Jump to content

Continuing the Conversation - Q&As from the Conference


Andrew Chen
 Share

Recommended Posts

Andrew Chen

During last month's virtual conference on Epigenetics in Neurobiology, we received many questions that unfortunately could not be answered live due to time constraint. The presenters were so excited to see all your great questions come in and wanted to make sure they could answer as many as possible, so they wrote out their responses here!

Session 1: Epigenetics as a Link Among Genes, the Environment, and Behavior

1. Can you say something about whether learned information acquired by parents can be transgenerationally transferred via epigenetic mechanisms?

Though fear responses, drug preferences, stress responses, immune responses and metabolic effects acquired by parents can be transmitted to offspring – the mechanisms are unclear and it is unlikely that complex memories can be transmitted. 

EXAMPLE: Dias BG, Ressler KJ (2014) Parental olfactory experience influences behavior and neural structure in subsequent generations. Nat Neurosci. 17(1):89-96.

2. What is the effect of breast feeding and pumping on the neuronal development?

The protein, hormone and immune contents of breast milk have been associated with beneficial effects on neurodevelopment.  In addition, the increased sensory and social interactions between caregiver and infant during breast feeding promote the development of neural systems involved in stress and emotional regulation.

Brown B. M. (2017) The Science of Breastfeeding and Brain Development. Breastfeed Med. 12(8):459-461.

3. As the mind is an emergent phenomena and not "determined" by the genetic underpinnings, isn't epigenetics/genetics more about coarse or potential influences where specific expressions may be facilitated by these influences but remain determined by the optimization of the self-state to environmental-state?

It’s important to separate genetics and epigenetics when considering this question.  Any biological measure is typically only probabilistic, not deterministic, in its prediction of an outcome. In the case of epigenetics, self-state or environmental state can lead to epigenetic changes – these are highly dynamic biological pathways.  This dynamic nature makes it difficult to determine the causal direction of the epigenetic-phenotype relationships. 

4. How would you determine whether transgenerational inheritance is truly…

This is a significant challenge.  In the case of paternal transgenerational effects, using the sperm to generate the offspring generations, and seeing the effects persist several generations is suggestive of a transgenerational inheritance – but for caveats/challenges, see:

Champagne FA (2019) Interplay between paternal germline and maternal effects in shaping development: the overlooked importance of behavioral ecology.  Functional Ecology 00:1– 13.

5. Thank you for such great talks. My question is about the impact of intermittent fasting on the epigenome- do you think it's possible diet can impact DNAme or acetylation which can affect neuronal regeneration after axonal injury?

Nutritional effects on the epigenome are very likely – with either acute or long-term effects, depending on the nutritional exposure.  There is evidence suggestive that these manipulations can have anti-aging effects, so it is likely that there are widespread biological implications.

Yong-Quan Ng G, Yang-Wei Fann D, Jo DG, Sobey CG, Arumugam TV (2019) Dietary Restriction and Epigenetics: Part I. Cond Med. 2(6):284-299.

6. Is there any evidence that early life stress during sensitive period can be reversed with late nurturing and care?

 In animal models (particularly rodents), the period of parental nurturing of offspring is very short and so this does not lend itself well to studying reversal via nurturing.  However, reversal is possible (using social enrichment in adolescence) and in humans, children who have been raised in institutions characterized by a lack of nurturing can experience considerable recovery when adopted into nurturing families.

Troller-Renfree S, McDermott JM, Nelson CA, Zeanah CH, Fox NA (2015) The effects of early foster care intervention on attention biases in previously institutionalized children in Romania. Dev Sci. 18(5):713-22.

7. Is there any known proteins or other epigenetics related materials that are transferred to the offspring from parental cells? And are there any specific gene promoters that are always methylated during mitosis and meiosis?

Incomplete erasure of DNA methylation within the parental genome (transferred to offspring), retained histones (and their associated post-translational modifications) and microRNA may be directly transferred from parental to offspring cells. For an annotation of the DNA methylation levels within the genome, see: https://www.encodeproject.org/

However, individual differences may exist in the fidelity of DNA methylation levels between cells.

8. One constraint of rodent literature is that rodents have limited complexity in their adult environment, and so potential reversibility of epigenetic effects of early life conditions can be hard to estimate (i.e. often effects are found to be reversible through environmental enrichment but how this relates to the human condition is unclear). Is this a valid concern or are there approaches we might be able to take while thinking about translation of rodent findings?

