Advancing Basic Science for Humanity
Recently, members of the NIH BRAIN Initiative Neuroethics Working Group (NEWG) and Multi-Council Working Group (MCWG) discussed racial inequities in neuroscience and biomedicine, COVID-19 impacts on research, and future BRAIN funding.
On August 20th, 2020 the Brain Research through Advancing Innovative Neurotechnologies (BRAIN) Initiative®’s Neuroethics Working Group (NEWG) held its eleventh meeting. The NEWG ensures that neuroethics is fully integrated into the BRAIN Initiative as new tools and technologies emerge. Meeting participants discussed how neuroethics can improve racial inequities in research and ethics related COVID impacts on BRAIN-funded research.
Dr. John Ngai, Director of the NIH Brain Initiative®, kicked off the NEWG meeting by emphasizing the importance of acknowledging and addressing racial disparities in the NIH workforce, among NIH grantees, and in research more broadly. Next, Dr. Kafui Dzirasa (Duke University) and Dr. Frances Shen (University of Minnesota) summarized recent papers on considering ancestry in brain research and the underreporting of race in neuroscience studies. The group proposed including more experts in racial disparities in neuroethics-focused meetings, fostering new research partnerships, and other efforts. There was also an update from the NIH Office of Science Policy on the recently-convened NASEM Committee on the Ethical, Legal, and Regulatory Issues Associated with Neural Chimeras and Organoids.
Next, there was a panel discussion on how COVID has impacted human subjects research. BRAIN-funded panelists Drs. Maria Franceschini (Harvard University), Leigh Hochberg (Brown University), and Sameer Sheth (Baylor College of Medicine) brought to light several ethical issues they are facing, such as balancing different risks and benefits for participants and researchers, appropriately communicating about COVID, and obligations to participants. For more details, please view the NEWG meeting summary (139KB) and archived videocast.
The next day, on August 21st, the BRAIN Initiative’s Multi-Council Working Group (MCWG) convened its seventeenth meeting. The MCWG, which maintains a coordinated effort across the NIH and oversees the long-term scientific vision of BRAIN, discussed the role of BRAIN in enhancing diversity and future BRAIN funding.
Dr. Ngai introduced the new Office of the BRAIN Director, summarized the virtual BRAIN Initiative Investigators Meeting, and highlighted four recent BRAIN-funded studies that used new tools to probe brain function across species. He also overviewed COVID-19 impacts on BRAIN research and NIH-wide efforts to develop vaccines, therapeutics, and tests for the virus. Dr. Ngai then led a conversation about the role of BRAIN in promoting diversity, equity, and inclusion in neuroscience. The group discussed many ways to enhance diversity, such as increasing training opportunities and access to new tools and technologies for those at institutions not typically funded by BRAIN.
Lastly, NIH staff presented one concept for future funding, which focuses on ramping up human brain cell census work. This concept builds upon the years of progress made thus far in non-human species under the BRAIN Initiative Cell Census Network (BICCN). To find out more, please read the MCWG meeting summary (115KB) and archived videocast.
The next NEWG and MCWG meetings will be held on January 26th and 27th, 2021, respectively.
This funding opportunity announcement supports the development of technologies, production efforts, and dissemination resources for a cell type-specific armamentarium to study brain function across species. Applications are due in February and October of 2021.
RFA-MH-20-556 Pilot resources for brain cell type-specific access and manipulation across vertebrate species (U01; clinical trial not allowed)
Molecular access to diverse brain cell types enables more detailed study of neuronal circuits. The purpose of this funding opportunity announcement (FOA) is to improve molecular access to diverse cell types through the evaluation of molecular or genetic technologies and the creation of pilot production and distribution resources for cell type-specific access and manipulation reagents for several vertebrate species. Applicants to this FOA should propose demonstration projects for reagent resource production, validation, and dissemination. The proposed projects should be scalable and must address the following three goals:
- Reagents enable unique access to many molecularly defined neural cell types that are found in a complex brain region or significant brain network of a vertebrate and that could exhibit distinct cellular, circuit, or behavioral functions.
- Reagents are easily produced, disseminated, utilized, and stored.
- Collection of reagents are catalogued for users in a brain atlas and registered to cell types based on molecular, anatomical, or other properties that can be referenced.
Examples of responsive research activities include, but are not limited to:
- Creation of a collection of systemically administered viral or non-viral reagents that efficiently cross the blood brain barrier and direct construct expression to molecularly defined brain cell types
- Bioinformatic analysis of genetic and epigenomic profiling data to identify candidate regulatory elements to confer brain cell type selectivity or specificity to constructs
- Production of transgenic vertebrate responder lines containing widely used reporter, monitoring, or manipulation constructs that can be combined with molecular access driver reagents to be expressed in specific brain neural cell types
This FOA is the first in a series comprising the BRAIN 2.0 transformative projects, which were motivated and inspired by the recommendations in “The BRAIN Initiative 2.0: From Cells to Circuits, Toward Cures” and “The BRAIN Initiative and Neuroethics: Enabling and Enhancing Neuroscience Advances for Society.” These large-scale projects, built on the early successes from the first half of the Initiative, will support the development and dissemination of important resources and data to propel neuroscience far into the future. Applications are due on February 11, 2021 and October 19, 2021. For more details and a list of relevant research topics, please see the full notice.
