Taming Stem Cells

It's a massive understatement to say that stem cells have picky tastes. These choosy micrometer-sized underpinnings of life have long maintained an air of mystery about their predilections, initially refusing to rest on anything but a comfortable bed of other cells and rebuffing any food except undefined cocktails of proteins and other biochemicals.

Choosy tendencies hold true for all flavors of stem cells: from the embryonic ones that can morph or "differentiate" into any one of the 200 cell types that make up the human body–an aptitude called pluripotency–to adult stem cells, which are limited to differentiating into a selection of cells within a particular tissue type, like brain or blood.

Given the right surroundings, stem cells will divide endlessly without dying, a characteristic often referred to as self-renewal. But finding the right conditions is complicated. Every type of stem cell, from blood stem cells to embryonic ones, has its own preferences. These preferences differ for human and mouse stem cells of exactly the same type. For example, a protein called BMP keeps mouse embryonic stem cells in a self-renewing state, whereas it does exactly the opposite in humans, inciting our embryonic cells to differentiate.

Widespread reliance on undefined growth conditions not only has been a nightmare for getting reproducible stem cell data but has also limited the therapeutic potential of these cells. Trying to cure degenerative disease by transplanting healthy stem cells into patients is one thing. Growing them for transplant with an entourage of mouse cells, as has been the norm in the past, is another. Mouse pathogens or mouse biomolecules that induce systemic rejection of the stem cells might be passed to the patient. To boot, embryonic stem cells grown in undefined conditions can differentiate into a heterogeneous mix of cells–patients would certainly not want heart cells accidentally transplanted into their brains.

But all this is changing. Researchers are teasing out the exact chemical preferences of stem cells, designing synthetic three-dimensional homes for them, and learning to steer the differentiation of an embryonic stem cell into specific cell types using small molecules. Chemists are also discovering new small molecules that achieve a form of cellular alchemy: the reprogramming of normal cells into stem cells.

"The body has a limited repertoire of chemical compounds to keep stem cells self-renewing and pluripotent," says Nadia Rosenthal, a stem cell researcher at the European Molecular Biology Laboratory in Monterotondo, Italy. "But once [stem cells are] outside the body, small molecules provide the opportunity to mimic and enhance what's going on inside the body."

To wholly tame stem cells, researchers must eventually figure out what gives rise to their privileged place in the cell hierarchy. In other words, they must answer this question: If every cell's destiny is provided for by the same genome, why do some cells get extra powers like eternal life and the ability to shift into a myriad of other types, while other cells must be content to live out a single fate?

At first, stem cells were tiny black boxes characterized solely by specific proteins found on their cell surface. Then a bewildering array of internal signaling pathways that control self-renewal or differentiation, many linked to the cell cycle or to cancer, began to emerge.

It turns out that the fate of embryonic stem cells lies with proteins in the nucleus. There, genes that code for proteins in the self-renewal and differentiation pathways are marked by a pattern of methylation and acetylation on the histone proteins around which genomic DNA is wound. The pattern of these so-called epigenetic modifications directs the transcription of genes (C&EN, July 17, 2006, page 13). Configurations of methylation and acetylation instruct nearby genes for self-renewal and differentiation to be turned on or off.

Chemical methods are helping researchers understand epigenetic modifications. For example, mass spectrometry is proving to be effective in characterizing unique patterns of modifications that reflect a biological directive about stem-cell-specific gene expression.

Small molecules may also prove useful for studying stem cell epigenetics. For example, 5-azacytidine, which causes the demethylation of histones, is used to differentiate mouse precursor cells into bone and could help researchers study the epigenetic mechanism of this process. Some researchers are screening small-molecule libraries to look for new ways to control the enzymes that catalyze these epigenetic modifications. Others are trying to use small molecules to sort out the behavior of proteins that recruit the enzymes that install epigenetic methyl marks on histones.

But what's really key to realizing the potential of stem cells is connecting the dots between a stem cell's epigenetic control panel–which controls the cell's destiny and is located deep within the nucleus–and the extracellular components that can manipulate a stem cell's fate. One step toward this goal is defining the chemistry of the media in which the cells are grown.

