected aspects of recent work with human embryonic and adult stem cells.  It is intended to be read in conjunction with the review articles cited above.  It emphasizes some general considerations involved in obtaining reproducible results in experiments with stem cell preparations.  It focuses on work with human stem cell preparations that can be reproducibly isolated and considerably expanded through growth in vitro, while maintaining the essential characteristics of stem cells.  It also describes the current state of progress toward development of stem cell-based therapies for some specific human diseases.

Stem cells share two characteristic properties: 1) unlimited or prolonged self-renewal capacity (i.e. the capability to maintain a pool of undifferentiated stem cells), and 2) the potential to produce two or more differentiated descendent cell types (see Figure 1).  As embryonic stem cells differentiate, they generally become more restricted in the differentiated cell types that they can generate. 

Figure 1. Schematic Diagram of Some Stages in Cell Differentiation

Under the influence of various cell differentiation signals, embryonic stem cellsⁱ differentiate into multiple descendent stem cells [dashed box area of Figure 1.]  These descendent stem cells can also self-renew, that is, they can undergo cell division to produce more descendent stem cells while retaining their ability to differentiate into multiple cell types.  With increasing frequency, such cells are being isolated from various "adult" tissues such as bone marrow (e.g. mesenchymal or stromal stem cells) and brain (e.g. neural stem cells) [see (6)]. 

Subsequent exposure to additional cellular differentiation signals can cause the descendent stem cells to differentiate further, so that they finally become the wide variety of differentiated tissue cells that make up the adult body (labeled A-D in Figure 1).  At each stage of the differentiation process, specific sets of genes are expressed (and other sets are repressed or turned off), so that hundreds of specific proteins, which give the cells their differentiated properties, can be produced.  Specialized proteins called transcription factors play key roles in determining which sets of genes are expressed and repressed.

The terminology used in describing the isolation and properties of different stem cell types can sometimes be confusing.  In this paper, adult stem cells [see (6)] are preparations from humans or animals isolated after birth that have the characteristic two properties described above.  Embryonic stem cell preparations [see (5)] isolated from the inner cell mass of early human embryos show these same two characteristic properties, and are abbreviated "hESCs".  Human cell preparations isolated from fetal germ cell tissue [see (1)] have similar properties to hESCs cells and are abbreviated "hEGCs".

Reproducible Results Using Stem Cell Preparations and Their Derivatives

A major goal of scientific research is the production of reliable knowledge based upon reproducible results.  To obtain reproducible results in experiments using stem cells, it is essential to produce, preserve and characterize stem cell preparations so that they generate reproducible results when the same experiment is repeated.  As described in more detail below, the tendency of stem cell preparations to differentiate into more specialized cells in vitro makes the task of obtaining homogeneous stem cell preparations that will generate reproducible results especially challenging.  Failure to address these problems with appropriate experimental methods may yield experimental results that are difficult or impossible to reproduce.

Initial Stem Cell Preparations Can Contain Multiple Cell Types

Isolation of adult stem cells from source tissues initially yields a heterogeneous cell preparation.  The initial preparation contains cells from the source tissue and may also include red blood cells, white blood cells and possibly circulating stem cells.  Initial mixtures of cells may then be treated in various ways to remove unwanted contaminating cells, thereby increasing the percentage of adult stem cells in the preparation.  Seldom (if ever) does one produce an adult stem cell preparation that is 100% adult stem cells, unless the adult stem cell preparation has been "single cell cloned" in vitro (see below).

Isolation of hESC cell preparations from ICM cells of blastocyst-stage embryos raises similar questions about the homogeneity of the initial cell preparations.  In the past, many hESC cell preparations were initially isolated by in vitro growth on top of irradiated "feeder" layers of mouse cells.  More recently, several groups have shown that it is possible to grow hESCs on feeder layers of human cells, including fibroblasts obtained from skin biopsies, or without any feeder cells¹⁰.  It is thought that the feeder cells secrete factor(s) that enable the stem cells to replicate while maintaining a relatively undifferentiated phenotype.  Subsequent in vitro growth of human stem cell preparations grown on mouse cells may have carried along some viable mouse cells.

