Saturday, January 12, 2008

Next Steps for Stem Cells

New methods to reprogram adult cells could create novel models of disease.

Reprogramming cells: Scientists at George Daley’s lab at Children’s Hospital Boston are using new methods to reprogram adult cells to develop stem-cell lines from patients. These cells can then be used as models to study disease. Daley, who is affiliated with the Harvard Stem Cell Institute, is shown here (right) with postdoctoral student In-Hyun Park.
Credit: Jon Chase/Harvard News Office

Searching the brain of an Alzheimer's patient for clues into the origin of the disease is like trying to find the cause of a plane crash in the wrecked aftermath. However, a recent breakthrough in stem-cell research could generate new cellular models that allow scientists to study disease with unprecedented accuracy, from its earliest inception to a cell's final biochemical demise.

Last November, two groups of scientists announced that they had independently achieved one of the stem-cell field's biggest goals: the ability to reprogram adult cells into embryonic-like stem cells without the need for human embryos. (See "Stem Cells without the Embryos.") The findings garnered extensive media attention, largely because the new method obviated the need for human embryos, a major ethical minefield that has stymied research.

But scientists at stem-cell labs around the world are excited for another reason. The technique creates cells that are genetically matched to an individual, meaning that it's now possible to create novel cell models that capture all the genetic quirks of complex diseases. "Being able to have human cells with human disease in a dish accessible for testing is a real boon to technology and to science," says Evan Snyder, director of the Stem Cells and Regeneration Program at the Burnham Institute, in La Jolla, CA.

While animal models exist for many human diseases, they typically only incorporate certain aspects of the disease and can't capture the complexity of human biology. In addition, some disorders known to have a significant genetic component, such as autism, have proved difficult to model in animals.

To reprogram cells, scientists from Wisconsin and Japan independently engineered skin cells to express four different genes known to be expressed in the developing embryo. For reasons not yet clear to scientists, this treatment turns back the developmental clock. The resulting cells are pluripotent, meaning that they can develop into any type of cell in the body, and they can apparently divide indefinitely in their undifferentiated state. The first two published studies on the new technique reprogrammed cells from a skin-cell line, while a third study, published last month, generated stem cells from the skin biopsy of a healthy volunteer.

No one has yet generated cell lines from a patient, although scientists have been talking about doing so for years. Previously, the only way to make such models for complex genetic diseases was through human therapeutic cloning, also known as nuclear transfer, which is fraught with technical and ethical issues and has not yet been achieved. (See "Stem Cells Reborn" and "The Real Stem Cell Hope.") "Assuming that these procedures are as easy to do as it seems, it's definitely more tractable than nuclear transfer," says Snyder. His own lab is trying to generate such models, as is "probably everyone else you could call on your rolodex," he says.

To generate a disease-specific cell model, scientists would take some cells from a patient with a particular disease and revert them to an embryonic state. The cells would then be prodded to develop into the tissue type damaged in that disease, such as dopamine neurons in Parkinson's disease or blood cells in sickle-cell anemia. By comparing the differentiation process in cells derived from healthy and diseased people, scientists could observe how that disease unfolds at a cellular level. They could also use the cells to test drugs that might correct those biochemical abnormalities. "We want to use these cells to ask and answer questions that can't be asked and answered any other way," says M. William Lensch, a research scientist at the Harvard Stem Cell Institute and Children's Hospital Boston.

The relative simplicity of the approach--and the fact that it can be supported by federal funding--means that many more scientists are likely to attempt reprogramming than cloning. (In 2001, President Bush limited federal funding for embryonic stem-cell research to embryonic stem-cell lines already in existence.) According to Story Landis, chair of the Stem Cell Task Force at the National Institutes of Health, in Bethesda, MD, the funding agency has already announced two programs to fund reprogramming research and would welcome applications to derive cell lines from patients.

