YBRBI 2433 No.

of Pages 8, Model 5G
30 August 2014

Brain, Behavior, and Immunity xxx (2014) xxx–xxx

1

Contents lists available at ScienceDirect

Brain, Behavior, and Immunity

journal homepage: www.elsevier.com/locate/ybrbi

2 Invited Review
6
4 The blood–brain barrier in neuroimmunology: Tales of separation
7
5 and assimilation
8 Q1 W.A. Banks ⇑
9 Geriatrics Research Education and Clinical Center, Veterans Affairs Puget Sound Health Care Center, Seattle, WA, United States
10 Q2 Division of Gerontology and Geriatric Medicine, University of Washington School of Medicine, Seattle, WA, United States

11
a r t i c l e i n f o a b s t r a c t
1 8
2 3
14 Article history: Neuroimmunology is concerned with the relations between the central nervous and immune systems 29
15 Received 7 August 2014 and with the mechanisms that drive those relations. The blood–brain barrier (BBB) employs mechanisms 30
16 Received in revised form 20 August 2014 that both separate and connect these two systems. In fact, the relative immune privilege of the central 31
17 Accepted 20 August 2014
nervous system (CNS) is largely attributable to the BBB’s ability to prevent the unregulated exchange 32
18 Available online xxxx
of immune cells and their secretions between the CNS and blood. Having separated the two systems, 33
the BBB then participates in mechanisms that allow them to influence, communicate, and interact with 34
19 Keywords:
one another. Likewise, the BBB itself is influenced by immune events that are occurring in the periphery 35
20 Blood–brain barrier
21 Cytokine
and in the CNS so that these three components (the BBB, the immune system, and the CNS) form neuro- 36
22 Neuroimmunology immune axes that adapt to physiological and pathological conditions. To date, four major themes have 37
23 Brain endothelial cell emerged by which the BBB participates in these neuroimmune axes. The first of these four, the formation 38
24 Pericyte of the barrier, acts to separate the immune and central nervous systems. The other three themes provide 39
25 Immune cells mechanisms for re-establishing communication: response of the BBB to immunomodulatory molecules 40
26 Central nervous system (e.g., prostaglandins, cytokines, chemokines, nitric oxide) secreted by immune and CNS cells; the con- 41
27
trolled, regulated exchange of chemokines, cytokines, and immune cells between the CNS and the blood 42
(i.e., transport across the BBB); the secretion of immunomodulatory molecules by the BBB, often in a 43
polarized fashion. Taken together, these mechanisms reveal the BBB to be a dynamic, interactive, and 44
adaptable interface between the immune system and the CNS, separating them on the one hand and fos- 45
tering their interaction on the other hand, adjusting to physiological changes, while being a target for dis- 46
ease processes. This review examines specific examples by which the BBB plays an interactive, defining 47
role in neuroimmunology. 48
Ó 2014 Published by Elsevier Inc. 49
50

51
52
53 Q3 1. Introduction was that a physical barrier existed between the brain and the blood 65
and the major contender for this site in adult mammals was the 66
54 The concept of a blood–brain barrier (BBB) arose from experi- cerebrovasculature. However, both grossly and by light micros- 67
55 ments done in Germany in the late half of the 19th and early part copy, the capillaries of the brain look no different than other cap- 68
56 of the 20th century. This included behavioral experiments, such as illary beds. It was not until the late 1960s that the ultrastructural 69
57 those of Biedl and Kraus (1898) who found that bile acids had studies of Reese and colleagues (Brightman and Reese, 1969; 70
58 effects after central but not after peripheral administration, and Reese and Karnovsky, 1967) showed that the endothelial cells of 71
59 anatomical experiments, most notably those of Paul Ehrlich who the brain differed from peripheral endothelial cells in three funda- 72
60 found that most dyes injected peripherally were unable to stain mental ways: (i) the presence of tight junctions fusing together the 73
61 the brain. Ehrlich maintained that this was because brain tissue membranes of endothelial cells in apposition; (ii) a greatly reduced 74
62 was unable to bind these dyes (Ehrlich, 1906), but later workers number of macropinocytotic vesicles; (iii) a greatly reduced num- 75
63 found that the dyes did strain brain when injected centrally ber of canaliculi and fenestrae. Thus, both the intercellular and 76
64 (Goldmann, 1913). One hypothesis to explain these phenomena transcellular routes of leakage are greatly reduced at the capillary 77
bed of the brain. 78
The lack of unregulated leakage at the BBB means that there 79
⇑ Address: VAPSHCS, 1660 S Columbian Way, Seattle, WA 98108, United States.
is no free passage of immunoactive substances from blood to 80
Tel.: +1 206 764 2701; fax: +1 206 764 2569.
brain, including immunoglobulins. The lack of production of an 81
E-mail address: wabanks1@uw.edu

http://dx.doi.org/10.1016/j.bbi.2014.08.007
0889-1591/Ó 2014 Published by Elsevier Inc.

Please cite this article in press as: Banks, W.A. The blood–brain barrier in neuroimmunology: Tales of separation and assimilation. Brain Behav. Immun.
(2014), http://dx.doi.org/10.1016/j.bbi.2014.08.007
 YBRBI 2433 No. of Pages 8, Model 5G
30 August 2014

2 W.A. Banks / Brain, Behavior, and Immunity xxx (2014) xxx–xxx

82 ultrafiltrate by the brain’s capillary bed means that the CNS does studies of their interactions involved BBB disruption (Fig. 1). For 115
83 not have a well-developed lymphatic system, a system that has over 80 years, BBB disruption in immune phenomena has been a 116
84 critical roles in immune functioning elsewhere in the body. The question of great interest, but the reasons for that interest, and 117
85 presence of a BBB also restricts the trafficking of immune cells into even how the question was posed, has shifted through the genera- 118
86 the CNS. For example, immediately after the intravenous injection tions. Some of the major questions have been: Does the ability of 119
87 of lymphocytes, about 100 times more lymphocytes are taken up bacterial pyrogen to induce fever depend on BBB disruption? What 120
88 by the axillary lymph nodes and about 800 times more by the role does BBB disruption play in the entry into the CNS of bacteria, 121
89 spleen than by the brain (Banks et al., 2012). These and other find- drugs, or immune cells (conversely, how does the trafficking into 122
90 ings led to the concept of the brain as an immune-privileged brain of bacteria, viruses, or immune cells affect measures of BBB 123
91 region, with this concept being applied early on in rather absolute integrity)? Does the vasogenic edema associated with meningitis 124
92 terms. Exceptions seemed to prove the rule as illustrated, for depend on BBB disruption and is that disruption caused by bacte- 125
93 example, by multiple sclerosis, where enhanced immune cell traf- rial pyrogen, endogenous pyrogen, cytokines, or other immunoac- 126
94 ficking was associated with dire consequences for the CNS. tive substances? What role(s) do neuroimmune processes play in 127
95 The BBB is best thought of as several barriers in parallel, includ- the disruption of the BBB seen in conditions associated with 128
96 ing the choroid plexus, which form the blood–cerebrospinal fluid immune/neuroimmune activation, such as stroke, vascular demen- 129
97 barrier and the tanycytes, which form a barrier around the circum- tia, and diabetes mellitus? Do physiological neuroimmune 130
98 ventricular organs. All these barriers, as well as the blood–spinal processes modulate a normative variance in BBB integrity? 131
99 cord barrier and the blood–retinal barrier, share common themes Many early studies investigated the appearance of pathogens in 132
100 of restricting to varying degrees the unregulated leakage of sub- the CNS and later studies were concerned with the effect that 133
101 stances between the blood and their tissue beds. Some of the pathogens had on the appearance in the CSF of antibodies, drugs, 134
102 mechanisms discussed below for the vascular BBB are known to or other circulating substances that were otherwise largely 135
103 be operational at the choroid plexus as well, so that it is likely that excluded from the CNS. Sepsis, infections, encephalitis, meningitis, 136
104 much of what is reviewed here for the vascular BBB reflects activ- and the experimental injection of bacteria or their toxins had all 137
105 ities at the other barriers as well. However, each of these barriers been associated with alterations in the BBB or changes in CSF com- 138
106 has unique adaptations that serve the special needs of their tissues position by the early half of the 20th century. For example, Skoog 139
107 and these adaptations likely extend to their neuroimmune func- reported in 1937 that allergic reactions could alter the BBB 140
108 tions as well. Discovering how these barriers integrate with each (Skoog, 1937) and that the pontobulbar region of the guinea pig 141
109 other and with the other great neuroimmune axis, the afferent was more susceptible in his model than other brain regions. 142
110 and efferent limbs of the nervous system, is a major challenge Eckman, in 1958, showed that 3–8 h after receiving an injection 143
111 for the field of neuroimmunology. into the carotid artery of 50 microg of E. coli endotoxin, rabbits 144
had increased brain staining of various substances (trypan blue, 145
Evans blue, fluorescein, and colloidal iron) that are normally 146
112 2. Barrier functions excluded by the BBB (Eckman et al., 1958). Allen showed that intra- 147
venous doses of bacterial pyrogen produced the onset of fever just 148
113 Given that the BBB is key to the separation of the central ner- as rapidly as intracarotid doses and that whereas low doses of bac- 149
114 vous and immune systems, it may not be surprising that early terial pyrogen produced fever without disrupting the BBB, high 150

