Intercellular Interactomics of Human Brain Endothelial Cells and Th17 Lymphocytes: A Novel Strategy for Identifying Therapeutic Targets of CNS Inflammation

Arsalan S. Haqqani; Danica B. Stanimirovic

doi:10.1155/2011/175364

Intercellular Interactomics of Human Brain Endothelial Cells and Th17 Lymphocytes: A Novel Strategy for Identifying Therapeutic Targets of CNS Inflammation

Haqqani, Arsalan S.;Stanimirovic, Danica B. 2011-06-13 00:00:00 Hindawi Publishing Corporation Cardiovascular Psychiatry and Neurology Volume 2011, Article ID 175364, 11 pages doi:10.1155/2011/175364 Hypothesis Intercellular Interactomics of Human Brain Endothelial Cells and Th17 Lymphocytes: A Novel Strategy for Identifying Therapeutic Targets of CNS Inﬂammation Arsalan S. Haqqani and Danica B. Stanimirovic Institute for Biological Sciences, National Research Council, Ottawa, ON, Canada K1A 0R6 Correspondence should be addressed to Arsalan S. Haqqani, arsalan.haqqani@nrc.ca Received 13 November 2010; Accepted 15 March 2011 Academic Editor: Daniela Kaufer Copyright © 2011 A. S. Haqqani and D. B. Stanimirovic. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Leukocyte inﬁltration across an activated brain endothelium contributes to the neuroinﬂammation seen in many neurological disorders. Recent evidence shows that IL-17-producing T-lymphocytes (e.g., Th17 cells) possess brain-homing capability and contribute to the pathogenesis of multiple sclerosis and cerebral ischemia. The leukocyte transmigration across the endothelium is a highly regulated, multistep process involving intercellular communications and interactions between the leukocytes and endothelial cells. The molecules involved in the process are attractive therapeutic targets for inhibiting leukocyte brain migration. We hypothesized and have been successful in demonstrating that molecules of potential therapeutic signiﬁcance involved in Th17-brain endothelial cell (BEC) communications and interactions can be discovered through the combination of advanced membrane/submembrane proteomic and interactomic methods. We describe elements of this strategy and preliminary results obtained in method and approach development. The Th17-BEC interaction network provides new insights into the complexity of the transmigration process mediated by well-organized, subcellularly localized molecular interactions. These molecules and interactions are potential diagnostic, therapeutic, or theranostic targets for treatment of neurological conditions accompanied or caused by leukocyte inﬁltration. 1. Leukocyte Inﬁltration in CNS Disorders A growing body of recent evidence shows that inﬁltration of a subset of IL-17-producing T-lymphocytes into the The central nervous system (CNS) has long been regarded as CNS contributes to the pathogenesis of multiple sclerosis, an “immune privileged” organ, being both immunologically and cerebral ischemia. In multiple sclerosis these cells are inert and immunologically separated from the peripheral CD4+ T helper 17 (Th17) lymphocytes that have CNS- immune system [1]. Current data, however, indicates that homing properties and mediate the immune response the CNS is both immune competent and actively interactive directed at the myelin sheath. Gyulveszi and coworkers with the peripheral immune system [2]. In physiological recently demonstrated that the CNS tropism of Th17 cells conditions, a limited number of peripheral immune cells is driven by IL23, since T cells defective in IL-23 signaling cross the blood-brain barrier (BBB) and enter the CNS in a fail to accumulate in the CNS in the mouse model of process called “immune surveillance” [1]. Many neurological experimental autoimmune encephalomyelitis (EAE) [6]. diseases are associated with a much higher rate of leukocyte These ﬁndings are consistent with the report by Prat and traﬃcking into the CNS, resulting in leukocyte inﬁltration colleagues, showing that IL-23-stimulated Th17 lymphocytes and leukocyte-mediated neuronal damage. CNS inﬂamma- promote BBB disruption in vitro and in the EAE, eﬃciently tion is a major contributor to the diverse forms of brain penetrate the BBB, kill neurons, and promote further injury seen in cerebral ischemia, multiple sclerosis, cerebral CNS inﬂammation through CD4+ lymphocyte recruitment infection, and epilepsy [3–5]. [7]. 2 Cardiovascular Psychiatry and Neurology In experimental ischemic stroke, T-lymphocytes are lo- route, that is, through the TJ of BEC [19]. A number of calized to the infarction boundary zones [8] and contribute intercellular interacting pairs (IIPs), including JAM1-JAM1, to the development of secondary inﬂammatory brain injury JAM1-LFA1, PECAM1-PECAM1, and CD99-CD99, have [9]. More recently, Shichita et al. have shown that in mice been identiﬁed between BEC TJ and leukocytes [20]. How- subjected to a transient middle cerebral artery occlusion ever, the paracellular migration route has been repeatedly an initial inﬁltration of IL-23-producing macrophages is challenged by histological and electron microscopy studies followed by subsequent recruitment/activation of IL-17- demonstrating that leukocytes can migrate through the producing γδTlymphocytes [10]. These T cells concomi- transcellular pathway, that is through the BEC themselves, tantly increase downstream proinﬂammatory and neurotoxic leaving the TJ morphologically intact [17, 18, 21]. While it factors and the infarct size after focal cerebral ischemia is likely that the leukocyte migration occurs through both [10]. Thus, as in multiple sclerosis, IL17-producing T cells pathways [17], the details of the molecular events involved activated by IL-23 are believed to “home-in” towards the in either of the pathways remain limited. CNS and induce injury during cerebral stroke. 3. Therapies against Leukocyte Migration 2. Molecular Mechanisms of Leukocyte into the Brain: Successes and Failures Migration through the Endothelial Layer Inhibiting the interactions between the migrating leukocytes The endothelial lining of brain capillaries exhibits a special- and the BEC is an attractive therapeutic approach for inhibit- ized phenotype, commonly referred to as the blood-brain ing tissue inﬂammation. A target is usually a molecule that is barrier (BBB). These endothelial cells (ECs) function as a essential in the transmigration process and is easily accessible restrictive gate to control the composition of extracellular to therapy. The IIPs discussed above are some of the key ﬂuid in the central nervous system (CNS), selectively restrict- players in diﬀerent stages of the leukocyte transmigration ing and/or controlling the access of blood-borne molecules process and are also blood-accessible molecules, that is, to the brain [11]. The brain capillary endothelium exhibits present on either the surface of leukocytes or the luminal unique anatomical and biochemical features, including tight membranes of BEC. Blocking antibodies against the selected junctions (TJ) that form a physical barrier for a majority leukocyte-brain BEC IIPs have already been developed and of hydrophilic molecules larger than 500 Da, low pinocytic used in preclinical studies and clinical trials. activity, and the polarized expression of transporters that Two IIPs have been primarily targeted for the inhibi- control both brain inﬂux and eﬄux of molecular substrates tion of CNS inﬂammation: VCAM1-VLA4 interactions or [11]. ICAM1-LFA1 interactions. Blocking antibodies and other The luminal surface of BEC (i.e., the side accessible to drug molecules against the α4-integrin part of VLA4 have the blood) contains a thick glycocalyx, which is enriched in been shown to inhibit or reverse the EAE in various proteins and glycoproteins involved in key BBB functions, animal models (summarized in [18]). These ﬁndings led including the transport of solutes and macromolecules, to the development of a humanized monoclonal anti-α4- BEC permeability, vasoreactivity, and interactions with cir- integrin antibody, Natalizumab, which was evaluated in culating cells and platelets [12–15]. Under inﬂammatory two randomized, double-blind, placebo-controlled trials in conditions, luminal adhesion molecules in the glycocalyx patients with multiple sclerosis. The results of the trials directly interact with leukocytes during the highly-regulated, demonstrated a large beneﬁt for patients taking Natalizumab, multistep transmigration process involving tethering and including a 42% reduction in the risk of sustained disability rolling, activation, arrest, and diapedesis [16, 17]. These steps progression and a 68% reduction in relapse rate [22]. are mediated by surface adhesion molecules and cytokines The drug was initially approved by FDA in 2004, but was on both leukocytes and BEC and the avidity of interactions subsequently withdrawn after it was linked with few cases among diﬀerent intercellular interacting molecular pairs. of progressive multifocal leukoencephalopathy (PML) when The initial low-aﬃnity contacts, leading to tethering and combined with immunomodulatory treatment [22, 23]. rolling, slow down the ﬂowing leukocytes and are mediated However, after a review of safety information, the drug was by the binding of selectins on the glycocalyx of BEC to returned to the market in 2006 because its clinical beneﬁts their paired ligands on leukocytes. Firm adhesion (arrest) of outweighed the risks involved. While Natalizumab has the leukocytes is mediated by the binding of cell adhesion proven to be very successful in controlling the disease, some molecules (CAMs) on EC to their respective integrins on limitations due to side eﬀects have been reported, including the leukocyte surface. ICAM1 and VCAM1 are two well- elevated lymphocyte, basophil, and eosinophil counts, with known CAMs that are overexpressed on brain EC in response 5% of drug recipients showing circulating nucleated erythro- to inﬂammatory insults. ICAM1 can pair with either αLβ2 cytes [23]; these side eﬀects are likely due to widespread (LFA1, CD11a/CD18) or αMβ2 (Mac1, CD11b/CD18) inte- functions of α4 integrin in the hematopoietic system grins, while VCAM1 interacts with α4β1 (VLA4, CD49d/ [18, 23]. CD29) integrin on leukocytes [18]. Blocking antibody therapies targeting ICAM1 and LFA1 Leukocyte diapedesis from the luminal surface of BEC have also been developed. In vitro studies using these to its abluminal side is the last step in the transmigration antibodies have consistently shown that the ICAM1-LFA1 process and is the least well understood. The step has been interaction is important for T-cell adhesion and diapedesis traditionally believed to take place through the paracellular across brain endothelium (summarized in [18]). However, Cardiovascular Psychiatry and Neurology 3 these therapies have produced highly conﬂicting results 6. Testing the Hypothesis in CNS inﬂammation in vivo. While some studies have The ﬂow chart of proposed strategy to test the above shown a beneﬁcial outcome of blocking either ICAM1 or hypothesis is shown in Figure 1.The ﬁrst phase of the LFA1, others could not ﬁnd a signiﬁcant inhibition or even strategy includes the development of appropriate methods showed worsening of the outcome and early mortality in for membrane and submembrane proteomics and glyco- the EAE animal models [18, 24]. Furthermore, clinical trails proteomics and appropriate model system(s) for testing using humanized anti-LFA1 (Rovelizumab, LeukArrest) and the hypothesis, including the generation and curation of anti-ICAM1 (Enlimomab) antibodies have shown lack of database(s) of relevant known protein-protein and protein- eﬃcacy or serious side eﬀects. Another anti-VLA1 antibody carbohydrate interactions. The second phase involves the (Efalizumab) has been approved for the therapy of psoriasis, application of membrane and submembrane proteomic a non-CNS inﬂammatory disorder [25]. Thus, while the methods to the selected model system to identify protein role of ICAM1-LFA1 interaction is controversial in CNS and glycoprotein changes induced in endothelial cells and inﬂammation, this interaction plays an important role CNS-tropic lymphocytes under pathological conditions (i.e., (similar to VCAM1-VLA4 interaction) in peripheral, non- inﬂammation or ischemia). In the third phase,intercellular CNS inﬂammation. interacting pairs (IIPs) between brain BEC and CNS-tropic lymphocytes, with potential roles in leukocyte-endothelium 4. Unmet Needs and the Hypothesis contact or communication (i.e., interactions among secreted proteins and cell surface proteins), are identiﬁed using in 4.1. Discovery of Novel, Speciﬁc Targets. Currently, only a silico interactomics. Finally, validation of the role of identiﬁed limited number of molecules on leukocytes and the BBB IIPs in leukocyte adhesion/transmigration is accomplished are known or known to interact. For example, ICAM1-LFA1 using in vitro assays and biological readouts. Elements and and VCAM1-VLA4 pairs have been known for more than phases of this strategy, and preliminary results obtained in two decades and are still pursued as the main targets for method and approach development, are described in detail inhibiting interactions between endothelium and leukocytes below. in the brain. Targeting these interactions alone does not provide complete protection, suggesting that other adhesion 6.1. Phase 1: Development and Validation of Methods molecules may “compensate” when these interactions are blocked [26]. From problems associated with the functional 6.1.1. Membrane and Submembrane Proteomics Methods. redundancy of leukocyte-BEC molecular interactions and Proteins present on cell membranes, facing the extracellular the inhibition of peripheral immune system function with environment, are the main site of contacts between BEC current treatments, an important need has emerged to and leukocytes during adhesion and diapedesis. The contact discover molecules speciﬁcally involved in the adhesion site on BEC is mainly the luminal membrane (glycocalyx), and diapedesis of speciﬁc subsets of leukocytes through which consists of a complex mixture of proteins, glyco- the BBB that are less important in the overall immune proteins, and other molecules. Since BEC have “polarized” system competency. With emerging evidence of the role of membranes, that is, molecules on the luminal and abluminal the IL-23-induced IL-17-producing T cells in the patho- membranes are diﬀerent, there is also a need to couple genesis of multiple sclerosis and cerebral ischemia, there submembrane fractionation with proteomics to identify is a need for identiﬁcation of more speciﬁc molecules the diﬀerences between the various fractions. Analyzing on these subpopulations of leukocytes, which confer their membrane molecules has usually been diﬃcult, especially brain-tropism, to develop more cell-selective targets. Fur- using proteomics. Traditional 2D gel-based proteomic meth- ther, there is a need to identify cell-speciﬁc (and human- ods lack the sensitivity, reproducibility, and the ability to speciﬁc) molecules as targets on brain EC to minimize non- analyze the complete spectrum of membrane proteins, due speciﬁc side eﬀects of current drugs (described above). We to inherent limitation of the technology (summarized in propose that human brain EC and human brain tropic [27]). A quantitative and reproducible method for analyzing T-cells should become key models to discover novel tar- glycosylated proteins on a global scale has been lacking. gets implicated in neuroinﬂammatory cell recruitment and Over the last 5 years, technological growth in the migration. proteomics ﬁeld has led to the development of advanced nanoLC-MS-based systems that are composed of highly reproducible and sensitive nanoﬂow ultra HPLC systems 5. Hypothesis (e.g., nanoAquity), coupled with sensitive and high mass- Using advanced methods that combine membrane and accuracy MS instruments (e.g., Orbitrap) [28]. We and oth- submembrane proteomics and glycoproteomics with meth- ers have also developed bioinformatics software that analyze ods of in silico interactomics, a novel set of intercel- nanoLC-MS data, allowing the quantiﬁcation of thousands lular interactions between human brain endothelial cells of proteins and glycoproteins in multiple biological samples and human CNS-homing T cells, can be identiﬁed and [29]. In addition, the recent development of hydrazide exploited as therapeutic targets for preventing brain inﬂam- capture technology [30, 31] has enabled selective enrichment mation caused by recruitment of peripheral inﬂammatory of glycoproteins from cells and tissues for large-scale iden- cells. tiﬁcation, using nanoLC-MS-based quantitative proteomics. 