Genetically homogeneous stem cell preparations through "single cell cloning"

Stem cell preparations growing in vitro are sometimes "single cell cloned" by placing a cylinder over a single cell located with a microscope.  Further cell division by this isolated single cell within the cylinder produces a cell preparation in which all the cells are descended from the originally isolated single cell.  The cells within the cylinder are then harvested and expanded further in vitro.  The resulting stem cell preparation is said to be "single cell cloned".  The stem cells within a "single cell cloned" preparation are homogeneous in the sense that they are all derived from the same original cell.  Some of the human embryonic stem cell preparations produced prior to August 9, 2001 have been "single cell cloned"¹¹, and are then called stem cell "lines".

In vitro propagation, freezing and storage at low temperature

Stem cell preparations can be grown in vitro so that the cells multiply, and many, presumably identical, vials of the cells can then be frozen and preserved at very low temperatures.  Frozen stem cell preparations can then be thawed and grown again in vitro to produce larger numbers of cells.  Selective pressures involved in both the in vitro growth process and the freezing step can increase the heterogeneity of a stem cell preparation by favoring the multiplication of cell variants in the preparation.  Whether or not several cycles of freezing and thawing change the phenotypic characteristics of stem cell preparations has not been much studied.  However, the practical advantages of preserving multiple samples of stem cell preparations by freezing are too large to ignore. These advantages are central to the ability to repeat an experiment with a very similar stem cell preparation, and to the ability to treat multiple patients with a stem cell preparation derived initially from a single donor sample.

Normal human stem cells have 46 chromosomes, 22 pairs plus two X chromosomes if they are from a female, or 22 pairs plus one X and one Y chromosome if they are from a male.  During the copying of chromosomal DNA and the separation of daughter chromosomes at cell division, rare mistakes occur that lead to the formation of abnormal chromosomes.  Cells with abnormal chromosomes can progress to malignancy, so retention of the normal human chromosome number and organization is an essential characteristic of useful stem cell preparations.  Human stem preparations generally have normal chromosome numbers and structure, and appear to retain this property better than mouse stem cell preparations^{7 12} .

In summary, researchers must address several factors in order to maximize the probability of obtaining reproducible results with stem cell preparations.  The more stem cell preparations are:

1. "single cell cloned", with a normal chromosome structure and number, and
2. multiple samples preserved by storage at very low temperature, and
3. compared in experiments where cells from the same lot of frozen material are used, and
4. well-characterized as to the absence of cellular, viral, bacterial, fungal and mycoplasma contaminants, and
5. tested to determine the proportion of stem cells and various differentiated cells in the cell preparation used in the experiments,

the greater the likelihood that results in experiments with stem cell preparations will be reproducible.  Human stem cell preparations with these properties will be in the forefront of stem cells that are useful in basic research and that will be investigated for possible clinical applications.

In the remainder of this paper, we discuss primarily stem cell preparations that exemplify the considerations listed above.  We focus on mesenchymal stem cells and neural stem cells among adult stem cell preparations, and on hESC and hEG cells among embryonic stem cell preparations.  For information on the wide variety of other stem cell preparations isolated from adult tissues, see reference 6.  Further research on some of these other adult stem cell preparations may reveal that they can also be "single cell cloned", grown extensively in vitro, cryopreserved, and characterized as to the absence of contaminants.  At that point, it would be very important to compare the properties of these other adult stem cells, and more differentiated cells that can be derived from them, with the already characterized embryonic and adult stem cell preparations.

Stem cell preparations that multiply extensively and continuously in vitro have been isolated from a variety of human tissues, including bone marrow, brain, cord blood, teeth and various organs (for details, see reference 6).  The focus in this section will be on two of the best studied of these, mesenchymal stem cells^iv from bone marrow, and on neural stem cells isolated from brain tissue.

The bone marrow contains at least two major stem cells, hematopoietic stem cells that give rise to the red cells and white cells of the blood⁸, and a cell type that adheres in vitro in plastic culture dishes called mesenchymal stem cells (MSC).  As described originally by A.J. Friedenstein, M.E. Owen and A.I. Caplan and their coworkers¹³, these MSCs were studied primarily as precursors of bone.  More recent work has indicated that these cells can be reproducibly isolated and expanded in vitro, and that they can give rise to cartilage, bone, adipose (fat), and muscle cell lineages¹⁴ in vitro.

