www.elsevier.com/locate/jneuroim

Journal of Neuroimmunolog

Innate immune response gene expression profiles of N9

microglia are pathogen-type specific

Clive S. McKimmie a, Douglas Roy b, Thorsten Forster b, John K. Fazakerley a,*

a Virology, Centre for Infectious Diseases College of Medicine and Veterinary Medicine, University of Edinburgh, Summerhall, Edinburgh EH9 1QH, UKb Scottish Centre for Genomic Technology and Informatics, College of Medicine and Veterinary Medicine, University of Edinburgh, UK

Received 24 January 2006; received in revised form 15 March 2006; accepted 20 March 2006

Abstract

Glial cells, particularly microglia, are thought to play a pivotal role in initiating and guiding innate immune responses to CNS infections

and in perpetuating inflammation and pathology in CNS diseases such as multiple sclerosis and Alzheimer’s disease. We describe here the

development and use of a new microarray designed to specifically profile transcript expression of innate immunity genes. Microarray analysis

validated by quantitative PCR demonstrated an extensive range of pattern recognition receptor gene expression in resting N9 microglia,

including Toll-like receptors, scavenger receptors and lectins. Stimulation with LPS or infection with virus modulated pattern recognition

receptor, cytokine, chemokine and other innate immune transcripts in a distinct and stimulus-specific manner. This study demonstrates that a

single glial cell phenotype has an innate capability to detect infection, determine its form and generate specific responses.

D 2006 Elsevier B.V. All rights reserved.

1. Introduction

The mechanisms that mediate innate immune responses

in the immunospecialised context of the CNS are gradually

starting to be understood. CNS tissue lacks a lymphatic

system and in the resting CNS, there is a marked absence of

leukocytes including lymphocytes, and dendritic cells;

MHC expression, antibodies, complement and proinflam-

matory cytokines are generally undetectable. Despite the

apparent absence of classical markers of immune surveil-

lance and response, florid inflammatory responses can be

initiated upon infection by pathogens and in response to

some autoantigens. Pathogen infection of the CNS can lead

to an influx of inflammatory cells and serum proteins,

including antibodies, which act to confine and to eliminate

the infection. Some CNS neurones are vital to life. The

majority of CNS neurones are post-mitotic and cannot be

replaced if damaged. Consequently, the CNS is particularly

susceptible to long-term damage mediated either by

pathogens or by immune responses. Infections of the CNS

0165-5728/$ - see front matter D 2006 Elsevier B.V. All rights reserved.

doi:10.1016/j.jneuroim.2006.03.012

* Corresponding author. Tel.: +44 131 650 6160; fax: +44 131 650 6511.

E-mail address: John.Fazakerley@ed.ac.uk (J.K. Fazakerley).

include important human and animal health problems;

bacterial meningitis, Japanese, West Nile, HIV and herpes

encephalitis, rabies and transmissible spongiform encepha-

lopathies.

Microglia are a resident population of cells found

throughout the CNS which play a pivotal role during CNS

immune responses. These cells remain in a resting quiescent

state until activated by injury or infection. Upon activation

microglial cells undergo hypertrophy and proliferation,

produce cytokines and change their array of cell surface

molecules. This response has often been considered to be

stereotypic and invariant. However, there may be subpopu-

lations of microglia in distinct CNS compartments which

have distinct phenotypes (Schmid et al., 2002) and different

subpopulations or all subpopulations may be capable of a

variety of responses to damage and infection. Microglia

express several key immunologically related receptors

including receptors for complement (Moller et al., 1997),

cytokines and chemokines (Hanisch, 2002), scavenger

receptors (Husemann et al., 2002) and Fc receptors (Peress

et al., 1993). Ex vivo microglia isolated from uninfected

brains express the co-stimulatory molecules B7-2, ICAM-1

and CD40 and can be induced to express high levels of

y 175 (2006) 128 – 141

C.S. McKimmie et al. / Journal of Neuroimmunology 175 (2006) 128–141 129

MHC-II molecules (Carson et al., 1998); microglia have

consequently been described as the resident antigen-

presenting cells of the CNS (Streit, 2002). Upon activation,

microglia can become active phagocytes, capable of

producing many of the same anti-microbial molecules as

activated macrophages and capable of stimulating T cells

(Mack et al., 2003; Olson and Miller, 2004). However, it

remains unclear how microglial cells are activated, whether

they are capable of discriminating between signals and

whether they are capable of making different responses to

different stimuli.

A growing consensus suggests that pattern recognition

receptors (PRRs) are key to the instigation of innate immune

responses (Akira and Takeda, 2004). PRRs recognise

distinctive molecular patterns characteristic of different types

of pathogen or altered self molecules such as oxidised low

density lipoproteins in arteriosclerosis (Gordon, 2002) and

the unique molecular patterns on the surface of apoptotic

cells (Savill et al., 2002). PRRs are a structurally diverse

grouping of receptors that include the scavenger receptors,

lectins, and Toll-like receptors. They are expressed on a

variety of leukocytes and other cells in a cell-specific manner

(Kadowaki et al., 2001; Mukhopadhyay et al., 2004). The

expression and function of these receptors on glial cells and

their role in initiating and perpetuating innate immune

responses to pathogens is an active area of research.

Microglia cultured in vitro demonstrate expression of most

TLR (McKimmie and Fazakerley, 2005), certain scavenger

receptors (Alarcon et al., 2005) and lectins such as CD14

(Bate et al., 2004). The initial response of innate immune

cells, such as microglia, is likely to be pivotal in initiating

appropriate adaptive immune responses (Napolitani et al.,

2005; Qi et al., 2003; Re and Strominger, 2001).

To dissect innate immune response gene expression

patterns in microglia at rest and upon activation, we have

designed and utilised a customised DNA microarray capable

of reporting the expression of PRRs, selected cytokines,

chemokines and stress-related transcripts. Microarray anal-

ysis has demonstrated that macrophages, dendritic cells, and

neutrophils respond to a broad range of microbial stimuli

with common transcriptional activation programmes termed

‘‘alarm signals’’ that are pathogen specific (Jenner and

Young, 2005). In this report we have used DNA microarrays

to characterise the gene expression profile of a single cloned

well-characterised microglial cell line, N9, at rest and upon

activation with bacterial and viral stimuli. Transcript levels

were assayed at rest; in response to stimulation with LPS, a

prominent component and known immune activator in many

bacterial infections; and following infection with Semliki

Forest virus (SFV) a well-characterised neuroinvasive virus

(Fazakerley, 2002). At rest, N9 cells expressed many innate

immune response genes. Upon stimulation, transcripts were

dynamically regulated in a manner specific to the microbial

stimulus, indicating that a single phenotype of microglial

cell can distinguish between pathogenic stimuli and initiate

different effector responses.

2. Results

In vivo there may be subpopulations of microglia in

distinct CNS compartments that have distinct resting and

response phenotypes (Schmid et al., 2002). In consequence,

primary cultures of microglia derived from whole brain

tissue may exhibit responses to pathogens that are the

aggregate of different microglial cell types within a culture.

To establish a system to characterise innate immune

response gene expression profiles in microglial cells and

to determine whether a single phenotype of microglial cell

can discriminate between and respond differently to diverse

pathogenic stimuli, an immortalised microglial cell line, N9,

derived from a sole glial progenitor was used in this study.

This cell line was chosen as it has been used in several

prominent studies to understand the biology of microglia

(Chen et al., 2006; Ferrari et al., 1996; McKimmie and

Fazakerley, 2005; Meda et al., 1995). To determine

expression patterns of genes involved in innate immune

responses a custom DNA microarray was first developed

and validated. The microarray was designed to assay

transcripts for most known PRRs, specific chemokines,

cytokines, interferon (IFN), stress-related and also a

collection of housekeeping genes. Control resting, LPS

stimulated or SFV infected microglial cDNAs were hybri-

dised to microarrays and gene transcript levels determined

(Forster et al., 2003). A robust quantitative PCR (QPCR)

technique was used to validate changes for selected

transcripts.

2.1. Development of a focused custom microarray to assay

innate immune gene transcripts

To rapidly and comprehensively screen a large number of

genes, a custom glass slide DNA oligonucleotide microarray

was developed. The microarray was designed principally to

characterise PRR expression. A total of 54 candidate PRRs

were assayed (see supplementary database available online

at http://www.gti.ed.ac.uk/GPX accession GPX-000057.1

and GPX-000049.1). These included Toll-like receptor

(TLR) 1 to 9 and other TLR-like proteins (such as Rp105

and Nod1), several lectins that bind microbial associated

sugars (such as CD14), secreted PRRs (such as mannose

binding lectin 1), scavenger receptors that bind both

exogenous and endogenous proteins (such as SR-AI), the

peptidoglycan recognition protein (PGRP) family, N-formyl

peptide receptors and other putative PRRs including the

Triggering Receptor Expressed on Myeloid cells (TREM)

family. Gene products that function to recognise the unique

molecular patterns of apoptotic cells, such as the phospha-

tidylserine receptor and integrins were also included. To

place PRR expression in context, transcripts for several key

genes involved in innate immune responses were also

assayed. These included pro-inflammatory cytokines (such

as TNF-a); anti-viral mediators (such as IFN-h), severalchemokines (such as CCL3/MIP-1a) and a variety of genes

C.S. McKimmie et al. / Journal of Neuroimmunology 175 (2006) 128–141130

that either mediate stress responses or are induced as a

reaction to conditions of cellular stress, such as inducible

nitric oxide synthase (iNOS), superoxide dismutase (SOD)

and cylco-oxygenases (COX). As positive controls a total of

14 housekeeping genes were assayed; these included genes

expressed at a high and at a low level (Warrington et al.,

2000). Finally, several genes with no homology to known

mammalian genes were included as negative controls.

