Scientists have discovered secret corridors that connect the bone marrow in the skull directly to the brain. According to a paper published August 27 in Nature Neuroscience, these tiny channels serve as an express route for marrow cells, including neutrophils, to cross into the cortex in response to injury. The researchers, led by Matthias Nahrendorf of Massachusetts General Hospital in Boston, challenge the prevailing view in the field that bone marrow-derived cells recruited to the brain hail primarily from the blood, which carries cells collectively released from all sources of bone marrow in the body.

  • Neutrophils infiltrating the brain hail from the skull bone marrow more so than other bones in the body.
  • They travel via small vascular channels connecting the skull marrow directly to the dura mater.
  • People and mice have the channels.

“This is a novel access route to the brain for inflammatory cells and its discovery, while providing new insight into the pathobiology of post-ischemic inflammation, contributes to debunk the myth of the immune privilege of the brain,” wrote Costantino Iadecola of Weill Cornell Medical College in New York. “Furthermore, inasmuch as innate and adaptive immune cells participate in neurodegeneration, the findings also have implications for Alzheimer’s disease and related conditions,” he added.

As the sole supplier of short-lived myeloid cells in the body, the bone marrow pumps out multitudes of leukocytes, including neutrophils, in response to injury or insult. Researchers have spotted such myeloid cells infiltrating the brain in response to stroke, traumatic brain injury, or in the context of neurodegenerative disease, but exactly which marrow sources supply those cells is unclear (Offner et al., 2006; Courties et al., 2014; Aug 2015 news). 

Bridges to the Brain.

Calcium-coated (blue) vascular microchannels (red) connect the marrow cavity to the dura mater. Myeloid cells in the marrow and brain are labeled green. [Courtesy of Herisson et al., Nature Neuroscience, 2018.]

First author Fanny Herisson and colleagues investigated whether neutrophils recruited to the brain in response to injury came from marrow in the skull, or equally from other marrow sources. To test this, they injected different cell-permeant fluorescent dyes into the tibia and the skull. They then monitored the neutrophils that took up the dyes and flooded the brain in response to stroke or aseptic meningitis. About twice as many neutrophils came from the skull as from the tibia. In contrast, triggering a brief cardiac arrest summoned neutrophils to the heart that derived equally from both bones. The data suggested that brain injury specifically recruits bone marrow cells from the skull.

The researchers hypothesized that a direct vascular connection could exist. Indeed, confocal microscopy of mouse brain sections revealed a dense network of 21-micron-diameter vessels crisscrossing the inner surface of the skull and connecting to the brain’s dura mater. The outsides of the vessels were speckled with microcalcifications from nearby osteoblasts, and their insides were lined with endothelial cells. Neutrophils and monocytes were seen in the lumen of these channels. Using time-lapse microscopy of freshly harvested sections of mouse skulls immersed in a bath of media, the researchers detected blood flowing through the channels from the brain toward the skull. Strikingly, neutrophils in the channels crawled against the current, making their way from the skull toward the brain. This cross-current traffic increased after a stroke.

Against the Current. A neutrophil (green) travels from the skull toward the brain through a calcium-coated (white) microchannel lined with endothelial cells (red). [Courtesy of Herisson et al., Nature Neuroscience, 2018.]

To search for the channels in people, the researchers imaged skull sections taken from three patients to relieve pressure caused by swelling from traumatic brain injury or other trauma. They found a similar network of channels feeding into the associated dura, though at about fivefold larger in diameter than those in mice. They have yet to determine whether they function similarly.—Jessica Shugart


  1. The meninges have emerged as a major reservoir of innate and adaptive immune cells in brain diseases associated with inflammation (Benakis et al., 2016; Perez-de-Puig et al., 2015). The present study indicates that, in acute stroke, the predominant source of meningeal neutrophils is not hematogenous, but is the skull bone marrow. This is a novel access route to the brain for inflammatory cells and its discovery, while providing new insight into the pathobiology of post-ischemic inflammation, contributes to debunk the myth of the immune privilege of the brain. The observation that similar channels are also present in the human skull, in conjunction with the well-established imaging finding of meningeal enhancement in human stroke (indicative of meningeal inflammation) suggests that this mechanism is also relevant to human stroke. Furthermore, inasmuch as innate and adaptive immune cells participate to neurodegeneration, the findings also have implications for Alzheimer’s disease and related conditions. Since the skull is permeable to small molecules (Roth et al., 2014) it is conceivable that agents targeting inflammatory cell trafficking or action could be effectively delivered to the skull for therapeutic purposes. A final consideration pertains to the use of head-shielding to protect the brain in bone marrow transplantation studies in rodents, a widely used method to investigate the contribution of hematogenous cells to disease mechanisms, including Alzheimer’s disease. Shielding the head from radiations will not remove the skull bone marrow, which, as shown in this paper, is a preferential source of inflammatory cells destined to the brain. The discovery of a transcranial entry route to the brain's coverings opens new avenues for the immunology of the brain and will undoubtedly stimulate much new work in this exciting research field. 


