Copy
Enhancers: Where evolution meets development and disease
Studies illuminate roles for non-coding DNA in human brain evolution and vulnerability to disease

Ron McKay, PhD, Chief Editor
Lieber Institute for Brain Development

Maltz Research Laboratories
In this time of COVID-19 pandemic, one may reasonably wonder why to bother reading this new issue of neuroDEVELOPMENTS. Keeping up with the latest numbers, epidemiological models, drug trials, or mask recommendations takes enough time, and people are rightly transfixed in hopes of keeping themselves and their loved ones healthy. Yet, here we hazard to say that reading this latest version of the neuroDEVELOPMENTS newsletter will capture your attention. The discovery that genetic variation is linked to psychiatric disorders and that most of the variation is in regions of the genome long thought to be “silent” is a profound insight to emerge from the Human Genome Project. Here we discuss three recent papers focused on the cell type and pathological specificity of genetic and epigenetic mechanisms in the human brain related to these disease-associated variations.

In the first issue since the death of our prescient friend, Steve Lieber, we are committed to follow his charge to gain insight into how the human nervous system functions. Steve first suggested this newsletter format and we trust you will enjoy this 5th issue. Three editorial members are authors of the papers we discuss, and each paper gives us insight into the genetic mechanisms important to brain evolution, development, and disease. Thanks to the Board for their comments that strengthen our understanding of the genesis of developmental brain disorders.

--RM
neuroDEVELOPMENTS Editorial Board
 
Fred 'Rusty' Gage, PhD
President, The Salk Institute for Biological Studies

Daniel Geschwind, MD, PhD
Professor, UCLA School of Medicine

Elizabeth Grove, PhD
Professor, University of Chicago
 
Jürgen Knoblich, PhD
Director, Institute of Molecular Biotechnology, Austrian Academy of Sciences 

Arnold Kriegstein, MD
Professor, UCSF

Pat Levitt, PhD
Professor, Keck School of Medicine of USC

Mu-Ming Poo, PhD
Director, Institute of Neuroscience, Chinese Academy of Sciences

John Rubenstein, MD, PhD
Professor of Psychiatry, UCSF

Nenad Sestan, MD, PhD
Professor, Yale University 

Flora Vaccarino, MD
Professor, Yale University 

Chris Walsh, MD, PhD
Chief, Division of Genetics & Genomics, Boston Children's Hospital


______________________________________

Venkata S. Mattay, MD
Managing Editor

Michele Solis, PhD
Science Writer 
neuroDEVELOPMENTS 

Programs for gene expression


Over 20 years, studies seeking genetic links to human disease have gotten really good at finding places in the genome linked to risk. But there’s a hitch: these risk variants mostly land in the uncharted territories of the non-coding portion of genome. Comprising nearly 99% of the genome, these mysterious regions have forced researchers to take new steps to understand their function. Efforts to “annotate” or “map” these vast and mysterious regions indicate that they contain dashboards of controls that regulate expression of the 1-2% of the genome containing the protein-coding sequences of actual genes.

This suggests risk for common disease manifests as a difference of degree, rather than kind: common risk variants don’t so much land in genes to damage resulting proteins, but rather affect how much or when or where a protein is produced. Gene regulation takes many forms, but one element involves short stretches of DNA, called enhancers, that are abundant and distributed throughout the genome. Enhancers serve as assembly sites for dynamic chromatin structures, which help bring together the molecular machinery needed to transcribe genes, i.e. turn them on in the right way.

Three recent studies fill out the picture of enhancer involvement in regulating programs of gene expression in the brain. In line with previous studies of pluripotent cells, their findings show that enhancers have a central role in orchestrating a cell’s particular transcriptional identity in the human brain. As you will see, these brain enhancers are a ready substrate for evolutionary transitions that link brain transcription to genetic risk for psychiatric and neurodegenerative conditions. The findings take us closer to understanding the network of gene regulatory elements that control brain function in sickness and in health.

