(Image caption: During development, nerve cells (shown in blue, green, violet and yellow) extend their axons to target leg muscles. If the EphA4 receptors of the growing nerve cells no longer encounter freely accessible ephrins, the axons of many nerve cells (violet) are no longer able to find their partner cells. Credit: © MPI of Neurobiology / Gatto)

Neurons in a forest of signposts

Our ability to move relies on the correct formation of connections between different nerve cells and between nerve and muscle cells. Growing axons of nerve cells are guided to their targets by signposts expressed on the surface of other cells. Very prominent are “do not enter” signs that push axons away. Cell culture studies suggest that protein-cutting enzymes (proteases) remove these signs as soon as they are recognized by the growing axons. In this way, the “bond of recognition” between the axon and the sign is quickly broken, and the axons are more easily guided in a new direction. Scientists from the Max Planck Institute of Neurobiology in Martinsried and the Institut de Recherches Cliniques de Montréal have now shown that proteases indeed control the navigation of growing axons. However, contrary to the current belief, they do so by regulating the number of existing signs. Without proteases, the signposts would be masked and the axons would grow in the wrong direction. These findings clarified how cells form connections during development and may also improve our understanding of neurodegenerative diseases such as amyotrophic lateral sclerosis (ALS).

The human brain consists of about 100 billion nerve cells. During embryonic development each of these cells connects with other cells by means of a long extension, known as axon. Some axons need to navigate long distances through the body to find their correct targets, for example from the spinal cord down to the foot. Only if all these connections are correctly established we can perform basic and fine-tuned movements, such as walking or playing the piano.

It is therefore essential that each nerve cell finds its correct target. But how does an axon navigate and find the appropriate partner cells among billions of other possible targets? “We have now identified a few dozen guidance molecules and their receptors that help axons orient themselves,” says Rüdiger Klein, Director at the Max Planck Institute of Neurobiology. “However, these few receptor-guidance molecule pairs need to control a very large number of navigational decisions. Therefore, there must be some mechanisms to amplify and modulate the effects of these protein pairs”.

Cutting for speed?

Over the last decade, Rüdiger Klein and his team have been studying how nerve cells find their way during development. They are focusing on “do not enter” signs, e.g. ephrin guidance molecules and their Eph receptors. Ephrins and Eph receptors, being present on almost all cell surfaces: on axons as well as on cells in the surrounding tissues, help the growing axons to explore their surroundings and locate their partner cells.

As an axon travels through the body, it docks again and again to other cells via the ephrin/Eph system. This triggers cellular processes, in one or both cells, that eventually cause the connection to be severed and the cells to repel each other, preventing the axon to grow in the wrong direction. It has been hypothesised that this cellular repulsion is accelerated by proteases. Proteases are enzymes that cut Eph receptors and/or ephrins, thus by severing the Eph/ephrin bond between two opposing cells they might expedite the repulsion process. “In this way, proteases could contribute to changes in the guidance process – but this has not yet been experimentally proven.” says Rüdiger Klein.

Not faster, but better

To address this question, the neurobiologists studied how proteases affect the rate of cellular repulsion controlled by EphA4 receptors and ephrins. “Although the experiments in cell culture initially appeared to confirm the theory, we discovered something quite different in living organisms,” states Rüdiger Klein. Contrary to expectations, cellular repulsion proceeded with undiminished accuracy in animals whose axons expressed EphA4 receptors resistant to protease severing. On the other hand, in animals whose axons and muscles expressed EphA4 receptors resistant to protease cutting many axons grew in the wrong direction. Because no cutting occurred, more and more functional EphA4 receptors accumulated on cell surfaces of the leg tissues. This accumulation caused EphA4 receptors to bind to the ephrins on the same cell surface, a phenomena termed as “masking”. Consequently, the ephrins could no longer act as “do not enter” signs for the growing axons. Thus, axons, being no longer repelled, are misguided in a “no entry zone” and are unable to find their correct targets.

These results show that the cleavage of Eph receptors by proteases does not, as expected, accelerate the repulsion reaction. Instead, it regulates the number of functioning receptors and indirectly the number of available ephrins on cells, where they serve as navigational aids. If the balance is disrupted, growing axons are misdirected.

