The present investigation
From understudied and ignored, pericytes flourished to prominence over the last decades. As explained in the preceding chapters several roles, various markers, and a spectrum of morphologies have been identified in multiple organs, in other words an intriguing cell type that deserves to be famous. In the brain the pericyte is involved in proper brain function and dysfunction, but especially in the latter there is room for advances to be made, especially their role in brain tumours. Therefore, in this thesis our main aim was to shed light on the significance of pericytes in the most aggressive of brain tumours, glioblastoma.
Through integrative studies at in silico, in vitro, and in vivo level we contribute pieces of the uncompleted pericyte puzzle. In paper 1, we utilize a combination of publicly available datasets on human glioblastoma patients and perform in silico analyses on pericyte-dominant gene signatures. To validate some of our findings in vivo, we investigated protocols to obtain a viable single cell suspension out of dissociated murine brain tissue in paper 2. In the same study, we performed scRNA-seq on the acquired cell suspension and detected distinct populations when we investigated the mural cell cluster at a higher resolution. Additional bioinformatic analysis on the sequenced brain tumour tissue from control and pericyte-poor mice in paper 3 shed light on the effects of pericytes on tumour composition. In addition to these investigations on pericytes in glioblastoma, the ongoing COVID-19 pandemic steered us into the project of paper 4, to investigate brain pericytes and their role in development of neurological symptoms in COVID-19 patients. The SARS-CoV-2 receptor ACE2 is expressed by pericytes, and we propose that the patient-specific expression of this receptor on brain pericytes can explain severity of neurological symptoms in COVID-19 patients.
At the start of this doctoral investigation, we did not make assumptions on whether pericytes are ‘good’ or ‘bad’, rather we speculated that instead of one type of pericyte, there exist several, which can be of either appearence or possibly switch between the two. Our in silico analysis of paper 1 revealed several pericyte-enriched gene (PEG) signatures that could be broadly discriminated as being overexpressed in glioblastoma patients (TCGA database) or showing no significant difference with pericytes in normal brain tissue. In addition, for both categories we
were able to designate five PEG-signatures based on gene ontology analyses of their common correlating genes (Table 2). The most impacting finding of paper 1 was that the PEG-signatures that are upregulated in glioblastoma patients are associated with a worse survival, suggesting that pericytes with those signatures can be considered ‘bad’. However, with the bioinformatic analyses of this paper we are not able to determine whether the PEG-signatures illustrate individual pericyte subtypes or rather illustrate separate states that a pericyte occupies at the time of tissue collection. Interestingly, in paper 4 we corroborated the distinct expression pattern of pericytes. We observed that ACE2 expression in brain pericytes coincided with PDGFR-β expression, but that not all PDGFR-β+ pericytes were positive for ACE2.
In addition, we noted that ACE2 expression is a patient-specific feature in brain pericytes, and that patients with strong ACE2 positivity around the brain vasculature experienced more severe neurological symptoms than those with a low or undetectable ACE2 expression.
Table 2: PEG-signatures.
To address the question of pericyte subtypes more comprehensively, we conducted an scRNA-seq study in paper 2 and 3. Where in bulk sequencing the complexity and diversity of a heterogeneous cell suspension gets diluted, single-cell transcriptomics has the capacity to resolve rare populations from a heterogeneous bulk. The technique of scRNA-seq is relatively young (Tang et al., 2009), but already developed to be indispensable for a paper of significance. The preferred scRNA-seq platform for testing subtypes of a rare cell population, such as pericytes, requires obtaining a pericyte-enriched suspension and to sequence these cells with a sufficient depth. For that reason, in paper 2, we initially attempted to sequence a pericyte-enriched cell suspension with Smart-seq2, a technique where single cells are sorted into pre-treated microtiter plates, so that in each well the mRNA content from one cell is reverse transcribed to cDNA (Picelli et al., 2013; Picelli et al., 2014).