This is a very valid point and an important consideration in translation of study findings.  The rodent literature is based on research designs that seek to remove much of the variability of the environment in an effort to make causal inferences.  However, rodent environments can be made more complex (even in the lab) and field studies provide a more naturalistic context to examine the stability and plasticity of environmental effects.

9. It was recently discovered that bone marrow transplantation leads to the replacement of the recipient's sperm DNA with the donor's. Could this suggest that circulating stem cells become sperm cells, and that this could play a role in transgenerational epigenetic programming?

Chimerism of genetic make-up is likely far more common than typically thought and certainly this could lead to phenotypic effects across generations.  There is an older literature on these effects (see Telegony) and in addition to the stem cell work referred to, it appears there can be transfer between twins in utero.

Ross CN, French JA, Ortí G (2007) Germ-line chimerism and paternal care in marmosets (Callithrix kuhlii). Proc Natl Acad Sci U S A. 104(15):6278-82.

Crean AJ, Kopps AM, Bonduriansky R (2014) Revisiting telegony: offspring inherit an acquired characteristic of their mother's previous mate. Ecol Lett. 17(12):1545-52.

10. Are there correlational studies that might support inherited memories in people. For example, people talk about instincts really being inherited memories (we're afraid of things that pose no risk to us, but that were a threat to our ancestors. I'm a science writer, not a scientist, so please forgive my use of simple language.

There are a number of processes that could lead to the phenomenon of fears being propagated across generations, including germline and non-germline (for example, prenatal maternal stress).  It’s unlikely that highly complex memories are transmitted via germline routes, since these would then have to be recapitulated within the developing brain. However, odor memories and generalized fear states certainly could be transferred across generations with a high degree of fidelity.

EXAMPLE: Dias BG, Ressler KJ (2014) Parental olfactory experience influences behavior and neural structure in subsequent generations. Nat Neurosci. 17(1):89-96.

Also:  Maternal Effects As Adaptations (1998) Mousseau & Fox

11. I am a Neuroscientist from Africa, with interest in Synaptic Plasticity, Learning and Memory. Does epigenetic effects of enriched environment in a generation cause epigenetic modifications in future generations?

There is certainly evidence to support this phenomenon:

Benito E, Kerimoglu C, Ramachandran B, Pena-Centeno T, Jain G, Stilling RM, Islam MR, Capece V, Zhou Q, Edbauer D, Dean C, Fischer A (2018) RNA-Dependent Intergenerational Inheritance of Enhanced Synaptic Plasticity after Environmental Enrichment. Cell Rep. 23(2):546-554. doi: 10.1016

Cutuli D, Berretta E, Laricchiuta D, Caporali P, Gelfo F, Petrosini L (2018) Pre-reproductive Parental Enriching Experiences Influence Progeny's Developmental Trajectories. Front Behav Neurosci. 12:254.

12. Do stress response in Parent and thereby in F1 generation transfer to F2 generation?

Stress exposure in the F0 generation (parent) can be observed in F1 and F2 generations on behavioral measures of stress reactivity.

EXAMPLE:  Morgan CP, Bale TL (2011) Early prenatal stress epigenetically programs dysmasculinization in second-generation offspring via the paternal lineage. J Neurosci. 31(33):11748-55.

13. Can the epigenetic transference be reversed in offspring by modulating behavior such as maternal LG?

Maternal care (including LG) can modulate the transmission of paternal effects on offspring development, see Mashoodh R, Habrylo IB, Gudsnuk KM, Pelle G, Champagne FA (2018) Maternal modulation of paternal effects on offspring development. Proc Biol Sci. 285(1874).

14. Are there any insights or studies regarding how parental enrichment or learning or any positive experience affect offspring?

There is a growing literature on parental enrichment effects and their impact on offspring:

Benito E, Kerimoglu C, Ramachandran B, Pena-Centeno T, Jain G, Stilling RM, Islam MR, Capece V, Zhou Q, Edbauer D, Dean C, Fischer A (2018) RNA-Dependent Intergenerational Inheritance of Enhanced Synaptic Plasticity after Environmental Enrichment. Cell Rep. 23(2):546-554. doi: 10.1016

Cutuli D, Berretta E, Laricchiuta D, Caporali P, Gelfo F, Petrosini L (2018) Pre-reproductive Parental Enriching Experiences Influence Progeny's Developmental Trajectories. Front Behav Neurosci. 12:254.

15. Why some orphans with lack of nurturing are better able to handle stress while some individuals are not and can you comment on drug abuse and addiction in the context of nurturing?