This is an important finding, as EphA4 receptors perform essential functions during the development of neural networks in the brain and in the spinal cord. They are also involved in neurodegenerative diseases such as amyotrophic lateral sclerosis (ALS). In the absence of EphA4 receptors, ALS manifests itself later and develops more slowly in a number of animal models. “It’s possible that the number of EphA4 receptors is kept low by the regulatory activity of proteases,” Rüdiger Klein reflects. “This could provide a way to exert a positive influence on the course of ALS.”

Scientific evidence does not support the brain game claims

The Stanford Center for Longevity joined today with the Max Planck Institute for Human Development in issuing a statement skeptical about the effectiveness of so-called “brain game” products. Signing the document were 69 scholars, including six from Stanford and cognitive psychologists and neuroscientists from around the world.

Laura Carstensen, a Stanford psychology professor and the director of the Center for Longevity, said as baby boomers enter their golden years, commercial companies are all too often promising quick fixes for cognition problems through products that are unlikely to produce broad improvements in everyday functioning.

"It is customary for advertising to highlight the benefits and overstate potential advantages of their products," she said. "But in the case of brain games, companies also assert that the products are based on solid scientific evidence developed by cognitive scientists and neuroscientists. So we felt compelled to issue a statement directly to the public."

One problem is that while brain games may target very specific cognitive abilities, there is very little evidence that improvements transfer to more complex skills that really matter, like thinking, problem solving and planning, according to the scholars.

While it is true that the human mind is malleable throughout a lifetime, improvement on a single task – like playing computer-based brain games – does not imply a general, all-around and deeper improvement in cognition beyond performing better on just a particular game.

"Often, the cited research is only tangentially related to the scientific claims of the company, and to the games they sell," said Carstensen, the Fairleigh S. Dickinson, Jr. Professor in Public Policy.

Agreeing with this view were the experts who signed the Stanford-Planck consensus statement, which reads in part:

"We object to the claim that brain games offer consumers a scientifically grounded avenue to reduce or reverse cognitive decline when there is no compelling scientific evidence to date that they do. … The promise of a magic bullet detracts from the best evidence to date, which is that cognitive health in old age reflects the long-term effects of healthy, engaged lifestyles."

Activity and cognition

As the researchers point out, the time spent on computer games takes away from other activities like reading, socializing, gardening and exercising that may benefit cognitive functions.

"When researchers follow people across their lives, they find that those who live cognitively active, socially connected lives and maintain healthy lifestyles are less likely to suffer debilitating illness and early cognitive decline," as the statement describes it.

"In psychology," the scientists note, "it is good scientific practice to combine information provided by many tasks to generate an overall index representing a given ability."

The same standards should be applied to the brain game industry, the experts maintain. But this has not been the case, they add.

"To date, there is little evidence that playing brain games improves underlying broad cognitive abilities, or that it enables one to better navigate a complex realm of everyday life," the participants state.

One reason is the so-called “file drawer effect,” which refers to the practice of researchers filing away studies with negative outcomes. For example, brain game studies proclaiming even modest positive results are more likely to be published, cited and publicized than ones that do not produce those affirming results.

The road ahead

In the statement, Carstensen and her fellow scientists offer recommendations for how people should view older adult life and issues like brain games:

• Legitimate research on brain games needs to be replicated and confirmed scientifically across multiple studies in different settings.
• Physical exercise is beneficial to both general and cognitive health.
• No studies have shown that brain games prevent diseases like Alzheimer’s or other forms of dementia.
• Brain games are not like “one shot” vaccines – the gains won’t last long after the end of the activity.
• People can cultivate their cognitive powers by leading physically active, intellectually challenging and socially engaged lives.

The Stanford Center on Longevity’s mission is to redesign long life. The center studies the nature and development of the human life span, looking for innovative ways to use science and technology to solve the problems of people over 50 by improving the wellbeing of people of all ages.

Scientists engineer toxin-secreting stem cells to treat brain tumors

Harvard Stem Cell Institute scientists at Massachusetts General Hospital have devised a new way to use stem cells in the fight against brain cancer. A team led by neuroscientist Khalid Shah, MS, PhD, who recently demonstrated the value of stem cells loaded with cancer-killing herpes viruses, now has a way to genetically engineer stem cells so that they can produce and secrete tumor-killing toxins.