Brain

IL-6

Blood-brain P-gp
Barrier

IgG LPS LPS

TNF

Mechanisms: Barrier Responder Transporter Secretor

Blood

Fig. 1. Illustrates with examples four mechanisms by which the BBB is central to neuroimmune phenomena. The barrier mechanism prevents the unregulated exchange of
substances between the brain and the blood; IgG is given as an example. The responder mechanism means that BBB functions are altered by neuroimmune events; the
downregulation of the brain-to-blood transporter p-glycoprotein is illustrated. The transporter mechanism allows certain molecules, most notably some cytokines/
chemokines, to cross the BBB because of the presence of specific saturable transporter; the transport of TNF is given as an example. The secretor mechanism endows barrier
cells with the ability to release immunoactive and immunomodulatory substances in response to neuroimmune events. The release of IL-6 in response to LPS is illustrated.

Please cite this article in press as: Banks, W.A. The blood–brain barrier in neuroimmunology: Tales of separation and assimilation. Brain Behav. Immun.
(2014), http://dx.doi.org/10.1016/j.bbi.2014.08.007
 YBRBI 2433 No. of Pages 8, Model 5G
30 August 2014

W.A. Banks / Brain, Behavior, and Immunity xxx (2014) xxx–xxx 3

151 doses did disrupt the BBB and were associated with increased induced or mediated by LPS, cytokines, prostaglandins, and nitric 217
152 hypothalamic damage (Allen, 1965). Allen also reviewed a series oxide. 218
153 of studies from the 1950s that failed to show any radioactively
154 tagged pyrogen entering the brain. Based on these findings, she
155 concluded that induction of fever by low doses of bacterial pyrogen 3. BBB responses to immunoactive molecules 219
156 did not require BBB disruption, but acted indirectly on the brain
157 through release of an endogenous substance, whereas high doses The BBB is able to respond to LPS because of the presence of 220
158 of bacterial endotoxin could disrupt the BBB and damage brain TLR4 and other toll-like receptors on the membranes of the cells 221
159 tissues. that constitute the BBB (Nagyoszi et al., 2010). Likewise, these cells 222
160 These early studies largely depended on the use of dyes, such as have receptors for cytokines, chemokines, and other immune- 223
161 Evans blue. These dyes in their free state readily cross the BBB, but related molecules (Cunningham and De Souza, 1993; Pan et al., 224
162 in the blood tightly bind to albumin and so do not cross the intact 2009, 2008; Takao et al., 1992). As a result, the immune system 225
163 BBB. Various problems with use of dyes have been recognized for is able to affect the functions of the BBB beyond that of disruption. 226
164 decades (Broman, 1950) and some of these problems, such as Perhaps the most widely studied effect of LPS on the BBB after 227
165 reduced serum albumin levels or intravascular coagulation, can disruption is that of alterations in BBB transport systems (Fig. 1). 228
166 be especially common when studying immune phenomena. Many substances with alterations in their brain/blood or CSF/blood 229
167 Tibbling et al. (1977) formalized the use of the CSF/serum ratio ratios that had been assumed to be caused by BBB disruption are 230
168 for albumin and for IgG, using them to identify BBB disruption more likely explained by alterations in the transporters for those 231
169 and excess immune activity in the CNS, respectively. Unfortunately substances. Immune-related alterations in BBB permeability not 232
170 for both the BBB and neuroimmune fields, the interpretation of the caused by BBB disruption occurs for cisplatin, TNF, insulin, HIV-1 233
171 IgG CSF/serum ratio has often been muddled, sometimes used as and its viral coat protein gp120, insulin, leukemia inhibitory factor, 234
172 an indicator of BBB disruption and sometimes as a measure of amyloid beta peptide, leptin, pituitary adenylate cyclase activating 235
173 intracranial antibody production. The use of CSF to evaluate CNS polypeptide, and immune cells (Banks et al., 2008, 1999, 2012; 236
174 events, a highly favored approach for decades, has become increas- Dohgu and Banks, 2008; Erickson et al., 2012; Minami et al., 237
175 ingly rare since the 1980s as the use of radioactively labeled sub- 1998b; Nonaka et al., 2004, 2005; Pan et al., 1996, 2008; 238
176 stances and imaging modalities have increasing allowed Persidsky et al., 1997; Xaio et al., 2001). The mechanisms for these 239
177 investigators to focus on brain tissue. This means that there has alterations vary, but basically fall into two broad categories. One 240
178 been a subtle shift from the study of phenomena that relate more category is that of a modulation of a transporter that is responsible 241
179 to the vascular BBB and perhaps less so to the blood–CSF barrier. for the rate at which the substance crosses. Another category is 242
180 This is an important bias to recognize as the BBB and blood–CSF that of induction of a pathological response, such as adsorptive 243
181 barriers likely play differing, but integrated, roles in transcytosis. 244
182 neuroimmunology. The effect of immune modulation of BBB transporters can take 245
183 As a shift from the study of CSF to the study of brain tissue several forms. In the cases of insulin, leukemia inhibitory factor, 246
184 occurred in the BBB field, a shift from the study of exogenous and TNF, uptake is dramatically enhanced because blood-to-brain 247
185 and endogenous pyrogens to the study of lipopolysaccharide transporters specific or selective for these substances have 248
186 (LPS) and cytokines occurred in the neuroimmunology field. Many increased activity. The enhanced transport is independent of any 249
187 of these cytokines, including tumor necrosis factor-alpha (TNF), disruption that occurs. In the cases of amyloid beta peptide, verap- 250
188 were quickly proposed as having the potential to disrupt the BBB amil, most anti-seizures drugs, and many anti-virals, uptake is 251
189 (Deli et al., 1995; Rosenberg et al., 1995; Sharief et al., 1993). enhanced because brain-to-blood transporters for these substances 252
190 LPS-induced BBB disruption can be blocked by pretreatment with have decreased activity (Deane et al., 2004; Ronaldson et al., 2008). 253
191 indomethacin and nitric oxide inhibitors, suggesting a dependence One transporter whose response to TNF and LPS has been well 254
192 on prostaglandins and nitric oxide (Candelario-Jalil et al., 2007; De worked out is that of P-gp. LPS acts directly on the brain endothe- 255
193 Vries et al., 1996; Minami et al., 1998a). lial cell (BEC) and all the components of the reaction are mediated 256
194 Immune-mediated disruption of the BBB is no longer simply a through pathways contained within the BEC (Hartz et al., 2006; Yu 257
195 concern for sepsis and CNS infections. An increasing number of dis- et al., 2008, 2007a,b). 258
196 eases are associated with BBB disruption; inflammatory or neuro- This ability of neuroimmune activation to alter BBB transporters 259
197 inflammatory events likely underlie those disruptions. Classic has many clinical consequences. For example, seizure activity 260
198 examples of BBB disruption include multiple sclerosis, neuro- tends to enhance P-gp activity; thus, the retention by brain of 261
199 trauma, vascular dementias, and stroke and all these now are P-gp substrates is reduced (Rizzi et al., 2002). Because most anti- 262
200 believed to involve neuroinflammatory events. More recent addi- seizure medications are P-gp substrates, this means that seizures 263
201 tions to the list of proinflammatory conditions with BBB disruption themselves act through P-gp to induce resistance to medication. 264
202 or inflammation-induced BBB disruption include diabetes mellitus This may explain in part why status epilepticus is especially resis- 265
203 (Huber, 2008; Starr et al., 2003), chronic pain (Huber et al., 2002; tant to treatment with anti-seizure medications. 266
204 Willis and Davis, 2008), and some neurodegenerative diseases Pathological reactions also likely have a diverse nature. HIV-1 267
205 (Erickson and Banks, 2013). Likewise, LPS-induced disruption of can enter the CNS as infected immune cells cross the BBB or can 268
206 the BBB can no longer be considered simply a model for disease, cross the BBB as free viruses. Inflammatory events, including expo- 269
207 as many conditions other than sepsis are associated with the pres- sure to LPS, enhance entry of HIV-1 into the nervous system by 270
208 ence of LPS in the circulation, including AIDS gastroenteropathy, both of these processes (Dohgu and Banks, 2008; Persidsky et al., 271
209 periodontal disease, peritoneal dialysis, depression, working in 1997). The cellular mechanisms that these two processes use are 272
210 the ‘‘bioprotein’’ industry, sporadic amyotrophic lateral sclerosis, different, with immune cell trafficking being dependent on JNK 273
211 long-distance running, and the ingestion of a high fat meal and STAT1 signaling (Chaudhiri et al., 2008; Ramirez et al., 2010). 274
212 (Amar et al., 2008; Ancuta et al., 2008; Brenchley et al., 2006; In comparison, LPS enhances free virus transport through a process 275
213 Ghoshal et al., 2009; Maes et al., 2008; Ng et al., 2008; Sikkeland dependent on MAPK (Dohgu and Banks, 2008). 276
214 et al., 2008; Szeto et al., 2008; Zhang et al., 2009). Taken together, An important characteristic of the response of BECs to immune 277
215 these suggest that immune-induced BBB disruption is more wide- stimuli is their polarized response. In other words, a substance may 278
216 spread than recently appreciated and that such disruption can be produce an effect when acting on the blood side of the BBB but not 279