However, after unanticipated results from the Smart-seq2 protocol we had to adjust our approach. We determined to single cell sequence whole brain tumour tissue from pericyte-poor mice and mice that have a normal pericyte coverage (Pdgfbret/ret
Signature name Gene signature
Non-significant PEGs Vascular COX4I2, CYTH3, GJA4, GPER1, HIGD1B, MIR143HG, MYO1B, PDGFRB, PEAR1, RGS5, SLC16A12 Immunological pericytes COLEC12, GPC3, MYLIP, PHLDB2
Upregulated PEGs Vascular CD248, EGFLAM, ENPEP, GGT5, HEYL, LAMC3, NDUFA4L2, NID1, NOTCH3
Immunological pericytes CALD1, LAMA2, LAMC1, MRC2
Macrophage ASPN, CTHRC1, ITGA4, KCNE4
and Pdgfbret/+, respectively) with the droplet-based technique of 10X genomics.
Although Smart-seq2 is the preferred technique to detect the highest number of genes and to detect genes with low expression levels, this method requires approximately 80 cells per cluster for sufficient power, where 10X genomics has the ability to detect rare cell populations due to the high throughput of cells (Wang et al., 2021; Ziegenhain et al., 2017). In our 10X scRNA-seq dataset, we detected several tumour cell clusters as well as cells comprising the tumour microenvironment, including a perivascular cell cluster that included pericytes (Figure 4).
In paper 2 we specifically addressed if the designated perivascular cell cluster (number 17 in Figure 4) could be additionally subdivided when applying a higher resolution. We were able to detect eight clusters among 471 cells in this perivascular cell cluster. Besides pericytes and cycling pericytes, we identified three clusters that expressed markers of mural cells, such as PVFs and VSMCs, but had a lower expression of prototypical pericyte markers, such as Kcnj8, Abcc9, and Rgs5. The perivascular cell clusters, with exception of the cycling pericytes, were not evidently separated from each other, and many genes displayed a gradient with the pericytes on one end, and PVFs on the other. The VSMCs displayed an intermediate expression in between pericytes and PVFs (Figure 5).
Figure 4. Distinct cell types in the mouse glioblastoma.
Uniform manifold approximation and projection (UMAP) of scRNA-seq (10X) of eleven mouse glioblastoma samples. A total of 21 clusters could be detected. Percentages indicate increase (green) or decrease (red) of that particular cluster in Pdgfbret/ret (4) over Pdgfbret/+ (7) mice.
Figure 5. Expression pattern of genes in perivascular cell clusters. Three genes are selected to exemplify the gradual overlap between mural cells. Kcnj8, Lum, and Acta2 represent pericytes, PVFs, and VSMCs, respectively.
Possibly, the PVFs that we detected on one end of our perivascular cell spectrum represent the same cell type that was identified by a single-cell transcriptomic study on healthy murine brain vasculature as ‘brain fibroblast-like cells’. These cells were identified by unique expression of Pdgfra, Lum, Dcn (Decorin), Cdh5, and Lama1, and the PVFs in our study share this expression pattern (Figure 6).
Figure 6. Selected genes expressed by PVFs.
Contrastingly, the same study in healthy mouse brain identified many collagen genes to be specific to the brain fibroblast-like cells, and although some of those (Col1a1, Col1a2, Col11a1, Col16a1) showed slight specificity towards the PVFs in our dataset, several other collagens were expressed throughout our perivascular cell cluster and did not show specificity for PVFs (Figure 7). The disparity between these two transcriptome datasets could possibly be explained by the fact that ours is derived from a pathologic, rather than a healthy, mouse brain.
Figure 7. Selection of collagens that are expressed throughout the perivascular cell clusters.
In addition to the pericyte cluster among the perivascular cells, we observed another distinct group of pericytes that in addition to pericyte marker genes was characterised by genes related to cell division. Future analysis will have to elucidate if these cycling pericytes have a biological meaning in glioblastoma, or that these pericytes merely happened to undergo cell division when they were captured at the time of tissue collection. A biological explanation for the dividing pericytes is not unlikely, because in the aforementioned transcriptome study of healthy mouse brain, where more mural cells were captured (1385, of which 1088 were pericytes) and sequenced deeper than in our study, there was no evidence of a cluster of cycling pericytes (Vanlandewijck et al., 2018). Again, the dissimilarity between this study and ours can be interpreted by the difference in tissue material. The excessive vascular proliferation in glioblastoma portrays activated pericytes which are highly proliferative, rather than quiescent pericytes in healthy tissue. Another interpretation of the dividing pericytes in our dataset could be that they are derived from glioblastoma cells that transdifferentiated into pericytes (Cheng et al., 2013).