The phenotypic characteristics (morphology, expressed proteins and biological properties) of these cells have been somewhat difficult to characterize, because they appear to vary depending upon the in vitro culture conditions and the specific cell preparation (see discussion in reference 15).  A molecular analysis of genes expressed in a single cell-derived colony of these cells provided evidence for the expression of genes also turned on in bone, cartilage, adipose, muscle, hematopoiesis-supporting stromal, endothelial and neuronal cells¹⁵.  It is possible that the cells within the colony entered into distinct cell differentiation pathways, resulting in a heterogeneous population composed of several different cell types.

hMSCs are important stem cells for research and therapy for several reasons.  First, because they can be differentiated in vitro into multiple cell types, they are making possible detailed research on the molecular events that underly differentiation into bone¹⁶ cartilage, and fat cell lineages.  Second, they have recently been shown to

support the in vitro growth of human embryonic stem cells¹⁷, thus replacing the mouse feeder cells used previously.  This would obviate concerns about xenotransplantation if the hESCs or their derivatives were ever used in human clinical transplantation therapies.  In addition, clinical studies are already underway in which hMSCs are co-transplanted with autologous hematopoietic stem cells to replace the blood cell-forming system of cancer patients who have received high dose chemotherapy¹⁸.  As summarized by Itescu [see (2)], MSCs may also prove to be important in modulating the immune system to more readily accept foreign tissue grafts.  Finally, hMSCs have the potential for cell replacement therapies in injuries involving bone, tendon or cartilage, and are, in fact, already being tested as experimental therapies for the human diseases osteogenesis imperfecta, metachromatic leukodystrophy, and Hurler syndrome (a lysosomal storage disease)¹⁹.

The relationship between hMSCs and the Multipotent Adult Progenitor Cells (MAPCs) described by Verfaillie and coworkers [see (7)] needs to be clarified by further research.  Both cell preparations are isolated as adherent cells from bone marrow aspirates.  Each can be differentiated in vitro into cells with cartilage, bone and fat cell properties.  They express several of the same cell antigens, but are reported to differ in some others⁷.  MAPCs have to be maintained at specific, low cell densities in vitro, otherwise they tend to differentiate into MSCs⁷.  Two recent publications make clear how the properties of the final cell preparations reflect both a heterogeneity in growth properties of the mesenchymal stem cells as isolated in vitro, and the emergence of different gene expression profiles as the cells go through multiple populations doublings in vitro²⁰.

Stem cells capable of differentiating into one or more neural cell lineage (i.e. astrocyte, oligodendrocyte, neuron) can be isolated from brain tissue (particularly the olfactory bulb²¹) and grown in vitro.  There appear to be at least two types of neural stem cells, one from the subventricular zone (SVZ), and another from other brain regions²².  In the presence of purified growth factor proteins, the cells can be expanded by growth in vitro as round clumps of cells called neurospheres.  However, many neurospheres in culture are developmentally heterogeneous, and the number of self-renewing stem cells is frequently low (<5%)²².

Human neural stem cells from developing human cortex, expanded in culture with epidermal growth factor (EGF), become senescent (cease dividing) after 30-40 population doublings²³.  Leukemia inhibitory factor (LIF) allowed the growth of a self-renewing neural stem cell preparation for up to 110 population doublings.  Studies using Affymetrix "gene chips" provided evidence for the expression of specific members of important growth factor and signal transduction gene families.  Withdrawal of LIF led to decreased expression of 200 genes²³.  The magnitude and specificity of the molecular information relevant to the regulation of neural stem cell multiplication obtained in this experiment powerfully illustrates the usefulness of stem cell preparations in basic biomedical research.