 

"The human brain seems to be particularly sensitive to gene dosage, suggesting the importance of the level and pattern of expression of a gene, rather than merely the presence of an intact gene." – Christopher Walsh

"Together these papers provide evidence that regulatory region variants contribute to the risk for different brain disorders, depending upon whether the elements control gene expression in neurons, oligodendrocytes or microglia." – John Rubenstein

"Variation and selection drive evolution, yet if the variation is too dramatic the organism may not survive, depending on the environmental conditions. Thus, variations that do penetrate or persist are more subtle." – Rusty Gage

"[T]he human evolved regulatory elements are enriched in specific cell types that underlie the evolutionary increase in brain size and expansion of the upper cortical layers.” - Arnold Kriegstein

Brain-focused

An early study led by neuroDEVELOPMENTS board member Christopher Walsh first proposed that enhancers were crucibles for human brain evolution as well as risk for brain disorders, based on DNA samples from blood cells (Doan et al., 2016). The new studies mentioned here address this question in brain cells for the first time, allowing a more refined and extensive picture of the brain loci involved. The first of these recent studies, led by Daniel Geschwind of the University of California Los Angeles (and neuroDEVELOPMENTS editorial board member) focused on identifying active gene regulatory elements in DNA obtained from precious human fetal brain at different stages of development, and also from adult cortical tissue (Won et al., 2019).

There are several methods to do this that involve identifying the parts of a DNA strand that are “open” and thus available to binding by transcription machinery. By comparing the openings found in humans to those in chimpanzees or rhesus macaques, the researchers defined different categories of human-specific changes in regulatory elements, including DNA sequence changes, and gain or loss of transcription-changing epigenetic chemical marks (Figure 1). They found these elements controlled different genes — which could be geographically far from the regulatory elements themselves — and were active at different points of development. Thus, these “human-evolved” gene regulatory elements seemed to represent three different pathways to control gene expression.

Yet, these regulatory elements showed some interesting convergences. They were especially active in layer 2/3 of the cortex, and in particular cell types, like neuron-supporting astrocytes in adult brain, and the neuron-birthing radial glial cells in fetal brain. This work suggests that these cell types are important venues for brain evolution. In addition, two regulatory element enhancers controlled genes related to autism and developmental disorders, which argues that disruptions to gene regulation very early in life can have consequences for brain health later.

They go on to discuss how this collection of enhancer elements, or the “enhancer-ome”, includes genes that play central roles in neuronal function and transcription factors known to be key regulators of forebrain development. They specifically demonstrate, in a few model in vitro cases, that one of the newly-defined human risk enhancers controls the expression of a distant transcriptional regulator. They suggest that these mechanisms of gene regulation will have important consequences for modeling the molecular and cellular origins of neuropsychiatric genetic risk.

Figure 1 - Human enhancer elements The authors defined four classes of regulatory elements found in humans, but not in other non-human primates. These “human-evolved” elements include: variants within the DNA sequence (“human accelerated regions” or HAR); genomic regions marked with epigenetic modifications (in green) in fetal brain tissue, specifically H3K27ac marks associated with increased transcription (“human gained enhancers” or HGE); HGEs obtained from adult brain tissue; and genomic regions marked by a loss of transcription-enhancing marks in humans (“human lost enhancers” or HLE) in adult brain. Image from Won et al., 2019.

Cell selection


The second study also draws a connection between non-coding regions of the genome and cell type, ultimately finding further cell-specific links to disease (Nott et al., 2019). Researchers, including Fred Gage of the University of California San Diego (and on the neuroDEVELOPMENTS editorial board), began with pools of specific cell types — neurons, microglia, oligodendrocytes, and astrocytes — from human brain tissue. They applied two methods for finding the active chromatin sites within each cell type, which yielded putative promoters and enhancers. While promoters were stationed close to the gene they are in charge of, the enhancers were often far from the genes they controlled. Interestingly the active promoters were not particularly cell-specific, but the enhancers were, and among these were nearly 3000 “super enhancers” — clusters of enhancers collectively bound by transcription factors. The study highlights the extensive role enhancers have in dictating the gene expression patterns central to cell identity in the human brain.

These regulatory elements also had interesting links to disease. The researchers found that many of the single nucleotide polymorphisms (SNPs) identified by genome-wide association studies (GWAS) of various diseases localized to promoters and enhancers. This was particularly true for SNPs associated with neuropsychiatric conditions like autism, bipolar disorder, and schizophrenia, which landed in promoters and enhancers active in neurons. In contrast, risk SNPs associated with Alzheimer’s Disease (AD) were enriched within enhancers active in microglia of AD brains, which adds a new cellular mechanism to the known pathological role of these non-neural cells in dementia patients (Figure 2). Piecing together enhancer-promoter-gene interactions revealed cell-specific combinations, such that a certain enhancer connected with a certain promoter only in one cell type, but not in others. Highlighting these three-way interactions between enhancers, promoters, and gene expression fills out the gene regulatory atlas by specifying a mechanism for many disease-related genetic variants landing in non-coding regions.