There are studies exploring the resiliency of individuals who experience early deprivation – and certainly foster care can support improved developmental outcomes.  Genetics could contribute to instances of resilience.  Other sources of social support may also augment the effects of caregiver deprivation.  Childhood maltreatment is associated with increased risk of drug abuse – and parenting styles can influence these outcomes as well.

Humphreys KL, Miron D, McLaughlin KA, Sheridan MA, Nelson CA, Fox NA, Zeanah CH (2018) Foster care promotes adaptive functioning in early adolescence among children who experienced severe, early deprivation. J Child Psychol Psychiatry. 59(7):811-821.

Niitsu K, Rice MJ, Houfek JF, Stoltenberg SF, Kupzyk KA, Barron CR (2019) A Systematic Review of Genetic Influence on Psychological Resilience. Biol Res Nurs. 21(1):61-71.

Eun JD, Paksarian D, He JP, Merikangas KR (2018) Parenting style and mental disorders in a nationally representative sample of US adolescents. Soc Psychiatry Psychiatr Epidemiol. 53(1):11-20.

16. What are the key stages of life that are more susceptible to epigenetic modifications in newborn vs adolescence etc

There appears to be lifelong epigenetic plasticity – however, early in development, there are higher levels of epigenetic plasticity as part of ongoing developmental processes (e.g. cell proliferation, differentiation, synapse formation) and disruption to these processes can have widespread consequences.  During adolescence, the hormonal changes occurring can impact the epigenome – however, we need to learn more about how the epigenetic changes associated with this period can impact outcomes.

17. How can we influence behavior with the understanding of epigenetic?

In the lab, we can now use epigenetic tools (pharmacological, genetic) to manipulate behavior and many of the drugs that are in use to treat behavioral/psychiatric disorders in humans target the epigenome.  If these epigenetic tools are able to target more precisely those gene(s) that regulate behavioral outcomes, then they may have potential for broader therapeutic use.

18. Can external effects such as pharmacological interventions result in altering biochemical/phsysiological baselines that result in long lasting responses to future stimuli?

The idea that transient stimuli like pharmacological interventions could have persistent effects on brain function is one of the most exciting possibilities suggested by epigenetic mechanisms of gene transcription. Some of the best evidence that this could be physiologically relevant has come from experimental animal studies with repeated exposure to psychostimulants or antidepressants. These drugs act rapidly at a pharmacological level to change levels of neuromodulators at synapses but they must be taken chronically to change behavior. Substantial data shows that these drugs can alter the epigenome, and ongoing studies are addressing whether those changes are causative for the persistent alterations in behavior.

19. What technological challenge do you think is the greatest hindrance to answering epigenetic questions in neurons?

 I think there are a few major challenges and my colleague Alex Nord and I have written more about this in our review PMID 31740812. A major one is that fact that epigenetic modifications of chromatin are a key part of making cell types different from one another. Therefore, any study that seeks to understand how the epigenome changes must find a way to interpret the findings in light of the huge heterogeneity of cell types in the brain. Large scale studies that are profiling the epigenome in neuronal cell types are helping to fill in these gaps. A second is the fact that it is easier to profile the epigenome than it is to overcome the challenge of experimentally determining whether and how changes to the chromatin landscape are actually meaningful either for changes in gene expression or neuronal function. Local experimental modifications of the epigenome and chromatin architecture that can be induced using dCas/CRISPR technology offer an important opportunity to address this question. A third major challenge is understanding the relevance of chromatin regulation for human disease. This is because the non-coding genome is far less conserved between model organisms and humans compared with the coding genome, yet we lack good human model systems for the study of brain function. iPSCs and organoids are good steps in this direction, but obviously you cannot study behavior in a culture dish.

20. How exactly are you able to take a peripheral cell and tell that it's epigenome, with respect to its neuronal genes, is due to environment rather than just the fact that it isn't a neuron?

You are exactly correct that one must take cell type into account in any evaluation of the epigenome, because chromatin regulation is the major way a single genome can be used to make many types of cells. Although the definition of what constitutes a neuron involves much more than just gene expression, programs of gene expression do underlie the cellular features of neurons and transcriptional patterns can give a good guess as to whether a cell is likely to be a neuron or not. As to the consequences of the environment on the epigenome, it is important to conduct experiments in which the environment is varied in a controlled way in a constant cell population to see what aspects of the epigenome are constant and which are dynamic.

21. Have mouse experiments shown correlation between peripheral tissues and neurobehavioural outcomes?

I think the question you are asking is whether changes in the function of organs of than the brain itself might contribute to behavior? This is definitely the case, and something to consider in all in vivo studies. For example germline mutations that affect brain and also affect muscle could impact motor-driven behaviors through non-neuronal effects on the organism.