    . Commensal microbiota affects ischemic stroke outcome by regulating intestinal γδ T cells. Nat Med. 2016 May;22(5):516-23. Epub 2016 Mar 28 PubMed.

    . Neutrophil recruitment to the brain in mouse and human ischemic stroke. Acta Neuropathol. 2015 Feb;129(2):239-57. Epub 2014 Dec 30 PubMed.

    . Transcranial amelioration of inflammation and cell death after brain injury. Nature. 2014 Jan 9;505(7482):223-8. Epub 2013 Dec 8 PubMed.

  2. Herisson et al. found that neutrophils take an unexpected “shortcut,” migrating toward the inflamed brain during stroke through microscopic channels that connect the skull marrow with the dura. The study results are exciting and point to a key role for direct vascular channels delivering neutrophils from skull marrow to the CNS surface during brain inflammation.

    Neutrophils are highly reactive immune system cells with a central role in host defense. They are equipped with a large array of cytotoxic mechanisms including the production of reactive oxygen species (ROS) and the release of enzymes, cytokines, and neutrophil extracellular traps (NETs). During low-grade chronic sterile inflammation, neutrophils contribute to long-term collateral tissue injury even if large numbers of neutrophils do not accumulate within tissues.

    A few studies, including our own, have shown neutrophil migration in the brains of patients with Alzheimer’s disease and in transgenic animals with AD-like disease, suggesting a role for these myeloid cells in AD (Subramanian et al., 2010; Baik et al., 2014; Brock et al., 2015; Zenaro et al., 2015). The presence of neutrophils in AD brains or other sites of chronic low-grade inflammation was previously underestimated for several reasons, potentially including their short life spans, high turnover, and phenotypic plasticity (Doring et al., 2015; Zenaro and Constantin, 2017). Furthermore, neutrophils invading the brain may rapidly disappear because they are engulfed by microglial cells, as previously reported in animal models of ischemic stroke (Neumann et al., 2008). Our recent studies demonstrated that neutrophils migrate into the brain in low numbers and play a role in the induction of neuropathological changes and memory deficit in 5xFAD and 3xTg-AD mice (Zenaro et al., 2015). Neutrophil depletion or blocking adhesion molecules controlling their migration reduced the severity of AD symptoms in animal models, suggesting that neutrophil-directed therapies may have a beneficial effect in AD patients.

    This study by Herisson et al. raises the interesting possibility that skull neutrophil supply could have a role in chronic neurodegenerative disorders like AD. However, future studies are needed to clarify the potential contribution of leptomeningeal vessels and neutrophils originating from the skull marrow in the pathogenesis of AD.


    . Microglia cells protect neurons by direct engulfment of invading neutrophil granulocytes: a new mechanism of CNS immune privilege. J Neurosci. 2008 Jun 4;28(23):5965-75. PubMed.

    . CCR6: a biomarker for Alzheimer's-like disease in a triple transgenic mouse model. J Alzheimers Dis. 2010;22(2):619-29. PubMed.

    . The antimicrobial protein, CAP37, is upregulated in pyramidal neurons during Alzheimer's disease. Histochem Cell Biol. 2015 Oct;144(4):293-308. Epub 2015 Jul 14 PubMed.

    . Migration of neutrophils targeting amyloid plaques in Alzheimer's disease mouse model. Neurobiol Aging. 2014 Jun;35(6):1286-92. Epub 2014 Jan 8 PubMed.

    . Neutrophils in atherosclerosis: from mice to man. Arterioscler Thromb Vasc Biol. 2015 Feb;35(2):288-95. Epub 2014 Aug 21 PubMed.

    . Neutrophils promote Alzheimer's disease-like pathology and cognitive decline via LFA-1 integrin. Nat Med. 2015 Aug;21(8):880-6. Epub 2015 Jul 27 PubMed.

    . Targeting neuaikroinflammation in the treatment and prevention of Alzheimer’s disease. Drugs of the Future 2017, 42(1): 21-42.

  3. This study is enchantingly novel and interesting. Before getting into the concepts, some acknowledgment is appropriate for the ingenuity of the experimental approach as well as for the creativity of the underlying research question. The authors started with a startlingly out-of-left-field hypothesis: It’s known that acute brain inflammation sends signals to bone marrow to mobilize leukocytes, which enter circulation, and that leukocytes are subsequently found in the inflammatory site in the brain. Question: When leukocytes from the bone marrow infiltrate a region of brain inflammation, do they come preferentially from the nearby bone marrow in the skull? Or just generally from all bone marrow sites? Behind this question is another, more subtle query: Do leukocytes in these regions of inflammation enter from the circulation (i.e., by leaving the blood vessels and entering the tissue), as is typically seen for any form of inflammation? Or do they find another pathway?