On September 23, 2020, the National Academies’ Forum on Neuroscience and Nervous System Disorders will host a public workshop on transcriptomic differences in the brains of women versus men with brain disorders.
In 2010, the National Academies hosted a workshop on sex differences and implications for translational research, which outlined the public health importance of studying sex differences in the nervous system and considered other important aspects of the field. Five years later, the National Institutes of Health introduced a new requirement for inclusion of sex as a biological variable in animal research. This policy change, among other initiatives, has led to increased recognition of sex differences across disease states, particularly in relation to clinical care and treatment outcomes in neuropsychiatric disorders. With the recent advent of unbiased genome-wide data, we now have more insight into the biological underpinnings of these differences than ever before.
In light of these advances, the upcoming Sex Differences in Brain Disorders: Emerging Transcriptomic Evidence and Implications for Therapeutic Development workshop will bring together experts from academia, government, industry, and non-profit organizations to discuss emerging evidence for differences in transcriptomic abnormalities that occur in the brains of women versus men with a range of brain disorders, including depression, drug addiction, neurodegenerative disorders, and others. John Ngai, BRAIN Initiative Director, will be speaking about considering sex differences in BRAIN Initiative Cell Census Network (BICCN) projects.
Workshop presentations and discussions are designed to:
- Review the current landscape of emerging evidence of these sex differences and consider how this can be used to advance our understanding of brain disorder pathophysiology.
- Explore ramifications for therapeutic development for brain disorders, including identification of new targets, implications for preclinical and clinical study design, and others.
- Discuss open research questions and identify opportunities to move the field forward.
The workshop will take place on Wednesday, September 23rd from 10:00 am to 5:00 pm (ET). Please view the full agenda provided by the National Academies.
Registration for the workshop is still open! Please register here.
The NIH is still seeking input on how to enhance scientific and workforce diversity through the BRAIN Initiative. Public comments will be accepted until August 31, 2020.
Diversity and scientific progress go hand in hand. Individuals from diverse backgrounds and life experiences bring different perspectives and creativity to addressing complex scientific problems.
As emphasized in BRAIN Initiative 2.0: From Cell to Circuits, Toward Cures, the report recommended that BRAIN “continue to recognize that enhancing diversity of the research workforce is a scientific imperative. It should continue to recruit and support students, postdocs, and investigators from diverse backgrounds in NIH BRAIN Initiative-funded projects. These include individuals from groups underrepresented in health-related research.”
Do you (or does anyone you know) have ideas as to how the BRAIN Initiative can broaden diversity in research? If so, let us know! The NIH recently issued a Request for Information (RFI) to seek input from BRAIN Initiative awardees and the broader scientific community on factors that may contribute to the lack of diversity in research teams. The BRAIN Initiative also seeks insight into new approaches to enhance diversity and inclusion. For instance, we seek input on strategies for offering new research opportunities for scientists with limited resources, or ways to benefit diverse trainees and established scientists at institutions traditionally not funded by BRAIN.
Stakeholder feedback is requested on several topics, including developing institutional partnerships and collaborations, creating effective outreach and recruitment, developing metrics to examine diversity outcomes, and others. Please see the previous BRAIN Blog post or RFI for a comprehensive list of topics.
The deadline for submitting public comments is Monday, August 31, 2020! Please send comments to BRAIN.Initiative.Training@mail.nih.gov. For more information, please view the full RFI Notice: https://grants.nih.gov/grants/guide/notice-files/NOT-MH-20-051.html.
Please join us for two upcoming virtual meetings: the NIH BRAIN Initiative Neuroethics Working Group on Thursday, August 20, 2020 and the Multi-Council Working Group on Friday, August 21, 2020. Videocast will be available for the open sessions of both meetings.
We hope that everyone is remaining safe and healthy. As we continue to navigate the evolving coronavirus public health crisis, many of our upcoming meetings are now being held virtually.
Neuroethics Working Group meeting – Thursday, August 20, 2020
This Thursday, the BRAIN Neuroethics Working Group (NEWG) will be holding its tenth meeting. The NEWG helps ensure that neuroethics is fully integrated into the BRAIN Initiative. The NEWG is the only working group at NIH charged with anticipating and identifying ethical challenges, and so the August meeting will feature two timely and important discussions. In one session on neuroethics and racial inequities, Drs. Kafui Dzirasa (Duke University) and Francis Shen (University of Minnesota) will present recent papers to begin a conversation on how neuroethics can contribute to improving racial injustices and underserved populations. In a later session, a panel of BRAIN-funded investigators conducting human subjects research (Drs. Maria Franceschini (Harvard University), Leigh Hochberg (Brown University), Cynthia Kubu (Cleveland Clinic), and Sameer Sheth (Baylor College of Medicine)) will provide first-hand perspectives on how the COVID pandemic has resulted in protocol changes and trade-offs for their research programs. In addition to these discussions, the open session will include an update from the NIH Office of Science Policy on the recently-convened NASEM Committee on the Ethical, Legal, and Regulatory Issues Associated with Neural Chimeras and Organoids.