22. What do we know about the interaction between epigenetic regulation and genetic variation (polygenic risk)?

This is an important question. Johannes Graff in session 3 of this meeting tells a very interesting story about how a variation near a gene implicated in Alzheimer’s disease influences the expression and function of this gene via changes in chromatin architecture. When genetic variations occur within transcriptional enhancers these can have direct effects on gene regulation. Genetic variations can also occur in between gene regulatory elements in way that effect chromatin architecture – the FKBP5 gene is an interesting example in this regard. Finally, genetic variations that affect transcription can feedforward to change the epigenome because chromatin regulators are recruited to genes based in part on their transcriptional activity. Methylation of the long repeat sequence in the FMR1 gene in Fragile X mental retardation is an extreme example of this process.

23. For Anne West: In the experiment where you showed transient placement of H3K27ac on grin2b, do you have a proposed mechanism of regulation for the acetylation writer so that it is recruited during that time (transcription, post-translation modification?)

We don’t know the mechanism that underlies the transient activation of enhancer on the Grin2b gene. DNA binding transcription factors are the major factors that coordinate gene transcription. We predict that the transiently activated enhancers are either bound by transiently expressed transcription factors or that the transcription factors that bind these sites are activated by signaling pathways that occur in a transient, developmentally regulated manner.

24. You mentioned that brain tissue is not normally available for epigenetic analysis. What about when brain tissue is removed during surgical procedures and fixed or frozen. Can these tissues be investigated for epigenetic profiles related to the disease.

Human brain tissue removed from surgery or harvested at autopsy is an important source for the study of the human epigenome. Obviously, these are in limited quantities, but they are nonetheless an essential piece of the research landscape. The challenges with these tissues are 1) tissues removed during surgery may well have been removed because of a pathological process (tumors, epilepsy) and therefore may not represent the normal brain, 2) human tissue has a mixture of cell types that is likely to vary between samples, and 3) the time from harvest to processing can vary and can affect the quality of the samples.

25. Is it possible to use epigenetics to modulate cognition? For example, intentionally modify certain gene(s) expression to increase cognitive capacity in an animal model?

This question reminds me of the Grin2b transgenic “smart mouse” made by Joe Tsien and colleagues. This study was intended to address the contribution of these NMDA receptor subunits not only to study synaptic plasticity but also learning and memory. Indeed it would be interesting to use genetic methods that experimentally manipulate the epigenome, such as CRISPR/dCas9-based epigenome editing, to control the epigenome at specific regulatory elements and determine the effects on cognition.

26. Is the H3 methylation function , can be restricted to a region or to the whole genome? there are mono, di and three methylation, there methylases associated to a particular "intensity" of the methylation is there a difference in the silencing/activation for the degree of methylation? And cofactors, FAD is one of them, is there the link to "metabolic state" and methylation

The methylation of histone H3 can occur at many different amino acids (e.g. lysine 4, lysine 9, lysine 27, lysine 36). Each of these has distinct consequences for the regulation of transcription because the different methylation events act to dock distinct complexes of transcriptional coregulators. There are a large set of enzymes that add or remove histone methylation, and these enzymes have specificity for the mono, di, or trimethylation of specific histones at specific amino acids. Different histone methylation states have distinct distributions across the genome due to the differential recruitment of the regulatory enzymes onto the genome - for example H3K4me3 is enriched at gene promoters and H3K36me3 is enriched in gene bodies of expressed genes. As you suggest, there is a relationship between methylation and cellular metabolism with the possibility that global shifts in histone methylation across the genome may be driven by changes in metabolic state.

27. For these methylations that occur based on experience, how stable are these alterations? Does it depend on whether there is more environmentally exposure or will the methylation be removed by some "homeostasis" mechanism to go back to the original genome?

The regulation of both histone and DNA methylation has long been thought to be relatively stable over time. This is both because the free energy of demethylation is high and because methylation can drive the recruitment of protein complexes that locally reinforce the methylated state.  However under some conditions the balance of biochemical processes that maintain stable methylation can be disturbed to result in dynamic changes. Whether or not the changed state will be maintained versus returning to baseline depends on the nature of the change. For example, DNA methylation patterns can be erased and re-established in new patterns when fibroblasts are transdifferentiated to neurons.

28. How many different types of epigenetic mechanisms interact to influence each other for indirect regulation of gene expression regulation? Are there other examples of this?

