Brain inflammatory cascade controlled by gut-derived molecules

Metabolite molecules produced by the gut’s microbes activate immune cells in the brain called microglia, which signal to astrocyte cells to mediate responses to inflammation in the central nervous system.
Hartmut Wekerle is at the Max Planck Institute of Neurobiology, D-82152 Martinsried, Germany.

Search for this author in:

Some immunologists regard the central nervous system (CNS) as a no-man’s-land, avoided by immune cells and therefore uninteresting. But, in fact, the CNS has a vigorous immune potential that remains dormant in normal conditions but is awakened after injury. The switch that controls the brain’s immune microenvironment involves non-neuronal cells called glia — not only microglia, which are sometimes called the immune cells of the CNS, but also multifunctional cells called astrocytes1. In a paper in Nature, Rothhammer et al.2 describe how these two glial cell types communicate on a molecular level to influence inflammation in the CNS, and show that this interaction is controlled remotely by microbes that inhabit the gut.

A decade ago, the group that performed the current study, along with another research group, discovered3,4 an unexpected immunoregulatory role for a ligand-activated transcription factor called the aryl hydrocarbon receptor (AHR), which at the time was best known as a receptor for environmental toxins5. The two groups showed that AHR modulates the progression of experimental autoimmune encephalomyelitis (EAE) — an autoimmune disease in mice in which the immune system becomes overactive and attacks the CNS. EAE is often used a model of multiple sclerosis (MS). Initially, the groups focused on how AHR might affect EAE by regulating pathogenic and protective subsets of immune cells outside the CNS. But it later emerged that AHR is also strongly expressed in the CNS, particularly in microglia and astrocytes6, raising the question of whether AHR in the CNS has a role in autoimmune diseases.

In the current study, Rothhammer et al. induced EAE in mice that had been genetically engineered so that AHR could be deleted in microglia (but not in other brain cells or immune cells) by a drug treatment. Elimination of microglial AHR substantially exacerbated EAE in the AHR-depleted mice, but left immune responses outside the CNS unaltered. This finding suggests that AHR activation in microglia inhibits inflammation in the CNS.

Microglia rarely act alone. Instead, they often team up with other cell types to respond to the stimuli that activate them. For example, after being activated, microglia can instruct certain astrocytes to attack local neurons7. Rothhammer and colleagues found that AHR-deficient microglia activated by EAE triggered exaggerated inflammatory responses in local astrocytes. Next, the authors used bioinformatics to analyse the gene-expression pathways altered in these glia. This analysis suggested that unexpected proteins signal from microglia to astrocytes.

The usual suspects in such cases are pro-inflammatory signalling molecules, but Rothhammer et al. showed that AHR in microglia directly regulates the expression of genes that encode the proteins TGF-α and VEGF-B (Fig. 1) — neither of which has previously received much attention from neuroimmunologists. Subsequent detailed in vitro and in vivo analyses confirmed that TGF-α and VEGF-B regulate the pro-inflammatory reactivity of astrocytes. TGF-α dampens astrocyte inflammatory responses to EAE, and its expression in microglia is inhibited by AHR deletion. Conversely, VEGF-B enhances responses to EAE, and its expression is promoted by AHR deletion.

Figure 1 | Long-distance regulation of immune cells in the brain. Gut bacteria process the dietary component tryptophan to produce metabolite molecules that enter the central nervous system (CNS). In the brain, these metabolites act as ligands for the aryl hydrocarbon receptor (AHR) — a transcription factor expressed in cells called microglia and astrocytes that mediate responses to inflammation in the CNS. Rothhammer et al.2 report that, when activated in microglia, AHR binds to the genes that encode the proteins VEGF-B and TGF-α, inhibiting the expression of the former and promoting that of the latter. Any VEGF-B released from microglia enhances the responsiveness of astrocytes to CNS inflammation. By contrast, TGF-α dampens astrocyte responsiveness.

So, the activity of microglia and astrocytes is modulated by AHR during brain inflammation in autoimmune disease. But which signals might modulate microglial AHR? In addition to environmental toxins, AHR is bound by a broad range of molecules, including dietary derivatives8. In particular, food plants such as broccoli and other members of the cabbage family contain components that bind AHR either directly or after being processed into metabolite molecules, such as derivatives of tryptophan (Trp), by gut microbes9. Rothhammer et al. fed their mice diets either depleted or enriched in Trp. Trp depletion exacerbated EAE in wild-type mice, whereas enrichment ameliorated the effects of the disease. By contrast, neither diet had any effect on the progress of EAE in AHR-deficient animals, as might have been predicted — in these animals, Trp cannot bind to AHR to dampen immune responses.