Multi-Council Working Group meeting – Friday, August 21, 2020
This Friday, the NIH BRAIN Initiative Multi-Council Working Group (MCWG) will convene to discuss the current state of the BRAIN Initiative and its future. The August meeting will feature updates from BRAIN Initiative Director Dr. John Ngai, and a summary of the virtual 6th Annual BRAIN Initiative® Investigators Meeting from Dr. Samantha White. Dr. Ngai will also lead a discussion with the MCWG on the role of BRAIN in promoting diversity, equity, and inclusion. Finally, the open session will include one concept clearance on Phase III of the BRAIN Initiative Cell Census.
Meeting summaries and archived videocast of both meetings will be made available at a later date.
Unraveling neural contributors to position tracking errors in aging humans… A switch to shut off pain… Inducing torpor in mice in order to understand hibernation states… A novel and noninvasive improvement for optogenetics…
Utilizing virtual environments and machine learning to understand path integration in young and aged populations
Specific neurons within the brain work together to form a mental map of our environment, and path integration enables the tracking of an individual’s position in space via neural processes that link mental maps to self-motion cues. However, this sense of direction tends to diminish as we age. One research collaboration between the United States and Germany, led by Drs. Ila Fiete and Thomas Wolbers, seeks to understand how path integration deteriorates with age. The research team conducted an immersive virtual reality path integration experiment in human participants. In the study, young (~22 years old) and aged (~69 years old) participants were asked to track their own position and orientation as they were guided through a virtual environment. Next, the participants’ responses were analyzed by a new, powerful mathematical approach developed by the research team to identify sources of path integration errors. Dr. Wolbers told DZNE – German Center for Neurodegenerative Diseases, “With the help of our mathematical model, we were able to unravel the contributions of various sources of error and identify what distorts position tracking the most and what has little effect. Such sources of error have never been investigated at this level of detail.” Specifically, the model suggests that errors in path integration are primarily caused by the accumulation of “internal noise” in information processing, which is likely a consequence of inaccuracies in an individual’s perception of movement speed. Path integration errors were particularly prominent in aged individuals. This important work helps us to better understand age-related navigational impairments and may ultimately aid in the development of diagnostic tools that distinguish normal age-related orientation problems from those caused by disorders like dementia and Alzheimer’s disease.The path integration task. (a) Example path from a top-down perspective. As participants navigated the path, there were four stopping points (red dots); at these points, participants were asked to report their estimate of the distance and angle to their starting point. (b) The virtual environment was experienced from first-person perspective via a head mounted display as shown here. (c) Three different virtual environments (left panel) used in the path integration task. A tile of each environment’s ground plane is also shown (right panel).
Identification of pain-suppression neurons within the central amygdala
What if we could more effectively suppress chronic pain? Researchers at Duke University led by Dr. Fan Wang have uncovered one potentially potent therapeutic target in the fight against chronic pain – the central amygdala. Although the amygdala is classically known for its role in fear, anxiety, and learning, the researchers have recently demonstrated that a specific neuronal population composed of inhibitory GABAergic cells within the central nucleus of the amygdala plays a critical role in pain processes and effectively acts as a “pain off” switch. The team first identified these cells (termed CeAGA neurons) by using a neuronal marker of activity, c-Fos, following administration of general anesthesia in mice. Next, the researchers employed in vivo calcium imaging and a Fos-based viral-genetic method known as CANE (Capturing Activated Neuronal Ensembles) to better characterize CeAGA neuronal ensembles. Using optogenetic techniques, the researchers demonstrated that activation or silencing of CeAGA neurons bidirectionally altered pain reflexes (i.e., withdrawal responses to mechanical, thermal, and cold sensory tests) and pain-elicited self-caring behaviors (i.e., paw licking and face wiping) in awake mice. Indeed, when inhibitory CeAGA neurons were optogenetically activated, Dr. Wang told The New Daily, “It’s so drastic. [The mice] just instantaneously stop licking and rubbing.” The long-term effects of recurrent activation of this neuronal population are unknown, but these results could help in the identification and development of next generation painkillers for humans and have important clinical implications for patients suffering from chronic pain.A specific population of neurons in the central amygdala (CeA) is associated with pain. (Top figures; a-h) General anesthesia activated a heterogeneous population of GABAergic cells in the central amygdala (CeAGA). Furthermore, the CANE technique was efficient at capturing CeAGA neurons. (Bottom line graphs, b) Compared to GFP control mice, optogenetic excitation of CeAGA neurons (ChR2; blue) decreased and inhibition of these neurons (eArch; purple) increased pain reflex behaviors in awake mice, demonstrating how pain can be bidirectionally modulated by a specific population of CeA neurons.