Epigenetic mechanisms of gene transcription are highly intertwined with respect to the regulation of gene expression. For example, non-coding RNAs help to recruit of the enzymes that methylate histone H3 at lysine 27, and the methylation status of CTCF binding insulator sites helps to determine which promoter-enhancer pairs are able to functionally interact in cells. Discovering the full set of regulatory mechanisms that control any single gene is an important area of research in this field. 

Session 2: New Methods and Directions for Psychiatric Disease

1. How does dopamine, amphetamine move into the nucleus? there is no known transporter in the nuclear membrane

While we originally believed that monoamine entry into the nucleus was achieved via passive diffusion, we now have data demonstrating that this actually occurs via a transport dependent mechanism. There are indeed select examples of high capacity, low affinity monoamine transporters that exist on the nuclear membrane, and we have found that their inhibition blocks monoamine uptake into the nucleus and inhibits deposition of the PTMs.

2. Any plans to look at how antipsychotic drugs affect histone dopaminylation?

We have not yet begun looking into whether antipsychotic drugs may affect these marks, but we do have a manuscript nearing submission that has examined regulation of the serotonyl modifications in response to chronic stress and reversals by antidepressant treatments. So be on the lookout.

3. When thinking about common comorbid psychiatric conditions, how would you interpret the impact of reducing a hyper monamine state in someone recently treated for a SUD, who is taking, for example, SSRIs/SNRIs for depression?

We have not yet begun looking into comorbidities, but this is a very interesting question. Of note, however, we do find similar patterns of regulation to what we see with H3 dopaminyl in the context of cocaine taking/withdrawal with alterations in H3 serotonyl in the context of chronic stress and regulation by antidepressants.

4. What are your thoughts on potential causal mechanisms for histone modification by monoamines, including specificity (why some histones are modified at certain genes but only under certain conditions)?

Very good question…we are currently exploring TGM2 (the monoaminylase) associated binding complexes in vivo, which we believe will shed greater light on the locus specificity of these PTMs.

5. Can dopaminylation affect the dynamics of proteins like HCN channels?

We do not have any data related to this.

6. What happens to dopamine secretion by VTA cells?

If you’re asking how our manipulations affect DA release by VTA neurons, we found that inhibiting H3Q5dop leads to reduced DA release into NAc in response to drug cues.

7. Any plans to look at how antipsychotic drugs affect histone dopaminylation?

We have not yet begun looking into whether antipsychotic drugs may affect these marks, but we do have a manuscript nearing submission that has examined regulation of the serotonyl modifications in response to chronic stress and reversals by antidepressant treatments. So be on the lookout.

8. When thinking about common comorbid psychiatric conditions, how would you interpret the impact of reducing a hyper monamine state in someone recently treated for a SUD, who is taking, for example, SSRIs/SNRIs for depression?

We have not yet begun looking into comorbidities, but this is a very interesting question. Of note, however, we do find similar patterns of regulation to what we see with H3 dopaminyl in the context of cocaine taking/withdrawal with alterations in H3 serotonyl in the context of chronic stress and regulation by antidepressants.

9. What are your thoughts on potential causal mechanisms for histone modification by monoamines, including specificity (why some histones are modified at certain genes but only under certain conditions)?

Very good question…we are currently exploring TGM2 (the monoaminylase) associated binding complexes in vivo, which we believe will shed greater light on the locus specificity of these PTMs.

10. Are there paradigms for the serotonergic system that give insight into processes modified by serotonin?

Yes, we have begun to explore regulation of H3 serotonylation in the DRN in the context of chronic stress/antidepressant treatments, as well as in postmortem tissues from subjects with major depressive disorder

11. Do you know what enzymes are involved in these reactions?

Yes…TGM2 is the monoaminylase. We also know of the removal mechanism, but I cannot comment on this further at this time. Be on the lookout for upcoming publications.

12. Blocking H3Q5Dop accumulation also reduced drug-seeking behavior?

Yes

13. Any data yet on the implications for medical use of SSRIs, L-DOPA etc.?

We do have a manuscript nearing submission that has examined regulation of the serotonyl modifications in response to chronic stress and reversals by antidepressant treatments. So be on the lookout. Nothing yet with L-DOPA.

14. How was the level of dopamine measured intracellularly?

DA ELISAs on fractionated nuclei

15. What are the best present methods for determining whether epigenetic effects have causal functional roles in animal behavior and disease processes?

Addressed during the Q&A (e.g., intein chemistry based methods, CRISPR targeting of chromatin modifying domains/enzymes to specific loci, etc.)

16. epigenetic and anxiety, depression?