Pending analyses on these pericytes will have to resolve this question.
In addition to the PVFs and mural cells (VSMCs and pericytes) in the perivascular cell cluster, we identified distinct clusters representing endothelial cells, fibroblasts, and microglia. Especially endothelial cells have been known to contaminate mural cell clusters due to their tight interaction in the brain vasculature (Vanlandewijck et al., 2018; Zeisel et al., 2015; Zhang et al., 2014). When we investigated the expression of prototypical pericyte and endothelial cell markers we observed that besides expression of genes characteristic for endothelial cells this cluster also expressed pericyte markers, suggesting pericyte-endothelial transcript contamination.
In skeletal muscle two types of pericytes have been identified based on their expression of Nestin. So-called type-2 pericytes expressed both Nestin and NG2,
while type-1 pericytes were negative for Nestin. In addition, both pericyte types expressed PDGFR-β, but only type-1 pericytes expressed PDGFR-α (Birbrair et al., 2013a). When we investigated the expression of these markers in our dataset, it is evident that Nes, as well as Pdgfrb, are abundantly expressed among the perivascular cells, while Cspg4 (the gene encoding for NG2) is primarily expressed by the pericytes (Figure 8). Together with the observation that Pdgfra is expressed by a few cells overlapping with the PVFs in our dataset (Figure 6) and brain fibroblast-like cells in healthy brain (Vanlandewijck et al., 2018), it is possible that the observed type-1 and 2 pericytes are equivalent to PVFs and pericytes, respectively. In fact, type-1 pericytes were found to accumulate around fibrotic tissue and to participate in scar tissue formation by producing collagen (Birbrair et al., 2014a)
Figure 8: Pericyte subtyping. Expression of three pericyte markers: Cspg4 (NG2), Pdgfrb, and Nes (Nestin). Nestin --/NG2+ (type-1) and Nestin+/NG2+ (type-2) pericytes are being distinguished in the literature (Birbrair et al., 2013a).
So far, our transcriptome dataset on mouse brain tumour pericytes does not provide a clear indication of the existence of pericyte subtypes, but rather a continuum of prototypical pericyte, VSMC, and fibroblast marker genes that gradually merge into one another. This could partially explain the variety of results on pericytes in health and disease. For instance, in addition to the genes discussed above, caution is advised when using CD146 (Mcam), which has been used to identify breast cancer CAFs, or CD13 (Anpep), to identify brain pericytes, in the context of murine glioblastoma, since in our dataset these markers show a reversed expression pattern (Figure 9) (Brechbuhl et al., 2020; Vanlandewijck et al., 2018).
Figure 9: Expression levels of Anpep and Mcam.CD13 (Anpep) is often used as brain pericyte-specific marker, CD146 (Mcam) is a popular pick for CAFs. Our dataset shows an opposite expression pattern.
In conclusion, based on our dataset the following genes are the most specific to pericytes: Kcnj8, Abcc9, Higd1b, Cox4i2, Gja4, Gper1, and Rgs5 (Figure 10).
Figure 10: Selection of genes with highest specificity to pericytes in our dataset of mouse glioblastoma.
Glioblastoma cells and GSCs
The complete transcriptome, comparative analyses of the tumours from Pdgfbret/ret and Pdgfbret/+ mice is discussed in paper 3. We obtained 21 distinct cell clusters, including twelve tumour cells, seven immune cell clusters, and two vascular clusters (perivascular and endothelial cells). Interestingly, while tumour cell clusters expressed typical glioblastoma genes, such as Olig2, Pdgfra, and Nes, only five out of twelve malignant cell clusters showed high expression of cell cycle gene signatures, while the other seven did not. A varying number of dividing tumour cells was observed previously in glioblastoma patients, where one sample only contained 1.4% of the cells matching to the cell cycle meta-signature and another had 21.9%.