In the AlphaMed Press journal STEM CELLS, Shah’s team shows how the toxin-secreting stem cells can be used to eradicate cancer cells remaining in mouse brains after their main tumor has been removed. The stem cells are placed at the site encapsulated in a biodegradable gel. This method solves the delivery issue that probably led to the failure of recent clinical trials aimed at delivering purified cancer-killing toxins into patients’ brains. Shah and his team are currently pursuing FDA approval to bring this and other stem cell approaches developed by them to clinical trials.

“Cancer-killing toxins have been used with great success in a variety of blood cancers, but they don’t work as well in solid tumors because the cancers aren’t as accessible and the toxins have a short half-life,” said Shah, who directs the Molecular Neurotherapy and Imaging Lab at Massachusetts General Hospital and Harvard Medical School.

“A few years ago we recognized that stem cells could be used to continuously deliver these therapeutic toxins to tumors in the brain, but first we needed to genetically engineer stem cells that could resist being killed themselves by the toxins,” he said. “Now, we have toxin-resistant stem cells that can make and release cancer-killing drugs.”

Cytotoxins are deadly to all cells, but since the late 1990s, researchers have been able to tag toxins in such a way that they only enter cancer cells with specific surface molecules; making it possible to get a toxin into a cancer cell without posing a risk to normal cells. Once inside of a cell, the toxin disrupts the cell’s ability to make proteins and, within days, the cell starts to die.

Shah’s stem cells escape this fate because they are made with a mutation that doesn’t allow the toxin to act inside the cell.  The toxin-resistant stem cells also have an extra bit of genetic code that allows them to make and secrete the toxins. Any cancer cells that these toxins encounter do not have this natural defense and therefore die. Shah and his team induced toxin resistance in human neural stem cells and subsequently engineered them to produce targeted toxins.

“We tested these stem cells in a clinically relevant mouse model of brain cancer, where you resect the tumors and then implant the stem cells encapsulated in a gel into the resection cavity,” Shah said. “After doing all of the molecular analysis and imaging to track the inhibition of protein synthesis within brain tumors, we do see the toxins kill the cancer cells and eventually prolonging the survival in animal models of resected brain tumors.”

Shah next plans to rationally combine the toxin-secreting stem cells with a number of different therapeutic stem cells developed by his team to further enhance their positive results in mouse models of glioblastoma, the most common brain tumor in human adults. Shah predicts that he will bring these therapies into clinical trials within the next five years.

(Figure 1: A magnified image of a mouse brain showing memory cells (red) that can be turned ‘on’ and ‘off’ using light delivered by a fiber optic cable (black). Credit: © Susumu Tonegawa)

Memories get the emotional switch

Memories of experiences are encoded in the brain along with contextual and emotional information such as where the experience took place and whether it was positive or negative. This allows for the formation of memory associations that might assist in survival. Just how this positive and negative encoding occurs, however, has remained unclear.

Susumu Tonegawa and colleagues from the RIKEN–MIT Center for Neural Circuit Genetics have now discovered that neurons in the hippocampus region of the brain can be artificially switched to encode memories as either positive or negative regardless of the original experience.

Tonegawa’s research team used genetic techniques to mark neurons in the dorsal dentate gyrus region of the hippocampus and the basolateral complex of the amygdala (BLA) in male mice. Memories are encoded in both these regions as specific groups of activated cells called ‘engrams’, but each region encodes the memory in slightly different ways: the BLA encodes positive and negative memory ‘valence’, while the dorsal dentate gyrus encodes contextual information such as emotion.

The genetic labeling, which involved using a light-sensitive ion channel called channelrhodopsin, was activated by the formation of either a positive memory, in this case exposure to females, or a negative memory associated with a foot shock. The cells that expressed this channel could be subsequently activated by exposure to light (Fig. 1); doing so induced aversive responses in mice that had experienced foot shocks, and appetitive responses in those that had experienced female interactions.