Identifying the brain region responsible for states of torpor
Torpor is a state of decreased activity that results from reductions in body temperature and metabolic processes and enables animals to survive periods of reduced food availability. While torpor is similar to hibernation in that it involves slowing of the body’s systems, it differs in that torpor is not a voluntary state, in contrast to hibernation. Torpor represents an adaptation that is a defining feature of mammalian and avian evolution, but its underlying neurobiology is poorly understood. A team of researchers from Harvard University led by Dr. Michael Greenberg recently shed light on this very issue using a model of fasting-induced torpor in mice. The researchers used excitatory Gq-DREADD (Gq-coupled Designer Receptor Exclusively Activated by Designer Drug) viral vectors in combination with a FosTRAP system (a tamoxifen-dependent form of Cre recombinase that is driven by Fos). Whole-brain imaging revealed which neurons and brain regions expressed the Gq-DREADD protein. Although simultaneous activation of several brain areas may be responsible for controlling torpor, the researchers revealed that the hypothalamus plays a major role in the regulation of torpor. Specifically, chemogenetic activation of the anterior and ventral portions of the medial and lateral preoptic area (avMLPA) of the hypothalamus resulted in rapid decreases in core body temperature and gross motor activity, which are behavioral characteristics of torpor. Further studies using single nucleus RNA sequencing suggested that avMLPA neurons that express Vglut2 or Adcyap1 regulate the decrease in body temperature that is associated with torpor. Altogether, these significant findings provide insight into how and why only certain species have the ability to enter torpor and the possibility of inducing this type of ancient adaptive behavior in other animal model systems.Inducing torpor in mice with chemogenetics. d) Chemogenetic activation of the anterior and ventral portions of the medial and lateral preoptic area (avMLPA) of the hypothalamus (blue) decreased core body temperature, inducing a state of torpor in mice. The same mice in a control condition (yellow) and mice lacking avMLPA TRAPed neurons (pink) did not show changes in torpor.
Introduction of a new, highly sensitive optogenetic technique
Over the last decade, optogenetics has been a popular technique to manipulate neuronal activity with high temporal and spatial precision, but the invasiveness of optical fiber implantation is less than ideal. In an effort to improve this method, a multi-university team led by Dr. Guoping Feng at the Massachusetts Institute of Technology developed a step-function opsin with ultra-high light sensitivity (SOUL), which enables external light delivery and offers a minimally invasive tool for manipulating neuronal activity in both rodent and primate models. This method is particularly advantageous because, after activation, it allows the animals to be released from the light source while the neurons remain activated. Further, this method allows for synchronized spiking in all opsin-expressing cells at a firing rate that is determined by the researcher. First, the researchers developed and validated the SOUL method utilizing a combination of techniques, including Cre-inducible mouse lines, whole-cell recordings, and photocurrent stimulation. The researchers also demonstrate that SOUL has sufficient photosensitivity to allow for transcranial stimulation within a deep brain region, the lateral hypothalamus (LH). Transcranial stimulation increased the expression of an immediate early gene, c-Fos, demonstrating that it is sufficient to induce neural activity in the LH. The researchers then showed that food-deprived LH SOUL-expressing mice displayed reduced food consumption after SOUL activation by blue light, which were then inactivated by an orange light and food consumption was restored. They further showed that blue and orange light transcranial stimulation does not increase brain inflammation in the cortex. Dr. Feng’s team also demonstrated that this system works in macaques; here, the light was delivered above the dura using an optical fiber to activate deep cortical areas. SOUL was found to induce local field potential oscillations in the macaque cortex. Thus, SOUL may be a promising new tool in the field of neuroscience for the study of the relationship between local or circuit activity and behavior in a minimally invasive and temporally precise way.Stimulation of SOUL-expressing neurons inhibits feeding behavior in mice. A) A schematic of transcranial optical stimulation of SOUL-expressing neurons in the mouse lateral hypothalamus (LH) (red). B) SOUL knock-in mice were injected with the virus used to drive SOUL expression (red) in the LH. C) A microscope image of opsin expression in LH neurons. D) Optogenetic activation of SOUL-expressing neurons in the LH reduced food consumption in food-deprived mice, demonstrating SOUL’s utility for simultaneous manipulation of neuronal activity and behaviors in a rodent model.
These three BRAIN Initiative funding opportunities span from developing new theories and computational models, to creating tools for gene editing in marmosets and next generation human brain imaging. Application due dates are coming up in September and October.
RFA-EB-20-001 Proof of Concept Development of Early Stage Next Generation Human Brain Imaging (U01; Reissue of RFA-EB-19-001)
High-risk, proof-of-concept studies are a critical first step to creating new neuroimaging tools and improving their spatial and temporal resolution. The goal of this FOA is to support the early stage development of brand-new noninvasive neuroimaging methods or exciting bold approaches for current methods that will lead to transformative advances in our understanding of human brain function. Innovative imaging tools often span numerous approaches. Examples of tools applicable to this award include, but are not limited to, new noninvasive tools that have not been used in humans, new approaches that increase the spatial resolution of existing methods, behaviorally active human neuroimaging, and advanced computational techniques for real-time image reconstruction. Tool development under this award will require diverse expertise arising from interdisciplinary teams capable of developing prototypes and conducting small-scale studies in mammals or humans. Thus, this FOA encourages collaborative academic-industrial partnerships. Applications are due on September 3, 2020. For more information, please view the full notice.