While no one has yet announced that he or she has derived a disease-specific cell model, George Daley's lab at Harvard may be in the lead. Last month, he and his team published a paper in Nature showing that they can reprogram cells from a skin biopsy from a healthy person, and they are already trying to repeat the feat with tissue from patients. Ultimately, they are interested in developing models of sickle-cell anemia and Fanconi anemia, a hereditary disease in which the bone marrow doesn't produce enough new cells to replenish the blood.

For example, patients with Fanconi anemia often suffer from skeletal problems, and their cells show an impaired ability to repair DNA. "We don't have any idea why kids with DNA repair defect would get a blood disease, and why they sometimes get these bone abnormalities," says Lensch, who works with Daley. But with stem-cell lines developed from a patient, "we could push the cells to develop into bone and blood, and try to learn about the links between the two."

Such models could also help resolve long-held debates about specific diseases, such as Alzheimer's. By differentiating reprogrammed cells from Alzheimer's patients into neurons and comparing them with neurons derived from healthy embryonic stem cells or with cells with mutations that mimic a rare, hereditary form of the disease, scientists will be able to determine how much of Alzheimer's is due to the environment versus genes, as well as how similar the sporadic form of the disease is to the hereditary form. (Most drugs on the market for Alzheimer's were developed using models that mimic the hereditary form of the disease and have shown limited efficacy in patients.) "This is a whole new world of investigation," says Lawrence Goldstein, a neuroscientist at the University of California, San Diego, whose lab is about to begin collecting skin cells from Alzheimer's patients.

Despite the excitement, Lensch and others caution against abandoning other embryonic stem-cell research, especially therapeutic cloning. "We're in the early stages of this research, where we're excited about the possibilities but still need to show it's both useful and representative of the disease," says Snyder. In addition, he says, embryonic stem cells and perhaps cloned stem cells will be needed as controls for future studies.

Scientists also say that it's too soon to tell how easy it will be to generate stem-cell lines from patients: the genetic variations that lead to the disease could also impact the reprogramming process. "With some genetic disease, I think it will be really difficult," says Lensch.

DNA Deletion Linked to Autism

A massive study takes a major step forward in explaining the genetics of autism.

Probing autism: Shown here is a DNA microarray. Each dot represents a specific sequence of donor DNA bound to DNA probes on the microarray. Bright green indicates an area where a chunk of DNA has been deleted, red indicates a duplication of DNA, and yellow indicates an area with no duplications or deletions.
Credit: Agilent Technologies

A specific structural variation on chromosome 16 dramatically boosts the risk of autism, according to a study published today in the New England Journal of Medicine. The finding--one of the most significant to date--permits the development of new diagnostic tests to identify children at risk, and could ultimately point to specific biochemical pathways to target in drug development.

"This is one of the single largest [influences] and most frequent genetic causes for autism identified so far," says Bai-Lin Wu,director of the Genetics Diagnostic Laboratory at Children's Hospital Boston and one of the senior authors on the study.

Autism spectrum disorder--or autism, as it is commonly called--refers to a group of developmental disabilities with wide-ranging language, social, and behavioral symptoms. The disorder is known to have a strong genetic influence, with up to 90 percent of cases thought to have a genetic component. However, because the disorder is linked to a combination of genetic variations, each playing a minor role, identifying specific genetic triggers has been difficult. Now new microarray technologies, which allow scientists to screen a million or more genetic variations in thousands of patients, are enabling the much larger studies needed to pinpoint these triggers.

In the new paper, scientists say that they used microarrays to scour the DNA of more than 2,000 individuals with autism. They found that deletion or duplication of approximately 500 of the same DNA letters on chromosome 16 was strongly linked to autism, accounting for about one percent of cases. "While that doesn't sound like a huge number, the fact that these people carry the identical spontaneous deletion or duplication would be incredibly unlikely to happen by chance," says Mark Daly, a geneticist at Massachusetts General Hospital's (MGH) Center for Human Genetic Research, in Boston, and at the Whitehead Institute, in Cambridge, and one of the study's senior authors.

The results were independently identified by three different groups--at MGH; Children's Hospital Boston; and deCODE Genetics, in Iceland--that are studying three different populations, giving added weight to the work.

