The Therapeutic Potential of Neural Stem Cells

He was of course talking about a cruise missile, the familiar and deadly self-guided denizen lying at the heart of American military might. Although this month's column is on molecular "warfare" in the theater of the brain, the idea of a self-guided weapon destroying its target with pinpoint precision is the subject. The "missiles" discussed here are a population of neural cells in the brain and the "enemies" are a variety of brain malignancies and various neurodegenerative disorders.

An unexpected property of neural stem cells will be described that is unrelated to their controversial potential as cell replacement therapy. They appear to have great promise as drug delivery agents, which hints at an unexpected clinical application. Beginning with a brief review of a few important biological characteristics of neural stem cells (NSCs), I will then discuss their delivery properties and potential therapeutic value.

NSCs are usually defined functional- ly: they are cells possessing the remark- able property of self-renewal; under proper conditions, they are fully capable of generating a wide variety of intermediate and mature neural cells. These intermediate cells include both glial and neuronal varieties.

Many types of NSCs exist. Some subpopulations present only embryonic tissues during certain stages of brain development while some are restricted to various regions of the mature brain. So many types exist that there is currently very little agreement on an overarching in vivo definition that is capable of characterizing all known members. To make matters worse, many of these cell types are taken from tissues and cultured in petri dishes, a manipulation that always changes certain biological properties of these fragile cells.

It is not at all clear whether these cultured cells remain true biological representatives of their original populations or whether they represent a new and unique subpopulation. These are not trivial definitional issues, especially considering the political minefield through which researchers examining the clinical applications of stem cells have to navigate.

In the mature brain, NSCs are maintained in cellular "nurseries"--microenvironments surrounded by endothelial cells, astroglia, or microglia (highly mobile cells involved in the maintenance of healthy brain tissue). These nurseries help keep the NSCs in an undifferentiated state until they are recruited to change, usually as a response to a specific molecular distress signal.

Depending on what is sensed, the cells surrounding the NSCs participate in their differentiation into particular populations. The type of glial or neural cell the stimulated NSC ultimately becomes depends on the signal and type of cell (endothelial, astroglia, or microglia) involved in the detection of the initial stimulus.

The use of stem cells for research and clinical purposes has received a great deal of attention in the media during the past few years. Although researchers have a hard time knowing exactly what they are, NSCs have received particular attention, mostly because of controversy regarding their use as a renewable cellular source for damaged tissues. The attention may have resulted from the fact that a form of this application is already in clinical use. The most famous example is probably in bone marrow transplantation, which is used to treat cancers such as leukemia and various blood disorders.

Perhaps lost in this flurry of media attention is the fact that NSCs exhibit another clinically relevant property that is just as extraordinary as their multipotency. When NSCs were implanted in a laboratory animal with a brain neoplasm, researchers found that these cells began to migrate. Their target? Like a cruise missile, they moved from their original transplantation site to the tumor itself.

This was made apparent by placing a "marker gene," such as luciferase, in cultured NSCs and then injecting the cells into the animal with the tumor. Luciferase is a gene that when injected into a cell makes a protein that causes the cell to glow when given the proper substrate (this is the same protein that makes fireflies glow in the dark). If that cell is capable of moving, its progress can be tracked with the light, which is how the NSCs were observed.

Even when these glow-in-the-dark cells were placed at distances far removed from the tumor, they acted like ants streaming toward a food source. Soon after transplantation, a line of moving, glowing cells could be observed migrating from the point of injection to the tumor site.

Researchers began to question whether other types of tumors could influence the NSCs' migratory instincts, and from what distances. It turns out that many types of tumors attract NSCs, and from locations sometimes very removed from the site. Researchers also wondered what other types of brain pathologies could attract NSCs; the answer has become one of the most promising new clinical opportunities in nearly half a century. The idea has to do with drug delivery. If an NSC can be injected with luciferase, why not also load it with a molecular "weapon" such as a chemotherapeutic agent? The NSC could then act as a tiny cruise missile, rocketing down some cellular street, turning left, and smashing directly into its tumorous target with no collateral damage.

Three different signals are capable of delivering targeted information to NSCs, all following the same general mechanism. It is to that mechanism, the specific signals, and the pathologies capable of generating them that I turn to next.

The overall mechanism guiding NSC migratory behavior involves the detection of a molecular gradient. The source event sends out a widely dispersed biochemical signal and the NSC detects it, usually when a surface receptor binds to it. This binding produces interior changes to the NSC and it begins its migratory behavior in the general direction of increased concentration of the gradient. Although such chemoattraction is a common characteristic of many cell types, it is extraordinary how sensitive NSCs can be to distress signals. The following are the 3 broad classes of signaling systems to which NSCs have been shown to respond.

Inflammation. The first category describes NSC's reaction to the inflammatory response. There are many causes of brain inflammation and microglial cells are the first responders, functioning as both a surveillance system and an initial line of defense. Once microglia arrive at the insult they produce cytokines and chemokines, part of the standard molecular repertoire developed in response to injury that includes specific interleukins, such as interleukin-1b, interleukin-6, and chemokine monocyte chemoattractant protein-1.