This is in contrast with in vitro analyses of glioblastoma, that often display 100% of actively dividing cells (Patel et al., 2014). Where in human glioblastoma cells it is possible to distinguish tumour cells from stromal cells based on copy number alterations, in our mouse model the developmental time of the tumour lesions is too short to develop such alterations. Interestingly, in line with the expression pattern of quiescence-associated genes (HES1, TSC22D1, KDM5B) of non-cycling tumour cells in human glioblastoma, the non-dividing tumour cells in our dataset are enriched for KDM5B (Figure 11) (Patel et al., 2014). Ongoing analyses will have to determine whether those clusters indeed contain GSCs.
Despite evidence of pericytes contributing to maintaining stemness of GSCs in the perivascular niche, in paper 1 we were unable to detect a PEG-signature that specifically associated with stemness or quiescence. However, in the upregulated vascular PEG-signature we detected NOTCH3. Nuclear translocation of Notch3 results in Nes transcription, after TMZ treatment-induced MMP14 expression, and
has been associated with enhanced stemness after treatment with TMZ (Ulasov et al., 2020). Notch3 has been shown to be involved in stemness of normal neural stem cells as well (Than-Trong et al., 2018).
Figure 11 : Expression of quiescence-associated genes.
Pericytes in the glioblastoma vasculature
In tumours that were harvested from the Pdgfbret/ret mice in paper 3, we observed a 50% decrease in pericytes and a 157% increase in endothelial cells. This is in accordance with the literature, where it has been shown previously that a lack of pericytes causes hyperproliferation of endothelial cells (Hellstrom et al., 2001).
Future analysis on our endothelial cell cluster will have to elucidate if there is an increase in the number of tip cells, as has been previously observed in an scRNA-seq study on healthy mouse brain from Pdgfbret/ret mice (Mae et al., 2021). In paper 1, upregulated signatures were clearly different from non-significant PEG-signatures, suggesting different gene expression profiles between tumour pericytes and normal pericytes. For instance, several PEGs from the upregulated vascular cluster have been described previously to have a negative effect on glioblastoma patient survival and even to contribute to TMZ resistance. One of these upregulated vascular PEGs is the oncogene NID1, encoding for the basement membrane protein Nidogen-1. Recent in silico analysis confirmed the high expression of NID1 in glioblastoma patients and in vitro experiments demonstrated that silencing NID1 improved sensitivity of TMZ (Zhang et al., 2021a). Our work contributes in a way that it shows that NID1 is not expressed by the malignant cell populations, but by pericytes in human glioblastoma patients (paper 1) and Nid1 by vascular cells in our murine glioblastoma transcriptomic dataset from paper 3 (Figure 12).
Figure 12: Nid1 expression is specific to vascular cells in murine glioblastoma.Expression of Nid1 in the complete dataset (left) and the perivascular clusters (right).
Immune regulation of glioblastoma pericytes
In paper 1 we described PEG-signatures that were associated with vascular- and immune-related biological processes. When we examined the three immune-related signatures in more detail, we observed that two of the immune PEG-signatures were enriched in pericytes rather than in immune cells, suggesting that the immune functions were executed by the pericytes themselves. Pericytes have been described before to exert immunological functions, such as phagocytosis (Balabanov et al., 1996). When we investigated the expression of those PEGs in the transcriptome dataset from paper 2 and 3, we could see that the immunological PEGs were enriched among the PVFs, suggesting that the annotated ‘immunological PEGs’
might in fact be PVFs. Since the PEGs of paper 1 were selected based on their high expression in pericytes, this underscores the thin line between the expression pattern of pericytes and PVFs.