Two 50-mer oligonucleotide microarray probes were

designed for each transcript. Probe pairs were carefully

balanced for their biophysical parameters and were selected

to be complementary to the gene of interest. Probes were

randomly printed across the array and hybridisation

efficiency was corrected using the SpotReport system

(Stratagene). cDNA samples were labelled with a single

dye, Cy3, hybridised to the microarray and the signal for

each probe pair compared. Probe signal intensities were

relatively constant within probe pairs. Where variations

were observed, transcript levels were examined by QPCR.

QPCR analysis demonstrated that array probe signals were

proportional to levels of cDNA, although in general, fold

Table 1

Expression profile of resting N9 microglia

Housekeeping PRR

h-Actin ATP-binding cassette 1 (ABC1)

Cofilin CD14

Elongation factor-1a CD36-like 2

Ferritin CD68

Guanine nucleotide binding

protein (GNAS)

Complement C1q receptor

Glycogen synthase kinase 3 hComplement receptor (CR)-1

Lactate dehydrogenase A

Complement receptor (CR)-3

Phosphatidylethanolamine

binding protein (Pbp)

Myeloid DAP12-associating

lectin-1 (MDL-1)

RAS-homolog enriched

in brain (Rheb)

Mincle (Clecsf9)

Tuba1

Formyl-peptide receptor

Vimentim

HM74 (G protein-coupled receptor 109B)

Y box protein 1

Integrin-h5

Zeta proteasome

Laminin receptor

Oxidised low density lipoprotein

(lectin-like) receptor 1 (Lox-1)

Mannose receptor 1

Peptidoglycan receptor protein (PGRP)

Phosphatidylserine receptor (PSR)

Sialoadhesin

Scavenger receptor (SR)-AI

Scavenger receptor (SR)-BI

Scavenger receptor (SR)-CL

Toll-like receptor (TLR) 1

Toll-like receptor (TLR) 2

Toll-like receptor (TLR) 4

Toll-like receptor (TLR) 7

Toll-like receptor (TLR) 9

Triggering receptor expressed on

myeloid cells (TREM) 2

Four cultures of microglia were grown in parallel for 48 h until semi-confluent, c

signal intensity from probes designed to be complementary to transcripts of non

hybridisation. The threshold for a positive signal was set as the mean plus two sta

probes that generated a signal higher than the threshold value in at least three of

changes in transcript levels were underestimated by the

array platform compared to QPCR.

2.2. Resting N9 microglia express a multitude of genes

involved in innate immune responses

To determine the expression profile of resting N9

microglia, four parallel cultures were grown for 48 h until

semi-confluent. RNA was extracted and cDNAs from

individual cultures were hybridised to replicate microarrays.

All probes that generated a signal greater than the detection

threshold are listed in Table 1. These results demonstrate

that resting N9 microglia expressed all housekeeping genes

assayed with the exception of c-yes, a gene previously

shown to be lowly expressed in some tissue types

(Warrington et al., 2000). Transcripts for typical markers

of microglia/macrophages such as CD14, CD68, comple-

ment receptors 1 and 3, and scavenger receptors (SR-AI and

SR-BI) were detected in resting N9 microglia. Gene

transcripts for the atypical scavenger receptor SR-CL,

formyl-peptide receptor, HM74, laminin receptor, peptido-

Stress Cytokine/chemokine

Catalase CCL3 (MIP-1a)

Cathepsin D CCL4 (MIP-1h)Cyclooxygenase-1 CCL21b (SLC)

Cyclooxygenase-2 CXCL10 (IP10)

Glutathione peroxidase 1 (GPX1) Fas ligand

Glutathione reductase 1 (GSR) Fibroblast growth factor 2

Glutathione synthetase (GSS) Glial maturation factor (GMF)-hGlutathione S-transferase

Pi (GST-Pi)

IL-1h

Heme oxygenase

Transforming growth

factor (TGF)-hHeat shock protein 25 Tumour Necrosis factor (TNF)-a

Neutrophil cytosolic factor (NCF)-1

Neutrophil cytosolic factor (NCF)-2

Oxidative stress induced (Osi)

Prion protein (PrP)

Superoxide dismutase (SOD)-1

Superoxide dismutase (SOD)-2

DNA prepared from each was hybridised to four separate microarrays. The

-animal origin (negative control probes) indicated the level of nonspecific

ndard deviations of the average signal of 36 negative control probes. Those

the four parallel cultures are shown here.

CR

3

LB

P

Mrc

1

Mrc

1

SR

B-I

SR

B-I

TL

R4

TR

EM

-2

TR

EM

-2

Cat

alas

e

Cat

alas

e

Min

cle

Min

cle

Fas

L

GS

R

Hem

eOx.

Hem

eOx.

IL-1

α

IL-1

α

IL-1

β

MA

RC

O

NC

F1

Osi

Osi

CC

L3

CC

L3

CC

L4

CC

L4

CX

CL

10

CX

CL

10

SO

D-2

SO

D-2

TN

F-α

TN

F-α

10

11

12

13

14

15

16

Downregulated Upregulated

Lo

g2

(pro

be

sig

nal

)

(A)C

R3

LB

P

Mrc

1

Mrc

1

SR

B-I

SR

B-I

TL

R4

TR

EM

-2

TR

EM

-2

Cat

alas

e

Cat

alas

e

Min

cle

Min

cle

Fas

L

GS

R

Hem

eOx.

Hem

eOx.

IL-1

α

IL-1

α

IL-1

β

MA

RC

O

NC

F1

Osi

Osi

CC

L3

CC

L3

CC

L4

CC

L4

CX

CL

10

CX

CL

10

SO

D-2

SO

D-2

TN

F-α

TN

F-α

0.25

0.5

1

2

4

8

16

PBSLPS

Downregulated Upregulated

Fo

ld c

han

ge

(pro

be

sig

nal

)

(B)

Fig. 1. LPS stimulation altered the expression of several gene transcripts. Parallel cultures of N9 microglia were grown until semi-confluent and then treated

with either LPS in PBS or PBS alone for 18 h. In parallel, cDNAs prepared from these cultures (PBS, n =3; LPS, n =5) were hybridised to separate microarrays

and the data sets normalised. Panel A demonstrates changes in probe signal intensities. As each probe differs in avidity for its target, values cannot be compared

between probes; however for any one probe, changes in intensity between samples can be compared. Each point represents the mean probe signalTS.E.M.,

g=PBS, ?=LPS. Changes in signal intensity can be clearly seen between samples from resting and LPS stimulated microglial cells. Only those probes that

generated a significantly (Student’s t-test, p <0.05) different signal in LPS stimulated cells compared to PBS treated cells are shown. Probes that gave a

significantly decreased mean signal in response to LPS are plotted on the left (probes for CR3 to TREM-2); probes that gave a significantly increased mean

signal in response to LPS are plotted on the right (probes for catalase to TNF-a). To facilitate comparisons between probes, panel B shows the same data plotted

as fold-change. For each probe, the mean value for the replicate LPS samples is shown relative to the mean value for the replicate unstimulated cells. The values

for the unstimulated cells are also plotted in order to show the variability in these replicate samples. Bars represent mean of fold changeTS.E.M.

C.S. McKimmie et al. / Journal of Neuroimmunology 175 (2006) 128–141 131

IL-1α IL-1β0.1

1

10

100

1000

0 Hour 18 hour LPS

Fo

ld c

han

ge

Fig. 2. QPCR for IL-1. The LPS stimulated N9 microglia RNA samples that

generated the cDNA used for the microarray hybridisations were also used

to generate full-length cDNA for validation purposes. The transcript levels

for IL-1a and IL-1h were measured using QPCR and plotted as fold-change

relative to PBS stimulated cultures. Each bar represents the mean of four

replicatesTS.E.M.

β-actin GAPDH SFV

108

107

106

105

104

103

102

0 Hours 3 Hours 12 Hours24 Hours

Ab

solu

te c

op

ies

tran

scri

pt

/ PC

R r

eact

ion

Fig. 3. Housekeeping gene transcripts were not significantly altered during

the course of SFV infection. Absolute copies of h-actin, GAPDH and SFV

transcripts were determined by QPCR at 3, 12 and 24 h after N9 infection

with SFV at a multiplicity of infection of 0.1. Bars show the mean of 4

replicatesTS.E.M. The dotted line indicates the level of detection of the

assay.

C.S. McKimmie et al. / Journal of Neuroimmunology 175 (2006) 128–141132

glycan receptor protein (PGRP), the lectins mincle and

myeloid DAP-12 associated lectin-1 (MDL-1), TLR 2, 4, 7

and 9 and TREM-2 were also expressed as were TGF-h,CCL3 (MIP-1a) and CCL4 (MIP-1h), and genes associated

with phagocytic responses such as neutrophil cytosolic

factor (NCF)-1 and NCF-2 (part of the NAPDH-oxidase

system that generates toxic free radicals), cathepsin D (a

lysosomal protease) and COX-1 and COX-2 that mediate

prostaglandin synthesis. Finally, basal levels of transcripts

associated with stress responses were also detectable, such

as Osi (also known as sequestosome 1), superoxide

dismutase (SOD)-1 and SOD-2 (scavengers of free radicals),

genes involved in glutathione synthesis (such as glutathione

synthetase (GSS), which acts to protect against free

radicals), heat shock protein 25 and heme oxygenase 1

(Table 1).