Please cite this article in press as: Banks, W.A. The blood–brain barrier in neuroimmunology: Tales of separation and assimilation. Brain Behav. Immun.
(2014), http://dx.doi.org/10.1016/j.bbi.2014.08.007
 YBRBI 2433 No. of Pages 8, Model 5G
30 August 2014

4 W.A. Banks / Brain, Behavior, and Immunity xxx (2014) xxx–xxx

280 on its brain side. For example, GM-CSF and IL-6 enhance HIV-1 and TNF entering the blood from the brain were sufficient to 340
281 transport across monolayers of BECs when applied to the luminal increase their blood levels (Chen et al., 1997, 2000; Chen and 341
282 side (AKA apical, blood side) but not when applied to the abluminal Reichlin, 1998; Reichlin et al., 2000; Romero et al., 1996). In these 342
283 side (AKA basolateral, brain side) of the cells (Dohgu et al., 2011). cases, this passage was not saturable, but occurred with the reab- 343
284 This can add a type of directionality to BBB-mediated neuroim- sorption of CSF, termed bulk flow. Such passage occurs for any sub- 344
285 mune responses. stance that is present in the CSF, but the small volume of total body 345
286 The effects of LPS on free HIV-1 transport and on P-gp function distribution typical of many cytokines favors the ability of this 346
287 are mediated at the endothelial cell. As discussed below, many mechanism to produce elevated levels in the blood. 347
288 effects of LPS on the BBB require crosstalk between the BEC and Only one cytokine has so far been shown to have a saturable 348
289 other cell types in the periphery or the CNS. component to its brain-to-blood efflux and that is IL-2 (Banks 349
et al., 2004). The combination of lack of a saturable blood-to-brain 350
transporter, robust enzymatic degradation, and the saturable efflux 351
290 4. BBB transport of immunoactive molecules
system means that the free (unbound) component of blood–borne 352
IL-2 enters the brain less well than even albumin. Besides helping 353
291 4.1. Blood to brain transport
to exclude their substrates from the brain, efflux systems have 354
been shown to play other roles as well. Efflux, along with the rates 355
292 A major function of the BBB is to prevent the leakage of sub-
of synthesis and degradation, can be a major factor in determining 356
293 stances from the blood into the CNS, thus protecting the brain from
the CNS level of a substance, as illustrated by amyloid beta peptide 357
294 a myriad of endogenous circulating substances that can have
(Zlokovic, 2005). Impairment of its efflux, both its saturable and 358
295 neurotoxic effects. Cytokines and chemokines are among the
bulk flow components, has been shown to occur and likely contrib- 359
296 substances whose unregulated access to the CNS is prevented by
utes to the amyloid burden of brain seen in Alzheimer’s disease. 360
297 the BBB. However, many cytokines and chemokines are trans-
Impaired efflux of amyloid beta peptide can be induced by inflam- 361
298 ported across the BBB in the blood-to-brain direction (Fig. 1),
mation, including treatment with LPS (Erickson et al., 2012; Jaeger 362
299 including the IL-1’s, IL-6, IL-15, TNF, CCL2, CCL11, cytokine-induced
et al., 2009). This suggests that neuroinflammation could play a 363
300 neutrophil chemoattractant-1, leukemia inhibitory factor-1, epi-
major role in the initiating events of Alzheimer’s disease. 364
301 dermal growth factor, and various fibroblast growth factors
Efflux is also a mechanism by which the brain can make sub- 365
302 (Banks et al., 1989, 1994, 1991; Cuevas et al., 1996; Erickson
stantial contributions to the blood levels of a substance. Efflux of 366
303 et al., 2014; Ge et al., 2008; Gutierrez et al., 1993; Hsuchou et al.,
corticotrophin, for example, elevates levels of this peptide to suffi- 367
304 2013; Pan and Kastin, 1999, 2001; Pan et al., 2000, 2009; Wagner
cient levels that they affect the release of beta-endorphin from the 368
305 et al., 1999). The difference between entry into the CNS of a
spleen (Martins et al., 1997). 369
306 neuro-active substance via unregulated leakage vs. a regulated
307 transporter is key to the general theme that runs throughout the
308 study of the BBB: the former is associated with BBB dysfunction
5. BBB secretion of immunoactive molecules 370
309 and can result in neurotoxicity; the latter is controlled ultimately
310 by the CNS, adjusting the properties of the BBB transporter to the
The BECs that comprise the vascular BBB and the epithelial cells 371
311 needs of the brain.
that comprise the blood–CSF barrier are capable of secreting 372
312 In the case of TNF, the transporter protein is believed to be the
immunoactive substances, including cytokines, chemokines, nitric 373
313 TNF receptors because receptor knockout mice do not transport
oxide and prostaglandins (Fig. 1). The secretion of these substances 374
314 TNF (Pan and Kastin, 2002). In the case of epidermal growth factor
can be either constitutive or inducible. The most compelling stud- 375
315 (Pan and Kastin, 1999) and IL-1 (Banks et al., 1991), the evidence
ies showing that barrier cells are secreting these substances are 376
316 suggests that the receptors are not part of the transport process,
those of in vitro cultures that contain the barrier cell and only 377
317 but as yet to be identified proteins are acting as the transporters.
the barrier cell. The presence of mRNA for the proteins or of the 378
318 Blood-derived cytokines crossing the BBB can act directly on
required enzymes for the synthesis of nitric oxide and the prosta- 379
319 their CNS receptors to alter function. For example, human IL-1
glandins are important supporting evidence. Such studies have 380
320 alpha injected intravenously into the mouse caused a cognitive
been conducted with isolated microvessels (which usually also 381
321 impairment that could be blocked if small amounts of antibody
contain pericytes), immortalized BECs, and primary cultures of 382
322 specific to human IL-1 alpha was injected into the posterior divi-
BECs. Culture of BBB cells has many difficulties including slow 383
323 sion of the septum, an area that avidly transports IL-1 alpha into
growth of primary BECs and dedifferentiation of immortalized cell 384
324 the brain (Banks et al., 2001). As the only source of human IL-1
lines. Not surprisingly, then, experiments often differ in the details 385
325 alpha in the mouse was the blood, this illustrates that blood–borne
of results, such as which cytokines are specifically secreted and to 386
326 IL-1 alpha crossed the BBB and did so in amounts sufficient to
what degree. Nevertheless, most studies agree that IL-6 is reliably 387
327 affect cognition. Blood-derived cytokines crossing the BBB can also
and robustly secreted, both constitutively and in response to 388
328 act indirectly by inducing release of more cytokine from brain
immune stimuli. The IL-1’s, IL8, IL-10, G-CSF, GM-CSF, interferon 389
329 stores. This mechanism has been demonstrated by Qin et al.
inducible protein-10, keratinocyte chemoattractant, TNF, MCP-1, 390
330 (2007), in which TNF injected into the blood stream crosses the
RANTES, and endothelin have also been shown to be secreted 391
331 BBB, inducing further release of TNF from brain stores, with this
(Didier et al., 2002; Fabry et al., 1993; Macvilay and Fabry, 1997; 392
332 centrally-derived TNF producing most of the CNS effect. Another
Mandi et al., 1998; McGuire et al., 2003; Simpson et al., 1998; 393
333 example of a cytokine crossing the BBB to induce a CNS effect is
Spranger et al., 2006; Vadeboncoeur et al., 2003; Verma et al., 394
334 that of basic fibroblast growth factor which crosses the BBB and
2006). These studies show that cytokine secretion can be modu- 395
335 accumulates in the hippocampus in amounts sufficient to protect
lated by LPS, HIV-1 and its surface glycoprotein, and adiponectin. 396
336 against ischemia–reperfusion injury (Cuevas et al., 1998).
The LPS-enhanced transport of HIV-1 as free virus depends on 397
the ability of BECs to secrete cytokines. LPS stimulates the BECs 398
337 4.2. Brain-to-blood transport to increase their release of GM-CSF and IL-6 (Dohgu et al., 2011). 399
These cytokines then act at their luminal receptors on the BEC, 400
338 Brain-to-blood passage of several cytokines has been noted. In a stimulating transcytosis mediated by the mannose-6-phosphate 401
339 series of papers, Reichlin showed that amounts of IL-1 beta, IL-6, receptor through a MAPK-dependent pathway (Dohgu et al., 2012). 402

