A Great Migration
By Elise Lamar, PhD
Illustration by Mark Miller
In vertebrates, bands of mobile cells fan out from the embryonic brain and spinal cord and then resettle in far-flung locales to form structures as diverse as bone or nerves. Paul Trainor and Paul Kulesa track their journeys in hopes of illuminating why birth defects occur when cells go off-track.
Neither scientist set out to become an expert on neural crest cells, as these inveterate migrants are known. Paul Trainor, PhD, spent his youth on the beaches of his native Australia, more interested in sports than science. As a University of Sydney undergrad he got hooked on biology only after discovering that through genetic manipulation you could produce fruit flies with too many wings or misplaced antennae. Halfway around the world, Paul Kulesa, PhD, a midwesterner, was captivated by mathematics and space exploration. After earning a bachelor’s degree in aerospace engineering at Notre Dame and a master’s in applied mathematics at the University of Southern California, he became interested in mathematical equations that could help explain complex biological patterns.
Now, after two decades of graduate school and postdoctoral training between them, both are at the Stowers Institute for Medical Research studying neural crest cells, albeit from vantage points that couldn’t be more different. Trainor, a Stowers investigator, is mainly interested in understanding the molecular intersections linking defects in the neural crest with craniofacial malformations. Kulesa, the director of Imaging, develops sophisticated imaging technology to understand how neural crest cells travel long distances and assemble various structures, such as the peripheral nervous system. Using mouse, chick or zebrafish models, both test what goes awry when mutations derail migrating neural crest cells in a developing embryo. In unique ways, each researcher’s work directly impacts a broad class of human birth defects.
One name: multiple disorders
Neural crest cells spring from the crest or dorsal ridge of the embryonic brain and spinal cord and then migrate to faraway regions of the face, heart or gut. Although neural in origin, their name is slightly misleading: They mature just as readily into bone, muscle or connective tissue as they do into a neuron. Failure of neural crest cells to either complete their journey or mature on arrival causes widely varying birth defects known collectively as neurocristopathies.
One of Trainor’s goals is to understand the molecular basis of these conditions. “Approximately one percent of all live births exhibit a minor or major congenital anomaly, and about a third of those display craniofacial abnormalities,” he says. “To have any hope of preventing these disorders you must understand how they originate.”
When Trainor arrived at Stowers, he initially focused on a birth defect called Treacher Collins syndrome (TCS). Babies born with TCS exhibit deformities of the cheek, eye or jaw, and some also suffer hearing loss or respiratory problems due to airway malformation. Scientists knew that the Tcof1 gene is mutated in association with most TCS cases. In a 2006 paper published in PNAS, the Trainor lab reported that mice with mutations in the Tcof1 gene also displayed severe facial deformities, mimicking the human condition. The group then analyzed mouse embryos using molecular markers and discovered that precursors of neural crest cells in the brain and spinal cord began dying even before crest cells destined to help build the face could start migrating.
In 2008, Trainor’s group reported in Nature Medicine that blocking a gene that promotes cell death, called p53, allowed nascent neural crest cells in Tcof1-mutant mice to survive, preventing manifestation of the animals’ craniofacial defects. This work is highly significant because, as a potential treatment option, erasing a mutation—in a whole embryo no less—will likely not be feasible. But this paper demonstrated that cells that harbor a disease-causing mutation can be rescued. “This outcome shows that we could potentially intervene and prevent defects, not just repair them,” Trainor says.
In parallel, his lab conducted a labor-intensive mouse genetic screen to discover novel genes required for normal neural crest activity. That work culminated in a 2011 Genesis paper identifying ten mutations that underlie conditions as diverse as holoprosencephaly (in which the forebrain fails to partition into two hemispheres), neural tube defects, and cleft palate. More recently, in a 2013 PLoS Genetics study, they reported that mice lacking a gene called Foxc1 model a human congenital defect called syngnathia, in which children are born with fused upper and lower jawbones.
These types of genetic analyses illustrate the challenge faced by researchers interested in preventing or treating neuroscristopathies. A single name may classify them in Wikipedia, but the field work of Trainor and colleagues demonstrates that multiple genes expressed either in a moving cell or somewhere along its path underlie neural crest-based defects. “In a condition like TCS we see a failure to make enough cells at the beginning of migration, while in syngnathia, crest cells migrate properly but make bone in the wrong place,” says Trainor. “Collectively, these studies suggest that there isn’t going to be any single way to prevent these disorders.”
Which is precisely where basic research comes in. “One of our goals is to identify the genetic instruction sets needed to make a face,” says Stowers Scientific Director Robb Krumlauf, PhD, whose ground-breaking research on Hox genes defined how neural crest cells, particularly those traveling toward the head, know where they’re going. “It takes hundreds of genes and likely thousands of control steps to make structures of a face. Thus identifying pathways useful for therapeutic interventions will require putting together a complex jigsaw puzzle, piece by piece.”
In the footsteps of trailblazers
By the mid-nineteenth century biologists had observed that a cell population migrated out of the vertebrate brain and spinal cord during embryogenesis, but an appreciation for where they went was possible only after scientists discovered how to track them. Building on the tissue transplantation techniques developed earlier in the twentieth century, renowned French embryologist Nicole Le Douarin grafted portions of the embryonic quail spinal cord into chick embryos prior to neural crest migration. After the neural crest cells had reached their final destination, she mapped where they ended up. Fifteen years later, Scott Fraser, a biophysicist at the California Institute of Technology and Kulesa’s former mentor, developed techniques to microinject fluorescent vital dyes into single premigratory neural crest cells. This allowed him to analyze their lineage and dynamic behaviors in embryonic mice, chicks, or frogs using time-lapse video and confocal microscopy.
