Microglial Physiology: Unique Stimuli, Specialized Responses

ANRV371-IY27-05 ARI 16 February 2009 8:25

Microglial Physiology: UniqueStimuli, Specialized ResponsesRichard M. Ransohoff1 and V. Hugh Perry2

1Neuroinflammation Research Center, Lerner Research Institute, Cleveland Clinic,Cleveland, Ohio 44195; email: [email protected] of Biological Sciences, University of Southampton, Southampton SO16 7PX, UK;email: [email protected]

Annu. Rev. Immunol. 2009. 27:119–45

First published online as a Review in Advance onDecember 3, 2008

The Annual Review of Immunology is online atimmunol.annualreviews.org

This article’s doi:10.1146/annurev.immunol.021908.132528

Copyright c© 2009 by Annual Reviews.All rights reserved

0732-0582/09/0423-0119$20.00

Key Words

microglia, macrophage, inflammation, neurodegenerative disease,brain cytology, activation, regulation, central nervous system

AbstractMicroglia, the macrophages of the central nervous system parenchyma,have in the normal healthy brain a distinct phenotype induced bymolecules expressed on or secreted by adjacent neurons and astrocytes,and this phenotype is maintained in part by virtue of the blood-brain bar-rier’s exclusion of serum components. Microglia are continually active,their processes palpating and surveying their local microenvironment.The microglia rapidly change their phenotype in response to any dis-turbance of nervous system homeostasis and are commonly referred toas activated on the basis of the changes in their morphology or expres-sion of cell surface antigens. A wealth of data now demonstrate that themicroglia have very diverse effector functions, in line with macrophagepopulations in other organs. The term activated microglia needs to bequalified to reflect the distinct and very different states of activation-associated effector functions in different disease states. Manipulatingthe effector functions of microglia has the potential to modify the out-come of diverse neurological diseases.

119

Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.


INTRODUCTION TOMICROGLIA

Resident tissue macrophages are present in alltissues of the body, and the central nervoussystem (CNS) is no exception. An extensiveliterature had debated the origin of the mostnumerous of the brain macrophages, the mi-croglia, specifically whether they are of myeloidlineage or of neuroectodermal origin. There islittle need to rehearse these different points ofview; the fact that it took so long to establishthe myeloid origin of microglia tells us thatthese cells have a phenotype that is distinct fromother tissue macrophages. Many lines of evi-dence have converged to establish that the mi-croglia are of myeloid lineage, which was con-clusively confirmed by their absence from theCNS of PU.1-null mice (1), although the brainsof PU.1-null mice are readily repopulated bymicroglia following bone marrow transplanta-tion (2). Because microglia and the other pop-ulations of macrophages associated with thedifferent structural compartments of the CNSare part of the mononuclear phagocyte system(MPS), we can identify more specific questions:When do macrophages first invade the CNS?Are microglia derived from a specific pool ofprogenitors or monocytes? How are their num-bers maintained in the steady state? What arethe signals that lead to microglial activation?What are the phenotypes and functions of thesecells in health and disease?

This review relies on studies in humans,mice, and rats. It is worth mention that cellsthat are functionally equivalent to microgliaare observed in the invertebrate CNS. Stud-ies carried out in the leech (Hirudo medici-nalis) reveal a population of neuroectodermalcells termed small glia or microglia. Leech mi-croglial cells move rapidly to nerve crush le-sions, at least partly via purinergic signaling,as described also for mammalian microglia (seebelow), and they phagocytose debris. Leechesshow axonal sprouting and accurate reestablish-ment of synaptic contacts after nerve crush. In-triguingly, blockade of microglial accumulationat the lesion site impairs axon sprouting (3, 4).

Macrophage Populations of the CNS

Microglia in the adult mammalian CNS havea small cell soma, little perinuclear cytoplasm,and a number of fine, branched processes thatare covered in fine protrusions. The cells arereadily revealed in rodent and human CNSby lectin cytochemistry or immunocytochem-istry for selected antigens, and they occupy aterritory that does not overlap with adjacentmicroglia. The absence or low levels of ex-pression of many antigens typically found onother macrophage populations indicate that theCNS microenvironment plays a critical rolein defining the phenotype (see the discussionbelow in Regulation of the Microglia Pheno-type in the Normal Healthy CNS). Despitethe downregulated phenotype of microglia inthe normal healthy brain, in vivo imaging stud-ies demonstrate that the fine processes of mi-croglia continually palpate and monitor theirlocal microenvironment (5, 6). Microglia aredistributed throughout the CNS and vary indensity in both rodents and humans, with sub-tle variations in morphology in different cytoar-chitectural regions (7). It is unclear what localfactors determine their numerical and morpho-logical variations or whether these variationsreflect functional differences, although thereare subtle regional diversities in expression ofimportant immune receptors (8).

In addition to the microglia, there aremacrophage populations associated withthe perivascular space [the perivascularmacrophages (PVMs)], the circumventricularorgans, the choroid plexus, and the meninges.These macrophages have different phenotypeswhen compared with the microglia and ap-pear differentially constrained by their localmicroenvironment. PVMs express antigensnot expressed on the microglia. For example,the mannose receptor is detected on PVMsof mouse and human (9) and CD163 inrat and human (10). Similarly, a significantproportion of meningeal macrophages expressthese receptors. The PVMs and meningealpopulations are more overtly phagocytic thanare the microglia, and the expression of major

120 Ransohoff · Perry

Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.


histocompatibility complex (MHC) antigens ismore widespread in these populations of cellsrelative to the microglia. Macrophages are alsopresent in regions of the brain where theyare exposed to the blood, including the cir-cumventricular organs and the choroid plexus,and the phenotype of these macrophages isdistinct from the microglia and PVMs, perhapsreflecting the role of serum proteins in theregulation of microglia phenotype (11).

An important question regarding the im-mune privilege of the CNS is whether thereexists a population of dendritic cells (DCs), theprofessional antigen-presenting cells (APCs) ofthe immune system (12). Since the first reportsdemonstrating a lack of MHC class II expres-sion in the CNS, there has been no convincingdemonstration of DCs, either by expression ofrelevant DC surface markers or by DC func-tion in the healthy brain parenchyma (13), al-though DCs are present in the meninges andthe choroid plexus. This situation is radicallychanged in the diseased brain, however (see thesection below on Multiple Sclerosis).

Origin of Microglia

In postnatal rodents, immunocytochemicalstudies using antibodies to F4/80 (ERM1),CR3, and FcγRII/III suggested that the mi-croglia entered the brain from the circulationand presumably were derived from circulatingmonocytes (14). At earlier stages of develop-ment (prior to the development of the vascu-lature), cells of myeloid origin can be foundwithin the embryonic mouse CNS (embryonicday 8), where they proliferate (15). In the ze-bra fish, yolk sac–derived macrophages enterthe developing brain and retina, where they de-velop into immature microglia (16). The lin-eage origin of these cells from the progenitorcells in the yolk sac is unclear, but evidencesuggests that these primitive macrophages de-velop along a PU.1-independent pathway (17).Whether any of these primitive macrophagescontribute to the microglial populations thatappear later in development or persist intoadulthood is not known. In one study (18), more

than 20% of microglia isolated from imma-ture brain expressed the hematopoietic stemcell marker CD34, and 90% expressed theB220 antigen, an isoform of CD45 typicallyassociated with B cells and a subset of DCs.These markers are subsequently downregulatedduring maturation.

Maintaining the Microglial Population

Investigators widely believe that the mi-croglia are a long-lived population of tissuemacrophages, but how the resident popula-tions of brain macrophages are maintainedin homeostasis and during disease is notyet resolved. Recent studies show that resi-dent tissue macrophage populations may arisefrom specific subsets of monocytes ratherthan simply from stochastic recruitment orentry of monocytes into tissues. At leasttwo distinct subsets of monocytes are foundin mouse, Gr-1hi/CCR2+/Ly6Chi/CX3CR1+

and Gr-1lo/CCR2−/Ly6Clo/CX3CR1++ (19).Geissmann et al. (20) reported that Gr-1lo/CCR2− cells may give rise to tissuemacrophages, including the microglia, in thesteady state. Those investigations did not clarifyhow many cells enter the brain and into whichcompartment; furthermore, the cells that ap-parently entered the parenchyma did not per-sist for more than a few days. It has beensuggested that either CX3CR1 or homeostatictissue macrophage turnover mediates recruit-ment, but Cardona et al. (21) demonstratedthat CX3CR1 is not required for entry intothe brain. Thymidine labeling studies in adultmice demonstrate that microglia undergo DNAsynthesis, divide, and contribute to the main-tenance of this population (22), akin to thesituation found for other long-lived residentmyeloid cell populations.

To address the question of the relative con-tribution of circulating monocytes or intrinsicdivision to the steady-state population, inves-tigators have typically used radiation chimerasor parabionts. However, the complications in-troduced by whole body irradiation, includ-ing irradiation of the brain, have led to some

www.annualreviews.org • Microglial Physiology 121

Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.


confusion. The use of radiation chimeras andparabiosis to study microglia turnover is dis-cussed below in the section on Irradiation BoneMarrow Chimerism and Parabiosis.

The extent to which other populations ofbrain macrophages, PVMs, or meningeal cellsturn over is also poorly understood. Severalbone marrow chimera studies have suggestedthat these cells turn over more rapidly than mi-croglia, but again there are technical issues toconsider (discussed below). By injecting Indiaink into the brain parenchyma, PVMs were la-beled that could be identified even two yearslater (23), suggesting that this population mayalso be long-lived.

RESEARCH METHODS—THEIRSTRENGTHS AND LIMITATIONS

In Vitro

In vitro microglial culture. In vitro systemsprovide a critical tool for exploring many as-pects of microglia biology. Removal of thesecells from their microenvironment will releasethem from the normal constraints that play sucha critical role in their phenotype. Thus, usingin vitro systems assumes we recognize eitherthat the cells must be studied rapidly after iso-lation or that the investigator must attempt torecapitulate features of the CNS microenviron-ment that will in turn give clues about the fac-tors regulating their phenotype. A significantproblem is the identification of the microgliaphenotype in vitro, as there is no single mi-croglia marker and the microglial phenotype isdefined in vivo by a combination of morphol-ogy and often lack of or low expression of mul-tiple macrophage antigens. This circumstancehas led to a relatively loose use of the term “pro-cess bearing” or other terms for the characteri-zation of microglia in vitro, and all too often noattempt is made to address whether they dif-fer from macrophages. Despite these caveats,many studies rely on the isolation and expan-sion of microglia from the neonatal brain as amodel for the study of resident microglia func-tion. Another limitation of studying neonatal

microglia is that they have not experienced theCNS milieu in vivo in the context of an intact,mature blood-brain barrier (BBB). Ponomarevet al. (24) recently described a protocol forisolating adult rodent microglia.

The importance of astrocyte-microglia in-teractions as part of the environmental reg-ulation is well illustrated in experiments byRosenstiel and colleagues (25), who have shownthat microglia and other macrophage popula-tions grown on an astrocyte monolayer developa highly ramified morphology that is associatedwith a downregulation of nuclear factor-κB(NF-κB). Whether expression of a spectrumof cell surface antigens is also downregu-lated has not been systematically studied.Isolation of adult microglia and neonatalmicroglia shows these cells to be poor or im-mature APCs, although treatment with gran-ulocyte macrophage colony-stimulating factor(GM-CSF) will lead to differentiation to a moreDC-like phenotype (26). Isolation of adult mi-croglia and subsequent culture in low lev-els of macrophage colony-stimulating factor-1(M-CSF) lead to proliferation of the surviv-ing cells that then develop a modestly ramifiedmorphology that could be maintained for manyweeks (24), consistent with the notion thatM-CSF is a key component of tissuemacrophage survival and phenotype. Mice lack-ing M-CSF do exhibit normal morphology andapparent function of microglia, however (27).

Elucidation of many aspects of macrophagesand microglia clearly can be readily accom-plished in vitro, but the direct relevance of theseobservations needs to be established throughin vivo studies before accepting them at facevalue. In most in vitro studies, there has beenno attempt to replicate the CNS microenviron-ment, and the state of the cells grown on glassor tissue culture plastic is likely more relevantto inflammatory cells rather than to steady-statemicroglia.

Brain slices. An intermediate approach to invivo characterization or the isolation of mi-croglia from their natural environment is toinvestigate their properties in accessible acute


Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.


brain slices or organotypic cultures. Acute brainslices suffer from the obvious consequences ofthe slicing process, and one might imagine thatthe microglia would be rapidly activated, butin contrast to isolated in vitro microglia, theion channel expression in the slice cultures isdistinct from that found in cells in vitro andconsistent with a downregulated or residentphenotype (28). Organotypic cultures permitthe study of microglia with a resident phenotypeafter several days in vitro, and subsequent injuryto these slices leads to somewhat predictable re-sponses given what we know of the in vivo con-dition and the responses of macrophages (29).Few reported studies have used slice culture,and wider exploitation of this approach mightbe productive, given the evident limitations ofin vitro culture of isolated microglia.