4 Cardiovascular Psychiatry and Neurology Phase 1: method development Development and Membrane and secreted Luminal/abluminal proteomics and curation of interactomics membrane separation database(s) glycoproteomics Phase 2: experimental approach Create brain endothelial cell Create T17 lymphocyte data data set of membrane and set of membrane and secreted secreted (glyco) proteome (glyco)proteome Phase 3: in silico interactomics of HBEC and T17 lymphocytes Identify putative IIPs involved in Identify putative IIPs involved in cell-cell contact from cell-cell communication from interactomics database(s) interactomics database(s) Phase 4: target validation Select IIPs from phase 3 that can be targeted therapeutically Raise antibodies against selected IIPs (or their domains) Validate the role of IIPs by showing that antibod(ies) block relevant cell-cell interactions (adhesion, transmigration) Optimize blocking antibodies as therapeutic leads Figure 1: Schematic ﬂowchart of the proposed approach to identify protein-protein interactions between Th17 lymphocytes and brain endothelial cells that are functionally implicated in the recruitment of inﬂammatory/immune cells across the blood-brain barrier. Recently, we also described methods to enrich for various of the molecules [32]. Since molecular interactions are the membrane fractions (e.g., luminal, abluminal) from human fundamentals of systems biology, a large amount of eﬀort has BEC and other cells, and coupled this with hydrazide capture been invested into generating protein-interaction databases and quantitative nanoLC-MS-based proteomics [28, 30]. through large-scale experimental interactomics and curation This combination of submembrane isolation methods and of the scientiﬁc literature [33]. We propose to use the curated advanced proteomics and glycoproteomics technology is protein-protein interaction databases and systems biology needed to discover novel proteins and glycoproteins on the approaches, to identify which molecules actually interact surfaces of BEC and T cells. between BEC and brain-speciﬁc lymphocytes. This in silico interactomics methodology has a potential to identify novel 6.1.2. Intercellular Interactomics Data Base(s). Proteomics IIPsastherapeutic targets for CNS inﬂammation. and other genomics methods generate overwhelming data Protein interactions have traditionally been measured sets of molecules that become diﬃcult to follow up. Since using immunoprecipitation and yeast two-hybrid systems. validating entire “lists” becomes costly and time consuming, However, high-throughput techniques have also been devel- a single-factor/reductionist approach is often undertaken to oped that identify aﬃnity pull-down complexes using select (or “cherry pick”) a couple of molecules for further advanced proteomics [34], and systematically constructed evaluation. As a result, a signiﬁcant portion of disease- double-knockout strains in yeast have proven to be useful implicated molecules are simply overlooked. Thus, there for constructing a large-scale view of molecular interaction is a need for development and application of alternative networks [35]. As a result, a number of publicly available methodology to identify the overlooked molecules, which protein-protein interaction databases currently exist, includ- may include novel and more speciﬁc targets. Systems biology ing BIND, Human Protein Reference Database, HiMAP, is the ﬁeld of biology that aims to provide a more “holistic” BioGRID, and EcoCyc. Utilizing these existing databases and view by examining the network of interactions between datasets, we reconstructed an in-house database consisting of multiple molecules, pathways, cells, and characteristics of the more than a million molecular interactions. For the current tissue, as they converge to determine the disease implication study, we limited the interactions to immunoprecipitation Cardiovascular Psychiatry and Neurology 5 and aﬃnity pull-down assays in mammalian systems to the validity of the model for further discovery of BEC-Th17 reduce the incidence of false interactions. The resulting IIPs. mammalian protein-protein interaction database (mPPI-db) The majority of the cellular, membrane, and secreted consists of more than 200,000 nonredundant interactions, proteins identiﬁed (>75%) were not previously described and was used for the discovery of novel IIPs between human to be associated with BEC, BBB, luminal or abluminal brain endothelial cells (BEC) and lymphocytes using the membranes, CNS inﬂammation or leukocyte traﬃcking; 50 experimental approaches described below. of these molecules were surface adhesion molecules that have not been previously reported in BEC in response to inﬂammation. Thus, the use of advanced proteomic and 6.2. Phase 2: Experimental Models and Their Validation glycoproteomic tools led to identiﬁcation of a large number 6.2.1. Activated Human BEC. Brain endothelium under- of novel molecules in the luminal glycocalyx and other goes signiﬁcant changes in response to inﬂammation and cellular compartments of human BEC. becomes more receptive to interactions towards immune cells. This “activation” of the endothelium is characterized 6.2.2. Brain-Homing T Cells. The same advanced tools were by molecular and physical changes in the luminal glycocalyx. also utilized to discover novel molecules in leukocytes. Because proposed studies are focused on BEC-lymphocyte T cells with encephalo-tropism were generated by activation of lymphocytes, isolated from multiple sclerosis patients, interactions relevant for human disease, to test the hypoth- esis we used the hCMEC/D3 human brain endothelial cell with IL-23 as recently described [7]. These IL-17-producing line as a stable human in vitro model of the BBB [36]. The T-helper (i.e., Th17) cells were utilized for further analysis, cells were grown in EBM-2 media (Lonza, Walkersville, MD) which included isolation of cellular and secreted proteins supplemented with 2% FBS and were activated under serum- and performing label-free proteomics/glycoproteomics to free conditions using various inﬂammatory insults, includ- identify molecular changes. More than 2850 cellular, 1875 ing TNFα/INFγ,IL-1β, or simulated ischemia/reperfusion membrane and 450 secreted proteins and glyoproteins were conditions as previously described [26, 37, 38]. Although identiﬁed in the IL-23-activated Th17 cells. These included the key glycoproteins IL-17, IL-22, INFγ,and TNFα in lymphocyte transmigration process in vivo occurs in post- capillary venules, there are currently no in vitro BBB models the secreted fraction, validating the adequate activation of that speciﬁcally distinguish endothelium from capillaries the cells. Well-known membrane glycoproteins involved in endothelium adhesion, such as LFA1, Mac1, and VLA4, were and venules; most models, including the one used in these studies are mixed population of endothelial cells originating also detected in the activated Th17 cells. Furthermore, a from both. The hCMEC/D3 have been demonstrated to number of proteins involved in intracellular signaling and be valid for studies of BBB function, responses of brain downstream pathways were also identiﬁed, including paxilin, endothelium to inﬂammatory and infectious stimuli, and talin, vinculin, Arp2/3, wasp, and cytoskeletal proteins. Since the interaction of brain endothelium with lymphocytes or these proteins are involved in cell migration and projection— other cells [26, 36]. From the activated cells, proteins from including formation of podosome-like structures for luminal (apical) and abluminal (basolateral) membranes invasion—the Th17 cells appear to be primed for the antic- ipated attachment, adhesion, and diapedesis process. The and from cellular and secreted fractions were isolated using recently described methods [28], enriched for glycoproteins majority of the molecules (>80%) identiﬁed in the activated using hydrazide capture [30], and analyzed using nanoLC- Th17 cells (in both membrane and secreted fractions) MS-based quantitative proteomics to identify diﬀerentially were not described in the literature in association with T expressed molecules [39]. In total, more than 4500 unique lymphocytes, inﬂammation, or leukocyte traﬃcking. These proteins and glycoproteins were identiﬁed in the human newly Th17-associated molecules include 10 additional BEC, with about 650 present on each the luminal and integrins and more than 200 additional adhesion molecules, abluminal membranes. In addition, about 25–30% of all suggesting the potential importance of new molecular interactions and interacting pairs between Th17 and BEC. the proteins responded signiﬁcantly to the inﬂammatory insults, several of which were well-known indicators of BEC The summary of the experimental approach used to activation. Upregulation of surface adhesion molecules on generate membrane and secreted proteome data sets from human BEC and brain-tropic Th17 lymphocytes is shown in luminal membranes (including ICAM1, VCAM1) changes in several TJ proteins [19] and integrins, including β1, β2, Figure 2. α4, αL, αM were observed. A number of proteins involved in intracellular signaling and downstream pathways were 6.3. Phase 3: Constructing Intercellular Networks between also detected, including those involved in the recruitment hBEC and T Cells. A computational approach was under- of the “transmigratory cup” machinery [20]tothe luminal taken to construct intercellular interaction networks of membranes, for example, ezrin, moesin, radixin, and other stimuli-responsive molecules identiﬁed in human BEC and cytoskeletal proteins, suggesting that the cells have been Th17 cells. This involved searching for each protein-protein “primed” for the anticipated process of leukocyte adhesion. interaction from the mPPI-db in the two proteomic datasets, Furthermore, about 20% of the proteins were associated such that one interactant is in the BEC and the other in the with BEC-alone or BEC activation as found by large-scale Th17 dataset. The resulting network consisted of more than literature mining. These results validate adequate activation 9000 interactions and was referred to as the master IIP net- of hCMEC/D3 cells with inﬂammatory insults and suggest work, since it represented all possible IIPs between the two 6 Cardiovascular Psychiatry and Neurology Th17 lymphocytes from multiple sclerosis patients hCMEC/D3 cells Activation by inﬂammatory insults Activation by IL-23 to induce CNS tropism (versus nonactivated controls) (versus nonactivated controls) Cellular, secreted, and membrane Cellular and secreted protein isolation (apical/basolateral) protein isolation NanoLC-MS-based proteomics NanoLC-MS-based proteomics and glycoproteomics and glycoproteomics HBEC data set Th17 data set Intercellular interactomics Novel therapeutic targets Figure 2: Schematic representation of the experimental approach used for described studies. Immortalized human brain endothelial cell line, hCMEC/D3 was exposed to various inﬂammatory cytokines or oxygen-glucose deprivation to identify diﬀerentially expressed proteins in response to these stimuli in luminal and abluminal membranes as well as in secreted proteins. Th17 lymphocytes derived from MS patients were exposed in vitro to IL-23toinducetheir CNStropism and toidentifyproteins regulated by this treatment. Respective databases of regulated proteins in inﬂammation-primed brain endothelial cells (BEC) and CNS-tropic Th17 lymphocytes were subjected to in sillico interactomics analyses to map putative protein-protein interactions that may contribute to Th17 cell adhesion and transmigration across the blood-brain barrier (BBB). datasets. However, most of these IIPs are unlikely to occur IIPs are identiﬁed than there are currently known in the in vivo, since the intercellular pathological and physiological literature. Great number of molecules showed a high degree interactions and communications between T cells and BEC of interactions, that is, they interacted with more than one are limited to only accessible molecules, that is, surface and protein and with as many as 18 proteins. If accurate, these secreted molecules. To reduce the complexity of the master results suggest that interactions between T cells and the network and identify more relevant IIPs, we limited the inter- glycocalyx of endothelium are likely much more complex, actions to cell-cell “contacts”and “communications.”The involving a signiﬁcantly larger number of molecules than intercellular “contacts” consist of interactions between BEC previously believed. and Th17 surface molecules, whereas intercellular “com- More than 180 interacting pairs (IIPs) were detected munications” consist of interactions between BEC secreted between 116 Th17 membrane proteins and 62 human BEC and Th17 surface molecules, and vice versa. For surface glycocalyx proteins. About half of the molecules (55%) have molecules, only proteins containing extracellular membrane been previously associated in the literature with endothelium domains were included. Furthermore, only molecules that or T-lymphocytes in general, but very few have been were expressed under activated conditions were included. implicated speciﬁcally with either BBB, Th17 cells, or BEC- These criteria signiﬁcantly reduced the master IIP network T cell adhesion/diapedesis. This suggests that many of the and produced a more relevant network of cell-cell contacts molecular changes in brain endothelium in response to (Figure 3) and communications (Figure 4) between T cells inﬂammation are common with changes in other endothelia, and brain endothelium during leukocyte traﬃcking. and similar is true for Th17 and T cells. Some of identiﬁed IIPs between human BEC glycocalyx and Th17 membranes included VCAM1-VLA4, ICAM1-LFA1, ICAM1-Mac1, P- 6.4. Identiﬁcation of Novel IIPs between selectin-PSGL1, and E-selectin-ESL1, respectively (Figure 3, Human BEC and T Cells table insert). Most of the molecules involved in these IIPs 6.4.1. Cell-Cell Contacts. A network of cell-cell contact had high number of intercellular interactions in the network. points between human BEC and Th17 cells was generated While connection points or “nodes” like these with high to identify potential IIPs involved in the adhesion and dia- numbers of interactions in the network can potentially pedesis processes (Figure 3). The network shows interactions be used to locate potential targets to interrupt interaction between cells, more often they indicate very common, between molecules on membranes of Th17 with molecules on either luminal membranes or in the TJ of BEC. In redundant, or nonselective interactions and consequently addition, it shows the molecules on the abluminal membrane may lead to drug side eﬀects [40]. The network however of BEC that can interact with Th17 surface molecules. It suggeststhat the remaining IIPsare novel and perhapsmore is apparent from the network (Figure 3)that many more speciﬁc to human BEC-Th17 interactions. Cardiovascular Psychiatry and Neurology 7 Nodes Degree Examples of intercellular Th17 interactions VLA4 12 VCAM1 Th17 membranes 2 CAM1, JAM1 LFA1 Mac1 2 CAM1 PECAM1 2 PECAM1 CD99 1 CD99 JAM1 1 JAM1 PSGL1 1 E-selectin ICAM1 LFA1, Mac1 BEC luminal membranes VCAM1 2 VLA4 ALCAM 1 CD6 E-selectin 1 ESL1 P-selectin 1 PSGL1 BEC TJ membranes JAM1 2 JAM1 PECAM1 1 PECAM1 CD99 1 CD99 Luminal TJ BEC Abluminal Noninteracting membrane proteins Th17 membrane proteins that can interact with BEC secreted proteins BEC luminal or TJ membrane proteins that can interact with Th17 proteins BEC abluminal membrane proteins that can interact with Th17 membrane proteins Figure 3: Shown is a visual depiction of the intercellular interaction network between proteins identiﬁed in Th17 lymphocytes and hCMEC/D3 brain endothelial cells using approach described in Figure 2. Each identiﬁed protein is represented by an oval; noninteracting proteins are shown in white, membrane proteins identiﬁed in Th17 lymphocytes that can interact with luminal membrane proteins of brain endothelial cells (BEC) are shown in blue, BEC luminal membrane- and tight junction proteins that can interact with Th17 membrane proteins are shown in red, and BEC abluminal proteins that can interact with Th17 membrane proteins are shown in green. Proteins were identiﬁed using membrane and subcellular proteomics and glycoproteomics and their interactions were catalogued using protein-protein databases as described in the text (each interaction is shown as a line connecting interacting proteins). Insert: the table shows a list of known proteins identiﬁed from the interactome network. Nodes represent the proteins or connection points in the membranes of Th17 and BEC in the network, and degree is the number of intercellular connections per node. Examples of these intercellular interactions are shown in the Table. 8 Cardiovascular Psychiatry and Neurology Th17 HBEC Luminal Th17 secreted proteins that can interact with BEC membrane proteins Noninteracting membrane proteins BEC secreted proteins that can interact Th17 membrane proteins that can interact with Th17 membrane proteins with BEC secreted proteins BEC luminal membrane proteins that can interact with Th17 secreted proteins Figure 4: Extracellular signaling (communication) between Th17 and BEC. Shown is a visual depiction of the intercellular interaction network between proteins secreted by either Th17 lymphocytes or BEC and their interacting membrane counterparts expressed on the other cell type. All identiﬁed membrane proteins are represented by ovals, whereas secreted proteins are represented as circles: in red are luminal BEC membrane proteins that interact with secreted T17 proteins (in blue circles); in blue are membrane T17 proteins that interact with secreted BEC proteins (in red circles). Proteins were identiﬁed using membrane and subcellular proteomics and glycoproteomics and their interactions were catalogued using protein-protein databases as described in the text (each interaction is shown as a line connecting interacting proteins). Not unexpectedly, IIPs between human BEC TJs and respectively. More than 50 additional IIPs between TJs and Th17 membranes were also discovered (Figure 3), consistent Th17 membranes were also identiﬁed, many of which have with the postulated mechanisms of leukocyte diapedesis. not been previously associated in this context. During the paracellular diapedesis process, leukocytes may Finally, 113 IIPs between human BEC abluminal mem- encounter homophilic interactions with TJ molecules, cre- branes and Th17 membranes were also identiﬁed (Figure 3). ating zipper-like contacts that replace the interendothelial Theseare of signiﬁcancesince they might be involved in junction [18]. Corroborating this and further validating the the diapedesis process through the transcellular pathway. In generated interactome network, we detected the known IIPs the ﬁnal stages of this process, the transmigrating leukocytes JAM1-LFA1, JAM1-JAM1, PECAM1-PECAM1, and CD99- need to transverse the abluminal membrane to enter the CD99 pairs between the BEC TJs and the Th17 membranes, perivascular space and thus may recruit these IIPs. Cardiovascular Psychiatry and Neurology 9 Taken together, the cell-cell contact network provides of majority of these proteins has been conﬁrmed in brain new insights into the complexity of the adhesion and vessels in human brain tissue sections in vivo [26, 42, 43]. diapedesis processes and underscores that the intercellular Most notably, activated leukocyte cell adhesion molecule interactions involved are likely not limited to just a few well- (ALCAM) identiﬁed in these studies has been shown to known IIPs, but rather extend to dozens of well-organized, promote leukocyte traﬃcking across the BBB through cell-domain-localized molecular interactions. homotypic interaction with leukocytes; inhibition of this interaction reduced the severity and delayed the time of onset of experimental autoimmune encephalomyelitis in 6.4.2. Cell-Cell Communication. A network of cell-cell animal model. These studies provide validation that some of paracrine communications between human BEC and Th17 identiﬁed BEC-Th17 cell-cell interactions could be success- cells was also generated (Figure 4)toidentify extracel- fully targeted in vivo to inhibit leukocyte migration across lular signals potentially involved in the cross-talks dur- the BBB. ing leukocyte recruitment, adhesion, and diapedesis. The The systematic validation of the role of BEC-Th17 cell- network shows the signals released from one cell which cell interactions catalogued by methods described above interact with their known receptors on the other cell’s in facilitating leukocyte