2.3. LPS initiated changes in N9 microglial transcript levels

To investigate N9 microglia responses to a bacterial

stimulus, cells were stimulated with 100 ng/ml LPS for 18 h.

Significant PRR transcript changes occur at this time point

in similarly stimulated monocytes (Zarember and God-

owski, 2002). Eight parallel cultures of N9 microglia were

grown until semi-confluent and were then treated with either

LPS in PBS or PBS alone for 18 h before collecting the

RNA. The resultant cDNAs were analysed using the

microarray platform (Fig. 1). LPS significantly (Student’s

t-test, p<0.05) upregulated a series of cytokine transcripts

including TNF-a, IL-1a and IL-1h, and the chemokines

CCL3, CCL4 and CXCL10 (IP10). Unexpectedly, there was

no significant increase in IL-6 transcripts. In a previous

study of LPS stimulation of N9 cells, we observed a

significant increase in IL-6 transcripts at 3 h and by

quantitative PCR (McKimmie and Fazakerley, 2005). This

difference most likely reflects a combination of the different

time points studied and the differential sensitivities of these

two assays. Several transcripts associated with stress were

significantly upregulated; the free-radical scavenger SOD-2

exhibited the largest fold-increase. Significant increases

were also seen for heme oxygenase, catalase, glutathione

reductase 1, Osi, the free radical generator NCF1, the lectin

mincle and in particular the scavenger receptor MARCO.

Several transcripts encoding cellular receptors were signif-

icantly (Student’s t-test, p <0.05) downregulated in LPS

stimulated cultures; these included, TREM-2, TLR 4,

scavenger receptor B-I and mannose receptor 1.

To verify the microarray results, selected gene transcripts

were quantified by real time quantitative (Q) PCR. The

array contained two probes for each gene, both probes

complementary to IL-1a reported increased levels of tran-

scripts, whilst only one of the probes for IL-1h demon-

strated significantly elevated levels in LPS stimulated

samples. QPCR analysis confirmed significant upregulation

of both IL-1a and IL-1h indicating that the microarray

correctly reported the increased expression of IL-1a and that

IL-h was also upregulated (Fig. 2). Readings from the probe

reporting no change for IL-1h and by similar analysis a

small number of other ‘‘null’’ probes on the array were

eliminated from this and all subsequent analyses. As

demonstrated by comparison of Figs. 1 and 2, the fold-

change reported by the array is an underestimate. Relative to

more sensitive techniques such as QPCR, microarrays

frequently underestimate fold changes in gene expression

(Kane et al., 2000; Reverter et al., 2005).

2.4. Virus infection triggered a different response to LPS

stimulation

Semliki Forest virus (SFV) is a neuroinvasive virus that

initiates encephalitis and activates glial cells in vivo (Amor et

al., 1996; Fazakerley, 2002). Parallel cultures of N9 micro-

glia were grown until semi-confluent, infected with SFVand

C1q

R

CD

68

CD

68

CR

1

GM

F-β

GP

X1

GS

T-P

i

GS

T-P

i

SR

B-I

Pb

p

Cat

alas

e

Cav

eolin

CD

40

CD

40

HM

74

HM

74

IFN

-α

IFN

-β

IFN

-β

IL-6

No

s2

PS

R

CC

L4

CC

L4

CX

CL

10

CX

CL

10

SR

A-I

TL

R2

TL

R2

TL

R3

TL

R3

TN

F-α

0.25

0.5

1

2

4

8

16

32

64

Downregulated Upregulated 0 hour12 hour

Fo

ld c

han

ge

(pro

be

sig

nal

)

Fig. 4. SFV infection altered the expression of several gene transcripts. Parallel cultures (n =8) of N9 microglia were grown until semi-confluent. Half were

infected with SFV4 for 12 h and half ‘‘mock infected’’ (PBS alone). cDNA were hybridised to microarrays and the data normalised. Only those probes that

generated a significantly different (Student’s t-test, p <0.05) signal in SFV infected samples compared to mock infected are shown. Those probes that gave a

significantly decreased mean signal are plotted on the left (probes for C1qR to Pbp), whilst those probes that gave a significantly increased mean signal are

plotted on the right (probes for catalase to TNF-a). Bars represent mean fold-change of four replicates in probe signal relative to 0 hTS.E.M.

C.S. McKimmie et al. / Journal of Neuroimmunology 175 (2006) 128–141 133

harvested for RNA at 0, 3, 12 and 24 h post-infection and the

resultant cDNAs hybridised to microarrays. To mimic

natural infection in vivo, cells were infected with a low

dose of SFV, a multiplicity of infection (MOI) of 0.1, so that

relatively few cells in the culture would be infected. It has

been reported that virus infection with a high MOI can

decrease levels of housekeeping gene transcripts rendering

C1q

R

GS

T-P

i

SR

B-I

TG

F-ββ

IFN

-α

IFN

-β

IL-6

CD

14

0.25

0.5

1

2

4

8

16

32

** * ** *

*

**

* *

0 Hours 12

Fo

ld c

han

ge

(pro

be

sig

nal

)

Fig. 5. Innate immune gene transcripts were altered at 12 and 24 h post-infection. F

0-h time pointTS.E.M. A single probe is shown for each gene. Bars marked with a

from the 0-h time point (Student’s t-test, p <0.05). All cultures (n =12) were gro

them inappropriate for normalisation (Nystrom et al., 2004).

With the exception of phosphatidylethanolamine binding

protein (Pbp) that was marginally downregulated, levels of

housekeeping transcripts were unaffected by the level of

infection used here. Further analysis of housekeeping tran-

scripts h-actin and GAPDH by QPCR confirmed levels were

not significantly altered by this infection (Fig. 3). This is in

CD

36

CC

L3

CC

L4

CX

CL

10

Min

cle

SO

D-2

SR

-AI

TL

R2

TL

R3

**

***

*

*

* *

*

*

*

Hours 24 Hours

*

old change in probe signal at 12 and 24 h post-SFV-infection, relative to the

n asterisk indicate that the probe signal intensity was significantly different

wn in parallel.

C.S. McKimmie et al. / Journal of Neuroimmunology 175 (2006) 128–141134

agreement with other studies in which infection of glial

cultures with virus at a low MOI had a minimal effect on

housekeeping transcript levels (Rempel et al., 2005). SFV is

an RNA virus that replicates rapidly in vitro, with new

virions released approximately 8 h post-infection. SFV RNA

was detected by QPCR at 3, 12 and 24 h post-infection, with

levels peaking at 12 h (Fig. 3).

Microarray analysis demonstrated that at 3 h post-infection

innate immune gene transcript levels were not significantly

different from mock-infected controls (data not shown).

However, substantial changes were observed by 12 h (Fig.

4). Significantly (Student’s t-test, p <0.05) downregulated

transcripts included the complement receptors C1qR and

CR1, scavenger receptors CD68 and SR-BI and glutathione

S-transferase Pi (GST-Pi, involved in glutathione synthesis).

Several transcripts were upregulated including some well

known tomediate responses to viruses. These included IFN-a

and IFN-h, IL-6, chemokines CCL4 and CXCL10, scavenger

receptor A-I, HM74, TLR 2 and 3, CD40 and iNOS. There

was no marked increase in the expression of pro-inflamma-

tory cytokine gene transcripts IL-1a, IL-1h or TNF-a, or in

genes associated with cellular stress such as heat shock

proteins, Osi or SOD-2. The greatest fold increases were

observed for the chemokines CCL4 and CXCL4 followed by

IFN-a.

At 24 h post-infection (Fig. 5) several PRR transcripts

were upregulated including TLR 2 and scavenger receptor A,

CD14 and CD36. Those chemokines which were elevated at

12 h remained upregulated at 24 h and in addition CCL3 was

significantly upregulated. The highest fold increases were

again observed with the chemokines CCL4 and CXCL10.

The free-radical scavenger SOD-2 also gave a marginal but

significantly higher signal at 24 h. In contrast, the

immunosuppressive cytokine TGF-h was downregulated.

All transcripts with decreased signal at 12 h remained

depressed at 24 h. Transcripts for type-I IFN and IL-6 were

decreased at 24 h relative to their levels at 12 h.

Levels of selected transcript were studied further using

QPCR (Fig. 6). The microarray results indicated a marginal

IL-6 IFN-α TLR2 TLR30.5

1

2

4

8

16

32

64

128

256

512

12Hr 24Hr0Hr

Fo

ld c

han

ge

Fig. 6. Selective QPCR analysis of SFV4 infected N9 cDNA confirms gene

transcript upregulation. Transcripts for selected genes were assayed by

QPCR and expressed as fold-increase relative to the 0-h time point. Each

bar represents the mean of four replicatesTS.E.M.

increase in IL-6 transcripts at 12 h post-infection, whilst

IFN-a levels were clearly increased at 12 h. QPCR analysis

confirmed this pattern of upregulation. The magnitude of

fold-increase for IFN-a at 12 h was substantially higher by

QPCR than by array. QPCR also validated the increase in

TLR 2 at 12 and 24 h and also demonstrated TLR 3

transcripts were significantly upregulated at both time

points.