    The approach was termed “spectrally resolved site-specific cell tagging” but in reality consisted of carefully injecting one type of cell-labeling dye into the skull bone marrow and another into the large tibia bone in the leg. Next the team carried out a formidable number of incisive control experiments to show that the labeling was efficient, specific, and due entirely to uptake by the cells, not dye leakage, as well as showing that tibia and skull equally released cells into the bloodstream. Then, the investigators generated acute brain inflammation, either focal (stroke) or more diffuse (chemical meningitis), and the prior labeling enabled the authors to identify infiltrating cells as coming from preferentially from skull or tibia, or equally from both. The surprising result was that in both conditions the skull marrow contributed about twice as many acute-inflammatory neutrophils as did the tibia. For a further control, they showed that a remote heart-attack inflammatory site incorporated equally the cells from both tibia and skull.

    This unexpected result raised many questions, some of which could be answered, while others will spur additional experimentation:

    1. How do the neutrophil cells get into the brain from the skull? Careful imaging showed that there are vascular channels that connect the skull bone marrow to the dura—a tough membrane that lies inside the skull and encases the brain. From the dura, there are additional pathways to the leptomeninges, which are isolated from the dura by the fluid-impermeable arachnoid membrane, and which constitute the subarachnoid space where spinal fluid circulates. These endothelial-lined channels support bulk flow from the brain to the marrow cavities in the skull, so that the inflammatory neutrophils must “swim upstream” (they’re known to be able to do so) to enter sites of brain inflammation.
    2. What signals come from the brain to mobilize the neutrophils from the skull marrow? Answering this question by analogy from other types of marrow-mobilization studies, the authors showed that a key ‘marrow retention factor’ CXCL12 is reduced in skull marrow as the neutrophils are leaving. It’s well established (Kim et al., 2006) that reduced CXCL12 mediates the mobilization of myeloid cells after inflammation throughout the body. The existence of such a signal from inflamed brain to skull marrow is implied by the extremely rapid appearance of leukocytes in these acute lesions, after only six hours. The specific signal(s) remain to be determined.

    Is this research finding relevant for neuroinflammation in the context of chronic neurodegeneration? This crucial point remains unaddressed in the present study. Fortunately, the way forward should be relatively straightforward. First, it should be noted that these findings aren’t related to the recent reports regarding dural lymphatics, which are proposed to be involved in handling soluble material and may connect to lymph nodes rather than to the brain itself. Second, the authors focused mainly on neutrophils whose role in neurodegeneration remains rather speculative. However, the same mechanisms as shown here may also pertain to monocytes, which can give rise to brain macrophages. The functions of monocytes during chronic neurodegeneration haven’t been unambiguously established and there’s ongoing controversy regarding their relative importance for pathogenesis. However, it will now be possible to ask the critical questions to move forward with this area of research:

    1. Are the skull marrow cells being released into direct channels that access the dura and the brain surface during chronic neuroinflammation? Are these cells monocytes, neutrophils, or both?
    2. What signals come from the brain to mediate this release? Will intervention in that process produce a beneficial effect? A deleterious effect? Or will it be neutral?
    3. Monocyte cells coming directly as progenitors from marrow to brain won't undergo the typical maturation process. Will their phenotype be altered from that of other infiltrating macrophages and will investigators still be able to distinguish them from resident microglia?

    Looking to the immediate future, having a new population of cells and novel set of signals to consider will be highly motivating in the search for therapeutic targets to prevent or ameliorate neurodegeneration.


    . G-CSF down-regulation of CXCR4 expression identified as a mechanism for mobilization of myeloid cells. Blood. 2006 Aug 1;108(3):812-20. Epub 2006 Mar 14 PubMed.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Could Neutrophils Be the Newest Players in Neurodegenerative Disease?

Paper Citations

  1. . Experimental stroke induces massive, rapid activation of the peripheral immune system. J Cereb Blood Flow Metab. 2006 May;26(5):654-65. PubMed.
  2. . The innate immune system after ischemic injury: lessons to be learned from the heart and brain. JAMA Neurol. 2014 Feb;71(2):233-6. PubMed.

Further Reading

No Available Further Reading

Primary Papers

  1. . Direct vascular channels connect skull bone marrow and the brain surface enabling myeloid cell migration. Nat Neurosci. 2018 Aug 27; PubMed.