Tissue Engineers out for Blood

Current methods of organ transplantation are less than optimal, in the view of Joseph P. Vacanti, director of pediatric transplantation at Massachusetts General Hospital. Vacanti, who is at the forefront of tissue engineering, compares organ transplantation to ripping a building from its foundation to move it to a new location--not the safest or most efficient method. A more appropriate tactic is to have architects--or physicians and engineers--design and build from the ground up a structure suited to the new location, Vacanti said.

To date, tissue engineers have accomplished this by building scaffolds and seeding them with cells that can grow. This approach works well for producing relatively small pieces of tissue but presents a problem as engineers try to design larger tissues, such as vital organs.

Growing tissue more than 100 mm thick has been challenging because blood--and its payload of oxygen and nutrients--doesn't reach the cells in the middle of the tissue. The problem is one of both biology and geometry. A surface is required for the exchange of nutrients and oxygen in the blood. The available surface for this exchange increases with the square of the radius of the blood vessel, but the volume of tissue that must be fed increases as the cube of the radius. "Nature solves the problem by using branched systems," Vacanti said, in which tiny capillaries provide the surface area to feed a large volume of tissue.

The need for oxygen is immediate and the need for nutrients only slightly less so, but the process of blood-vessel growth, or angiogenesis, can take anywhere from three to five days. "A cell suffocates in three to five minutes," Vacanti said. "You have to provide sufficient oxygen immediately, and then you have to provide nutrition in an ongoing way. A cell cannot 'hold its breath' for three to five days."

In a session at the U.S.-Japan Symposium on Drug Delivery Systems, held last December in Lahaina, Hawaii, Vacanti and other researchers described new approaches for solving this vascularization problem.

The usual approach to stimulate angiogenesis is to add growth factors to the tissue, the most common of which is VEGF, or vascular endothelial growth factor. The problem with adding only VEGF is that it forms what look like capillaries, but they tend to leak. Mature blood vessels require a type of cell called a pericyte in addition to the endothelial cells produced with VEGF. Pericytes are musclelike cells that stabilize blood-vessel walls.

VEGF is "only one growth factor out of dozens involved in blood-vessel development," said Michael Sefton, professor of chemical engineering and applied chemistry and director of the Institute of Biomaterials & Biomedical Engineering at the University of Toronto. Sefton presented his research at the drug delivery meeting. The "backup plan" has been to add another growth factor, such as platelet-derived growth factor (PDGF) or angiopoietin-1.

Adding one or two more growth factors may not be enough. "How many do you actually need to recapitulate true blood-vessel growth? What are the ratios of those growth factors? Is there a timing effect?" Sefton asked. "As you need to have more and more growth factors in some coordinated fashion, it's not clear you can actually get more benefit relative to the complexity."

Sefton's group is exploring several strategies for providing blood vessels in three-dimensional tissue. The first approach builds on the growth factor method. Rather than delivering the protein VEGF by itself or the gene for VEGF, the researchers are genetically modifying cells to produce VEGF. They encapsulate the cells in a polymer to isolate them from the immune system.

USING THESE microcapsules, the researchers are getting what looks like blood vessels, and they even have "hints that they're functional," Sefton said. However, he would not be surprised to find that they don't work. The next step is to add a second set of modified cells to produce another growth factor. If the combination of the two factors doesn't work well, the subsequent options get very complicated, he added.

In another method that Sefton is developing, polymeric beads made of compounds dubbed "theramers," for therapeutic polymers, induce angiogenesis without the addition of growth factors. The beads are primarily implanted in the skin, where they appear to harness and accelerate the natural wound-healing process. Potential applications include improving skin grafts and healing chronic wounds in the elderly or in diabetics.

Sefton's group discovered the angiogenic properties of the beads serendipitously. They were making capsules containing insulin-producing islet cells and noticed that even when they used empty capsules, blood vessels formed. "We were puzzled by that and thought it might be related to the small amount of methacrylic acid that was part of the copolymer," Sefton said. By increasing the amount of methacrylic acid in the copolymer, they found that blood-vessel development increased in proportion to the amount of methacrylic acid.

These angiogenic beads are made of 45% methacrylic acid in a random copolymer with methyl methacrylate. Sefton suspects that the methacrylic acid is acting as a "sink for growth factors" that are produced as part of the wound-healing process. The growth factors are absorbed, stabilized, and released over a long period of time. The vessels persist for several months.