3. Discussion

The study utilised a custom microarray to describe the

profile of innate immune gene expression in a single

phenotype of microglial cell at rest and the changes in

response to two distinct pathogenic stimuli, LPS and an

RNA virus infection. Resting N9 microglia expressed a

broad range of PRRs including TLRs, scavenger receptors

and lectins. Several gene transcripts not previously associ-

ated with microglia were shown to be constitutively

expressed these included, myeloid DAP-12-associated lectin

(MDL-1), mincle, HM74, PGRP and scavenger receptor

with C-type lectin (SRCL). MDL-1 and mincle are lectins,

sugar-binding proteins that can mediate both pathogen

recognition and cell–cell interactions. MDL-1, also known

as Clecsf 5, is expressed in peripheral blood monocytes and

macrophages and together with DAP-12 may signal to

initiate an oxidative burst, cytokine production and induc-

tion of co-stimulatory molecules (Bakker et al., 1999).

Mincle, also known as Clecsf 9, is strongly induced in

murine peritoneal macrophages in response to several

inflammatory stimuli such as LPS and TNF-a and may

play a role in the function of macrophages infiltrating sites

of inflammation (Matsumoto et al., 1999). The LPS

stimulated upregulation of mincle and MDL-1 reported

here in N9 cells suggests these lectins have similar functions

in microglia. HM74, also known as PUMA-G in mice, is a

novel chemokine receptor best known as a nicotinic receptor

in adipose tissue. Its potential role in the CNS remains to be

elucidated (Tunaru et al., 2003). PGRP, also known as

(PGRP)-S is a member of a family of pattern recognition

molecules conserved from insects to mammals that bind to

bacteria and their peptidoglycan components. Mice deficient

in PGRP have an increased susceptibility to infection with

some gram-positive bacteria and possess neutrophils that

exhibit defective intracellular killing of pathogens (Dziarski

et al., 2003; Dziarski, 2004). SRCL is a novel member of

the scavenger receptor family expressed in the macrophage

cell line J774A.1 and can bind gram-negative bacteria when

transfected into CHO-K1 cells (Nakamura et al., 2001).

TLR 2, 4, 7 and 9 have been previously reported in both N9

microglia (McKimmie and Fazakerley, 2005) and primary

microglial cultures (Olson and Miller, 2004). N9 microglia

thus express a wide variety of PRRs, some of which have

not been previously associated with this cell type. In

addition, a variety of PRRs previously associated with

C.S. McKimmie et al. / Journal of Neuroimmunology 175 (2006) 128–141 135

microglia were detected in resting N9 microglia including;

mannose receptor 1; scavenger receptor-AI (SR-AI), scav-

enger receptor-BI (SR-BI), triggering receptor expressed on

myeloid cells (TREM)-2 and CD14. Taken together, these

results suggest that N9 microglia have the capacity to detect

the presence of a considerable range of pathogen-associated

molecules as well as endogenous altered or oxidised

proteins. The expression of COX-1 and COX-2 suggests

these cells have an innate capability to generate pro-

inflammatory prostaglandins. If replicated in vivo, the broad

expression of innate immune transcripts on glial cells may

provide a potent pathogen sensing system for a tissue that at

rest is otherwise devoid of recognisable immune processes.

Before considering the gene expression changes induced

by LPS or virus infection it is worth noting that a change in

gene transcripts in a population of cells, even a cell line,

does not imply that all cells alter their transcripts. In the case

of LPS all cells should have received the same stimulus at

the same time, but this would not have been the case with

virus infection. With virus (MOI=0.1), some cells would

have been infected within the first few hours and additional

cells several hours later. Uninfected cells could have

responded to soluble factors (e.g. IFN) or cell surface

changes produced by infected cells. The response of the

population of cells to SFV infection must therefore be

LPS Binding proteinTriggering receptor Complement C1q reComplement receptComplement receptMannose receptor 1HM74Phosphatidylserine CD68Scavenger receptorScavenger receptorMARCOMincleToll-like receptor 2Toll-like receptor 3Toll-like receptor 4Fas LigandCD40IFN-αIFN-βIL-1 αIL-1 βCCL3CCL4CXCL10TNF-αNitric oxidase synthNeutrophilc cytosoliCatalseCaveolinGlutathione reductaHeme oxygenaseOxidative stress indSuperxide dismutasGlutathione S-transGlial maturation facEongation factor 1αLactate dehydrogen

LPS SFV

Fig. 7. N9 microglia have an innate capability to generate different transcript

significantly different signal upon LPS or RNA virus stimulation relative to unstim

cells, blue indicates downregulation, white no change and red upregulation. The

inflammatory genes and stress-associated genes. Also included are the housekeepi

these is not affected by either stimulus.

interpreted as the amalgamation of all these responses, early

and late, direct and indirect. The issue could clearly be

dissected further; however, our aim was to determine

whether this single phenotype of microglial cells would

make differential responses to bacterial and viral stimuli.

LPS stimulation and SFV infection invoked qualitatively

and quantitatively different changes in gene expression.

This is illustrated in Fig. 7. Importantly, these changes were

relevant to the stimulus: LPS invoked substantial pro-

inflammatory cytokine, bactericidal and stress-associated

gene expression and downregulation of the TREM-2 gene;

whilst SFV infection induced a strong type-I IFN response

and downregulation of the TGF-h gene. Both stimuli

induced chemokine transcripts, CCL3, CCL4 and CXCL10;

these are well-characterised potent attractants of monocytes

and other leukocytes that in vivo would actively recruit a

CNS inflammatory response. Considering these changes in

more detail, LPS also induced TNF-a and IL-1a/h,proinflammatory cytokines that activate leukocytes but

which can also be neurotoxic and which may contribute to

neuronal damage during inflammatory responses of the

brain (Kielian, 2004; Streit, 2002; Wang et al., 2004). These

cytokines represent a group of genes that are strongly and

consistently upregulated during inflammatory responses in

numerous cell types (Jenner and Young, 2005). In contrast,

expressed on myeloid cells (TREM)-2ceptoror 1or 3

receptor

-AI-BI

ase 2c factor 1

se 1

uced (osi)e-2ferase pi (GST-Pi)tor-β

ase A

Pattern

Recognition

Receptors

Immune /

Inflammatory

Stress

Glial growth factor

Housekeeping

ional responses specific to different stimuli. Those probes that reported

ulated cells are shown here as a ‘‘heat’’ diagram. Relative to unstimulated

probes have been clustered into functional relationships; PRRs, immune/

ng genes, elongation factor-1a and lactate dehydrogenase A; expression of

C.S. McKimmie et al. / Journal of Neuroimmunology 175 (2006) 128–141136

during SFV infection, N9 microglia exhibited a less

pronounced pro-inflammatory response, with no significant

increase in TNF-a, or IL-1a/h; however, gene expression of

the anti-inflammatory cytokine TGF-h was reduced. TGF-his constitutively expressed within the resting CNS and

prevents induction of expression of chemokines and other

genes involved in cell migration (Paglinawan et al., 2003).

The large increases in chemokine gene expression observed

after virus infection are consistent with the observed

downregulation of TGF-h. Virus infection induced large

fold-increases in the levels of type-I IFN transcripts; these

exert powerful anti-viral effects and are a classic response to

virus infection (Muller et al., 1994). Transcripts for iNOS

were also uniquely upregulated by virus. iNOS mediates the

production of anti-microbial nitric oxide. iNOS is upregu-

lated by microglia derived from embryonic CNS cultures

following infection with Theiler’s virus, another RNA virus,

or stimulation with LPS, peptidoglycan and CpG DNA

(Olson and Miller, 2004).

Pathogen responses in leukocyte populations have been

shown to be dependent on a set of PRRs, in particular the

TLRs. Whilst different PRRs upregulate many of the same

genes, PRRs also mediate pathogen-specific responses

(Jenner and Young, 2005). For example, in dendritic cells,

Escherichia coli upregulates a cluster of inflammatory

cytokine genes that are only weakly induced by infection

with influenza virus (Huang et al., 2001). These responses

observed in professional antigen presenting cells appear to

mirror the apparent pathogen-specific changes in microglial

transcripts observed here.

LPS upregulated several heat shock protein transcripts

and increased transcript levels for NCF1 (also known as

p40phox). The NCF1 gene product activates the NADPH

oxidase system that generates superoxide anions, a precur-

sor of microbicidal oxidants (Tsunawaki et al., 1996).

Perhaps in a coordinated gene response, or in response to

an increased production of free radicals, N9 cells upregu-

lated several transcripts such as SOD-2 and catalase that

counteract free radical damage. SOD-2 acts to convert

oxygen free radicals to less reactive compounds (Zelko et

al., 2002), whilst catalase plays a major role in cellular

antioxidant defence (Ho et al., 2004). With the exception of

a slight increase in SOD-2 at 24 h, SFV infection did not

upregulate these stress-associated gene transcripts.

SFV but not LPS upregulated the CD40 gene. Expression

of CD40 has been reported in microglia in inflammatory

lesions of CNS demyelination in multiple sclerosis and in

Alzheimer’s disease. Importantly, ablation of the CD40

ligand (CD40L) or pharmacological intervention amelio-

rates pathology or attenuates progression in models of both

diseases (Howard et al., 1999; Tan et al., 2002). Interaction

of CD40 on microglia with CD40L, expressed on CD4+ T-

cells, may enhance inflammation by stimulating production

of cytokines, chemokines, prostaglandins and nitric oxide

(D’Aversa et al., 2002; Jana et al., 2001; Okuno et al.,

2004). Upregulation of microglial CD40 in response to virus

infection may represent a mechanism by which activated

microglia amplify inflammation in the presence of T-cells.