Please cite this article in press as: Banks, W.A. The blood–brain barrier in neuroimmunology: Tales of separation and assimilation. Brain Behav. Immun.
(2014), http://dx.doi.org/10.1016/j.bbi.2014.08.007
 YBRBI 2433 No. of Pages 8, Model 5G
30 August 2014

W.A. Banks / Brain, Behavior, and Immunity xxx (2014) xxx–xxx 5

403 Many cells are capable of secreting substances into their envi- in depth, such as the effect of LPS on free HIV-1 transport across 469
404 ronment. But BECs are unique in that they face into two environ- the BBB, largely involved only the barrier cells. But the concept 470
405 ments and do so simultaneously: the periphery and the CNS. of the neurovascular unit with its cross talk among the barrier 471
406 BECs, for example, can secrete cytokines from their luminal side cells, other CNS cells, and other peripheral tissues (with blood as 472
407 into the blood stream or from their abluminal side into the intersti- its proxy) implies that many neuroimmune functions will involve 473
408 tial fluid of the brain. More interestingly, BECs can respond to an interactions of the barrier cells with other cells. Many of these 474
409 immune stimulus at one of its sides by secreting an immunoactive cell–cell interactions are likely to be exceedingly complex and 475
410 substance from the other. For example, LPS applied to the ablumi- tedious to work out, but there are some interesting examples 476
411 nal surface of monolayers of BECs greatly stimulates the release of already in the literature. 477
412 IL-6 from the luminal side (Verma et al., 2006). The directionality A few of these examples have already been discussed. Qin et al. 478
413 can go the other way as well as illustrated by the ability of luminal have shown that the BBB transports TNF that then acts on microg- 479
414 adiponectin to decrease the abluminal release of IL-6 (Spranger lia to release more TNF that is then toxic to dopaminergic cells in 480
415 et al., 2006) or by the ability of gp120 at the luminal surface to the substantial nigra (Qin et al., 2007). The LPS-enhanced transport 481
416 increase endothelin secretion from the abluminal surface (Didier of free HIV across BBB monolayers does not require other cells to 482
417 et al., 2002). This endows the BBB with the ability to translate an participate, but it is further enhanced in the presence of pericytes 483
418 immune event in the CNS into a peripheral one or, conversely, to (Dohgu and Banks, 2013). Since LPS does not cross the BBB and 484
419 translate a peripheral event into a CNS one. The ability to translate pericytes are on the abluminal side of the BBB, this action of 485
420 neuroimmune input from one side to the other likely forms a fun- pericytes likely involves an LPS-induced release of factors from 486
421 damentally important neuroimmune axis. endothelial cells with those factors then inducing a release of sub- 487
422 Integration of the BBB with Other Components of the Neurovas- stances from the pericyte, and these pericyte-derived substances 488
423 cular Unit, the Neuroimmune Axes, and the Peripheral Tissues. then interacting with the BEC to enhance HIV-1 transport. Consis- 489
424 The BBB is dynamic, changing throughout the life cycle and tent with this, we found that pericytes that were exposed to the 490
425 with disease. Under physiologic conditions, it is slave to the brain, abluminal fluids of monolayers of BEC that had LPS added to their 491
426 adapting to the changing needs of the brain during development, luminal chambers secreted a unique pattern of cytokines, a pattern 492
427 aging, and other life events. Under pathologic conditions, it may that differed from that of the added abluminal fluid and from that 493
428 be the target of disease and it may give rise to or promote disease secreted by LPS-exposed pericytes (Dohgu and Banks, 2013). 494
429 when its adjustments do not match the needs of the brain or ade- One approach to elucidating complex immune interactions is to 495
430 quately adapt to events in the periphery. This responsive dyna- determine whether monocultures of BECs recapitulate the in vivo 496
431 mism has long been appreciated, and it has been increasingly BBB response or whether recapitulation requires co-culture with 497
432 understood that its roots lie in the crosstalk with the cells that con- other cells. For example, functional expression of the blood-to- 498
433 stitute the physical BBB (e.g., the endothelial cells of the vascular brain insulin transporter does not occur in monocultures of BECs, 499
434 BBB and the epithelial cells of the blood–CSF barrier), the other but requires co-culture with pericytes (Nakaoke et al., 2007). 500
435 cells of the brain (e.g., neurons, microglia, astrocytes, pericytes, Effects of LPS and nitric oxide on insulin transport are a further 501
436 immune cells within the CNS), and the cells in the periphery illustration of the requirement for crosstalk between BECs and 502
437 (e.g., circulating immune cells, peripheral tissues and organs via other cells. In vivo studies show that LPS enhances insulin trans- 503
438 their hormonal secretions) (Abbott et al., 2006; Greenwood et al., port across the BBB (Xaio et al., 2001). However, LPS has no effect 504
439 2011; Neuwelt et al., 2008; Willis and Davis, 2008). Although the on the rate of insulin transport across monolayer monocultures of 505
440 responsive dynamism has been long appreciated conceptually, BECs (Banks et al., 2008). This suggests that LPS, although it acts 506
441 the mechanisms that underlie it have been less clear. directly on BECs to induce cytokine release and to affect free HIV 507
442 The previous sections go a long way towards outlining mecha- transport, is in this case not acting directly on the BEC to affect 508
443 nisms for this crosstalk (Fig. 2). To recapitulate: The ability of the insulin transport. Instead, LPS must be acting on some other cell 509
444 BEC to respond to and release immunoactive substances provides type to induce a substance that then acts on the BEC to alter insulin 510
445 a mechanism for intercellular crosstalk. The barrier function of transport. To test this, serum from LPS-treated mice was incubated 511
446 these cells means that they are polarized, permitting interactions with the BECs, but this did not alter insulin transport. This suggests 512
447 with the cells of the CNS and simultaneous, independent interac- either that the substance is not circulating but of central origin or 513
448 tions with the peripheral tissues while preventing the cells of the that it is a blood–borne substance that is not stable, such as nitric 514
449 CNS and peripheral cells from directly interacting with one oxide. When cervical lymphocytes were placed in the luminal 515
450 another. The barrier function coupled with BBB transporters means chamber of the BEC monolayer culture and LPS added to it, the sat- 516
451 that substances such as cytokines either do not enter the CNS or do urable transport of insulin across the monolayers was enhanced. 517
452 so under regulated conditions that are ultimately under the control Thus, the findings show that LPS acts through another cell type 518
453 of the CNS. Communication between the CNS cells and the periph- to enhance insulin transport across the BBB and that immune cells 519
454 ery is controlled by neuroimmune axes, two of which have been are one cell type that can fulfill that role. 520
455 elucidated so far: an indirect relay in which the BBB cell receives Nitric oxide effects on insulin transport also demonstrate com- 521
456 input from one cell surface and secretes immunoactive substances plex cellular interactions (Banks et al., 2008). Treatment of mice 522
457 from the other and a direct transport of cytokines and chemokines with a general inhibitor of nitric oxide synthase (NOS) or an inhib- 523
458 across the BBB. A third axis involving the BBB, the trafficking of itor specific for neuronal NOS (nNOS) further enhances the LPS 524
459 immune cells into the CNS, is also a highly regulated process that effect on insulin transport. This suggests that nitric oxide has an 525
460 depends on crosstalk between the immune cells and the cells con- inhibitory effect on insulin transport. However, inhibition of endo- 526
461 stituting the BBB (Engelhardt, 2008; Greenwood et al., 2011). A thelial NOS (eNOS) and inducible NOS (iNOS) inhibits insulin trans- 527
462 study using simple mathematical modeling has shown that port and treating animals with L-arginine, the precursor of nitric 528
463 immune cell distribution between the peripheral tissues and the oxide, stimulates insulin transport. These findings show that nitric 529
464 CNS is variably dependent on which cell type dominates in this oxide can also stimulate insulin transport. Since nitric oxide syn- 530
465 cross talk and that those relations can be readily altered by thesized by any enzyme is the same molecule, it is unlikely that 531
466 immune events and genetics (Banks et al., 2012). both the inhibitory and stimulatory effects of nitric oxide are both 532
467 The above sections have discussed some of the direct actions of mediated directly at the BEC. Rather, it is likely that either the 533
468 LPS and immunoactive substances on the BBB. The cases reviewed inhibitory or stimulatory effect occurs indirectly by acting on 534