Sefton has a "fuzzier" understanding of another effect of the beads. The blood vessels generated are not being resorbed. Usually the capillary density increases while the wound is healing but returns to normal after the wound has healed. "We think we've driven the macrophages, the key cells in wound healing, to express a pro-angiogenic phenotype," he said. "They're producing more pro-angiogenic factors than anti-angiogenic factors. We've disturbed that balance. That's part of the mechanism we're trying to explore and understand."

The 45% methacrylic acid represents an optimal balance between the angiogenic properties and the material properties. "We could make 65% methacrylic acid, but we didn't seem to get any better angiogenic effect," Sefton said. Plus, at the higher methacrylic acid levels, the material was more difficult to use. The angiogenic beads are being developed by a spin-off company called Rimon Therapeutics, which Sefton cofounded in 2000. The beads are currently being prepared for clinical trials.

The third method that Sefton is working on--"endothelial seeding"--creates a different type of vascular network than the first two methods. "In the first two cases, the blood vessels are sprouting from the existing vasculature," Sefton said. "In endothelial seeding, we're not using angiogenesis. We're preforming the endothelial cells in the form of pseudocapillaries, but we are left with the problem of how to connect those pseudocapillaries to the rest of the vascular system."

With endothelial seeding, they are adapting a method used to make blood-compatible linings for vascular grafts. They place endothelial cells on the surface of small modules that contain the cells they are interested in, for example, liver cells. Those modules are then packed in a larger column such as a chromatography column.

"If the endothelial cells work as they are supposed to be working, they allow blood to flow through the column, essentially creating a vascularized network," Sefton said. "We're at the proof-of-principle stage. One of the ways it could be used is by taking the column and connecting it to an artery and a vein and then allowing the blood to flow directly through the column, in one end and out the other."

The columns could potentially be used as bioartificial organs. For example, liver cells could produce a protein that's picked up by the blood as it passes through. "Whether this design is practical for that really depends on the size of the device you would need [for it] to be a clinically practical liver," Sefton said.

Vacanti also is taking a more mechanical approach to issues of vascularization. Rather than rely on growth factors to generate the blood vessels that the tissue needs to survive, his group decided to engineer every blood vessel. They generate 3-D computer models of the circulation that are converted into designs on silicon wafers by collaborator Jeffrey Borenstein at Draper Laboratory, Cambridge, Mass. Photolithography is the only technology that is precise enough for individual capillaries, Vacanti said. The designs on the silicon wafers are translated into channels in polymer films. Vascular endothelial cells line the capillary system.

The vascular channels are oriented horizontally with the channels in one layer that is connected with those in the next via "through-holes." The tissue of interest, such as liver cells, is sandwiched between the vascular layers.

"We have sort of a Dagwood sandwich or a set of pancakes that alternate the liver cells with the vascular channels," Vacanti said. "By alternating these, we can stack it--theoretically--as thick as we want."

WHEN THE NUMBER of vertical communication channels is increased, the size of the horizontal channels can be decreased and thus their number increased. However, as the density of the horizontal channels increases, they become more difficult to manufacture.

"Ideally, we want the same density of capillaries as a normal piece of tissue or liver," Vacanti said. With the multiple layers interconnected, "we can settle for a lot less because we don't need that [ideal density] to get the job done."

Other engineering technologies available for exchanging nutrients or waste products include hollow-fiber cartridges and membranes. Vacanti said his group's device has a much larger surface area available for exchange than the other technologies have, which means that devices could be much smaller and potentially wearable or implantable.

Vacanti used the kidney as an example. Currently in the U.S., about 75,000 people are waiting for kidney transplants and are being sustained by renal dialysis. "If we could use our designs and just provide better dialysis, [the dialysis device] might be wearable like a fanny pack," Vacanti said. Cells might not even be needed in a first-generation device. Subsequent generations could be improved with kidney cells, and an ultimate version could be a completely tissue-engineered kidney.

"We see this as a staged approach, because these are very difficult problems," Vacanti said. "Along the way, we may be able to improve things greatly, even though it isn't a complete living replacement."

So far, Vacanti and his coworkers have demonstrated their system in culture conditions for both liver and kidney cells. They are moving into animal studies to see if the systems are biocompatible and achieve good blood flow.

One of the issues that Vacanti has yet to resolve is whether the approach is scalable. "We've been able to successfully stack 10 layers," he said. "We don't know whether we can therefore do 100 and have it just as accurate as 10."