Theiler’s virus infection also upregulates microglial CD40

(Olson et al., 2001).

Stimulation with LPS and infection with SFV both

altered the expression of several key scavenger receptor

(SR) genes. SRs are defined by their ability to bind

chemically modified low density lipoprotein but can also

bind a variety of other ligands (Franc et al., 1999; Krieger

and Stern, 2001). Reports suggest that expression of SR-AI

is restricted to mononuclear phagocytes at sites of immune

surveillance, such as the spleen (Yamada et al., 1998). SR-

AI is also expressed in microglia in the embryonic brain, but

expression is terminated as the brain matures (El Khoury et

al., 1996; Honda et al., 1998). SR-AI can recognise a variety

of ligands such as modified lipoproteins, apoptotic cells and

gram-negative bacteria (Husemann et al., 2002). Following

LPS exposure, SR-AI can induce the production of pro-

inflammatory cytokines and acts to clear LPS in vivo

(Ashkenas et al., 1993; Hampton et al., 1991). The

upregulation of the SR-AI gene in LPS treated N9 microglia

is in agreement with previous studies in primary rat

microglia (Grewal et al., 1997). The upregulation of SR-

AI during SFV infection is a novel finding and may

represent a response to apoptotic cells. SFV is known to

induce apoptosis in many continuously cultured cell lines

(Scallan et al., 1997). The phosphatidylserine receptor

(PSR), which recognises phosphatidylserine on the surface

of apoptotic cells, was also upregulated.

SR-BI is expressed on macrophages and endothelia in

cholesterol-rich tissues and participates in reverse choles-

terol transport to the liver (Hirano et al., 1999; Krieger,

1999). Like SR-AI, SR-BI appears to be developmentally

regulated; being expressed at high levels in microglia of the

immature brain and at low levels in the adult brain

(Husemann et al., 2001). The function of this receptor in

the brain is yet to be fully elucidated, however it has a

putative function in lipid metabolism and the clearance of

apoptotic cells and denatured proteins (Husemann et al.,

2002). SR-BI is downregulated in macrophages following

exposure to TNF-a, although there is no work reporting

transcript levels in microglia (Buechler et al., 1999). The

results here suggest that either the LPS stimulus or

subsequent TNF-a production acts to downregulate this

receptor. However, since SR-BI downregulation also oc-

curred during SFV infection, in the absence of LPS or

significant TNF-a production, molecules other than these

may also modulate its expression. The expression of both

SR-AI and SR-BI in resting cultures of N9 microglia further

highlights the likely embryonic origin of this cell line.

CD36 is another SR related to SR-BI that is expressed in

both macrophages and vascular endothelium (Febbraio et

al., 2001). CD36 is expressed in human microglia in vivo

and on N9 microglia (Coraci et al., 2002). This receptor has

the capacity to recognise h-amyloid and initiate the

production of reactive oxygen species, and may play a role

C.S. McKimmie et al. / Journal of Neuroimmunology 175 (2006) 128–141 137

in the pathogenesis of Alzheimer’s disease (Coraci et al.,

2002; El Khoury et al., 2003). The finding that the CD36

gene is highly upregulated in N9 microglia following virus

infection is novel.

MARCO, also known as macrophage receptor with

collagenous structure, is a scavenger receptor expressed in

a limited subset of macrophages and in granulocyte-

macrophage-colony stimulating factor (GM-CSF)-treated

microglia. LPS stimulation can induce MARCO expression

in macrophage populations and in dendritic cells (Palecanda

et al., 1999; van der Laan et al., 1999). N9 microglia

dramatically upregulated this receptor upon LPS exposure

but not SFV infection.

TREM-2 is expressed on monocyte-derived dendritic

cells and upon activation dramatically increases the expres-

sion of molecules associated with antigen presentation to T-

cells (Bouchon et al., 2001). Whilst TREM-2 expression is

lacking on monocytes and macrophages, the receptor was

among the few mRNAs identified as being expressed by

resting microglia but downregulated by LPS. Thus, TREM-

2 is considered a marker of resting microglia in vivo, and

brain regions with incomplete blood–brain barrier exhibit

reduced expression. Indeed, regional variations in TREM-2

CNS expression may contribute to the varying sensitivities

of different brain regions to pathogenic signals (Schmid et

al., 2002). The substantial expression of TREM-2 in resting

N9 cultures further underscores their phenotype as a resting

microglial cell line. TREM-2 was downregulated in LPS

stimulated N9 microglia, but was not significantly altered

during SFV infection. This indicates TREM-2 expression is

downregulated in a pathogen specific fashion and is not

indicative of microglial activation per se.

The gene expression profile of resting N9 cells suggests

they possess an embryonic (e.g. SR-AI), partially inacti-

vated (e.g. TREM-2) phenotype equivalent to microglia

found in the developing brain. Expression of several genes

not previously associated with N9 cells has been observed

and serves to further characterise these cells. LPS stimu-

lation generated a substantial pro-inflammatory cytokine

and chemokine response with the induction of several

bactericidal and stress-associated genes, whilst SFV infec-

tion induced substantial increases in type-I IFN and

chemokines. It has been suggested that subpopulations of

microglia derived from separate CNS compartments may

differ in their response to infection (Carson, 2002; Schmid

et al., 2002). Primary microglia cultures derived from

whole brain tissue may therefore exhibit divergent

responses to pathogens due to their heterogeneity. The

present study, taken together with our recent observation

that N9 cells dynamically and differentially alter TLR

expression in response to LPS stimulation and SFV

infection (McKimmie et al., 2005; McKimmie and Faza-

kerley, 2005), indicate that a single microglia cell pheno-

type has an innate capability to detect infection, determine

its form and generate responses that are specific and

relevant to the pathogenic stimulus.

4. Materials and methods

4.1. Cell culture and reagents

N9 microglia were a kind gift from Dr. Alun Williams

(Institute of Comparative Medicine, University of Glasgow,

UK). N9 cells express typical markers of resting mouse

microglia and have been extensively used as representative

of mouse microglial cells (Ferrari et al., 1996; McKimmie

and Fazakerley, 2005; Righi et al., 1989). N9 cells were

grown in RPMI supplemented with 5% heat-inactivated

foetal calf serum, 2 mM glutamine, 100 units/ml penicillin

and 100 Ag/ml streptomycin at 37 -C and 5% CO2.

4.2. Stimulation of cells with LPS and infection with SFV4

Cells were seeded into T25 flasks (Corning) and allowed

to grow for at least 48 h until 80% confluent. Cells were

then stimulated either by the addition of LPS (E. coli

O55:B5, Sigma) 100 ng/ml or by infection with SFV4 at a

multiplicity of infection (MOI) of 0.1. As controls, parallel

cultures were grown and treated with an equal volume PBS

in place of pathogenic stimuli. All cells were grown in

parallel and had undergone 5 to 10 passages post thawing so

that the length of culture prior to stimulation was minimal.

At specific times post stimulation, cells were lysed and

RNA extracted using RNeasy columns (Qiagen) with on

column DNAse digestion as per the manufacturer’s instruc-

tions.

4.3. Microarray design and fabrication

Customised DNA arrays were fabricated by The Scottish

Centre for Genomic Technology and Informatics (ScGTI),

University of Edinburgh, UK. Two 50-mer oligonucleotide

probes were designed for each transcript using Oligo6

software. From the list of probes suggested by the software

for an individual gene, two were selected as probe sequences

based on the predicted thermodynamics of the probe, its

probability of forming unwanted secondary structures, its

hybridisation target on the cDNA sequence and its specific-

ity. All probes possessed a Tm between 87 -C and 94 -C(90 -C optimal), a DG between �84 and �96 kcal/mol, a

hairpin Tm that did not exceed 45 -C, a DG for probe

duplexes that was not above�10 kcal/mol and a binding site

close to the 5Vpoly T tail of the cDNA. Probes were screened

by BLAST searches to ensure that no probe possessed a

stretch of 16 or more contiguous bases complementary to

any mouse mRNA sequence other than the target sequence.

In total, each array contained 282, 50-mer long probes

complementary to 161 genes printed in triplicate and 10

SpotReport probes (Stratagene, SpotReporti) printed 12

times each. Probes were dissolved in 3� SSC buffer and

arrayed in a randomly generated order onto Corning-Gap iglass slides, using a Microgrid II arrayer (Genomic

Solutions).

C.S. McKimmie et al. / Journal of Neuroimmunology 175 (2006) 128–141138

To determine the success of each print-run a represen-

tative microarray was hybridised to SpotCheck, a Cy-3

labelled random 9-mer solution (Genetix). 20 Al of Spot

Check was denatured at 95 -C for 5 min, followed by

cooling on ice for 5 min then applied to the array surface.

The SpotCheck solution was allowed to hybridise for

20 minutes in the dark at room temperature and then

washed with vigorous stirring in 1� SSC/0.2% SDS for

5 min, 0.1� SSC/0.2% SDS for 5 min and finally 0.1� SSC

for 5 min. Slides were then spun dry and read using an

Affymetrix 418 scanner.