In Vivo

Immunohistochemistry and in situ hy-bridization. Given the marked limitations im-posed by studying microglia in vitro (see above),much effort has been devoted to evaluating mi-croglial biology in the intact CNS. Immuno-histochemistry (IHC) has been widely used tocharacterize microglia, both in human tissuesand in those from experimental animals. Thestrengths of this approach are manifold, andno other technique can provide a comparablewealth of information. Different from RNAanalysis [such as in situ hybridization (ISH),Northern blotting, or reverse transcriptase-coupled polymerase chain reaction (RT-PCR)],IHC detects protein, which is more likely tobe functionally relevant than mRNA. Further-more, unlike Western blotting or enzyme-linked immunosorbent assay (ELISA), IHCenables protein both to be detected and to belocalized to specific cell types with regional dif-ferentiation. With proper controls, IHC sup-ports quantitative morphometry. In addition,using high-resolution techniques such as con-focal microscopy with immunofluorescence,subcellular localization can be achieved, occa-sionally with functional implications. For exam-ple, stimulus-dependent transcription factors,

such as the signal transducers and activatorsof transcription (STATs) and the NF-κB com-ponents, translocate to the nucleus upon lig-and engagement of upstream receptors. There-fore, the detection of STATs in microglial nucleiimplies ligation of relevant cytokine receptors(30). IHC can accordingly yield an incompara-bly vivid depiction, albeit static, of physiologicaland pathological processes.

Against these strengths lies the cardinalweakness that IHC is fraught with artifact,much of which is deceptively subtle and thecauses of which are legion. Artifactual im-munostaining can appear crisp and specific withlow or absent background. Because of IHC’simportance in clarifying neurobiological pro-cesses, guidelines for validating antibody stain-ing in neuroscience have been proposed (31,32). These guidelines are useful both for thoseperforming experiments and for those evalu-ating experimental results. At the outset, it isevident that simple technical controls such asomission of primary antibody represent only afirst step in this crucial process. Pertinent sug-gestions include use of Western immunoblot-ting to verify presence of the antigenic targetin tissue lysates and application of multiple in-dependent antibodies directed against a singletarget to verify cellular localization. Where pos-sible, antibodies can be used in negative controlexperiments to stain tissues of animals [suchas gene-targeted mice or patients with genemutations (33)] that definitively lack the targetantigen. These guidelines also recommend thatsubcellular localization of target antigen shouldcorrespond to the known function of the pro-tein; immunoreactivity of receptor antibodiesshould typically demonstrate plasma membranedistribution, for example.

Parallel performance of ISH with IHC isextremely useful for confirming that target-encoding mRNA resides in the same cell,region, and physiological context as the im-munoreactivity. This analysis can be assisted insilico using the Allen Brain Atlas (http://www.brain-map.org/welcome.do), a searchableonline compendium of ISH experimental re-sults, that can be used to localize mRNAs in the


Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.


mouse brain with impressive anatomical detail.Definitive experiments combine IHC and ISH.

In summary, tissue IHC is a centrally impor-tant technique for microglial research becausethe most valid observations of these enigmaticcells need to be conducted in the intact CNS.At the same time, it has been famously difficultto apply IHC to characterize microglia owingto their relatively low expression of MPS mark-ers. Adherence to guidelines for validating IHCresults is therefore critical for research intomicroglia.

Two-photon imaging. Excitation fluores-cence imaging allows visualization of neu-ral structures to a depth of 100–200 μm,through a thinned-skull preparation. This ap-proach has been successfully applied to neu-robiological questions related to dendriticphysiology for more than a decade (34). Amajor advantage is that this technique affordsprolonged imaging sessions, without photo-bleaching or photodamage. As noted above, twogroups used two-photon fluorescence imagingto study mice in which CX3CR1, the recep-tor for chemokine CX3CL1, was replaced withenhanced green fluorescent protein (EGFP)(Cx3cr1+/GFP mice). Both groups made the samesurprising observation: that microglia in the in-tact, healthy CNS continually remodel theirprocesses, in apparent surveillance of the ex-tracellular milieu (5, 6). This landmark findingprovided an entirely new view of parenchy-mal cells previously stigmatized as resting mi-croglia (35). Their dynamic surveillance ofthe CNS involved processes at all levels ofbranching, without movement of the soma,and was calculated to monitor the entire tissueevery several hours. Contacts with astrocytes,vascular elements, and neurons were visu-alized. Blocking voltage-gated sodium chan-nels (and thus neuronal action potentials) withtetrodotoxin did not affect process motility.Further research motivated by these findingsshowed that microglia walled off laser lesionsextremely rapidly through process extension,again with minimal initial displacement of thesoma (5). Extracellular nucleotides/nucleosides

such as ATP or ADP were required for mi-croglial process movement in response to in-jury (but not in surveillance of the healthyCNS) and signaled through the metabotropicpurine receptor P2Y12 (5, 36). The signals thatunderlie physiological process movement re-main to be established (35). Cx3cr1+/GFP micewere ideally suited for these studies, as theEGFP fluorescence was expressed in all corti-cal microglia, as judged by colocalization withmicroglial markers such as ionized calcium-binding adapter molecule 1 (iba-1), and wasrestricted to microglial cells, as determined bylack of colocalization with lineage markers forneurons and glial subpopulations (21). Inter-pretation of data gained by the application oftwo-photon imaging to microglial biology willdepend on which promoter-reporter is usedand on detailed knowledge of promoter activ-ity under physiological and pathological con-ditions. In this context, it bears repeating thatno microglial-specific promoter has yet beenidentified.

Irradiation bone marrow chimerism andparabiosis. Hematopoiesis research has longrelied on introducing labeled donor cells, withor without functional alterations, into recipientanimals. One widely used technique is to subjecta recipient host to lethal irradiation and pro-vide rescue by transferring bone marrow con-taining hematopoietic stem cells. An alterna-tive is parabiosis, involving anastamosis of thecirculations of host and recipient, allowing formixing of the circulating elements. Irradiationbone marrow chimerism to study the immunol-ogy of the CNS was first reported in the early1980s and addressed questions such as the iden-tification of CNS cells that expressed the MHCclass II antigens (37). Irradiation chimerism wassuitable for studies of this type because thepolymorphic MHC antigens enabled differen-tiation between the host and donor cells, andthe experiments gave unambiguous answers. Itwas noted early on that parenchymal microgliadid not display the markers of transferred cells,although perivascular MPS cells did (38), yield-ing a preliminary suggestion that perivascular


Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.


and parenchymal MPS cells turned over withdiffering kinetics.

The technique of irradiation chimerism wassoon adapted to address key questions in CNSimmunity and inflammation. A wide spec-trum of research employed the model diseaseexperimental autoimmune encephalomyelitis(EAE), involving sensitization of rodents withmyelin protein fragments or adoptive transferof myelin-specific T cells. In 1988, a seminalstudy used irradiation chimerism and adoptive-transfer EAE to demonstrate that PVMs of theCNS could stimulate myelin-reactive T lym-phocytes adequately to bring about demyelina-tion (39). These results were updated recently,using additional controls to show that PVMsof the CNS were necessary and sufficient foradoptive-transfer EAE (40).

Irradiation chimerism was used to addresscentral questions regarding the turnover andproliferative capacity of microglia and perivas-cular cells (41–43). The response of engrafted(infiltrating) and resident cells to varied dis-ease or injury models was extensively compared.Irradiation chimerism seemed well adapted forexamining the comparative roles of the resi-dent microglial cells and the infiltrating blood-derived elements.

It now appears that data from these earlierirradiation chimerism experiments need reeval-uation in light of studies that directly addresspotential confounds arising from the prepara-tion (44–46). Three sources of confound arerecognized: (a) Use of bone marrow cells to re-constitute the hematopoietic system might leadto nonphysiological numbers of hematopoieticstem cells or progenitors in the circulation.(b) Lethal irradiation-induced cell death in thehematopoietic system is associated with enor-mous fluxes of cytokines through the circula-tion and in tissues. (c) Irradiation of the CNScauses vascular changes, permanently affectingcompetence of the BBB. Together, these con-stituents of the irradiation chimerism protocolappear to lead to nonphysiological transmigra-tion of cells into the CNS parenchyma.

Informative studies have used the facial-axotomy model, which results in brisk mi-

croglial reaction in the ipsilateral facial nervenucleus, after facial nerve crush in the periph-ery (47, 48). Previous studies showed that fa-cial axotomy of irradiation chimeric rodentsleads to incorporation of donor-derived cellsin the microglial reactive population (43), andparabiotic animals showed few if any donor-derived microglia under physiological condi-tions (49). To address mechanisms for re-cruiting donor microglia into the axotomizedfacial nerve nucleus, parabiotic animals weregenerated, and the CNS vasculature of the re-cipient was conditioned by lethal irradiation,yielding recipients in which actin-GFP-labeleddonor cells predominated in the circulation ofthe recipient but did not contain bone mar-row elements beyond those found circulatingphysiologically (44). After facial axotomy, noGFP+ microglia were found in the reactivemicroglial population of the facial nerve nu-cleus. These results indicated that irradiationalone was not sufficient to allow circulating cellsto become parenchymal microglia after thistype of injury (44). In complementary exper-iments, irradiation chimerae were generated,using a protocol that spared the cranial vas-culature, and then subjected to facial axotomy(45). Again, no donor-derived microglia werefound in the axotomized facial nerve nucleus.The interpretation was that irradiation wasrequired (although not sufficient) (44) to en-able circulating cells into the parenchymalmicroglial pool. Therefore, based on di-rect comparisons between parabiotic andirradiation-chimeric preparations, investiga-tors now recognize that nonphysiological pro-cesses seed the CNS with donor microgliafollowing irradiation chimerism (46). Thesetwo complementary observations pointed tothe conclusion that both cranial irradiation andtransfer of bone marrow cells into the circu-lation were required for the nonphysiologicalentry of donor cells into the parenchymal mi-croglial population (46). Corollary conclusionsare that parenchymal microglia are replacedby proliferation of resident cells and that mi-crogliosis can be exceedingly impressive in theabsence of a contribution from circulating cells.


Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.


PET scan. Microglia are embedded in a densenetwork of neural interactions and are in-timately associated with neurons. Owing toits networked nature, the CNS demonstratesunique, distributed pathological patterns, bycontrast to other tissues. Damage to an axonexcites a brisk (retrograde) response by its re-mote cell body; damage to an axon also causes an(anterograde) reaction by local microglia thatinvests the terminus of the injured cell process;most remarkably, when a nerve cell loses in-put owing to elimination of a distant projectingneuron, there is a trans-synaptic response, in-cluding microglial reaction in the vicinity of thedeprived neuron.

Neuropathological description of these phe-nomena captures neither their extent northeir kinetics. Positron emission tomography(PET) scanning has been a useful adjunctin neurological practice, showing, for exam-ple, hypometabolism in affected regions ofbrain in neurodegenerative disease throughthe application of [18F] 2-fluoro 2-deoxy-d-

glucose (FDG) as a PET tracer. LabeledR-enantiomer of an isoquinolone [11C] (R)-PK11195 binds selectively to a mitochondrialmembrane translocator protein (TP)-18, partof a protein complex previously termed the pe-ripheral benzodiazepine-binding site (PBBS).This binding activity is expressed selectivelyby myeloid cells in the CNS, as shown bycorrelation of IHC and autoradiography. PETstudies using PK11195 take advantage of thepartition of this moiety into brain across thecompetent BBB, as well as its selectivity foractivated, in contrast to surveillant, microglia.The binding activity remains relatively constantacross the spectrum of morphological varia-tions of microglial activation (activated, rod-like, amoeboid) and also binds to infiltratedmacrophages. The technique requires signifi-cant postprocessing and is challenged by con-founds in circumstances of BBB disruption.Having accepted these limitations that, to date,preclude routine clinical application, PET withPK11195 has provided proof-of-principle formicroglial activation in neurodegenerative dis-orders, where trans-synaptic processes must be

operative, as well as in the lateral geniculateprojection fields of optic nerves of multiple scle-rosis (MS) patients suffering from severe opticneuritis (50–52). Furthermore, this techniquedemonstrates early microglial activation as wellas remarkably prolonged activation in neurode-generation. Documenting microglial activationin these clinical settings has been a prominentcatalyst for ongoing research, but the pheno-type associated with PK11195 upregulation hasnot been defined.

Genomics, Proteomics,and Other “Omics”

MPS cells originate from a clonogenicbone marrow progenitor that gives rise tomacrophages and DCs (53). Despite theircommon origin, components of the MPS aremarkedly heterogeneous, beginning with thecirculating elements: As noted above, inflam-matory and resident populations of bloodmonocytes have been tentatively identified inboth humans and mice using surface markers(20). Although the two monocyte populationsare believed to originate from a bone marrowprogenitor by cell-autonomous mechanisms,differential properties of tissue macrophagesprobably derive mainly from environmentalcues (54). The properties of tissue macrophagesreflect the demands placed on them by hosttissues. Cutaneous Langerhans cells expressfunctions associated with pathogen recogni-tion and entrainment of host defenses. In bone,osteoclasts reflect the requirement for con-tinual tissue remodeling. Pulmonary alveolarmacrophages express host-defense functions as-sociated with airborne pathogens and particu-late matter, whereas thymic macrophages needhigh-capacity ability to engulf apoptotic cells.Growth factors related to the generation andmaintenance of some of these populations havebeen described: Mice lacking M-CSF exhibitgrossly defective osteoclast development butonly subtle alterations in microglia, whereasthose deficient for GM-CSF show faulty alve-olar macrophage function resulting in alveo-lar proteinosis. Deficiency for the ets family


Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.


transcription factor PU.1 results in failed de-velopment of the myeloid lineage and, as notedabove, provided proof of the myeloid origin ofmicroglia (1).

