Brain Endothelial- and Epithelial-Specific Interferon Receptor Chain 1 Drives Virus-Induced Sickness Behavior and Cognitive Impairment, 2016, Blank +

[Hutan]

Link

Spoiler: Authors

Thomas Blank 1 ∙ Claudia N. Detje 2 ∙ Alena Spieß 1 ∙ Nora Hagemeyer 1 ∙ Stefanie M. Brendecke 1 ∙ Jakob Wolfart 3 ∙ Ori Staszewski 1 ∙ Tanja Zöller 1 ∙ Ismini Papageorgiou 4 ∙ Justus Schneider 4 ∙ Ricardo Paricio-Montesinos 1 ∙ Ulrich L.M. Eisel 5∙ Denise Manahan-Vaughan 6 ∙ Stephan Jansen 6 ∙ Stefan Lienenklaus 2, 7 ∙ Bao Lu 8 ∙ Yumiko Imai 9 ∙ Marcus Müller 10 ∙ Susan E. Goelz 11 ∙ Darren P. Baker 12 ∙ Markus Schwaninger 13 ∙ Oliver Kann 4 ∙ Mathias Heikenwalder 14, 15 ∙ Ulrich Kalinke 2∙ Marco Prinz 1, 16

Highlights
• Viruses induce depressive behavior and ISG15 expression at the blood-brain barrier
• IFNAR1 expression on neural cells is not involved in IFN-β-induced sickness behavior
• IFNAR1 expression on brain endothelial and epithelial cells drives behavioral changes
• Brain endothelia- and epithelia-derived CXCL10 inhibits hippocampal synaptic plasticity

Summary
Sickness behavior and cognitive dysfunction occur frequently by unknown mechanisms in virus-infected individuals with malignancies treated with type I interferons (IFNs) and in patients with autoimmune disorders. We found that during sickness behavior, single-stranded RNA viruses, double-stranded RNA ligands, and IFNs shared pathways involving engagement of melanoma differentiation-associated protein 5 (MDA5), retinoic acid-inducible gene 1 (RIG-I), and mitochondrial antiviral signaling protein (MAVS), and subsequently induced IFN responses specifically in brain endothelia and epithelia of mice.

Behavioral alterations were specifically dependent on brain endothelial and epithelial IFN receptor chain 1 (IFNAR). Using gene profiling, we identified that the endothelia-derived chemokine ligand CXCL10 mediated behavioral changes through impairment of synaptic plasticity. These results identified brain endothelial and epithelial cells as natural gatekeepers for virus-induced sickness behavior, demonstrated tissue specific IFNAR engagement, and established the CXCL10-CXCR3 axis as target for the treatment of behavioral changes during virus infection and type I IFN therapy.

Graphical abstract

 

[Hutan]

This paper was mentioned by @jnmaciuch on another thread. It suggests a mechanism for sickness behaviour and cognitive impairment arising during acute infections. Sickness behaviour is the set of behaviours that animals exhibit when they are sick e.g. social withdrawal, reduced activity along with feelings of pain and malaise. Given the similarities between sickness behaviour and ME/CFS, it is interesting to consider whether mechanisms causing it might be relevant to a disease mechanism in ME/CFS.

This is a relatively old paper. Perhaps there are more recent papers that finesse the mechanisms.

 

[Peter T]

This seems to imply there are two components to ill health, the triggering biomedical condition and an additional ‘sickness behaviour’ response on top, and further that it would theoretically be possible to be ‘ill’ without feeling ‘ill’.

A useful comparator might be my experience of bing seasick. A friend who is a very experienced long distanced yachtsman gets seasick, but can unconcernedly continue with the tasks of sailing his boat, just pausing now and again to vomit over the side. In contrast in the same sea conditions I am trapped lying flat below deck with just the thought of trying to move triggering further vomiting. Indeed feeling so bad that the thought of stepping off the boat into the waves is very appealing. Objective measurement of say number of vomiting episodes comparing the two of us might reveal little difference but in practice it has little impact on my friend’s ability to function but I am completely decked feeling as if I am on death’s door.

The suggestion is ‘sickness behaviour’ does not have a direct biological relationship to the situation, rather it is an indirect behavioural control mechanism ensuring we get appropriate levels of rest, are hidden away from potential predators and perhaps less likely to transmit infection to our wider social group.

My knowledge of sleep studies is half a century out of date, but there was also the idea that the purpose of feeling tired and sleeping was similarly a way of regulating our behaviour, that we sleep to keep us quiet at the times we have no need to be active, ensuring we avoid predators, and tiredness functions to enforce sleep.