"The most striking finding is that enhancer variants associated with the risk of sporadic AD and autism are associated with regulatory elements for gene expression in microglia and oligodendrocytes, respectively."-Mu-ming Poo

"It is a general principle that the most recent evolved traits are the most liable to disease or disorder, and it is not surprising that the rapid evolution of non-coding regions is associated with the manifestation of diseases such as autism." – Arnold Kriegstein
 

Figure 2 - Checkerboard of signals The plot shows how genetic variants associated with risk for neurological diseases, psychiatric disorders, and traits (x-axis) land in regulatory regions obtained from different cell types (y-axis), with promoters in green and enhancers in yellow. Enhancer regions identified in neurons were significantly enriched (dark blue) for risk variants associated with autism, major depressive disorder, bipolar disorder, and schizophrenia. Enhancers in microglia were enriched for risk variants associated with Alzheimer’s disease. Image from Nott et al., 2019.

“[These] papers lay the foundation for functional analysis of those enhancer elements, a project that will require genetically accessible human model systems, such as cultured neurons or organoids.”-Jürgen Knoblich

“Moving forward, in addition to physical mapping and chromatin interactions maps, functional validation of enhancer activity by orthogonal assays in an in vitro “living system” (including expression of enhancer libraries linked to a reporter, or targeted activation/repression using CRISPR) may be necessary.”- Flora Vaccarino

“[W]e are now finding that mutations in enhancers—which are in charge of controlling these levels and patterns of gene expression—can cause almost as much dysfunction as mutation of the gene itself.” – Christopher Walsh
 

Regulatory rewire

The third study also knits together the themes of recently evolved gene regulators, brain disorders, and cell type specificity (Castelijns et al., 2020). Reasoning that the brain size expansion characteristic to humans began earlier in primate evolution, the researchers took a comparative approach, looking for gene regulatory elements in marmoset, macaque, chimpanzee, and human brain. In both prefrontal cortex and cerebellum, they found a shared pattern of enrichment for certain regulatory elements between humans and chimps, but not with marmosets or macaques. Specifically, they found humans and chimps shared many of the same active enhancers and promoters, which they called “hominin gains.” Another set of “hominin losses” was also identified, for which humans and chimps showed lost enrichment relative to marmosets and macaques (Figure 3).

Enrichment for hominin-specific gains — an index of active regulatory elements found only in humans and chimps — was specific to cell type and time. These elements were particularly pronounced in oligodendrocytes during postnatal development, but not so much in astrocytes, microglia, and neurons. The researchers suggest that this connection between oligodendrocytes and postnatal development reflects the importance of protracted axon myelination in the hominin brain. Along these lines, the study found signs of dysfunctional gene regulation in autism: post-mortem brain tissue from people with autism did not share the pattern of enrichment seen in hominin-specific elements in controls. For example, two regulatory regions deemed hominin-specific gains had reduced activation in autism brain, and these were found to physically interact with genes previously associated with autism. Ultimately, the work suggests that the evolving genome rewires interactions between its regulatory elements to produce different outcomes.

Figure 3 - Access gained and lost In prefrontal cortex samples obtained from humans (far left), chimpanzees (left of center), rhesus macaques (right of center), and marmosets (far right), changes in regulatory elements are plotted: at the top, enriched access to a regulatory element in humans and chimps is shown in a row (solid red), whereas the same genomic region in macaques and marmosets did not show much access (sparse red). Increased access means that the regulatory element was available to help with transcription. Image from Castelijns et al., 2020.

An active enhancer catalog

The first enhancer was defined 40 years ago in studies of SV40, a monkey virus similar to JC virus, a normally benign human counterpart. In the subsequent decades we have developed powerful tools to catalog the vast number of potential enhancer sequences in the human genome. The three papers we discuss here present a contemporary view of gene enhancers as a canvas for human evolution and the loci of vulnerability for developmental brain disorders. As we work to understand the full catalog of enhancers and their place in human gene regulatory systems, these new studies argue for their central role linking genome evolution and the dynamic epigenetics of brain disease.
###
The full commentary from our Editorial Board on the two papers highlighted in this issue of neuroDEVELOPMENTS.
 