Please cite this article in press as: Banks, W.A. The blood–brain barrier in neuroimmunology: Tales of separation and assimilation. Brain Behav. Immun.
(2014), http://dx.doi.org/10.1016/j.bbi.2014.08.007
 YBRBI 2433 No. of Pages 8, Model 5G
30 August 2014

6 W.A. Banks / Brain, Behavior, and Immunity xxx (2014) xxx–xxx

Brain/
Abluminal Side

TNF
Substana
Microglia nigra
Pericyte

3 +
+ 2
Insulin Brain Endothelial
Transporter Cells

1
+ TNF

Free Lymphocyte
HIV-1 LPS
LPS

Blood/Luminal Side

Fig. 2. Integration of the four mechanisms of barrier, responder, transporter, and secretor allows intercellular crosstalk that is important to the adaption of the BBB to the
needs of the CNS, adaption to neuroimmune events, and response to disease. The far right panel shows that transport of TNF across the BBB allows blood–borne TNF to
interact with microglia, inducing the microglia to release more TNF, which can induce apoptosis in dopaminergic cells at the substantia nigra. The middle panel shows that
LPS enhances insulin transport across the BBB by acting indirectly on other cell types, including circulating lymphocytes, to induce them to release substances that then act at
the brain endothelial cell. The final panel shows that pericytes induce the insulin transporter in monolayers of brain endothelial cells. It also illustrates that a complex, three
step cross-talk between brain endothelial cells pericytes can further induce the LPS-enhanced passage of HIV-1 across the BBB: (1) LPS acts on the luminal side of the brain
endothelial cell to (2) induce it to release factors from its abluminal side that act on the pericyte (3) inducing the pericyte to release substances that then act on the brain
endothelial cell to accelerate HIV-1 passage across the BBB.

535 another cell type that then secretes a substance to act on the BBB. immune axis, translating the signals it receives on one of its sides 567
536 To determine whether nitric oxide could act directly on BBB cells, by secreting from its other side. The transport of cytokines across 568
537 in vitro monocultures of BECs were incubated with SNAP, a sub- the BBB creates another type of neuroimmune axis, providing a 569
538 stance that releases nitric oxide. These studies showed SNAP mechanism by which peripheral events can directly influence 570
539 applied to the luminal side of the monolayers inhibited insulin brain. These functions, events, mechanisms, and axes are not static, 571
540 transport. SNAP applied to the abluminal side also inhibited insulin but are modulated by the environment in which the BBB is embed- 572
541 transport, showing that the paradoxic effect of nitric oxide was not ded. In some cases, these modulations are mediated entirely at the 573
542 because it had different effects on the polarized BEC. Thus, nitric BBB, but other modulations require crosstalk between the cells 574
543 oxide generated from nNOS could be acting directly on the BEC that form the BBB and the other cells found in the CNS and periph- 575
544 to inhibit insulin transport, whereas eNOS and iNOS are likely act- ery. Thus, the BBB is a dynamic, interactive, regulatory tissue that 576
545 ing on other cells to induce them to release a substance that stim- is involved both in making the CNS an immunoprivileged organ 577
546 ulates insulin transport across the BBB. and in fostering immune-CNS communications by participating 578
547 The most parsimonious summary of these findings is that LPS in the formation of neuroimmune axes. 579
548 does not act directly on the BBB to alter transport of insulin, but
549 indirectly on cells such as lymphocytes to enhance BBB transport.
550 LPS-induced release of nitric oxide from non-BBB cells has complex Funding source 580

551 effects on insulin transport, with nitric oxide generated from nNOS
552 from non-BECs acting directly on BECs to inhibit insulin transport Supported by VA Merit Review, Friends of Alzheimer’s Research, 581

553 and iNOS- and eNOS-generated nitric oxide acting on non-BECs to NIA R0-1 AG029834, NIDDK R0-1 DK083485, and NINDS NS051134. 582

554 release substances that enhance BBB transport of insulin.

References 583
555 6. Summary
Abbott, N.J., Ronnback, L., Hansson, E., 2006. Astrocyte–endothelial interactions at 584
the blood–brain barrier. Nat. Rev. 7, 41–53. 585
556 The BBB plays various roles in neuroimmunology. By virtue of Allen, I.V., 1965. The effect of bacterial pyrogen on the blood–brain barrier for 586
557 its barrier functions, it separates the CNS from the peripheral com- trypan blue. J. Pathol. Bacteriol. 89, 481–494. 587
Amar, J., Burcelin, R., Ruidavets, J.B., Cani, P.D., Fauvel, J., Alessi, M.C., Chamontin, B., 588
558 ponents of the immune system, making the brain an immunopriv- Ferrieres, J., 2008. Energy intake is associated with endotoxemia in apparently 589
559 ileged organ. It then is involved in a re-establishment of CNS- healthy men. Am. J. Clin. Nutr. 87, 1219–1223. 590
560 immune communications by various mechanisms. Its ability to Ancuta, P., Kamat, A., Kinstman, K.J., Kim, E.Y., Autissier, P., Wurcel, A., Zaman, T., 591
Stone, D., Mefford, M., Morgello, S., Singer, E.J., Wolinsky, S.M., Gabuzda, D., 592
561 respond to and to secrete immunoactive substances provides a 593
2008. Microbial translocation is associated with increased monocyte activation
562 mechanism by which it can be influenced by events in both the and dementia in AIDS patients. PLoS ONE 23, e2516. 594
563 CNS and peripheral tissues; this ability also provides a mechanism Banks, W.A., Dohgu, S., Nakaoke, R., Lynch, J.L., Fleegal-DeMotta, M.A., Erickson, 595
M.A., Vo, T.Q., 2008. Nitric oxide isoenzymes regulate LPS-enhanced insulin 596
564 by which it can influence the CNS and peripheral tissues. The com-
transport across the blood–brain barrier. Endocrinology 149, 1514–1523. 597
565 bination of barrier function with ability to respond to and secrete Banks, W.A., Farr, S.A., La Scola, M.E., Morley, J.E., 2001. Intravenous human 598
566 immunoactive substances allows the BBB to form a unique neuro- interleukin-1 alpha impairs memory processing in mice: dependence on blood– 599

Please cite this article in press as: Banks, W.A. The blood–brain barrier in neuroimmunology: Tales of separation and assimilation. Brain Behav. Immun.
(2014), http://dx.doi.org/10.1016/j.bbi.2014.08.007
 YBRBI 2433 No. of Pages 8, Model 5G
30 August 2014