Printed microarrays were stored in desiccators at room

temperature prior to use. Microarrays were prepared for

hybridisation to labelled cDNA as previously described

(Ebrahimi et al., 2003). Oligonucleotide probes were

firstly cross-linked to Corning-Gap i slides under UV

light and the array area demarcated using a diamond pen.

Arrays were then re-hydrated in a humidity chamber

containing 3� SSC for 5 min at room temperature and

then placed on a hotplate at 70 -C for 10 min. Slides

were then submerged in 170 mM succinic anhydride/N-

methyl-2-pyrrolidinone buffered by 40 mM borate buffer

for 15 min, washed in 95 -C water for 2 min, followed

by a short wash in absolute alcohol and dried by

centrifugation.

4.4. Labelling of cDNA and microarray hybridisation

All samples were labelled with a single dye, dCTP-

Cy3, and absolute probe intensities measured. For each

sample to be analysed 25 Ag of purified RNA was spiked

with 0.2 ng Stratagene SpotReporti mRNA, carried in

10 ng/Al of yeast tRNA buffer. RNA was then precipi-

tated, pelleted and resuspended in DEPC-treated water and

cDNA synthesised using the FluoroScripti cDNA

Labelling System (Invitrogen) according to the manufac-

turer’s instructions. Labelled cDNA was mixed with 20 AgCOT-1 human DNA (Invitrogen) as a carrier and to

reduce nonspecific binding of DNA to the microarray

surface. Excess Cy3-dCTP label was removed from the

cDNA by elution through S.N.A.Pi columns as per

manufacturer’s instructions (Invitrogen). YM-30 micro-

filtration columns (Amicon) were used to reduce the final

volume of labelled cDNA to < 15 Al. Samples were

immediately used for hybridisation to the microarray.

Labelled cDNA samples were mixed with 3.46 Al 20�SSC, 0.58 Al 10% SDS and nuclease-free water to give a

final volume of 20 Al and loaded onto the array surface

on the underside of a LifterSlip (Merck) cover slip to

distribute an equal volume across the array. Arrays were

placed in a humidified chamber, sealed and incubated at

57.5 -C overnight in the dark. Arrays were washed with

vigorous stirring in 1� SSC/0.2% SDS for 5 min, 0.1�SSC/0.2% SDS for 5 min and finally 0.1� SSC for 5 min.

Slides were then dried by centrifugation and read using an

Affymetrix 428 scanner.

4.5. Microarray data processing and analysis

Image quantification was carried out as described

previously (Forster et al., 2003). Probe signal and

background signal were determined using Quantarrayi(PerkinElmer, Boston, USA). The signal of each probe

was converted to a log base2 prior to data manipulation.

Signal variation across the array surface, also known

as ‘‘within-array normalisation’’, due to localised fluctua-

tions in hybridisation efficiency or other global effects

were corrected using the SpotReport system (Stratagene)

as per manufacturer’s instructions. Briefly, 10 different

exogenous Arabidpopsis mRNA transcripts were spiked

into samples prior to labelling. Upon hybridisation these

bound to specific probes printed at multiple locations

across the array surface. The array was divided into four

sub-grids, in which each probe was printed in triplicate.

The 10 SpotReport probes were also printed in triplicate

alongside the custom probes in each of the four sub-

grid, so that there was in total 120 SpotReport probes

across the entire array. Within-array normalisation

was then carried out by scaling the 75th percentile of

the SpotReport probes in each sub-grid to an identical

level, and so normalising probe signals across the array

surface.

Normalisation for array-to-array variation, due to

hybridisation or other global effects, was carried out by

scaling the 75th percentile of the 28 housekeeping probes,

each printed in triplicate, to an identical level.

The average signal of all negative control probes plus

2 standard deviations were used to calculate the threshold

for positive signals. Probes producing a signal at or lower

than this threshold were considered below the level of

detection. The median value of each probe spot (printed

in triplicate) was selected to represent each probe for a

particular array. Selecting the median prevents probes that

produce outlying signal intensities from contributing to

the data (Forster et al., 2003). All microarray data has

been entered into a MIAME compliant database at http://

www.gti.ed.ac.uk/GPX, accession GPX-000057.1 (tran-

scriptional profile of N9 microglia at 8, 24 and 72 h post

infection with Semliki Forest Virus), GPX-000049.1 (gene

expression profile of microglia following 18 h LPS

stimulation).

4.6. Reverse transcription of RNA for Quantitative Poly-

merase Chain Reaction (QPCR)

cDNA was synthesised from 5 Ag total RNA by

reverse transcription (RT) using Superscript II as per

manufacturer’s instructions (Invitrogen). Samples were

then diluted 1:5 in nuclease free water and used as

template for QPCR. Samples that were to be directly

compared within a QPCR analysis were reverse tran-

scribed using the same master mix to avoid inter-assay

variability.

C.S. McKimmie et al. / Journal of Neuroimmunology 175 (2006) 128–141 139

All primers were designed for use in the QPCR

machine (Rotorgene 3000, Corbett Research, Australia)

and an ‘‘outer’’ set of primers was designed to generate a

PCR product to act as a standard template as described

previously (McKimmie and Fazakerley, 2005). All primers

were designed using Primer3 software (available online at:

http://www.broad.mit.edu/cgi-bin/primer/primer3_www.cgi/)

and verified for low spontaneity to form duplexes and

hairpin loops using Oligo6 software (Molecular Biology

Insights). Both sets of primers were tested for specificity

using standard PCR and products characterised using an

Agilent Bioanalyser (Agilent Technologies). The QPCR

platform was accurately able to determine cDNA copy

numbers between 500 and 2�108 copies per reaction

using FastStart Taq polymerase (Roche) and SYBR green

(1:30,000) (BioGene Ltd.). The absolute amount of each

standard was determined by analysing the mass of each

PCR product on an Agilent Bioanalyser. The number of

copies of each transcript was then calculated from the

PCR product size. Alternatively for SFV RNA quantifi-

cation, a plasmid containing the SFV structural genes was

quantified using a spectrophotometer and dilutions used

for quantification. Following a hot start at 95 -C for

10 min to activate the hot-start Taq polymerase, an

amplification cycle was carried out for 40 cycles at the

following temperatures; 95 -C for 30 s, 62 -C for 20 s,

72 -C for 20 s. In each case, a serial dilution of PCR

fragment or plasmid containing template of interest was

assayed in parallel to convert cycle number of exponential

amplification to copies of template (Brown et al., 2003).

Sequences of primers used in QPCR were as follows;

T L R 2 ( 5 V- C G T TG T T CCCTGTG T TGC T- 3 V,AAAGTGGTTGTCGCCTGCT); TLR 3 (5V-TTGC-

GTTGCGAAGTGAAG-3V, 5V-TAAAAAGAGCGAGGG-GACAG-3V); IFN-a (5V-AGGACAGGAAGGATTTTGGA-3V, 5V-GCTGCTGATGGAGGTCATT-3 V); IL-1a (5V-G A T G T C C A A C T T C A C C T T C A - 3 V, 5 V-ACAAACTTCTGCCTGACGA-3V); IL-1h (5V-GTGTAAT-GAAAGACGGCACA-3V, 5V-AGAAACAGTCCAGCCCA-TAC-3V); SFV E1 (5V-CGCATCACCTTCTTTTGTG-3V, 5V-CCAGACCACCCGAGATTTT-3V); h-actin (5V-GTA-

CTCCTGTTGCTGATCC-3V, 5V-GTACTCCTGCTTGCT-GATCC-3V); GAPDH (5V-TGTCTCCTGCGACTTCAA-3V,5V-TGCAGCGAACTTTATTGATG-3V).

Acknowledgments

This work was supported by the award of a Wellcome

Trust Studentship to CSM, the European Union (SFVEC-

TORS) and the Scottish Higher Education Funding Council.

We thank in particular Klemens Vierlinger, Marie Craigon

and Alan Ross for help with microarray design and all work

involved in printing microarrays and quality control. We also

thank Graeme Grimes for help with curation of the micro-

array data using the GPX: Macrophage Expression Atlas.

References

Akira, S., Takeda, K., 2004. Toll-like receptor signalling. Nat. Rev.,

Immunol. 4, 499–511.

Alarcon, R., Fuenzalida, C., Santibanez, M., von Bernhardi, R., 2005.

Expression of scavenger receptors in glial cells. Comparing the

adhesion of astrocytes and microglia from neonatal rats to surface-

bound beta-amyloid. J. Biol. Chem. 280, 30406–30415.

Amor, S., Scallan, M.F., Morris, M.M., Dyson, H., Fazakerley, J.K., 1996.

Role of immune responses in protection and pathogenesis during

Semliki Forest virus encephalitis. J. Gen. Virol. 77 (Pt. 2), 281–291.

Ashkenas, J., Penman, M., Vasile, E., Acton, S., Freeman, M., Krieger, M.,

1993. Structures and high and low affinity ligand binding properties of

murine type I and type II macrophage scavenger receptors. J. Lipid Res.

34, 983–1000.

Bakker, A.B., Baker, E., Sutherland, G.R., Phillips, J.H., Lanier, L.L., 1999.

Myeloid DAP12-associating lectin (MDL)-1 is a cell surface receptor

involved in the activation of myeloid cells. Proc. Natl. Acad. Sci. U. S. A.

96, 9792–9796.

