People who lose a limb to war, accident, or disease can choose from a remarkable array of prosthetic replacements, including legs specialized for cyclists and sprinters or arms with hands that can grasp and manipulate playing cards and paring knives.
Bioengineers continue to refine prosthetic limbs, but they still can’t replicate the entire constellation of capabilities provided by flesh and blood. So a few determined scientists are pursuing a different solution: They are seeking the recipe for regrowing a missing limb.
The ability to assemble a biological limb cell by cell in a lab requires so much detailed knowledge and such advanced technical capabilities that it may be a century off, according to biologist Michael Levin, who directs the Tufts Center for Regenerative & Developmental Biology at Tufts University. “But asking the host organism to build a limb is a much more achievable goal,” he says. After all, the body already knows how to create limbs, having done so during embryonic development. So researchers are attempting to reactivate the body’s existing biochemical programs for limb growth without micromanaging all the details of the process. Taking this shortcut “is very important for bringing this to biomedical applications sooner rather than later,” Levin says.
For illustrative models, researchers in the field turn to animals that can replace body parts easily. Nonvertebrates including hydras and flatworms can grow new brains, digestive organs, nerve cords, and muscle even when huge chunks of their tissue are removed, Levin says.
Regenerative capabilities are rarer and much less powerful among vertebrates. Fish can regrow fins—including skin, bone, blood vessels, nerves, and connective tissue—as well as kidney tissue and heart muscle. Deer routinely sprout new antlers. And humans can repair flesh wounds, the liver, and nerves outside the brain and spinal cord, according to Jeremy P. Brockes, who studies cellular and molecular mechanisms that underlie regeneration at University College London.
But these fairly simple repair jobs differ considerably from “regenerating something of the complexity of a limb, which involves formation of the bones and the digits and the muscles and so on in a very intricate pattern,” Brockes says.
The closest that humans come to regenerating a limb is the ability of young children to regrow a fingertip including bone, blood vessels, nerves, skin, and the fingernail. Mice share this ability, even in adulthood, although if too much tissue is lost, neither mouse nor child can rebuild a digit.
Molecular biologist Ken Muneoka, who studies limb development and regeneration at Tulane University, in New Orleans, is exploring how a digit loss that can be repaired differs from one that can’t.
Working with mice, his team has discovered that cells release bone morphogenetic proteins (BMPs) at regeneration-capable amputation sites but not at sites that fail to regenerate. The researchers also determined that treating a normally nonregenerative amputation site with these growth factors enables the wound to grow a replacement digit. And they found that mice in different stages of development—newborn versus embryonic, for example—utilize different BMPs to promote regeneration (Development 2010, 137, 551). BMP production is turned on by MSX genes, whose expression increases during regeneration of mouse digits, as well as during regeneration of tadpole tails and zebrafish fins, according to Muneoka.
An adult mouse can replace a digit tip, but the only adult vertebrate that can regenerate an entire limb is the salamander, and it completes the task in 10 weeks or less. In fact, salamanders, which include newts, “really are the champions of regeneration among adult vertebrates,” Brockes says. “As well as the limb, they can regenerate the tail, the jaw, ocular tissue like the lens and retina, and sections of the heart and intestine.”
After a salamander loses a limb, epidermal cells move to seal the wound within about six hours. The wound epidermis emits signaling compounds including fibroblast growth factors that attract connective tissue cells known as fibroblasts to the wound site, Muneoka explains. Once there, the cells revert to a less-developed state, coalescing and dividing to form a “blastema,” a mound of stemlike cells that differentiate into multiple types of cells. The blastema cells “show striking positional identity by regenerating just the structures that were removed,” Brockes says. “For example, wrist cells give rise to a hand” and not to a second forearm. Muneoka believes that Hox genes preserve this positional information in salamanders’ fibroblasts.
Blastema formation—and thus the entire process of regeneration—can succeed only if the severed nerves regenerate as well, Brockes notes. A few years ago, his team discovered that both the regenerating nerves within the amputated stump and skin near the wound express a protein they dubbed newt anterior gradient (nAG). The researchers believe that the binding of nAG to the cell-surface protein Prod 1 promotes growth of appropriate cells in the stump (Science 2007, 318, 772).
Other biochemicals that appear to help the blastema form and grow include matrix metalloproteinases. These enzymes may function in part by preventing scar formation, according to University of Utah neurobiologist Shannon J. Odelberg.
It’s still a mystery why some animals can regenerate and others can’t, Brockes adds. Researchers don’t even know whether salamanders are unique among vertebrates in having evolved their extensive regenerative ability, or whether ancestral animals possessed the ability but salamanders are unique in retaining it.
Most scientists favor this second hypothesis, Brockes says. They think that once they identify the factors that stop most mammals from regenerating, such as inflammation caused by immune cells or scar formation caused by connective tissue during wound healing, they can remove that block. In that way, they can access “some ancestral property, some ‘newt within,’ that’s just waiting to be unlocked,” and reactivate regenerative capabilities, he says.
But regeneration involves the same signaling pathways, transcription factors, and other biochemical machinery used in development, tissue turnover, wound healing, and other activities common to all vertebrates, Brockes notes. He believes that the salamander, unlike vertebrates that are incapable of regeneration, underwent evolutionary changes that modified these common pathways “to obtain a regenerative outcome” (Integr. Comp. Biol., DOI: 10.1093/icb/icq022).