transmigration could be achieved surface (Figure 4), depicting the potential cross-talk between using a high-throughput antibody development approach leukocytes and the BBB during inﬂammation. Most of the that targets the relevant epitopes of interacting molecules. signaling molecules identiﬁed were cytokines, chemokines, For example, this can be done using antibody display hormones, and/or growth factors. While the majority of the approaches (phage or ribosome display) or in vivo immu- paracrine signaling molecules were not previously associ- nization against expressed epitopes of interacting molecules. ated in the literature with leukocyte-BBB communication, Resulting antibody “libraries” could then be screened for several expected interacting pairs were also detected. These antibodies with interaction blocking properties using var- included MCP1, RANTES, and CCL19 cytokines in the ious in vitro assays (Surface Plasmon Resonance—SPR, BEC secretome and their respective receptors on the Th17 ELISA etc.) or biological readouts that determine biological surface. Likewise, IL17, IL22, and others were detected in functions of the antibodies, for example, an eﬃcacy in the Th17 secreted milieu and their receptors on the BEC inhibiting leukocyte adhesion to brain endothelial cells glycocalyx. More than 25 additional extracellular signaling or leukocyte transmigration across in vitro BBB mod- molecules from BEC were found to have receptors on Th17. els. Antibodies eﬃcacious in modulating and inhibiting Some of these receptors are known to be involved in T- leukocyte adhesion and transmigration across the in vitro cell recruitment/activation during the process of transmigra- BBB model simultaneously validate the physiological or tion. Furthermore, more than 30 signals from Th17 were pathological role of identiﬁed IIPs. Identiﬁed “function- found to have receptors on human BEC, some of which blocking” antibodies in vitro, could proceed into testing are involved in receptor-mediated endocytosis/transcytosis. in various animal models, including EAE model [26]. The Overall, it was apparent that the paracrine communication resulting lead molecule(s) could become amenable for fur- between leukocytes and BBB under inﬂammatory condi- ther development as diagnostic, therapeutic, or theranostic tions involves a large number of complex soluble ligand- for the treatment of neurological conditions accompanied, receptor interactions which might lead to “priming” of the or caused by, leukocyte/lymphocyte inﬁltration. As a proof receptive cell for the anticipated diapedesis/transmigration of concept for such an approach, we have generated a library process. of anti-ICAM-1 single-domain (VhH) antibodies from an immunized llama phage-displayed library, among which two 6.5. Phase 4: Validation of IIPs That Can Be Targeted Thera- antibodies displayed high aﬃnity and ICAM-1 blocking peutically. Validation of human BEC-T17 IIPs that may activity in leukocyte adhesion assays (unpublished). One of be suitable for the development of (blocking) therapeu- these antibodies is currently being developed as molecular imaging agent for vascular inﬂammatory activation in stroke. tic approaches could be done using various approaches. We propose that, after selection of potentially important Current advances in antibody engineering enable generation interactions based on various bioinformatics algorithms of bispeciﬁc antibodies that could simultaneously target more than one interacting molecule involved in the process that determine “strengths” of interactions in silico, poten- tially “drugable” interactions could be identiﬁed based on of T-cell brain entry. predetermined set of criteria including target molecule The proof of concept for clinical translation of proposed approach has already been achieved with Natalizumab anti- expression in brain vessels in vivo, accessibility from systemic compartment, brain/disease speciﬁcity, and so forth. In body, showing that the inhibition of leukocyte recruitment previous studies that compared the expression of proteins into the brain by antibody that blocks one of molecu- identiﬁed by 2D-gel and ICAT proteomics in BBB model(s) lar interactions between leukocytes and BEC resulted in in vitro [38] and those identiﬁed using ICAT proteomics clinically successful control of inﬂammatory brain disease. The hypothesis and experimental approach described in in laser-captured microvessels from animals in vivo [41], we identiﬁed and validated using immunochemistry and this paper provide the opportunity to identify and clin- enzyme assays 19 commonly expressed proteins. In another ically target other important leukocyte-BEC interactions that are more speciﬁc/selective for brain-tropic leukocyte proteomics study that identiﬁed more than 40 lipid-raft- speciﬁc proteins in human BEC in vitro, the expression subsets. 10 Cardiovascular Psychiatry and Neurology 7. Conclusions from BEC and T cells also target other cell types, including cells in the neurovascular unit, as well as in the peripheral To develop new therapies for inhibiting CNS inﬂammation, tissues. It would be interesting to further elaborate these there is a need to identify novel interacting molecules intercellular networks to decipher the complex communica- between CNS-homing T cells and activated BEC. Currently tions within the CNS, as well as between the CNS and the peripheral environment. only a limited number of molecules on leukocytes and BEC are known or known to interact. We have hypothesized here, and have been successful in demonstrating, that inter- Acknowledgments acting molecules of potential therapeutic signiﬁcance can be discovered through the combination of advanced mem- The authors thank Dr. Alexandre Prat, Centre hospitalier brane/submembrane proteomic and interactomic methods. ´ ´ de l’Universitede Montreal (CHUM), for providing IL-23- A number of novel protein-protein interactions identiﬁed activated Th17 cells. They also thank Christie Delaney and between BEC luminal and Th17 membrane molecules by James Mullen for their technical assistance. these methods are key targets for inhibiting various stages of T-cell entry into brain, including T cell/BEC attachment, ﬁrm adhesion, and diapedesis. In addition, formation of References specialized domain, including recruitment of the “transmi- [1] M.J.Carson, J. M. Doose, B. Melchior,C. D. Schmid, and gratory cup” complex at luminal membranes in BEC and C. C. Ploix, “CNS immune privilege: hiding in plain sight,” evidence of podosome-associated molecules on the activated Immunological Reviews, vol. 213, no. 1, pp. 48–65, 2006. T cells, not only suggests that the cells are preparing and [2] R.M.Ransohoﬀ,P. Kivisakk, ¨ and G. Kidd, “Three or more ready for interactions, but also identiﬁes these structures routes for leukocyte migration into the central nervous as important therapeutic targets. Furthermore, interactions system,” Nature Reviews Immunology, vol. 3, no. 7, pp. 569– identiﬁed between BEC tight junction and Th17 membrane 581, 2003. molecules are key therapeutic targets for inhibiting the [3] B. V. Zlokovic, “The blood-brain barrier in health and chronic paracellular diapedesis route. The identiﬁed TJ molecules neurodegenerative disorders,” Neuron, vol. 57, no. 2, pp. 178– may also be involved in the transcellular process since 201, 2008. Muller and coworkers recently showed that the membrane [4] P. F. Fabene, M. G. Navarro, M. Martinello et al., “A role from a parajunctional reticulum of interconnected vesicles, for leukocyte-endothelial adhesion mechanisms in epilepsy,” containing JAM1, PECAM1, and CD99, is recruited to Nature Medicine, vol. 14, no. 12, pp. 1377–1383, 2008. surround the transcellularly migrating leukocytes and aids [5] G. J. del Zoppo, “Inﬂammation and the neurovascular unit in the setting of focal cerebral ischemia,” Neuroscience, vol. 158, in the process [44]. Interactions identiﬁed between BEC no. 3, pp. 972–982, 2009. abluminal and Th17 membrane molecules are also inter- [6] G. Gyulveszi, S. Haak, and B. Becher, “IL-23-driven esting as therapeutic targets since they may be involved in encephalo-tropism and Th17 polarization during CNS- the ﬁnal stages of the diapedesis process (by either route). inﬂammation in vivo,” European Journal of Immunology,vol. Finally, intercellular paracrine signaling between T cells and 39, no. 7, pp. 1864–1869, 2009. BEC is also considered an important target for inhibiting [7] H.Kebir,K.Kreymborg,I.Ifergan et al., “Human TH17 recruitment and activation of the leukocytes. Collectively, lymphocytes promote blood-brain barrier disruption and identiﬁcation of novel human BEC-Th17 IIPs not only central nervous system inﬂammation,” Nature Medicine,vol. allows more in-depth understanding of the interactions and 13, no. 10, pp. 1173–1175, 2007. communications between CNS-homing T cells and activated [8] M. Schroeter, S. Jander, O. W. Witte, and G. Stoll, “Local BEC, but also allows discovery of novel targets for therapy at immune responses in the rat cerebral cortex after middle diﬀerent stages of the CNS inﬂammation. cerebral artery occlusion,” Journal of Neuroimmunology,vol. The combination of advanced methodologies described 55, no. 2, pp. 195–203, 1994. here also has other applications. The use of emerging [9] G.Yilmaz, T. V. Arumugam,K. Y.Stokes, and D. N. Granger, membrane/submembrane proteomic and glycoproteomic “Role of T lymphocytes and interferon-γ in ischemic stroke,” Circulation, vol. 113, no. 17, pp. 2105–2112, 2006. methods is applicable to other systems to allow identiﬁca- [10] T. Shichita, Y. Sugiyama, H. Ooboshi et al., “Pivotal role of tion of blood-accessible luminal and secreted targets. The cerebral interleukin-17-producing gammadelta T cells in the intercellular interactomics has applications in understanding delayed phase of ischemic brain injury,” Nature Medicine,vol. the complex interactions and communications among the 15, no. 8, pp. 946–950, 2009. various cell types in other systems. [11] D. J. Begley and M. W. Brightman, “Structural and functional The signiﬁcance of the molecules discovered by the aspects of the blood-brain barrier,” Progress in Drug Research, advanced “omic” methodologies described here also goes vol. 61, pp. 39–78, 2003. beyond the scope of this hypothesis. Novel molecules [12] S. Weinbaum, X.Zhang, Y.Han,H.Vink, and S. C.Cowin, identiﬁed on the luminal surfaces of the activated BEC are “Mechanotransduction and ﬂow across the endothelial glyco- also attractive targets for molecular imaging-based diagnosis calyx,” Proceedings of the National Academy of Sciences of the of inﬂammatory insults in the brain since they are blood- United States of America, vol. 100, no. 13, pp. 7988–7995, 2003. accessible. Molecules identiﬁed in IL-23 stimulated Th17 [13] S. Weinbaum, J. M.Tarbell, and E.R.Damiano, “The structure cells may clarify the mechanism of tropism of these cells and function of the endothelial glycocalyx layer,” Annual toward the CNS. Furthermore, secreted signaling molecules Review of Biomedical Engineering, vol. 9, pp. 121–167, 2007. Cardiovascular Psychiatry and Neurology 11 [14] A. W. Vorbrodt,D.H.Dobrogowska, A. S.Lossinsky, and H. [30] J. J. Hill, M. J. Moreno,J. C. Y. Lam, A.S.Haqqani,and J. F. M. Wisniewski, “Ultrastructural localization of lectin recep- Kelly, “Identiﬁcation of secreted proteins regulated by cAMP tors on the luminal and abluminal aspects of brain micro- in glioblastoma cells using glycopeptide capture and label-free blood vessels,” Journal of Histochemistry and Cytochemistry, quantiﬁcation,” Proteomics, vol. 9, no. 3, pp. 535–549, 2009. vol. 34, no. 2, pp. 251–261, 1986. [31] H. Zhang, X. J. Li, D. B. Martin, and R. Aebersold, [15] J. Vogel, M. Sperandio, A. R. Pries, O. Linderkamp, P. “Identiﬁcation and quantiﬁcation of N-linked glycoproteins Gaehtgens, and W. Kuschinsky, “Inﬂuence of the endothelial using hydrazide chemistry, stable isotope labeling and mass glycocalyx on cerebral blood ﬂow in mice,” Journal of Cerebral spectrometry,” Nature Biotechnology, vol. 21, no. 6, pp. 660– Blood Flow and Metabolism, vol. 20, no. 11, pp. 1571–1578, 666, 2003. [32] S. M. Studer and N. Kaminski, “Towards systems biology [16] B. Rossi and G. Constantin, “Anti-selectin therapy for of human pulmonary ﬁbrosis,” Proceedings of the American the treatment of inﬂammatory diseases,” Inﬂammation and Thoracic Society, vol. 4, no. 1, pp. 85–91, 2007. Allergy—Drug Targets, vol. 7, no. 2, pp. 85–93, 2008. [33] K. A. Pattin and J. H. Moore, “Role for protein-protein [17] B. Engelhardt and H. Wolburg, “Mini review: transendothelial interaction databases in human genetics,” Expert Review of migration of leukocytes: through the front door or around the Proteomics, vol. 6, no. 6, pp. 647–659, 2009. side of the house?” European Journal of Immunology,vol.34, [34] M. Pellegrini, D. Haynor, and J. M. Johnson, “Protein no. 11, pp. 2955–2963, 2004. interaction networks,” Expert Review of Proteomics,vol.1,no. [18] C. Coisne, R. Lyck, and B. Engelhardt, “Therapeutic target- 2, pp. 239–249, 2004. ing of leukocyte traﬃcking across the blood-brain barrier,” [35] A. H. Y. Tong,M.Evangelista,A. B.Parsons et al., “Systematic Inﬂammation and Allergy—Drug Targets, vol. 6, no. 4, pp. 210– genetic analysis with ordered arrays of yeast deletion mutants,” 222, 2007. Science, vol. 294, no. 5550, pp. 2364–2368, 2001. [19] S. M. Stamatovic, R.F.Keep,and A. V. Andjelkovic, “Brain [36] B. B. Weksler, E. A. Subileau, N. Perrie`re et al., “Blood-brain endothelial cell-cell junctions: how to ”open” the blood brain barrier-speciﬁc properties of a human adult brain endothelial barrier,” Current Neuropharmacology, vol. 6, no. 3, pp. 179– cell line,” FASEB Journal, vol. 19, no. 13, pp. 1872–1874, 2005. 192, 2008. [37] D. Stanimirovic, A. Shapiro, J. Wong, J. Hutchison, and [20] E. S. Wittchen, “Endothelial signaling in paracellular and tran- J. Durkin, “The induction of ICAM-1 in human cere- scellular leukocyte transmigration,” Frontiers in Bioscience,vol. bromicrovascular endothelial cells (HCEC) by ischemia-like 14, pp. 2522–2545, 2009. conditions promotes enhanced neutrophil/HCEC adhesion,” [21] J. Greenwood, R. Howes, and S. Lightman, “The blood-retinal Journal of Neuroimmunology, vol. 76, no. 1-2, pp. 193–205, barrier in experimental autoimmune uveoretinitis: leukocyte interactions and functional damage,” Laboratory Investigation, [38] A. S. Haqqani, J. Kelly, E. Baumann, R. F. Haseloﬀ, I.E.Blasig, vol. 70, no. 1, pp. 39–52, 1994. and D. B. Stanimirovic, “Protein markers of ischemic insult in [22] B. A. Brown, “Natalizumab in the treatment of multiple brain endothelial cells identiﬁed using 2D gel electrophoresis sclerosis,” Therapeutics and Clinical Risk Management,vol.5, and ICAT-based quantitative proteomics,” Journal of Proteome no. 3, pp. 585–594, 2009. Research, vol. 6, no. 1, pp. 226–239, 2007. [23] R. M. Ransohoﬀ,“Natalizumab and PML,” Nature Neuro- [39] A. S. Haqqani, J. F. Kelly, and D. B. Stanimirovic, “Quantitative science, vol. 8, no. 10, p. 1275, 2005. protein proﬁling by mass spectrometry using label-free pro- teomics,” Methods in Molecular Biology, vol. 439, pp. 241–256, [24] C. T. Welsh, J. W. Rose, K. E. Hill, and J. J. Townsend, “Augmentation of adoptively transferred experimental allergic encephalomyelitis by administration of a monoclonal anti- [40] T. Hase, H. Tanaka, Y. Suzuki, S. Nakagawa, and H. Kitano, body speciﬁc for LFA-1α,” Journal of Neuroimmunology,vol. “Structure of protein interaction networks and their implica- 43, no. 1-2, pp. 161–167, 1993. tions on drug design,” PLoS Computational Biology,vol.5,no. 10, Article ID e1000550, 2009. [25] C. L. Leonardi, “Efalizumab in the treatment of psoriasis,” Dermatologic Therapy, vol. 17, no. 5, pp. 393–400, 2004. [41] A. S. Haqqani, M. Nesic, ED. Preston, E. Baumann, J. Kelly, [26] R. Cayrol,K.Wosik, J. L.Berard etal.,“Activated leukocyte and D. Stanimirovic, “Characterization of vascular protein expression patterns in cerebral ischemia/reperfusion using cell adhesion molecule promotes leukocyte traﬃcking into the laser capture microdissection and ICAT-nanoLC-MS/MS,” central nervous system,” Nature Immunology, vol.9,no. 2, pp. FASEB Journal, vol. 19, no. 13, pp. 1809–1821, 2005. 137–145, 2008. [27] A. S. Haqqani,J.F.Kelly, and D. B. Stanimirovic, “Quantitative [42] A. Dodelet-Devillers, R. Cayrol, H. J. Van et al., “Functions of lipid raft membrane microdomains at the blood-brain protein proﬁling by mass spectrometry using isotope-coded aﬃnity tags,” Methods in Molecular Biology, vol. 439, pp. 225– barrier,” Journal of Molecular Medicine, vol. 87, no. 8, pp. 765– 774, 2009. 240, 2008. [28] A. S. Haqqani,J.J.Hill, J. Mullen, and D. Stanimirovic, [43] R. Cayrol,A. S.Haqqani, I. Ifergan,A.Dodelet-Devillers, and A. Prat, “Isolation of human brain endothelial cells “Methods to Study Glycoproteins at the Blood-Brain Barrier Using Mass Spectrometry,” in Methods in Molecular Biology: and characterization of lipid raft-associated proteins by mass spectroscopy,” Methods in Molecular Biology, vol. 686, pp. 275– The Blood-Brain and Other Neural Barriers, S.Nag,Ed., The Humana Press, 2011. 295, 2011. [44] Z. Mamdouh, A. Mikhailov, and W. A. Muller, “Transcellular [29] A. S. Haqqani,J.F.Kelly, and D. B. Stanimirovic, “Quantitative protein proﬁling by mass spectrometry using label-free pro- migration of leukocytes is mediated by the endothelial lat- teomics,” Methods in Molecular Biology, vol. 439, pp. 241–256, eral border recycling compartment,” Journal of Experimental 2008. Medicine, vol. 206, no. 12, pp. 2795–2808, 2009. MEDIATORS of INFLAMMATION The Scientific Gastroenterology Journal of World Journal Research and Practice Diabetes Research Disease Markers Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 International Journal of Journal of Immunology Research Endocrinology Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Submit your manuscripts at http://www.hindawi.com BioMed PPAR Research Research International Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Journal of Obesity Evidence-Based Journal of Journal of Stem Cells Complementary and Ophthalmology International Alternative Medicine Oncology Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Parkinson’s Disease Computational and Behavioural Mathematical Methods AIDS Oxidative Medicine and in Medicine Research and Treatment Cellular Longevity Neurology Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Cardiovascular Psychiatry and Neurology Hindawi Publishing Corporation http://www.deepdyve.com/lp/hindawi-publishing-corporation/intercellular-interactomics-of-human-brain-endothelial-cells-and-th17-KLofHnNpNZ