It is tempting to consider how con-cepts related to macrophage heterogeneitycould be informative for microglial physiology.Macrophage functions are evidently adapted tohost tissues. Even within a single organ suchas spleen, regional macrophage heterogeneityhas been described. The CNS MPS popula-tions exhibit phenotypic heterogeneity, plausi-bly related to function. For example, parenchy-mal microglia may infrequently be exposed toviral pathogens but rarely to bacteria, fungi,parasites, or particulate matter. Their Toll-like receptor (TLR) complement comprisesat minimum TLR1 through TLR9, but rela-tive expression levels and signaling efficiencyhave not been defined (55). Adult parenchy-mal microglia would, however, be confrontedwith lipid-rich tissue debris and, to a lesserextent, with apoptotic cells. Appropriate re-ceptors, including scavenger receptors (SRs)and apoptosis-recognition components, appearto be expressed by parenchymal microglia.Surveillant microglia also express P2Y12, a re-ceptor for ATP and ADP, which mediates pro-cess extension in reaction to tissue damage (36).

Parenchymal microglia encounter T lym-phocytes only under pathological conditions(56). Other CNS MPS cells, including PVMs,choroid plexus, epiplexus, and meningealmacrophages, may, however, be called uponto present antigen to T lymphocytes withinthe subarachnoid space (57). Accordingly,parenchymal microglia express MHC class IIdeterminants at a low level compared with otherCNS MPS cells; furthermore, MHC class IIon parenchymal microglia appears cytoplasmicrather than associated with the plasma mem-brane when localized by confocal microscopy(58). Therefore, one can envision diverse MPScells in the CNS, analogous to those foundthroughout the organism.

Defining how the varied but closely relatedcell populations of the MPS differ at the levelsof gene expression, protein production, and

signaling has proven a tantalizing and produc-tive field of investigation (59). As one example,the Alliance for Cell Signaling (AfCS, http://www.signaling-gateway.org) conducted anextensive analysis of the responses of RAW264.7 cells (a murine macrophage-like, Abelsonleukemia virus–transformed cell line on theBALB/c background). For this study, RAW cellswere exposed individually to each of 22 separateligands and, subsequently, to 231 pairwise com-binations of ligands (60). Evaluation of responseincluded cytokine secretion, second messengergeneration, and signaling protein phosphoryla-tion. Applying matrix analysis of these outputsto address how cells respond to multiple simul-taneous inputs was revealing: There were re-markably selective cytokine outputs in responseto combinations such as TLR- plus G protein–coupled receptor (GPCR)-mediated signaling.Evaluation following varied ligand combina-tions including purinergic- and prostaglandin-responsive GPCRs provided sufficientinformation to derive a dynamic scheme forcross-regulation of cAMP and Ca2+ mobi-lization. Initial application of the brute-forceapproach was required to obtain this type of in-formation, as existing data sets would not haveallowed generation of a broad concept relatingcAMP levels and Ca2+ in a dynamic interaction.

Use of these multifold approaches to an-alyze CNS MPS cells is constrained byconsiderations described above for in vitroapproaches to microglial biology. The most ro-bust physiological data can likely be generatedonly from cells analyzed immediately ex vivo,but such cell preparations are both labor inten-sive and resource consumptive (61). Keepingthese limitations in mind, data sets combininghigh-density oligonucleotide microarrays andproteomics platforms have been described forcultured microglia exposed to disease-relevantstimuli such as nitrated alpha-synuclein aggre-gates, a moiety present in the CNS of individ-uals with Parkinson’s disease (62). Comparingresults from such studies with those obtainedthrough examination of other MPS populationssubjected to other stimuli will be informative(63).


Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.


Systems biological studies, in which com-putational bioinformatics tools are applied toexploit fully the information in unbiased geneexpression data sets, have provided novel in-sights into the MPS and will also be essentialfor research into microglia. One instructive re-port began by compiling a kinetic descriptionof LPS-induced genes in macrophages. Clusteranalysis yielded the unexpected prediction thatthere was an identifiable group of transcriptspredicted to be regulated by the cAMP responseelement binding-protein (CREB) family mem-ber activating transcription factor (ATF)-3 (64).Network analysis suggested that ATF-3, notpreviously ascribed an immune or inflamma-tory function, modulated a subset of NF-κB-induced transcripts that were induced by TLR4ligation in macrophages. Specifically, ATF-3seemed to be a negative regulator of the latewave of NF-κB-induced genes such as IL-6.Biochemical studies validated this hypothesisboth by showing closely approximated bindingsites for ATF-3 and NF-κB transcription fac-tor complexes in the promoters of the putativeregulated genes and by demonstrating physicalinteractions between ATF-3 and NF-κB com-ponents. This type of research, beginning withunbiased descriptive compilation of regulatedgenes and culminating in delineation of reg-ulatory pathways, should be applicable to dis-secting how microglia respond to the distinctivestimuli with which they are confronted withintheir unique environment. Microarray experi-ments using cultured microglia have been re-ported but have not taken systematic advantageof the strengths of kinetic description of gene-expression changes after stimulation and havenot been comprehensively exploited with bioin-formatics tools (65, 66).

Integrated molecular, bioinformatics, andbiochemical approaches will be essentialfor characterizing microglial responses tostimuli and will benefit from existing richdata sets. As one example, the Innate ImmuneDatabase (http://www.innateimmunity-systemsbiology.org) contains a searchabledatabase of more than 150 microarray exper-iments in which TLR ligands were used to

stimulate murine macrophages from a singleinbred strain (67). The pathway to use thesetools to explore microglial biology has beenably blazed by the MPS research communityand now awaits utilization for CNS MPSstudies.

AFFERENT SIGNALS FORMICROGLIAL ACTIVATION—NEURODEGENERATIONAND INFLAMMATION

Regulation of the Microglia Phenotypein the Normal Healthy CNS

One of the striking features of microglia beyondtheir morphology is their distinct downregu-lated phenotype when compared with other tis-sue macrophage populations. Surprisingly, onlyrecently have factors emerged that might be re-sponsible for this phenotype.

Soluble Factors

A first place to seek factors that regulate mi-croglia phenotype is in the interstitial fluid, thesolution that bathes them. In contrast to mostother macrophages, these cells are shieldedfrom serum proteins that might lead to theiractivation, and there is evidence that serumconstituents can selectively and potently acti-vate macrophages (68). Furthermore, the cy-tokine mediator profile in the healthy adultCNS is consistent with a downregulated pheno-type with relatively readily detectable levels ofcytokines such as transforming growth factor-β (TGF-β) and prostaglandin E2 (PGE2). Acytokine both soluble and bound that plays akey role in tissue macrophage survival, pro-liferation, and differentiation is M-CSF. Micedeficient in M-CSF (op/op) show reduced num-bers of microglia with minor alterations in mor-phology (27), but no difference in the PVMpopulation.

Cellular Interactions

The microglia are in intimate contact withcells in their immediate environment, and the


Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.


expression of receptor-ligand pairs on the mi-croglia and their neighbors will have a potenteffect on their phenotype. The state of acti-vation of myeloid cells is determined in partby the relative levels of expression of receptorsexpressing immunoreceptor tyrosine-based ac-tivation motifs (ITAMs) and immunoreceptortyrosine-based inhibition motifs (ITIMs) (69).Following receptor ligation, receptors bearingan intracellular ITAM consensus sequence arephosphorylated by Src family protein kinases,which leads to docking of SH2 domains of Sykprotein kinases and an intracellular activationcascade. In contrast, ITIM-bearing receptorsare phosphorylated by Src with recruitment ofSHP-1 and SHP-2 phosphatases, with conse-quent decreased intracellular activation. Thereceptor CD200R is expressed on microglia andhas an ITIM motif. In mice lacking CD200,normally expressed on neurons, the microgliashow constitutive upregulation with changesin morphology and expression of MHC classII, and these changes also lead to more severedisease in EAE (70). There are a number ofother receptor-ligand pairs that might be ex-pected to play a similar role. CD172a/SIRPaexpressed on macrophages binds the relativelyubiquitous ligand CD47, leading to downregu-lation of phagocytosis via an ITIM motif (71).Neumann and colleagues (72) have shown thatmicroglia express TREM2 (triggering receptorexpressed on myeloid cells-2) and that this re-ceptor is involved in the phagocytosis of de-bris and downregulation of proinflammatorycytokine expression. Loss or ablation of thisreceptor leads to deficient removal of cellu-lar debris from apoptotic cells but enhancedexpression of inflammatory mediators. Pre-cisely how signaling via TREM2 leads to in-activation of the macrophage is as yet unclearbecause this receptor signals via the adaptormolecule DAP12 normally associated with cellactivation. Recent studies show that deletion ofTREM2 or of DAP12 from macrophages leadsto identical phenotypes (namely, enhanced in-flammatory cytokine production) when chal-lenged with TLR agonists (73). Importantly,homozygous deficiency of either TREM2 or

DAP12 in patients leads to adult-onset dement-ing leukoencephalopathy, providing proof-of-principle that microglial functions are requiredfor CNS homeostasis. The expression of thesereceptors and others such as the Siglecs (74)will likely contribute to maintaining the state ofinactivation of the microglia. Accordingly, dis-turbances or loss of their ligands during patho-logical processes will contribute to microglialactivation.

Defining Microglial Activation

2008 marked the 150th anniversary of a sem-inal description of microglia by Virchow, andmorphological accounts of resting and activatedmicroglia were embedded in that first report.Subsequent studies of microglial activation be-gan with a premise that activated microgliaemerged from a resting state and underwentmorphological transformation from ramifiedto varied activated forms, including amoeboid,rodlike, phagocytic, and so on. The notionof alternative forms of microglial activationseemed implicit in these studies but remain in-completely defined. At present, one can be-gin to integrate contemporary macrophage bi-ology with further understanding of microgliato attempt to reframe concepts of microglialactivation (Table 1).

Studies of the MPS have led to varied for-mulations of macrophage activation. Early dur-ing a pathological process, tissue macrophagesmay be stimulated either by nonself pathogens(stranger signals) or by injured-self compo-nents (danger signals). Both stimuli can acti-vate pattern-recognition receptors such as theTLRs, SRs, and the NOD system. Effectoroutputs of these stimuli focus on clearance oftissue debris, generation of cues for tissuerestoration, and resistance to pathogens. To-gether, these reactions comprise innate immuneresponses. Subsequent events may require es-tablishment of responses, including lymphocyteeffector functions, antibodies, and immunolog-ical memory, collectively termed adaptive im-munity. MPS cells contribute to this process byantigen presentation, including the instruction


Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.


Tab

le1

Alt

erna

teac

tiva

tion

ofm

acro

phag

esan

dm

icro

glia

a

Mac

roph

age

Stim

ulus

Rec

epto

rSi

gnal

ing

Act

ivat

ion

patt

ern

Out

put

LP

S,pe

ptid

ogly

can,

dsR

NA

TL

RN

F-κ

B/I

RFs

Inna

test

rang

erIn

flam

mat

ory

cyto

kine

s,R

OS,

RN

S

End

ogen

ous

TL

Rlig

ands

(dsD

NA

,HSP

s)

TL

RN

F-κ

B/I

RFs

Inna

teda

nger

Infla

mm

ator

ycy

toki

nes,

RO

S,iN

OS/

RN

S

IFN

-γ+

LP

SIF

NG

Rpl

usT

LR

4Ja

k/ST

AT+

NF-

κB

/IR

FsC

lass

ical

(M1)

Infla

mm

ator

ycy

toki

nes,

RO

S,iN

OS/

RN

S,in

crea

sed

antig

enpr

esen

tatio

nvi

aup

regu

late

dM

HC

IIan

dco

stim

ulat

ory

mol

ecul

es,

bact

erio

cida

lact

ivity

,iN

OS

IL-4

and

IL-1

3IL

-4/I

L-1

3RJa

k/ST

ATA

ltern

ativ

e(M

2a)

End

ocyt

icac

tivity

,man

nose

rece

ptor

,β

-glu

can

rece

ptor

,fibr

ogen

iccy

toki

nes,

para

sito

cida

lact

ivity

,ar

gina

se

TG

F-β

,IL

-10,

gluc

ocor

ticoi

ds,

CD

200

TG

F-β

R,

IL-1

0R,G

C-R

,C

D20

0L

Smad

s,Ja

k/ST

AT,

nucl

ear

horm

one

rece

ptor

,unk

now

n

Alte

rnat

ive-

deac

tivat

ing

(M2c

)

Red

uced

MH

Ccl

ass

II,r

educ

edin

flam

mat

ory

cyto

kine

s,an

ti-in

flam

mat

ory

pros

tagl

andi

ns

Mic

rogl

iaSt

imul

usR

ecep

tor

Sign

alin

gA

ctiv

atio

npa

tter

nO

utpu

t

Seru

mpr

otei

nssu

chas

fibri

noge

n

CD

11b/

CD

18(M

ac-1

)R

ho-f

amily

GT

Pas

esD

ange

rC

ytos

kele

talr

earr

ange

men

t,en

hanc

edph

agoc

ytos

is

Glu

tam

ate

mG

luR

2G

PC

R/N

F-κ

BN

euro

toxi

cT

NF-

α

ATP,

AD

P(b

rief

expo

sure

)P

2Y12

GP

CR

Dan

ger

Pro

cess

exte

nsio

n

UD

P(p

rolo

nged

expo

sure

)P

2Y6

GP

CR

Dan

ger

Pha

gocy

tosi

s

Per

iphe

raln

erve

inju

ryU

nkno

wn

Src-

fam

ilyki

nase

sA

xoto

my

and

loss

ofin

put

Via

P2X

4,m

edia

tene

urop

athi

cpa

in

CX

3CL

1C

X3C

R1

GP

CR

Inhi

bitio

nR

educ

edIL

-1,R

OS,

iNO

S/R

NS

CD

200

CD

200L

ITIM

Con

tact

-dep

ende

ntin

hibi

tion

Red

uced

infla

mm

ator

ycy

toki

nes,

less

MH

Ccl

ass

II

a Abb

revi

atio

ns:G

PC

R,G

prot

ein–

coup

led

rece

ptor

s;iN

OS,

indu

cibl

eni

tric

oxid

esy

ntha

se;I

RF,

inte

rfer

onre

gula

tory

fact

or;I

TIM

,im

mun

orec

epto

rty

rosi

ne-b

ased

inhi

bitio

nm

otif;

LP

S,lip

opol

ysac

char

ide;

RN

S,re

activ

eni

trog

ensp

ecie

s;R

OS,

reac

tive

oxyg

ensp

ecie

s;T

LR

,Tol

l-lik

ere

cept

or.


Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.


of T cells to adopt varied effector programs(Th1, Th2, Th17) and in some cases direct-ing them to the tissue from which pathogenicmaterial originated. As part of the immunolog-ical effector program, MPS cells respond to theproducts of activated T cells, including IFN-γ,TNF, and IL-1.

One useful exercise has been to organizeMPS effector responses along varied pathwaysby which macrophages can be polarized byexposure to pathogen or to T cell–derivedproducts (Table 1). Classically (or M1) acti-vated macrophages respond to LPS plus IFN-γ,whereas alternatively activated macrophages“see” either IL-4/IL-13 (M2a); immune com-plexes along with IL-1 (M2b); or a variety ofmodulatory stimuli, such as glucocorticoids,TGF-β, or IL-10 (M2c) (75). Microglia differdecisively from peripheral macrophages, andtheir activation likely does not follow theseprecise pathways; nevertheless, it is instruc-tive to consider how the underlying concept ofmacrophage heterogeneity might apply to mi-croglial responses.

Given the results discussed above from two-photon imaging (35), microglia may now beregarded as surveying the healthy CNS andengaging varied modes of progression fromthe surveillant state to effector microglia. Thisconcept should replace the notion that mi-croglia proceed from resting to activated (76).Next, stimuli for microglia are analogous to, butclearly distinct from, those confronted by pe-ripheral macrophages. Three (at least) clear dis-tinctions between microglia and macrophagesapply here:

1. Microglia reside behind the BBB so thatserum products represent danger signalsindicating BBB breach.

2. Microglia are MPS family members butneed also to be equally regarded as brainglial cells, so that altered synaptic ac-tivity with perturbation of neurotrans-mitter availability will affect microglialactivation states and effector properties.Regulation by neurotransmitter signal-

ing may intermittently affect peripheralmacrophages, as described for the cholin-ergic anti-inflammatory response (77),but neurotransmitter effects are more di-verse and pervasive for microglia.

3. Microglial cells, perhaps as a conse-quence of living among fragile, nonre-newing neurons, exhibit an actively re-pressed phenotype, so that removal of(mainly neuronally derived) inhibitoryconstraints constitutes a type of dangersignal, indicating that neuronal functionis impaired.

Here, we cite examples of each of these specialconsiderations for microglial activation.

BBB disruption as a danger signal. Serumconstituents activate microglia (which is a keylimitation of using in vitro cultures to exam-ine their physiology). One molecular interac-tion that underlies microglial activation duringdemyelinating disease may be exposure to fib-rinogen, which engages CD11b/CD18 integrinheterodimers (68). Suppressing this stimulatorypathway reduces the severity of EAE.

Microglial responses to neurotransmitteralterations. Glutamate is an excitatory neuro-transmitter, and excess glutamate stimulation isneurotoxic. Glutamate excess can result fromincreased release (by neurons) or decreasedclearance (which is mainly the responsibility ofastrocytes). Glutamate receptors can be cou-pled to calcium entry (ionotropic) or to GPCRsignaling (metabotropic or mGluRs). Microgliaexpress mGluRs and respond to fluxes in ex-tracellular glutamate. Microglial responses toacute and chronic injury are regulated in part byselective activation of mGluRs, culminating ei-ther in neurotoxic or neuroprotective outcomes(78, 79). Adding complexity, extrinsic inflam-matory stimuli regulate mGluR expression.

Danger signals from damaged cells andfrom altered neurotransmitter levels. As acomposite example of specialized danger re-sponses by microglia, ATP is a purinergic


Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.


neurotransmitter, signaling both to P2Xionotropic and P2Y metabotropic receptors.Microglia are endowed with both P2X and P2Ypurinergic receptors. As noted above, extracel-lular ATP, released in part from damaged neu-rons, acts through microglial P2Y12 to medi-ate process extension after a cortical laser lesion(5, 36). Therefore, P2Y12 can be regarded asa specialized PRR for danger signaling withinthe CNS. Neuronal cell death or injury con-fronts microglia with a requirement to functionas phagocytes, and another purinergic receptor,P2Y6, acting in a longer time frame than P2Y12(which transduces signals within seconds), ap-pears to play a significant role in microglialphagocytosis (80). Intriguingly, the endogenousdanger ligand for P2Y6 is extracellular UDP,again providing an instance of selectivity in themicroglial reaction to injury.

Microglial activation by loss of neural input.P2X4 receptors are induced on spinal microgliavia Src family kinases after peripheral nerve in-jury and are critically involved in the CNS me-diation of neuropathic pain (tactile allodynia) inpain models, although the specific mechanismby which this effect occurs is uncertain (81–83).This phenomenon, given its specificity, can beviewed as an alternative mode of microglial ac-tivation, evidenced by selective upregulation ofa purinergic receptor.

Microglia are actively repressed. CX3CL1is a membrane-tethered chemokine that is ton-ically released from CNS neurons through theaction of ADAM proteases. Microglia receivetonic inhibitory inputs through the CX3CL1receptor, CX3CR1 (21). If this tonic inhibi-tion is removed, microglial neurotoxicity isunleashed in response either to systemic in-flammatory stimuli or to damage to residentneurons (21). Release from constitutive inhi-bition is also believed to underlie the devastat-ing neurological consequences of deficiency ei-ther for microglial TREM2 or its intracellularadaptor DAP12 (84, 85). Although TREM2 isa transmembrane receptor–like molecule withclear-cut inhibitory function for varied stimuli,

its endogenous ligands remain to be fully char-acterized.

EFFECTOR FUNCTIONS OFMICROGLIA: DEVELOPMENTAND REPAIR

The early entry of macrophages into the em-bryonic brain was described above (Originof Microglia section). It is uncertain whetherthese cells persist and give rise to adult mi-croglia, but their potential roles in developmenthave attracted interest. Microglial developmen-tal functions may be informative for roles indisease, and conversely.

Synaptic Remodeling: Pruning,Stripping, and Plasticity

During CNS development, axonal connectionswith synaptic targets exceed those required andare reduced in a process termed pruning. Forsome years, it has been clear that synaptic prun-ing is aberrant in mice lacking specific MHCclass I determinants, although mechanisms con-tinue to unfold (86). Recently, Stevens et al.(87) found that mice lacking complement com-ponents C1q or C3 exhibit defects in visualsystem synaptic refinement that are similar tothose found in MHC class I–deficient animals.They and others (86, 87) proposed assigningmicroglia an effector role in this aspect of CNSdevelopment, as microglial cells are the CNSelements that express the appropriate comple-ment receptors.

Stevens et al. (87) also provided evidencethat synaptic loss might be associated withaberrant engagement of complement pathwaysin a mouse glaucoma model. The oppositepossibility has also been proposed: specifically,that microglia assist in synaptic repair afterbrain lesions. Synaptic repair is incorporatedalong with learning under the rubric of synapticplasticity. By extrapolation from developmen-tal synaptic pruning, Cullhein & Thams(88) suggested that microglia contribute toplasticity after lesions, by pathways similar tothose used for pruning excess synapses duringdevelopment and facial axotomy, but directevidence is lacking.


Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.


Inducing Apoptosis inSupernumerary Purkinje CellNeurons During Development

CNS development is characterized by waste:excess production of synaptic connections andgeneration of a surplus of cells. Supernumer-ary cells undergo apoptotic cell death andare engulfed without generating inflammationby microglia (14, 89, 90). It was long specu-lated that microglia might also provide inputsthat promote developmental neuronal apopto-sis (89), and data have been forthcoming to sup-port this contention. Cerebellar slice culturesfrom neonatal mice demonstrate apoptosis ofPurkinje cell neurons, which are elaborately in-vested with microglial processes. Furthermore,elimination of microglia with toxic liposomesreduces the efficiency of Purkinje cell commit-ment to apoptosis (91).

Do Microglia Produce NeuralGrowth Factors?

Can microglia also provide growth factors forneurons during development? Here the dataare indirect and rely solely on observations inadult mice. PU.1-deficient mice lacking mi-croglia exhibit grossly intact CNS structure andfunction. However, studies of adult neural stemcell niches provide tantalizing hints of sup-portive microglial-neural progenitor relation-ships. In adult mice, neural stem/progenitorcells (NSPCs) are localized in the hippocampalsubgranular zone (SGZ) and in the rostral sub-ventricular zone (SVZ). In culture, NSPCs giverise to both neurons and glia, thereby fulfillingcriteria of multipotency. Conditioned mediumof microglial cultures contained factors capa-ble of supporting prolonged neurogenesis, im-plying an instructive function for microgliatoward NSPCs of the adult SVZ (92). Mi-croglia secrete factors that direct NSPC mi-gration in vitro (93). Microglia also promotedneuronal differentiation in these studies (93).Taken together, these data suggest a role formicroglia in promoting CNS lesion repair byNSPCs. This concept was given relevance forneurodegenerative disease, through studies of

microglia from transgenic mice expressing mu-tant forms of human presenilin-1 (PS1) asso-ciated with familial Alzheimer’s disease (AD).NSPCs in these mice fail to exhibit environ-mental enrichment-mediated proliferation andneurogenesis. In vitro, NSPCs from transgenicmice expressing either mutant or wild-type PS1proliferated and differentiated equally. By con-trast, co-culture of microglia from mice ex-pressing mutant PS1 with wild-type NSPCs re-capitulated the phenotype of impaired prolifer-ation and neuronal differentiation. This effectwas mediated by microglial-derived secretedfactors. The data supported an important rolefor microglia in regulating NSPCs during phys-iology and pathology (94, 95).

Vascular DevelopmentIt is not clear whether microglia exert functionsthat promote CNS vascularization. The no-tion that they might exert such functions comesby inference from two observations: first, theprovocative finding that microglia are presentin the CNS before vasculogenesis and that in-vading vessels are closely invested with mi-croglia; second, that experimental CNS tissueimplants are colonized by microglia before theyare vascularized (89).

PATHOLOGY

Introduction

Microglia are exquisitely sensitive to distur-bances of their microenvironment, and theyhave been dubbed the sensors of pathology(96). The early and rapid response of microgliais entirely consistent with the role of tissuemacrophages as the first line of defense againstinfection or injury. The response of the mi-croglia is not a linear process varying simply bydegree, but rather, as has been documented inother macrophage populations, their responseis dictated by the nature of the stimulus, thereceptor repertoire that is engaged, and theprior state of the macrophage (97) (see sectionon Defining Microglial Activation). In the con-text of neurodegeneration, a number of impor-tant variables must also be considered in the


Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.


temporal and spatial domains. The two mostcommon neurodegenerative conditions of theCNS are stroke and AD. In the former, thedeath of neurons is extremely rapid, a matter ofminutes; in the latter, the pathology driven bythe presence of a misfolded protein may lead todeath of a slowly increasing number of neuronsover years or even decades. The former stimu-lus is a step function, the latter a slow ramp func-tion. The slow degeneration of neurons—theirsynapses, soma, axons, and myelin sheath—maycontinue for many years, providing a stimulusthat must lead to adaptive changes in the sur-rounding microglia. The adaptive changes maythen be influenced by other comorbidities andsystemic influences that communicate with thebrain (98).

