ARTICLE IN BRIEF
A magnetic resonance image (left) showing cerebral cavernous malformations (CCMs) in dark spots in a patient; a micro computed tomography image shows CCMs in the mouse model.
Researchers reported that bacteria in the gut microbiome may drive the formation of cerebral cavernous malformations (CCMs), and suggest that altering the microbiome may be a new route to preventive therapy for CCMs.
Two chance observations have unlocked a major mystery in the etiology of cerebral cavernous malformations (CCMs) — that exposure to lipopolysaccharide (LPS) from gram-negative bacteria in the gut can trigger disease in mice carrying the same gene mutations that cause the disease in humans.
The finding may explain the wide clinical variability among patients with the disease-causing mutations, and suggests that altering the microbiome may be a new route to preventive therapy, wrote the authors of the new study in the May 18 edition of the journal Nature.
CCMs are knots of abnormal blood vessels that develop in the brain due to inherited or acquired mutations in any one of three genes: krev interaction trapped protein 1 (KRIT1), CCM2, and programmed cell death protein 10 (PDCD10). Those with the sporadic form of the disease typically have only a single malformation, while those with familial CCM, comprising about 20 percent of patients, often develop several to several hundred malformations. CCMs are a cause of hemorrhagic stroke, seizure, and focal neurologic deficits.
“One of the mysterious aspects of the familial form of the disease is that even in patients with the same germline mutations, the clinical course is highly variable, much more than for most genetic diseases,” said Mark L. Kahn, MD, professor of medicine at the Perelman School of Medicine at the University of Pennsylvania in Philadelphia, and lead author on the study.
“One patient may have no lesions at age 70, while another patient with the exact same mutation may have hundreds of lesions at age 10. This suggests that there is non-genetic input that is very important.”
As a leader in the field of understanding the development of CCMs, Dr. Kahn was nothing if not prepared to take advantage of a chance observation. Through his work in mice that carry gene mutations, combined with other important developments in the field, he knew that the three CCM genes each encode one part of a three-protein adapter complex within endothelial cells in the brain's vasculature.
DR. MARK L. KAHN: “What we think is happening in people with the familial form of the disease, who are born with a CCM complex mutation in one copy of the gene, is that cells do OK as long as the other copy is still working. Probably over a lifetime, in a number of endothelial cells, you get a second hit, which inactivates the last working copy of the gene. And then that cell will now be CCM-deficient. That's when you begin to develop the disease.”
He also knew that the complex inhibits a downstream signaling pathway called mitogen-activated protein kinase kinase kinase 3 (MEKK3) that regulates vascular development, and that loss of the complex through mutation leads to over-activity of the MEKK3 pathway and development of vascular malformations (the mechanism of which is still unclear). But the upstream and extracellular controllers of the MEKK3 pathway were unknown.
AN UNEXPECTED DISCOVERY
In 2016, as Dr. Kahn was putting the finishing touches on experiments elucidating the role of those downstream pathways, he was also moving his colony of CCM-forming mice into a new building. The experiments involved intraperitoneal injection of tamoxifen to induce mutant gene expression, a standard tool for studying mutations too lethal to express continuously; injected mice would develop CCMs within days — but not the mice in the new building.
“The offspring were getting lots of lesions in the old building, but we noticed that gradually, they stopped,” Dr. Kahn said. “It wasn't subtle — it went away completely. We thought we might not be able to do experiments anymore.”
But as annoying and potentially catastrophic as it was at first, this turn of events became the first lucky break. The second came some weeks later, when the team noticed that the few mice that still developed lesions also had peritoneal abscesses, which likely developed because the tamoxifen needle was passed accidentally through the bowel.
“In a litter of eight or 10 pups, one would have CCM lesions, and that's the one that had developed an abscess,” Dr. Kahn said. “When that happened several times, it was a big clue. We realized it wasn't by chance.”