Researchers have been trying to chemically define stem cell media since James Thomson and his colleagues at the University of Madison, Wisconsin, isolated and grew the first human embryonic stem cell line in 1998. They initially grew their cells on a bed of mouse "feeder layer" cells and fed them a nourishing broth. Researchers eventually found ways to grow the precious cells without the bed of mouse cells. But as recently as 2005, one still needed to expose the growth medium to mouse feeder cells, to "condition" it with an unknown mix of exported mouse proteins before it could be used to coax embryonic stem cells to stay self-renewing.

"This type of conditioned culture is called feeder-free, which is misleading because you are still dependent on the mouse cells," says Mark Levenstein, a biochemist at WiCell Research Institute, in Madison. Using ideas pulled from in-house proteomics studies and extensive literature searches, Levenstein, Tenneille E. Ludwig, and 10 colleagues reported the first derivation of embryonic stem cells in completely defined, mouse-free conditions in early 2006. "As you can imagine, we had far too many variables to cope with initially," Ludwig says. "It was tedious. It took four years."

There are now about 10 recipes for chemically defined media to derive embryonic stem cells. They contain a variety of basic ingredients such as vitamins, trace metals, antioxidants, salt to control osmolarity, and buffer to control pH. The toil and the trophy was clarifying the precise protein and chemical factors that yield self-renewing embryonic stem cells and using only human-derived or pure, recombinant proteins to do so.

One common ingredient that "coaxes cells to self-renewal" is a protein called basic fibroblast growth factor (bFGF), Levenstein says. In collaboration with Lloyd M. Smith, an analytical chemist at UW Madison, Levenstein is now studying how this extracellular protein initiates intracellular signals to trigger and maintain self-renewal. He pins it on "exquisite interactions at the membrane," where two bFGF molecules complex with two receptors and a glycoprotein cofactor to pass the message across the cell wall.

Other researchers are studying what happens inside the cell as a consequence of these sorts of membrane interactions with proteins in the media. Joshua J. Coon, another chemist at UW Madison, is using mass spectrometry to determine which proteins get phosphorylated inside the cell in response to media ingredients outside the cell. Phosphorylation events act like a middle man between the membrane, where media components trigger differentiation signals, and the nucleus, where epigenetic modifications clinch the deal, Coon says.

Other chemists are hunting for small molecules that can be placed in media to direct everything from self-renewal of embryonic stem cells to differentiation of adult ones. Peter A. Schultz, Sheng Ding, and their colleagues at Scripps Research Institute recently found a molecule that can maintain self-renewal in mouse embryonic stem cells. The pyrimidine derivative, called SC1 or pluripotin, acts dually in the cell. It indirectly activates a protein implicated in self-renewal, and it inhibits proteins known to inhibit self-renewal.

Ding and Schultz also have discovered a myriad of small molecules that initiate differentiation of embryonic and adult stem cells into cardiac cells, as well as molecules that lead to neurons, germ cells, or bone. They've even used small molecules to convince stem cells to proliferate, which is essential if stem cells are to be manufactured in large amounts for therapeutic use or for biochemical characterization.

In addition to the chemical sustenance stem cells need to self-renew or differentiate, they also need a home, and mimicking their natural home is another hot area of stem cell research.

For more than a century, scientists have been culturing cells of all types in the 2-D space of a petri dish. The physiological niche occupied by stem cells, however, is a tight 3-D network of sugar and protein called the extracellular matrix. If physiology wasn't convincing enough, geneticists have found that the genes expressed by ordinary cells in a 2-D matrix are different from those by cells in a 3-D one.

Shuguang Zhang, a biomedical engineer at Massachusetts Institute of Technology, and many others argue that it's important to move away from cell-culture technology "that predates the last century." But how? Zhang likes to quote his MIT colleague David E. Housman, who once said, "What we need is a 3-D culture system–something between a petri dish and a mouse."