Neurodegeneration versusNeuroprotection

A major theme in studies of the role of microgliain neuropathology is the dichotomy betweentheir contributions to neurodegeneration ver-sus neuroprotection. The role of macrophagesin wound repair is well documented and part ofthe innate immune response (99). The thinkingbehind this dichotomization is that if we canunderstand the two components we can min-imize the harmful and capitalize on the ben-eficial (100, 101). This is a hugely ambitiousaim when we recognize that this has yet to beachieved for significant clinical advantage inany organ, let alone within the complex con-text of the CNS. Furthermore, the CNS is animmune-privileged organ in which evolution-ary pressures have ensured that the innate andacquired immune responses are tightly con-trolled. Overcoming the immune privilege perse is unlikely to benefit the host except per-haps in combating infections. Finally, microgliathemselves might be targeted by a pathologicalprocess, which also affects adjacent tissues. Re-cent data from studying the genetically deter-mined inflammatory demyelinating metabolicdisorder X-linked adrenoleukodystrophy sug-gest that such a mechanism might play a part inthe cerebral form of this disease (102).

SPECIFIC NEUROLOGICALDISORDERS

Alzheimer’s Disease

It is now nearly 20 years ago that an in-nate inflammatory response in AD, the pres-ence of morphologically activated microglia,was described, rapidly followed by studies in-dicating that people taking nonsteroidal anti-inflammatory drugs were in some way protectedfrom the onset or progression of AD. The roleof inflammation in AD has been extensively re-searched and exhaustively reviewed (103, 104),but the contributions to the disease of the mi-croglia and the associated innate inflammatoryresponse are by no means clear. A wide varietyof inflammatory mediators are expressed by mi-croglia when they are challenged in vitro withthe Aβ peptides derived from the amyloid pre-cursor protein (APP). However, there is a lackof consensus on which cytokines or other me-diators are indeed present in the postmortembrain tissues of patients that died with AD. InAPP transgenic mice (which mimic the depo-sition of Aβ amyloid in the brain, a hallmarkof AD pathology), the microglia do indeed ap-pear morphologically activated, particularly inthe immediate vicinity of the plaques. But again,there is little agreement on the cytokine profileassociated with these microglia (98), which isconsistent with the difficulty in performing as-says of low abundance and highly soluble pro-teins such as the cytokines (see above).

A more direct route to assessing preciselywhich molecules are synthesized by microgliathat might either promote or delay disease isto study their impact on the plaque load. Theplaques in various mouse models, and in humanAD patients, inexorably increase with age, andthe conclusion must be that the microglia arerather inefficient phagocytes, which is perhapsnot surprising given their very slow removal ofneuronal and myelin debris after a stroke (105).However, indirect activation of the microglia byintracranial challenge with LPS leads to a de-crease in plaque load (106). Others suggest thatbone marrow–derived macrophages selectivelyinvade the region of the plaques and play a role


Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.


in removal of the amyloid material (107). Howsuch findings translate to the clinical situation isunclear because these models are not progres-sive neurodegenerative disease models. The useof bone marrow chimeras in these studies willalso have to be reassessed.

An exciting development in the attemptsto manipulate the immune system for bene-fit in AD comes from studies in APP trans-genics, which show that systemic immuniza-tion against Aβ leads to removal of the amyloidplaques (108). Several hypotheses have beensuggested as to how systemic antibodies mightachieve this, including microglia phagocyto-sis via FcRs. Alternatively, antibody-mediatedclearance of systemic/circulating Aβ wouldpromote increased drainage of Aβ from thebrain and inhibit Aβ fibril aggregation (109).These mechanisms are not mutually exclusive,and it is by no means clear that the microgliaare essential for effective removal of the amy-loid. This intriguing experimental approach hasbeen rapidly translated into human: Despite thefact that it is unclear how the amyloid is re-moved, this approach does permit a direct testof the hypothesis that removal of the amyloidplaques might slow disease progression. A re-cent neuropathological analysis of the brainsfrom a small number of patients who were vac-cinated in the AN1792 Elan Pharmaceuticalstrial shows a positive correlation between thedegree of plaque removal and the antibody re-sponse. However, clinical analysis of the immu-nized cohort revealed evidence neither of im-proved survival nor of a delay in the time tosevere dementia (109a). It is too early to con-clude that a vaccination strategy will not work,and it is possible that earlier or different vacci-nation strategies will be required to achieve abenefit.

Multiple Sclerosis

MS is an acquired inflammatory demyelinat-ing disorder of the CNS and is the leadingcause of nontraumatic disability among youngadults in the United States and Europe. Thecause of MS is unknown but includes both en-

vironmental and genetic factors, some of whichhave been identified (110). MS research hasdepended largely on an autoimmune animalmodel, EAE, that exhibits features of CNS in-flammatory demyelination and paralysis thatresemble aspects of MS. The model disease be-gins with peripheral immunization, usually withpeptide epitopes of myelin proteins. EAE is aT cell–dependent disease, most commonly me-diated by CD4+ T cells, including those thatexpress both Th1 and Th17 cytokine profiles.EAE proceeds either by monitoring animals un-til disease onset around two weeks postimmu-nization (active EAE) or by isolation and trans-fer of primed cells from draining lymph nodesor spleen (adoptive-transfer EAE).

Roles of microglia in EAE illustrate theirability to contribute to adaptive (auto)immunereactions that target CNS antigens. We havelong known that, to elicit EAE after adoptivetransfer, T cells require restimulation with anti-gen in the CNS compartment. Classic exper-iments used radiation bone marrow chimerasto generate animals in which bone marrow–derived cells expressed MHC determinants dis-tinct from those present on parenchymal mi-croglia. Adoptive transfer of myelin-specific Tcells that were restricted by the parenchymalcells did not cause disease, whereas those thatrecognized antigen in the context of PVMsderived from the bone marrow inoculum ledto EAE (39). In follow-up studies, eliminatingall potential APCs except for CD11c+ PVMsproduced mice that remained susceptible toadoptive-transfer EAE (40). A variant form ofadoptive transfer, in which EAE was primedby one myelin epitope and mice with ongoingdisease received unprimed TCR transgenic Tcells specific for a different epitope, was used toshow that myelin-specific T cell proliferationoccurred mainly in the CNS after disease on-set (111). This experiment extended conceptsof the capabilities of CNS DCs, which arise inthe context of active CNS inflammation (112).Nevertheless, it has yet to be shown that thehealthy CNS parenchyma contains DCs com-petent to initiate immune responses by present-ing antigen to naive T cells in vivo (113).


Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.


Following priming, do parenchymal mi-croglia contribute to neurobehavioral impair-ment in EAE? This issue was addressed by gen-erating CD11b-HSV-TK mice that expresseda lineage-restricted Herpes virus thymidine ki-nase (HSV-TK) suicide gene in myeloid cells,so that administration of gancyclovir wouldeliminate proliferating CD11b+ cells. Radia-tion chimerism with wild-type bone marrowproduced animals in which only the radiation-resistant parenchymal microglia remained sus-ceptible to gancyclovir-mediated death. Com-bining EAE immunization with gancyclovirtreatment led to mice exhibiting EAE with mi-croglial paralysis. These mice exhibited mildEAE signs, indicating a role for microglia inthe severity of EAE (114).

Are these studies informative for MS? Thejourney from EAE mechanism to MS patho-genesis is notoriously tortuous, the route baf-fling, and the destination uncertain. Proof-of-principle that MS is autoimmune has not beenforthcoming. For that reason, immunopatho-logic mechanisms of EAE seem more directlypertinent for MS than do applications of theprinciples of autoimmunity. From EAE, wecould predict that MPS cells would exhibit DC-like characteristics in MS tissues, a hypothesiswith experimental support (115, 116). In whitematter, actively demyelinating MS lesions areidentified by the presence of myelin debris inmacrophages. Early-active zones in which de-myelination began within a few days can bedistinguished from late-active zones in whichdemyelination has been present for 2–4 weeks,by defining whether labile or relatively stablemyelin breakdown products are found withinphagocytes (117). Early-active MS lesions con-tain a mixture of infiltrating hematogenousmonocytes and microglia, which are differen-tiated by morphology, localization, and sur-face markers; late-active MS lesions are typifiedby a monomorphic population of phagocyticmacrophages (117). Numbers of MPS cells inindividual early-active and late-active MS le-sions are identical, suggesting that phagocyticmacrophages in late-active regions arose bothfrom monocytes and microglia. Enumeration of

monocytic and microglial cells in early-activezones suggests that between one-half and two-thirds of phagocytic macrophages in late-activeMS white matter lesions arise from microglia(117, 118).

Recently, it has become clear that MS is notexclusively a white matter disease: Remarkably,cortical (or gray matter) demyelination is moreextensive, as a fraction of total myelin, than iswhite matter demyelination. The largest andmost mysterious demyelinating foci of MS graymatter affect the most superficial three layersof the cortical ribbon and are termed subpiallesions (119). Among inflammatory demyeli-nating diseases, subpial lesions are relativelyspecific for MS (120). Although the mecha-nisms underlying subpial lesions remain un-clear, the inflammatory characteristics of suchlesions in chronic-progressive disease are note-worthy: There is a paucity of lymphocyte ormonocyte infiltration, as evidenced by sparse-ness of perivascular cuffs, whereas microglialactivation is remarkably robust (119). The datasuggest the possibility that microglia are keyeffectors in the process of subpial demyeli-nation, possibly activated by diffusible factorsfrom the cerebrospinal fluid. Subpial demyeli-nation may be a critical determinant of dis-ability in MS patients (121), so its mechanismsare crucially important for therapeutic devel-opment. Equally vital, data to date compriseonly chronic material; imaging studies indicatethat cortical damage occurs very early duringthe course of MS, and the mechanisms underly-ing cortical demyelination in the initial phasesof disease have yet to be characterized. Rolesof microglia, which seem prominent in chroniccortical demyelination, may be different in theearlier stages of MS.

Can microglia promote repair of MS le-sions? Remyelination is clearly a prevalentand poorly understood attribute of MS lesions(122). Many factors elaborated by activatedmicroglia might support oligodendroglial sur-vival or function (123). However, present MStherapy aims at complete elimination of MS-associated neuroinflammation. Until this goalis shown by evidence to be impossible, it is most


Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.


logical and straightforward to strive to abrogateCNS inflammation, rather than allow damagewhile trying simultaneously to bias the processtoward enhanced repair.

Prion Disease

Prion disease offers a tractable laboratorymodel to study many aspects of fatal neurode-generation associated with protein misfoldingand has the additional attraction that it is thesame disease in both humans and animals. Inmurine prion disease, the microglia become ac-tivated early in the disease process, even in theabsence of widespread histologically detectableamyloid protease–resistant PrPsc deposits(124). The cytokine profile associated with thisactivated phenotype was not dissimilar fromthat observed in APP transgenics, featuringlow levels of inflammatory cytokines but, in ad-dition, readily detectable levels of TGF-β andPGE2, a phenotype previously associated withmacrophages that had phagocytosed apoptoticcells (125). This phenotype appears prior to theonset of neuronal apoptosis but in associationwith synaptic loss (126), indicating first thatprion disease is a member of diseases with anearly synaptic degeneration (a synaptopathy),and second that this may lead to an atypicalmicroglial activation that has been referredto as anti-inflammatory or benign. In priondisease induced on a wild-type background,there is no evidence that the enhanced levelsof PGE2 are detrimental, nor that TGF-βis injurious. However, when TGF-β is neu-tralized by decorin, delivered by an otherwiseharmless adenoviral vector, this leads to acuteand marked neuronal degeneration, high-lighting the importance of anti-inflammatoryregulation in chronic neurodegenerativedisease (127).

To investigate how the activated microgliaphenotype might be affected by systemicinflammation, a common occurrence in clini-cal neurodegenerative disease, mice were chal-lenged systemically with endotoxin to mimican intercurrent infection. This maneuver led toa dramatic switch in the microglial phenotype

with an aggressive inflammatory cytokine pro-file, exacerbation of the sickness behavior asso-ciated with endotoxin challenge, and increasedneuronal apoptosis (128). Investigators haveproposed that the microglia in the prion dis-eased brain are primed by the ongoing pathol-ogy and that a secondary stimulus, the signal-ing of systemic inflammation across the BBB,switches these cells to an aggressive phenotype(98). Whether microglia switched to an inflam-matory phenotype phagocytose or degrade theprion protein by secretion of relevant proteasesremains unclear.

The concept of the rapid switching of themicroglia phenotype is of course entirely inkeeping with what we know of the consid-erable degree of plasticity of the cells of themacrophage lineage. Systemic inflammationhas a profound impact on a number of otheranimal models of neurological disease (98)and accelerates cognitive decline in Alzheimer’spatients (129).

Amyotrophic Lateral Sclerosis (ALS)

ALS is a fatal neurodegenerative disease thatcan involve toxic gain-of-function of a mutantprotein. In contrast to prion diseases, inwhich the protein accumulates predominantlyextracellularly, the toxic protein in ALS isan intracellular protein and exerts its actionintracellularly. Approximately 5–10% of ALSis an autosomal-dominant familial form ofthe disease, of which a significant proportionhas mutations in the enzyme superoxidedismutase-1 (SOD-1). Microglial activationis present in the vicinity of the degeneratingmotorneurons (130). Transgenic mice express-ing the mutant SOD-1 have been a valuablemodel for understanding disease pathogenesis,revealing that the demise of the motorneuronsis not a cell-autonomous event, but ratherdepends on the surrounding non-neuronalcells (131). Dissection of the role of microgliahas come from several different approaches.