Kulesa walks in these pioneers’ footsteps: Soon after arriving at Stowers, he began working on techniques to deliver a cocktail of rainbow-colored fluorescent proteins into neural crest cells and utilize a newly emerging microscopy tool called multispectral-imaging to distinguish each color. By labeling the cell nucleus, cell membrane and cytoplasm with different colors, single neural crest cells could be more efficiently identified and tracked within the complex cellular architecture of a living chick embryo. That initial work was reported in 2005 in Biotechniques and then refined in a 2010 BMC Developmental Biology paper.
“In the past scientists could only take snapshots of cells at an early stage and then at their final destination,” says Krumlauf. “Now, at Stowers we are fortunate in having both cutting edge imaging and experts who can make movies of a cell’s life in a living organism. It’s revealing an amazing world with subtle complexities.”
Remembering their roots
Given how rapidly the imaging field is evolving, Kulesa feels the next frontier will exploit multispectral-imaging to better visualize 3-D cell dynamics and link cell behaviors with molecular data. To do so, he now pairs spectral imaging with a method called laser capture microdissection (LCM) and quantitative polymerase chain reaction (qPCR). These approaches allow researchers to excise a small number of cells at specific points during migration and analyze their gene expression profile. Using this technique, Kulesa’s team literally cut the leading cells from a pack of streaming neural crest cells in a chick embryo and found that they express unique genes associated with invasive behavior. These findings, which define factors expressed by neural crest trailblazers, were reported in Development in 2012.
Single cell identification and tracking is more accurate using multicolor cell labeling and multispectral 3D confocal imaging. Here, individual cell nuclei within the early chick brainstem and migrating neural crest cells are distinctly visualized using a cocktail of fluorescent proteins.
Since then the Kulesa team has applied that strategy to explore a suspected link between neural crest migration and metastatic melanoma. Melanoma is an invasive cancer of pigment cells called melanocytes, and vertebrate melanocytes are yet another cell type born from neural crest cells. To determine if the embryonic neural crest microenvironment alters melanoma cell invasiveness, Kulesa transplanted human melanoma cells into chick embryos and then waited to see if they would invade along neural crest migratory pathways.
As detailed in a series of papers culminating in a 2012 Pigment Cell & Melanoma Research report, not only did human metastatic melanoma cells start migrating but they followed host chick neural crest migratory pathways and reached peripheral targets. That study also revealed that melanoma cells transplanted into the chick microenvironment exploit neural crest genes to facilitate invasion. Overall, these discoveries show that when placed in a cellular neighborhood that reminds them of their birthplace, even cancer cells recognize what street they’re on.
Kulesa now wonders whether other neural crest-related cancers hijack a neural crest cell gene expression program. One of those is neuroblastoma, the most common cancer in infants. Neuroblastomas are malignancies of sympathetic nervous system structures, such as adrenal glands.
“In neuroblastoma, neural crest cells fail to mature but instead remain like a multipotent progenitor cell,” says Kulesa, referring to the fact that some cancer cells exhibit properties of normal, “good” stem cells. His lab is now testing whether human neuroblastoma cells respond to microenvironmental cues like the melanoma cells did, with a goal of thwarting them. “We are excited about creating in vivo systems in which we can transplant, visualize and profile human neuroblastoma cells and identify molecular signals that may reprogram them to become less destructive.”
A move in a different direction
Trainor remains focused on the molecular dissection of cranial neural crest cells but has extended his analyses to a neural crest population headed the opposite way: cells that march toward the gut. At journey’s end, those cells mature into neurons that innervate the colon, comprising the enteric nervous system. Mutations that slow or block their progress cause a congenital defect called Hirschsprung disease. Children born with Hirschsprung lack innervation of variable regions of the gastrointestinal tract and typically require surgery to repair bowel function.
Last year, Trainor reported in Human Molecular Genetics that mice mutants in a specific pair of genes expressed in migrating crest cells mimicked human Hirschsprung disease. One of those genes was Tcof1 (also a culprit in TCS) and the other was Pax3, a prime suspect in neurocristopathies affecting ear, face, and heart development. In mice expressing about half the normal allotment of both genes, fewer neural crest cells migrated out of the brain and spinal cord, and many died on the way to the gut. Significantly, mice expressing suboptimal levels of just
one of those genes did not exhibit equivalent defects.
According to Trainor, these findings confirm the lesson that genome sequencing has taught us over the last decade: that many genetically based diseases emerge due to the malfunction of more than one gene. “Some neurocristopathies are multigenic,” he says, noting this finding’s biological and clinical significance. “As therapeutic interventions become available, proper diagnosis of these conditions will require genetic testing for combinations of mutations.”
Trainor’s interest in neurocristopathies is not confined to the lab. He helps write fact sheets for the National Organization for Rare Disorders to update TCS patient families and physicians on basic research and provide information about screening. And he draws inspiration from interacting personally with the TCS community. “It’s the medical problems that keep us motivated,” he says. “It makes us hopeful that somewhere down the track we will figure out how to prevent or at least ameliorate these conditions.”