Neuroscience

How can the immune system be reprogrammed once it goes on the attack against its own body? EPFL scientists retrained T-cells involved in type I diabetes, a common autoimmune disease. Using a modified protein, they precisely targeted the white blood cells (T-lymphocytes, or T-cells) that were attacking pancreatic cells and causing the disease. When tested on laboratory mice, the therapy eliminated all signs of the pathology. This same method could be a very promising avenue for treating multiple sclerosis as well. The scientists have just launched a start-up company, Anokion SA, on the Lausanne campus, and are planning to conduct clinical trials within the next two years. Their discovery has been published in the journal PNAS (Proceedings of the National Academy of Science).

To retrain the rebellious white blood cells, the researchers began with a relatively simple observation: every day, thousands of our cells die. Each time a cell bites the dust, it sends out a message to the immune system. If the death is caused by trauma, such as an inflammation, the message tends to stimulate white blood cells to become aggressive. But if the cell dies a programmed death at the end of its natural life cycle, it sends out a soothing signal.

In the human body there is a type of cell that dies off en masse, on the order of 200 billion per day – red blood cells. Each of these programmed deaths sends a soothing message to the immune system. The scientists took advantage of this situation, and attached the pancreatic protein targeted by T-cells in type I diabetes to red blood cells.

"Our idea was that by associating the protein under attack to a soothing event, like the programmed death of red blood cells, we would reduce the intensity of the immune response," explains Jeffrey Hubbell, co-author of the study. To do this, the researchers had to do some clever bioengineering and equip the protein with a tiny, molecular scale hook, that is able to attach itself to a red blood cell. Billions of these were manufactured and then simply injected into the body.

Complete eradication of diabetes symptoms

As these billions of red blood cells died their programmed death, they released two signals: the artificially attached pancreatic protein, and the soothing signal. The association of these two elements, like Pavlov’s dog, who associates the ringing of a bell with a good or bad outcome, essentially retrained the T lymphocytes to stop attacking the pancreatic cells. “It was a total success. We were able to eliminate the immune response in type I diabetes in mice,” explains Hubbell.

Minimizing risks and side effects

Co-author Stephan Kontos adds that the great advantage of this approach is its extreme precision. “Our method carries very little risk and shouldn’t introduce significant side effects, in the sense that we are not targeting the entire immune system, but just the specific kind of T-cells involved in the disease.”

The scientists are planning to conduct clinical trials in 2014, at the earliest. To demonstrate the potential of their method, they plan to first test applications that would counteract the immune response to a drug known for its effectiveness against gout. “We chose to begin with this application before we tackled diabetes or multiple sclerosis, since we knew and were in control of all the parameters,” explains Hubbell.

Currently, the researchers are also testing the potential of this method in treating multiple sclerosis. In this disease, T-cells destroy myelin cells, which form a protective sheath around nerve fibers. They are also studying the potential of their method with another kind of white blood cell, B-lymphocytes, that are involved in many other autoimmune diseases.

Autistic-like behaviors can be partially remedied by normalizing excessive levels of protein synthesis in the brain, a team of researchers has found in a study of laboratory mice. The findings, which appear in the latest issue of Nature, provide a pathway to the creation of pharmaceuticals aimed at treating autism spectrum disorders (ASD) that are associated with diminished social interaction skills, impaired communication ability, and repetitive behaviors.

"The creation of a drug to address ASD will be difficult, but these findings offer a potential route to get there," said Eric Klann, a professor at NYU’s Center for Neural Science and the study’s senior author. "We have not only confirmed a common link for several such disorders, but also have raised the exciting possibility that the behavioral afflictions of those individuals with ASD can be addressed."

The study’s other co-authors included researchers from the University of California, San Francisco (UCSF) and three French institutions: Aix-Marseille Universite’; Institut National de la Santé et de la Recherche Médicale (INSERM); and Le Centre National de la Recherche Scientifique (CNRS).

The researchers focused on the EIF4E gene, whose mutation is associated with autism. The mutation causing autism was proposed to increase levels of the eIF4E, the protein product of EIF4E, and lead to exaggerated protein synthesis. Excessive eIF4E signaling and exaggerated protein synthesis also may play a role in a range of neurological disorders, including fragile X syndrome (FXS).