The researchers then used light to activate the hippocampal or BLA neurons that had been labeled during the formation of a positive memory while exposing the mice to foot shocks. The next time the animals were tested, light activation of those hippocampal neurons that had initially induced appetitive responses instead led the mice to exhibit aversive responses. However, BLA neurons could not be switched in this way, indicating that only neurons in the hippocampus have plasticity in their encoding of positive or negative memories.

The valence of hippocampal neurons, the researchers found, could be switched from both good to bad and bad to good using this technique, with the switch attributed to a change in the strength of connections between the hippocampal and BLA neurons of each engram.

The findings provide new insight into how memories can be altered after they are formed. The possibility of inducing similar changes to memory valence in humans could also offer hope of a treatment for those suffering from conditions such as post-traumatic stress disorder.

Reminiscing can help boost mental performance

To solve a mental puzzle, the brain’s executive control network for externally focused, goal-oriented thinking must activate, while the network for internally directed thinking like daydreaming must be turned down to avoid interference – or so we thought.

New research led by Cornell University neuroscientist Nathan Spreng shows for the first time that engaging brain areas linked to so-called “off-task” mental activities (such as mind-wandering and reminiscing) can actually boost performance on some challenging mental tasks. The results advance our understanding of how externally and internally focused neural networks interact to facilitate complex thought, the authors say.

“The prevailing view is that activating brain regions referred to as the default network impairs performance on attention-demanding tasks because this network is associated with behaviors such as mind-wandering,” said Spreng. “Our study is the first to demonstrate the opposite – that engaging the default network can also improve performance.”

There are plenty of neuroimaging studies showing that default network activation interferes with complex mental tasks – but in most, Spreng explained, the mental processes associated with default network conflict with task goals. If you start thinking about what you did last weekend while taking notes during a lecture, for example, your note-taking and ability to keep up will suffer.

Spreng and his team developed a new approach in which off-task processes such as reminiscing can support rather than conflict with the aims of the experimental task. Their novel task, “famous faces n-back,” tests whether accessing long-term memory about famous people, which typically engages default network brain regions, can support short-term memory performance, which typically engages executive control regions.

While undergoing brain scanning, 36 young adults viewed sets of famous and anonymous faces in sequence and were asked to identify whether the current face matched the one presented two faces back. The team found participants were faster and more accurate when matching famous faces than when matching anonymous faces and that this better short-term memory performance was associated with greater activity in the default network. The results show that activity in the default brain regions can support performance on goal-directed tasks when task demands align with processes supported by the default network, the authors say.

“Outside the laboratory, pursuing goals involves processing information filled with personal meaning – knowledge about past experiences, motivations, future plans and social context,” Spreng said. “Our study suggests that the default network and executive control networks dynamically interact to facilitate an ongoing dialogue between the pursuit of external goals and internal meaning.”

(Image caption: Neurons with the Unc5-receptor send their axons in a cell culture in all directions. The processes avoid the parallel orientated stripes containing the FLRT3-protein (red). Credit: ©Seiradake et al, Neuron 2014)

Navigation for nerve cells

During brain development, the precursors of nerve cells sometimes have to migrate long distances from their place of origin to their destination. In this process, proteins, such as FLRTs (pronounced “flirts”), act as guide molecules. Researchers at the Max Planck Institute of Neurobiology in Martinsried, together with colleagues at the Universities of Oxford and Frankfurt have now discovered that FLRT proteins on the surface of progenitor cells can induce repellent and attractant signals depending on its binding partner. The scientists used X-ray crystallography to reveal the structural bases for both FLRT-mediated adhesion and repulsion. They applied this knowledge to elucidate how these opposed signals control cellular migration. Which signal predominates depends on the particular type of cell migration. The results further show that FLRTs also exert attractant and repellent effects in the walls of blood vessels and therefore control the development of other tissue types as well.

Pyramidal cells are the central nerve cells in the cerebral cortex. During embryonic development, the precursors of pyramidal cells follow the paths of glial cell axons to migrate from their original location to the surface of the cerebral cortex. As soon as they reach their intended layer, they develop into mature pyramidal cells and interlink to form a functional network. Pyramidal cells also spread to a limited extent within these layers, though the importance of such tangential migration is still poorly understood.