Loading next page...

References (46)

S. Stamatovic, R. Keep, A. Andjelkovic (2008)
Brain Endothelial Cell-Cell Junctions: How to “Open” the Blood Brain Barrier
Current Neuropharmacology, 6
M. Carson, Jonathan Doose, B. Melchior, C. Schmid, C. Ploix (2006)
CNS immune privilege: hiding in plain sight
Immunological Reviews, 213
G. Yilmaz, T. Arumugam, K. Stokes, D. Granger (2006)
Role of T Lymphocytes and Interferon-γ in Ischemic Stroke
Circulation, 113
A. Haqqani, J. Kelly, D. Stanimirovic (2008)
Quantitative protein profiling by mass spectrometry using label-free proteomics.
Methods in molecular biology, 439
A. Haqqani, Jennifer Hill, J. Mullen, D. Stanimirovic (2011)
Methods to study glycoproteins at the blood-brain barrier using mass spectrometry.
Methods in molecular biology, 686
S. Weinbaum, Xiaobing Zhang, Yuefeng Han, H. Vink, S. Cowin (2003)
Mechanotransduction and flow across the endothelial glycocalyx
Proceedings of the National Academy of Sciences of the United States of America, 100
A. Haqqani, J. Kelly, D. Stanimirovic (2008)
Quantitative protein profiling by mass spectrometry using isotope-coded affinity tags.
Methods in molecular biology, 439
A. Haqqani, J. Kelly, Ewa Baumann, R. Haseloff, I. Blasig, D. Stanimirovic (2007)
Protein markers of ischemic insult in brain endothelial cells identified using 2D gel electrophoresis and ICAT-based quantitative proteomics.
Journal of proteome research, 6 1
B. Zlokovic (2008)
The Blood-Brain Barrier in Health and Chronic Neurodegenerative Disorders
Neuron, 57
C. Welsh, J. Rose, K. Hill, J. Townsend (1993)
Augmentation of adoptively transferred experimental allergic encephalomyelitis by administration of a monoclonal antibody specific for LFA-1α
Journal of Neuroimmunology, 43
M. Schroeter, S. Jander, O. Witte, G. Stoll (1994)
Local immune responses in the rat cerebral cortex after middle cerebral artery occlusion
Journal of Neuroimmunology, 55
A. Wettstein (1960)
Fortschritte der Arzneimittelforschung
Experientia, 16
S. Weinbaum, J. Tarbell, E. Damiano (2007)
The structure and function of the endothelial glycocalyx layer.
Annual review of biomedical engineering, 9
(2009)
Pivotal role of cerebral interleukin-17-producing gammadelta T cells in the delayed phase of ischemic brain injury
Nature Medicine, 15
Erika Wittchen (2009)
Endothelial signaling in paracellular and transcellular leukocyte transmigration.
Frontiers in bioscience, 14
A. Tong, M. Evangelista, A. Parsons, H. Xu, Gary Bader, N. Pagé, M. Robinson, S. Raghibizadeh, C. Hogue, H. Bussey, B. Andrews, M. Tyers, C. Boone (2001)
Systematic Genetic Analysis with Ordered Arrays of Yeast Deletion Mutants
Science, 294
Brandon Brown (2009)
Natalizumab in the treatment of multiple sclerosis
Therapeutics and Clinical Risk Management, 5
T. Hase, Hiroshi Tanaka, Yasuhiro Suzuki, S. Nakagawa, H. Kitano (2009)
Structure of Protein Interaction Networks and Their Implications on Drug Design
PLoS Computational Biology, 5
Z. Mamdouh, A. Mikhailov, W. Muller (2009)
Transcellular migration of leukocytes is mediated by the endothelial lateral border recycling compartment
The Journal of Experimental Medicine, 206
A. Dodelet-Devillers, R. Cayrol, J. Horssen, A. Haqqani, H. Vries, B. Engelhardt, J. Greenwood, A. Prat (2009)
Functions of lipid raft membrane microdomains at the blood–brain barrier
Journal of Molecular Medicine, 87
R. Cayrol, A. Haqqani, I. Ifergan, A. Dodelet-Devillers, A. Prat (2011)
Isolation of human brain endothelial cells and characterization of lipid raft-associated proteins by mass spectroscopy.
Methods in molecular biology, 686
Hui Zhang, Xiao-Jun Li, Daniel Martin, R. Aebersold (2003)
Identification and quantification of N-linked glycoproteins using hydrazide chemistry, stable isotope labeling and mass spectrometry
Nature Biotechnology, 21
Caroline Coisne, R. Lyck, B. Engelhardt (2007)
Therapeutic targeting of leukocyte trafficking across the blood-brain barrier.
Inflammation & allergy drug targets, 6 4
R. Ransohoff (2005)
Natalizumab and PML
Nature Neuroscience, 8
J. Greenwood, R. Howes, S. Lightman (1994)
The blood-retinal barrier in experimental autoimmune uveoretinitis. Leukocyte interactions and functional damage.
Laboratory investigation; a journal of technical methods and pathology, 70 1
P. Fabene, G. Mora, Marianna Martinello, Barbara Rossi, F. Merigo, L. Ottoboni, S. Bach, Stefano Angiari, D. Benati, A. Chakir, L. Zanetti, F. Schio, A. Osculati, P. Marzola, E. Nicolato, J. Homeister, Lijun Xia, J. Lowe, R. McEver, F. Osculati, A. Sbarbati, E. Butcher, G. Constantin (2008)
A role for leukocyte-endothelial adhesion mechanisms in epilepsy
Nature Medicine, 14
B. Weksler, E. Subileau, N. Perrière, P. Charneau, K. Holloway, M. Lévêque, H. Tricoire-Leignel, A. Nicotra, S. Bourdoulous, P. Turowski, D. Male, F. Roux, J. Greenwood, I. Romero, P. Couraud (2005)
Blood‐brain barrier‐specific properties of a human adult brain endothelial cell line
The FASEB Journal, 19
R. Ransohoff, P. Kivisäkk, G. Kidd (2003)
Three or more routes for leukocyte migration into the central nervous system
Nature Reviews Immunology, 3
B. Engelhardt, H. Wolburg (2004)
Mini‐review: Transendothelial migration of leukocytes: through the front door or around the side of the house?
European Journal of Immunology, 34
S. Studer, N. Kaminski (2007)
Towards systems biology of human pulmonary fibrosis.
Proceedings of the American Thoracic Society, 4 1
D. Stanimirovic, A. Shapiro, J. Wong, J. Hutchison, J. Durkin (1997)
The induction of ICAM-1 in human cerebromicrovascular endothelial cells (HCEC) by ischemia-like conditions promotes enhanced neutrophil/HCEC adhesion
Journal of Neuroimmunology, 76
(2004)
Efalizumab in the treatment of psoriasis
Dermatologic Therapy, 17
Jennifer Hill, M. Moreno, Jean Lam, A. Haqqani, J. Kelly (2009)
Identification of secreted proteins regulated by cAMP in glioblastoma cells using glycopeptide capture and label‐free quantification
PROTEOMICS, 9
D. Begley, M. Brightman (2003)
Structural and functional aspects of the blood-brain barrier.
Progress in drug research. Fortschritte der Arzneimittelforschung. Progres des recherches pharmaceutiques, 61
A. Haqqani, Momir Nesic, E. Preston, Ewa Baumann, J. Kelly, D. Stanimirovic (2005)
Characterization of vascular protein expression patterns in cerebral ischemia/reperfusion using laser capture microdissection and ICAT‐nanoLC‐MS/MS
The FASEB Journal, 19
M. Pellegrini, D. Haynor, Jason Johnson (2004)
Protein Interaction Networks
Expert Review of Proteomics, 1
Gabor Gyülveszi, Stefan Haak, B. Becher (2009)
IL‐23‐driven encephalo‐tropism and Th17 polarization during CNS‐inflammation in vivo
European Journal of Immunology, 39
G. Zoppo (2009)
Inflammation and the neurovascular unit in the setting of focal cerebral ischemia
Neuroscience, 158
T. Shichita, Y. Sugiyama, H. Ooboshi, H. Sugimori, R. Nakagawa, I. Takada, T. Iwaki, Y. Okada, M. Iida, D. Cua, Y. Iwakura, A. Yoshimura (2009)
Pivotal role of cerebral interleukin-17–producing γδT cells in the delayed phase of ischemic brain injury
Nature Medicine, 15
Hania Kebir, Katharina Kreymborg, I. Ifergan, A. Dodelet-Devillers, R. Cayrol, Monique Bernard, F. Giuliani, N. Arbour, B. Becher, A. Prat (2007)
Human TH17 lymphocytes promote blood-brain barrier disruption and central nervous system inflammation
Nature Medicine, 13
K. Pattin, J. Moore (2009)
Role for protein–protein interaction databases in human genetics
Expert Review of Proteomics, 6
W. Boehncke (2007)
Efalizumab in the treatment of psoriasis
Biologics : Targets & Therapy, 1
J. Vogel, M. Sperandio, A. Pries, O. Linderkamp, P. Gaehtgens, W. Kuschinsky (2000)
Influence of the Endothelial Glycocalyx on Cerebral Blood Flow in Mice
Journal of Cerebral Blood Flow & Metabolism, 20
Barbara Rossi, G. Constantin (2008)
Anti-selectin therapy for the treatment of inflammatory diseases.
Inflammation & allergy drug targets, 7 2
R. Cayrol, K. Wosik, J. Berard, A. Dodelet-Devillers, I. Ifergan, Hania Kebir, A. Haqqani, Katharina Kreymborg, S. Krug, R. Moumdjian, A. Bouthillier, B. Becher, N. Arbour, S. David, D. Stanimirovic, A. Prat (2008)
Activated leukocyte cell adhesion molecule promotes leukocyte trafficking into the central nervous system
Nature Immunology, 9
A. Vorbrodt, D. Dobrogowska, A. Lossinsky, H. Wiśniewski (1986)
Ultrastructural localization of lectin receptors on the luminal and abluminal aspects of brain micro-blood vessels.
The journal of histochemistry and cytochemistry : official journal of the Histochemistry Society, 34