Human neural stem cell preparations can be transduced in vitro with foreign genes so that the treated cells now express readily visualized specific proteins, such as Green Fluorescent Protein (GFP).   This makes it possible to track GFP-marked human neural stem cells after they are injected into experimental animals, to determine whether they survive and migrate following injection.  Studies of this type have provided evidence that human neural cells can migrate extensively after injection²⁴.   In addition, such cells can be injected into animal models of human diseases such as intracerebral hemorrhage and Parkinson's Disease (PD) to study their effect on the progression of the disease²⁵.  Although human neural stem cells may not yet be as well characterized as mesenchymal stem cells or hESCs, they are being actively studied with the hope that they can be used in future treatments for devastating neurological diseases such as stroke, Alzheimer's Disease and PD.

Prentice [see (6)] has summarized a large amount of current information on stem cell preparations isolated from amniotic fluid, peripheral blood, umbilical cord blood, umbilical cord, brain tissue, muscle, liver, pancreas, cornea, salivary gland, skin, tendon, heart, cartilage, thymus, dental pulp and adipose tissue.  In several of these cases, the stem cells isolated are of animal rather than human origin.  In other cases, the long-term expandability in vitro of the stem cell preparations was not demonstrated.   Studies of many of the stem cell preparations from these sources are just getting started, and further work is needed to determine their biological properties and their relatedness to other stem cell types.  The demonstration that they can be isolated from such tissue compartments in animals should spur the search for similar human stem cell types.

Human embryonic stem cell (hESC) preparations have been isolated from the inner cell masses of blastocyst-stage human embryos in multiple laboratories throughout the worldⁱⁱⁱ.  As of July 2003, 12 hESC preparations are available for shipment to recipients of U.S. federal research grants, out of a total of 78 eligible hESC preparations^iv.  Limited characterization data for most of the 12 currently available hESC preparations is summarized in reference 12.  The review by Ludwig and Thomson⁵ lists more than 40 peer-reviewed hESC primary research papers that have been published since the initial publication in 1998. 

Although isolated from different blastocyst-stage embryos in laboratories in different parts of the world, hESC preparations have a number of properties in common.

These include expression of common cell surface antigens recognized by binding of specific antibodies, expression of the enzyme alkaline phosphates, and production of the gene-regulating transcription factor Oct-4¹².  Many hESC preparations have been cryopreserved at low temperatures so they can be well characterized and compared with one another.

hESC preparations are initially isolated as colonies of undifferentiated cells on "feeder" layers of mouse or human cells.  The feeder cells are believed to supply protein and perhaps other cell products that are needed by the ESCs to retain the property of symmetrical self-renewal.  Appropriately supplemented, the culture medium from feeder layer cells can sustain ESCs in vitro.  The culture medium from mouse feeder layer cells is very complex, containing hundreds if not thousands of different proteins²⁶.  Purified Leukemia Inhibitory Factor (LIF) can replace the feeder cell layer requirement for mouse ESCs, but not for human ESCs. 

hESC preparations have been differentiated in vitro into neural (neurons, astrocytes, and oligodendrocytes), cardiac (synchronously contracting cardiomyocytes), endothelial (blood vessels), hematopoietic (multiple blood cell lineages), hepatocyte (liver cell), and trophoblast (placenta) lineages⁵.  In the case of neural and cardiac lineages, similar results have been obtained in different laboratories using different hESC preparations, thus fulfilling the "reproducible results" criterion described above.  For other lineages, the results described have not yet been reproduced in another laboratory.

Somatic cell nuclear transfer (SCNT) creates cloned embryos from which cloned embryonic stem cells can be isolated [see (4)].  In theory, using cloned embryonic stem cells from individual patients might provide a way around the possible immune rejection problem (see below).  To date, such experiments have only been successful with mouse cells.  However, they have made possible an experiment demonstrating the potential of cloned embryonic stem cells in the possible future treatment of genetic diseases.  Rideout et al²⁷used a mutant mouse strain that was deficient in immune system function.  They produced a cloned mouse stem cell line carrying the mutation, and then specifically repaired that gene mutation in vitro.  The repaired cloned stem cell preparation was then differentiated in vitro into bone marrow precursor cells. When these precursor cells were injected back into the genetically mutant mice, they produced an observed partial restoration of immune system function.