RFA-EB-20-002 Theories, Models and Methods for Analysis of Complex Data from the Brain (R01; Reissue of RFA-EB-17-005)
Advances in novel neurotechnologies, largely made possible by the BRAIN Initiative, are increasingly producing enormous, complex datasets. Now, novel theoretical and analytical approaches are needed to harness the power of these data, making them useful for understanding how the brain and its circuits function. This funding opportunity announcement (FOA) promotes the development of new theories, computational models, and analytical tools to enhance our knowledge of the human brain from existing neuroscience data. Proposed projects could develop tools to integrate current theories or formulate new ones, new frameworks or multiscale/ multiphysics models that generate testable hypotheses used to drive new experiments, and/ or new analysis methods to support or challenge a hypothesis about the brain. Projects must focus on theories about neural circuit mechanisms, models of circuit structure and function, and/ or computational methods of analysis of neural activity data. Theories can include the use of specialized animal models. Applications should include plans for how tools and software will be disseminated to the neuroscience research community. Applications are due on September 14, 2020. For more details on topics relevant to this award, please see the full notice.
RFA-DA-21-006 Tools for Germline Gene Editing in Marmosets (U01; New)
Innovative technological advances in germline and brain-specific gene editing, especially transgenic studies in rodents, have provided us valuable tools for interrogating complex neural circuits responsible for animals’ behaviors. With relatively developed frontal cortices and neural structures unique to primates, common marmosets (Callithrix jacchus) possess sophisticated cognitive functions and exhibit cooperative social behavior. Now that they can be made amenable to genetic engineering, marmosets are well-suited for rigorous research involving the genetic study of behavior and/or brain disorders. Last year, the NIH released two FOAs focused on developing marmoset colonies and coordination centers. Building on these efforts, this new FOA aims to develop tools and technologies to conduct scientifically rigorous, ethical, efficient, and cost-effective research that supports gene editing studies in marmosets. Proposed studies should answer critical scientific questions that require the use of genetically modified marmosets. Further, study design must balance potential benefits to human health and/ or scientific advancement with the ‘Three R’s’ (replace, reduce, and refine) of ethically-guided experimental methodology. Applications should propose to optimize breeding technologies, develop germline and somatic gene editing tools (e.g., CRE-driver tools, CRISPR technologies), or promote infrastructure to support these resources. Applications are due on October 15, 2020 and October 14, 2021. For more details and a list of relevant research topics, please see the full notice.
As the COVID-19 pandemic continues to evolve, NIH has revised its page on Frequently Asked Questions (FAQs) to align with implementation of recent guidance from the Office of Management and Budget (OMB) for recipients affected by COVID-19.
As we continue to navigate the current COVID-19 global pandemic, NIH has resources for Frequently Asked Questions (FAQs) on COVID-19-related funding flexibilities. Recently authorized guidance from the Office of Management and Budget (OMB) extends short-term administrative relief to grant recipients that have been directly impacted by COVID-19 due to loss of operations. NIH has implemented this guidance and revised the FAQs accordingly, with new/updated FAQs indicated as such on the page.
These FAQs address issues that may be relevant to the BRAIN community, including:
- Funding flexibilities and extensions (at the time of submission of application and for current awardees), including an exception to our standard policy on post-submission materials and accepting preliminary data as post-submission materials for applications submitted under all activity codes, with caveats that are subject to change (NOT-OD-20-123).
- Commitment to supporting the workforce, including charging salaries to NIH grants when no work is being performed, which is allowable through September 30, 2020, if your organization’s policy allows for the charging of salaries and benefits to currently active awards (under unexpected or extraordinary circumstances).
- Support for training, including salary support for international students and post-doctoral trainees who can continue to be supported on NIH grant projects, even if they are not able to return to the U.S. because of COVID-19 related travel restrictions.
To view these and all of the updated FAQs, please visit https://grants.nih.gov/faqs#/covid-19.htm.