Publisher: Hindawi Publishing Corporation
Copyright: Copyright © 2011 Arsalan S. Haqqani and Danica B. Stanimirovic. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
ISSN: 2090-0163
DOI: 10.1155/2011/175364
Publisher site: See Article on Publisher Site

Abstract

Hindawi Publishing Corporation Cardiovascular Psychiatry and Neurology Volume 2011, Article ID 175364, 11 pages doi:10.1155/2011/175364 Hypothesis Intercellular Interactomics of Human Brain Endothelial Cells and Th17 Lymphocytes: A Novel Strategy for Identifying Therapeutic Targets of CNS Inﬂammation Arsalan S. Haqqani and Danica B. Stanimirovic Institute for Biological Sciences, National Research Council, Ottawa, ON, Canada K1A 0R6 Correspondence should be addressed to Arsalan S. Haqqani, arsalan.haqqani@nrc.ca Received 13 November 2010; Accepted 15 March 2011 Academic Editor: Daniela Kaufer Copyright © 2011 A. S. Haqqani and D. B. Stanimirovic. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Leukocyte inﬁltration across an activated brain endothelium contributes to the neuroinﬂammation seen in many neurological disorders. Recent evidence shows that IL-17-producing T-lymphocytes (e.g., Th17 cells) possess brain-homing capability and contribute to the pathogenesis of multiple sclerosis and cerebral ischemia. The leukocyte transmigration across the endothelium is a highly regulated, multistep process involving intercellular communications and interactions between the leukocytes and endothelial cells. The molecules involved in the process are attractive therapeutic targets for inhibiting leukocyte brain migration. We hypothesized and have been successful in demonstrating that molecules of potential therapeutic signiﬁcance involved in Th17-brain endothelial cell (BEC) communications and interactions can be discovered through the combination of advanced membrane/submembrane proteomic and interactomic methods. We describe elements of this strategy and preliminary results obtained in method and approach development. The Th17-BEC interaction network provides new insights into the complexity of the transmigration process mediated by well-organized, subcellularly localized molecular interactions. These molecules and interactions are potential diagnostic, therapeutic, or theranostic targets for treatment of neurological conditions accompanied or caused by leukocyte inﬁltration. 1. Leukocyte Inﬁltration in CNS Disorders A growing body of recent evidence shows that inﬁltration of a subset of IL-17-producing T-lymphocytes into the The central nervous system (CNS) has long been regarded as CNS contributes to the pathogenesis of multiple sclerosis, an “immune privileged” organ, being both immunologically and cerebral ischemia. In multiple sclerosis these cells are inert and immunologically separated from the peripheral CD4+ T helper 17 (Th17) lymphocytes that have CNS- immune system [1]. Current data, however, indicates that homing properties and mediate the immune response the CNS is both immune competent and actively interactive directed at the myelin sheath. Gyulveszi and coworkers with the peripheral immune system [2]. In physiological recently demonstrated that the CNS tropism of Th17 cells conditions, a limited number of peripheral immune cells is driven by IL23, since T cells defective in IL-23 signaling cross the blood-brain barrier (BBB) and enter the CNS in a fail to accumulate in the CNS in the mouse model of process called “immune surveillance” [1]. Many neurological experimental autoimmune encephalomyelitis (EAE) [6]. diseases are associated with a much higher rate of leukocyte These ﬁndings are consistent with the report by Prat and traﬃcking into the CNS, resulting in leukocyte inﬁltration colleagues, showing that IL-23-stimulated Th17 lymphocytes and leukocyte-mediated neuronal damage. CNS inﬂamma- promote BBB disruption in vitro and in the EAE, eﬃciently tion is a major contributor to the diverse forms of brain penetrate the BBB, kill neurons, and promote further injury seen in cerebral ischemia, multiple sclerosis, cerebral CNS inﬂammation through CD4+ lymphocyte recruitment infection, and epilepsy [3–5]. [7]. 2 Cardiovascular Psychiatry and Neurology In experimental ischemic stroke, T-lymphocytes are lo- route, that is, through the TJ of BEC [19]. A number of calized to the infarction boundary zones [8] and contribute intercellular interacting pairs (IIPs), including JAM1-JAM1, to the development of secondary inﬂammatory brain injury JAM1-LFA1, PECAM1-PECAM1, and CD99-CD99, have [9]. More recently, Shichita et al. have shown that in mice been identiﬁed between BEC TJ and leukocytes [20]. How- subjected to a transient middle cerebral artery occlusion ever, the paracellular migration route has been repeatedly an initial inﬁltration of IL-23-producing macrophages is challenged by histological and electron microscopy studies followed by subsequent recruitment/activation of IL-17- demonstrating that leukocytes can migrate through the producing γδTlymphocytes [10]. These T cells concomi- transcellular pathway, that is through the BEC themselves, tantly increase downstream proinﬂammatory and neurotoxic leaving the TJ morphologically intact [17, 18, 21]. While it factors and the infarct size after focal cerebral ischemia is likely that the leukocyte migration occurs through both [10]. Thus, as in multiple sclerosis, IL17-producing T cells pathways [17], the details of the molecular events involved activated by IL-23 are believed to “home-in” towards the in either of the pathways remain limited. CNS and induce injury during cerebral stroke. 3. Therapies against Leukocyte Migration 2. Molecular Mechanisms of Leukocyte into the Brain: Successes and Failures Migration through the Endothelial Layer Inhibiting the interactions between the migrating leukocytes The endothelial lining of brain capillaries exhibits a special- and the BEC is an attractive therapeutic approach for inhibit- ized phenotype, commonly referred to as the blood-brain ing tissue inﬂammation. A target is usually a molecule that is barrier (BBB). These endothelial cells (ECs) function as a essential in the transmigration process and is easily accessible restrictive gate to control the composition of extracellular to therapy. The IIPs discussed above are some of the key ﬂuid in the central nervous system (CNS), selectively restrict- players in diﬀerent stages of the leukocyte transmigration ing and/or controlling the access of blood-borne molecules process and are also blood-accessible molecules, that is, to the brain [11]. The brain capillary endothelium exhibits present on either the surface of leukocytes or the luminal unique anatomical and biochemical features, including tight membranes of BEC. Blocking antibodies against the selected junctions (TJ) that form a physical barrier for a majority leukocyte-brain BEC IIPs have already been developed and of hydrophilic molecules larger than 500 Da, low pinocytic used in preclinical studies and clinical trials. activity, and the polarized expression of transporters that Two IIPs have been primarily targeted for the inhibi- control both brain inﬂux and eﬄux of molecular substrates tion of CNS inﬂammation: VCAM1-VLA4 interactions or [11]. ICAM1-LFA1 interactions. Blocking antibodies and other The luminal surface of BEC (i.e., the side accessible to drug molecules against the α4-integrin part of VLA4 have the blood) contains a thick glycocalyx, which is enriched in been shown to inhibit or reverse the EAE in various proteins and glycoproteins involved in key BBB functions, animal models (summarized in [18]). These ﬁndings led including the transport of solutes and macromolecules, to the development of a humanized monoclonal anti-α4- BEC permeability, vasoreactivity, and interactions with cir- integrin antibody, Natalizumab, which was evaluated in culating cells and platelets [12–15]. Under inﬂammatory two randomized, double-blind, placebo-controlled trials in conditions, luminal adhesion molecules in the glycocalyx patients with multiple sclerosis. The results of the trials directly interact with leukocytes during the highly-regulated, demonstrated a large beneﬁt for patients taking Natalizumab, multistep transmigration process involving tethering and including a 42% reduction in the risk of sustained disability rolling, activation, arrest, and diapedesis [16, 17]. These steps progression and a 68% reduction in relapse rate [22]. are mediated by surface adhesion molecules and cytokines The drug was initially approved by FDA in 2004, but was on both leukocytes and BEC and the avidity of interactions subsequently withdrawn after it was linked with few cases among diﬀerent intercellular interacting molecular pairs. of progressive multifocal leukoencephalopathy (PML) when The initial low-aﬃnity contacts, leading to tethering and combined with immunomodulatory treatment [22, 23]. rolling, slow down the ﬂowing leukocytes and are mediated However, after a review of safety information, the drug was by the binding of selectins on the glycocalyx of BEC to returned to the market in 2006 because its clinical beneﬁts their paired ligands on leukocytes. Firm adhesion (arrest) of outweighed the risks involved. While Natalizumab has the leukocytes is mediated by the binding of cell adhesion proven to be very successful in controlling the disease, some molecules (CAMs) on EC to their respective integrins on limitations due to side eﬀects have been reported, including the leukocyte surface. ICAM1 and VCAM1 are two well- elevated lymphocyte, basophil, and eosinophil counts, with known CAMs that are overexpressed on brain EC in response 5% of drug recipients showing circulating nucleated erythro- to inﬂammatory insults. ICAM1 can pair with either αLβ2 cytes [23]; these side eﬀects are likely due to widespread (LFA1, CD11a/CD18) or αMβ2 (Mac1, CD11b/CD18) inte- functions of α4 integrin in the hematopoietic system grins, while VCAM1 interacts with α4β1 (VLA4, CD49d/ [18, 23]. CD29) integrin on leukocytes [18]. Blocking antibody therapies targeting ICAM1 and LFA1 Leukocyte diapedesis from the luminal surface of BEC have also been developed. In vitro studies using these to its abluminal side is the last step in the transmigration antibodies have consistently shown that the ICAM1-LFA1 process and is the least well understood. The step has been interaction is important for T-cell adhesion and diapedesis traditionally believed to take place through the paracellular across brain endothelium (summarized in [18]). However, Cardiovascular Psychiatry and Neurology 3 these therapies have produced highly conﬂicting results 6. Testing the Hypothesis in CNS inﬂammation in vivo. While some studies have The ﬂow chart of proposed strategy to test the above shown a beneﬁcial outcome of blocking either ICAM1 or hypothesis is shown in Figure 1.The ﬁrst phase of the LFA1, others could not ﬁnd a signiﬁcant inhibition or even strategy includes the development of appropriate methods showed worsening of the outcome and early mortality in for membrane and submembrane proteomics and glyco- the EAE animal models [18, 24]. Furthermore, clinical trails proteomics and appropriate model system(s) for testing using humanized anti-LFA1 (Rovelizumab, LeukArrest) and the hypothesis, including the generation and curation of anti-ICAM1 (Enlimomab) antibodies have shown lack of database(s) of relevant known protein-protein and protein- eﬃcacy or serious side eﬀects. Another anti-VLA1 antibody carbohydrate interactions. The second phase involves the (Efalizumab) has been approved for the therapy of psoriasis, application of membrane and submembrane proteomic a non-CNS inﬂammatory disorder [25]. Thus, while the methods to the selected model system to identify protein role of ICAM1-LFA1 interaction is controversial in CNS and glycoprotein changes induced in endothelial cells and inﬂammation, this interaction plays an important role CNS-tropic lymphocytes under pathological conditions (i.e., (similar to VCAM1-VLA4 interaction) in peripheral, non- inﬂammation or ischemia). In the third phase,intercellular CNS inﬂammation. interacting pairs (IIPs) between brain BEC and CNS-tropic lymphocytes, with potential roles in leukocyte-endothelium 4. Unmet Needs and the Hypothesis contact or communication (i.e., interactions among secreted proteins and cell surface proteins), are identiﬁed using in 4.1. Discovery of Novel, Speciﬁc Targets. Currently, only a silico interactomics. Finally, validation of the role of identiﬁed limited number of molecules on leukocytes and the BBB IIPs in leukocyte adhesion/transmigration is accomplished are known or known to interact. For example, ICAM1-LFA1 using in vitro assays and biological readouts. Elements and and VCAM1-VLA4 pairs have been known for more than phases of this strategy, and preliminary results obtained in two decades and are still pursued as the main targets for method and approach development, are described in detail inhibiting interactions between endothelium and leukocytes below. in the brain. Targeting these interactions alone does not provide complete protection, suggesting that other adhesion 6.1. Phase 1: Development and Validation of Methods molecules may “compensate” when these interactions are blocked [26]. From problems associated with the functional 6.1.1. Membrane and Submembrane Proteomics Methods. redundancy of leukocyte-BEC molecular interactions and Proteins present on cell membranes, facing the extracellular the inhibition of peripheral immune system function with environment, are the main site of contacts between BEC current treatments, an important need has emerged to and leukocytes during adhesion and diapedesis. The contact discover molecules speciﬁcally involved in the adhesion site on BEC is mainly the luminal membrane (glycocalyx), and diapedesis of speciﬁc subsets of leukocytes through which consists of a complex mixture of proteins, glyco- the BBB that are less important in the overall immune proteins, and other molecules. Since BEC have “polarized” system competency. With emerging evidence of the role of membranes, that is, molecules on the luminal and abluminal the IL-23-induced