Bate, C., Veerhuis, R., Eikelenboom, P., Williams, A., 2004. Microglia kill

amyloid-beta1-42 damaged neurons by a CD14-dependent process.

NeuroReport 15, 1427–1430.

Bouchon, A., Hernandez-Munain, C., Cella, M., Colonna, M., 2001. A

DAP12-mediated pathway regulates expression of CC chemokine

receptor 7 and maturation of human dendritic cells. J. Exp. Med. 194,

1111–1122.

Brown, A.R., Webb, J., Rebus, S., Walker, R., Williams, A., Fazakerley,

J.K., 2003. Inducible cytokine gene expression in the brain in the

ME7/CV mouse model of scrapie is highly restricted, is at a strikingly

low level relative to the degree of gliosis and occurs only late in disease.

J. Gen. Virol. 84, 2605–2611.

Buechler, C., Ritter, M., Quoc, C.D., Agildere, A., Schmitz, G., 1999.

Lipopolysaccharide inhibits the expression of the scavenger receptor

Cla-1 in human monocytes and macrophages. Biochem. Biophys. Res.

Commun. 262, 251–254.

Carson, M.J., 2002. Microglia as liaisons between the immune and central

nervous systems: functional implications for multiple sclerosis. Glia 40,

218–231.

Carson, M.J., Reilly, C.R., Sutcliffe, J.G., Lo, D., 1998. Mature microglia

resemble immature antigen-presenting cells. Glia 22, 72–85.

Chen, K., Iribarren, P., Hu, J., Chen, J., Gong, W., Cho, E.H., Lockett, S.,

Dunlop, N.M., Wang, J.M., 2006. Activation of Toll-like receptor 2 on

microglia promotes cell uptake of Alzheimer disease-associated

amyloid beta peptide. J. Biol. Chem. 281, 3651–3659.

Coraci, I.S., Husemann, J., Berman, J.W., Hulette, C., Dufour, J.H.,

Campanella, G.K., Luster, A.D., Silverstein, S.C., El Khoury, J.B.,

2002. CD36, a class B scavenger receptor, is expressed on microglia in

Alzheimer’s disease brains and can mediate production of reactive

oxygen species in response to beta-amyloid fibrils. Am. J. Pathol. 160,

101–112.

D’Aversa, T.G., Weidenheim, K.M., Berman, J.W., 2002. CD40–CD40L

interactions induce chemokine expression by human microglia: impli-

cations for human immunodeficiency virus encephalitis and multiple

sclerosis. Am. J. Pathol. 160, 559–567.

Dziarski, R., 2004. Peptidoglycan recognition proteins (PGRPs). Mol.

Immunol. 40, 877–886.

Dziarski, R., Platt, K.A., Gelius, E., Steiner, H., Gupta, D., 2003. Defect in

neutrophil killing and increased susceptibility to infection with

nonpathogenic gram-positive bacteria in peptidoglycan recognition

protein-S (PGRP-S)-deficient mice. Blood 102, 689–697.

Ebrahimi, B., Dutia, B.M., Roberts, K.L., Garcia-Ramirez, J.J., Dickinson,

P., Stewart, J.P., Ghazal, P., Roy, D.J., Nash, A.A., 2003. Transcriptome

profile of murine gammaherpesvirus-68 lytic infection. J. Gen. Virol.

84, 99–109.

El Khoury, J., Hickman, S.E., Thomas, C.A., Cao, L., Silverstein, S.C.,

Loike, J.D., 1996. Scavenger receptor-mediated adhesion of microglia

to beta-amyloid fibrils. Nature 382, 716–719.

C.S. McKimmie et al. / Journal of Neuroimmunology 175 (2006) 128–141140

El Khoury, J.B., Moore, K.J., Means, T.K., Leung, J., Terada, K., Toft, M.,

Freeman, M.W., Luster, A.D., 2003. CD36 mediates the innate host

response to beta-amyloid. J. Exp. Med. 197, 1657–1666.

Fazakerley, J.K., 2002. Pathogenesis of Semliki Forest virus encephalitis. J.

Neurovirology 8 (Suppl. 2), 66–74.

Febbraio, M., Hajjar, D.P., Silverstein, R.L., 2001. CD36: a class B

scavenger receptor involved in angiogenesis, atherosclerosis, inflam-

mation, and lipid metabolism. J. Clin. Invest. 108, 785–791.

Ferrari, D., Villalba, M., Chiozzi, P., Falzoni, S., Ricciardi-Castagnoli, P., Di

Virgilio, F., 1996. Mouse microglial cells express a plasma membrane

pore gated by extracellular ATP. J. Immunol. 156, 1531–1539.

Forster, T., Roy, D., Ghazal, P., 2003. Experiments using microarray

technology: limitations and standard operating procedures. J. Endocri-

nol. 178, 195–204.

Franc, N.C., White, K., Ezekowitz, R.A., 1999. Phagocytosis and

development: back to the future. Curr. Opin. Immunol. 11, 47–52.

Gordon, S., 2002. Pattern recognition receptors: doubling up for the innate

immune response. Cell 111, 927–930.

Grewal, R.P., Yoshida, T., Finch, C.E., Morgan, T.E., 1997. Scavenger

receptor mRNAs in rat brain microglia are induced by kainic acid

lesioning and by cytokines. NeuroReport 8, 1077–1081.

Hampton, R.Y., Golenbock, D.T., Penman, M., Krieger, M., Raetz, C.R.,

1991. Recognition and plasma clearance of endotoxin by scavenger

receptors. Nature 352, 342–344.

Hanisch, U.K., 2002. Microglia as a source and target of cytokines. Glia 40,

140–155.

Hirano, K., Yamashita, S., Nakagawa, Y., Ohya, T., Matsuura, F.,

Tsukamoto, K., Okamoto, Y., Matsuyama, A., Matsumoto, K.,

Miyagawa, J., Matsuzawa, Y., 1999. Expression of human scavenger

receptor class B type I in cultured human monocyte-derived macro-

phages and atherosclerotic lesions. Circ. Res. 85, 108–116.

Ho, Y.S., Xiong, Y., Ma, W., Spector, A., Ho, D.S., 2004. Mice lacking

catalase develop normally but show differential sensitivity to oxidant

tissue injury. J. Biol. Chem. 279, 32804–32812.

Honda, M., Akiyama, H., Yamada, Y., Kondo, H., Kawabe, Y., Takeya, M.,

Takahashi, K., Suzuki, H., Doi, T., Sakamoto, A., Ookawara, S., Mato,

M., Gough, P.J., Greaves, D.R., Gordon, S., Kodama, T., Matsushita,

M., 1998. Immunohistochemical evidence for a macrophage scavenger

receptor in Mato cells and reactive microglia of ischemia and

Alzheimer’s disease. Biochem. Biophys. Res. Commun. 245, 734–740.

Howard, L.M., Miga, A.J., Vanderlugt, C.L., Dal Canto, M.C., Laman, J.D.,

Noelle, R.J., Miller, S.D., 1999. Mechanisms of immunotherapeutic

intervention by anti-CD40L (CD154) antibody in an animal model of

multiple sclerosis. J. Clin. Invest. 103, 281–290.

Huang, Q., Liu, D., Majewski, P., Schulte, L.C., Korn, J.M., Young, R.A.,

Lander, E.S., Hacohen, N., 2001. The plasticity of dendritic cell

responses to pathogens and their components. Science 294, 870–875.

Husemann, J., Loike, J.D., Kodama, T., Silverstein, S.C., 2001. Scavenger

receptor class B type I (SR-BI) mediates adhesion of neonatal murine

microglia to fibrillar beta-amyloid. J. Neuroimmunol. 114, 142–150.

Husemann, J., Loike, J.D., Anankov, R., Febbraio, M., Silverstein, S.C.,

2002. Scavenger receptors in neurobiology and neuropathology: their

role on microglia and other cells of the nervous system. Glia 40,

195–205.

Jana, M., Liu, X., Koka, S., Ghosh, S., Petro, T.M., Pahan, K., 2001.

Ligation of CD40 stimulates the induction of nitric-oxide synthase in

microglial cells. J. Biol. Chem. 276, 44527–44533.

Jenner, R.G., Young, R.A., 2005. Insights into host responses

against pathogens from transcriptional profiling. Nat. Rev., Microbiol.

3, 281–294.

Kadowaki, N., Ho, S., Antonenko, S., Malefyt, R.W., Kastelein, R.A.,

Bazan, F., Liu, Y.J., 2001. Subsets of human dendritic cell precursors

express different toll-like receptors and respond to different microbial

antigens. J. Exp. Med. 194, 863–869.

Kane, M.D., Jatkoe, T.A., Stumpf, C.R., Lu, J., Thomas, J.D., Madore, S.J.,

2000. Assessment of the sensitivity and specificity of oligonucleotide

(50mer) microarrays. Nucleic Acids Res. 28, 4552–4557.

Kielian, T., 2004. Microglia and chemokines in infectious diseases of the

nervous system: views and reviews. Front. Biosci. 9, 732–750.

Krieger, M., 1999. Charting the fate of the ‘‘good cholesterol’’: identifica-

tion and characterization of the high-density lipoprotein receptor SR-BI.

Annu. Rev. Biochem. 68, 523–558.

Krieger, M., Stern, D.M., 2001. Series introduction: multiligand receptors

and human disease. J. Clin. Invest. 108, 645–647.

Mack, C.L., Vanderlugt-Castaneda, C.L., Neville, K.L., Miller, S.D., 2003.

