Depending on whom you ask, pericytes are either vital vascular regulators that boost blood flow to needy areas of the brain, or mere bystanders that let other cells take on that job. A new study published January 30 in Nature Neuroscience bolsters the evidence for the former. Researchers led by Berislav Zlokovic of the Keck School of Medicine of the University of Southern California, Los Angeles, found that mice with fewer pericytes than normal had diminished blood flow and lower oxygen levels in the brain. As they grew older, the animals also lost neurons and had trouble carrying out normal mouse activities. Pericytes shrink and die in patients with Alzheimer’s disease, and the study supports the idea that the loss of these cells promotes the disease’s pathology. “It makes a direct link between blood flow dysregulation and neuronal degeneration and loss,” Zlokovic told Alzforum.

Blood Vessel Basics. Pericytes make their homes on capillaries, whereas smooth muscle cells reside on thicker vessels. [Courtesy of Costantino Iadecola and Neuron.]

The paper presents a compelling case, in part because it includes a variety of techniques, Gareth Howell of the Jackson Laboratory in Bar Harbor, Maine, told Alzforum. “It’s almost [always] impossible to be definitive, but the number of experiments they did points to a key role of pericytes” in regulating brain blood flow, said Howell, who was not connected to the study. But Jaime Grutzendler of the Yale School of Medicine in New Haven, Connecticut, disagreed. “The study is not designed to determine the precise contributions of the different cell types to blood flow control because the genetic manipulation used is not exclusive to either smooth muscle cells or pericytes,” he wrote to Alzforum. 

When a neuron in the brain feels hungry for oxygen or nutrients, it orders room service, spurring nearby vessels to deliver more blood. Ensuring that supply meets demand is known as neurovascular coupling. A klatch of interacting cells is involved, including neurons, astrocytes, smooth muscle cells, endothelial cells, and pericytes. Researchers still debate which cells adjust blood flow. Smooth muscle cells, which surround arteries and arterioles, are one candidate (see image). Pericytes that sit on capillaries, which lack smooth muscle cells, are another. Some findings suggest pericytes may open the circulatory taps in the brain by allowing capillaries to relax (Yemisci et al., 2009; Hall et al., 2014). But a 2015 study by Grutzendler and colleagues implicated smooth muscle cells that encircle large arterioles (see Jun 2015 news). 

In the Red.

Partial oxygen pressure is higher in the cortex of a normal mouse (top) than in one that has fewer pericytes (bottom). Blood vessels are gray. [Courtesy of Berislav Zlokovic and Nature Neuroscience.]

Nailing down the role of pericytes in metering blood flow has taken on added importance with the discovery that these cells deteriorate and die in people with AD and other neurodegenerative illnesses. This attrition could contribute to altered blood flow patterns detected by blood oxygen level-dependent (BOLD) fMRI in patients who have early AD.

With this in mind, co-first authors Kassandra Kisler, Amy Nelson, Sanket Rege, and colleagues tested mice that carry only one copy of the gene that encodes platelet-derived growth factor receptor-β. Pericytes need PDGFRβ for survival. The mice have fewer of the cells, which cover about 25 percent less capillary surface area than they would in normal mice. Overall blood delivery to the cerebrum was down 30 percent in these pericyte-starved mice.

To test how blood flow responds to a stimulus, the researchers gave small electric shocks to one of the mice’s hind legs and then monitored changes in cerebral blood flow using in vivo two-photon laser scanning microscopy. They homed in on the portion of the somatosensory cortex that corresponds to the stimulated leg. Capillaries in the genetically modified mice dilated more slowly, reaching 50 percent peak diameter 6.5 seconds later than did vessels in normal mice. The surge in capillary blood flow that follows the stimulus was also tardy in the mice.

To confirm that individual capillaries behaved differently if they didn’t sport a pericyte, the researchers crossed pericyte-deficient mice with animals that express the pericyte marker NG2-dsRed. They then tested the offspring. After the same stimulus, only capillaries with pericytes on them relaxed, Kisler and colleagues found.

