RES E AR C H Open Access

A three-dimensional collagen construct to model

lipopolysaccharide-induced activation of BV2

microglia

Randy Tatt Yhew Haw

1†

, Chih Kong Tong

1†

, Andrew Yew

, Han Chung Lee

, James B Phillips

and Sharmili Vidyadaran

Abstract

Background: We report a novel method of culturing microglia in three dimension (3D) using collagen as a

substrate. By culturing microglia within a matrix, we aim to emulate the physical state of microglia embedded

within parenchyma.

Methods: BV2 microglia cell suspensions were prepared with type I collagen and cast into culture plates. To

characterise the BV2 microglia cultured in 3D, the cultures were evaluated for their viability, cell morphology and

response to lipopolysaccharide (LPS) activation. Conventional monolayer cultures (grown on uncoated and

collagen-coated polystyrene) were set up concurrently for compa rison.

Results: BV2 microglia in 3D collagen matrices were viable at 48 hrs of culture and exhibit a ramified morphology

with multiplanar cytoplasmic projections. Following stimulation with 1 μg/ml LPS, microglia cultured in 3D collagen

gels increase their expression of nitric oxide (NO) and CD40, indicating their capac ity to become activated within

the matrix. Up to 97.8% of BV2 microglia grown in 3D cultures gained CD40 positivity in response to LPS, compared

to approximately 60% of cells grown in a monolayer (P < .05). BV2 microglia in 3D collagen gels also showed

increased mRNA and protein expression of inflammatory cytokines IL-6, TNF-α and the chemoattractant MCP-1

following LPS stimulation.

Conclusions: In summary, BV2 microglia cultured in 3D collagen hydrogels exhibit multiplanar cytoplasmic

projections and undergo a characteristic and robust activation response to LPS. This culture system is accessible to

a wide range of analyses and provides a useful new in vitro tool for research into microglial activation.

Keywords: Microglia, Lipopolysaccharide, Collagen matrix, Three-dimensional cultures

Background

Microglia are tissue-specific macrophages of the central

ner vous system (CNS) and derive from primitive haem-

atopoietic progenitors of erythromyeloid origin [1,2].

These mononuclear phagocytes are the resident im-

mune cells of the CN S, along with other subsets of

mononuclear phagocytes including meningeal macro-

phages, choroid plexus macrophages and perivascular

macrophages [3]. Microglia in the brain are disseminated

throughout the parenchyma and are highly motile. In the

healthy mature CNS, microglia exist mainly in a ramified

form, continuously traversing the CNS and using their cyto-

plasmic processes to sample the tissue environment [4].

Although microglia have long been thought of as merely

a stromal cell, microglia research has now revealed crucial

roles for these cells in inflammation and in different stages

of neurodevelopment. Homeostatic changes in the CNS

rapidly trigger a reactive form of microglia, characterised

by a shift to amoeboidal morphology, increased motility,

proliferation and release of inflammatory mediators [5].

In the embryonic brain, microglia associate closely with

apoptotic cells, presumably to promote developmental

neuron death and phagocytose the ensuing cellular debris

* Correspondence: [email protected]

†

Equal contributors

Neuroinflammation Group, Immunology Laboratory, Department of

Pathology, Faculty of Medicine & Health Sciences, Universiti Putra Malaysia,

43400 Serdang, Selangor, Malaysia

Full list of author information is available at the end of the article

JOURNAL OF

NEUROINFLAMMATION

Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and

reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain

Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article,

unless otherwise stated.

Haw et al. Journal of Neuroinflammation 2014, 11:134

http://www.jneuroinflammation.com/content/11/1/134

[6]. During postnatal development, microglia play a role

in synaptic pruning by engulfing synaptic material [7,8]. In

adult hippocampal neurogenesis, microglia provide an

important housekeeping role by phagocytosing and clear-

ing apopto tic ne wborn neurons [9]. These function s of

the microglia have led to the recognition of the signifi-

cance of this cell in brain rese arch.

Homogenous cell cultures are a valuable approach for

neuroscience research that allows monitoring of the cell

population of interest in a carefully controlled environ-

ment. Monolayer cultures fail to recreate the 3D spatial

arrangement of cells and matrices present in tissues, and

stiff plastic substrates do not resemble the physical environ-

ment of the CNS. Culture systems that better mimic the

behaviour of microglia within a 3D milieu would allow

scrutiny of these cells in a more relevant tissue-like micro-

environment. Brain slice culture allows in situ examination

of cellular responses, and has been used to study microglial

responses to carcinomas [10] and cerebral amyloidosis [11].

However, downstream analysis of individual cell popula-

tions can be complicated and the local cellular environment

is complex.

Here, we describe a 3D culture system utilising type I

collagen as the basal substrate to provide a matrix for the

culture of microglial cells. Type I collagen is a matrix

material that is easily manipulated, is widely used in

culture models [12], and has previously been used to

develop 3D cultures for astrocytes and neurons [13-15].

Being a simple matrix, it also serves as a suitable baseline

scaffold on which the deposition of other extracellular

matrix (ECM) molecules can be detected. With control

over the seeding density and the chemical environment

within the gels, we are able to examine specific features of

microglia in a 3D matrix with ease of monitoring the cells

compared to in vivo models or brain slice cultures.