Production of cloned human embryonic stem cell preparations remains technically very difficult and ethically controversial.  Recently however, Chen and coworkers²⁸ have reported that fusion of human fibroblasts with enucleated rabbit oocytes leads to the development of embryo-like structures from which cell preparations with properties similar to human embryonic stem cells can be isolated.  This work needs to be confirmed by repetition in other laboratories. In addition, further work is needed to decisively settle the question of whether rabbit mitochondrial DNA and rabbit mitochondrial proteins persist in the embryonic stem cell preparations.  Persistence of rabbit mitochondrial proteins in the cloned stem cell preparations could possibly enhance the possibility of immune rejection of the cells, thus limiting their clinical application.

Stem cell preparations are proving useful in basic research in at least two ways.  First, they are useful in unraveling the complex molecular pathways governing cell differentiation.  For example, preparations of embryonic stem cells can be induced to become more differentiated cells in vitro.  In one case, this has made available quantities of human trophoblast-like cells that are difficult to obtain from other sources³⁰.  Since, in the best circumstances, these can be quite homogeneous cell preparations, their gene expression profiles can be compared in detail^{16, 23}.  In addition, cultures of differentiated cells derived from stem cells could be used to test new chemical compounds for toxicity and mutagenicity³¹.  As experience with these differentiated derivatives of hESCs grows, it may become possible to reduce or eliminate the use of live animals for toxicity and mutagenicity testing of new chemical compounds.

The ability to grow large, relatively homogeneous populations of human stem cells and their more differentiated descendants in vitro means that one can now study certain molecular aspects of cell differentiation in vitro.  The genes expressed by the stem cell population can be compared with the genes expressed by the differentiated cell population in order to characterize the differentiated cell not just by one or two biological markers, but by the expression pattern of hundreds or thousands of genes.  Genes specifically expressed by stem cells that are important to maintaining their specific properties are being identified³² Research of this kind is providing insight as to the role of specific transcription factors in switching gene sets off and on as cell differentiation progresses.

In general, one approach to using human stem cells in future therapies starts from the isolation of a cell that can be maintained in the undifferentiated state and expanded substantially by growth in vitro.  A large number of undifferentiated stem cells could then be stored in the frozen state, characterized in vitro, and used as a reproducible starting material from which to prepare differentiated cell preparations that express beneficial properties when transplanted into patients with specific diseases.  To make more concrete both the potential and problems with this approach, we discuss the potential problem of immune rejection of the transplanted cells, and summarize some current information on the properties of cells derived from human stem cell populations in an animal model of Type-1 diabetes as a case-study example.

One type of experiment that has been done to examine this possibility was to examine human embryonic and mesenchymal stem cell preparations growing in vitro for the expression of gene products known to play important roles in the immune rejection process.  Drukker et al³³ showed that hESC preparations in vitro express very low levels of major histocompatibility complex (MHC) class I proteins on their cell surface. Expression of MHC-I proteins increased moderately when the hESC cells differentiated in vitro or in vivo.  A more pronounced increase in MHC-I protein expression was observed when the hESCs were treated with ³-interferon.  These experiments indicate that expression of MHC-I proteins that can play important roles in immune rejection can depend on the differentiated state of the cells and upon the presence of specific immune effector molecules like ³-interferon in the environment.

Similarly, Majumdar et al³⁴ showed that human mesenchymal stem cell preparations in vitro express multiple proteins on their cell surfaces that would enable them to bind to, and interact with, T lymphocytes.  The presence of such molecules helps to explain the observed ability of hMSCs to modulate immune reactions in vitro and in vivo (see references in 2).  Majumdar et al also observed that ³-interferon increased both human leukocyte antigen (HLA) class I and class II molecules on the surface of hMSCs.  These studies indicate that it will probably not be possible to predict the likelihood that transplanted stem cell preparations will trigger immune rejection processes in vivo, solely on the basis of in vitro experiments.  The in vivo environment encountered by transplanted embryonic and mesenchymal stem cell preparations will likely contain different concentrations of immunomodulatory molecules like ³-interferon than are contained in the growth medium in a petri dish.

Case Study: Stem cells in the future treatment of Type-1 Diabetes?