Could changes in other cells account for the capillaries’ responses? Not likely, the scientists concluded. They measured the behavior of arterioles, which are swaddled by smooth muscle cells, and detected no differences between the pericyte-lacking mice and normal animals. Endothelial cells, astrocytes, and microglia were also comparable in the two types of mice.

Kisler and colleagues then asked whether the diminished blood flow led to less oxygen availability in the brain. They found that by several measures, including fluorescence of NADH, neurons in the pericyte-deficient mice were indeed receiving less oxygen.

This hypoxia didn’t seem to faze mice under two months old. Their neurons remained alive and responded normally to stimulation. The young mice burrowed, made nests, and explored new objects placed in their cages just as normal mice did.

By the time the mice were six to eight months old, however, they had begun to pay the price for their pericyte shortage. Leg stimulation evoked weak responses with long latent periods from somatosensory neurons, which had begun dying. The animals scored poorly on all three behavior tests. Cerebrovascular flow had diminished further, with the cerebrum receiving 58 percent less blood than usual.

Even so, Grutzendler remained unconvinced that pericytes on capillaries are directly responsible for local changes in vessel diameter. The criterion for distinguishing arterioles from capillaries—vessel size—was not specific enough, Grutzendler claimed. “Using only size as a criterion doesn’t tell you if a blood vessel is covered by smooth muscle cells or pericytes,” he said. Other researchers have noted that pericytes and smooth muscle cells express some of the same markers, including PDGFRb and NRG2. In addition, the pericyte-deficient mice show other defects, such as leaks in the blood-brain barrier and reductions in microvascular density, that could account for their neural deterioration, said Grutzendler. “That makes it difficult to interpret the data in the context of neurovascular coupling,” he told Alzforum.

But Costantino Iadecola of Weill Cornell Medicine in New York said the paper makes a valuable contribution. “They show that in these mice there is a level of hypoxia that has never been connected to pericytes before,” he told Alzforum. Howell said that by linking blood flow changes to neurodegeneration, the study “adds to the growing weight of evidence that vascular dysfunction is likely to be a major contributor to increased susceptibility to Alzheimer’s disease.” He added that determining what happens to the mice between the ages of two and eight months, when the effects of the blood flow alterations start to show up, could help researchers identify molecular pathways that foster neurodegeneration—and possibly reveal ways to stop it. —Mitch Leslie

Mitch Leslie is a freelance writer based in Tucson, Arizona.


  1. In response to Dr. Grutzendler’s comments, we would like to emphasize that two recent exceptional papers (Mishra et al., 2016; Biesecker et al., 2016) from two different groups have independently confirmed the role of pericytes in neurovascular coupling, as we also showed. These two papers demonstrated that astrocytic calcium regulates neurovascular coupling to pericytes, but not to arteriolar smooth muscle cells. Moreover, a recent single-cell RNA-seq study demonstrated expression of several contractile proteins in pericytes derived from mouse cortex or hippocampus including skeletal muscle actin, vimentin, desmin, calponin, non-muscle myosin variants, and a low SMA (smooth muscle actin) expression (Zeisel et al., 2015). This study confirmed earlier findings showing expression of contractile proteins in pericytes using immunocytochemical staining and immunogold labeling at the ultrastructural level. Additionally, two recent optogenetic studies, both presented at the Society for Neuroscience 2016 annual meeting, one from Andy Shih’s group (Hartmann et al., 2016) and the other from our group (Nelson et al., 2016), have shown that pericytes contract after single- or two-photon stimulation. It remains unclear to us why Dr. Grutzendler was not able to see pericyte contractility in his optogenetic study (Hill et al., 2015). As we discussed in this paper, we think that the future studies should carefully address whether different Cre drivers can lead to differential channelrhodopsin 2 expression in different pericyte subpopulations in transgenic mice, and whether stimulation of pericytes in these studies is model-dependent or optogenetic light stimulus source- and duration-dependent, which could account for differences. Others have noted that Grutzendler’s Neuron 2015 study did not carefully distinguish between pericytes and smooth muscle cells based on their distinct morphology, and very likely might have overlooked pericyte constriction of capillaries by pericyte processes that extend along the capillary walls (Atwell et al., 2016). These processes and the shape of pericyte cell body do not share any morphological similarities with typical arteriolar ring-like smooth muscle cells (Hartmann et al., 2015). 