We do have a manuscript nearing submission that has examined regulation of the serotonyl modifications in response to chronic stress and reversals by antidepressant treatments. So be on the lookout.

17. Chromatin conformation is dynamic and there are different levels of "loops" and their tissue specificity and conservation. How do you filter HiC data to find chromatin interactions that provide reliable, replicable readout of distal enhancer annotation to their target genes?

We used chromatin interactions that are anchored to promoters and/or exons. We acknowledge that there are roughly two types of definition for loops: (1) loops that are engaged in TAD boundaries: ~20,000 loops typically discovered, cell-type invariant, (2) statistically significant loops that represent long-range interactions: >100,000 loops typically discovered, cell-type specific. We used the second loop definition because (1) we wanted to connect genetic variants with cell-type specific regulatory landscape, and (2) a typical GWAS identifies association statistics for >1M SNP, the majority of which we wanted to link to cognate genes. Before we combine Hi-C data with SNPs, we used variable functional genomic datasets (enhancers, genes, eQTLs) to verify that chromatin interactions can predict gene regulation (e.g. gene promoters often interact with enhancers, genes connected to a higher number of enhancers are more highly expressed, genes that interact with enhancers are more highly expressed than genes that interact with repressive marks).

18. Can you elaborate more on degenerative diseases show more glial cells affected than neurons in your analysis? Especially for Multiple Scleroisis and Alzheimer’s?

We found that multiple sclerosis-associated genes and Alzheimer’s disease-associated genes were highly expressed in microglia but neurons, suggesting that microglia-mediated (neuro)inflammation may be an important mechanism in understanding pathophysiology of the two disorders.

19. Is the SNP database searchable?

NCBI offers a SNP database that is searchable: https://www.ncbi.nlm.nih.gov/snp/

 20. Question for Hyejung Won: What is mechanism by which cortical neurons are more sensitive than interneurons in polygenic psychiatric disorders; are they molecularly more diverse?

Our findings do not necessarily suggest that glutamatergic neurons are more sensitive to psychiatric disorders. What we suggest is that the genes associated with genetic risk factors of psychiatric disorders are more highly expressed in glutamatergic neurons than in GABAergic neurons. Given that psychiatric disorder associated genes are highly expressed in glutamatergic neurons, we predict that their altered regulation will impact glutamatergic neurons more so than GABAergic neurons. However, your idea – whether glutamatergic neurons are more sensitive because they are molecularly diverse – is super interesting!

 21. Are the risk genes more susceptible to epigenetic intervention to prevent or reverse the disease development? For example, could we silence or change specific gene expression if we knew what and at what point in development those specific genes would play a role in the rise of depression or schizophrenia?

Given that psychiatric disorder associated genes are often enriched for epigenetic regulators that are highly expressed during neurodevelopment, we believe that epigenetic regulation during brain development will be a key player for the majority of psychiatric disorders. Your idea – whether we could alter the risk gene expression in the right developmental window and see its role in the rise of psychiatric symptoms – is indeed a dream experiment.

22. My question is to Dr. Won. Have you come across any loci (from Hi-C data) that is linked to different genes in a disorder specific manner?

This is something that we are definitely interested! However, identifying the distinction between two disorders is much more difficult than identifying the shared pathways, because each GWAS is differently powered. We will be able to ask this question once psychiatric disorder GWAS are more equally powered.  

23. What is the next step to do in GWAS research?

We just began identifying genome-wide significant (GWS) loci from GWAS. Therefore, a larger sample size is required for the majority of GWAS. Given that the current GWAS is largely conducted from European descents, having more samples from diverse ethnicities will also be essential. Once we identify GWS loci, functional genomic datasets can help predict the biological impact of genetic variation, but what they predict is association, not causality. To reveal causal links, we would need experimental validation such as CRISPR/Cas9-based genome engineering or massively parallel reporter assays.  

24. Many of these diseases have a sex bias, did you analyze sex differences in these data?

Sex biases in psychiatric disorder are indeed an important subject. Unfortunately, there is not much sex-specific GWAS available. Even when there are sex-specific GWAS available, often a sample size causes another issue because sex-specific GWAS is less powered than sex-agnostic GWAS. For example, ADHD shows a sex bias between males and females and sex-specific GWAS is available. However, ADHD GWAS with and without sex as a covariate are well correlated, suggesting that we are yet to be powered to identify sex-specific effects from GWAS. Once we have a better powered GWAS in a sex-specific manner, we will be able to employ H-MAGMA to identify potential sex-specific mechanisms in psychiatric disorders.