Novel neural network translates cortical activity to text… Low-carb diet may prevent and reverse brain aging… Transcriptomic characterization of rare von Economo neurons… Deep learning model predicts Parkinson’s symptomology…
Encoder-decoder framework translates human cortical activity to text
Brain-machine interfaces (BMIs) allow tetraplegics to speak by decoding brain activity into text or speech. Direct decoding of speech, however, is typically restricted to single phonemes or monosyllables, and decoding continuous speech is both limited in vocabulary size and prone to errors. By leveraging advances in machine translation – the computational translation of text from one language to another – Dr. Edward Chang and his research team at the University of California, San Francisco crafted a machine learning framework to translate cortical brain activity to text with high accuracy and at natural-speech rates. Rather than base their decoding on single words, researchers trained a neural network model to translate brain signals from electrocorticograms (ECoGs) into whole written English sentences. In the study, four consenting participants undergoing treatment for epilepsy read a series of sentences displayed on a screen out loud. While they read, researchers recorded their brain activity using ECoG arrays of 120-250 electrodes implanted over each participant’s peri-Sylvian cortex. Then, these data sequences were fed into an ‘encoder-decoder’-style artificial neural network. The network processes the sequences in three stages: (1) it extracts temporal or timing patterns in the signals (i.e., temporal convolution); (2) the data are passed to an encoder recurrent neural network (RNN), which learns to summarize sequences in a single abstract state; and (3) the abstract state is decoded or transformed into written text by a second RNN, which learns to predict the next word in the sequence. For each participant, speech was limited to a restricted ‘language’ of 30-50 sentences. Nevertheless, the framework was highly accurate, as word error rates were as low as three percent on datasets with 250-word vocabularies. Further, speech decoded with minimal data was improved by transfer learning, meaning that the network can potentially learn new English words that it had not trained on from other participants or speaking tasks. The novel framework created by Dr. Chang and his team represents significant progress in using artificial neural networks to decode natural speech and has the potential to improve current speech prosthetics in patients.The decoding pipeline. Each participant read sentences out loud while their neural signals were recorded with an ECoG array covering their cortices. Sentence sequences are processed in three stages: (1) temporal convolution; (2) encoding into a hidden state by an encoder recurrent neural network (RNN); and (3) decoding into speech by a second RNN.
Neuroimaging study shows that a low-carb diet may prevent, reverse brain aging
Brain aging, and especially dementia, is associated with hypometabolism – the inability of neurons to use glucose as an energy source. Converging evidence in animal models also suggests that switching the brain’s fuel source from glucose to ketones could restore and maybe even prevent the neurobiological changes associated with cognitive decline. However, little is known about the time course of these age-related changes and how diet influences brain aging in humans. In a recent neuroimaging study, Dr. Lilianne R. Mujica-Parodi and her colleagues at Stony Brook University showed that the neurobiological changes associated with aging are observed at a younger age than previously thought and that this process can be prevented or reversed by a diet low in carbohydrates. First, Dr. Mujica-Parodi and her team used two large-scale human fMRI datasets from 928 individuals across the life span (ages 18 to 88) to establish network stability as a robust whole-brain biomarker associated with aging. They found that functional communication between brain regions destabilizes with age, typically beginning in the late 40s, and that this biomarker is linked to poorer cognition and accelerates insulin resistance. Next, researchers conducted two neuroimaging experiments in 42 adults under the age of 50 to test the effects of manipulating fuel type: glucose versus ketone bodies, on the biomarker. For the first group, each participant as scanned while on a standard diet, overnight fasting, and ketogenic diet. To isolate the effects of fuel type, a second group fasted overnight and was scanned before and after receiving a calorie-matched glucose and ketone ester supplement. Dr. Mujica-Parodi and her team found that brain networks were destabilized by glucose and stabilized by ketones, regardless of whether ketosis was achieved via changes in diet or the calorie-rich supplement. Taken together, these findings suggest that brain network destabilization may reflect early signs of hypometabolism and that simple dietary interventions may protect the aging brain. To learn more about these fascinating results, please read the Stony Brook University news release.Brain networks stabilize with ketones. Each participant was scanned three separate times: while on a standard diet, after an overnight fast, and following a weeklong ketogenic diet. Ketosis induced by a ketogenic diet and by drinking an exogenous ketone supplement (data not shown) showed similar network stabilization effects to those seen with fasting.
Transcriptomics and cross-species homology mapping used to characterize von Economo neurons
von Economo neurons (VENs) are large, spindle-shaped neurons in the frontal cortex that are particularly vulnerable to neuropsychiatric and neurodegenerative diseases. In humans, these rare neurons only reside in the anterior cingulate or insular cortices and are not found in rodent brains, making their functional properties challenging to study. Recent studies using postmortem brain tissue data from the Allen Human Brain Atlas have identified several genetic markers of VENs. Other studies have identified possible projection targets, but distinct features of VENs remain unknown. To more thoroughly characterize VENs, Dr. Ed Lein and his research team at the Allen Institute for Brain Science used modern RNA-sequencing technologies in human and mouse brains to determine that VENs are a regionally distinct type of extratelencephalic-projecting excitatory neuron. Researchers first used single-nucleus RNA-sequencing (snRNA-seq) to closely examine the genetic profile of cells from the frontoinsular cortex of two postmortem human brain specimens. SnRNA-seq is a valuable tool for classifying and characterizing human brain cells because it can be used on postmortem frozen tissue and it enables the alignment of cell type datasets across brain areas and species. By using this method, researchers identified 13 excitatory neuron types and found that VENs can be localized to a single transcriptomic cell type. This classification included cells with fork and pyramidal morphologies, however, researchers also identified new genetic markers for VENs. Next, to further characterize the cellular properties of VENs, the team carried out cross-species homology mapping by aligning the human snRNA-seq data with mouse cortical single-cell RNA-sequencing data. They found that the human cells were homologous to extratelencephalic (ET) neurons in the mouse, enabling them to predict that, just like mouse ET neurons, human VENs send projections subcortically. Researchers also performed the very first ex vivo electrophysiological recordings of VENs in postmortem human brain tissue and showed that these cells had distinct intrinsic membrane properties of nearby pyramidal neurons. These findings demonstrate how advanced transcriptomic approaches that enable precise cell-type identification and mapping across species enable a better understanding of human brain cells and their circuitry.Identifying and visualizing transcriptomic cell types of von Economo neurons (VENs) in situ. Fluorescent in situ hybridization (ISH) and double chromogenic ISH for marker genes of transcriptomic cell types corresponding to VENs were used to visualize cells with pyramidal (P), VEN (V), and fork (F) morphologies in human brain tissue.