In contrast, the third immunological signature that we detected was expressed by myeloid cells, in addition to their expression by pericytes. Interestingly, this signature was only observed among the upregulated PEGs, and their strong association with anti-inflammatory cytokine IL-10, suggests an immunosuppressive function for tumour pericytes. Moreover, IL-10 has been demonstrated to enhance glioma cell proliferation and migration, and IL10 mRNA is higher expressed in glioblastoma compared to less invasive gliomas (Nitta et al., 1994; Widodo et al., 2021). IL-10 is primarily secreted by macrophages and microglia (Wagner et al., 1999), so possibly pericytes in glioblastoma stimulate tumour-associated macrophages and microglia to secrete increased levels of IL-10, thereby contributing to a more infiltrative and immunosuppressive tumour environment.
Interestingly, in paper 3, we observed a marked increase in microglia in pericyte-deficient mice (Figure 4). Future pathway analysis can shed light on the interactions between pericytes and microglia in glioblastoma.
Also in paper 4, where we investigated the significance of pericytes in neurological symptomatology in COVID-19 patients, the relationship between pericytes and immune cells became evident. We detected specific perivascular inflammation in the brain vasculature in SARS-CoV-2 infected patients, rather than the hypothesised neuro-inflammation. Specifically, infiltration of CD68+ macrophages and CD4+ and CD8+ T cells was observed around a subset of brain vessels in COVID-19 patients.
Future analysis will have to elucidate if this is due to active communication from ACE2+ pericytes or simply resulted as a by-stander effect due to a compromised BBB.
Pericytes in COVID-19 patients
In paper 4 we aim to elucidate the significance of pericytes in neurological symptoms of COVID-19 patients. Although we were not able to make conclusive statements that pericytes cause neurological manifestations, we unequivocally demonstrated that ACE2 expression in the context of the brain is limited to pericytes. At that time, CNS expression of ACE2 was attributed to different cell types, including glial cells, endothelial cells, and neurons (Doobay et al., 2007;
Gallagher et al., 2006; Hamming et al., 2004; Xu and Lazartigues, 2020).
Interestingly, corresponding to the highly investigated heterogeneity of pericytes, we observed that ACE2 expression was patient-specific, and moreover, that patients with strong ACE2 positivity in the CNS displayed severe neurological symptoms prior of their decease. Of note, in our in silico analysis of paper 1 we noticed very low, yet specific, expression levels of Ace2 in mouse brain pericytes. Similarly, in the same investigation we could detect a few pericytes barely positive for ACE2 in human glioblastoma patients. This discrepancy between high ACE2 protein levels and nearly undetectable mRNA levels has been observed previously in a study of murine olfactory bulb, where Ace2 transcripts were more rarely detected than protein (Brann et al., 2020).
Pericytes have been underappreciated for a long time, often being forgotten, neglected, or misinterpreted, but earned increased recognition of their significant roles over the years. However, the difficulty of distinguishing between different perivascular cells, i.e. pericytes, PVFs, and VSMCs still exists to date and utilisation of a combination of markers remains necessary to undoubtedly visualise either one of them. Although we observed differences between perivascular cells, we were not able to detect pericyte subtypes. More importantly, we contributed evidence that pericytes express distinct gene signatures in the tumour context, and we suggest that functional pericyte subsets can engage in different biological processes, such as angiogenesis or immune regulation.
Figure S1. Dutch translation of Erasmus’ letter. Letter (june 1489) from the Dutch philosopher and humanist Desiderius Erasmus Roterodamus to his dear friend Cornelis Gerards from Gouda, where he describes that his great affection for him evokes the desire to write him letters. Moreover, ‘’the more he writes of them, the more he wants to write’’, freely translated by Theo Steens from the original in Latin “Crescit scribendo scribendi studium” (Steens, 2004).
Figure S2. Early observations of pericytes. Different pericyte morphologies were already observed in the early days of vascular research. Left: Capillaries in the eye of adult frog (a capillary wall, b nucleus, c adventitia cell, d, encircling shoots of the adventitia cells. Middle, top: Capilary of adult frog (a,b adventitia cells). Note that in these first drawings by Eberth in 1871, pericytes were referred to as ‘adventia cells’ (Eberth, 1871). Abb. 114: Capillary pericyte in heart of decapitated 45 yro human. Abb. 171-187: Pre-capillary pericytes (173, 174), capillary pericytes (187), and post-capillary pericytes (171, 175) in several tissues of cats (tongue: 171, 187 kidney: 174, 175. Heart: 173) (Zimmermann, 1923).