In the laboratory we routinely culture BV2 microglia, a cell

line of murine origin immortalised with v-raf/v-myc onco-

genes and commonly used in microglia studies. The BV2

microglia are similar in morphology to isolated microglia,

express inflammatory mediators and display phagocytic

activity [16]. To stimulate the BV2 microglia into an inflam-

matory phenotype, the bacterial cell wall component lipo-

polysaccharide (LPS) is used. By examining morphology,

viability and activation status (by evaluating nitric oxide

production along with CD40 and inflammatory cytokine

expression) of BV2 microglia in 3D constructs and com-

paring them to conventional monolayer cultures, we char-

acterise microglia cultured in 3D and report a model for

microglial activation in a 3D collagen matrix using LPS.

Methods

BV2 cells

BV2, an immortalised mouse microglia cell line, was cul-

tured in high glucose Dulbecco modified Eagle medium

(DMEM; Gibco, Carls bad USA) supplemented with 5%

foetal bovine serum (Gibco, Carlsbad USA), 6.25 μg/ml

insulin (Sigma-Aldrich, St. Louis USA), 1X non essen-

tial amino acid (Gib co, Carlsbad USA), 1% penicillin

and streptomycin ( i-DNA , Singapore), 0.5% fungizone

(Gibco, Carlsbad USA) and 0.1% gentamicin (Gib co,

Carlsbad USA). The cultures were maintained in a hu-

midified incubator at 37°C with 5% CO

:95% air.

Culture of BV2 cells on monolayer surfaces and in

three-dimensional collagen gels

BV2 cells were har vested at 70-90% confluency and

counted u sing a haemocytometer. Cultures up to pas-

sage 20 were used for experiments. The seeding density

for all culture conditions and downstream assays was

0.3 × 10

cells/well. For uncoated and collagen-coated

monolayer cultures, cells were seeded in 6-well plates.

For coated monolayer cultures, culture plates were pre -

coated with type I rat tail collagen (First Link UK Ltd,

Birmingham, UK) in 0.6% acetic acid for 30 min. The

culture flasks were then rinsed with supplemented DMEM

to remove any traces of acid.

For 3D cultures, collagen gels were prepared by adding

10% v/v cell suspension in supplemented DMEM, 10% v/v

10X minimum essential medium (MEM; Sigma-Aldrich St.

Louis USA) and 80% v/v type I rat tail collagen (2 mg/ml in

0.6% acetic acid; First Link, Birmingham UK). The MEM

and collagen were mixed and neutralised using sodium

hydroxide as assessed by colour change of the phenol red

indicator. Upon neutralisation, the collagen-MEM mixture

was gently mixed with the BV2 cell suspension and

transferred to culture plates (0.3 ml/well in 24-well

plates and 1.2 ml/well in 6-well plates or 35 mm WillCo

(WillCo Wells, Amsterdam, Netherlands) dishes; resulting

gels were approximately 2-mm thick). The gels were

allowed to set in a humidified incubator at 37°C with 5%

:95% air for 5 to 10 min and subsequently covered

with 2 ml supplemented DMEM. The cultures were main-

tained for up to 48 hrs and cells were retrieved for analysis

using 0.25% trypsin for monolayer cultures and 0.125%

type I collagenase for 3D cultures.

To stimulate the cultures, BV2 microglia were treated

with 1 μg/ml lipopolysaccharide (LPS; E. coli serotype

O26:B6; Sigma-Aldrich St. Louis USA, Cat. No. L2762) in

supplemented DMEM. Controls were subjected to media

change only.

Scanning electron microscopy

Collagen gels with and without BV2 cells were processed

based on a mo dified version of Lizárrag a and col-

leagues [17]. Briefly, samples were p refixed in 2.5% v/v

glutaraldehyde at 4°C for 4 hrs and washed with 0.1 M

sodium cacodylate buffer. After overnight incubation,

samples were post-fixated in 1% osmium tetraoxide.

Haw et al. Journal of Neuroinflammation 2014, 11:134 Page 2 of 10

http://www.jneuroinflammation.com/content/11/1/134

Samples were then dehydrated using a graded series of

ethanol followed by immersion in 100% acetone. The

samples were then transferred to a Baltec CPD 030 Critical

Point Dryer for critical point drying. Samples were coated

with gold-palladium in a Baltec SCD 005 Sputter Coater

and examined under JEOL JSM-6400 SEM.

DAPI/propidium iodide staining

BV2 cell viability in all culture conditions was assessed

using propidium iodide ( PI) and a 4',6-d iamidino-2-

phenylindole, dihydrochlorid e (DAPI) counterstain. In

brief, 20 μg/ml PI (Molecular Probes, Oregon USA)

wasaddedtoculturesat24and48hrsandincubated

for 10 min at 37°C. Supernatant was t hen removed and

the cultures rinsed in 1× PBS thrice for 5 min to remove

PI residue. Cells were fixed with 4% paraformaldehyde

(PFA) at 4°C for 1 hr. The cultures were then incubated

with 1 μg/ml DAPI (Molecular Probes, Oregon USA) in

1X PBS with 0.1% Triton-X for 10 min. For the positive

control, Triton-X (0.2% in basic DMEM) was used to treat

the cells for 5 min prior to PI staining.

Lactate dehydrogenase assay

Lactate dehydrogenase (LDH) assay is a colourimetric assay

that specifically detects the enzyme lactate dehydrogenase.