The human body requires glucose for cell energy, central nervous system functioning, and other critical tasks. Glucose is produced by the liver and generated from dietary carbohydrates. A protein hormone called insulin, produced by the beta cells in the islets of Langerhans of the pancreas, regulates the use of glucose by the body and the amount of glucose in the blood. 

Type-1A diabetes is caused by a known autoimmune response. Up to 90% of cases of type-1 diabetes are type-1A.  Other forms of type-1 diabetes include Type-1B (idiopathic, or of unknown origin), and latent autoimmune diabetes of adulthood (LADA).  People with type-1A diabetes have little to no insulin production because the islet beta cells of their pancreas have been damaged by their own immune system. They have a type of self-allergy that causes the T lymphocytes of the immune system to attack their pancreatic beta cells as if they were a foreign invader. Eventually, this results in the destruction of most or all of the insulin-producing pancreatic beta cells.

When insufficient insulin is produced because the beta cells of the pancreas have been destroyed, tight regulation of blood glucose concentrations is lost.  Abnormally high blood glucose concentrations result in modification of cellular proteins and pathologic changes in blood vessels, degeneration of the retina, and kidney failure.  Abnormally low blood glucose concentrations can lead to diabetic coma and death in severe, untreated cases.

Type-1 diabetes is a devastating, lifelong condition that currently affects an estimated 550,000-1,110,000 Americans³⁵.  As such, it exacts a significant burden on the US healthcare system and economy as a whole, over and above the drastic quality-of-life reductions that are borne by its individual sufferers. Recent estimates suggest that treatment of all forms of diabetes cost Americans a total of $132 billion per year³⁶.  Since Type-1 diabetes comprises roughly 5-10% of all diabetes cases, its costs can be estimated as $6.5 - $13 billion per year.

Measuring blood glucose and injecting human insulin preparations several times a day is the current treatment for type-1 diabetes.  Although this treatment is life prolonging for type-1 diabetics, its procedures are painful and in many cases they do not result in adequate control of blood glucose concentrations.  Whole pancreas transplants can essentially cure type-1 diabetes, but less than 2000 donor pancreases become available for transplantation in the U.S. each year³⁷, and they are primarily used to treat patients who also need a kidney transplant.  Pancreas transplant patients have to continuously take powerful drugs to suppress the immunological rejection of the transplanted pancreas.

Recently, a less intensive immunosuppressive treatment (the Edmonton protocol),

coupled with islet cell transplantation into the liver, has produced encouraging responses

in early clinical trials³⁸.  Expanded clinical trials of this procedure are currently underway.  Also being evaluated are methods of slowing the autoimmune destruction of beta cells in the pancreas that subsequently results in disease.

Whole pancreas and islet cell transplants ameliorate Type-1 diabetes, but there are insufficient quantities of these materials to treat all Type-1 diabetics.  It is hopeful to think that, if human stem cells could be induced to differentiate in vitro into functional pancreatic beta cells, this material limitation might be overcome.  Of course, it would also be crucial to prevent immunological destruction of the newly transplanted beta cells. 

Initial experiments in mice suggested that insulin-producing cells could be obtained from stem cells following in vitro differentiation³⁹.  Can this approach be extended to human stem cells?  In a recent paper, Lechner and Habener provided a list of six criteria to define a pancreas-derived cell that could be potentially useful in treatment of Type-1 diabetes⁴⁰.  We have used these six criteria to facilitate assessment of the current state of progress toward development of functional beta cells that might eventually be tested in human Type-1 diabetes patients (see Table 1, which summarizes the properties of human cells studied by Abraham et al⁴¹ Zulewski et al⁴², Assady et al⁴³, Zhao et al⁴⁴, and Zalzman et al⁴⁵).  The evidence for insulin synthesis in each paper is specifically listed, since Rajagopal et al⁴⁶ have shown that bovine insulin, taken up by human stem and differentiated cells when they are grown in medium containing 10% fetal bovine serum, can be mistaken for human insulin synthesized by the human cells.    