    We disagree with Dr. Grutzendler that the vessel size for capillaries was not specific in our study and that we could not distinguish them from arterioles. We showed in Supplementary Fig. 3a that 57 total capillaries from 10 control mice and 40 total capillaries from 10 pericyte-deficient mutant mice had averaged baseline diameters of 4.4 μm and 4.5 μm, respectively. This diameter is considered by hundreds of studies and hundreds of laboratories to be typical for mouse brain capillaries, but not the arterioles, which have been shown to have larger diameters. For example, the average basal diameter of small arterioles in our study was nearly three times larger than capillaries (Supplementary Fig. 3a).

    In addition to identifying capillary-sized vessels by diameter, in a subset of experiments pericytes were labeled with dsRed. Here, we observed that capillaries with typical mid-capillary shaped pericytes (bump-on-a-log appearance with processes running along the capillaries) dilated ahead of locations that lacked a pericyte cell body or pericyte processes (Fig. 1j, k) confirming regulation of capillary flow by pericytes.

    Naturally, we were also particularly concerned as to whether arteriolar and smooth muscle cell functional responses are affected in pericyte-deficient mice. To clarify, we would like to list the evidence that we think convincingly demonstrates that arteriolar responses and smooth muscle cells are not affected in our pericyte-deficient loss-of-function model. The time to 50 percent peak arteriolar dilation (Fig. 1e-f), the smooth muscle thickness on studied arterioles (Supplementary Fig. 3c, d),  the number of smooth muscle cells, and the stimulus-driven red blood cell velocity increase in arterioles (Fig. 3a) were all similar in one- to two-month-old controls and pericyte-deficient mutants. Arteriolar constric­tion in response to phenylephrine was unchanged in the mutant mice (Fig. 3c, d) as was smooth mus­cle relaxation induced by adenosine, an endothelium-independent vasodilator that acts as a direct vascular smooth muscle cell relaxant (Fig. 3c, e). In vivo cerebral blood flow in response to adenosine was the same in pericyte mutant mice and in littermate controls (Fig. 3f). Collectively, these data present, in our opinion, a compelling case that the smooth muscle cell function is unaffected in the pericyte-deficient mice, and therefore does not contribute to neurovascular uncoupling.

    We do agree with Dr. Grutzendler that it is difficult to determine the exact contributions of impaired hemodynamic responses and blood-brain barrier breakdown to the pathophysiological process of neurodegeneration, either in humans or mouse models, as stated in our discussion. This remains, however, an area of intense research by our group and others who study humans with genetic risk factors for sporadic and autosomal dominant Alzheimer’s disease, as well as in new transgenic rodent models of pericyte ablation. 


    . Astrocytes mediate neurovascular signaling to capillary pericytes but not to arterioles. Nat Neurosci. 2016 Dec;19(12):1619-1627. Epub 2016 Oct 24 PubMed.

    . Glial Cell Calcium Signaling Mediates Capillary Regulation of Blood Flow in the Retina. J Neurosci. 2016 Sep 7;36(36):9435-45. PubMed.

    . Brain structure. Cell types in the mouse cortex and hippocampus revealed by single-cell RNA-seq. Science. 2015 Mar 6;347(6226):1138-42. Epub 2015 Feb 19 PubMed.