IL-17-producing T cells in the patho- membranes are diﬀerent, there is also a need to couple genesis of multiple sclerosis and cerebral ischemia, there submembrane fractionation with proteomics to identify is a need for identiﬁcation of more speciﬁc molecules the diﬀerences between the various fractions. Analyzing on these subpopulations of leukocytes, which confer their membrane molecules has usually been diﬃcult, especially brain-tropism, to develop more cell-selective targets. Fur- using proteomics. Traditional 2D gel-based proteomic meth- ther, there is a need to identify cell-speciﬁc (and human- ods lack the sensitivity, reproducibility, and the ability to speciﬁc) molecules as targets on brain EC to minimize non- analyze the complete spectrum of membrane proteins, due speciﬁc side eﬀects of current drugs (described above). We to inherent limitation of the technology (summarized in propose that human brain EC and human brain tropic [27]). A quantitative and reproducible method for analyzing T-cells should become key models to discover novel tar- glycosylated proteins on a global scale has been lacking. gets implicated in neuroinﬂammatory cell recruitment and Over the last 5 years, technological growth in the migration. proteomics ﬁeld has led to the development of advanced nanoLC-MS-based systems that are composed of highly reproducible and sensitive nanoﬂow ultra HPLC systems 5. Hypothesis (e.g., nanoAquity), coupled with sensitive and high mass- Using advanced methods that combine membrane and accuracy MS instruments (e.g., Orbitrap) [28]. We and oth- submembrane proteomics and glycoproteomics with meth- ers have also developed bioinformatics software that analyze ods of in silico interactomics, a novel set of intercel- nanoLC-MS data, allowing the quantiﬁcation of thousands lular interactions between human brain endothelial cells of proteins and glycoproteins in multiple biological samples and human CNS-homing T cells, can be identiﬁed and [29]. In addition, the recent development of hydrazide exploited as therapeutic targets for preventing brain inﬂam- capture technology [30, 31] has enabled selective enrichment mation caused by recruitment of peripheral inﬂammatory of glycoproteins from cells and tissues for large-scale iden- cells. tiﬁcation, using nanoLC-MS-based quantitative proteomics. 4 Cardiovascular Psychiatry and Neurology Phase 1: method development Development and Membrane and secreted Luminal/abluminal proteomics and curation of interactomics membrane separation database(s) glycoproteomics Phase 2: experimental approach Create brain endothelial cell Create T17 lymphocyte data data set of membrane and set of membrane and secreted secreted (glyco) proteome (glyco)proteome Phase 3: in silico interactomics of HBEC and T17 lymphocytes Identify putative IIPs involved in Identify putative IIPs involved in cell-cell contact from cell-cell communication from interactomics database(s) interactomics database(s) Phase 4: target validation Select IIPs from phase 3 that can be targeted therapeutically Raise antibodies against selected IIPs (or their domains) Validate the role of IIPs by showing that antibod(ies) block relevant cell-cell interactions (adhesion, transmigration) Optimize blocking antibodies as therapeutic leads Figure 1: Schematic ﬂowchart of the proposed approach to identify protein-protein interactions between Th17 lymphocytes and brain endothelial cells that are functionally implicated in the recruitment of inﬂammatory/immune cells across the blood-brain barrier. Recently, we also described methods to enrich for various of the molecules [32]. Since molecular interactions are the membrane fractions (e.g., luminal, abluminal) from human fundamentals of systems biology, a large amount of eﬀort has BEC and other cells, and coupled this with hydrazide capture been invested into generating protein-interaction databases and quantitative nanoLC-MS-based proteomics [28, 30]. through large-scale experimental interactomics and curation This combination of submembrane isolation methods and of the scientiﬁc literature [33]. We propose to use the curated advanced proteomics and glycoproteomics technology is protein-protein interaction databases and systems biology needed to discover novel proteins and glycoproteins on the approaches, to identify which molecules actually interact surfaces of BEC and T cells. between BEC and brain-speciﬁc lymphocytes. This in silico interactomics methodology has a potential to identify novel 6.1.2. Intercellular Interactomics Data Base(s). Proteomics IIPsastherapeutic targets for CNS inﬂammation. and other genomics methods generate overwhelming data Protein interactions have traditionally been measured sets of molecules that become diﬃcult to follow up. Since using immunoprecipitation and yeast two-hybrid systems. validating entire “lists” becomes costly and time consuming, However, high-throughput techniques have also been devel- a single-factor/reductionist approach is often undertaken to oped that identify aﬃnity pull-down complexes using select (or “cherry pick”) a couple of molecules for further advanced proteomics [34], and systematically constructed evaluation. As a result, a signiﬁcant portion of disease- double-knockout strains in yeast have proven to be useful implicated molecules are simply overlooked. Thus, there for constructing a large-scale view of molecular interaction is a need for development and application of alternative networks [35]. As a result, a number of publicly available methodology to identify the overlooked molecules, which protein-protein interaction databases currently exist, includ- may include novel and more speciﬁc targets. Systems biology ing BIND, Human Protein Reference Database, HiMAP, is the ﬁeld of biology that aims to provide a more “holistic” BioGRID, and EcoCyc. Utilizing these existing databases and view by examining the network of interactions between datasets, we reconstructed an in-house database consisting of multiple molecules, pathways, cells, and characteristics of the more than a million molecular interactions. For the current tissue, as they converge to determine the disease implication study, we limited the interactions to immunoprecipitation Cardiovascular Psychiatry and Neurology 5 and aﬃnity pull-down assays in mammalian systems to the validity of the model for further discovery of BEC-Th17 reduce the incidence of false interactions. The resulting IIPs. mammalian protein-protein interaction database (mPPI-db) The majority of the cellular, membrane, and secreted consists of more than 200,000 nonredundant interactions, proteins identiﬁed (>75%) were not previously described and was used for the discovery of novel IIPs between human to be associated with BEC, BBB, luminal or abluminal brain endothelial cells (BEC) and lymphocytes using the membranes, CNS inﬂammation or leukocyte traﬃcking; 50 experimental approaches described below. of these molecules were surface adhesion molecules that have not been previously reported in BEC in response to inﬂammation. Thus, the use of advanced proteomic and 6.2. Phase 2: Experimental Models and Their Validation glycoproteomic tools led to identiﬁcation of a large number 6.2.1. Activated Human BEC. Brain endothelium under- of novel molecules in the luminal glycocalyx and other goes signiﬁcant changes in response to inﬂammation and cellular compartments of human BEC. becomes more receptive to interactions towards immune cells. This “activation” of the endothelium is characterized 6.2.2. Brain-Homing T Cells. The same advanced tools were by molecular and physical changes in the luminal glycocalyx. also utilized to discover novel molecules in leukocytes. Because proposed studies are focused on BEC-lymphocyte T cells with encephalo-tropism were generated by activation of lymphocytes, isolated from multiple sclerosis patients, interactions relevant for human disease, to test the hypoth- esis we used the hCMEC/D3 human brain endothelial cell with IL-23 as recently described [7]. These IL-17-producing line as a stable human in vitro model of the BBB [36]. The T-helper (i.e., Th17) cells were utilized for further analysis, cells were grown in EBM-2 media (Lonza, Walkersville, MD) which included isolation of cellular and secreted proteins supplemented with 2% FBS and were activated under serum- and performing label-free proteomics/glycoproteomics to free conditions using various inﬂammatory insults, includ- identify molecular changes. More than 2850 cellular, 1875 ing TNFα/INFγ,IL-1β, or simulated ischemia/reperfusion membrane and 450 secreted proteins and glyoproteins were conditions as previously described [26, 37, 38]. Although identiﬁed in the IL-23-activated Th17 cells. These included the key glycoproteins IL-17, IL-22, INFγ,and TNFα in lymphocyte transmigration process in vivo occurs in post- capillary venules, there are currently no in vitro BBB models the secreted fraction, validating the adequate activation of that speciﬁcally distinguish endothelium from capillaries the cells. Well-known membrane glycoproteins involved in endothelium adhesion, such as LFA1, Mac1, and VLA4, were and venules; most models, including the one used in these studies are mixed population of endothelial cells originating also detected in the activated Th17 cells. Furthermore, a from both. The hCMEC/D3 have been demonstrated to number of proteins involved in intracellular signaling and be valid for studies of BBB function, responses of brain downstream pathways were also identiﬁed, including paxilin, endothelium to inﬂammatory and infectious stimuli, and talin, vinculin, Arp2/3, wasp, and cytoskeletal proteins. Since the interaction of brain endothelium with lymphocytes or these proteins are involved in cell migration and projection— other cells [26, 36]. From the activated cells, proteins from including formation of podosome-like structures for luminal (apical) and abluminal (basolateral) membranes invasion—the Th17 cells appear to be primed for the antic- ipated attachment, adhesion, and diapedesis process. The and from cellular and secreted fractions were isolated using recently described methods [28], enriched for glycoproteins majority of the molecules (>80%) identiﬁed in the activated using hydrazide capture [30], and analyzed using nanoLC- Th17 cells (in both membrane and secreted fractions) MS-based quantitative proteomics to identify diﬀerentially were not described in the literature in association with T expressed molecules [39]. In total, more than 4500 unique lymphocytes, inﬂammation, or leukocyte traﬃcking. These proteins and glycoproteins were identiﬁed in the human newly Th17-associated molecules include 10 additional BEC, with about 650 present on each the luminal and integrins and more than 200 additional adhesion molecules, abluminal membranes. In addition, about 25–30% of all suggesting the potential importance of new molecular interactions and interacting pairs between Th17 and BEC. the proteins responded signiﬁcantly to the inﬂammatory insults, several of which were well-known indicators of BEC The summary of the experimental approach used to activation. Upregulation of surface adhesion molecules on generate membrane and secreted proteome data sets from human BEC and brain-tropic Th17 lymphocytes is shown in luminal membranes (including ICAM1, VCAM1) changes in several TJ proteins [19] and integrins, including β1, β2, Figure 2. α4, αL, αM were observed. A number of proteins involved in intracellular signaling and downstream pathways were 6.3. Phase 3: Constructing Intercellular Networks between also detected, including those involved in the recruitment hBEC and T Cells. A computational approach was under- of the “transmigratory cup” machinery [20]tothe luminal taken to construct intercellular interaction networks of membranes, for example, ezrin, moesin, radixin, and other stimuli-responsive molecules identiﬁed in human BEC and cytoskeletal proteins, suggesting that the cells have been Th17 cells. This involved searching for each protein-protein “primed” for the anticipated process of leukocyte adhesion. interaction from the mPPI-db in the two proteomic datasets, Furthermore, about 20% of the proteins were associated such that one interactant is in the BEC and the other in the with BEC-alone or BEC activation as found by large-scale Th17 dataset. The resulting network consisted of more than literature mining. These results validate adequate activation 9000 interactions and was referred to as the master IIP net- of hCMEC/D3 cells with inﬂammatory insults and suggest work, since it represented all possible IIPs between the two 6 Cardiovascular Psychiatry and Neurology Th17 lymphocytes from multiple sclerosis patients hCMEC/D3 cells Activation by inﬂammatory insults Activation by IL-23 to induce CNS tropism (versus nonactivated controls) (versus nonactivated controls) Cellular, secreted, and membrane Cellular and secreted protein isolation (apical/basolateral) protein isolation NanoLC-MS-based proteomics NanoLC-MS-based proteomics and glycoproteomics and glycoproteomics HBEC data set Th17 data set Intercellular interactomics Novel therapeutic targets Figure 2: Schematic representation of the experimental approach used for described studies. Immortalized human brain endothelial cell line, hCMEC/D3 was exposed to various inﬂammatory cytokines or oxygen-glucose deprivation to identify diﬀerentially expressed proteins in response to these stimuli in luminal and abluminal membranes as well as in secreted proteins. Th17 lymphocytes derived from MS patients were exposed in vitro to IL-23toinducetheir CNStropism and toidentifyproteins regulated by this treatment. Respective databases of regulated proteins in inﬂammation-primed brain endothelial cells (BEC) and CNS-tropic Th17 lymphocytes were subjected to in sillico interactomics analyses to map putative protein-protein interactions that may contribute to Th17 cell adhesion and transmigration across the blood-brain barrier (BBB). datasets. However, most of these IIPs are unlikely to occur IIPs are identiﬁed than there are currently known in the in vivo, since the intercellular pathological and