However I am unsure about this interpretation, it seems to imply that in ME if we chose to we could push through the sickness behaviour, that we could learn it was for us a maladaptive behaviour. In contrast it is the case that for ME fighting ‘sickness behaviour’ results in a worsening of our symptoms, both short term and long term. It might be a short term exaggeration of symptoms is not incompatible with this idea, but I would have thought long term deterioration in the underlying condition is.

 

[Hutan]

If fighting through sickness behaviour did something to make more sickness behaviour (like more interferon? sensitised brain endothelial and epithelial cells? upregulated chemokine), then fighting it would have consequences.

Peripheral IFN-β activated interferon receptor chain 1 (IFNAR) expressed on brain endothelia and epithelia, which released the cytokine CXCL10 into the brain parenchyma where neuronal function was compromised.

Next, we examined the synaptic plasticity of adult hippocampal neurons after challenge with CXCL10. Field potentials were recorded in the stratum radiatum of the hippocampal CA1 subregion by stimulation of the fibers between CA3 and CA1 (Figure 6M). Hippocampal slices were incubated with CXCL10 causing reduced paired-pulse facilitation (ratio of second pulse to the first pulse), which is indicative of impaired presynaptic transmitter release (Zucker and Regehr, 2002). The presence of CXCL10 also markedly weakened synaptic long-term potentiation (LTP), an electrophysiological model of learning and memory (Figure 6N). In freely behaving mice, high-frequency afferent stimulation resulted in robust LTP in the hippocampal CA1 region of vehicle-treated animals and persisted for at least 24 hr. Treatment with CXCL10 resulted in a significant impairment of LTP. The impairment obtained after high-frequency stimulation (HFS) was significant throughout the whole recording period, which lasted for 25 hr

Compromised neuronal function sounds hard to push through.

Many chemokines, especially CXCL9, CXCL10, and CXCL11, were strongly upregulated, whereas no overt induction of genes indicative of pro-inflammatory responses were detectable, suggesting a primary chemokine-mediated response of brain endothelial cells upon IFN-β challenge.

Induction of chemokines was confirmed using qRT-PCR (Figure 6B) and ELISA (Figure 6C) with particularly high release of CXCL10 (Figure 6D). Intra-endothelial production of CXCL10 was evident upon VSV-M2 or virus ligand challenge and after IFN-β incubation (Figure 6E).

In response to VSV-M2 infection, elevated CXCL10 protein concentrations were detectable in brain homogenates, in spleen and in blood serum (Figure S3C). CXCR3 as joint receptor for CXCL9, CXCL10, and CXCL11 was found to be expressed in the brain on neurons and microglia (Figure S5A). Mice lacking CXCR3 or CXCL10 subjected to IFN-β treatment were protected from depressive-like behavior (Figure 6F) and impairment of spatial learning and memory (Figures 6G and 6H),

Elevation of CXCL10 in the brain (Figure S3C) did not cause activation of microglia (Figure S5B) nor the recruitment of immune cells to the brain (Figure S6).

 

[Kitty]

Many chemokines, especially CXCL9, CXCL10, and CXCL11, were strongly upregulated, whereas no overt induction of genes indicative of pro-inflammatory responses were detectable, suggesting a primary chemokine-mediated response of brain endothelial cells upon IFN-β challenge.

The bit I've bolded is interesting, maybe making it harder to detect that something's going on?

 

[jnmaciuch]

Thanks @Hutan for making the thread. I’ve been trying to look for other more recent papers that follow up on this idea but unfortunately haven’t found much yet.

When I have a free moment, I will try to do a layman’s explanation of some of the methodological details to supplement the bits you’ve pulled out already.

 

[jnmaciuch]

The bit I've bolded is interesting, maybe making it harder to detect that something's going on?

Yes that’s something that caught my eye as well—not only because you’d have to look for those specific chemokines if you wanted to detect anything, but also because it gives a potential mechanism for how “sickness behavior” could be without any detectable inflammation. You’d have to either have only interferon and not other cytokines present at the blood brain barrier, or have the endo/epi-thelial cells producing only these chemokines on their own without interferon stimulation.

 

[rvallee]

Sickness behaviour is the set of behaviours that animals exhibit when they are sick e.g. social withdrawal, reduced activity along with feelings of pain and malaise

That seems weird to me. It's the pain and malaise and other symptoms that cause the behavior, not something that happens alongside it.

 

[Utsikt]

That seems weird to me. It's the pain and malaise and other symptoms that cause the behavior, not something that happens alongside it.