This enzyme is particularly stable and is present in the cul-

ture supernatant when cells are damaged. For this assay,

three controls were established, namely, the background

control (only media), low control (untreated cells) and posi-

tive control (cells treated with lysis solution). All cultures

were supplied with an equal volume (1.5 ml) of culture

medium. Following overnight incubation at 37°C and 5%

:95% air, cells were treated with 1 μg/ml LPS in phenol

red-free, supplemented DMEM or subjected to a media

change for untreated cells. At 48 hrs, lysis solution obtained

from the Cytotoxicity LDH kit (Roche, Mannheim Germany)

was added into the positive control wells. Next, 500 μlof

media from each well was aliquoted into tubes and centri-

fuged at 1,500 rpm for 5 min to remove any debris. Next,

100 μl of the supernatant was transferred into a 96-well

plate in triplicate. A reaction mixture was prepared by

mixing the catalyst solution and dye solution from the kit

and 100 μl of the reaction mixture was added into each

well. The 96-well plate was incubated for 20 min prior to

measuring the absorbance at a 490-nm wavelength using

a microplate reader (Dynex, Virginia United States). Read-

ings from background controls were subtracted from all

reaction absorbance readings.

Lectin staining

BV2 cell morphology was assessed using lectin histochem-

ical staining. Briefly, cells were fixed with 4% paraformalde-

hyde (PFA) in 1X PBS for 1 hr followed by permeabilisation

with 0.2% Triton- X in 1× PBS for 30 min. Cells were then

incubated with fluorescein isothiocyanate (FIT C)-conjugated

tomato lectin (1:300 dilution in 0.2% of Triton- X in 1X PBS;

Sigma-Aldrich, St Louis USA) for 1 hr. Nuclei were counter-

stained with 0.1 μg/ml DAPI for 5 min before cells were

viewed and photographed with an inverted fluorescence

microscope (Olympus, Tokyo Japan) and laser scanning

confocal microscope (Leica DMIL, Wetzlar Germany).

CD40 immunophenotyping

For immunopheno typing, cells were harvested from cul-

tures and res uspended in 100 μl of 1× PBS. They were

then incubated with Fixable Viability Dye eFluor™ 780

(eBioscience, San Diego U SA) at 1:1000 dilution for

15 min, followed by an anti-mouse CD40-FIT C antibody

(1:100 dilution; BD Pharmigen, San Diego USA) for 30 min

at 4°C. Cells were then washed and resuspended in 1× PBS

before being analysed by a FACS Fortessa Cytometer

(BD Biosciences, San Jose, CA USA). Gating was applied

to identify intact cells according to forward a nd side

scatter plots; dead cells were excluded from analysis using

the Fixable Viability Dye eFluor™ 780 staining. Data were

analysed using the FACS Diva software.

Griess assay

This assay was performed to study the activation of BV2

cells in a resting state and LPS-treated state by determining

the production of nitric oxide (NO) at 36 and 48 hrs, re-

spectively. BV2 cells were seeded in 6-well plates for mono-

layer cultures and 24-well plates for 3D collagen cultures at

a seeding density of 0.3 × 10

cells/well. All cultures were

supplied with an equal volume (2 ml) of culture medium.

Following overnight incubation at 37°C and 5% CO

,cells

were treated with 1 μg/ml LPS in phenol red-free, supple-

mented DMEM. For the Griess assay, 200 μlofmediafrom

each well was aliquoted into tubes and centrifuged at

1,500 rpm for 5 min to remove any debris. Next, 50 μlof

the supernatant was transferred into a 96-well plate in trip-

licate. A series of sodium nitrite (NaNO

)standardsforthe

Griess assay were prepared by serial dilution, ranging from

0 μMto100μM. Griess reagent was freshly prepared by

dissolving sulphanilamide (Sigma-Aldrich, United States)

and N-(1Naphtyl)ethylenediamine (Sigma-Aldrich, St Louis

United States) with phosphoric acid. Fifty microliters of

Griess reagent was added to each well before absor-

bances were measured with a microplate s pec trometer

(Dynex, Virginia United States) at 530 nm wavelength.

The NO

−

concentr ation was then evaluated by normalising

the absorbance reading using the graph equation obtained

from the plotted standard graph.

Detection of cytokine expression with reverse transcriptase

quantitative PCR and a cytokine bead array

The mRNA expression of MCP-1, IL-6, IL-1β,IL-12β

and TNF-α of BV2 mic roglia was assessed with reverse-

Haw et al. Journal of Neuroinflammation 2014, 11:134 Page 3 of 10

http://www.jneuroinflammation.com/content/11/1/134

transcriptase quantitative P CR (RT-qP CR). The RNA

of BV2 microglia wa s isolated using the RNeas y Plus

Mini Kit (Qiagen, Limburg German y) after a 6-hr

stimulation with LPS. The isolatio n process wa s con-

ducted according to the kit’s manual. The yield of

total RNA was quantified by optical density (OD)

readings at 260 nm, and the purity wa s estimated by

the 260:280 nm ratio. Using the SuperScript III

Reverse Transcriptase Kit (Invitrogen, Carlsbad U SA),

an equivalent amount of RNA samples (500 ng) were

then re verse transcribed into single stranded cDNA in

a reaction mixture consisting of 0.5 mM dN TP mix,

2.5uM olig o d( T)

,0.01MDTT,40URNAseOUT™,

1× RT buffer and 200U Superscript™ RT prepared in

20uL per reaction, according to the manufacturer’s

protocol. The primers and housekeeping genes (Hmbs,

Pgk1 and Psmb2) were designed and probes selec ted

using ProbeFinder Version 2.49 (Universal Probe Library

Assay (UPL) Design Center, Roche). Details on the primer

sequences are in Additional file 1. PCR wa s then per-

formed on a LightCycler 480 System (Roche, B a sel

Switzerland).Conditions for the RT-qPCR were a pre-

denaturing step of 95°C for 10 min; 45 cycles of 95°C

for 10 sec , 60°C for 30 se c, and 72°C for 1 sec; and

finishing with a cooling step at 40°C for 30 sec [18].