The results described in Table 1 indicate that cells derived from some human stem cell preparations were able to reverse hyperglycemia in mouse models of human diabetes⁴⁷.  Although these results are encouraging, the likely immune rejection of the transplanted human cells was prevented in these experiments by using immunodeficient (SCID) mice.   Because most cases of Type-1 diabetes are caused by immune destruction of the pancreatic beta cells, prospective new cell transplantation therapies for Type-1 diabetes will also need to assess whether there is rapid or chronic immune reaction to the transplanted cells.  It will be important to determine whether the candidate human beta-like cells express MHC class I and class II antigens in vitro and in vivo in animal models of diabetes.  The studies described in Table 1 did not address this point.  

In addition, although no tumors were observed in the transplanted mice, the experiments were terminated after about 2 months, so there was a very limited time for tumor development to occur.  Because many Type-1 diabetes patients are children and there is a currently effective therapy, there will need to be a high degree of certainty that any cells transplanted into such patients will not become malignant over the course of their lives.  This means that stringent tests of the tumorigenic potential of candidate cell preparations for Type-1 diabetes treatment will be required.  These tests will need to include multi-year studies in longer-lived animals than mice or rats.  Long-term follow-up of children and adult patients who received whole bone marrow transplants years ago has revealed an increased risk of severe neurologic complications⁴⁸and a variety of types of cancer⁴⁹.

Before stem cell-based therapies are used to treat human diseases, they will have to gain approval through the FDA regulatory process.  The first step in this process is filing an Investigational New Drug Exemption (IND) application.  As of July 2003 four IND applications have been filed for clinical applications of mesenchymal stem cells.  The disease indications include: 1) providing MSC support for peripheral blood stem cell transplantation in cancer treatment, 2) providing MSC support for cord blood transplantation in cancer treatment, 3) using MSCs to stimulate regeneration of cardiac tissue post acute myocardial infarction (heart attack), and 4) using MSCs to stimulate regeneration of cardiac tissue in cases of congestive heart failure.  Each of the first two applications is currently in Phase II of the regulatory process, with pivotal Phase III trials scheduled to begin in 2004⁵⁰.

Legend to Table 1: Six benchmark criteria for assessing the characteristics of candidate replacements for functional ß-cells are addressed above. Clonally isolated or marked cells are thought to provide a well-characterized, and thus consistent, source of cells for replicable tissue engineering. Of interest, the ES cell line used by Assady et al (the H9 stem cell line) is approved for federal funding under NIH guidelines. ß-cell-specific markers include: PDX-1 (a.k.a. IPF-1), a regulatory gene that has been shown to be important for pancreatic ß-cell function; glucokinase (GK), a gene encoding an enzyme that is largely responsible for detecting high levels of glucose and modulating insulin release in the pancreas; GLUT-2, a gene associated with glucose-responsive insulin secretion; and CK19, a marker for pancreatic ducts. Insulin production was judged by a number of criteria: synthesis of the messenger RNA for insulin, or its precursor, preproinsulin; tests for the presence of insulin protein itself; and ultrastructural studies (like electron microscopy) to determine whether insulin was present in secretory granuoles, which would suggest its intracellular production. In addition, the glucose-responsiveness of insulin-producing stem cell derivatives was assessed in a number of the studies described above and is considered an essential characteristic of normal pancreatic ß-cell function.

Two mouse models of Type-1 Diabetes were used in the in vivo studies referenced above. Both used mice that had a condition known as Severe Combined Immunodeficiency (SCID) and that were treated with streptozoticin (STZ), a drug that induces destruction of insulin-producing cells. The mice in the Zalzman study were also born with a form of mouse diabetes, and are called Non-Obese Diabetic (NOD) mice. Hyperglycemia refers to high levels of blood sugar (glucose) seen in patients unable to synthesize insulin.

Embryonic stem cell-based therapies for human diseases do not appear to have progressed as far toward human clinical application as have therapies based on MSCs.  As of July 2003, no IND applications had been filed for clinical trials of embryonic stem cell-based therapies.  A possible candidate for the first clinical application with these cells is using oligodendrocyte-like cells derived from human embryonic stem cells to treat cases of spinal cord injury (Keirstad, H., personal communication).  An IND application may be filed for this indication in late 2004 or 2005⁵¹.

