Inhibition of Histone Deacetylation Induces Constitutive Derepression of the Beta Interferon Promoter [...] , 2001, Shestakova et al.

jnmaciuch

Senior Member (Voting Rights)

Inhibition of Histone Deacetylation Induces Constitutive Derepression of the Beta Interferon Promoter and Confers Antiviral Activity​

Elena Shestakova, Marie-Thérèse Bandu, Janine Doly, Eliette Bonnefoy

Abstract
[Paragraph breaks added]
The induction of alpha/beta interferon (IFN-α/β) genes constitutes one of the first responses of the cell to virus infection. The IFN-β gene is constitutively repressed in uninfected cells and is transiently activated after virus infection. In this work we demonstrate that histone deacetylation regulates the silent state of the murine IFN-β gene. Using chromatin immunoprecipitation (ChIP) assays, we show a direct in vivo correlation between the transcriptionally silent state and a state of hypoacetylation of histone H4 on the IFN-β promoter region.

Trichostatin A (TSA), a specific inhibitor of histone deacetylases, induced strong, constitutive derepression of the murine IFN-β promoter stably integrated into a chromatin context, as well as the hyperacetylation of histone H4, without requiring de novo protein synthesis. We also show in this work that TSA treatment strongly enhances the endogenous IFN level and confers an antiviral state to murine fibroblastic L929 cells.

Inhibition of histone deacetylation with TSA protected the cells against the lost of viability induced by vesicular stomatitis virus (VSV) and inhibited VSV multiplication. Using antibodies neutralizing IFN-α/β, we show that the antiviral state induced by TSA is due to TSA-induced IFN production. The demonstration of the predominant role of histone deacetylation during the regulation of the constitutive repressed state of the IFN-β promoter constitutes an interesting advance on the understanding of the negative regulation of this gene and opens up the possibility of new therapeutic perspectives.

Link | PDF (Open access)
 
Explain like I'm brain-foggy:

Type I interferons can be produced by nearly all cell types. They are one of the main cellular responses to viral infection, triggered by detection of viral DNA or RNA. In order to quickly ramp up interferon production when needed, cells maintain a constitutive (i.e. constant, without a trigger) level of interferon production. Since interferon generates its own positive feedback loop with interferon stimulated genes [edit: aka ISGs], constitutive interferon production is actively suppressed by several mechanisms to keep it at low levels.

The main difference between interferon production during this constitutive state vs. during active infection is the transcriptional regulation of the interferon gene--meaning the particular transcription factors [edit: bound to the DNA] and local chromatin accessibility (all of which determine how much the gene encoding interferon alpha/beta gets transcribed).

This study determined that a particular epigenetic modification--histone acetylation, which opens up a region of DNA and makes it more accessible for transcription--is an important regulator of constitutive interferon response. Specifically, they added a drug (Trichostatin A) that prevents histone deacetylation to a cell culture and found that this resulted in a much higher level of constitutive interferon production and better protection against later viral infections.

An interesting finding is that histone deacetylation appeared to be necessary for transitioning between "active" interferon production during viral infection to suppressed constitutive interferon production.
 
Last edited:
Recently I've been interested in the specific biological processes that "change over" in cells once an infection clears. For the vast majority of immune responses, the "active" state is continually triggered by ongoing detection of a virus [edit: and, of course, downstream cascades from that detection], and so those responses just fade away once the virus is no longer detected.

However, the interferon response is an interesting example of an immune response that never fully shuts off. Therefore, the question of how cells make that switch between the high vs. low interferon production states could be a potentially relevant area for post-viral illnesses. I'll also note that the mechanism identified in this paper is only one of many, [edit: and there are a couple of differences between mouse and human].
 
Last edited:
So, we could look at the interferon B gene to see if ME/CFS cells are functioning as if there is a viral threat - or not? The histone deacylation could be quantified?

What sort of samples would be needed? Would blood cells be okay, or could it be that there are some tissue cells somewhere that are carrying on, either correctly reacting to a chronic infection, or somehow didn't get the memo that an infection was over and they could stand down? And they are pumping out interferon that acts only locally? In which case, biopsying all the possibilities becomes impossible. Would it be worth looking at gut biopsies? Is this an investigation that needs a dead body?