physiological literature. Great number of molecules showed a high degree interactions and communications between T cells and BEC of interactions, that is, they interacted with more than one are limited to only accessible molecules, that is, surface and protein and with as many as 18 proteins. If accurate, these secreted molecules. To reduce the complexity of the master results suggest that interactions between T cells and the network and identify more relevant IIPs, we limited the inter- glycocalyx of endothelium are likely much more complex, actions to cell-cell “contacts”and “communications.”The involving a signiﬁcantly larger number of molecules than intercellular “contacts” consist of interactions between BEC previously believed. and Th17 surface molecules, whereas intercellular “com- More than 180 interacting pairs (IIPs) were detected munications” consist of interactions between BEC secreted between 116 Th17 membrane proteins and 62 human BEC and Th17 surface molecules, and vice versa. For surface glycocalyx proteins. About half of the molecules (55%) have molecules, only proteins containing extracellular membrane been previously associated in the literature with endothelium domains were included. Furthermore, only molecules that or T-lymphocytes in general, but very few have been were expressed under activated conditions were included. implicated speciﬁcally with either BBB, Th17 cells, or BEC- These criteria signiﬁcantly reduced the master IIP network T cell adhesion/diapedesis. This suggests that many of the and produced a more relevant network of cell-cell contacts molecular changes in brain endothelium in response to (Figure 3) and communications (Figure 4) between T cells inﬂammation are common with changes in other endothelia, and brain endothelium during leukocyte traﬃcking. and similar is true for Th17 and T cells. Some of identiﬁed IIPs between human BEC glycocalyx and Th17 membranes included VCAM1-VLA4, ICAM1-LFA1, ICAM1-Mac1, P- 6.4. Identiﬁcation of Novel IIPs between selectin-PSGL1, and E-selectin-ESL1, respectively (Figure 3, Human BEC and T Cells table insert). Most of the molecules involved in these IIPs 6.4.1. Cell-Cell Contacts. A network of cell-cell contact had high number of intercellular interactions in the network. points between human BEC and Th17 cells was generated While connection points or “nodes” like these with high to identify potential IIPs involved in the adhesion and dia- numbers of interactions in the network can potentially pedesis processes (Figure 3). The network shows interactions be used to locate potential targets to interrupt interaction between cells, more often they indicate very common, between molecules on membranes of Th17 with molecules on either luminal membranes or in the TJ of BEC. In redundant, or nonselective interactions and consequently addition, it shows the molecules on the abluminal membrane may lead to drug side eﬀects [40]. The network however of BEC that can interact with Th17 surface molecules. It suggeststhat the remaining IIPsare novel and perhapsmore is apparent from the network (Figure 3)that many more speciﬁc to human BEC-Th17 interactions. Cardiovascular Psychiatry and Neurology 7 Nodes Degree Examples of intercellular Th17 interactions VLA4 12 VCAM1 Th17 membranes 2 CAM1, JAM1 LFA1 Mac1 2 CAM1 PECAM1 2 PECAM1 CD99 1 CD99 JAM1 1 JAM1 PSGL1 1 E-selectin ICAM1 LFA1, Mac1 BEC luminal membranes VCAM1 2 VLA4 ALCAM 1 CD6 E-selectin 1 ESL1 P-selectin 1 PSGL1 BEC TJ membranes JAM1 2 JAM1 PECAM1 1 PECAM1 CD99 1 CD99 Luminal TJ BEC Abluminal Noninteracting membrane proteins Th17 membrane proteins that can interact with BEC secreted proteins BEC luminal or TJ membrane proteins that can interact with Th17 proteins BEC abluminal membrane proteins that can interact with Th17 membrane proteins Figure 3: Shown is a visual depiction of the intercellular interaction network between proteins identiﬁed in Th17 lymphocytes and hCMEC/D3 brain endothelial cells using approach described in Figure 2. Each identiﬁed protein is represented by an oval; noninteracting proteins are shown in white, membrane proteins identiﬁed in Th17 lymphocytes that can interact with luminal membrane proteins of brain endothelial cells (BEC) are shown in blue, BEC luminal membrane- and tight junction proteins that can interact with Th17 membrane proteins are shown in red, and BEC abluminal proteins that can interact with Th17 membrane proteins are shown in green. Proteins were identiﬁed using membrane and subcellular proteomics and glycoproteomics and their interactions were catalogued using protein-protein databases as described in the text (each interaction is shown as a line connecting interacting proteins). Insert: the table shows a list of known proteins identiﬁed from the interactome network. Nodes represent the proteins or connection points in the membranes of Th17 and BEC in the network, and degree is the number of intercellular connections per node. Examples of these intercellular interactions are shown in the Table. 8 Cardiovascular Psychiatry and Neurology Th17 HBEC Luminal Th17 secreted proteins that can interact with BEC membrane proteins Noninteracting membrane proteins BEC secreted proteins that can interact Th17 membrane proteins that can interact with Th17 membrane proteins with BEC secreted proteins BEC luminal membrane proteins that can interact with Th17 secreted proteins Figure 4: Extracellular signaling (communication) between Th17 and BEC. Shown is a visual depiction of the intercellular interaction network between proteins secreted by either Th17 lymphocytes or BEC and their interacting membrane counterparts expressed on the other cell type. All identiﬁed membrane proteins are represented by ovals, whereas secreted proteins are represented as circles: in red are luminal BEC membrane proteins that interact with secreted T17 proteins (in blue circles); in blue are membrane T17 proteins that interact with secreted BEC proteins (in red circles). Proteins were identiﬁed using membrane and subcellular proteomics and glycoproteomics and their interactions were catalogued using protein-protein databases as described in the text (each interaction is shown as a line connecting interacting proteins). Not unexpectedly, IIPs between human BEC TJs and respectively. More than 50 additional IIPs between TJs and Th17 membranes were also discovered (Figure 3), consistent Th17 membranes were also identiﬁed, many of which have with the postulated mechanisms of leukocyte diapedesis. not been previously associated in this context. During the paracellular diapedesis process, leukocytes may Finally, 113 IIPs between human BEC abluminal mem- encounter homophilic interactions with TJ molecules, cre- branes and Th17 membranes were also identiﬁed (Figure 3). ating zipper-like contacts that replace the interendothelial Theseare of signiﬁcancesince they might be involved in junction [18]. Corroborating this and further validating the the diapedesis process through the transcellular pathway. In generated interactome network, we detected the known IIPs the ﬁnal stages of this process, the transmigrating leukocytes JAM1-LFA1, JAM1-JAM1, PECAM1-PECAM1, and CD99- need to transverse the abluminal membrane to enter the CD99 pairs between the BEC TJs and the Th17 membranes, perivascular space and thus may recruit these IIPs. Cardiovascular Psychiatry and Neurology 9 Taken together, the cell-cell contact network provides of majority of these proteins has been conﬁrmed in brain new insights into the complexity of the adhesion and vessels in human brain tissue sections in vivo [26, 42, 43]. diapedesis processes and underscores that the intercellular Most notably, activated leukocyte cell adhesion molecule interactions involved are likely not limited to just a few well- (ALCAM) identiﬁed in these studies has been shown to known IIPs, but rather extend to dozens of well-organized, promote leukocyte traﬃcking across the BBB through cell-domain-localized molecular interactions. homotypic interaction with leukocytes; inhibition of this interaction reduced the severity and delayed the time of onset of experimental autoimmune encephalomyelitis in 6.4.2. Cell-Cell Communication. A network of cell-cell animal model. These studies provide validation that some of paracrine communications between human BEC and Th17 identiﬁed BEC-Th17 cell-cell interactions could be success- cells was also generated (Figure 4)toidentify extracel- fully targeted in vivo to inhibit leukocyte migration across lular signals potentially involved in the cross-talks dur- the BBB. ing leukocyte recruitment, adhesion, and diapedesis. The The systematic validation of the role of BEC-Th17 cell- network shows the signals released from one cell which cell interactions catalogued by methods described above interact with their known receptors on the other cell’s in facilitating leukocyte transmigration could be achieved surface (Figure 4), depicting the potential cross-talk between using a high-throughput antibody development approach leukocytes and the BBB during inﬂammation. Most of the that targets the relevant epitopes of interacting molecules. signaling molecules identiﬁed were cytokines, chemokines, For example, this can be done using antibody display hormones, and/or growth factors. While the majority of the approaches (phage or ribosome display) or in vivo immu- paracrine signaling molecules were not previously associ- nization against expressed epitopes of interacting molecules. ated in the literature with leukocyte-BBB communication, Resulting antibody “libraries” could then be screened for several expected interacting pairs were also detected. These antibodies with interaction blocking properties using var- included MCP1, RANTES, and CCL19 cytokines in the ious in vitro assays (Surface Plasmon Resonance—SPR, BEC secretome and their respective receptors on the Th17 ELISA etc.) or biological readouts that determine biological surface. Likewise, IL17, IL22, and others were detected in functions of the antibodies, for example, an eﬃcacy in the Th17 secreted milieu and their receptors on the BEC inhibiting leukocyte adhesion to brain endothelial cells glycocalyx. More than 25 additional extracellular signaling or leukocyte transmigration across in vitro BBB mod- molecules from BEC were found to have receptors on Th17. els. Antibodies eﬃcacious in modulating and inhibiting Some of these receptors are known to be involved in T- leukocyte adhesion and transmigration across the in vitro cell recruitment/activation during the process of transmigra- BBB model simultaneously validate the physiological or tion. Furthermore, more than 30 signals from Th17 were pathological role of identiﬁed IIPs. Identiﬁed “function- found to have receptors on human BEC, some of which blocking” antibodies in vitro, could proceed into testing are involved in receptor-mediated endocytosis/transcytosis. in various animal models, including EAE model [26]. The Overall, it was apparent that the paracrine communication resulting lead molecule(s) could become amenable for fur- between leukocytes and BBB under inﬂammatory condi- ther development as diagnostic, therapeutic, or theranostic tions involves a large number of complex soluble ligand- for the treatment of neurological conditions accompanied, receptor interactions which might lead to “priming” of the or caused by, leukocyte/lymphocyte inﬁltration. As a proof receptive cell for the anticipated diapedesis/transmigration of concept for such an approach, we have generated a library process. of anti-ICAM-1 single-domain (VhH) antibodies from an immunized llama phage-displayed library, among which two 6.5. Phase 4: Validation of IIPs That Can Be Targeted Thera- antibodies displayed high aﬃnity and ICAM-1 blocking peutically. Validation of human BEC-T17 IIPs that may activity in leukocyte adhesion assays (unpublished). One of be suitable for the development of (blocking) therapeu- these antibodies is currently being developed as molecular imaging agent for vascular inﬂammatory activation in stroke. tic approaches could be done using various approaches. We propose that, after selection of potentially important Current advances in antibody engineering enable generation interactions based on various bioinformatics algorithms of bispeciﬁc antibodies that could simultaneously target more than one interacting molecule involved in the process that determine “strengths” of interactions in silico, poten- tially “drugable” interactions could be identiﬁed based on of T-cell brain entry. predetermined set of criteria including target molecule The proof of concept for clinical translation of proposed approach has already been achieved with Natalizumab anti- expression in brain vessels in vivo, accessibility from systemic compartment, brain/disease speciﬁcity, and so forth. In body, showing that the inhibition of leukocyte recruitment previous studies that compared the expression of proteins into the brain by antibody that blocks one of molecu- identiﬁed by 2D-gel and ICAT proteomics in BBB model(s) lar interactions between leukocytes and BEC resulted in in vitro [38] and those identiﬁed using ICAT proteomics clinically successful control of inﬂammatory brain disease. The hypothesis and experimental approach described in in laser-captured microvessels from animals in vivo [41], we identiﬁed and validated using immunochemistry and this paper provide the opportunity to identify and clin- enzyme assays 19 commonly expressed proteins. In another ically target other important leukocyte-BEC interactions that are more speciﬁc/selective for brain-tropic leukocyte proteomics study that identiﬁed more than 40 lipid-raft- speciﬁc proteins in human BEC in vitro, the expression subsets. 