    . Regional Blood Flow in the Normal and Ischemic Brain Is Controlled by Arteriolar Smooth Muscle Cell Contractility and Not by Capillary Pericytes. Neuron. 2015 Jun 23; PubMed.

    . What is a pericyte?. J Cereb Blood Flow Metab. 2016 Feb;36(2):451-5. Epub 2015 Oct 14 PubMed.

    . Pericyte structure and distribution in the cerebral cortex revealed by high-resolution imaging of transgenic mice. Neurophotonics. 2015 Oct;2(4):041402. Epub 2015 May 27 PubMed.

  2. I believe that the debate is not about which cells control blood flow—there is enough evidence that both vascular smooth muscle cells and pericytes can do this (movies of pericytes altering capillary diameter can be found here and here).

    The real debate should be about quantifying their relative contributions and identifying under what circumstances each plays a role. Kisler et al. show that the consequence of pericytes not working may not be pronounced right away, but can snowball into disease-like pathology over time. Recent evidence from our own work (Mishra et al., 2016) and from Eric Newman's lab (Biesecker et al., 2016) shows that capillaries, defined as blood vessels that are both narrow and lack smooth muscle cells, can dilate in response to neuronal activity in a manner independent of arterioles. Kisler et al.'s work confirms that such capillary-level regulation of cerebral vasculature is essential for healthy brain function and warrants more research into the functions of pericytes in health and disease, including capillary flow regulation and, as pointed out by Dr. Grutzendler, in maintaining the blood brain barrier.


    . Astrocytes mediate neurovascular signaling to capillary pericytes but not to arterioles. Nat Neurosci. 2016 Dec;19(12):1619-1627. Epub 2016 Oct 24 PubMed.

    . Glial Cell Calcium Signaling Mediates Capillary Regulation of Blood Flow in the Retina. J Neurosci. 2016 Sep 7;36(36):9435-45. PubMed.

  3. Kisler et al. have performed a tour de force study on pericyte-deficient mice, providing strong evidence that even a mild loss of pericytes (22 percent) leads to impaired neurovascular coupling and brain oxygenation. This is a significant step beyond a key 2010 Neuron paper from the Zlokovic lab showing that pericyte deficiency leads to progressive BBB degeneration (Bell et al., 2010). Both BBB and blood flow changes caused by pericyte loss may therefore be contributors to neurodegeneration in Alzheimer’s disease, and likely other cerebrovascular diseases such as stroke and vascular dementia.

    Inextricably linked with these new findings is the long-standing question of whether capillary diameter is regulated by contractility of pericytes in vivo. Past studies seeking to answer this question have arrived at opposing conclusions. As mentioned by Professor Grutzendler’s response to this article, and discussed by Kisler, et al., more studies are needed to address this issue head-on. Below, we briefly discuss why more studies are needed, and what type of information would help moving forward.

    Does PDGFRβ heterozygosity only affect pericytes?
    In their discussion, Kisler et al. describe the difficulties in proving that neurovascular uncoupling is a direct effect of pericyte roles in controlling capillary diameter. Their 2010 paper showed that endothelial cells are damaged in young PDGFRβ+/- mice and a leaky blood-brain barrier could affect other cells of the neurovascular unit. This makes it difficult to say for certain whether impaired blood flow responses are due to altered pericyte contractility, or some other change in neuron-to-vessel signaling. However, Kisler et al. performed an impressive breadth of experiments to show that many requirements for neurovascular coupling were actually unchanged in PDGFRβ+/- mice, including arteriole dilation and smooth muscle function. Thus, pericyte defects are likely a major factor in the neurovascular decoupling they observed. However, whether loss of pericyte contractile function is directly responsible requires further investigation.

