Map Drawn of Molecular Network Underlying ‘Stemness’

One of the key challenges of stem cell biology is defining pluripotency, the ability of stem cells to turn into many cell types. The lab of Stuart Orkin, a Howard Hughes investigator and the David G. Nathan professor of pediatrics at Dana–Farber Cancer Institute, has created a map of protein networks built around a key protein in mouse embryonic stem cells. The resulting map, published Nov. 16 in Nature, gives an overall view of an interconnected network of proteins that gives stem cells their chameleon-like properties.

Image courtesy of Stuart Orkin

Geography of a stem cell. A map of protein interactions in mouse embryonic stem cells shows certain proteins acting as hubs in a tight-knit network. Proteins with red labels were used as “bait” for fishing out other proteins that physically associate with them in cells.

Orkin’s team, led by research fellow Jianlong Wang, focused on the protein Nanog, which previous studies showed to be necessary for maintaining pluripotency in embryonic stem cells. Orkin said that while studies had demonstrated that Nanog and other proteins play a role in stem cells, little was known about the overall context in which the proteins function. “We wanted to do a comprehensive analysis in a way that had not been done before,” he said.

Their approach was to attach a short peptide “hook” to Nanog expressed in mouse embryonic stem cells, allowing the researchers to fish Nanog out along with any proteins attached to it. Using mass spectrometry, they identified more than a dozen proteins that associate with Nanog, only some of which were known or suspected to help maintain stem cells. Using RNA interference, the team found that inhibiting several of these proteins in embryonic stem cells led to a loss of stem cell characteristics.

With this same approach, the team then developed a network by mapping interactions with some of the Nanog partners identified in the first screen, as well as with another stem cell marker protein, Rex1. Orkin said that the resulting map looks something like an airline route map; certain proteins, such as Nanog, form hubs that have many connections to other nodes, while some are outliers that only connect to one or two other nodes.

One of the most striking aspects of this network, Orkin said, is that most of the proteins interact with or regulate the expression of other proteins in the model. “It’s a very tight network; it’s like a self-contained unit,” Orkin said. “It supports the notion that this is a specific regulatory module in the cell.” It also suggests that loss of one or more components can quickly cause the cell to lose its pluripotency.

Orkin said that the map can be used to begin to study interactions between proteins in the network more closely and guide experiments aimed at reprogramming other cell types into stem cells. Similar mapping studies may also reveal important similarities and differences between mouse and human embryonic stem cells. Ultimately, Orkin hopes, this systems-level view will be useful for answering the overarching question of stem cell biology: How does a cell get its ID and how can you reprogram it from one cell type to another?

—Courtney Humphries

Role of Bone-growth Protein Described in Fracture Healing

Since their discovery 40 years ago, bone morphogenetic proteins (BMPs) have been studied for their ability to induce bone formation. BMP2, the most abundant of the group, is a crucial component in osteogenesis; its absence causes embryonic lethality. BMP2 is also required for fracture healing and, due to this function, is widely used in treating spinal fusion. Yet the role the protein plays in healing fractures and its position in the healing pathway is not well understood. Vicki Rosen, HSDM professor of developmental biology; Kunikazu Tsuji, a research fellow in Rosen’s lab; and colleagues have now reported for the first time a distinct function for BMP2 in the early stages of the healing process. Their study appears in the December issue of Nature Genetics.

In the investigation, the authors created transgenic mice with a BMP2 allele under the control of a temporally controlled promoter element, Prx1, which is shut off prior to the onset of skeletal growth during normal limb development. This design allowed the researchers to switch off BMP2 expression specifically in the limbs and study fracture healing while avoiding the mortality associated with global loss of BMP2.

The authors showed that BMP2 is dispensable for skeletal patterning and early bone development, at least in the limb, by demonstrating that the skeletons of mice lacking BMP2 were near normal at birth. Yet a loss of BMP2 function impaired proper maturation of bones and caused an inability to initiate fracture repair. At the cellular level, incomplete maturation was reflected in a decreased production of type II collagen, a component of the extracellular matrix needed to strengthen the bone. This eventually caused the mice to develop spontaneous fractures, specifically in the growing limbs, which lacked BMP2 function. These fractures failed to heal long after similar fractures were repaired in mice that retained normal function of BMP2.

The primary reason for the lack of healing appeared to be an absence of new bone formation at the fracture site. A detailed histological examination of the site showed a marked absence of repair tissue and subsequent bone formation. The researchers discovered that cells at the fracture site were primed for repair through their expression of surface receptors that when bound by BMP2 lead the cells to differentiate into those vital for repair.

The authors are now assessing the contribution of other key bone morphogenetic proteins, BMP4 and 7, in the fracture healing process.

—Amita Joshi

Fatigue Found Behind Critical Medical Errors

Doctors’ mistakes are no small matter. In 2000, the Institute of Medicine reported that between 48,000 and 98,000 deaths occur each year due to medical errors. To address the problem, in 2003, the Accreditation Council for Graduate Medical Education (ACGME) capped the number and scope of extended-duration shifts residents can work: no more than 30 consecutive hours and no more than twice a week. The new guidelines were intended to minimize the risk of fatigue-related medical errors and improve safety for interns and patients alike, but a study in the December PLoS Medicine suggests that they are not sufficiently stringent to achieve that goal. It is one of a series of related studies from senior author Charles Czeisler, the Frank Baldino Jr., PhD, professor of sleep medicine at HMS and Brigham and Women’s Hospital and director of the HMS Division of Sleep Medicine (see Focus, Sept. 15, 2006).

The study examined the effects that a 24-hour workday has on intern performance. Interns were 3.5 times more likely to report at least one fatigue-related medical error during months in which they worked between one and four extended-duration shifts compared to months in which they worked none. For months in which they worked five or more such shifts, these odds jumped to 7.5 times more likely. The ACGME guidelines allow for nine extended-duration shifts over the course of a month.

Although studies have shown that overworked doctors and interns make more mistakes than well-rested ones, Laura Barger, the study’s lead author and a research fellow in medicine, Czeisler, and their colleagues looked at a much broader cohort than had previously been examined. The group conducted a web-based survey of 2,737 interns from a wide range of specialties across the United States, who submitted monthly reports detailing their work hours, sleeping habits, and the medical errors they had made.

Extended-duration work shifts—those running at least 24 hours—increased the likelihood that an intern would make significant medical errors, some of which had an adverse effect on patients. Patient fatalities also increased with the number of extended-duration shifts, but the total number of these was comparatively small. The correlation between long shifts and more mistakes was seen in both men and women, regardless of age.

These long work hours were also linked to nodding off on the job. During months with five or more extended-duration shifts, interns were significantly more likely to fall asleep during surgery, while talking to or examining patients, during rounds or during lectures and seminars. The authors compare the adverse effects of 24 hours without sleep on neurobehavioral performance to a blood alcohol concentration of 0.10 percent. They point out that the amount of sleep reported by most interns is less than what has been shown to be necessary for satisfactory performance of the basic cognitive tasks necessary for learning.

—Jeneen Interlandi

top