I saw that the immune system of mice differs in some ways from humans. Is there potential to work with an animal model to see what happens when the IFN-B switch stays on after a viral infection?

This article seems like good review of interferon - link. There seems to be lots of different sorts of interferons. Perhaps it's another sort of interferon that didn't get the 'stand down' memo?
 
So, we could look at the interferon B gene to see if ME/CFS cells are functioning as if there is a viral threat - or not? The histone deacylation could be quantified?
I'm already on the case ;). I can't give many details since some of the data and work involved may come from other teams, but to generally answer your questions, yes and yes.

I think the key is that ME/CFS cells would not necessarily be functioning as if there is a viral threat. The difference between "active" interferon production and constitutive levels is quite vast (here's a good review if anyone's interested). Getting "stuck" in the viral infection pedal-to-the-metal interferon loop is characteristic of interferonopathies, and that features uncontrolled inflammation and substantial tissue damage which is lethal if untreated.

However, this histone acetylation mechanism could very well lead to a level of interferon that is higher than the tightly-suppressed normal baseline without any pathogenic trigger--high enough to cause chronic illness symptoms--but low enough to avoid tissue damage. We already know that such a threshold exists from people who go on interferon therapies. And histone acetylation is only one option for a mechanism that has gone awry. It could very well be one of the other suppressive transcription factors, or some protein kinase or ubiquitin ligase that modulates those transcription factors...who's to say it would even be the same thing for every pwME?

Histone deacetylation could be measured via ChIP-seq as in this study, or you could measure chromatin accessibility of interferon genes via ATAC-seq, or you could just go straight to the hypothesized pathological mechanism and measure interferon/interferon-stimulated genes directly.

What sort of samples would be needed? Would blood cells be okay, or could it be that there are some tissue cells somewhere that are carrying on, either correctly reacting to a chronic infection, or somehow didn't get the memo that an infection was over and they could stand down? And they are pumping out interferon that acts only locally? In which case, biopsying all the possibilities becomes impossible. Would it be with looking at gut biopsies? Is this an investigation that needs a dead body?
My sense is that between Rob Phair's negative preliminary data and the PBMC studies we already have, it's unlikely to be a strong signature in the blood. In PBMCs from lupus, where we know there's high levels of interferon in the blood, you see a very clear signature of interferon stimulated genes across many circulating cell types. But higher-than-normal constitutive interferon levels in tissue cells has already been observed in other illnesses--most notably interferon kappa in skin cells in lupus. So it's a short leap to think that something similar may be happening in one or more specific tissues in ME/CFS.

I've mentioned on other threads that I'm really interested in muscle cells. It would be the easiest to collect (besides skin, but it doesn't seem like skin issues are very common in ME/CFS), and I think there's enough justification in the literature to suspect that something is probably happening in muscle tissue (even if exact findings haven't been replicated).

I saw that the immune system of mice differs in some ways from humans. Is there potential to work with an animal model to see what happens when the IFN-B switch stays on after a viral infection?
Theoretically yes! One of the most commonly used mouse models for psoriasis is an IRF2 knockout, which is an interferon stimulated gene that functions as the main suppressive transcription factor for type I interferons. But like you already mentioned, there are tons of different interferons, and even more different mechansisms by which they are regulated, many of which are specific to certain cell types. I think it's unlikely that IRF2 is our specific culprit given its strong association with psoriasis, but there are plenty of other candidates which would result in different symptom patterns.

I think the first step would be to confirm whether higher-than-normal constitutive interferon production is actually occurring in ME/CFS tissue across many participants. Then you could do a deep dive into samples from a handful of pwME to see if you can pinpoint at least one "upstream" dysregulated protein or pathway that leads to an ME/CFS phenotype (even if it isn't representative of all pwME). Once that's identified, then you could try knocking out that protein in a mouse model and see if it induces ME/CFS-like behavior and recapitulates other findings in the field. That's my long-term plan, anyways ;)
 
Thanks so much for you detailed and helpful reply @jnmaciuch. I really hope you get support for your proposed study.

Yes, as you say, the very similar symptoms people got after interferon treatment is a good clue that interferon of some sort is part of the pathology of ME/CFS.