The findings build on previous reports that autism is linked to genetic deletions or duplications that arise spontaneously, rather than being passed down through generations. In almost all cases, parents of the affected people did not carry the chromosome 16 variation.

One of the most immediate clinical benefits of the research will be the development of inexpensive diagnostic tests. "Because the variation occurs so frequently, you could directly test for the presence or absence of a duplication or deletion as part of standardized genetic testing for autism," says James Gusella, a neurogeneticist at Harvard Medical School, in Boston, who participated in the research. For example, children who show developmental delays but are too young to undergo clinical autism testing could be screened for this variation, allowing parents and doctors to prescribe intervention for those who test positive. "We will be able to find at-risk children early on so that language and behavior problems can be treated much earlier," says Yiping Shen, director of research and development at Children's Hospital's Genetics Diagnostic Laboratory, who was also involved in the work.

Such testing could also predict if parents with one autistic child are at greater risk of having another; if their child's autism is linked to a spontaneous variation, they are at no greater risk than the general population. Researchers at Children's Hospital, which provides genetic testing to families, are already developing a clinical diagnostic test.

Scientists are also trying to pinpoint the specific gene or genes within this section of DNA that underlie the increased risk. Daly and his collaborators plan to sequence this region of the genome in another group of people with autism, in search of single-letter mutations that might disrupt the function of specific genes. "Genetics provides us with the only opportunity to gain insight into the biological mechanisms that underlie autism," says Daly. "We can look at individual gene discovery as a small first step in the overall path to develop treatments."

Previous studies have identified autism risk genes. However, these studies have focused on people with genetic disorders that often co-occur with autism, such as Fragile-X syndrome, complicating the role those genes play in the disorder. "Up until now, we haven't had the capacity to look at a single gene that is associated with pure autism," says Gusella.

The findings could point to additional spots in the human genome to search for autism risk genes. The variation on chromosome 16 lies within a genetic "hot spot," an area that is predisposed to undergoing structural duplications due to the architecture of the DNA, says Evan Eichler, a geneticist at the University of Washington in Seattle, who wrote an editorial accompanying the paper. "Every time we produce gametes, there's a finite probability of this region to duplicate," he says. In addition, the region has a high concentration of genes that are rapidly evolving in humans. While the significance of that finding is not yet clear, it may explain autism's status as a relatively young disease.

Creating a Web of Worlds

Metaplace builds a different architecture for virtual worlds.

Web worlds: Metaplace wants to enable its users to build virtual worlds, such as the one shown above, that could exist anywhere on the Web.
Credit: Metaplace

Many of today's virtual worlds have been influenced by science-fiction writer Neal Stephenson's vision of a Metaverse, described in his novel Snow Crash. Stephenson's Metaverse swallows up the Web and Internet into a 3-D space that users navigate with avatars. But Raph Koster, president of Metaplace, based in San Diego, and former creative lead for the influential game Ultima Online, believes that the Metaverse should look decidedly different.

Metaplace is building a system that's designed to treat virtual worlds like other content on the Web, Koster says. A virtual world, he explains, is simply a place where multiple users can interact with one another or with objects built for that world. Metaplace is designed to allow users to host these places on the Web the way they might host embedded video, and to build them the way they might build other content on the Web.

"We think virtual worlds are just a new medium," Koster says. "That means that like other media--like pictures, audio, and video--virtual worlds are eventually going to start being ubiquitous on all sorts of Web pages."

With Metaplace, designers can build worlds using a markup language, style sheets, modules, and a scripting language. Every world acts like a Web server, Koster says, and every object in a world has a URL. What this means for users of these worlds is that they can move seamlessly from the rest of the Web into the virtual world and back again, he says. A user can browse to any object in a Metaplace world from outside, and every object can be linked to the rest of the Web and exchange information with Web services. With this architecture, Koster says, he plans for users to be able to build worlds with games as simple as a two-dimensional Tetris game, or as complex as the World of Warcraft, a massive, multiplayer, online role-playing game. Users might also build widgets, such as a virtual weatherman who could deliver the latest news from, or a Coke machine that gives them a real-world coupon whenever they drink a virtual Coke. Koster says that users should be able to stage up a basic world with chat functionality and a map within about five minutes.