He and his colleagues found that a gram-negative bacterium, Bacteroides fragilis, and the membrane-derived LPS that it produced, was responsible for the effect: Injection of LPS into lesion-resistant mutant mice from the new facility caused robust CCM lesion formation.
“We don't yet know what was different about the buildings,” Dr. Kahn said, but he speculated that it may have simply been that the new facility was cleaner and harbored fewer sources of B. fragilis.
LPS interacts with an innate immune system receptor called toll-like receptor 4 (TLR4) to trigger inflammation and other cellular defense responses. TLR4 is found on the surface of many mammalian cells, including the vascular endothelium in the brain; variants in TLR4 had previously turned up in a genome-wide association study (GWAS) of familial CCM patients as a susceptibility factor for greater disease burden, although the significance of the finding was not clear at the time.
Dr. Kahn showed that by knocking down the level of TLR4 in the mutant mice, he could greatly reduce the development of lesions, indicating that LPS was acting through TLR4.
The model he proposes from this work is that gut bacteria release LPS, which enters the circulation and activates TLR4 on the surface of blood vessels in the brain. TLR4 signaling then drives MEKK3 signaling. LPS in small quantities does not normally lead to vascular malformations, because a functioning CCM complex acts to inhibit MEKK3. But when one of the genes of the complex is mutated, it ceases to act as a restraint on MEKK3, and LPS exposure can trigger formation of CCM lesions.
“What we think is happening in people with the familial form of the disease, who are born with a CCM complex mutation in one copy of the gene, is that cells do OK as long as the other copy is still working. Probably over a lifetime, in a number of endothelial cells, you get a second hit, which inactivates the last working copy of the gene. And then that cell will now be CCM-deficient. That's when you begin to develop the disease,” Dr. Kahn said.
As for the large differences in clinical manifestation in people with the same mutation, he said, “My guess is that this is related to differences in their biomes, and that's just the luck of the draw. If we can characterize these differences, and identify the biomes that are relatively protective, we might be able to take someone with the bad luck of an aggressive bacterial population, give them antibiotics, and then give them a fecal transplant, to give them a different biome. It might be a lifetime therapy.”
“The beauty of this study is that it puts together so many things about CCMs that might otherwise have stayed incomprehensible,” from the known mutations to the GWAS finding of TLR4 to the clinical variability in patients, said Douglas Marchuk, PhD, professor of molecular genetics and microbiology at Duke University in Durham, NC, who was not involved in the study. In addition, he said, the paper drives home the point that the inflammatory aspect hinted at by the GWAS results is likely central to the disease. “These results suggest it's not secondary — it may be fundamental for formation of the lesions,” Dr. Marchuk said.
Amy Akers, PhD, chief scientific officer of the Angioma Alliance, which supports research on CCMs, added that it was an exciting finding for the families of those with inherited CCM lesions, particularly the discovery of a potential role for the microbiome. The possibility of a new treatment approach may reduce the likelihood of developing lesions. “This gives us a lot of hope,” she said.
Will neurologists now need to become microbiologists as well? Dr. Kahn doesn't think so, at least not yet. While there are microbiome associations with other neurologic diseases, they may be mainly through immune system responses, rather than direct microbial action, he said. “In our case, it's a remarkably concrete and direct connection between the biome and the disease, since the biome is the source of the molecule that activates a receptor on the brain endothelium. I don't think we are going to find a dozen other diseases of the brain that are so directly linked.”
EXPERTS: ON THE ROLE OF GUT BACTERIA IN CEREBRAL CAVERNOUS MALFORMATIONS
DR. DOUGLAS MARCHUK said the paper drives home the point that the inflammatory aspect hinted at by the genome-wide association study results are likely central to the disease. “These results suggest it's not secondary — it may be fundamental for formation of the lesions.”
DR. AMY AKERS said that it was an exciting finding for the families of those with inherited CCM lesions, particularly the discovery of a potential role for the microbiome. The possibility of a new treatment approach may reduce the likelihood of developing lesions. “This gives us a lot of hope.”