Researchers are doing their best to re-create elements of a stem cell's typical environment. "In its niche, a stem cell is going to be influenced by neighboring cells, by the extracellular matrix, and by secreted factors," explains Laura L. Kiessling, a biochemist at UW Madison. "We try to create synthetic niches where you present molecules to the stem cell to make it think that it is adhering to a certain kind of extracellular matrix or to another cell."

Within the synthetic and biomimetic scaffolds being developed to provide stem cells with an imitation of their 3-D niche, researchers conjugate fragments of proteins a stem cell might normally be exposed to. The idea is to provide the stem cells with multivalent signals from their normal niche to make them flourish in a synthetic one.

One source of inspiration is human extracellular matrix proteins. Another source is Matrigel, the undefined extracellular matrix harvested from mouse cancer cells that has predominantly been used to house stem cells. Researchers know the major protein elements of Matrigel, such as collagen and fibronectin, and they have grown cells on purified human versions of these proteins. But the cost of these purified human versions is enormous, and their efficacy is lower than with the Matrigel cocktail.

Besides being inexpensive, an ideal stem cell matrix would be one in which the scaffold would degrade if you wanted to implant the cells as therapeutics or remain intact if the goal was long-term housing.

One scaffold being developed by researchers is based on self-assembled monolayers of peptides. Zhang's version decorates self-assembling peptides with functional motifs that appeal to stem cells. The peptides form ordered nanoscale fibers that mimic the extracellular matrix environment in which stem cells normally reside, and he's successfully grown neural and bone stem cells therein. Kiessling is also developing self-assembling peptides to house embryonic stem cells.

The benefit, Zhang says, of a peptide-based matrix is that it could be degraded by proteases, an advantage when stem cell transplantation therapies come to fruition. The challenge is making sure this degradation doesn't happen too soon.

Other researchers are using entirely synthetic polymers for long-term housing of stem cells. For example, the matrix devised by chemical engineer David V. Schaffer and his colleague Kevin Healy at the University of California, Berkeley, is made of an outer network of polyacrylamide, copolymerized with a polyethylene glycol network inside. The polyethylene glycol units are tipped with peptides or chemical signals appealing to a stem cell.

"The polyacrylamide provides the mechanics," Schaffer explains. "By changing its cross-link density, you can change the stiffness of the material. It turns out that the mechanical modulus is an important signal that the cells respond to."

Indeed, using a polyacrylamide gel matrix, Dennis E. Discher, a chemical engineer at the University of Pennsylvania, recently showed that the differentiation of adult mouse stem cells exposed to the same chemical media was determined by the mechanical properties of the matrix (Cell 2006, 126, 677). It's an idea that makes intuitive sense: Bone cells normally differentiate in a more rigid microenvironment, whereas blood cells would logically prefer a soft, liquid one. For neurons, the tension requirements are somewhere in between.

"Instead of cells just sitting there like a droplet in which molecules diffuse in and out, cells actively feel their environment and respond to it," Discher says.

Despite major advances to demarcate media and find new chemicals to manipulate stem cell fate, stem cell tastes are still not completely defined. For example, to derive adult blood stem cells from embryonic stem cells, "we're using everything but the kitchen sink," says George Q. Daley, a stem cell researcher at Harvard University.

As researchers get better at making stem cells feel at home outside their normal environment, they are also widening opportunities to study these mystical cells in culture and to provoke them to act in ways that defy textbook truths.

One such truth holds that organ- or tissue-specific stem cells are restricted to differentiating into cell types of the tissue in which they reside. In other words, a brain stem cell can't make heart cells–beating cells in your brain would make for a massive headache. Furthermore, fully differentiated cells, such as those on your skin, don't have the ability to step back developmentally to become a stem cell.

In August 2005, two Harvard stem cell researchers set the field abuzz when they showed that a normal human skin cell could be reprogrammed to become an embryonic stem cell (Science 2005, 309, 1369). Cell biologists Kevin Eggan, Douglas A. Melton, and their colleagues took human embryonic cells and, using polyethylene glycol, fused them into a human fibroblast. The resulting cells behaved like embryonic stem cells, even though they had twice the genetic material.