Koster envisions users coming to a Metaplace world by clicking on a link in a Web page. That link launches a page where the user finds herself inside a world, perhaps using a default avatar, but no log-in or registration is immediately required. "They don't make you log in to play a YouTube video," Koster points out.

The Metaplace client is basically a Flash application, he says, and, consequently, is available to nearly everyone who uses the Internet. Currently, Metaplace does not allow users to build 3-D worlds, but Koster says that he expects Flash to add 3-D capabilities in the near future. The client will work anywhere on the Web, and Koster adds that he hopes to see user-generated clients built for mobile devices such as iPhones.

Cory Ondrejka, who was a cofounder of Linden Lab, makers of Second Life, and is now a visiting professor at the University of Southern California, in Annenberg, has been testing Metaplace's system in its early stages. He says that Metaplace is betting that the Web will continue to evolve increasingly better capabilities for real-time interaction, such as 3-D capabilities for Flash, that will allow it to continue improving its system. At heart, Ondrejka says, Metaplace is a lightweight protocol for lightweight communication through the Web, and one of the ways that he sees designers using Metaplace is as a way of letting users experience each other's presence online. "Anything that causes the two of us to know we're both on the Web together makes the Web a better place," Ondrejka says. "A big part of what makes interaction in virtual worlds so compelling compared to the Web is the fact that we both know we're there. It isn't the same as leaving bread crumbs on a blog to show that you were there."

Ondrejka says that at this point, Metaplace gives users far simpler capabilities than those in worlds like Second Life. "But simple doesn't mean bad," he notes. "Simple can mean approachable." While Ondrejka says that Second Life gives users far more design power, he also says that Metaplace could allow a great deal of flexibility: Metaplace worlds can be anywhere on the Web, or even within Second Life.

Koster doesn't intend Metaplace to only be used to design games: he says that he intends it to be used to build virtual places, and that users and designers can choose what activities will be hosted in those places. While he wants the worlds to be capable of including games, Koster says that they could also be used for training, education, and other activities.

Metaplace expects to make money by charging designers for premium hosting services. The system is currently being tested in a closed early release, but Koster expects to open the release to more people by around April. In the meantime, he says, it's possible that the company may make certain features, such as the client that allows people to play games built with Metaplace, available before then.

Guiding Light

A new fabrication technique brings us closer to optical chips.

Lighting the way: A new process for creating three-dimensional silicon structures that can manipulate and trap light could lead to all-optical integrated circuits.
Credit: Stephen Eisenmann

Getting optical signals to bend around sharp corners has remained an obstacle to developing all-optical integrated circuits and better opto-electronic devices. But now researchers have created a new process for making complex miniature waveguides that can steer optical signals in three dimensions through solid materials.

Paul Braun, a materials scientist at the University of Illinois at Urbana-Champaign, and his colleagues have demonstrated a technique that uses focused laser light to carve out intricate waveguides within photonic crystals--materials that can be used to manipulate photons in much the way that semiconductors direct electrons. Recently, there have been advances in using conventional lithography to create two-dimensional optical waveguides, says Braun. "But what's very hard to do is take light and manipulate it in 3-D."

The work will spur the development of a range of optical devices, says Steven Johnson, an applied mathematician at MIT who has carried out research on the use of photonic crystals as waveguides. A 3-D waveguide carved into photonic crystals, he says, "can be used to trap and control light, and has potential applications in everything from more-efficient lasers to optical signal processing for telecommunications or other applications," he says.

Photonic crystals can be made by packing together beads of silica. When they're packed together in a precise three-dimensional arrangement, it is possible to create what is known as a complete photonic bandgap material. This material, says Braun, will act as a perfect reflector for a particular narrow band of light--dictated by the size of the beads. "It's a perfect reflector for all angles of incidence."