While it might be argued that it is too soon to attempt to draw any conclusions about the state of a field that is changing as rapidly as stem cell research, it may be useful to attempt to draw some preliminary conclusions regarding the current state of the field.

1. Human stem cell preparations can be reproducibly isolated from a variety of embryonic, fetal and adult tissue sources. 

2. Some human stem cell preparations (e.g. hESCs, hEGs, hMSCs and MAPCs) can be reproducibly expanded to substantially larger cell numbers in vitro, the cells can be stored frozen and recovered, and they can be characterized and compared by a variety of techniques.  These cells are receiving a large share of the attention regarding possible future non-hematopoietic stem cell transplantation therapies.

3. hESC, hEGC, hMSC and MAPC preparations can be induced to differentiate in vitro into cells with properties similar to those of cells in differentiated tissues.

4. Research using these human stem cell preparations holds promise for a) increased understanding of the basic molecular process underlying cell differentiation, b) increased understanding of the early stages of genetic diseases and possibly cancer, and c) future cell transplantation therapies for human diseases.

5. The case study of developing stem cell-based therapies for Type-1 diabetes illustrates that, although insulin-producing cells have been made, we could still have a long way to go before stem cell-based therapies are developed for this disease.  This appears to be true irrespective of whether one starts from human embryonic stem cells or from human adult stem cells.

6. Human mesenchymal stem cell preparations are currently being evaluated in pre-clinical studies and clinical trials for several specific diseases.

_________________

Footnotes

_________________

Endnotes

• In vitro differentiation to a fully functional beta cell should be unequivocally established.  Insulin expression per se does not make a particular cell a beta cell.  The expression of other markers of beta cells (e.g. Pdx1/Ipf1, GLUT2, and glucokinase) or other endocrine islet cells should be demonstrated.
• Ultrastructural studies should confirm the formation of mature endocrine cells by identification of characteristic insulin secretory granules.
• The in vitro function of endocrine cells, differentiated from stem cells, should be reminiscent of the natural counterparts.  For beta cells, this would imply a significant glucose-responsive insulin secretion, adequate responses to incretin hormones and secretagogues, and the expected electrophysiological properties.
• In vivo studies in diabetic animals should demonstrate a reproducible and durable effect of the stem/progenitor-derived tissue on the attenuation of the diabetic phenotype.  It should also be demonstrated that removal of the stem cell-derived graft after a certain period of time leads to reappearance of the diabetes.
• For future clinical use, the tumorigenicity of stem/progenitor tissue should be determined. Additionally, immune responses toward the transplanted cells should be examined.

1. Rambhatla, L., et al., "Generation of Hepatocyte-Like Cells from Human Embryonic Stem Cells", Cell Transplantation, 12, 1-11 (2003)
2. Carpenter, M.K., et al., "Enrichment of Neurons and Neural Precursors from Human Embryonic Stem Cells", Experimental Neurology, 172, 383-397 (2001)
3. Sottile, V., et al., "In Vitro Osteogenetic Differentiation of Human ES Cells", Cloning and Stem Cells, 5, 1-7 (2003)
4. Chadwick, K., et al., "Cytokines and BMP-4 promote hematopoietic differentiation of human embryonic stem cells", Blood, 102, 906-915 (2003)
5. Carpenter, M.K., et al., "Protocols for the Isolation and Maintenance of Human Embryonic Stem Cells" NO CITATION
6. Lebkowski, J., et al., "Human Embryonic Stem Cells: Culture, Differentiation, and Genetic Modification for Regenerative Medicine", The Cancer Journal, 7 (Supplement 2), S83-S93 (2001)
7. Xu, C., et al., "Characterization and Enrichment of Cardiomyocytes Derived From Human Embryonic Stem Cells", Circulation Research, 91, 501-8 (2002)
8. Lysaght, M.J. and A.L. Hazelhurst, "Private Sector Development of Stem Cell Technology and Therapeutic Cloning", Tissue Engineering, 9, 555-561 (2003)
9. Mitalipova, M., et al., "Human Embryonic Stem Cell Lines Derived from Discarded Embryos", Stem Cells, 21, 521-526 (2003)