Figure S3. Pericyte investigation in the early days required tissue from many subjects. On page 78, Zimmermann describes that he investigated human lungs, heart, kidneys, and liver from a 45 year old decapitated individual, as well as tissues from a female bear, dogs, young and adult cats, hedgehogs, rabbits, rats, guinea pigs, pigs, and rhesus monkeys. He specifically mentions that he only described the tissue where he was successful in his ‘pericyte hunt’, which was just a minor part of the many samples he investigated over the years (top). Finally, he states states that he appreciates the most his results from the human heart (bottom) (Zimmermann, 1923).
Figure S4. Stubborn Ehrlich. P. Ehrlich refuses to believe that the endothelium could be structurally different among organs (top) and was glad that others had his back (bottom) (Ehrlich, 1906). Now it is widely accepted that the BBB barrier, that Ehrlich himself observed for the first time in 1885, has a very special endothelium (Ehrlich, 1885).
Figure S5. Interesting source of pericytes.Section from the materials and methods from a study on endothelial-pericyte interactions in cutaneous tissue (Braverman and Sibley, 1990). I will leave the isolation of endothelial-pericytes from buttock skin to my successors.
(2014). Method of the year 2013. Nat Methods 11, 1.
Abramsson, A., Berlin, O., Papayan, H., Paulin, D., Shani, M., and Betsholtz, C. (2002).
Analysis of mural cell recruitment to tumor vessels. Circulation 105, 112-117.
Aguilera, K. Y., and Brekken, R. A. (2014). Recruitment and retention: factors that affect pericyte migration. Cell Mol Life Sci 71, 299-309.
Aird, W. C. (2005). Spatial and temporal dynamics of the endothelium. J Thromb Haemost 3, 1392-1406.
Alarcon-Martinez, L., Yilmaz-Ozcan, S., Yemisci, M., Schallek, J., Kilic, K., Can, A., Di Polo, A., and Dalkara, T. (2018). Capillary pericytes express alpha-smooth muscle actin, which requires prevention of filamentous-actin depolymerization for detection.
Antonelli-Orlidge, A., Saunders, K. B., Smith, S. R., and D'Amore, P. A. (1989). An activated form of transforming growth factor beta is produced by cocultures of endothelial cells and pericytes. Proc Natl Acad Sci U S A 86, 4544-4548.
Armulik, A., Genove, G., and Betsholtz, C. (2011). Pericytes: developmental,
physiological, and pathological perspectives, problems, and promises. Dev Cell 21, 193-215.
Armulik, A., Genove, G., Mae, M., Nisancioglu, M. H., Wallgard, E., Niaudet, C., He, L., Norlin, J., Lindblom, P., Strittmatter, K., et al. (2010). Pericytes regulate the blood-brain barrier. Nature 468, 557-561.
Attwell, D., Mishra, A., Hall, C. N., O'Farrell, F. M., and Dalkara, T. (2016). What is a pericyte? J Cereb Blood Flow Metab 36, 451-455.
Baisiwala, S., Auffinger, B., Caragher, S. P., Shireman, J. M., Ahsan, R., Lee, G., Hasan, T., Park, C., Saathoff, M. R., Christensen, A. C., and Ahmed, A. U. (2019).
Chemotherapeutic Stress Induces Transdifferentiation of Glioblastoma Cells to Endothelial Cells and Promotes Vascular Mimicry. Stem Cells Int 2019, 6107456.
Baker, G. J., Yadav, V. N., Motsch, S., Koschmann, C., Calinescu, A. A., Mineharu, Y., Camelo-Piragua, S. I., Orringer, D., Bannykh, S., Nichols, W. S., et al. (2014).
Mechanisms of glioma formation: iterative perivascular glioma growth and invasion leads to tumor progression, VEGF-independent vascularization, and resistance to antiangiogenic therapy. Neoplasia 16, 543-561.