Mice in which the microglia are no longerable to respond to the downregulatory influ-ence of the chemokine CX3CL1 have been


Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.


crossed with SOD-1-expressing mice, and thesemice show a modest shortening of survival time(21). More dramatic effects have been uncov-ered in studies in which mice expressing Cre-recombinase driven by the CD11b promoterare crossed with a LoxSOD1G37R mouse, lead-ing to levels of mutant SOD-1 that are signifi-cantly reduced in microglia (132). These micelived longer than LoxSOD1G37R mice, and, inparticular, the late stage of the disease wasprolonged. The role for microglia expressingSOD-1 in accelerating disease progression wasalso demonstrated by another group using a dif-ferent approach in which SOD-1 mice crossedwith PU.1 mice were transplanted with eitherwild-type or SOD-1-expressing bone marrowthat would then populate the tissues, includ-ing the brain, with either wild-type or SOD-1-expressing macrophages (2). The mice trans-planted with wild-type cells lived longer thanthose transplanted with SOD-1 bone marrow,and again it was the later stage of the disease thatappeared to be prolonged. SOD-1-expressingmicroglia generated more superoxide and nitricoxide when challenged with LPS in vitro thandid wild-type microglia. The general principle,namely that action of a mutant gene in microgliacan promote selective loss of neuron subpopula-tions, has since been validated in other diseases(133).

In both these studies, the impact of theCre-recombinase or bone marrow transplanta-tion will affect not only the microglia, but alsothe systemic macrophage populations and anymonocytes that might be recruited to the vicin-ity of the degenerating motor neurons. Thedifferential contributions of the local microgliaand of the systemic macrophages have yet to beseparately assessed and may be consequentialfor studies on prion disease, as described above.

Tumor

Primary brain tumors are primarily derivedfrom the astroglial lineage. These lesions areinfiltrative and typified by vigorous angiogen-esis; their metastatic potential is relatively low.Considerations of the roles of microglia in

brain tumor follow from a robust literatureabout tumor-associated macrophages (TAMs)(75). TAMs are considered to provide a varietyof proneoplastic functions, while residing in animmunosuppressive environment, which re-strains antitumor properties such as capacity togenerate reactive oxygen and nitrogen species.Investigators have proposed that TAMs repre-sent variant M2 macrophages that have beenexposed to immunosuppressive prostaglandins,TGF-β, and IL-10. Beyond inhibiting theirantitumor potential, these environmental cuescould stimulate TAMs to produce growthand angiogenic factors for tumor cells. Suchprinciples translate readily to an evaluation ofthe roles of microglia in primary brain tumors.Microglia were detected in brain tumorsmore than 80 years ago (134). Subsequentresearch focused on three issues: (a) Why dotumor-associated microglia appear quiescentrather than tumoricidal? (b) What microglialfactors promote tumor growth or survival? (c)How can microglia be stimulated to expresstumoricidal functions? As a generalization,immunosuppressive factors present in othertumor beds are also detected in brain tumors,prominently including TGF-β, prostaglandins,and IL-10. The origin of tumor-associated mi-croglia is uncertain. Glioma cells exuberantlyexpress CCL2, so much so that the humanprotein was first isolated from glioma culturefluid (135). CCL2 acts toward hematogenousmonocytes, but CCR2, the relevant receptor, isnot expressed on microglia. Furthermore, flowcytometric analysis of human tumor samplesrevealed infiltrating CD45hi cells, unlike themixed population of CD45hi and CD45dim

cells found in rodent glioma models (134). It isplausible therefore that tumor-infiltrating MPScells are mainly monocytic in origin, attractedthrough an impaired BBB by chemoattractantssuch as CCL2. Tumor-infiltrating microglia,stimulated by local IL-6, produce IL-10, whichsuppresses macrophage antitumor effectorfunctions and also promotes Type 2–related an-giogenic factors. Type 2 macrophage/microgliamay also support invasiveness by expressionof matrix metalloproteases. The presence of


Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.


MPS cells within brain tumors has elicitedtherapeutic endeavors, mainly aimed at stimu-lating antitumor immunity through peripheralimmunization with DCs. To present tumorantigens, DCs are pulsed with antigenicpeptides or programmed to express gliomacell cDNAs (136). One particularly interestingapproach involves use of cDNA from gliomastem cells (136). These strategies rely inpart on the APC properties of resident MPScells to restimulate tumor-specific CD4+ Tcells within the CNS. Microglia may also becalled upon to execute tumoricidal functionsfollowing stimulation by activated CD4+

T cells.

Human ImmunodeficiencyVirus (HIV)

Infection with HIV-1 imposes a vast burden ofneurological disease, exclusive of that causedby opportunistic infection and metaboliccompromise (137, 138). During early yearsof the pandemic, HIV-associated dementia(HAD) was the most feared and most promi-nent of these complications. HAD is a clinicalsyndrome comprising cognitive, affective, andmotor symptoms. With highly active antiretro-viral therapy (HAART) and prolonged survival,the incidences of HAD have lessened and thoseof subtle cognitive impairment along with thatof painful sensory neuropathy have increased.Two considerations placed microglia andinfiltrating MPS cells squarely in the center

of HAD pathogenesis: First, mononuclearphagocytes are the major infected populationin the CNS, and second, neurons, whosefunction is most evidently impaired, are verypoorly infectable and are not grossly reducedin number. Given their prevalence and severity,HAD and related conditions represent themost significant disease for which primarypathogenic pathways involve microglia.

HIV-1, like other lentiviruses (139), entersthe CNS in Trojan horse mode within traffick-ing MPS cells. In HAD tissue sections, mostcells containing viral antigens are macrophagesand microglia. In vitro studies suggest thatmicroglial physiology is disrupted by infec-tion, so that potentially neurotoxic functionsare engaged and neuroprotective responses areblunted. Microglia can elevate levels of excito-toxins such as glutamate and quinolinic acid bydiverse mechanisms, many of which seem ac-tive in HAD; combined with oxidative stress,these components may lead to neuronal struc-tural pathology and functional compromise (62,140). Microglial activation may also operatethrough production of inflammatory cytokinesto recruit astrocytes and BBB elements into theHAD process. There is evidence for unappar-ent BBB dysfunction and for the participationof astrocytes in mediating neuronal dysfunc-tion. Understanding mechanisms underlyingHAD is regarded as significant for therapeutics,for extension to other dementing illnesses, andfor addressing neuron-glial interactions duringdisease.

DISCLOSURE STATEMENT

The authors are not aware of any affiliations, memberships, funding, or financial holdings thatmight be perceived as affecting the objectivity of this review.

ACKNOWLEDGMENTS

Research in the Ransohoff lab has been supported by the U.S. National Institutes of Health, theNational Multiple Sclerosis Society, the Charles A. Dana Foundation, the Nancy Davis CenterWithout Walls, and the Williams Family Fund for Multiple Sclerosis Research. Research in thePerry lab has been supported by the Medical Research Council (UK), the Wellcome Trust, theMultiple Sclerosis Society (UK), and the European Union. We thank Dr. Natalia M. Moll andAnna Rietsch for generating images shown in the Table.


Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.


LITERATURE CITED

1. McKercher SR, Torbett BE, Anderson KL, Henkel GW, Vestal DJ, et al. 1996. Targeted disruption ofthe PU.1 gene results in multiple hematopoietic abnormalities. EMBO J. 15:5647–58

2. Beers DR, Henkel JS, Xiao Q, Zhao W, Wang J, et al. 2006. Wild-type microglia extend survival inPU.1 knockout mice with familial amyotrophic lateral sclerosis. Proc. Natl. Acad. Sci. USA 103:16021–26

3. Morgese VJ, Elliott EJ, Muller KJ. 1983. Microglial movement to sites of nerve lesion in the leechCNS. Brain Res. 272:166–70

4. Ngu EM, Sahley CL, Muller KJ. 2007. Reduced axon sprouting after treatment that diminishes mi-croglia accumulation at lesions in the leech CNS. J. Comp. Neurol. 503:101–9

5. Davalos D, Grutzendler J, Yang G, Kim JV, Zuo Y, et al. 2005. ATP mediates rapid microglial responseto local brain injury in vivo. Nat. Neurosci. 8:752–58

6. Nimmerjahn A, Kirchhoff F, Helmchen F. 2005. Resting microglial cells are highly dynamic surveillantsof brain parenchyma in vivo. Science 308:1314–18

7. Lawson LJ, Perry VH, Dri P, Gordon S. 1990. Heterogeneity in the distribution and morphology ofmicroglia in the normal adult mouse brain. Neuroscience 39:151–70

8. de Haas AH, Boddeke HW, Biber K. 2008. Region-specific expression of immunoregulatory proteinson microglia in the healthy CNS. Glia 56:888–94

9. Galea I, Palin K, Newman TA, Van Rooijen N, Perry VH, Boche D. 2005. Mannose receptor expressionspecifically reveals perivascular macrophages in normal, injured, and diseased mouse brain. Glia 49:375–84

10. Fabriek BO, Van Haastert ES, Galea I, Polfliet MM, Dopp ED, et al. 2005. CD163-positive perivascularmacrophages in the human CNS express molecules for antigen recognition and presentation. Glia51:297–305

11. Perry VH, Gordon S. 1987. Modulation of CD4 antigen on macrophages and microglia in rat brain.J. Exp. Med. 166:1138–43

12. Steinman RM, Banchereau J. 2007. Taking dendritic cells into medicine. Nature 449:419–2613. Galea I, Bechmann I, Perry VH. 2007. What is immune privilege (not)? Trends Immunol. 28:12–1814. Perry VH, Hume DA, Gordon S. 1985. Immunohistochemical localization of macrophages and mi-

croglia in the adult and developing mouse brain. Neuroscience 15:313–2615. Alliot F, Godin I, Pessac B. 1999. Microglia derive from progenitors, originating from the yolk sac, and

which proliferate in the brain. Brain Res. Dev. Brain Res. 117:145–5216. Herbomel P, Thisse B, Thisse C. 2001. Zebrafish early macrophages colonize cephalic mesenchyme

and developing brain, retina, and epidermis through a M-CSF receptor-dependent invasive process.Dev. Biol. 238:274–88

17. Lichanska AM, Browne CM, Henkel GW, Murphy KM, Ostrowski MC, et al. 1999. Differentiation ofthe mononuclear phagocyte system during mouse embryogenesis: the role of transcription factor PU.1.Blood 94:127–38

18. Davoust N, Vuaillat C, Cavillon G, Domenget C, Hatterer E, et al. 2006. Bone marrow CD34+/B220+

progenitors target the inflamed brain and display in vitro differentiation potential toward microglia.FASEB J. 20:2081–92

19. Tacke F, Randolph GJ. 2006. Migratory fate and differentiation of blood monocyte subsets. Immunobi-ology 211:609–18

20. Geissmann F, Jung S, Littman DR. 2003. Blood monocytes consist of two principal subsets with distinctmigratory properties. Immunity 19:71–82

21. Cardona AE, Pioro EP, Sasse ME, Kostenko V, Cardona SM, et al. 2006. Control of microglial neu-rotoxicity by the fractalkine receptor. Nat. Neurosci. 9:917–24

22. Lawson LJ, Perry VH, Gordon S. 1992. Turnover of resident microglia in the normal adult mousebrain. Neuroscience 48:405–15

23. Kida S, Steart PV, Zhang ET, Weller RO. 1993. Perivascular cells act as scavengers in the cerebralperivascular spaces and remain distinct from pericytes, microglia and macrophages. Acta Neuropathol.85:646–52


Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.


24. Ponomarev ED, Novikova M, Maresz K, Shriver LP, Dittel BN. 2005. Development of a culturesystem that supports adult microglial cell proliferation and maintenance in the resting state. J. Immunol.Methods. 300:32–46

25. Rosenstiel P, Lucius R, Deuschl G, Sievers J, Wilms H. 2001. From theory to therapy: implicationsfrom an in vitro model of ramified microglia. Microsc. Res. Tech. 54:18–25

26. Aloisi F. 2001. Immune function of microglia. Glia 36:165–7927. Wegiel J, Wisniewski HM, Dziewiatkowski J, Tarnawski M, Kozielski R, et al. 1998. Reduced number

and altered morphology of microglial cells in colony stimulating factor-1-deficient osteopetrotic op/opmice. Brain Res. 804:135–39

28. Boucsein C, Zacharias R, Farber K, Pavlovic S, Hanisch UK, Kettenmann H. 2003. Purinergic receptorson microglial cells: functional expression in acute brain slices and modulation of microglial activationin vitro. Eur. J. Neurosci. 17:2267–76

29. Hailer NP, Vogt C, Korf HW, Dehghani F. 2005. Interleukin-1β exacerbates and interleukin-1 receptorantagonist attenuates neuronal injury and microglial activation after excitotoxic damage in organotypichippocampal slice cultures. Eur. J. Neurosci. 21:2347–60

30. Kerr BJ, Patterson PH. 2004. Potent proinflammatory actions of leukemia inhibitory factor in the spinalcord of the adult mouse. Exp. Neurol. 188:391–407

31. Saper CB, Sawchenko PE. 2003. Magic peptides, magic antibodies: guidelines for appropriate controlsfor immunohistochemistry. J. Comp. Neurol. 465:161–63

32. Rhodes KJ, Trimmer JS. 2006. Antibodies as valuable neuroscience research tools versus reagents ofmass distraction. J. Neurosci. 26:8017–20

33. Kantarci OH, Morales Y, Ziemer PA, Hebrink DD, Mahad DJ, et al. 2005. CCR5Δ32 polymorphismeffects on CCR5 expression, patterns of immunopathology and disease course in multiple sclerosis.J. Neuroimmunol. 169:137–43

34. Svoboda K, Tank DW, Denk W. 1996. Direct measurement of coupling between dendritic spines andshafts. Science 272:716–19

35. Raivich G. 2005. Like cops on the beat: the active role of resting microglia. Trends Neurosci. 28:571–7336. Haynes SE, Hollopeter G, Yang G, Kurpius D, Dailey ME, et al. 2006. The P2Y12 receptor regulates

microglial activation by extracellular nucleotides. Nat. Neurosci. 9:1512–1937. Ting JP, Nixon DF, Weiner LP, Frelinger JA. 1983. Brain Ia antigens have a bone marrow origin.