This migration of precursor pyramidal cells is controlled by FLRTs (fibronectin-leucine-rich transmembrane proteins) located on the cell surface. According to the researchers at the Max Planck Institute in Martinsried, FLRTs and the Unc5 receptor form a group of guidance proteins with opposing effects on cell migration. On one hand, they act as a repellent. This is the case when a FLRT molecule binds to an Unc5 receptor on the surface of a progenitor cell. “In this way, as the precursor cell migrates radially, it receives a signal to continue migrating at an adjusted speed to not move prematurely into outer layers,” explains Rüdiger Klein from the Max Planck Institute of Neurobiology.

However, if two identical FLRT molecules bind to each other, this triggers an adhesive signal. The scientists’ results show that pyramidal cells are guided in this manner as they spread tangentially, without affecting their ability to find their target layer. Thus, there are proteins with attractant and repellent effects located on the surface of precursor pyramidal cells. “By integrating these opposing signals, cells can navigate through brain tissue. During radial migration FLRTs induce repulsion; during tangential dispersion FLRT attraction dominates,” says Klein.

In their study the scientists also investigated whether the mechanisms of FLRT adhesion and repulsion are present in other cell types. Their findings show that cells in the walls of blood vessels in the retina and the umbilical cord are also controlled by a combination of attractant and repellent signals modulated by FLRT and Unc5 proteins.

New insight on why people with Down syndrome invariably develop Alzheimer’s disease

A new study by researchers at Sanford-Burnham Medical Research Institute reveals the process that leads to changes in the brains of individuals with Down syndrome—the same changes that cause dementia in Alzheimer’s patients. The findings, published in Cell Reports, have important implications for the development of treatments that can prevent damage in neuronal connectivity and brain function in Down syndrome and other neurodevelopmental and neurodegenerative conditions, including Alzheimer’s disease.

Down syndrome is characterized by an extra copy of chromosome 21 and is the most common chromosome abnormality in humans. It occurs in about one per 700 babies in the United States, and is associated with a mild to moderate intellectual disability. Down syndrome is also associated with an increased risk of developing Alzheimer’s disease. By the age of 40, nearly 100 percent of all individuals with Down syndrome develop the changes in the brain associated with Alzheimer’s disease, and approximately 25 percent of people with Down syndrome show signs of Alzheimer’s-type dementia by the age of 35, and 75 percent by age 65. As the life expectancy for people with Down syndrome has increased dramatically in recent years—from 25 in 1983 to 60 today—research aimed to understand the cause of conditions that affect their quality of life are essential.

"Our goal is to understand how the extra copy of chromosome 21 and its genes cause individuals with Down syndrome to have a greatly increased risk of developing dementia," said Huaxi Hu, Ph.D., professor in the Degenerative Diseases Program at Sanford-Burnham and senior author of the paper. "Our new study reveals how a protein called sorting nexin 27 (SNX27) regulates the generation of beta-amyloid—the main component of the detrimental amyloid plaques found in the brains of people with Down syndrome and Alzheimer’s. The findings are important because they explain how beta-amyloid levels are managed in these individuals."

Beta-Amyloid, Plaques and Dementia

Xu’s team found that SNX27 regulates beta-amyloid generation. Beta-amyloid is a sticky protein that’s toxic to neurons. The combination of beta-amyloid and dead neurons form clumps in the brain called plaques. Brain plaques are a pathological hallmark of Alzheimer’s disease and are implicated in the cause of the symptoms of dementia.

"We found that SNX27 reduces beta-amyloid generation through interactions with gamma-secretase—an enzyme that cleaves the beta-amyloid precursor protein to produce beta-amyloid," said Xin Wang, Ph.D., a postdoctoral fellow in Xu’s lab and first author of the publication. "When SNX27 interacts with gamma-secretase, the enzyme becomes disabled and cannot produce beta-amyloid. Lower levels of SNX27 lead to increased levels of functional gamma-secretase that in turn lead to increased levels of beta-amyloid."