One example of the salamander’s repurposing of a shared biochemical pathway involves three-finger proteins. Many different animals utilize these proteins—which extend three fingerlike loops from a hydrophobic core—for purposes as diverse as enhancing sperm motility and instilling snake venom with toxins. But one of the three-finger proteins in salamanders is Prod 1, the protein that interacts with nAG to help orchestrate the regenerative process, Brockes says.
Further complicating the attempt to unravel regeneration is the fact that these capabilities change over the lifetime of a single organism. A tadpole, for example, can generally replace a missing tail or limb but loses this ability after its metamorphosis into a frog. Likewise, higher vertebrates such as mammals can regenerate much better during their embryonic and fetal stages than after they have become adults, Muneoka says.
At least some of this loss in mammalian regenerative capacity might be because of a change in environment after birth. “During embryonic stages, we live in a very different environment than we do as adults,” Levin says. “One of the major issues is that it’s aqueous” inside the womb. pH and oxygen exposure in the womb also differ from postnatal conditions.
Levin and his collaborator David Kaplan, a Tufts biomedical engineer, are now testing whether a replicated amniotic environment can promote regeneration in adult mammals. Kaplan has already developed a small, cylindrical “regenerative sleeve” that can be filled with an aqueous solution and fastened onto the stump of a rat’s amputated limb. The sleeve is fitted with a variety of ports and electrical connections so the researchers can sample and alter the container’s chemical contents and also control ion flows and voltage gradients that might affect regeneration.
Bioelectric signaling—which an organism detects through means including voltage-sensing domains on proteins, changes in electrophoretic transfer of neurotransmitters, and movement of calcium ions—is in fact crucial for this process, Levin says (Semin. Cell Dev. Biol. 2009, 20, 543). Changes in the electric potential of cells near the wound control differentiation and proliferation of cells in the blastema. Electrical signaling by the epithelium that covers the wound also provides a directional cue for growth so that, for instance, a replacement arm grows outward rather than into the animal’s trunk.
Unlike nonregenerative organisms, salamanders that lose a limb maintain “a very strong ‘current of injury’ for weeks at the wound edge” while the new limb grows, Levin explains. The current is created by the movement of sodium, chloride, and potassium ions through the wound epithelium. Researchers have demonstrated that applying a particular electric field to an amputated stump allows normally nonregenerative animals such as adult frogs to start regrowing limbs, he says.
Levin’s team used a different, molecular genetics technique to regenerate tadpole tails—complex appendages that contain muscles and spinal cord. First, the researchers determined that regeneration in tadpoles requires the expression of V-ATPase, an enzyme that pumps protons out of cells at the amputation site. This pumping action alters membrane voltage at the wound and also creates a long-range electric field that promotes nerve growth into the site, Levin says.
Next, the researchers turned to tadpoles that had reached a stage of development when they are normally unable to regrow a tail. Levin’s group showed that V-ATPase is ineffective during this phase. In a striking experiment, the researchers nevertheless induced these tadpoles to regenerate complete tails by inserting proton pumps from yeast into the tadpoles’ cell membranes (Development 2007, 134, 1323). The work showed that “any convenient electrogenic protein can be used for regenerative medicine approaches, not just the ones that are natively expressed in the host,” Levin says.
Levin and Kaplan’s teams are drawing on these previous findings as they work out the optimal conditions for regrowing an amputated rat limb with the help of their new regenerative sleeve. After attaching the device to the stump of a rat’s limb, the researchers hope to create a regenerative current at the stump’s surface by adjusting the ionic composition of the solution inside the sleeve and by adding drugs that open or close ion channels in the membranes of the cells at the wound site.
The sleeve will offer some additional benefits. The aqueous environment it provides will prevent the scarring that normally develops in a mammalian wound exposed to air. The researchers might also use it to bathe the wound with scar-reducing compounds, immune-modulating drugs, and more traditional growth factors, Levin says.
Clearly, many details must be worked out before success with tadpoles and rats can be extended to people. In the meantime, researchers are realizing they need to take a closer look at human patients.
The conventional medical response to an amputation in humans has been to seal the wound site to prevent complications, rather than to investigate the site’s growth potential, Muneoka explains. As a result, “we don’t know very much about what happens when you amputate a human limb,” he says. “We don’t know what the different stem cells are doing or what the bone is doing, or the muscles or the nerves. And there are literally hundreds of cell types there. So there’s a lot that we have to characterize before we can try to manipulate the wound to cause bone regrowth or muscle regrowth.”
Given that a limb is so complex, it’s amazing that any organism can replace one on its own. “It’s the equivalent of forming a baby out of an embryo,” Muneoka says. In fact, regeneration is even harder, “because you’re working with adult tissue that is going backward in time,” he adds. “After you get the injury site organized to a certain point, the re-creation of the structure is largely a reiteration of development.” But that new structure has to connect up with the existing, adult body parts, Muneoka says. “For the most part, that interface has been ignored by researchers, and it’s as critical as being able to regenerate the new structure.”