A PCR efficiency of between 90 and 110% and an

R-squared value >0.98 were used to define successful

assays. Relative quantification of target g ene e xpression

in our samples wa s carried out using the comparative

Ct method. We performed intra-sample data normalisa-

tion against the three endogenous control reference

genesasmentionedabove.

For the bead array, culture supernatants were assesse d

at 48 hrs for expression of IL-6, IL-10, MCP-1, IFN-γ,

TNF, and IL-12p70 using the multiplex bead array kit

(BD Cytometric Bead Array mouse inflammation kit,

BD Biosciences, San Jose, C A, USA). The samples were

assayed according to the manufacturer ’s instructions

with the FACS Fortessa flow cytometer (BD Bioscienc e,

San Jose, CA, USA). The resulting data were analysed

with FC AP array software (BD Bioscience, San Jose, CA,

USA). Expression of the cytokines wa s e valuated by

determining their respe ctive concentration (pictograms

per millilitre) using individual standard curves.

Statistical analysis

Signific ance was assessed using GraphPad Prism version 6

(GraphPad Software, CA, USA, http://www.graphpad.com/).

Results

Scanning electron microscopy of BV2 microglia and

collagen gel structure

Examination of the collagen gel structure was performed

using scanning electron microscopy. The matrix consists

of a dense fibrillar network, with numerous interfibrillar

spaces (Figure 1A). Microglia were detected on the surface

of the gels and also beneath the surface (arrows and

arrowheads in Figure 1B,C ).

Morphology and viability of BV2 microglia cultured in

monolayer and three dimensions

To examine the morphology of microglia in the various

culture systems , we performed staining of cells with

FITC-tagged lectin. M icroglia cultured on uncoated

monolayer surfaces mostly exhibited round cytopla sm

with some bipolar project ions (Figure 2A). In collagen-

coated monolayer cultures , the extent of amoeboidal

morphology seemed increased (Figure 2B) with cyto-

plasmic area appearing minimal, indicating the extent

of deramification. The morphology of cells cultured in

3D collagen wa s distinct from monolayer, with clear

multiplanar projections (Figure 2C). The microglia

were suspended within the collagen matrix and ev enly

distributed across the width of the gel. When viewed

with confocal microscopy, the extent of ramification of

microglia cultured in the 3D matrix was evident, with

cells displaying long and multi directional cellular pro-

jections (Figure 2F and video in Additional file 2).

To determine whether microglia cultured in 3D were

viable, the lactate dehydrogenase (LDH) assay and DAPI/PI

staining were performed. Compared to the respective posi-

tive controls, BV2 microglia in all three culture formats

Figure 1 Scanning electron micrographs of three-dimensional (3D) collagen gel and microglia cultured in 3D. (A) Structure and

organisation of fibrils of 2 mg/ml collagen gels. (B,C) Low magnification images show microglia on the surface (arrows) and embedded

(arrowheads) within the collagen gels. Magnification as indicated on each micrograph.

Haw et al. Journal of Neuroinflammation 2014, 11:134 Page 4 of 10

http://www.jneuroinflammation.com/content/11/1/134

showed low LDH release (Figure 3A). Ev en following treat -

ment with 1 μg/ml LPS, LDH levels in the culture superna-

tants remained low. Additionally, BV2 microglia showed

negligible PI staining at 24 (data not sh own) and 48 hrs

post-culture in 3D colla gen (Figure 3B).

CD40 expression of microglia cultured in the

three-dimensional c ollagen matrix

We routinely assess expression of the co-stimulatory

molecule CD40 as a measure of the activation status for

microglia. Upon exposure to lipopolysaccharide (LPS),

microglia secrete inflammatory mediators and upregulate

expression of major histocompatibility complex (MHC)

class II receptors and CD40 to facilitate antigen presenta-

tion to T lymphocytes [19]. Due to these effects, LPS is a

common stimulus for activating microglia in vitro [20-22],

including for our previous work [23-25].

The number of microglia expressing basal CD40 in 3D

cultures was higher (27.6 ± 14.6%) compared to cells

grown in uncoated monolayer cultures (5.1 ± 3.39%)

(P < .05; Kruskal-Wallis with Dunn’smultiplecom-

parisontest).WhenstimulatedwithLPS,thenumber

of CD40

BV2 microglia increased by 70% (P < .001),

with almost the entire population of BV2 cells cul-

tured in 3D collagen shifting to a CD40

phenotype at

24 hrs (97.8 ± 1.5%; Figure 4A). CD40

BV2 cells in

uncoated and collagen-coated monolayer cultures also in-

creased from 5.1 ± 3.4% and 12.8 ± 11.2% to 60.2 ± 20.0%

and 62.3 ± 11.8%, respectively (Figure 4A; P < .05). The

increase in CD40

cells was sig nificantly higher f or 3D

cultures compared to uncoated and collagen-coated

monolayer cultures (P < .05; Kruskal-Wallis with Dunn’s

multiple comparison test). Achie ving a hom ogenous

population shift may indicate a uniform le vel of activa-

tion amongst the microglia , an effect that appears to

be rendered by the 3D collagen culture.