Could mass spectroscopy be used to assess interferon levels (and differentiate different sorts of interferon quickly)? I see that it can be applied to even a single cell, so a biopsy need not be highly invasive.
 
I really hope you get support for your proposed study.
Thank you! I'm really doing everything I can to make it happen, I just need a lucky break or two.

Could mass spectroscopy be used to assess interferon levels (and differentiate different sorts of interferon quickly)? I see that it can be applied to even a single cell, making a biopsy less invasive.
Mass spec would unfortunately not work here since interferons are notoriously difficult to detect. They are normally present at quantities that are orders of magnitude lower than other cytokines typically detected by mass spec--the reason they have any physiologically relevant effect at all is because the amplification of that positive feedback loop is so powerful.

However, there are other methods to detect it. ELISpot is considered the best option nowadays [edit: (and is what Rob Phair used on blood samples, I believe)], but even then you might be below the level of detection. The most straightforward way is actually to just to do transcriptomics [edit: on a needle biopsy sample or cultured muscle cells from that needle biopsy]. Sometimes you get lucky and detect mRNA transcripts of the interferons themselves. But even if not, the set of downstream genes triggered by interferon is pretty well defined and their expression levels will be higher than interferon itself.

It's not the preferred method since the presence of mRNA doesn't always mean those proteins are being produced, but if you see the whole suite of interferon-stimulated genes being differentially expressed, that's a pretty reliable indicator.
 
Has there been transcriptomics done on muscle cells of people with ME/CFS already?
I was shocked to discover that the answer was no! Hence why I'm very eager to get this going.

Maureen Hanson's group is currently doing spatial transcriptomics on muscle biopsies, but I'm pretty sure the spatial transcriptomic method they're using requires pre-selecting a very small amount of genes you want to measure. So unless they're already 10 steps ahead of me on this interferon hypothesis I don't think they would detect it.
 
Wiggle,

I also find this from Cornell. it references a paper that describes a methodology….

That's from Maureen Hanson's group. Spatial transcriptomics is a great technology because you can see the actual spatial arrangements of different cells with the levels of genes they are expressing, which can sometimes allow you to infer cell-to-cell interactions. But it is a very new technology with very limited capacity. You pretty much have to know exactly which genes you expect to see before you go in, which makes it no longer an "untargeted" approach like bulk or single-cell transcriptomics. Even if they happened to screen for some interferon-stimulated genes, I don't think they would get enough read depth to detect a statistically significant difference.
 
That's from Maureen Hanson's group. Spatial transcriptomics is a great technology because you can see the actual spatial arrangements of different cells with the levels of genes they are expressing, which can sometimes allow you to infer cell-to-cell interactions. But it is a very new technology with very limited capacity. You pretty much have to know exactly which genes you expect to see before you go in, which makes it no longer an "untargeted" approach like bulk or single-cell transcriptomics. Even if they happened to screen for some interferon-stimulated genes, I don't think they would get enough read depth to detect a statistically significant difference.
The article states they are also doing single cell analysis as well to see what genes are present. Also, the text lists the researchers working on the project - perhaps they might answer an email?
 
Ah @wigglethemouse you might be referring to this on the Cornell page:
The goal is to identify molecular and cellular alterations present in ME/CFS skeletal muscles through innovative approaches to capture the transcriptome and epigenome with single-cell and spatial resolution in human biopsies. Using single nuclei isolated from muscle biopsies, the genes expressed in each cell, as well as the configuration of chromosomes in each cell, will be determined. Gene expression information will also be obtained from small regions of cross-sections (slices) of muscle tissue (spatial transcriptomics). All of this information will be used to determine whether specific types of muscle-resident cells are dysregulated at the transcriptional and epigenetic levels in ME/CFS subjects vs. controls.
Apologies on behalf of the field for the confusing terminology. It's talking about "single cell resolution" here because spatial transcriptomics is inherently done on a "cell-by-cell" basis. But that is actually different from scRNA-seq, which involves sample digestion and [edit: untargeted primer ligation]. The method that they're using here still uses spatial probes for a limited gene set, which has the same read depth problems I was referring to.
 
Last edited:
Back
Top Bottom