To determine whether their work is likely to have implications for humans, the authors verified basic elements of their analyses in tissue samples from people with MS, in which an autoimmune attack drives glial inflammation, destruction of nerve processes and their insulating myelin sheaths, and ultimately scar formation10. The group found that AHR, TGF-α and VEGF-B were expressed in microglia-like cells in MS tissues. Levels of the proteins were higher in newly inflamed regions than in old scar tissue or unaffected surrounding tissue. This suggests (but does not prove) that TGF-α and VEGF-B have a role in the formation of MS scar tissue.

Rothhammer and colleagues’ work sheds light on the complex regulation of inflammatory reactivity in the CNS and adds another facet to our understanding of the gut–brain connection. Robust regulation of inflammatory responsiveness is essential for proper CNS function. Deficient regulation, with unrestrained inflammatory episodes, leads to sickness, irreversible cell loss and scar formation11, whereas compromised inflammatory reactivity can result in tumour formation and opportunistic infection12. The authors’ findings are therefore likely to have implications beyond MS.

The interactions between the gut, microglia and astrocytes outlined by Rothhammer et al. are not the only mechanisms that safeguard inflammatory responses in the brain13. It will be of interest to examine how other regulators of the CNS immune microenvironment modulate the newly identified signalling pathway. These factors include the cells associated with the cerebral blood vessels, as well as active neurons. Indeed, pharmacological silencing of neurons leads to the activation of neighbouring microglia14.

That the behaviour of microglia can be controlled remotely by intestinal products is intriguing, although not without precedent. A flurry of observations previously linked the CNS to the gut and its microbial contents. Neuronal pathways, hormones, microbial molecules and metabolites are all involved in signalling between these regions15. Specifically, short-chain fatty acids produced by gut bacteria can modulate microglia cells16, and tryptophan metabolites act directly on astrocytes6. Nonetheless, the current findings broaden our understanding of the gut–brain connection. The authors speculate that this pathway might support the repair of injured neural cells.

As Rothhammer and colleagues point out, their experimental observations might lead to new therapeutic approaches to quelling unwanted CNS inflammation, and possibly to supporting neuronal repair. First, enhancement of TGF-α and blockade of VEGF-B might reduce CNS inflammation to an acceptable, non-toxic level. Second, clinical CNS inflammation could be dampened indirectly by means of the gut. Dietary protocols that promote anti-inflammatory regulation could be a promising non-invasive approach to treating brain inflammation. It is to be hoped that diets that have been proposed as effective medications for diseases such as MS, but whose effectiveness has yet to be formally proved17, will now be re-examined.

doi: 10.1038/d41586-018-05113-0
Nature Briefing

Sign up for the daily Nature Briefing email newsletter

Stay up to date with what matters in science and why, handpicked from Nature and other publications worldwide.

Sign Up


  1. 1.

    Salter, M. W. & Stevens, B. Nature Med. 23, 1018–1027 (2017).

  2. 2.

    Rothhammer, V. et al. Nature (2018).

  3. 3.

    Quintana, F. J. et al. Nature 453, 65–71 (2008).

  4. 4.

    Veldhoen, M. et al. Nature 453, 106–109 (2008).

  5. 5.

    Okey, A. B., Riddick, D. S. & Harper, P. A. Trends Pharmacol. Sci. 15, 226–232 (1994).

  6. 6.

    Rothhammer, V. et al. Nature Med. 22, 586–597 (2016).

  7. 7.

    Liddelow, S. A. et al. Nature 541, 481–487 (2017).

  8. 8.

    Stockinger, B., Di Meglio, P., Gialitakis, M. & Duarte, J. H. Annu. Rev. Immunol. 32, 403–432 (2014).

  9. 9.

    Veldhoen, M. & Brucklacher-Waldert, V. Nature Rev. Immunol. 12, 696–708 (2012).

  10. 10.

    Lassmann, H. & Wekerle, H. in McAlpine’s Multiple Sclerosis 4th edn (eds Compston, A. et al.) 557–600 (Churchill Livingstone, 2006).

  11. 11.

    Sofroniew, M. V. Neuroscientist 20, 160–172 (2014).

  12. 12.

    Monaco, M. C. G. & Major, E. O. Front. Immunol. 6, 159 (2015).

  13. 13.

    Li, Q. & Barres, B. A. Nature Rev. Immunol. 18, 225–242 (2018).

  14. 14.

    Neumann, H. Glia 36, 191–199 (2001).

  15. 15.

    Sampson, T. R. & Mazmanian, S. K. Cell Host Microbe 17, 565–576 (2015).

  16. 16.

    Erny, D. et al. Nature Neurosci. 18, 965–977 (2015).

  17. 17.

    Riemann-Lorenz, K. et al. PLoS ONE 11, e0165246 (2016).

Download references