The median fluorescence intensity (MFI) readouts on

a flow cytometer indicate whether cells within a positive

population express the marker in question at different

intensities. Using MFI, we show that not only do the

Figure 2 Morphology of microglia cultured in monolayer and three-dimensional (3D) systems. Microglia cultured in monolayer (A), coated

monolayer (B) and 3D collagen gels (C,D,E,F) were stained with fluorescein isothiocyanate (FITC)-tagged lectin and 4',6-diamidino-2-phenylindole,

dihydrochloride (DAPI) and viewed with fluorescent (A-C) or confocal (D-F) microscopy.

Haw et al. Journal of Neuroinflammation 2014, 11:134 Page 5 of 10

http://www.jneuroinflammation.com/content/11/1/134

number of CD40

BV2 microglia increase in 3D culture,

but the degree of their CD40 expression also increases

3-fold compared with cells grown on uncoated mono-

layer and collagen-coated monolayer surfaces (P < .05;

Figure 4B). Also, microglia in monolayer cultures appear

to respond to LPS by acquiring CD40 expression, and

not by increasing the level of expression (Figure 4B).

Collectively, microglia in 3D culture appear to be acti-

vated in a more homogenous fashion, and to a greater

extent than cells in monolayer cultures. Interestingly,

the intensity of CD40 expression in unstimulated cul-

tures remained similar between all three culture formats.

Therefore, although the number of CD40

cells is higher

in 3D cultures compared to monolayer, the degree of

expression per cell is similar.

Nitric oxide and inflammatory cytokine expression by

microglia cultured in three dimensions

Disease o r damage within the CNS triggers production

of nitric oxide (NO) by microglia and a strocytes [26].

The inducible nitric oxide synt hase (iNOS) iso form

of NO synthase is responsible for secretion of large,

continuous amounts o f NO by these glial cells during

inflammation, as compared to neuronal NOS (nNOS)

that is constitutively expressed in neurons and is

believed to have a physiological role in the brain [26].

To assess NO production by BV2 microglia in mono-

layer and 3D cultures, cells were seeded in the different

culture formats with equal seeding number per well.

Microglia in monolayer and 3D cultures had negligible

NO production (<3 μM at both 36 and 48 hrs; Figure 5),

indicating that the culture conditions alone do not in-

duce NO expres sion. Following stimulation with LPS,

BV2 microglia in all culture formats showed an induc-

tion of NO expression, approximately 4 to 8 times higher

at 36 hrs and 6 to 10 times higher at 48 hrs compared to

untreated BV2 (P < .05). Negligible NO induction was

detected at 24 hrs post-LPS (data not shown). At 36 hrs,

BV2 grown as uncoated monolayer cultures produced

the highest amount of NO (20.03 ± 0.76 μM), followed

by collagen-coated monolayer cultures (6.59 ± 2.39 μM)

and 3D cultures (10.54 ± 1.46 μM). These levels in-

creased for uncoated monolayer, collagen-coated mono-

layer and 3D culture a t the 48-hr time point to 26.96 ±

1.76 μM, 17.87 ± 2.27 μM and 24.47 ± 1.75 μMNO,

respectively. The NO expression by LPS-treated BV2

cells in 3D cultures was not significantly different com-

pared to microglia cultured in uncoated or collagen-

Figure 3 BV2 microglia are viable in three-dimensional (3D) collagen gels. (A) BV2 microglia in monolayer, coated monolayer and 3D

culture conditions were assessed for lactate dehydrogenase (LDH) activity at 48 hrs. Data are mean ± SD from three independent experiments.

(B) Cells were stained with 4',6-diamidino-2-phenylindole, dihydrochloride (DAPI) and propidium iodide (PI) to assess viability at 48 hrs

post-culture. LPS, lipopolysaccharide.

Haw et al. Journal of Neuroinflammation 2014, 11:134 Page 6 of 10

http://www.jneuroinflammation.com/content/11/1/134

coated monolayer cultures at either time points (P >.05;

Kruskal-Wallis with Dunn’s multiple comparison test).

The response of microglia in 3D cultures to LPS was

also evaluated by determining expression of inflammatory

cytokines mRNA and protein. At 6 hrs, mRNA expression

of all inflammatory cytokines tested (MCP-1, IL-6, IL-1β,

IL-12β and TNF-α) was significantly upregulated in BV2

microglia cultured in 3D, relative to housekeeping gene

expression (Figure 6 and Additional file 3). Similar to

mRNA expression, protein levels for IL-6, TNF-α and

MCP-1 were increased following LPS stimulation in 3D

cultures. At 48 hrs of LPS stimulation, BV2 microglia in

3D cultures re corded signif icantly higher expression of

IL-6, TNF-α and MCP-1 compared to untreated BV2

cells (P < .05; Mann Whitney Test) (Table 1). Levels of

IL-6, TNF-α and MCP-1 increa sed by 1 ,998.3 pg/ml,

1,735.9 pg/ml and 5,119.0 pg/ml res pe ctively compared

to basal (untreated) levels. As we have shown before [23],

levels of IFN-γ,IL-10and IL-12p70 were unaffected by

LPS stimulation. Additionally it was observed that un-

stimulated BV2 microglia in 3D cultures demonstrated

markedly lower MCP-1 levels compared to monolayer

cultures (P < .05; Kruskal-Wallis with Dunn’s multiple

comparison test), which was significantly upregulated with

LPS stimulation. MCP-1 expression for LPS-treated BV2

cells in monolayer and coated monolayer cultures was not

significantly different from unstimulated cultures.