Microglia are activated to become competent antigen presenting and

effector cells in the inflammatory environment of the Theiler’s virus

model of multiple sclerosis. J. Neuroimmunol. 144, 68–79.

Matsumoto, M., Tanaka, T., Kaisho, T., Sanjo, H., Copeland, N.G., Gilbert,

D.J., Jenkins, N.A., Akira, S., 1999. A novel LPS-inducible C-type

lectin is a transcriptional target of NF-IL6 in macrophages. J. Immunol.

163, 5039–5048.

McKimmie, C.S., Fazakerley, J.K., 2005. In response to pathogens, glial

cells dynamically and differentially regulate Toll-like receptor gene

expression. J. Neuroimmunol. 169, 116–125.

McKimmie, C.S., Johnson, N., Fooks, A.R., Fazakerley, J.K., 2005. Viruses

selectively upregulate Toll-like receptors in the central nervous system.

Biochem. Biophys. Res. Commun. 336, 925–933.

Meda, L., Cassatella, M.A., Szendrei, G.I., Otvos, L. Jr., Baron, P., Villalba,

M., Ferrari, D., Rossi, F., 1995. Activation of microglial cells by beta-

amyloid protein and interferon-gamma. Nature 374, 647–650.

Moller, T., Nolte, C., Burger, R., Verkhratsky, A., Kettenmann, H., 1997.

Mechanisms of C5a and C3a complement fragment-induced [Ca2+]i

signaling in mouse microglia. J. Neurosci. 17, 615–624.

Mukhopadhyay, S., Peiser, L., Gordon, S., 2004. Activation of murine

macrophages by Neisseria meningitidis and IFN-{gamma} in vitro:

distinct roles of class A scavenger and Toll-like pattern recognition

receptors in selective modulation of surface phenotype. J. Leukoc. Biol.

Muller, U., Steinhoff, U., Reis, L.F., Hemmi, S., Pavlovic, J., Zinkernagel,

R.M., Aguet, M., 1994. Functional role of type I and type II interferons

in antiviral defense. Science 264, 1918–1921.

Nakamura, K., Funakoshi, H., Miyamoto, K., Tokunaga, F., Nakamura, T.,

2001. Molecular cloning and functional characterization of a human

scavenger receptor with C-type lectin (SRCL), a novel member of a

scavenger receptor family. Biochem. Biophys. Res. Commun. 280,

1028–1035.

Napolitani, G., Rinaldi, A., Bertoni, F., Sallusto, F., Lanzavecchia, A.,

2005. Selected Toll-like receptor agonist combinations synergistically

trigger a T helper type 1-polarizing program in dendritic cells. Nat.

Immunol. 6, 769–776.

Nystrom, K., Biller, M., Grahn, A., Lindh, M., Larson, G., Olofsson, S.,

2004. Real time PCR for monitoring regulation of host gene expression

in herpes simplex virus type 1-infected human diploid cells. J. Virol.

Methods 118, 83–94.

Okuno, T., Nakatsuji, Y., Kumanogoh, A., Koguchi, K., Moriya, M.,

Fujimura, H., Kikutani, H., Sakoda, S., 2004. Induction of cyclo-

oxygenase-2 in reactive glial cells by the CD40 pathway: relevance to

amyotrophic lateral sclerosis. J. Neurochem. 91, 404–412.

Olson, J.K., Miller, S.D., 2004. Microglia initiate central nervous system

innate and adaptive immune responses through multiple TLRs. J.

Immunol. 173, 3916–3924.

Olson, J.K., Girvin, A.M., Miller, S.D., 2001. Direct activation of innate

and antigen-presenting functions of microglia following infection with

Theiler’s virus. J. Virol. 75, 9780–9789.

Paglinawan, R., Malipiero, U., Schlapbach, R., Frei, K., Reith, W., Fontana,

A., 2003. TGFbeta directs gene expression of activated microglia to an

anti-inflammatory phenotype strongly focusing on chemokine genes

and cell migratory genes. Glia 44, 219–231.

Palecanda, A., Paulauskis, J., Al Mutairi, E., Imrich, A., Qin, G., Suzuki,

H., Kodama, T., Tryggvason, K., Koziel, H., Kobzik, L., 1999. Role of

the scavenger receptor MARCO in alveolar macrophage binding of

unopsonized environmental particles. J. Exp. Med. 189, 1497–1506.

Peress, N.S., Fleit, H.B., Perillo, E., Kuljis, R., Pezzullo, C., 1993.

Identification of Fc gamma RI, II and III on normal human brain

C.S. McKimmie et al. / Journal of Neuroimmunology 175 (2006) 128–141 141

ramified microglia and on microglia in senile plaques in Alzheimer’s

disease. J. Neuroimmunol. 48, 71–79.

Qi, H., Denning, T.L., Soong, L., 2003. Differential induction of

interleukin-10 and interleukin-12 in dendritic cells by microbial toll-

like receptor activators and skewing of T-cell cytokine profiles. Infect.

Immun. 71, 3337–3342.

Re, F., Strominger, J.L., 2001. Toll-like receptor 2 (TLR2) and TLR4

differentially activate human dendritic cells. J. Biol. Chem. 276,

37692–37699.

Rempel, J.D., Quina, L.A., Blakely-Gonzales, P.K., Buchmeier, M.J.,

Gruol, D.L., 2005. Viral induction of central nervous system innate

immune responses. J. Virol. 79, 4369–4381.

Reverter, A., McWilliam, S.M., Barris, W., Dalrymple, B.P., 2005. A rapid

method for computationally inferring transcriptome coverage and

microarray sensitivity. Bioinformatics 21, 80–89.

Righi, M., Mori, L., De Libero, G., Sironi, M., Biondi, A., Mantovani, A.,

Donini, S.D., Ricciardi-Castagnoli, P., 1989. Monokine production by

microglial cell clones. Eur. J. Immunol. 19, 1443–1448.

Savill, J., Dransfield, I., Gregory, C., Haslett, C., 2002. A blast from the

past: clearance of apoptotic cells regulates immune responses. Nat.

Rev., Immunol. 2, 965–975.

Scallan, M.F., Allsopp, T.E., Fazakerley, J.K., 1997. bcl-2 acts early to

restrict Semliki Forest virus replication and delays virus-induced

programmed cell death. J. Virol. 71, 1583–1590.

Schmid, C.D., Sautkulis, L.N., Danielson, P.E., Cooper, J., Hasel, K.W.,

Hilbush, B.S., Sutcliffe, J.G., Carson, M.J., 2002. Heterogeneous

expression of the triggering receptor expressed on myeloid cells-2 on

adult murine microglia. J. Neurochem. 83, 1309–1320.

Streit, W.J., 2002. Microglia as neuroprotective, immunocompetent cells of

the CNS. Glia 40, 133–139.

Tan, J., Town, T., Crawford, F., Mori, T., DelleDonne, A., Crescentini,

R., Obregon, D., Flavell, R.A., Mullan, M.J., 2002. Role of CD40

ligand in amyloidosis in transgenic Alzheimer’s mice. Nat. Neurosci.

5, 1288–1293.

Tsunawaki, S., Kagara, S., Yoshikawa, K., Yoshida, L.S., Kuratsuji, T.,

Namiki, H., 1996. Involvement of p40phox in activation of

phagocyte NADPH oxidase through association of its carboxyl-

terminal, but not its amino-terminal, with p67phox. J. Exp. Med. 184,

893–902.

Tunaru, S., Kero, J., Schaub, A., Wufka, C., Blaukat, A., Pfeffer, K.,

Offermanns, S., 2003. PUMA-G and HM74 are receptors for nicotinic

acid and mediate its anti-lipolytic effect. Nat. Med. 9, 352–355.

van der Laan, L.J., Dopp, E.A., Haworth, R., Pikkarainen, T., Kangas, M.,

Elomaa, O., Dijkstra, C.D., Gordon, S., Tryggvason, K., Kraal, G.,

1999. Regulation and functional involvement of macrophage scavenger

receptor MARCO in clearance of bacteria in vivo. J. Immunol. 162,

939–947.

Wang, T., Town, T., Alexopoulou, L., Anderson, J.F., Fikrig, E., Flavell,

R.A., 2004. Toll-like receptor 3 mediates West Nile virus entry into the

brain causing lethal encephalitis. Nat. Med. 10, 1366–1373.

Warrington, J.A., Nair, A., Mahadevappa, M., Tsyganskaya, M., 2000.

Comparison of human adult and fetal expression and identification of

535 housekeeping/maintenance genes. Physiol. Genomics 2, 143–147.

Yamada, Y., Doi, T., Hamakubo, T., Kodama, T., 1998. Scavenger receptor

family proteins: roles for atherosclerosis, host defence and disorders of

the central nervous system. Cell. Mol. Life Sci. 54, 628–640.

Zarember, K.A., Godowski, P.J., 2002. Tissue expression of human Toll-

like receptors and differential regulation of Toll-like receptor mRNAs in

leukocytes in response to microbes, their products, and cytokines. J.

Immunol. 168, 554–561.

Zelko, I.N., Mariani, T.J., Folz, R.J., 2002. Superoxide dismutase multigene

family: a comparison of the CuZn-SOD (SOD1), Mn-SOD (SOD2), and

EC-SOD (SOD3) gene structures, evolution, and expression. Free

Radic. Biol. Med. 33, 337–349.