In their experiments, the researchers examined mice with increased levels of eIF4E. They found that these mice had exaggerated levels of protein synthesis in the brain and exhibited behaviors similar to those found in autistic individuals—repetitive behaviors, such as repeatedly burying marbles, diminished social interaction (the study monitored interactions with other mice), and behavioral inflexibility (the afflicted mice were unable to navigate mazes that had been slightly altered from ones they had previously solved). The researchers also found altered communication between neurons in brain regions linked to the abnormal behaviors.

To remedy to these autistic-like behaviors, the researchers then tested a drug, 4EGI-1, which diminishes protein synthesis induced by the increased levels of eIF4E. Through this drug, they hypothesized that they could return the afflicted mice’s protein production to normal levels, and, with it, reverse autistic-like behaviors.

The subsequent experiments confirmed their hypotheses. The mice were less likely to engage in repetitive behaviors, more likely to interact with other mice, and were successful in navigating mazes that differed from those they previously solved, thereby showing enhanced behavioral flexibility. Additional investigation revealed that these changes were likely due to a reduction in protein production—the levels of newly synthesized proteins in the brains of these mice were similar to those of normal mice.

"These findings highlight an invaluable mouse model for autism in which many drugs that target eIF4E can be tested," added co-author Davide Ruggero, an associate professor at UCSF’s School of Medicine and Department of Urology. "These include novel compounds that we are developing to target eIF4E hyperactivation in cancer that may also be potentially therapeutic for autistic patients."

The causes of learning problems associated with an inherited brain tumor disorder are much more complex than scientists had anticipated, researchers at Washington University School of Medicine in St. Louis report.

The disorder, neurofibromatosis 1 (NF1), is among the most common inherited pediatric brain cancer syndromes. Children born with NF1 can develop low-grade brain tumors, but their most common problems are learning and attention difficulties.

“While one of our top priorities is halting tumor growth, it’s also important to ensure that these children don’t have the added challenges of living with learning and behavioral problems,” says senior author David H. Gutmann, MD, PhD, the Donald O. Schnuck Family Professor of Neurology. “Our results suggest that learning problems in these patients can be caused by more than one factor. Successful treatment depends on identifying the biological reasons underlying the problems seen in individual patients with NF1.”

The study appears online in Annals of Neurology.

According to Gutmann, who is director of the Washington University Neurofibromatosis Center, scientists are divided when considering the basis for NF1-associated learning abnormalities and attention deficits.

Mutations in the Nf1 gene can disrupt normal regulation of an important protein called RAS in the hippocampus, a brain region critical for learning. Initial work from other investigators had shown that increased RAS activity due to defective Nf1 gene function impairs memory and attention in some Nf1 mouse models.

However, earlier studies by Gutmann and collaborator David F. Wozniak, PhD, research professor in psychiatry, showed that a mutation in the Nf1 gene lowers levels of dopamine, a neurotransmitter involved in attention. In this Nf1 mouse model, Gutmann and his colleagues found that the branches of dopamine-producing nerve cells were unusually short, limiting their ability to make and distribute dopamine and leading to reduced attention in those mice.

The new research suggests that both sides may be right.

In the latest study, postdoctoral fellow Kelly Diggs-Andrews, PhD, found that the branches of dopamine-producing nerve cells that normally extend into the hippocampus are shorter in Nf1 mice. As a result, dopamine levels are lower in that part of the brain.

Charles F. Zorumski, MD, the Samuel B. Guze Professor and head of the Department of Psychiatry, showed that the low dopamine levels disrupts the ability of nerve cells in the hippocampus to modulate the way they communicate with each other. These communication adjustments are a primary way the brain creates memories.

Researchers then found that giving Nf1 mice L-DOPA, which increases dopamine levels, restored their nerve cell branch lengths to normal and corrected the hippocampal communication defect. L-DOPA also eliminated the memory and learning deficits in these mice.

“These results and the earlier findings suggest that there are a variety of ways that NF1 may cause cognitive dysfunction in people,” Gutmann says. “Some may have problems caused only by increased RAS function, others may be having problems attributable to reduced dopamine, and a third group may be having difficulties caused by both RAS and dopamine abnormalities.”

To customize patient therapy, Gutmann and his colleagues are now working to develop ways to quantify the contributions of dopamine and RAS to NF1-related learning disorders.