Discussion

Conventional monolayer cell cultures involve the growth

of cells on a plastic surface, often coated with extracellu-

lar matrix proteins such as collagen and laminin to

encourage adherence. Several fundamental disparities

exist between these conventional monolayer cultures

and cells in situ, namely cells cultured in monolayer

grow flat, may receive cues from the stiff matrix, do not

grow in a stratified manner and have only one side

adhering to the plastic surface. This also means that

perfusion of nutrients for the cells only occurs via the

non-adhered side. For highly reactive cells such as the

microglia, these culture conditions could affect their

behaviour. The BV2 microglia cell line is commonly

used for microglia research, is well-characterised and is

routinely studied in our laboratory. To approach in vitro

BV2 microglia cultures in a more relevant manner, we

sought to culture BV2 microglia within a collagen matrix

to mimic the mechanical relationship of these cells with

tissue. We demonstrate that BV2 microglia grown in 3D

collagen are viable, embedded and distributed within

the matrix, with a ramified morphology. The ability of

microglia to grow within collagen matrices also allows

Figure 4 BV2 microglia in three-dimensional (3D) collagen gels

shift as a population to express CD40 in response to lipopoly-

saccharide (LPS) stimulation. BV2 microglia in monolayer, coated

monolayer and 3D culture conditions were activated with 1 μg/ml

LPS (+LPS) and CD40 expression analysed with flow cytometry at

24 hrs. (A) Histograms show percentage of CD40

cells. Data are

mean ± SD from three independent experiments. **P < .01, ****P

< .001; one-tailed Mann-Whitney U test. (B) Histograms show median

fluorescence intensity (MFI) of CD40

. Data are mean ± SD from three

independent experiments. *P <.05,**P < .01 ; Kruskal Wallis with Dunn’s

Multiple Comparison test.

Figure 5 BV2 microglia cultured in three-dimensional (3D)

collagen gels express nitric oxide (NO) in response to

lipopolysaccharide (LPS) stimulation. BV2 microglia in monolayer,

coated monolayer and 3D culture conditions were activated with

1 μg/ml LPS and assessed for NO

−

expression at 36 and 48 hrs with the

Griess assay. Data are mean ± SD from three independent experiments.

Haw et al. Journal of Neuroinflammation 2014, 11:134 Page 7 of 10

http://www.jneuroinflammation.com/content/11/1/134

for multiplanar projections of the cytoplasm, which

would be impossible to achieve with conventional mono-

layer cultures.

Using lipopolysaccharide (LPS), we developed an acti-

vation model for BV2 microglia cultured in 3D. LPS is a

bacterial cell wall component that triggers microglia

activation via Toll-like receptor 4 [19]. With 1 μg/ml

LPS, a dose ro utinely used to activate micr oglia in con-

ventional monolayer cultures [20,24,25], BV2 microglia

in 3D cultures significantly increa sed expression of the

cell surface re ceptor CD40, nitric oxide (NO), IL-6,

TNF-α and MC P - 1. This demonstrates that both cells

and supernatant of 3D collagen cultures can be assayed,

demonstrating the suitability of this cul ture system to

accommodate a range of tests. For CD40 expression,

microglia cultured in 3D show ed a higher number of

cells with ba sal e xpression of CD40 compared to the

monolayer cultures. The NO levels among a ll three

cultures in unstimulated conditions, however, were

similar, indicating that collagen alone (in monolayer

or 3D format s) does not trigger NO product ion in

unstimulated BV2 cells. The extent of CD40 expression

per cell (shown by MFI readings) was also similar

between microglia cultured in monolayer and 3D.

Importantly, LPS stimulation triggers a uniform upreg-

ulation of CD40 in 3D cultures compared to cells

cultured in monolayer formats, with the entire population

of microglia acquiring CD40 expression. Therefore, it

appears that BV2 microglia cultured in 3 D acquire

CD40 towards LPS in a more homogenous manner. The

microglia cultured in 3D also expressed significant

amounts of NO in response to LPS.

Beyond the parameters assayed here, the 3D collagen

culture model may be appropriate for studying micro-

glial deposition of ECM material as it offers a matrix of

simple composition compared to other more complex

substrates such as Matrigel™ [14]. With the data now to

show that BV2 microglia are activated by LPS within 3D

collagen cultures, we are keen to utilise this model for

our main research approach of modulating inflammatory

responses of microglia. The growth of microglia within

an environment that is more physically and spatially

relevant than conventional flat plastic culture plates,

along with the ability to stimulate the microglia into an

activated phenotype, gives us access to a more refined

in vitro tool for microglia research.

Conclusions

By cultur ing micro glia w ithin a simple matrix , we offer

amorerelevantin v itro model compared to conven-

tional monolayer cultures where microglia grow flat on

a plastic surface. BV2 micro glia cultured in 3D co lla-

gen constructs render cells that are viable, well distrib-

uted within the matrix, ramified in morphology and

activated into an inflammatory phenotype following

LPS stimulation.