 1. From John Rubenstein, MD, PhD, University of California, San Francisco:
These three papers have made important advances in identifying regulatory elements (enhancers and promoters) that appear to be evolutionarily gained or lost in the human genome, compared to monkeys and primates, and that are associated with different aspects of neural development and different brain cell types, and different human disorders. Won et al., 2019 identified such elements associated with genes controlling patterning and cell specification in the developing forebrain (Gli2&3 and Tbr). Variants in the elements are associated with neuropsychiatric disorders. Castelijns et al., 2020 found evidence for such regulatory elements to be associated with genes expressed in oligodendrocytes (myelin producing cells), and that these were disrupted in autism patients. Nott et al., 2019 found variants in regulatory regions that normally are implicated in the control of gene expression in neurons. The genes associated with these regulatory elements are in many cases implicated in several psychiatric disorders. By contrast, regulatory regions variants identified in schizophrenia were associated with genes that are expressed in microglia. Together these papers provide evidence that regulatory region variants contribute to the risk for different brain disorders, depending upon whether the elements control gene expression in neurons, oligodendrocytes or microglia.

2. From Rusty Gage, PhD, Salk Institute:
Variation and selection drive evolution, yet if the variation is too dramatic the organism may not survive, depending on the environmental conditions. Thus, variations that do penetrate or persist are more subtle. It follows that mutations (variations) in coding regions of the genome are risky, while mutations in non-coding regions are less so. The challenge has been to find how non-coding variations can have any effect on behavior, on which selection can act. The papers selected in the current edition of neuroDEVELOPMENTS highlight both technical and conceptual advances in deciphering the underlying mechanisms that select for both positive and negative consequences of variation. While these are excellent first efforts in this challenging field, I predict that there are other clues (loci of DNA damage and repair) which will add to our understanding of how these non-coding sites of variation are selected and what coding genes they regulate.

3. From Arnold Kriegstein, MD, University of California, San Francisco:
In this issue of neuroDEVELOPMENTS we review three important papers that highlight the importance of regulatory elements in driving both evolutionary change and disease susceptibility in the human brain. While most protein-coding gene sequences are conserved across human and non-human primate species, evolutionary change has been most significantly driven by alterations in the non-coding regulatory sequences and has played out in a cell-type specific manner. As elegantly shown by Won et al., 2019, the human evolved regulatory elements are enriched in specific cell types that underlie the evolutionary increase in brain size and expansion of the upper cortical layers. Cells in the supragranular layers, those layers superficial to layer IV, are largely cortical projecting cells thought to underlie cognitive abilities and these have undergone a particular expansion in human and non-human primate evolution. Particularly interesting is the pattern of cell enrichment of human accelerated regions and human gained enhancers in the developing and adult brain. In the developing brain, these human regulatory elements converge on radial glia cells, in particular the outer radial glia, that are expanded in the developing primate brain. In the adult they converge on upper cortical layer neurons, which are the progeny of the outer radial glia. It is a general principle that the most recent evolved traits are the most liable to disease or disorder, and it is not surprising that the rapid evolution of non-coding regions is associated with the manifestation of diseases such as autism. Hopefully, the finding of cell type specific vulnerabilities to disease mediated by regulatory gene elements will eventually lead to a better understanding of disease pathology and targeted therapies.

4. From Jürgen Knoblich, PhD, Institute of Molecular Biotechnology, Austrian Academy of Sciences:
This issue of neuroDEVELOPMENTS highlights how modern computational genome analysis can shed light on what was long thought to be the “dark matter” of the genome – the non-coding areas of our DNA that regulate gene expression. For the longest time, genetics has focused on genes as units of transcription and disease mutations as genetic defects inactivating their final product, the proteins generated from transcripts. Therefore, most human genetic analysis relied on the “exome”, the combined sequences of only the coding part in the genome. In a time of more and more complete genome sequences, however, we realize that most neurological and psychiatric diseases, however, do not follow that simple rule. Often, their genetic component rests on a combination of many risk variants that reside outside the known transcription units. Understanding those genetic variants requires new ways of genetic analysis.

The three papers outlined in the current issue present such analysis, coming to important, yet simple and easily understandable biological messages. Won et al., 2019 point out that regulatory elements only found in humans are particularly important for neurological disease and show that these elements are particularly active in the outer layers of our cortex. Nott et al., 2019 analyzed various disease-relevant enhancers and show that neuro-psychiatric disorders map to neurons, while Alzheimer’s disease variants map to microglia. Finally, Castelijns et al., 2020 show that elements particularly important for human evolution map to oligodendrocytes, highlighting the importance of non-neuronal cells in the brain. Together, those papers lay the foundation for functional analysis of those enhancer elements, a project that will require genetically accessible human model systems, such as cultured neurons or organoids.