We actually don’t know the causal pathways here. I think it’s reasonable that the pain and malaise might cause some changes in behaviour, but I wouldn’t be surprised if disease can cause changes in behaviour directly as well.

 

[jnmaciuch]

Explain like I'm brain-foggy:

This study hypothesized that sickness behavior in mice was induced by interferon. After infection with their model virus (VSV-M2), they detected high levels of ISG15, a gene that gets highly expressed in response to interferon, in brain endothelial cells (the cells making up the blood brain barrier). A mouse model deficient in pDCs, which are the primary cells that produce high amounts of interferon during infection, did not show sickness behavior.

To confirm that it was specifically interferon causing these behavioral differences and not something else from pDCs, they tested a mouse model that does not have the receptor for interferon, and found that it did not display sickness behavior. They confirmed that it was specifically brain endothelial cells detecting interferon and leading to sickness behavior by creating a mouse model where the interferon receptor was knocked out only in those cells.

To figure out what signaling molecules these brain endothelial cells were generating in response to interferon, they cultured these cells with interferon and measured which genes were induced by treatment compared to control cells. Several chemokines (CXCL9, 10, etc.) came up as strongly upregulated. They determined that it was specifically CXCL10 causing these behavioral changes by directly injecting CXCL10 into the brain, which recapitulated the sickness behavior. As further confirmation, mice that lacked the receptor for CXCL10 did not show sickness behavior.

There were a couple more experiments on some tangential questions and verifying that their models worked as expected, but these are the main findings relevant to their conclusions.

 

[duncan]

We cannot query the mice about their motivation.

Fortunately, that limitation does not apply to humans.

So I'd discard the entire sickness behavior model when humans are involved. Too much erudite/academic speculation regardless. We don't do shit because we feel crappy; it's a much more simple model, although admittedly limited by common sense.

 

[Hutan]

A mouse model deficient in pDCs, which are the primary cells that produce high amounts of interferon during infection, did not show sickness behavior.

To confirm that it was specifically interferon causing these behavioral differences and not something else from pDCs, they tested a mouse model that does not have the receptor for interferon, and found that it did not display sickness behavior.

pDCs are plasmacytoid dendritic cells

Plasmacytoid dendritic cells (pDCs) are a rare type of immune cell that are known to secrete large quantities of type 1 interferon (IFNs) in response to a viral infection.[1] They circulate in the blood and are found in peripheral lymphoid organs. They develop from bone marrow hematopoietic stem cells and constitute < 0.4% of peripheral blood mononuclear cells (PBMC).[1][2] Other than conducting antiviral mechanisms, pDCs are considered to be key in linking the innate and adaptive immune systems. However, pDCs are also responsible for participating in and exacerbating certain autoimmune diseases like lupus.[3

 

[V.R.T.]

pDCs are plasmacytoid dendritic cells

The lupus connection again...

However, pDCs are also responsible for participating in and exacerbating certain autoimmune diseases like lupus.[3

 

[Hutan]

Here's a 2022 paper about plasmacytoid dendritic cells:
Plasmacytoid Dendritic Cells, a Novel Target in Myeloid Neoplasms

They seem relevant to our discussions - professional producers of interferon, part of the immune response. They have some features of myeloid and some features of lymphoid cells. In fact it seems as though some are produced in bone marrow and some in lymph tissue.

Endosomal nucleic acid-sensing Toll-like receptors (TLRs) 7 and 9 are two innate sensors highly expressed in pDCs that recognize viruses and nucleic acids [46]. TLR 7 and TLR9 recognize single-stranded RNA and unmethylated CpG motif-containing DNA, respectively [6]. The recognition of nucleic acid induces a swift and massive secretion of IFN-I. In addition to IFN-β and IFN-α secretions (IFN-I), pDCs are able to produce IFN-III (IFN-λ) and other pro-inflammatory cytokines and cytokines including TNF-α, interleukine-6 (IL-6), IL-12, CXC-chemokine ligand 8 (CXCL8), CXCL10, CC-chemokine ligand 3 (CCL3) and CLL4 (Figure 1) [1,47].

 

[jnmaciuch]

pDCs are a good candidate for thinking about interferon production in the short term, though in the case of lupus they seem to lose their effector function after prolonged stimulation and don’t seem to drive the disease in later stages (https://www.nature.com/articles/s41467-020-19918-z). So if they are involved in ME/CFS at all, their more likely role would be something in early disease rather than maintaining disease—at least using their behavior in other diseases as a reference.