Figure 6 Reverse-transcriptase quantitative PCR ( RT-qPCR) demonstrates the level of inflammatory cytokines by BV2 microglia 6 hrs

after lipopolysaccharide (LPS) stimulation. The values have been normalised to three housekeeping genes (Hmbs, Pgk1 and Psmb2). TNF,

MCP-1, IL-1b, IL-6 and IL-12b were upregulated in stimulated BV2 microglia for three-dimensional (3D) cultures. Data are mean ± SEM from three

independent experiments. *P < .05, **P < .01, ***P < .001, ****P < .0001; Welch’s t test.

Table 1 BV2 microglia in three-dimensional (3D) cultures show increased expression of IL-6, TNF-α and MCP-1 at

48 hours post-lipopolysaccharide (LPS) stimulation

Untreated LPS-treated

IL-6 TNF-α MCP-1 IL-6 TNF-α MCP-1

Monolayer 1.7 ± 0.3 252.3 ± 120.5 9 447.0 ± 226.8 2 476.0 ± 833.6** 3 823.0 ± 1189.0** 10 459.0 ± 1, 639.0

Coated monolayer 1.6 ± 0.99 217.6 ± 181.8 9 231.0 ± 2268.0 1 764.0 ± 324.3** 2 357.0 ± 500.9** 10 936.0 ± 1292.0

3D 0.7 ± 0.8 8.1 ± 3.1 284.0 ± 73.5 1 999.0 ± 685.2* 1 744.0 ± 911.6* 5 403.0 ± 517.6*

Values are expressed in pg/ml, and assayed using the Cytometric Bead Array. Data are mean ± SD from five independent experiments. *p < .05, **p < .01;

two-tailed Mann-Whitney U test, as compared to untreated controls. ns, not significant.

Haw et al. Journal of Neuroinflammation 2014, 11:134 Page 8 of 10

http://www.jneuroinflammation.com/content/11/1/134

Additional files

Additional file 1: List of primers and UPL probes used for

reverse-transcriptase quantitative PCR (RT-qPCR) validations.

Additional file 2: Video demonstrating three-dimensional (3D)

image of a BV2 microglia with the 3D collagen gel.

Additional file 3: Reverse-transcriptase quantitative PCR (RT-qPCR)

analyses of TNF, MCP-1, IL-b, IL-12b and IL-6 mRNA.

Abbreviations

CNS: Central nervous system; DAPI: 4',6-diamidino-2-phenylindole,

dihydrochloride; ECM: Extracellular matrix; FITC: Fluorescein isothiocyanate;

iNOS: inducible nitric oxide synthase; LDH: Lactate dehydrogenase;

LPS: Lipopolysaccha ride; MFI: Median fluorescence intensity; nNOS: neuronal

nitric oxide synthase; NO: Nitric oxide; OD: Optical density;

PFA: Paraformaldehyde; PI: Propidium iodide; 3D: three-dimensional.

Competing interests

The authors declare that they have no competing interests.

Authors’ contributions

RHTY performed the cell cultures, LDH assay, Griess assay and cytokine bead

array. TCK established the 3D collagen cultures, performed the confocal

microscopy, immunophenotyping and viability staining. AY performed the

scanning electron microscopy. LHC performed the RT-qPCR experiments. SV

and JBP conceptualised the study. SV wrote the manuscript. All authors

analysed the data and read and approved the final manuscript.

Acknowledgements

The authors would like to thank Hi-Tech Instruments Sdn. Bhd. and the Brain

Research Institute Monash Unive rsity Sunway for use of their confocal

microscope, staff of the Electron Micro scopy Unit, Universiti Putra Malaysia

for assistance with the electron microscopy, and Dr. Michael Ling King Hwa

for his expert advice on RT-qPCR. This study was funded by the Research

University Grant Scheme (UPM) [04-02-12-1796RU] and the Exploratory Research

Grant Scheme (Ministry of Higher Education Malaysia) [ERGS/1/2012/5527106].

Randy Haw Tatt Yhew, Tong Chih Kong an d Lee Ha n Chung are supp orted

by MyBr ain15 postgraduate scholarship programmes by the Ministry of

Education (MOE), Malaysia.

Author details

Neuroinflammation Group, Immunology Laboratory, Department of

Pathology, Faculty of Medicine & Health Sciences, Universiti Putra Malaysia,

43400 Serdang, Selangor, Malaysia.

Genetic & Regenerative Medicine

Research Centre (GRMRC) & Department of Obstetrics & Gynaecology,

Faculty of Medicine & Health Sciences, Universiti Putra Malaysia, 43400 UPM

Serdang, Selangor, Malaysia.

Department of Biomaterials & Tissue

Engineering, University College London, UCL Eastman Dental Institute,

256 Gray’s Inn Road, London WC1X 8LD, UK.

Received: 16 December 2013 Accepted: 16 July 2014

Published: 30 July 2014

References

1. Ginhoux F, Greter M, Leboeuf M, Nandi S, See P, Gokhan S, Mehler MF,

Conway SJ, Ng LG, Stanley ER, Samokhvalov IM, Merad M: Fate mapping

analysis reveals that adult microglia derive from primitive macrophages.

Science 2010, 330:841–845.

2. Kierdorf K, Erny D, Goldmann T, Sander V, Schulz C, Perdiguero EG,

Wieghofer P, Heinrich A, Riemke P, Holscher C, Muller DN, Luckow B,

Brocker T, Debowski K, Fritz G, Opdenakker G, Diefenbach A, Biber K,

Heikenwalder M, Geissmann F, Rosenbauer F, Prinz M: Microglia emerge

from erythromyeloid precursors via Pu.1- and Irf8-dependent pathways.