    How can we examine pericyte contractility in vivo?
    As mentioned by the Zlokovic group, and pioneered by the Grutzendler lab (Hill et al., 2015), optogenetic depolarization of pericytes enables acute ”cause and effect”-type studies of whether pericytes can regulate blood flow in vivo. Data we presented at the 2016 Society for Neuroscience annual meeting suggest that selective activation of Channelrhodopsin-2 in pericytes (even those deep in the capillary bed lacking smooth muscle actin) can reduce capillary diameter and red blood cell velocity in vivo. Our data supports the notion that most pericytes have the capacity to alter capillary diameter.

    We note, however, that the level of optogenetic activation used in our studies was higher than that used by Hill et al., which suggests that the threshold of depolarization needed for contraction is higher in pericytes than in smooth muscle cells. Contraction may only occur under scenarios of more intense pericyte depolarization (likely pathological circumstances such as stroke). Optogenetics can nevertheless be used to delineate mechanisms of pericyte contractility, which will help design better investigations of normal pericyte physiology. It also remains to be tested whether specific hyperpolarization of pericytes by halorhodopsin, for example, can drive responses resembling functional hyperemia.

    Pericyte diversity
    Pericytes are a diverse group of cells both with respect to their appearance and their physiological roles. Moving forward, we will need to elucidate the identity and function of the various pericyte types that exist in the brain. In the past, there was some confusion about what cells should be considered pericytes or smooth muscle cells (Attwell et al., 2015). The use of new technologies, such as RNAseq, and more careful attention to the topological location, morphology, and biochemical profiles of mural cells will help to address this issue. Further, the development of new transgenic mice that can genetically target only pericytes, or better yet, specific pericyte subtypes, will be invaluable.

    For the study of Kisler et al., we were curious to know if all pericyte types were reduced in PDGFRβ+/- mice, even those that share both features of smooth muscle cells and pericytes (i.e., transitional pericytes, also referred to as smooth muscle cells by Hill et al.). This would be provide insight on the blood flow changes they saw, as transitional pericytes are ideally located near arterioles to regulate flow into the rest of the capillary bed. Also, we wondered if vessel diameter was a sufficient means to separate between mural cell types, or whether other biochemical and topological criterion were used.

    Despite these detailed questions, the work of Kisler et al. has added a significant amount of new knowledge on pericyte roles in the brain, and has provoked many interesting new questions moving forward.


    David A. Hartmann contributed to this comment.


    . Pericytes control key neurovascular functions and neuronal phenotype in the adult brain and during brain aging. Neuron. 2010 Nov 4;68(3):409-27. PubMed.

    . Regional Blood Flow in the Normal and Ischemic Brain Is Controlled by Arteriolar Smooth Muscle Cell Contractility and Not by Capillary Pericytes. Neuron. 2015 Jun 23; PubMed.

    . What is a pericyte?. J Cereb Blood Flow Metab. 2016 Feb;36(2):451-5. Epub 2015 Oct 14 PubMed.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Smooth Muscle Cells, Not Pericytes, Control Brain Blood Flow

Paper Citations

  1. . Pericyte contraction induced by oxidative-nitrative stress impairs capillary reflow despite successful opening of an occluded cerebral artery. Nat Med. 2009 Sep;15(9):1031-7. Epub 2009 Aug 30 PubMed.
  2. . Capillary pericytes regulate cerebral blood flow in health and disease. Nature. 2014 Apr 3;508(7494):55-60. Epub 2014 Mar 26 PubMed.

Further Reading


  1. . Pericyte loss influences Alzheimer-like neurodegeneration in mice. Nat Commun. 2013;4:2932. PubMed.
  2. . Pericytes in capillaries are contractile in vivo, but arterioles mediate functional hyperemia in the mouse brain. Proc Natl Acad Sci U S A. 2010 Dec 21;107(51):22290-5. Epub 2010 Dec 6 PubMed.

Primary Papers

  1. . Pericyte degeneration leads to neurovascular uncoupling and limits oxygen supply to brain. Nat Neurosci. 2017 Mar;20(3):406-416. Epub 2017 Jan 30 PubMed.