10 Cardiovascular Psychiatry and Neurology 7. Conclusions from BEC and T cells also target other cell types, including cells in the neurovascular unit, as well as in the peripheral To develop new therapies for inhibiting CNS inﬂammation, tissues. It would be interesting to further elaborate these there is a need to identify novel interacting molecules intercellular networks to decipher the complex communica- between CNS-homing T cells and activated BEC. Currently tions within the CNS, as well as between the CNS and the peripheral environment. only a limited number of molecules on leukocytes and BEC are known or known to interact. We have hypothesized here, and have been successful in demonstrating, that inter- Acknowledgments acting molecules of potential therapeutic signiﬁcance can be discovered through the combination of advanced mem- The authors thank Dr. Alexandre Prat, Centre hospitalier brane/submembrane proteomic and interactomic methods. ´ ´ de l’Universitede Montreal (CHUM), for providing IL-23- A number of novel protein-protein interactions identiﬁed activated Th17 cells. They also thank Christie Delaney and between BEC luminal and Th17 membrane molecules by James Mullen for their technical assistance. these methods are key targets for inhibiting various stages of T-cell entry into brain, including T cell/BEC attachment, ﬁrm adhesion, and diapedesis. In addition, formation of References specialized domain, including recruitment of the “transmi- [1] M.J.Carson, J. M. Doose, B. Melchior,C. D. Schmid, and gratory cup” complex at luminal membranes in BEC and C. C. Ploix, “CNS immune privilege: hiding in plain sight,” evidence of podosome-associated molecules on the activated Immunological Reviews, vol. 213, no. 1, pp. 48–65, 2006. T cells, not only suggests that the cells are preparing and [2] R.M.Ransohoﬀ,P. Kivisakk, ¨ and G. Kidd, “Three or more ready for interactions, but also identiﬁes these structures routes for leukocyte migration into the central nervous as important therapeutic targets. Furthermore, interactions system,” Nature Reviews Immunology, vol. 3, no. 7, pp. 569– identiﬁed between BEC tight junction and Th17 membrane 581, 2003. molecules are key therapeutic targets for inhibiting the [3] B. V. Zlokovic, “The blood-brain barrier in health and chronic paracellular diapedesis route. The identiﬁed TJ molecules neurodegenerative disorders,” Neuron, vol. 57, no. 2, pp. 178– may also be involved in the transcellular process since 201, 2008. Muller and coworkers recently showed that the membrane [4] P. F. Fabene, M. G. Navarro, M. Martinello et al., “A role from a parajunctional reticulum of interconnected vesicles, for leukocyte-endothelial adhesion mechanisms in epilepsy,” containing JAM1, PECAM1, and CD99, is recruited to Nature Medicine, vol. 14, no. 12, pp. 1377–1383, 2008. surround the transcellularly migrating leukocytes and aids [5] G. J. del Zoppo, “Inﬂammation and the neurovascular unit in the setting of focal cerebral ischemia,” Neuroscience, vol. 158, in the process [44]. Interactions identiﬁed between BEC no. 3, pp. 972–982, 2009. abluminal and Th17 membrane molecules are also inter- [6] G. Gyulveszi, S. Haak, and B. Becher, “IL-23-driven esting as therapeutic targets since they may be involved in encephalo-tropism and Th17 polarization during CNS- the ﬁnal stages of the diapedesis process (by either route). inﬂammation in vivo,” European Journal of Immunology,vol. Finally, intercellular paracrine signaling between T cells and 39, no. 7, pp. 1864–1869, 2009. BEC is also considered an important target for inhibiting [7] H.Kebir,K.Kreymborg,I.Ifergan et al., “Human TH17 recruitment and activation of the leukocytes. Collectively, lymphocytes promote blood-brain barrier disruption and identiﬁcation of novel human BEC-Th17 IIPs not only central nervous system inﬂammation,” Nature Medicine,vol. allows more in-depth understanding of the interactions and 13, no. 10, pp. 1173–1175, 2007. communications between CNS-homing T cells and activated [8] M. Schroeter, S. Jander, O. W. Witte, and G. Stoll, “Local BEC, but also allows discovery of novel targets for therapy at immune responses in the rat cerebral cortex after middle diﬀerent stages of the CNS inﬂammation. cerebral artery occlusion,” Journal of Neuroimmunology,vol. The combination of advanced methodologies described 55, no. 2, pp. 195–203, 1994. here also has other applications. The use of emerging [9] G.Yilmaz, T. V. Arumugam,K. Y.Stokes, and D. N. Granger, membrane/submembrane proteomic and glycoproteomic “Role of T lymphocytes and interferon-γ in ischemic stroke,” Circulation, vol. 113, no. 17, pp. 2105–2112, 2006. methods is applicable to other systems to allow identiﬁca- [10] T. Shichita, Y. Sugiyama, H. Ooboshi et al., “Pivotal role of tion of blood-accessible luminal and secreted targets. The cerebral interleukin-17-producing gammadelta T cells in the intercellular interactomics has applications in understanding delayed phase of ischemic brain injury,” Nature Medicine,vol. the complex interactions and communications among the 15, no. 8, pp. 946–950, 2009. various cell types in other systems. [11] D. J. Begley and M. W. Brightman, “Structural and functional The signiﬁcance of the molecules discovered by the aspects of the blood-brain barrier,” Progress in Drug Research, advanced “omic” methodologies described here also goes vol. 61, pp. 39–78, 2003. beyond the scope of this hypothesis. Novel molecules [12] S. Weinbaum, X.Zhang, Y.Han,H.Vink, and S. C.Cowin, identiﬁed on the luminal surfaces of the activated BEC are “Mechanotransduction and ﬂow across the endothelial glyco- also attractive targets for molecular imaging-based diagnosis calyx,” Proceedings of the National Academy of Sciences of the of inﬂammatory insults in the brain since they are blood- United States of America, vol. 100, no. 13, pp. 7988–7995, 2003. accessible. Molecules identiﬁed in IL-23 stimulated Th17 [13] S. Weinbaum, J. M.Tarbell, and E.R.Damiano, “The structure cells may clarify the mechanism of tropism of these cells and function of the endothelial glycocalyx layer,” Annual toward the CNS. Furthermore, secreted signaling molecules Review of Biomedical Engineering, vol. 9, pp. 121–167, 2007. Cardiovascular Psychiatry and Neurology 11 [14] A. W. Vorbrodt,D.H.Dobrogowska, A. S.Lossinsky, and H. [30] J. J. Hill, M. J. Moreno,J. C. Y. Lam, A.S.Haqqani,and J. F. M. Wisniewski, “Ultrastructural localization of lectin recep- Kelly, “Identiﬁcation of secreted proteins regulated by cAMP tors on the luminal and abluminal aspects of brain micro- in glioblastoma cells using glycopeptide capture and label-free blood vessels,” Journal of Histochemistry and Cytochemistry, quantiﬁcation,” Proteomics, vol. 9, no. 3, pp. 535–549, 2009. vol. 34, no. 2, pp. 251–261, 1986. [31] H. Zhang, X. J. Li, D. B. Martin, and R. Aebersold, [15] J. Vogel, M. Sperandio, A. R. Pries, O. Linderkamp, P. “Identiﬁcation and quantiﬁcation of N-linked glycoproteins Gaehtgens, and W. Kuschinsky, “Inﬂuence of the endothelial using hydrazide chemistry, stable isotope labeling and mass glycocalyx on cerebral blood ﬂow in mice,” Journal of Cerebral spectrometry,” Nature Biotechnology, vol. 21, no. 6, pp. 660– Blood Flow and Metabolism, vol. 20, no. 11, pp. 1571–1578, 666, 2003. [32] S. M. Studer and N. Kaminski, “Towards systems biology [16] B. Rossi and G. Constantin, “Anti-selectin therapy for of human pulmonary ﬁbrosis,” Proceedings of the American the treatment of inﬂammatory diseases,” Inﬂammation and Thoracic Society, vol. 4, no. 1, pp. 85–91, 2007. Allergy—Drug Targets, vol. 7, no. 2, pp. 85–93, 2008. [33] K. A. Pattin and J. H. Moore, “Role for protein-protein [17] B. Engelhardt and H. Wolburg, “Mini review: transendothelial interaction databases in human genetics,” Expert Review of migration of leukocytes: through the front door or around the Proteomics, vol. 6, no. 6, pp. 647–659, 2009. side of the house?” European Journal of Immunology,vol.34, [34] M. Pellegrini, D. Haynor, and J. M. Johnson, “Protein no. 11, pp. 2955–2963, 2004. interaction networks,” Expert Review of Proteomics,vol.1,no. [18] C. Coisne, R. Lyck, and B. Engelhardt, “Therapeutic target- 2, pp. 239–249, 2004. ing of leukocyte traﬃcking across the blood-brain barrier,” [35] A. H. Y. Tong,M.Evangelista,A. B.Parsons et al., “Systematic Inﬂammation and Allergy—Drug Targets, vol. 6, no. 4, pp. 210– genetic analysis with ordered arrays of yeast deletion mutants,” 222, 2007. Science, vol. 294, no. 5550, pp. 2364–2368, 2001. [19] S. M. Stamatovic, R.F.Keep,and A. V. Andjelkovic, “Brain [36] B. B. Weksler, E. A. Subileau, N. Perrie`re et al., “Blood-brain endothelial cell-cell junctions: how to ”open” the blood brain barrier-speciﬁc properties of a human adult brain endothelial barrier,” Current Neuropharmacology, vol. 6, no. 3, pp. 179– cell line,” FASEB Journal, vol. 19, no. 13, pp. 1872–1874, 2005. 192, 2008. [37] D. Stanimirovic, A. Shapiro, J. Wong, J. Hutchison, and [20] E. S. Wittchen, “Endothelial signaling in paracellular and tran- J. Durkin, “The induction of ICAM-1 in human cere- scellular leukocyte transmigration,” Frontiers in Bioscience,vol. bromicrovascular endothelial cells (HCEC) by ischemia-like 14, pp. 2522–2545, 2009. conditions promotes enhanced neutrophil/HCEC adhesion,” [21] J. Greenwood, R. Howes, and S. Lightman, “The blood-retinal Journal of Neuroimmunology, vol. 76, no. 1-2, pp. 193–205, barrier in experimental autoimmune uveoretinitis: leukocyte interactions and functional damage,” Laboratory Investigation, [38] A. S. Haqqani, J. Kelly, E. Baumann, R. F. Haseloﬀ, I.E.Blasig, vol. 70, no. 1, pp. 39–52, 1994. and D. B. Stanimirovic, “Protein markers of ischemic insult in [22] B. A. Brown, “Natalizumab in the treatment of multiple brain endothelial cells identiﬁed using 2D gel electrophoresis sclerosis,” Therapeutics and Clinical Risk Management,vol.5, and ICAT-based quantitative proteomics,” Journal of Proteome no. 3, pp. 585–594, 2009. Research, vol. 6, no. 1, pp. 226–239, 2007. [23] R. M. Ransohoﬀ,“Natalizumab and PML,” Nature Neuro- [39] A. S. Haqqani, J. F. Kelly, and D. B. Stanimirovic, “Quantitative science, vol. 8, no. 10, p. 1275, 2005. protein proﬁling by mass spectrometry using label-free pro- teomics,” Methods in Molecular Biology, vol. 439, pp. 241–256, [24] C. T. Welsh, J. W. Rose, K. E. Hill, and J. J. Townsend, “Augmentation of adoptively transferred experimental allergic encephalomyelitis by administration of a monoclonal anti- [40] T. Hase, H. Tanaka, Y. Suzuki, S. Nakagawa, and H. Kitano, body speciﬁc for LFA-1α,” Journal of Neuroimmunology,vol. “Structure of protein interaction networks and their implica- 43, no. 1-2, pp. 161–167, 1993. tions on drug design,” PLoS Computational Biology,vol.5,no. 10, Article ID e1000550, 2009. [25] C. L. Leonardi, “Efalizumab in the treatment of psoriasis,” Dermatologic Therapy, vol. 17, no. 5, pp. 393–400, 2004. [41] A. S. Haqqani, M. Nesic, ED. Preston, E. Baumann, J. Kelly, [26] R. Cayrol,K.Wosik, J. L.Berard etal.,“Activated leukocyte and D. Stanimirovic, “Characterization of vascular protein expression patterns in cerebral ischemia/reperfusion using cell adhesion molecule promotes leukocyte traﬃcking into the laser capture microdissection and ICAT-nanoLC-MS/MS,” central nervous system,” Nature Immunology, vol.9,no. 2, pp. FASEB Journal, vol. 19, no. 13, pp. 1809–1821, 2005. 137–145, 2008. [27] A. S. Haqqani,J.F.Kelly, and D. B. Stanimirovic, “Quantitative [42] A. Dodelet-Devillers, R. Cayrol, H. J. Van et al., “Functions of lipid raft membrane microdomains at the blood-brain protein proﬁling by mass spectrometry using isotope-coded aﬃnity tags,” Methods in Molecular Biology, vol. 439, pp. 225– barrier,” Journal of Molecular Medicine, vol. 87, no. 8, pp. 765– 774, 2009. 240, 2008. [28] A. S. Haqqani,J.J.Hill, J. Mullen, and D. Stanimirovic, [43] R. Cayrol,A. S.Haqqani, I. Ifergan,A.Dodelet-Devillers, and A. Prat, “Isolation of human brain endothelial cells “Methods to Study Glycoproteins at the Blood-Brain Barrier Using Mass Spectrometry,” in Methods in Molecular Biology: and characterization of lipid raft-associated proteins by mass spectroscopy,” Methods in Molecular Biology, vol. 686, pp. 275– The Blood-Brain and Other Neural Barriers, S.Nag,Ed., The Humana Press, 2011. 295, 2011. [44] Z. Mamdouh, A. Mikhailov, and W. A. Muller, “Transcellular [29] A. S. Haqqani,J.F.Kelly, and D. B. Stanimirovic, “Quantitative protein proﬁling by mass spectrometry using label-free pro- migration of leukocytes is mediated by the endothelial lat- teomics,” Methods in Molecular Biology, vol. 439, pp. 241–256, eral border recycling compartment,” Journal of Experimental 2008. Medicine, vol. 206, no. 12, pp. 2795–2808, 2009. MEDIATORS of INFLAMMATION The Scientific Gastroenterology Journal of World Journal Research and Practice Diabetes Research Disease Markers Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 International Journal of Journal of Immunology Research Endocrinology Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Submit your manuscripts at http://www.hindawi.com BioMed PPAR Research Research International Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Journal of Obesity Evidence-Based Journal of Journal of Stem Cells Complementary and Ophthalmology International Alternative Medicine Oncology Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Parkinson’s Disease Computational and Behavioural Mathematical Methods AIDS Oxidative Medicine and in Medicine Research and Treatment Cellular Longevity Neurology Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Journal

Cardiovascular Psychiatry and Neurology – Hindawi Publishing Corporation

Published: Jun 13, 2011

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Intercellular Interactomics of Human Brain Endothelial Cells and Th17 Lymphocytes: A Novel Strategy for Identifying Therapeutic Targets of CNS Inflammation

Intercellular Interactomics of Human Brain Endothelial Cells and Th17 Lymphocytes: A Novel Strategy for Identifying Therapeutic Targets of CNS Inflammation

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Intercellular Interactomics of Human Brain Endothelial Cells and Th17 Lymphocytes: A Novel Strategy for Identifying Therapeutic Targets of CNS Inflammation

Intercellular Interactomics of Human Brain Endothelial Cells and Th17 Lymphocytes: A Novel Strategy for Identifying Therapeutic Targets of CNS Inflammation

References (46)

Abstract

Journal

Recommended Articles

There are no references for this article.

Our policy towards the use of cookies