|
||||||||
|
TISSUE ENGINEERING Organoids of the Third DimensionDown-to-Earth Structure Built that Mimics Thymus, Matures T Cells
This is not a coral reef but a scanning electron microscope close-up of the CellFoam matrix where the mouse stroma and human precursor cells can interact in three dimensions. Scanning electron microscopy courtesy of Cytomatrix Astonishment, everyone agrees, was not the intent—yet there was something startling about a paper in the July Nature Biotechnology by a team of Harvard-affiliated researchers at MGH unveiling a "tissue-engineered thymic organoid" that churns out human T cells. Perhaps it was the word organoid? "It is different but it didn't spring from the head of Zeus," says David Scadden, associate professor of medicine and codirector of the Partners AIDS Research Center, who along with Mark Poznansky, instructor in medicine; Richard H. Evans, research fellow in medicine; Ivona Olszak, research assistant, and others engineered the thymic organoid using a biocompatible material developed for bone grafts as a matrix for mouse–human cell coculture. "We wanted to find materials that would help recreate the organ functions on a different platform," Scadden explains. "It came from a lot of hard work from other people, and we just happened to fit together some of the different pieces in a different way." That way was three-dimensional. The Holy GrailThe extra dimension sprang from CellFoam, a synthetic material manufactured by Cytomatrix of Woburn, Mass. It is a carbon skeleton onto which tantalum, a heavy, noncorrosive metal is aerosolized at high temperature. The result is chalky, rigid foam with a spongy appearance that can be sliced, autoclaved, and popped into well boxes, ready for tissue culturing. Seeded with the stroma, or structural, cells of a mouse thymus, the matrix is plated two weeks later with hematopoietic stem cells drawn from human bone marrow. Two weeks after that, the coculture is generating mature functional human T cells, duplicating ex vivo the chief function of the thymus.The team's first thymic organoid is xenogeneic; that is, it uses mouse stroma and human stem cells. The lab is already pressing ahead on an organoid that uses human stroma and stem cells. Such a human–human organoid could be either allogeneic, from donor cells, or autologous, with the patient's own cells repackaged. The systems could open a dizzying array of possibilities, not only in the area of AIDS but also in cancer, organ transplants, and other pathologies that hinge on the body's ability to recognize "self" and "nonself."
The organoid organizers: (l to r) Richard H. Evans, Ivona Olszak, Mark Poznansky, and David Scadden. Photo by Steve Gilbert "If someone was getting an organ transplant," Scadden says, "you might want the stem cells to come from the recipient but you might want the stroma cells to come from the donor so that you could 'tolerize' these T cells so they would not reject a subsequent organ graft." An autologous organoid could be used to rebuild an individual's depleted repertoire of T cells after a bone marrow transplant, chemotherapy, or AIDS. Another organoid might be able to generate ex vivo T cells that react against a specific target, for example, HIV or a cancer strain. "Another possibility is, could this be an implantable device? There are circulating stem cells," Scadden wonders. "Could it do what the organ does by itself?" Raison d'Être"The thymus is primarily involved in T cell development so it's almost a unifunctional organ," Poznansky says. "We reconstructed the one function of the thymus and thus it's what we call an organoid. We picked a good organ to imitate because it has this one very specific function."It may have been a good choice but it was not an easy job. The lab had been struggling with a flat plate thymic system that did express a few new T cells. "We had been doing this on a two-dimensional system," says Scadden. "We had shown you could grow human fetal thymus and then onto that put stem cells and get them to undergo differentiation, but it was a much less predictable process and much less efficient." The clue that the problem might be geometrical not biochemical came from the stroma tissue itself, says Poznansky. "The observation was made that if you grow thymic stroma on a two-dimensional plate, over time it actually tries to organize itself into three-dimensional structures. It grows much more efficiently. I thought, it's naturally trying to do that. If you could give it a scaffolding on which to grow, would you enhance the effect?" CellFoam seemed promising. Under the electron microscope, the matrix looks like a coral reef with an intricate 3-D landscape of nooks, crannies, and passages. Stroma cells should be able to organize themselves inside the CellFoam matrix far more naturally than they can on a flat dish, says Evans. "One of the problems with a two-dimensional system with mono layers is that if they just sit on top, they really only have the chance to interact in one plane. But if you have it in three dimensions, the precursor cell can interact with several different cell types at any one time so it can recognize the signals it needs to become a T cell." Green ThumbAt the outset, though, getting mouse stroma to grow on the matrix was difficult. "In the lab at that time, people hadn't had much luck in growing single cells on it," Poznansky recalls. "It wasn't until we started to seed it with fragments of mouse thymus that it took. It turned out that the less we handled the tissue—just cut it up into pieces and plated it onto the scaffolding—the more efficiently it grew."Within days, the fragments disappeared from view, melting into the pores and spreading out as a tissue coating the intricate galleries of the material. Then the stem cells were plated directly on top of the matrix. Two weeks later, mature T cells were pouring out. Non-adherent cells were harvested and subjected to staining and immunological phenotyping to ensure that human, not murine, T cells were being generated. Over 70 percent of the new cells were recognized as T cells by their surface expression of CD3, of both the CD4 and CD8 subtypes. An assay for T-cell receptor excision circles showed strong de novo T cell generation rather than an expansion of T cells that might have been present in the hema-topoietic precursor population as a contaminant. There are critical hurdles ahead. The thymic organoid has to lend itself to scaling up, says Scadden, and the T cells it generates have to demonstrate "specificity of reactivity," which is the bottom line in immunology. "As a T cell matures, there's a random recombination with a T-cell receptor," he explains. "It may be the right fit for things on normal cells, which you don't want it to respond to. It's essential that those T cells be destroyed or at least be tolerized so they express the receptor but don't respond by killing the normal cell. You also don't want a lot of T cells hanging around that are unnecessary. Given that the number of possible useless combination events is so high, relative to the useful ones, you can imagine how you could crowd out the useful ones." –John Fleischman |