If channels can be created within the material, any light entering the material via these channels will not be able to escape, except through the channels. So once in the material, it becomes possible to manipulate the light in unusual ways, such as by trapping it or bending it around very sharp corners without fear of it escaping.

A number of research groups have been working on using the materials to create 3-D optical waveguides, says Johnson. But one of the problems has been the low refractive properties of the polymer materials used, which makes them unsuitable for completely trapping light, he says.

Braun's group has gotten around this by using the polymer as a template for creating a complete photonic bandgap material out of silicon, which has a higher refractive index.

The group starts off with a stack of precisely arranged silica beads. The stack is then immersed in a light-sensitive monomer, which solidifies into a polymer when hit by pulses of focused laser light.

By carefully directing pulses of laser light, it is therefore possible to create continuous paths out of the polymer material. Then, after the researchers rinse out the remaining monomer, they fill the voids between the beads and the polymer material with a silicon-based material using a process called chemical vapor deposition. The entire structure is then bathed in hydrofluoric acid to dissolve all but the silicon.

What's left is a solid structure of silicon with a network of waveguides within it, says Braun.

In the current issue of Nature Photonics, the group reports its findings and shows that by starting off with beads that are 725 nanometers in diameter, it is possible to create waveguides for a narrow band of wavelengths in the near-infrared range. This is potentially extremely useful, since this is the range that is currently employed for most optical communications, says Braun.

The work is more of an evolution than a revolution, says Johnson. And he notes that while Braun's structures are not yet useful for making working devices, they are an important first step toward creating more complex and functional optical devices.

Mitsubishi Unveils Laser TV, 3-D Home Theater

Expected to be available by the end of the year, laser TV promises twice the color of HD.

Last night, at a Consumer Electronics Show event at the Palms Hotel, in Las Vegas, Mitsubishi gave a first look at its forthcoming line of flat-panel, high-definition displays. The company claims that the display "delivers a range of color never seen before in home entertainment." The display, called laser TV, uses laser as the light source, unlike liquid-crystal displays, which use a white backlight, and plasma displays, which use cells of charged gas to illuminate the screen. Mitsubishi representatives didn't supply a lot of details; they said only that the TV will ship to retailers later this year.

At the event, Mitsubishi showed off three 65-inch laser displays, which are currently being manufactured. (Gadget blog Engadget posted nice pictures here.) In addition, the company demonstrated how its laser TV could be used as a 3-D home theater. The company played clips from Beowulf, a football game, and U2's 3-D concert on its laser display. Viewers wore shutter glasses from RealD, a supplier of 3-D technology. Shutters on the lenses switched on and off--imperceptibly--60 times a second, synchronizing to a signal emitted from the display. (See "Hollywood's New 3-D Age.")

The basic premise behind laser TV is not entirely new. (See "Ultra-Colorful TV.") It's essentially a variant of digital light projection (DLP) technology developed at Texas Instruments. DLP chips are in most of the projectors used in business presentations, and they're found in home projection displays. A laser display is built a little differently, however. Instead of projecting light onto a screen from the front, lasers and the DLP chip are in the rear of the display, which allows it to be manufactured thinner than traditional front-projection systems.

The main difference with a laser display, however, is that it uses lasers as the light source. Usually, projection displays shine white light through a color wheel, and then it's projected onto the screen. This approach is inefficient, filtering out much of the original brightness. Laser displays use red, blue, and green lasers to directly deliver the color to the screen. Lasers not only have a brightness and color advantage over filtered white light, but they also have an advantage over light-emitting diode (LED) technology, another up-and-coming display backlight. The color produced by a laser is much more pure than that produced by an LED because the former allows for more-precise color combinations. The net result is an extremely crisp, lifelike image in which white is many times brighter than standard high-definition displays, and black is many times darker.

The laser displays at the Palms looked impressive to me, although Mitsubishi didn't show a side-by-side comparison with other displays. One of the more exciting aspects of these new displays, however, is that they use much less energy than other flat panels do, and they should quickly become less expensive than plasmas since the lasers can be mass-produced in semiconductor facilities.