Balabanov, R., Washington, R., Wagnerova, J., and Dore-Duffy, P. (1996). CNS
microvascular pericytes express macrophage-like function, cell surface integrin alpha M, and macrophage marker ED-2. Microvasc Res 52, 127-142.
Baluk, P., Morikawa, S., Haskell, A., Mancuso, M., and McDonald, D. M. (2003).
Abnormalities of basement membrane on blood vessels and endothelial sprouts in tumors. Am J Pathol 163, 1801-1815.
Bao, S., Wu, Q., McLendon, R. E., Hao, Y., Shi, Q., Hjelmeland, A. B., Dewhirst, M. W., Bigner, D. D., and Rich, J. N. (2006a). Glioma stem cells promote radioresistance by preferential activation of the DNA damage response. Nature 444, 756-760.
Bao, S., Wu, Q., Sathornsumetee, S., Hao, Y., Li, Z., Hjelmeland, A. B., Shi, Q.,
McLendon, R. E., Bigner, D. D., and Rich, J. N. (2006b). Stem cell-like glioma cells promote tumor angiogenesis through vascular endothelial growth factor. Cancer Res 66, 7843-7848.
Bardin, N., Blot-Chabaud, M., Despoix, N., Kebir, A., Harhouri, K., Arsanto, J. P., Espinosa, L., Perrin, P., Robert, S., Vely, F., et al. (2009). CD146 and its soluble form regulate monocyte transendothelial migration. Arterioscler Thromb Vasc Biol 29, 746-753.
Batchelor, T. T., Mulholland, P., Neyns, B., Nabors, L. B., Campone, M., Wick, A., Mason, W., Mikkelsen, T., Phuphanich, S., Ashby, L. S., et al. (2013). Phase III randomized trial comparing the efficacy of cediranib as monotherapy, and in combination with lomustine, versus lomustine alone in patients with recurrent glioblastoma. J Clin Oncol 31, 3212-3218.
Batchelor, T. T., Sorensen, A. G., di Tomaso, E., Zhang, W. T., Duda, D. G., Cohen, K. S., Kozak, K. R., Cahill, D. P., Chen, P. J., Zhu, M., et al. (2007). AZD2171, a pan-VEGF receptor tyrosine kinase inhibitor, normalizes tumor vasculature and alleviates edema in glioblastoma patients. Cancer Cell 11, 83-95.
Bautch, V. L., and Caron, K. M. (2015). Blood and lymphatic vessel formation. Cold Spring Harb Perspect Biol 7, a008268.
Berger, M., Bergers, G., Arnold, B., Hammerling, G. J., and Ganss, R. (2005). Regulator of G-protein signaling-5 induction in pericytes coincides with active vessel remodeling during neovascularization. Blood 105, 1094-1101.
Bergers, G., and Song, S. (2005). The role of pericytes in blood-vessel formation and maintenance. Neuro Oncol 7, 452-464.
Bergers, G., Song, S., Meyer-Morse, N., Bergsland, E., and Hanahan, D. (2003). Benefits of targeting both pericytes and endothelial cells in the tumor vasculature with kinase inhibitors. J Clin Invest 111, 1287-1295.
Berthiaume, A. A., Grant, R. I., McDowell, K. P., Underly, R. G., Hartmann, D. A., Levy, M., Bhat, N. R., and Shih, A. Y. (2018). Dynamic Remodeling of Pericytes In Vivo Maintains Capillary Coverage in the Adult Mouse Brain. Cell Rep 22, 8-16.
Birbrair, A., Zhang, T., Files, D. C., Mannava, S., Smith, T., Wang, Z. M., Messi, M. L., Mintz, A., and Delbono, O. (2014a). Type-1 pericytes accumulate after tissue injury and produce collagen in an organ-dependent manner. Stem Cell Res Ther 5, 122.
Birbrair, A., Zhang, T., Wang, Z. M., Messi, M. L., Enikolopov, G. N., Mintz, A., and Delbono, O. (2013a). Role of pericytes in skeletal muscle regeneration and fat accumulation. Stem Cells Dev 22, 2298-2314.