W.A. Banks / Brain, Behavior, and Immunity xxx (2014) xxx–xxx 7

600 brain barrier transport into posterior division of the septum. J. Pharmacol. Exp. Dohgu, S., Ryerse, J.S., Robinson, S.M., Banks, W.A., 2012. Human immunodeficiency 686
601 Ther. 299, 536–541. virus-1 uses the mannose-6-phosphate receptor to cross the blood–brain 687
602 Banks, W.A., Kastin, A.J., Brennan, J.M., Vallance, K.L., 1999. Adsorptive endocytosis barrier. PLoS ONE 7, e41623. 688
603 of HIV-1gp120 by blood–brain barrier is enhanced by lipopolysaccharide. Exp. Eckman, P.L., King, W.M., Brunson, J.G., 1958. Studies on the blood–brain barrier. I. 689
604 Neurol. 156, 165–171. Effects produced by a single injection of gram-negative endotoxin on the 690
605 Banks, W.A., Kastin, A.J., Durham, D.A., 1989. Bidirectional transport of interleukin-1 permeability of the cerebral vessels. Am. J. Pathol. 34, 631–643. 691
606 alpha across the blood–brain barrier. Brain Res. Bull. 23, 433–437. Ehrlich, P., 1906. Collected studies on immunity. In: Bolduan, C. (Ed.), John Wiley & 692
607 Banks, W.A., Kastin, A.J., Gutierrez, E.G., 1994. Penetration of interleukin-6 across Sons, London, pp. 404–442. 693
608 the blood–brain barrier. Neurosci. Lett. 179, 53–56. Engelhardt, B., 2008. The blood–central nervous system barriers actively control 694
609 Banks, W.A., Niehoff, M.L., Ponzio, N.M., Erickson, M.A., Zalcman, S.S., 2012. immune cell entry into the central nervous system. Curr. Pharm. Des. 14, 1555– 695
610 Pharmacokinetics and modeling of immune cell trafficking: quantifying 1565. 696
611 differential influences of target tissues versus lymphocytes in SJL and Erickson, M.A., Banks, W.A., 2013. Blood–brain barrier dysfunction as a cause and 697
612 lipopolysaccharide-treated mice. J. Neuroinflamm. 9, 231. consequence of Alzheimer’s disease. J. Cereb. Blood Flow Metab. 33, 1500–1513. 698
613 Banks, W.A., Niehoff, M.L., Zalcman, S., 2004. Permeability of the mouse blood–brain Erickson, M.A., Hartvigson, P.E., Morofuji, Y., Owen, J.B., Butterfield, D.A., Banks, 699
614 barrier to murine interleukin-2: predominance of a saturable efflux system. W.A., 2012. Lipopolysaccharide impairs amyloid beta efflux from brain: altered 700
615 Brain Behav. Immun. 18, 434–442. vascular sequestration, cerebrospinal fluid reabsorption, peripheral clearance 701
616 Banks, W.A., Ortiz, L., Plotkin, S.R., Kastin, A.J., 1991. Human interleukin (IL) 1alpha, and transporter function at the blood–brain barrier. J. Neuroinflamm. 9, 150. 702
617 murine IL-1alpha and murine IL-1 beta are transported from blood to brain in Erickson, M.A., Morofuji, Y., Owen, J.B., Banks, W.A., 2014. Rapid transport of CCL11 703
618 the mouse by a shared saturable mechanism. J. Pharmacol. Exp. Ther. 259, across the blood–brain barrier: regional variation and importance of blood cells. 704
619 988–996. J. Pharmacol. Exp. Ther. 349, 497–507. 705
620 Biedl, A., Kraus, R., 1898. Uber einer bisher unbekannte toxische Wirking der Fabry, Z., Fitzsimmons, K.M., Herlein, J.A., Moninger, T.O., Dobbs, M.B., Hart, M.N., 706
621 Gallensauren auf das Zentralnervensystem. Zentralblatt inn. Med. 19, 1185– 1993. Production of the cytokines interleukin 1 and 6 by murine brain 707
622 1200. microvessel endothelium and smooth muscle pericytes. J. Neuroimmunol. 47, 708
623 Brenchley, J.M., Price, D.A., Schacker, T.W., Asher, T.E., Silvestri, G., Rao, S., Kazzaz, Z., 23–34. 709
624 Bornstein, E., Lambotte, O., Altmann, D., Blazar, B.R., Rodriguez, B., Teixeira- Ge, S., Serwanski, D.R., Kuziel, W.A., Pachter, J.S., 2008. Transcellular transport of 710
625 Johnson, L., Landay, A., Martin, J.N., Hecht, F.M., Picker, L.J., Lederman, M.M., CCL2 across brain microvascular endothelial cells. J. Neurochem. 104, 1219– 711
626 Deeks, S.G., Douek, D.C., 2006. Microbial translocation is a cause of systemic 1232. 712
627 immune activation in chronic HIV injection. Nat. Med. 12, 1365–1371. Ghoshal, S., Witta, J., Zhong, J., de Villiers, W., Eckhardt, E., 2009. Chylomicrons 713
628 Brightman, M.W., Reese, T.S., 1969. Junctions between intimately apposed cell promote intestinal absorption of lipopolysaccharides. J. Lipid Res. 50, 90–97. 714
629 membranes in the vertebrate brain. J. Cell Biol. 40, 648–677. Goldmann, E.E., 1913. Vitalfarbung am Zentral-nervensystem. Abh. Preuss. Akad. 715
630 Broman, T., 1950. Supravital analysis of disorders in the cerebrovascular Wiss. Phys.-Math KL I, 1–60. 716
631 permeability. Acta Psychiatr. 25, 19–31. Greenwood, J., Heasman, S.J., Alvarez, J.I., Pratt, A., Lyck, R., Engelhardt, B., 2011. 717
632 Candelario-Jalil, E., Taheri, S., Yang, Y., Sood, R., Grossetete, M., Estrada, E.Y., Fiebich, Review: leukocyte–endothelial cell crosstalk at the blood–brain barrier: a 718
633 B.L., Rosenberg, G.A., 2007. Cyclooxygenase inhibition limits blood–brain prerequisite for successful immune cell entry to the brain. Neuropathol. Appl. 719
634 barrier disruption following intracerebral injection of tumor necrosis factor-a Neurobiol. 37, 24–39. 720
635 in the rat. J. Pharmacol. Exp. Ther. 323, 488–498. Gutierrez, E.G., Banks, W.A., Kastin, A.J., 1993. Murine tumor necrosis factor alpha is 721
636 Chaudhiri, A., Yang, B., Gendelman, H.E., Persidsky, Y., Kanmogne, G.D., 2008. STAT1 transported from blood to brain in the mouse. J. Neuroimmunol. 47, 169–176. 722
637 signaling modulates HIV-1-induced inflammatory responses and leukocyte Hartz, A.M.S., Bauer, B., Fricker, G., Miller, D.S., 2006. Rapid modulation of P- 723
638 transmigration across the blood–brain barrier. Blood 111, 2062–2072. glycoprotein-mediated transport at the blood–brain barrier by tumor necrosis 724
639 Chen, G., Castro, W.L., Chow, H.H., Reichlin, S., 1997. Clearance of 125 I-labelled factor-alpha and lipopolysaccharide. Mol. Pharmacol. 69, 462–470. 725
640 interleukin-6 from brain into blood following intracerebroventricular injection Hsuchou, H., Pan, W., Kastin, A.J., 2013. Fibroblast growth factor 19 entry into brain. 726
641 in rats. Endocrinology 138, 4830–4836. Fluids Barriers CNS 10, 32. 727
642 Chen, G., McCuskey, R.S., Reichlin, S., 2000. Blood interleukin-6 and tumor necrosis Huber, J.D., 2008. Diabetes, cognitive function, and the blood–brain barrier. Curr. 728