Deep learning neural network uncovers vital role of brain regions in Parkinson’s symptomology
Machine learning (ML) has emerged as a useful tool for neurodegenerative disease diagnosis and progression. Recent ML studies have used deep-learning models to predict Alzheimer’s symptomology and observable cognitive and motor symptoms in Parkinson’s disease (PD) patients. However, the high variability of PD brains, small sample sizes, and lack of long-term brain atrophy make estimating future PD symptomology with deep-learning networks a challenge. In this study, Dr. Ashish Raj and his research team at the University of California, San Francisco employed an autoencoder-based deep learning model to predict future cognitive and motor impairments in PD patients. Researchers selected structural fMRI data from 42 healthy controls and 116 PD patients from the Parkinson’s Progression Markers Initiative database. Using this dataset, they designed a neural network model that predicted a patients’ cognitive and motor score over a few years simply based on anatomical features of brain regions at baseline. Importantly, training and testing the model was quicker and computationally inexpensive compared to existing neural networks that perform similarly. Overall, their model identified salient brain regions involved in cognitive and motor decline, such as the substantia nigra and caudate nucleus. Dr. Raj and his team also found that the contributions of various brain regions to these symptoms changed over time, suggesting that the regions responsible for certain impairments shifts as the disease progresses. For instance, their model revealed that the red nucleus plays a primary role in cognitive scores early on but diminishes by the fourth year when the substantia nigra is most involved in cognitive impairments. As for motor scores, the palladium is most dominant in the second year and the role of the subcallosal area grows over time. These results offer new insights about PD progression and symptomology and highlight the growing importance of using ML methods as powerful diagnostic tools in medicine.A deep learning neural network model predicts the emergence of salient regions involved in cognitive and motor impairments of Parkinson’s patients over four years. Numbered regions represent areas when they first appear, compared to prior years. Montreal Cognitive Assessment (MoCA) scores reflect cognitive impairment and Movement Disorder Society-Unified Parkinson’s Disease Rating Scale Part III (UPDRS) scores represent motor dysfunction.
Wireless control of stress hormone secretion in rats … Cortical organizational mechanisms underlie brain-computer interface learning … Dynamic advances in noninvasive imaging technology aided by machine learning … Amputees control prosthetic limbs with a new regenerative peripheral nerve interface …
Remote, transgene-free magnetothermal regulation of adrenal hormones
Scientists can use implantable devices to modulate electrical activity within the body and current less invasive techniques lack precision and rely on transgenes – but what if there was an alternative, transgene-free noninvasive strategy? The robust, endogenous expression of a heat-sensitive cation channel, transient receptor potential vanilloid family member 1 (TRPV1), allows for the possibility of applying magnetothermal control to remotely modulate organ function without relying on transgenes. Dr. Polina Anikeeva and her colleagues at the Massachusetts Institute of Technology (MIT) developed a magnetothermal switch to wirelessly control the release of the adrenal hormones epinephrine and corticosterone in a genetically unaltered rat. After verifying that the TRPV1 receptor was highly expressed in the adrenal gland using immunofluorescent labeling procedures, researchers characterized the magnetic nanoparticles to determine if these particles were necessary and sufficient to trigger a robust calcium ion influx into TRPV1-expressing adrenal cells. First, they verified that the new magnetothermal control system worked in vitro and then in vivo. Altogether, this study demonstrates that these new magnetic nanoparticles can persist within tissue for up to six months and allow for chronic stimulation of the adrenal gland over days and weeks without damaging tissues or altering cellular function. Dr. Anikeeva and her team’s work paves the way for technological advancements in remote, noninvasive control of organ function and may help inform bioelectronic medicines in humans. Eventually, given the central role of adrenal glands in regulating the stress response in mammals, technology like this could one day be used to treat stress-related disorders by modulating peripheral organ function.Magnetothermal stimulation of adrenal cells in vitro. (A) Schematic of magnetothermal triggering of TRPV1, which is recorded via calcium (Ca2+) imaging using a fluorescent indicator, Fluo-4. (E) Ca2+ imaging shows the influx of Fluo-4 fluorescence in TRPV1-expressing cells in response to capsaicin (E) and magnetic nanoparticles (H) in adrenal cell culture.