"The paper got everybody pretty excited," Melton says. Because deriving embryonic stem cell lines from embryos is so contentious, the ability to reprogram normal cells into embryonic stem cells would sidestep many people's ethical concerns.

Scientists continue to chase the reprogramming dream. Last year, Japanese researchers announced a four-protein cocktail that turned fibroblast cells into pluripotent cells with embryonic stem cell features (Cell 2006, 126, 663). The four proteins are all transcription factors that are normally involved in embryonic development or pluripotency. They are also factors associated with carcinogenesis, adding further weight to the tantalizing theory that tumors originate from malignant stem cells (see page 28).

Chemists have also played the game of cellular reprogramming by screening for small molecules that induce mature cells to take a step back in primitiveness. Schultz found a small molecule that morphs muscle cells into a more primitive cell called a myoblast. Then Ding and Shultz identified reversine, which morphs myoblasts into an even more primitive progenitor cell that can then differentiate into bone and fat cells.

"We are now looking at multiple systems where you take a lineage-committed cell and induce it to a more multipotent cell and then convert that to a cell type that's normally disallowed," Schultz says. "If one can reprogram cells, then you don't have to use embryonic stem cells, which avoids problems associated with stem cell source, compatibility, ethical issues, and cell or gene therapy," he says.

Researchers in many disciplines are now plunging into stem cell science. These new converts are using RNA interference to dissect self-renewal, leveraging bioinformatics to identify protein interaction networks that define pluripotency, and providing new innovations such as microchips on which stem cells can be differentiated. At the same time, stem cell biologists are identifying new sources of stem cells. Earlier this month, researchers announced the discovery of stem cells in amniotic fluid (Nat. Biotechnol., DOI:10.1038/nbt1274). Slightly more mature than embryonic stem cells, amniotic stem cells are stillable to differentiate into myriad cell types.

Amniotic stem cells may sidestep some of the contentious ethical issues involved in embryonic stem cell research because amniotic stem cells are isolated without destroying an embryo. Many researchers, however, including the amniotic stem cell study's author, do not believe amniotic stem cells are a substitute for embryonic stem cells. Even with a continuous flux of promising stem cell discoveries, many believe a lot of groundwork remains before stem cells achieve the therapeutic potential often touted about them.

For chemists like Ding, one lure of stem cell science is its open playing field. "Stem cell science is really in its infancy," Ding says. "It's unlike well-developed fields where you can do a lot of hypothesis-driven studies because of existing knowledge. In stem cells, there's not much known." One can start with an unbiased discovery-based approach to screen for molecules with a desired function that can be used to study biological mechanism in stem cells, he says. The challenge is breaking in. "To work with stem cells requires a lot of training and expertise," says Ding, who has been working with stem cells for several years. "In early days, we made mistakes and learned. These days we are a lot more sophisticated." And in the process, his group derived its own chemically defined media.

Because doing stem cell science "requires a very substantial investment," Ding suggests that chemists collaborate with biologists.

But the benefits of diving into the field are huge. "I see chemistry having a huge impact on stem cell biology," Daley says. "It's arguable that the first productive stem-cell-based therapies may be small molecules-drugs that act on endogenous stem cells. I think that's what we'll see in the near term, perhaps long before we'll see cell-based therapies."

Schultz agrees. "Small molecules are certainly a complementary approach to gene therapy," he says. "But using small molecules as drugs is far more precedented than using genetic approaches."

"The sad truth is, at the moment, we've really only got one working stem cell therapy, and that's using adult blood stem cells, and it's been around for years," says Graham C. Parker, a stem cell biologist at Wayne State University. Despite rhetoric in the popular press, the clinical utility of stem cells remains limited, something chemists may be able to change, he says.

Parker recently wrote an opinion piece titled "Stem Cells Need Chemical Solutions" in the Australian Journal of Chemistry. He points out that chemists can design new materials for matrices, find new small molecules that control the fate of stem cells, and provide a means to deliver stem cells to the right location in patients. As Parker puts it, chemists are "frightfully clever" at making everything from molecules to matrices. The "definite opportunity" is there for the taking.