643 factor-alpha elevation after intracerebroventricular injection of Escherichia Pharm. Des. 14, 1594–1600. 729
644 coli endotoxin in the rat is determined by two opposing factors: peripheral Huber, J.D., Hau, V.S., Borg, L., Campos, C.R., Egleton, R.D., Davis, T.P., 2002. Blood– 730
645 induction by LPS transferred from brain to blood and inhibition of peripheral brain barrier tight junctions are altered during a 72-h exposure to lambda- 731
646 response by a brain-mediated mechanism. NeuroImmunoModulation 8, 59– carrageenan-induced inflammatory pain. Am. J. Physiol. 283, H1531–H1537. 732
647 69. Jaeger, J.B., Dohgu, S., Lynch, J.L., Fleegal-DeMotta, M.A., Banks, W.A., 2009. Effects of 733
648 Chen, G., Reichlin, S., 1998. Clearance of [125 I]-tumor necrosis factor-à from the lipopolysaccharide on the blood–brain barrier transport of amyloid beta 734
649 brain into the blood after intracerebroventricular injection into rats. protein: a mechanism for inflammation in the progression of Alzheimer’s 735
650 NeuroImmunoModulation 5, 261–269. disease. Brain Behav. Immun. 23, 507–517. 736
651 Cuevas, P., Carceller, F., Munoz-Willery, I., Gimenez-Gallego, G., 1998. Intravenous Macvilay, S., Fabry, Z., 1997. TGF cytokine production by murine brain microvessel 737
652 fibroblast growth factor penetrates the blood–brain barrier and protects endothelial cells. Neural Notes 3, 21–23. 738
653 hippocampal neurons against ischemia–reperfusion injury. Surg. Neurol. 49, Maes, M., Kubera, M., Leunis, J.C., 2008. The gut–brain barrier in major depression: 739
654 77–83. intestinal mucosal dysfunction with an increased translocation of LPS from 740
655 Cuevas, P., Fermamdez-Ayerdi, A., Carceller, F., Colin, S., Mascarelli, F., Munoz- gram negative enterobacteria (leaky gut) plays a role in the inflammatory 741
656 Willery, I., Gimenez-Gallego, G., 1996. Central nervous system distribution of pathophysiology of depression. Neuro Endocrinol. Lett. 29, 117–124. 742
657 fibroblast growth factor injected into the blood stream. Neurol. Res. 18, Mandi, Y., Ocsovszki, I., Szabo, D., Nagy, Z., Nelson, J., Molnar, J., 1998. Nitric oxide 743
658 267–272. production and MDR expression by human brain endothelial cells. Anticancer 744
659 Cunningham Jr., E.T., De Souza, E.B., 1993. Interleukin 1 receptors in the brain and Res. 18, 3049–3052. 745
660 endocrine tissues. Immunol. Today 14, 171–176. Martins, J.M., Banks, W.A., Kastin, A.J., 1997. Transport of CRH from mouse brain 746
661 De Vries, H.E., Blom-Roosemalen, M.C.M., De Boer, A.G., van Berkel, T.J., Breimer, directly affects peripheral production of beta-endorphin by the spleen. Am. J. 747
662 D.D., Kuiper, J., 1996. Effect of endotoxin on permeability of bovine cerebral Physiol. 273, E1083–E1089. 748
663 endothelial cell layers in vitro. J. Pharmacol. Exp. Ther. 277, 1418–1423. McGuire, T.R., Trickler, W.J., Hock, L., Vrana, A., Hoie, E.B., Miller, D.W., 2003. Release 749
664 Deane, R., Wu, Z., Sagare, A., Davis, J., Du Yan, S., Hamm, K., Xu, F., Parisi, M., LaRue, of prostaglandin E-2 in bovine brain endothelial cells after exposure to three 750
665 B., Hu, H.W., Spijkers, P., Guo, H., Song, X., Lenting, P.J., Van Nostrand, W.E., unique forms of the antifungal drug amphotericin-B: role of COX-2 in 751
666 Zlokovic, B., 2004. LRP/amyloid beta–peptide interaction mediates differential amphotericin-B induced fever. Life Sci. 72, 2581–2590. 752
667 brain efflux of Abeta isoforms. Neuron 43, 333–344. Minami, T., Okazaki, J., Kawabata, A., Kawaki, H., Okazaki, Y., Tohno, Y., 1998a. Roles 753
668 Deli, M.A., Descamps, L., Dehouck, M.P., Cecchelli, R., Joo, F., Abraham, C.S., Torpier, of nitric oxide and prostaglandins in the increased permeability of the blood– 754
669 G., 1995. Exposure of tumor necrosis factor alpha to luminal membrane of brain barrier caused by lipopolysaccharide. Environ. Toxicol. Pharmacol. 5, 755
670 bovine brain capillary endothelial cells cocultured with astrocytes induces a 35–41. 756
671 delayed increase of permeability and cytoplasmic stress fiber formation of actin. Minami, T., Okazaki, J., Kawabata, A., Kuroda, R., Okazaki, Y., 1998b. Penetration of 757
672 J. Neurosci. Res. 41, 717–726. cisplatin into mouse brain by lipopolysaccharide. Toxicology 130, 107–113. 758
673 Didier, N., Banks, W.A., Creminon, C., Dereuddre-Bosquet, N., Mabondzo, A., 2002. Nagyoszi, P., Wilhelm, I., Farkas, A.E., Fazakas, C., Dung, N.T., Hasko, J., Krizbai, I.A., 759
674 HIV-1-induced production of endothelin-1 in an in vitro model of the human 2010. Expression and regulation of toll-like receptors in cerebral endothelial 760
675 blood–brain barrier. NeuroReport 13, 1179–1183. cells. Neurochem. Int. 57, 556–564. 761
676 Dohgu, S., Banks, W.A., 2008. Lipopolysaccharide-enhanced transcellular transport Nakaoke, R., Verma, S., Niwa, M., Dohgu, S., Banks, W.A., 2007. Glucose-regulated 762
677 of HIV-1 across the blood–brain barrier is mediated by the p38 mitogen- blood–brain barrier transport of insulin: pericyte–astrocyte–endothelial cell 763
678 activated protein kinase pathway. Exp. Neurol. 210, 740–749. cross talk. Int. J. Neuroprot. Neuroregener. 3, 195–200. 764
679 Dohgu, S., Banks, W.A., 2013. Brain pericytes increase the lipopolysaccharide- Neuwelt, E., Abbott, N.J., Abrey, L., Banks, W.A., Blakley, B., Davis, T., Engelhardt, B., 765
680 enhanced transcytosis of HIV-1 free virus across the in vitro blood–brain Grammas, P., Nedergaard, M., Nutt, J., Pardgridge, W., Rosenberg, G.A., Smith, Q., 766
681 barrier: evidence for cytokine-mediated pericyte–endothelial cell cross talk. Drewes, L.R., 2008. Strategies to advance translational research into brain 767
682 Fluids Barriers CNS 10, 23. barriers. Lancet Neurol. 7, 84–96. 768
683 Dohgu, S., Fleegal-DeMotta, M.A., Banks, W.A., 2011. Lipopolysaccharide-enhanced Ng, Q.Y., Lee, K.W., Byrne, C., Ho, T.F., Lim, C.L., 2008. Plasma endotoxin and immune 769
684 transcellular transport of HIV-1 across the blood–brain barrier is mediated by responses during a 21-km road race under a warm and humid environment. 770
685 luminal microvessel IL-6 and GM-CSF. J. Neuroinflamm. 8, 167. Ann. Acad. Med. Singapore 37, 307–314. 771