Immunogenetics 17:295–30138. Matsumoto Y, Fujiwara M. 1987. Absence of donor-type major histocompatibility complex class I

antigen-bearing microglia in the rat central nervous system of radiation bone marrow chimeras.J. Neuroimmunol. 17:71–82

39. Hickey WF, Kimura H. 1988. Perivascular microglial cells of the CNS are bone marrow-derived andpresent antigen in vivo. Science 239:290–92

40. Greter M, Heppner FL, Lemos MP, Odermatt BM, Goebels N, et al. 2005. Dendritic cells permitimmune invasion of the CNS in an animal model of multiple sclerosis. Nat. Med. 11:328–34

41. Hickey WF, Vass K, Lassmann H. 1992. Bone marrow-derived elements in the central nervous system:an immunohistochemical and ultrastructural survey of rat chimeras. J. Neuropathol. Exp. Neurol. 51:246–56

42. Ponomarev ED, Shriver LP, Maresz K, Dittel BN. 2005. Microglial cell activation and proliferationprecedes the onset of CNS autoimmunity. J. Neurosci. Res. 81:374–89

43. Flugel A, Bradl M, Kreutzberg GW, Graeber MB. 2001. Transformation of donor-derived bone mar-row precursors into host microglia during autoimmune CNS inflammation and during the retrograderesponse to axotomy. J. Neurosci. Res. 66:74–82

44. Ajami B, Bennett J, Krieger C, Tetzlaff W, Rossi F. 2007. Local self-renewal sustains CNS microgliamaintenance and function throughout adult life. Nat. Neurosci. 10:1538–43

45. Mildner A, Schmidt H, Nitsche M, Merkler D, Hanisch U-K, et al. 2007. Microglia in the adult brainarise from Ly-6ChiCCR2+ monocytes only under defined host conditions. Nat. Neurosci. 10:1544–53

46. Ransohoff RM. 2007. Microgliosis: the questions shape the answers. Nat. Neurosci. 10:1507–947. Blinzinger K, Kreutzberg G. 1968. Displacement of synaptic terminals from regenerating motoneurons

by microglial cells. Z. Zellforsch. Mikrosk. Anat. 85:145–57


Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.


48. Graeber MB, Lopez-Redondo F, Ikoma E, Ishikawa M, Imai Y, et al. 1998. The microglia/macrophageresponse in the neonatal rat facial nucleus following axotomy. Brain Res. 813:241–53

49. Massengale M, Wagers AJ, Vogel H, Weissman IL. 2005. Hematopoietic cells maintain hematopoieticfates upon entering the brain. J. Exp. Med. 201:1579–89

50. Banati RB, Myers R, Kreutzberg GW. 1997. PK (‘peripheral benzodiazepine’)—binding sites in theCNS indicate early and discrete brain lesions: microautoradiographic detection of [3H]PK11195 bind-ing to activated microglia. J. Neurocytol. 26:77–82

51. Banati RS. 2002. Visualizing microglia activation in vivo. Glia 40:206–1752. Cagnin A, Kassiou M, Meikle SR, Banati RB. 2007. Positron emission tomography imaging of neu-

roinflammation. Neurotherapeutics 4:443–5253. Fogg DK, Sibon C, Miled C, Jung S, Aucouturier P, et al. 2006. A clonogenic bone marrow progenitor

specific for macrophages and dendritic cells. Science 311:83–8754. Gordon S, Taylor PR. 2005. Monocyte and macrophage heterogeneity. Nat. Rev. Immunol. 5:953–6455. Olson JK, Miller SD. 2004. Microglia initiate central nervous system innate and adaptive immune

responses through multiple TLRs. J. Immunol. 173:3916–2456. Piccio L, Rossi B, Scarpini E, Laudanna C, Giagulli C, et al. 2002. Molecular mechanisms involved

in lymphocyte recruitment in inflamed brain microvessels: critical roles for P-selectin glycoproteinligand-1 and heterotrimeric Gi-linked receptors. J. Immunol. 168:1940–49

57. Kivisakk P, Imitola J, Rasmussen S, Elyaman W, Zhu B, et al. 2008. Localizing central nervous systemimmune surveillance: Meningeal antigen-presenting cells activate T cells during experimental autoim-mune encephalomyelitis. Ann. Neurol. In press

58. Bo L, Mork S, Kong PA, Nyland H, Pardo CA, Trapp BD. 1994. Detection of MHC class II-antigenson macrophages and microglia, but not on astrocytes and endothelia in active multiple sclerosis lesions.J. Neuroimmunol. 51:135–46

59. Ravasi T, Wells CA, Hume DA. 2007. Systems biology of transcription control in macrophages. Bioessays29:1215–26

60. Natarajan M, Lin KM, Hsueh RC, Sternweis PC, Ranganathan R. 2006. A global analysis of cross-talkin a mammalian cellular signalling network. Nat. Cell Biol. 8:571–80

61. Cardona AE, Huang DR, Sasse ME, Ransohoff RM. 2006. Isolation of murine microglial cells for RNAanalysis for flow cytometry. Nat. Protoc. 1:1947–51

62. Reynolds AD, Glanzer JG, Kadiu I, Ricardo-Dukelow M, Chaudhuri A, et al. 2008. Nitrated alpha-synuclein-activated microglial profiling for Parkinson’s disease. J. Neurochem. 104:1504–25

63. Baker CA, Lu ZY, Zaitsev I, Manuelidis L. 1999. Microglial activation varies in different models ofCreutzfeldt-Jakob disease. J. Virol. 73:5089–97

64. Gilchrist M, Thorsson V, Li B, Rust AG, Korb M, et al. 2006. Systems biology approaches identifyATF3 as a negative regulator of Toll-like receptor 4. Nature 441:173–78

65. Duke DC, Moran LB, Turkheimer FE, Banati R, Graeber MB. 2004. Microglia in culture: What genesdo they express? Dev. Neurosci. 26:30–37

66. Moran LB, Duke DC, Turkheimer FE, Banati RB, Graeber MB. 2004. Towards a transcriptome defi-nition of microglial cells. Neurogenetics 5:95–108

67. Korb M, Rust AG, Thorsson V, Battail C, Li B, et al. 2008. The Innate Immune Database (IIDB).BMC. Immunol. 9:7

68. Adams RA, Bauer J, Flick MJ, Sikorski SL, Nuriel T, et al. 2007. The fibrin-derived γ377−395 peptideinhibits microglia activation and suppresses relapsing paralysis in central nervous system autoimmunedisease. J. Exp. Med. 204:571–82

69. Barrow AD, Trowsdale J. 2006. You say ITAM and I say ITIM, let’s call the whole thing off: the ambiguityof immunoreceptor signalling. Eur. J. Immunol. 36:1646–53

70. Hoek RM, Ruuls SR, Murphy CA, Wright GJ, Goddard R, et al. 2000. Down-regulation of themacrophage lineage through interaction with OX2 (CD200). Science 290:1768–71

71. van Beek EM, Cochrane F, Barclay AN, Van Den Berg TK. 2005. Signal regulatory proteins in theimmune system. J. Immunol. 175:7781–87


Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.


72. Takahashi K, Prinz M, Stagi M, Chechneva O, Neumann H. 2007. TREM2-transduced myeloid pre-cursors mediate nervous tissue debris clearance and facilitate recovery in an animal model of multiplesclerosis. PLoS Med. 4:e124

73. Turnbull IR, Gilfillan S, Cella M, Aoshi T, Miller M, et al. 2006. Cutting edge: TREM-2 attenuatesmacrophage activation. J. Immunol. 177:3520–24

74. Crocker PR, Paulson JC, Varki A. 2007. Siglecs and their roles in the immune system. Nat. Rev. 7:255–6675. Martinez FO, Sica A, Mantovani A, Locati M. 2008. Macrophage activation and polarization. Front

Biosci. 13:453–6176. Hanisch UK, Kettenmann H. 2007. Microglia: active sensor and versatile effector cells in the normal

and pathologic brain. Nat. Neurosci. 10:1387–9477. Pavlov VA, Tracey KJ. 2004. Neural regulators of innate immune responses and inflammation. Cell.

Mol. Life Sci. 61:2322–3178. Kaushal V, Schlichter LC. 2008. Mechanisms of microglia-mediated neurotoxicity in a new model of

the stroke penumbra. J. Neurosci. 28:2221–3079. Taylor DL, Diemel LT, Pocock JM. 2003. Activation of microglial group III metabotropic glutamate

receptors protects neurons against microglial neurotoxicity. J. Neurosci. 23:2150–6080. Koizumi S, Shigemoto-Mogami Y, Nasu-Tada K, Shinozaki Y, Ohsawa K, et al. 2007. UDP acting at

P2Y6 receptors is a mediator of microglial phagocytosis. Nature 446:1091–9581. Tsuda M, Tozaki-Saitoh H, Masuda T, Toyomitsu E, Tezuka T, et al. 2008. Lyn tyrosine kinase is

required for P2X4 receptor upregulation and neuropathic pain after peripheral nerve injury. Glia 56:50–58

82. Tsuda M, Inoue K, Salter MW. 2005. Neuropathic pain and spinal microglia: a big problem frommolecules in “small” glia. Trends Neurosci. 28:101–7

83. Tsuda M, Shigemoto-Mogami Y, Koizumi S, Mizokoshi A, Kohsaka S, et al. 2003. P2X4 receptorsinduced in spinal microglia gate tactile allodynia after nerve injury. Nature 424:778–83

84. Klunemann HH, Ridha BH, Magy L, Wherrett JR, Hemelsoet DM, et al. 2005. The genetic causes ofbasal ganglia calcification, dementia, and bone cysts: DAP12 and TREM2. Neurology 64:1502–7

85. Paloneva J, Manninen T, Christman G, Hovanes K, Mandelin J, et al. 2002. Mutations in two genesencoding different subunits of a receptor signaling complex result in an identical disease phenotype.Am. J. Hum. Genet. 71:656–62

86. Fourgeaud L, Boulanger LM. 2007. Synapse remodeling, compliments of the complement system. Cell131:1034–36

87. Stevens B, Allen NJ, Vazquez LE, Howell GR, Christopherson KS, et al. 2007. The classical complementcascade mediates CNS synapse elimination. Cell 131:1164–78

88. Cullheim S, Thams S. 2007. The microglial networks of the brain and their role in neuronal networkplasticity after lesion. Brain Res. Rev. 55:89–96

89. Streit WJ. 2001. Microglia and macrophages in the developing CNS. Neurotoxicology 22:619–2490. Hume DA, Perry VH, Gordon S. 1983. Immunohistochemical localization of a macrophage-specific

antigen in developing mouse retina: phagocytosis of dying neurons and differentiation of microglialcells to form a regular array in the plexiform layers. J. Cell Biol. 97:253–57

91. Marin-Teva JL, Dusart I, Colin C, Gervais A, van Rooijen N, Mallat M. 2004. Microglia promote thedeath of developing Purkinje cells. Neuron 41:535–47

92. Walton NM, Sutter BM, Laywell ED, Levkoff LH, Kearns SM, et al. 2006. Microglia instruct subven-tricular zone neurogenesis. Glia 54:815–25

93. Aarum J, Sandberg K, Haeberlein SL, Persson MA. 2003. Migration and differentiation of neuralprecursor cells can be directed by microglia. Proc. Natl. Acad. Sci. USA 100:15983–88

94. Choi SH, Veeraraghavalu K, Lazarov O, Marler S, Ransohoff RM, et al. 2008. Non-cell-autonomouseffects of presenilin 1 variants on enrichment-mediated hippocampal progenitor cell proliferation anddifferentiation. Neuron 59(4):568–80

95. Villeda S, Wyss-Coray T. 2008. Microglia—a wrench in the running wheel? Neuron 59(4):527–2996. Kreutzberg GW. 1996. Microglia: a sensor for pathological events in the CNS. Trends Neurosci. 19:312–

18


Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.