25. Will H-MAGMA be implemented in FUMA?

That’s what we are discussing with FUMA team right now. For now, you can visit our H-MAGMA website: https://github.com/thewonlab/H-MAGMA.

Session 3: Epigenetics in CNS Plasticity

1. Does your work reveal how the PM20D1 promoter becomes methylated? Is methylation dependent on interaction with rs708727?

The methylation depends on the genotype of rs708727 and on the interaction between the two. That is, if we block the interaction by shRNAs against CTCF, we also alter the DNA methylation. However, we do not know yet how PM20D1 becomes methylated. We presume that it is through the recruitment and interaction of other proteins.

2. Do animals with PM20D1 overexpression have regular accumulation of plaques/tangles?

When we overexpress PM20D1, plaque load is reduced. That is what we reported in the Nat Med paper. However, we have not yet looked at the effects on tangles, in part because APP/PS1 mice do not develop tangles.
 

3. What are the relative effects on PMS20D1 expression of SNP v effect of b-amyloid? Is the effect of the SNP larger?

In order to investigate this, one would have to expose AD genotype-matched cells to amyloid fibrils. This could be done in iNs, but it is something we have not done.
 

4. Do you find different enhancer-promoter interactions depending on the type of neuronal activity (strong vs weak for example)?

That’s something we are currently looking into, how the upstream regulatory loop reacts to different external stimuli, including neuronal activity. We don’t have the answers yet.
 

5. Early life stress has also been hypothesized as a factor in AD. Are there any ideas on how this may impact epigenetic differences in AD?

Generally speaking, there are multiple papers indicating that early life stress can impact the epigenome in AD models (e.g., Lahiri et al. 2009, Sierksma et al. 2013). For PM20D1, there is also one paper indicating that early life stress may alter its expression (Suderman et al. 2014). 
 

6. Is the AA locus in CTCF binding site? Intuitively AA SNP would control

the binding of CTCF at its locus. I was curious as to how AA locus is influencing the binding of CTCF at the DMR region. Do you have any insights on this?

The rs708727 SNP doesn’t fall within a CTCF binding site, but there is one adjacent to it. Our ChIP analysis shows that even so, CTCF binding is reduced in AA carriers, although we haven’t investigated the underlying mechanism in detail.
 

7. Have you looking into whether the motif containing rs708727 is generalizable to other sites in the genome?

Not yet, but that’s an excellent idea. 
 

8. What are some other notable mechanisms linking toxic conformers of neurodegenerative diseases (like AD) with epigenetic modulation?

HDAC2, for example, has been shown to be unregulated following oxidative stress and amyloid exposure via the transcription factor GR1 (Gräff et al., 2012). Conversely, HDACis have been shown to reduce amyloid load (Volmar et al. 2017), and to rescue against HD toxicity (Siebzehnrübl et al., 2018). 
 

9. How does rs708727 develop specificity for PM20D1? Are other genes in the vicinity inactive and protected from interactions with rs708727?

We have been intrigued by this question as well. First, the loop seems to be specific for PM20D1 (Sanchez-Mut et al. 2018). Second, the methylation is also specific for PM20D1, as well as the expression change (Sanchez-Mut et al. 2020). This latter study is a follow-up paper that we just published, it precisely addresses this question.

10. Do you find different enhancer-promoter interactions depending on the type of neuronal activity (strong vs weak for example)

While we only tested for genome architecture one pattern of neuronal activation using optogenetics, and agree that modulating the strength of stimulation would be interesting. We have tested a battery of sensorimotor stimuli using different behavior paradigms and profiled gene expression in the cerebellum. We found that both common and distinct activity-dependent gene expression programs were regulated with different behaviors, suggesting that diverse activity patterns and changes in brain state (mood, stress, attention, etc) may all have important roles in gene regulation. As suggested, it will be important to determine the roles of enhancer-promoter looping in each of these cases.

11. How is Rad21 recruited to activity dependent genes in an activity dependent manner (post translation modifications? Expression?)

Rad21 and the cohesin complex are recruited to highly activate genes via the cohesin-loading factor Nipbl. While the precise mechanisms in neurons remain to determined, it has been shown in cell lines that Nipbl preferentially binds active gene promoters and enhancers. Once Nipbl loads the cohesin complex, the cohesin ring via loop extrusion would then bring together promoters with active distal enhancers. In our data, activity-regulated genes have strong increases in histone acetylation at both their gene promoters and enhancers. In the most simplistic model, this should be sufficient to recruit the cohesin complex … although surely there might be additional brain-specific mechanisms!