5. From Flora Vaccarino, MD, Yale University:
Both Nott et al., 2019 and Won et al., 2019 identify regulatory relationships between upstream elements—promoter, enhancers—and target genes. The two studies use a variety of methods to solidify these relationships, which include not only a physical mapping of their location in the human genome by ChIP-seq, but also chromatin conformation analyses to assess their interactions in 3D space (HI-C and PLAC-seq). This is important since proximity to a promoter is not always a good indicator of an enhancer with regulatory relationships. Indeed, Nott et al., 2019 identified risk variants using the PLAC-seq interactome that were linked to a distal promoter but not to the closest promoter. In addition, the landscape of regulatory elements is likely very diverse—encompassing different kinds of elements identified by the ENCODE project many years ago in cell lines, but yet to be identified in human brain—which include bivalent/poised enhancers, flanking promoter/poised promoters, and repressed enhancers.

Won et al., 2019 described human accelerated regions (HAR) and human-gained enhancers (HGE) that are putative regulators of genes active in the developing and adult human brain, and expressed in upper cortical layers and outer radial glia cells, cell types known to be crucial for human cortical expansion. It remained more difficult, however, to show that these target genes are actually differentially expressed between human and primates during brain development. Together, all three papers give invaluable insights into how cell type-specific putative regulatory elements may differentially contribute to specific diseases. Moving forward, in addition to physical mapping and chromatin interactions maps, functional validation of enhancer activity by orthogonal assays in an in vitro “living system” (including expression of enhancer libraries linked to a reporter, or targeted activation/repression using CRISPR) may be necessary. This in turn highlights the difficulty in selecting an appropriate human cellular model, given that enhancers exhibit far more cell type-specificity than genes.

6. From Mu-ming Poo, PhD, University of California, Berkeley, and Chinese Academy of Sciences, Shanghai, China:
These three papers have convincingly demonstrated that non-coding genetic variation could be a major driver of phenotypic diversity, and variants in enhancers and promoters in specific cell types, including neurons and glial cells, could be involved directly in conferring psychiatric and neurodegenerative phenotypes. The most striking finding is that enhancer variants associated with the risk of sporadic AD and autism are associated with regulatory elements for gene expression in microglia and oligodendrocytes, respectively. This supports the notion that specific glia cell types are involved in the pathogenesis of different diseases. Given that modification of a variant regulatory element could alter the expression of a different set of genes that differ from those resulting from modification of the coding region of specific genes, a brave new world of genetically modified animal models targeting regulatory elements is now opened for exploration. The finding of emerging evolutionary changes of regulatory elements in primates and hominins also argue that new animal models need to include non-human primates, such as marmosets and macaque monkeys.

7. From Christopher Walsh, MD, PhD, Boston Children’s Hospital:
These three new papers illustrate the rapid progress in understanding the regulation of gene expression in the brain, and how this can differ dramatically between species. The human brain seems to be particularly sensitive to gene dosage, suggesting the importance of the level and pattern of expression of a gene, rather than merely the presence of an intact gene. We see this repeatedly in the large number of neurological and psychiatric conditions that reflect damage to just one of two copies of a gene. Many forms of autism spectrum disorders, or epilepsy, reflect de novo (i.e., spontaneous) mutations of one copy of a gene expressed in brain, while leaving a second functional copy of the gene intact. Moreover, these heterozygous mutations are sufficient to cause devastating disorders of brain, without affecting other organs in most cases. Even more surprisingly, many of these genes, when inactivated in mouse models, don’t cause similar phenotypes unless both copies of the gene are impaired, which suggests that this dependence of gene dosage is most developed in humans.

What explains the particular dependence of human brain on having two functional copies of a gene? Although the answer is not yet known, it most likely lies in the control of the specific levels and patterns of gene expression, which might not be adequately maintained without two functioning gene copies. Similarly, we are now finding that mutations in enhancers—which are in charge of controlling these levels and patterns of gene expression—can cause almost as much dysfunction as mutation of the gene itself. These three articles illustrate new methods that are beginning to provide a catalog of functional elements regulating gene expression. This already provides glimpses of a landscape onto which we will be able to map increasing numbers of genetic variants that underlie differences in brain function between normal people, in developmental brain disorders, and as a dynamic target of human brain evolution.


Share
Tweet
Forward
Copyright © 2020 Lieber Institute for Brain Development, All rights reserved.


Want to change how you receive these emails?
You can update your preferences or unsubscribe from this list.

Email Marketing Powered by Mailchimp