97. Gordon S. 2003. Alternative activation of macrophages. Nat. Rev. 3:23–3598. Perry VH, Cunningham C, Holmes C. 2007. Systemic infections and inflammation affect chronic

neurodegeneration. Nat. Rev. 7:161–6799. Martin P, Leibovich SJ. 2005. Inflammatory cells during wound repair: the good, the bad and the ugly.

Trends Cell Biol. 15:599–607100. Crutcher KA, Gendelman HE, Kipnis J, Perez-Polo JR, Perry VH, et al. 2006. Debate: “Is increasing

neuroinflammation beneficial for neural repair?” J. Neuroimmune Pharmacol. 1:195–211101. Popovich PG, Longbrake EE. 2008. Can the immune system be harnessed to repair the CNS? Nat.

Rev. Neurosci. 9:481–93102. Eichler FS, Ren JQ, Cossoy M, Rietsch AM, Nagpal S, et al. 2008. Is microglial apoptosis an early

pathogenic change in cerebral X-linked adrenoleukodystrophy? Ann. Neurol. 63(6):729–42103. Akiyama H, Barger S, Barnum S, Bradt B, Bauer J, et al. 2000. Inflammation and Alzheimer’s disease.

Neurobiol. Aging 21:383–421104. Heneka MT, O’Banion MK. 2007. Inflammatory processes in Alzheimer’s disease. J. Neuroimmunol.

184:69–91105. Miklossy J, Van Der Loos H. 1991. The long-distance effects of brain lesions: visualization of myelinated

pathways in the human brain using polarizing and fluorescence microscopy. J. Neuropathol. Exp. Neurol.50:1–15

106. Herber DL, Roth LM, Wilson D, Wilson N, Mason JE, et al. 2004. Time-dependent reduction in Aβ

levels after intracranial LPS administration in APP transgenic mice. Exp. Neurol. 190:245–53107. Simard AR, Soulet D, Gowing G, Julien JP, Rivest S. 2006. Bone marrow-derived microglia play a

critical role in restricting senile plaque formation in Alzheimer’s disease. Neuron 49:489–502108. Schenk D, Barbour R, Dunn W, Gordon G, Grajeda H, et al. 1999. Immunization with amyloid-β

attenuates Alzheimer-disease-like pathology in the PDAPP mouse. Nature 400:173–77109. Gelinas DS, DaSilva K, Fenili D, St George-Hyslop P, McLaurin J. 2004. Immunotherapy for

Alzheimer’s disease. Proc. Natl. Acad. Sci. USA 101(Suppl. 2):14657–62109a. Holmes C, Boche D, Wilkinson D, Yadegarfar G, Hopkins V, et al. 2008. Long-term effects of Aβ42

immunisation in Alzheimer’s disease: follow-up of a randomised, placebo-controlled phase I trial. Lancet372:216–23

110. Ransohoff RM. 2007. Natalizumab for multiple sclerosis. N. Engl. J. Med. 356:2622–29111. McMahon EJ, Bailey SL, Castenada CV, Waldner H, Miller SD. 2005. Epitope spreading initiates in

the CNS in two mouse models of multiple sclerosis. Nat. Med. 11:335–39112. McMahon EJ, Bailey SL, Miller SD. 2006. CNS dendritic cells: critical participants in CNS inflamma-

tion? Neurochem. Int. 49:195–203113. Galea I, Bechmann I, Perry VH. 2007. What is immune privilege (not)? Trends Immunol. 28:12–18114. Heppner FL, Greter M, Marino D, Falsig J, Raivich G, et al. 2005. Experimental autoimmune en-

cephalomyelitis repressed by microglial paralysis. Nat. Med. 11:146–52115. Serafini B, Rosicarelli B, Magliozzi R, Stigliano E, Capello E, et al. 2006. Dendritic cells in multiple scle-

rosis lesions: maturation stage, myelin uptake, and interaction with proliferating T cells. J. Neuropathol.Exp. Neurol. 65:124–41

116. Kivisakk P, Mahad DJ, Callahan MK, Sikora K, Trebst C, et al. 2004. Expression of CCR7 in multiplesclerosis: implications for CNS immunity. Ann. Neurol. 55:627–38

117. Bruck W, Porada P, Poser S, Rieckmann P, Hanefeld F, et al. 1995. Monocyte/macrophage differen-tiation in early multiple sclerosis lesions. Ann. Neurol. 38:788–96

118. Trebst C, Sorensen TL, Kivisakk P, Cathcart MK, Hesselgesser J, et al. 2001. CCR1+/CCR5+ mononu-clear phagocytes accumulate in the central nervous system of patients with multiple sclerosis. Am. J.Pathol. 159:1701–10

119. Peterson JW, Bo L, Mork S, Chang A, Trapp BD. 2001. Transected neurites, apoptotic neurons, andreduced inflammation in cortical multiple sclerosis lesions. Ann. Neurol. 50:389–400

120. Moll NM, Rietsch AM, Ransohoff AJ, Cossoy MB, Huang D, et al. 2007. Cortical demyelination inPML and MS: similarities and differences. Neurology 70:336–43

121. Kutzelnigg A, Lucchinetti CF, Stadelmann C, Bruck W, Rauschka H, et al. 2005. Cortical demyelinationand diffuse white matter injury in multiple sclerosis. Brain 128:2705–12


Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.


122. Patrikios P, Stadelmann C, Kutzelnigg A, Rauschka H, Schmidbauer M, et al. 2006. Remyelination isextensive in a subset of multiple sclerosis patients. Brain 129:3165–72

123. Stadelmann C, Kerschensteiner M, Misgeld T, Bruck W, Hohlfeld R, Lassmann H. 2002. BDNFand gp145trkB in multiple sclerosis brain lesions: neuroprotective interactions between immune andneuronal cells? Brain 125:75–85

124. Williams AE, Lawson LJ, Perry VH, Fraser H. 1994. Characterization of the microglial response inmurine scrapie. Neuropathol. Appl. Neurobiol. 20:47–55

125. Perry VH, Cunningham C, Boche D. 2002. Atypical inflammation in the central nervous system inprion disease. Curr. Opin. Neurol. 15:349–54

126. Cunningham C, Deacon R, Wells H, Boche D, Waters S, et al. 2003. Synaptic changes characterizeearly behavioural signs in the ME7 model of murine prion disease. Eur. J. Neurosci. 17:2147–55

127. Boche D, Cunningham C, Docagne F, Scott H, Perry VH. 2006. TGFβ1 regulates the inflammatoryresponse during chronic neurodegeneration. Neurobiol. Dis. 22:638–50

128. Cunningham C, Wilcockson DC, Campion S, Lunnon K, Perry VH. 2005. Central and systemicendotoxin challenges exacerbate the local inflammatory response and increase neuronal death duringchronic neurodegeneration. J. Neurosci. 25:9275–84

129. Holmes C, El-Okl M, Williams AL, Cunningham C, Wilcockson D, Perry VH. 2003. Systemic in-fection, interleukin 1β, and cognitive decline in Alzheimer’s disease. J. Neurol. Neurosurg. Psychiatry74:788–89

130. Troost D, Van Den Oord JJ, Vianney de Jong JM. 1990. Immunohistochemical characterization of theinflammatory infiltrate in amyotrophic lateral sclerosis. Neuropathol. Appl. Neurobiol. 16:401–10

131. Clement AM, Nguyen MD, Roberts EA, Garcia ML, Boillee S, et al. 2003. Wild-type nonneuronalcells extend survival of SOD1 mutant motor neurons in ALS mice. Science 302:113–17

132. Boillee S, Yamanaka K, Lobsiger CS, Copeland NG, Jenkins NA, et al. 2006. Onset and progression ininherited ALS determined by motor neurons and microglia. Science 312:1389–92

133. Giorgini F, Guidetti P, Nguyen Q, Bennett SC, Muchowski PJ. 2005. A genomic screen in yeastimplicates kynurenine 3-monooxygenase as a therapeutic target for Huntington disease. Nat. Genet.37:526–31

134. Watters JJ, Schartner JM, Badie B. 2005. Microglia function in brain tumors. J. Neurosci. Res. 81:447–55135. Yoshimura T, Robinson EA, Tanaka S, Appella E, Kuratsu J, Leonard EJ. 1989. Purification and amino

acid analysis of two human glioma-derived monocyte chemoattractants. J. Exp. Med. 169:1449–59136. Yamanaka R. 2008. Cell- and peptide-based immunotherapeutic approaches for glioma. Trends Mol.

Med. 14:228–35137. Ances BM, Ellis RJ. 2007. Dementia and neurocognitive disorders due to HIV-1 infection. Semin.

Neurol. 27:86–92138. Kopnisky KL, Bao J, Lin YW. 2007. Neurobiology of HIV, psychiatric and substance abuse comorbidity

research: workshop report. Brain Behav. Immun. 21:428–41139. Gendelman HE, Narayan O, Molineaux S, Clements JE, Ghotbi Z. 1985. Slow, persistent replication

of lentiviruses: role of tissue macrophages and macrophage precursors in bone marrow. Proc. Natl. Acad.Sci. USA 82:7086–90

140. Ciborowski P, Gendelman HE. 2006. Human immunodeficiency virus-mononuclear phagocyte inter-actions: emerging avenues of biomarker discovery, modes of viral persistence and disease pathogenesis.Curr. HIV. Res. 4:279–91


Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.

AR371-FM ARI 16 February 2009 15:37

Annual Review ofImmunology

Volume 27, 2009Contents

FrontispieceMarc Feldmann � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � x

Translating Molecular Insights in Autoimmunity into EffectiveTherapyMarc Feldmann � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �1

Structural Biology of Shared Cytokine ReceptorsXinquan Wang, Patrick Lupardus, Sherry L. LaPorte,and K. Christopher Garcia � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 29

Immunity to Respiratory VirusesJacob E. Kohlmeier and David L. Woodland � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 61

Immune Therapy for CancerMichael Dougan and Glenn Dranoff � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 83

Microglial Physiology: Unique Stimuli, Specialized ResponsesRichard M. Ransohoff and V. Hugh Perry � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �119

The Liver as a Lymphoid OrganIan Nicholas Crispe � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �147

Immune and Inflammatory Mechanisms of AtherosclerosisElena Galkina and Klaus Ley � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �165

Primary B Cell Immunodeficiencies: Comparisons and ContrastsMary Ellen Conley, A. Kerry Dobbs, Dana M. Farmer, Sebnem Kilic,Kenneth Paris, Sofia Grigoriadou, Elaine Coustan-Smith, Vanessa Howard,and Dario Campana � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �199

The Inflammasomes: Guardians of the BodyFabio Martinon, Annick Mayor, and Jürg Tschopp � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �229

Human Marginal Zone B CellsJean-Claude Weill, Sandra Weller, and Claude-Agnes Reynaud � � � � � � � � � � � � � � � � � � � � � �267

v

Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.

AR371-FM ARI 16 February 2009 15:37

AireDiane Mathis and Christophe Benoist � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �287

Regulatory Lymphocytes and Intestinal InflammationAna Izcue, Janine L. Coombes, and Fiona Powrie � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �313

The Ins and Outs of Leukocyte Integrin SignalingClare L. Abram and Clifford A. Lowell � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �339

Recent Advances in the Genetics of Autoimmune DiseasePeter K. Gregersen and Lina M. Olsson � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �363

Cell-Mediated Immune Responses in TuberculosisAndrea M. Cooper � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �393

Enhancing Immunity Through AutophagyChristian Munz � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �423

Alternative Activation of Macrophages: An Immunologic FunctionalPerspectiveFernando O. Martinez, Laura Helming, and Siamon Gordon � � � � � � � � � � � � � � � � � � � � � � � �451

IL-17 and Th17 CellsThomas Korn, Estelle Bettelli, Mohamed Oukka, and Vijay K. Kuchroo � � � � � � � � � � � � � �485

Immunological and Inflammatory Functions of the Interleukin-1FamilyCharles A. Dinarello � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �519

Regulatory T Cells in the Control of Host-Microorganism InteractionsYasmine Belkaid and Kristin Tarbell � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �551

T Cell ActivationJennifer E. Smith-Garvin, Gary A. Koretzky, and Martha S. Jordan � � � � � � � � � � � � � � �591

Horror Autoinflammaticus: The Molecular Pathophysiology ofAutoinflammatory DiseaseSeth L. Masters, Anna Simon, Ivona Aksentijevich, and Daniel L. Kastner � � � � � � � � �621

Blood Monocytes: Development, Heterogeneity, and Relationshipwith Dendritic CellsCedric Auffray, Michael H. Sieweke, and Frederic Geissmann � � � � � � � � � � � � � � � � � � � � � � � �669

Regulation and Function of NF-κB Transcription Factors in theImmune SystemSivakumar Vallabhapurapu and Michael Karin � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �693

vi Contents

Ann

u. R

ev. I

mm

unol

. 200

9.27

:119

-145

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

WIB

6402

- B

ER

EIC

H M

ED

IZIN

(C

HA

RIT

E)

DE

R o

n 01

/11/

10. F

or p

erso

nal u

se o

nly.

Documents

Microglial Physiology: Unique Stimuli, Specialized Responses