Please cite this article in press as: Banks, W.A. The blood–brain barrier in neuroimmunology: Tales of separation and assimilation. Brain Behav. Immun.
(2014), http://dx.doi.org/10.1016/j.bbi.2014.08.007
 YBRBI 2433 No. of Pages 8, Model 5G
30 August 2014

8 W.A. Banks / Brain, Behavior, and Immunity xxx (2014) xxx–xxx

772 Nonaka, N., Hileman, S.M., Shioda, S., Vo, P., Banks, W.A., 2004. Effects of Sikkeland, L.I., Skogstad, M., Ovstebo, R., Brusletto, B., Haug, K.B., Kongerud, J., 827
773 lipopolysaccharide on leptin transport across the blood–brain barrier. Brain Eduard, W., Kierulf, P., 2008. Circulating lipopolysaccharides in the blood from 828
774 Res. 1016, 58–65. ‘‘bioprotein’’ production workers. Occup. Environ. Med. 65, 211–214. 829
775 Nonaka, N., Shioda, S., Banks, W.A., 2005. Effect of lipopolysaccharide on the Simpson, J.E., Newcombe, J., Cuzner, M.L., Woodrofe, M.N., 1998. Expression of 830
776 transport of pituitary adenylate cyclase activating polypeptide across the monocyte chemoattractant protein-1 and other beta-chemokines by resident 831
777 blood–brain barrier. Exp. Neurol. 191, 137–144. glia and inflammatory cells in multiple sclerosis. J. Neuroimmunol. 84, 238–249. 832
778 Pan, W., Banks, W.A., Kennedy, M.K., Gutierrez, E.G., Kastin, A.J., 1996. Differential Skoog, T., 1937. On the vital staining of the central nervous system. Acta 833
779 permeability of the BBB in acute EAE: enhanced transport of TNF-alpha. Am. J. Otolaryngol. 25, 365–378. 834
780 Physiol. 271, E636–E642. Spranger, J., Verma, S., Gohring, I., Bobbert, T., Seifert, J., Sindler, A.L., Pfeiffer, A., 835
781 Pan, W., Kastin, A.J., 1999. Entry of EGF into brain is rapid and saturable. Peptides Hileman, S.M., Tschop, M., Banks, W.A., 2006. Adiponectin does not cross the 836
782 20, 1091–1098. blood–brain barrier, but modifies cytokine expression of brain endothelial cells. 837
783 Pan, W., Kastin, A.J., 2001. Changing the chemokine gradient: CINC1 crosses the Diabetes 55, 141–147. 838
784 blood–brain barrier. J. Neuroimmunol. 115, 64–70. Starr, J.M., Wardlaw, J., Ferguson, K., MacLullich, A., Deary, I.J., Marshall, I., 2003. 839
785 Pan, W., Kastin, A.J., 2002. TNF alpha transport across the blood–brain barrier is Increased blood–brain barrier permeability in type II diabetes demonstrated by 840
786 abolished in receptor knockout mice. Exp. Neurol. 174, 193–200. gadolinium magnetic resonance imaging. J. Neurol. Neurosurg. Psychiatry 74, 841
787 Pan, W., Kastin, A.J., Brennan, J.M., 2000. Saturable entry of leukemia inhibitory 70–76. 842
788 factor from blood to the central nervous system. J. Neuroimmunol. 106, Szeto, C.C., Kwan, B.C., Chow, K.M., Lai, K.B., Chung, K.Y., Leung, C.B., Li, P.K., 2008. 843
789 172–180. Endotoxemia is related to systemic inflammation and atherosclerosis in 844
790 Pan, W., Yu, C., Hsuchou, H., Khan, R.S., Kastin, A.J., 2009. Cerebral microvascular IL- peritoneal dialysis patients. Clin. J. Am. Soc. Nephrol. 3, 431–436. 845
791 15 is a novel mediator of TNF action. J. Neurochem. 111, 819–827. Takao, T., Culp, S.G., Newton, R.C., De Souza, E.B., 1992. Type I interleukin-1 846
792 Pan, W., Yu, C., Hsuchou, H., Zhang, Y., Kastin, A.J., 2008. Neuroinflammation receptors in the mouse brain–endocrine–immune axis labeled with [125 I] 847
793 facilitates LIF entry into brain: role of TNF. Am. J. Physiol. Cell Physiol. 294, recombinant human interleukin-1 receptor antagonist. J. Neuroimmunol. 41, 848
794 C1436–C1442. 51–60. 849
795 Persidsky, Y., Stins, M., Way, D., Witte, M.H., Weinand, M., Kim, K.S., Bock, P., Tibbling, G., Link, H., Thman, S., 1977. Principles of albumin and IgG analyses in 850
796 Gendelman, H.E., Fiala, M., 1997. A model for monocyte migration through the neurological disorders. I. Establishment of reference values. Scand. J. Clin. Lab. 851
797 blood–brain barrier during HIV-1 encephalitis. J. Immunol. 158, 3499–3510. Invest. 37, 385–390. 852
798 Qin, L., Wu, X., Block, M.L., Liu, Y., Breese, G.R., Hong, J.S., Knapp, D.J., Crews, F.T., Vadeboncoeur, N., Segura, M., Al-Numani, D., Vanier, G., Gottschalk, M., 2003. 853
799 2007. Systemic LPS causes chronic neuroinflammation and progressive Proinflammatory cytokine and chemokine release by human brain 854
800 neurodegeneration. Glia 55, 453–462. microvascular endothelial cells stimulated by Streptococcus suis serotype 2. 855
801 Ramirez, S.H., Fan, S., Dykstra, H., Reichenbach, N., Del Valle, L., Potula, R., Phipps, FEMS Immunol. Med. Microbiol. 35, 49–58. 856
802 R.P., Maggirwar, S.B., Persidsky, Y., 2010. Dyad of CD40/CD40 ligand fosters Verma, S., Nakaoke, R., Dohgu, S., Banks, W.A., 2006. Release of cytokines by brain 857
803 neuroinflammation at the blood–brain barrier and is regulated via JNK endothelial cells: a polarized response to lipopolysaccharide. Brain Behav. 858
804 signaling: implications for HIV-1 encephalitis. J. Neurosci. 30, 9454–9464. Immun. 20, 449–455. 859
805 Reese, T.S., Karnovsky, M.J., 1967. Fine structural localization of a blood–brain Wagner, J.P., Black, I.B., DiCicco-Bloom, E., 1999. Stimulation of neonatal and adult 860
806 barrier to exogenous peroxidase. J. Cell Biol. 34, 207–217. brain neurogenesis by subcutaneous injection of basic fibroblast growth factor. 861
807 Reichlin, S., Chen, G., Nicolson, M., 2000. Blood to brain transfer of leptin in normal J. Neurosci. 19, 6006–6016. 862
808 and interleukin-1beta treated male rats. Endocrinology 141, 1951–1954. Willis, C.L., Davis, T.P., 2008. Chronic inflammatory pain and the neurovascular unit: 863
809 Rizzi, M., Caccia, S., Guiso, G., Richichi, C., Gorter, J.A., Aronica, E., Aliprandi, M., a central role for glia in maintaining BBB integrity? Curr. Pharm. Des. 14, 864
810 Bagnati, R., Fanelli, R., D’Incalci, M., Samanin, R., Vezzani, A., 2002. Limbic 1625–1643. 865
811 seizures induce P-glycoprotein in rodent brain: functional implications for Xaio, H., Banks, W.A., Niehoff, M.L., Morley, J.E., 2001. Effect of LPS on the 866
812 pharmacoresistance. J. Neurosci. 22, 5833–5839. permeability of the blood–brain barrier to insulin. Brain Res. 896, 36–42. 867
813 Romero, L.I., Kakucska, I., Lechan, R.M., Reichlin, S., 1996. Interleukin-6 (IL-6) is Yu, C., Argyropoulos, G., Zhang, Y., Kastin, A.J., Hsuchou, H., Pan, W., 2008. 868
814 secreted from the brain after intracerebroventricular injection of IL-1 in rats. Neuroinflammation activates Mdr1b efflux transport through NFkappaB: 869
815 Am. J. Physiol. 270, R518–R524. promoter analysis in BBB endothelia. Cell. Physiol. Biochem. 22, 745–756. 870
816 Ronaldson, P.T., Persidsky, Y., Bendayan, R., 2008. Regulation of ABC membrane Yu, C., Kastin, A.J., Tu, H., Waters, S., Pan, W., 2007a. TNF activates P-glycoprotein in 871
817 transporters in glial cells: relevance to the pharmacotherapy of brain HIV-1 cerebral microvascular endothelial cells. Cell. Physiol. Biochem. 20, 853–858. 872
818 infection. Glia 56, 1711–1735. Yu, C., Pan, W., Tu, H., Waters, S., Kastin, A.J., 2007b. TNF activates MDR1 873
819 Rosenberg, G.A., Estrada, E.Y., Dencoff, J.E., Stetler-Stevenson, W.G., 1995. Tumor (p-glycoprotein) in cerebral microvascular endothelial cells. Cell. Physiol. 874
820 necrosis factor-alpha-induced gelatinase B causes delayed opening of the Biochem. 20, 853–858. 875
821 blood–brain barrier: an expanded therapeutic window. Brain Res. 703, 151– Zhang, R., Miller, R.G., Gascon, R., Champion, S., Katz, J., Lancero, M., Narvaez, A., 876
822 155. Honrada, R., Ruvalcaba, D., McGrath, M.S., 2009. Circulating endotoxin and 877
823 Sharief, M.K., Noori, M.A., Ciardi, M., Cirelli, A., Thompson, E.J., 1993. Increased levels systemic immune activation in sporadic amyotrophic lateral sclerosis (sALS). J. 878
824 of circulating ICAM-1 in serum and cerebrospinal fluid of patients with active Neuroimmunol. 206, 121–124. 879
825 multiple sclerosis. Correlation with TNF-alpha and blood–brain barrier damage. Zlokovic, B.V., 2005. Neurovascular mechanisms of Alzheimer’s neurodegeneration. 880
826 J. Neuroimmunol. 43, 15–22. Trends Neurosci. 28, 202–208. 881
882

Please cite this article in press as: Banks, W.A. The blood–brain barrier in neuroimmunology: Tales of separation and assimilation. Brain Behav. Immun.
(2014), http://dx.doi.org/10.1016/j.bbi.2014.08.007