12. Does LTD/LTP in Purkinje cells alter CGC transcription? 

Certainly all signaling in the anterograde (mossy fiber to granule neuron) and retrograde (Purkinje cell to parallel fiber) direction would be expected to regulate long-term changes in granule neurons via gene transcription. However, while this is predicted, we lack experimental evidence that this is true. Current protocols for directly triggering LTP/LTD at parallel fiber/PC synapses involve electrical or optogenetic stimulation of parallel fibers. However, this would also lead to antidromic spiking in the granule neuron soma, thus opening L-type VGCCs and directly driving Ca2+ mediated gene transcription. Dissociating signaling pathways and studying their relative roles on gene transcription remains an active area of molecular neuroscience.

Session 4: Technological Approaches for the Advancement of Epigenetics Research in Neurobiology

1. I am interested in the co expression of neurogenic genes for reprograming. My question is when you screened transcription factors with co-expression of Ascl1 and Ngn3, do you find all neurogenic factors that you found in first analysis in your single factor screening or do you see different factors when you did coexpression?

The list was only partially overlapping.

2. So did the EVs you delivered have both pro and anti-inflammatory miRs and how do miRs that have multiple targets impact specific pathways

The function of miRNAs delivered by EVs will depend on the target genes expressed in the recipient cells as well as the levels/abundance of miRNAs delivered. Other factors such as alternate splicing impacting 3’UTR, presence of RNA binding protein influencing secondary structure of mRNAs limiting access for miRNA binding, can contribute to how pro or anti-inflammatory miRNAs within the EVs mediate gene regulation in a recipient cell type/state-specific manner.

3. What's your vision for using epigenome editing in neurological/neurodegenerative diseases?

It’s a long-term vision, but there are many interesting examples of how this could be used, and also a lot of progress in AAV delivery to the CNS that could facilitate this. As a starting example of how it could be applied, take a look at the pipeline of Sangamo Therapeutics and their recent partnership with Biogen.

4. Can dCas9 library be used in a "state-dependent" or temporally regulated manner?

Yes – the easiest way to do this is to regulate Cas9 expression – either with a Cre-dependent system that could control dCas9 expression with tissue-specific Cre or estrogen-inducible CRe, or a drug- or light-inducible promoter, such as the tetracycline-inducible expression system. There are also several systems in which the dCas9 is directly activated or inactivated by drugs or light – we published one example of this (Polstein et al., Nature Chemical Biology, 2015). Stanley Qi at Stanford and David Liu at Broad have also published some nice inducible dCas9 systems, as have many other groups.

5. This question is for Dr. Ajit, have you looked at other TLR agonist-mediated changes in miRNAs? For example, TLR2 or TLR3? Are there signature miRNAs that are involved in multiple TLR pathways?

We have only looked at LPS induced changes in miRNAs in EVs. Yes, there are miRNAs involved in multiple TLR pathways, for example miR-146, miR-155, miR-147 and miR-9.

6. What technological challenge do you think is the greatest hindrance to answering epigenetic questions in neurons?

Definitely delivery and being able to precisely restrict activity to the cell type of interest. Another technical challenge for epigenome editing is that we still don’t have great predictors for what makes a good or bad gRNA, or whether when epigenome editing fails – is it a technical obstacle or real biology?

7. How were the EVs used in the in vivo studies prepared? In other words, what were the contents? Are they from the stimulated RAW cells?

We used ultracentrifugation for isolating EVs from RAW 264.7 cells. The method and composition of these EVs are described in McDonald et al. Pain. 2014 Aug;155(8):1527-39. doi: 10.1016/j.pain.2014.04.029

We used both unstimulated and LPS stimulated EVs in our studies.

 8. Are micro-droplets expelled when speaking, coughing or sneezing, exosomes? Does coronavirus use exosomes to produce infectious airborne particles?

This is difficult to answer because there are no published reports on this.

We asked Dr. Kenneth W. Witwer, Associate Professor at Johns Hopkins University School of Medicine, and his response was the following:

Since SARS-CoV-2 is a positive-stranded RNA virus, it could theoretically infect a cell by introduction of RNA alone, without viral proteins. However, there is currently no evidence that this happens for this virus. There is certainly no evidence to date that host endosomal origin EVs are involved, or that EVs from other origins would or would not be involved. In any case, a host exosome or ectosome would be a very poor way for a virus like SARS-CoV-2 to replicate. The genome is very large to be efficiently incorporated into a particle without the specific mechanisms that are used for virion formation, and fusion is very inefficient for an EV compared with a virion because of the viral surface proteins.

Link to comment
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in

 Share

×
×
  • Create New...