Birbrair, A., Zhang, T., Wang, Z. M., Messi, M. L., Enikolopov, G. N., Mintz, A., and Delbono, O. (2013b). Skeletal muscle pericyte subtypes differ in their differentiation potential. Stem Cell Res 10, 67-84.
Birbrair, A., Zhang, T., Wang, Z. M., Messi, M. L., Mintz, A., and Delbono, O. (2013c).
Type-1 pericytes participate in fibrous tissue deposition in aged skeletal muscle. Am J Physiol Cell Physiol 305, C1098-1113.
Birbrair, A., Zhang, T., Wang, Z. M., Messi, M. L., Olson, J. D., Mintz, A., and Delbono, O. (2014b). Type-2 pericytes participate in normal and tumoral angiogenesis. Am J Physiol Cell Physiol 307, C25-38.
Brann, D. H., Tsukahara, T., Weinreb, C., Lipovsek, M., Van den Berge, K., Gong, B., Chance, R., Macaulay, I. C., Chou, H. J., Fletcher, R. B., et al. (2020). Non-neuronal expression of SARS-CoV-2 entry genes in the olfactory system suggests mechanisms underlying COVID-19-associated anosmia. Sci Adv 6.
Brat, D. J., and Van Meir, E. G. (2004). Vaso-occlusive and prothrombotic mechanisms associated with tumor hypoxia, necrosis, and accelerated growth in glioblastoma. Lab Invest 84, 397-405.
Braverman, I. M., and Sibley, J. (1990). Ultrastructural and three-dimensional analysis of the contractile cells of the cutaneous microvasculature. J Invest Dermatol 95, 90-96.
Brechbuhl, H. M., Barrett, A. S., Kopin, E., Hagen, J. C., Han, A. L., Gillen, A. E., Finlay-Schultz, J., Cittelly, D. M., Owens, P., Horwitz, K. B., et al. (2020). Fibroblast subtypes define a metastatic matrisome in breast cancer. JCI Insight 5.
Bruns, R. R., and Palade, G. E. (1968). Studies on blood capillaries. I. General organization of blood capillaries in muscle. J Cell Biol 37, 244-276.
Calabrese, C., Poppleton, H., Kocak, M., Hogg, T. L., Fuller, C., Hamner, B., Oh, E. Y., Gaber, M. W., Finklestein, D., Allen, M., et al. (2007). A perivascular niche for brain tumor stem cells. Cancer Cell 11, 69-82.
Cancer Genome Atlas Research, N. (2008). Comprehensive genomic characterization defines human glioblastoma genes and core pathways. Nature 455, 1061-1068.
Cao, R., Xue, Y., Hedlund, E. M., Zhong, Z., Tritsaris, K., Tondelli, B., Lucchini, F., Zhu, Z., Dissing, S., and Cao, Y. (2010). VEGFR1-mediated pericyte ablation links VEGF and PlGF to cancer-associated retinopathy. Proc Natl Acad Sci U S A 107, 856-861.
Cao, Y. (2004). Antiangiogenic cancer therapy. Semin Cancer Biol 14, 139-145.
Carmeliet, P., Ferreira, V., Breier, G., Pollefeyt, S., Kieckens, L., Gertsenstein, M., Fahrig, M., Vandenhoeck, A., Harpal, K., Eberhardt, C., et al. (1996). Abnormal blood vessel development and lethality in embryos lacking a single VEGF allele. Nature 380, 435-439.
Caruso, R. A., Fedele, F., Finocchiaro, G., Pizzi, G., Nunnari, M., Gitto, G., Fabiano, V., Parisi, A., and Venuti, A. (2009). Ultrastructural descriptions of pericyte/endothelium peg-socket interdigitations in the microvasculature of human gastric carcinomas.
Anticancer Res 29, 449-453.
Chasseigneaux, S., Moraca, Y., Cochois-Guegan, V., Boulay, A. C., Gilbert, A., Le Crom, S., Blugeon, C., Firmo, C., Cisternino, S., Laplanche, J. L., et al. (2018). Isolation and differential transcriptome of vascular smooth muscle cells and mid-capillary pericytes from the rat brain. Sci Rep 8, 12272.