SNX27’s Role in Brain Function

Previously, Xu and colleagues found that SNX27 deficient mice shared some characteristics with Down syndrome, and that humans with Down syndrome have significantly lower levels of SNX27. In the brain, SNX27 maintains certain receptors on the cell surface—receptors that are necessary for neurons to fire properly. When levels of SNX27 are reduced, neuron activity is impaired, causing problems with learning and memory. Importantly, the research team found that by adding new copies of the SNX27 gene to the brains of Down syndrome mice, they could repair the memory deficit in the mice.

The researchers went on to reveal how lower levels of SNX27 in Down syndrome are the result of an extra copy of an RNA molecule encoded by chromosome 21 called miRNA-155. miRNA-155 is a small piece of genetic material that doesn’t code for protein, but instead influences the production of SNX27.

With the current study, researchers can piece the entire process together—the extra copy of chromosome 21 causes elevated levels of miRNA-155 that in turn lead to reduced levels of SNX27. Reduced levels of SNX27 lead to an increase in the amount of active gamma-secretase causing an increase in the production of beta-amyloid and the plaques observed in affected individuals.

"We have defined a rather complex mechanism that explains how SNX27 levels indirectly lead to beta-amyloid," said Xu. "While there may be many factors that contribute to Alzheimer’s characteristics in Down syndrome, our study supports an approach of inhibiting gamma-secretase as a means to prevent the amyloid plaques in the brain found in Down syndrome and Alzheimer’s."

"Our next step is to develop and implement a screening test to identify molecules that can reduce the levels of miRNA-155 and hence restore the level of SNX27, and find molecules that can enhance the interaction between SNX27 and gamma-secretase. We are working with the Conrad Prebys Center for Chemical Genomics at Sanford-Burnham to achieve this," added Xu.

Brain simulation raises questions

What does it mean to simulate the human brain? Why is it important to do so? And is it even possible to simulate the brain separately from the body it exists in? These questions are discussed in a new paper published in the scientific journal Neuron today.

Simulating the brain means modeling it on a computer. But in real life, brains don’t exist in isolation. The brain is a complex and adaptive system that is seated within our bodies and entangled with all the other adaptive systems inside us that together make up a whole person. And the fact that the brain is a brain inside our bodies is something we can’t ignore when we attempt to simulate it realistically. Today, two Human Brain Project (HBP) researchers, Kathinka Evers, philosopher at the Centre for Research Ethics and Bioethics at Uppsala University and Yadin Dudal, neuroscientist at the Weizmann Institute of Science, publish a paper in Neuron that discusses the questions raised by brain simulations within and beyond the EU flagship project HBP.

For many scientists, understanding means being able to create a mental model that allows them to predict how a system would behave under different conditions. For the brain sciences, this type of understanding is currently only possible for a limited number of basic functions. In the article, Kathinka Evers and Yadin Dudal discuss the goal of simulation. In broad terms it has to do with understanding. But what does understanding mean in neuroscience?

As it dwells inside our bodies, the brain is always a result of what the individual has experienced up to that point. That is why, when we simulate the brain, we have to take this ‘experienced brain’ into account and try and reflect that.

According to Kathinka Evers, leader of the Ethics and Society part of the Human Brain Project, neglecting this experience would severely limit the outcome of any brain simulation. But if we are to include experience we have to simulate real-life situations.

“That is a daunting task: a large part of that experience is the brain’s interaction with the rest of the human body existing and interacting in a still larger social context”, says Kathinka Evers.

What outcome would be realistic to hope for in the Human Brain Project’s simulation? In neuroscience, computer simulations of specific systems are already in use. These simulations are a complement to other tools scientists use.

But there are some warnings to issue here. According to Kathinka Evers and Yadin Dudal, our knowledge to date is still very limited. There are many neuroscientists who think that it is too early for large scale brain simulations. Collecting the data we need for this is not an easy task. Another problem is whether we truly can understand what we are about to build. There are also technical limitations: there simply isn’t enough computing power available today.

But if we do manage to simulate the brain, would that mean we have created artificial consciousness? And can a computer be conscious at all? According to Kathinka Evers and Yadin Dudal, that depends on what consciousness is: If it is the result of certain types of organization or functions of biological matter, like the cells in the human body, then a computer can never gain consciousness. But if it is a matter of organization alone, without the need for biological matter, then the answer could be yes. But it is still a very hypothetical stance.