Interestingly, microglia also produce neurotrophins, a class of molecules usually associated with promoting neural growth. Each class of molecule creates a gradient capable of mobilizing NSCs into action. This has been shown both in petri dishes and in vivo.

Reactive astrocytosis. The presence of the cytokines released at the site of injury stimulates a process known as reactive astrocytosis. This is a complex reaction by local glial cells that is not well understood, but involves, among other reactions, hyperplasia and localized hypertrophy. Astrocytosis is also associated with the release of a protein called glial fibrillary acidic protein and another called SDF-1 (stromal cell-derived factor-1), which has powerful chemoattractant properties for NSCs.

Angiogenic signals. One of the most interesting NSC signaling processes is also one of the subtlest, involving endothelial cells and their response to injury. While the mechanism is not well characterized, there is growing consensus that NCS chemoattraction involves injury-induced localized angiogenesis. During injury, new blood vessels grow in response to the release of VEGF (vascular endothelial growth factor) by local endothelial cells. At the same time, these cells also release SDF-1. Both VEGF and SDF-1 are powerful chemoattractants for NSCs. It is possible that disruption of vessel integrity due to injury allows this release, establishing a gradient that the NSCs can follow and to which they can migrate.

There are hints that blood vessels may interact with critical aspects of stem cell response in another fashion. The basal lamina that endothelial cells produce is a complex mixture of molecules, some of which have chemoattractant properties. Recent experiments demonstrated that NSCs are capable of reacting to injury-activated endothelial cells from the luminal side, possibly mediated by the embedded chemoattractants. The cells act as leukocytes, sticking to the inner surface of the blood vessel and transmigrating into relevant tissue. This has opened up the possibility that some NSCs can use the bloodstream to get around, with endothelial-derived signals creating a perivascular neural niche for immigrating NSCs.

From strokes to neurodegenerative disease, these 3 signaling systems are involved in a wide variety of brain pathologies. It is one of the reasons why researchers are excited about the potential of using NSCs in the clinical setting. What follows next are descriptions of recent work that has attempted to exploit NSCs as sophisticated drug delivery systems.

Although the applications of NSCs to brain pathologies are broad, most follow the same multistep pattern. First, NSCs are cultured in a dish, expanding their numbers in preparation for the introduction of a molecular "weapon." They are then genetically engineered to express that weapon. Finally, the modified cells are injected into the patient. They then migrate to the site of injury and deliver their payload.

One of the most interesting delivery ideas involves the use of a nontoxic prodrug stitched into the NSC chromosomal compartment via retroviral transfer (Figure). Although the cell begins to immediately express the prodrug, no harm is done as the cell travels to its target. The reason is that the prodrug remains in an inert form until it is activated by the presence of a secondary signal that is unique to the site of injury, perhaps by some sufficiently concentrated inflammatory molecule. This ensures that the molecular weapon "detonates" only at the relevant site. A molecular cruise missile indeed!

The clinical opportunities to use models like these are immense. For example, in fighting cancer, attacking brain malignancies represents some of the most intensely investigated lines of therapy. A wide variety of weapons have been proposed to equip NSCs to kill cancer cells, and a few have even been tested. These weapons include antitumor cytokines and the enzymatically activated prodrug ideas depicted in the Figure. Proteins with anti-angiogenic properties and even lytic viruses have also been researched.

Another important research focus addresses genetic defects that affect the function of the CNS. A great deal of progress has been made in targeting lysosomal storage diseases such as Krabbe disease with NSCs. Caused by a deficiency in the enzyme galactocerebrosidase, this disease is characterized by widespread inflammation in the brain. Unleashing NSCs equipped with anti-inflammatory agents represents a promising approach toward eliminating its deleterious effects.

From amyotrophic lateral sclerosis (ALS) to Alzheimer disease, a large number of research projects are also attempting to use NSCs to ameliorate the devastating effects of neurodegenerative disorders. Research has shown that the overexpression of growth factors, such as insulin-like growth factor-1, has surprisingly positive effects on the course of ALS in animal models. The big problem has been getting enough of these large molecules across the blood-brain barrier to make the treatment effective. NSCs, aside from providing target specificity, make this delivery concern irrelevant.

The effective delivery of neurotrophic factors such as nerve growth factor-1 in the treatment of Alzheimer disease has also been investigated. These growth factors exert a powerful neuroprotective effect in the brain and can even reinvigorate damaged circuitry that is common in conditions such as Alzheimer disease. Could specific delivery of these growth factors to affected regions of the brain alter the course of the disease? The use of NSCs in the treatment of other brain disorders such as neuropathic pain and Parkinson disease are also being investigated.

Like all new technologies, unabashed enthusiasm for NSC delivery systems is tempered by the turbulent reality of the clinical setting. To date, nothing has worked, although this probably represents the newness of the findings more than the failure of a large number of well-funded experiments.

Which stem cells to use, which weapon works best, and which targets are most susceptible to specific vehicle-weapon combinations represent major hurdles to overcome before these technologies can find a useful home in clinical settings. But the very notion that conditions such as stroke and Alz- heimer disease could be addressed in a site-specific manner is cause for wonder--not unlike the reaction of the Iraqi businessman who was awestruck by the powerful technology that made a missile go past his shop and take a left turn.