Performance in brain-computer interface learning is predicted by the organization of cortical areas
Brain-computer interface (BCI) learning can be used to improve neurofeedback devices that directly regulate an external device, but how can we improve on their usability? One way to improve BCI technology is to gain a better understanding of the neural processes that enable BCI control, such as neuroplasticity. Recently, Dr. Danielle Bassett at the University of Pennsylvania and her colleagues at Sorbonne Université in Paris, France, characterized the evolution of large-scale cortical network changes that occur during BCI training. Specifically, researchers predicted that BCI learning would be accompanied by a decrease in functional integration in cortical regions strongly associated with cognitive processing as BCI users transition from using a deliberate mental strategy to processing nearly automatically. To determine precisely how cortical connectivity changes, the researchers simultaneously recorded high-density electroencephalographic (EEG) and magnetoencephalographic (MEG) signals during BCI training, which consisted of four sessions over a period of two weeks. They found that at the end of the two weeks, event-related desynchronizations – which reflect sensorimotor brain activity – were more localized in the contralateral paracentral lobule, precentral gyrus, and superior parietal lobule. These regions are typically involved in hand motor tasks and motor imagery. Dr. Bassett and her team also observed that functional connectivity was reduced in the orbital part of the inferior frontal gyrus, an area involved in mental rotation and working memory. This area was also strongly predictive of BCI scores. Altogether, understanding the large-scale neural mechanisms underlying motor plasticity is fundamental to our understanding of the cognitive aspects of human learning and may inform the development of better machine learning interfaces in the future.Cortical connectivity changes during BCI training. Connectivity networks obtained with MEG-source reconstructed signals in the α2 (8-12 Hz) and β1 (14-29 Hz) frequency ranges are represented here on circular graphs across four sessions involving motor imagery tasks. Within each graph, the red and blue edges correspond to higher functional connectivity levels in the motor imagery and rest conditions, respectively. The color of each node corresponds to a different brain region of interest, as shown in the bottom.
New shifts in dynamic and noninvasive human brain imaging technology
Human brain imaging has come a long way in the nearly 100 years since the discovery of the electroencephalogram (EEG), but every next step towards higher resolution processing is a step towards improving patient outcomes. Dr. Gregory Worrell from the Mayo Clinic in Minnesota and his colleagues from the Carnegie Mellon University in Pennsylvania created a novel non-invasive, high-spatiotemporal resolution source imaging approach to map the brain networks of individuals with epilepsy. Specifically, they demonstrated the effectiveness of the fast spatiotemporal iteratively reweighted edge sparsity (FAST-IRES) technique, which uses unbiased machine learning to estimate signal sources and activity across time. Here, they used this approach to estimate the epileptogenic zone (EZ), or the minimum amount of brain tissue that needs to be removed to stop seizures. The FAST-IRES technique was further validated through a direct comparison to intracranial EEG findings and surgical outcomes in 36 patients with focal epilepsy. In total, 1,027 spikes and 86 seizures were analyzed. Altogether, the new FAST-IRES approach could be applied to both medical and research settings to objectively and noninvasively, with high precision, acquire human brain imaging data via scalp high-density EEG recordings. FAST-IRES can work in collaboration with existing hardware to improve, for instance, the management of epilepsy. For more information, check out Carnegie Mellon University’s news article on this exciting technology.Imaging seizures with FAST-IRES. To determine the epileptogenic tissue, the dominant frequency of the middle of the seizure (i.e., the ictal phase) is determined and the source signal distribution is filtered at this frequency; then the energy of the source signals is calculated at this frequency. FAST-IRES provides a spatiotemporal estimate of the underlying brain region, or brain source, which is then compared to clinical findings to confirm the reliability of the approach.
A new regenerative peripheral nerve interface allows for real time control of prosthetic limbs in amputees
Restoration of movement is something that many upper limb amputees struggle with – but new advances in regenerative nerve interface technology allows for the real time control of prosthetic limbs. Here, Dr. Cynthia Chestek at the University of Michigan and her team reveal a newly designed, biologically stable regenerative peripheral nerve interface (RPNI). This RPNI is a peripheral nerve bioamplifier that produces high-amplitude surface electromyography (EMG) signals and can be used to control a prosthetic device. Its use also indicated that RPNIs can reduce phantom limb pain. This nerve regeneration interface was applied successfully even in cases where years had elapsed since the initial injury. Overall, the authors demonstrate in multiple independent cases of upper limb loss that these RPNIs might be an effective peripheral nerve interface for control of prosthetic devices. Dr. Chestek and her colleagues also demonstrate that RPNIs are a viable alternative to existing prosthetic control approaches that interface with intact muscles, where directly extracting efferent motor action potentials from the peripheral nerve is possible. A device such as this has the potential to revolutionize the future of clinical prosthetic technology, as well as greatly improve the quality of life for patients with limb loss. For more information, check out The Michigan Engineer News Center article on this innovative work.RPNI sonograms captured 19 months after RPNI surgery. Highlighted areas on the sonograms show which regions of the median and ulnar RPNIs were contracting during cued finger movements.