Nat Neurosci 2013, 16:273–280.

3. Ransohoff RM, Cardona AE: The myeloid cells of the central nervous

system parenchyma. Nature 2010, 468:253–262.

4. Nimmerjahn A, Kirchhoff F, Helmchen F: Resting microglial cells are highly

dynamic surveillants of brain parenchyma in vivo. Science 2005,

308:1314–1318.

5. Kettenmann H, Hanisch UK, Noda M, Verkhratsky A: Physiology of

microglia. Physiol Rev 2011, 91:461–553.

6. Swinnen N, Smolders S, Avila A, Notelaers K, Paesen R, Ameloot M, Brone B,

Legendre P, Rigo JM: Complex invasion pattern of the cerebral cortex

bymicroglial cells during development of the mouse embryo. Glia 2013,

61:150–163.

7. Paolicelli RC, Bolasco G, Pagani F, Maggi L, Scianni M, Panzanelli P, Giustetto

M, Ferreira TA, Guiducci E, Dumas L, Ragozzino D, Gross CT: Synaptic

pruning by microglia is necessary for normal brain development. Science

2011, 333:1456–1458.

8. Schafer DP, Lehrman EK, Kautzman AG, Koyama R, Mardinly AR, Yamasaki R,

Ransohoff RM, Greenberg ME, Barres BA, Stevens B: Microglia sculpt

postnatal neural circuits in an activity and complement-dependent

manner. Neuron 2012, 74:691–705.

9. Sierra A, Encinas JM, Deudero JJ, Chancey JH, Enikolopov G,

Overstreet-Wadiche LS, Tsirka SE, Maletic-Savatic M: Microglia shape adult

hippocampal neurogenesis through apoptosis-coupled phagocytosis.

Cell Stem Cell 2010, 7:483–495.

10. Chuang HN, Van Rossum D, Sieger D, Siam L, Klemm F, Bleckmann A,

Bayerlova M, Farhat K, Scheffel J, Schulz M, Dehghani F, Stadelmann C,

Hanisch UK, Binder C, Pukrop T: Carcinoma cells misuse the host tissue

damage response to invade the brain. Glia 2013, 61:1331–1346.

11. Krabbe G, Halle A, Matyash V, Rinnenthal JL, Eom GD, Bernhardt U, Miller KR,

Prokop S, Kettenmann H, Heppner FL: Functional impairment of microglia

coincides with Beta-amyloid deposition in mice with Alzheimer-like

pathology. PLoS One 2013, 8:e60921.

12. Brown RA: In the beginning there were soft collagen-cell gels:

towards better 3D connective tissue models? Exp Cell Res 2013,

319:2460–24 69.

13. East E, De Oliveira DB, Golding JP, Phillips JB: Alignment of astrocytes

increases neuronal growth in three-dimensional collagen gels and is

maintained following plastic compression to form a spinal cord repair

conduit. Tissue Eng Part A 2010, 16:3173–3184.

14. East E, Golding JP, Phillips JB: A versatile 3D culture model facilitates

monitoring of astrocytes undergoing reactive gliosis. J Tissue Eng Regen

Med 2009, 3:634–646.

15. Phillips JB, Brown R: Structured Materials and Mechanical Cues in 3D

Collagen Gels. In 3D Cell Culture: Methods and Protocols, Methods in

Molecular Biology. Edited by Haycock JW. New York Dordrecht Heidelberg

London: Springer Science and Business Media; 2011:183–196.

16. Blasi E, Barluzzi R, Bocchini V, Mazzolla R, Bistoni F: Immortalization of murine

microglial cells by a v-raf/v-myc carrying retrovirus. JNeuroimmunol1990,

27:229–237.

17. Lizarraga F, Poincloux R, Romao M, Montagnac G, Le Dez G, Bonne I, Rigaill

G, Raposo G, Chavrier P: Diaphanous-related formins are required for

invadopodia formation and invasion of breast tumor cells. Cancer Res

2009, 69:2792–2800.

18. Hewitt CA, Ling KH, Merson TD, Simpson KM, Ritchie ME, King SL, Pritchard

MA, Smyth GK, Thomas T, Scott HS, Voss AK: Gene network disruptions

and neurogenesis defects in the adult Ts1Cje mouse model of Down

syndrome. PLoS One 2010, 5:e11561.

19. Olson JK, Miller SD: Microglia initiate central nervous system innate and

adaptive immune responses through multiple TLRs. J Immunol 2004,

173:3916–3924.

20. Dheen ST, Jun Y, Yan Z, Tay SS, Ling EA: Retinoic acid inhibits

expression of TNF-alpha and iNOS in activated rat microglia.

Glia 2005, 50:21–31.

21. Kim J, Hematti P: Mesenchymal stem cell-educated macrophages: a novel

type of alternatively activated macrophages. Exp Hematol 2009,

37:1445–1453.

22. Loane DJ, Stoica BA, Pajoohesh-Ganji A, Byrnes KR, Faden AI: Activation of

metabotropic glutamate receptor 5 (mGLUR5) modulates microglial

reactivity and neurotoxicity by inhibiting NADPH oxidase. J Biol Chem

2009, 284:15629–15639.

23. Ooi YY, Ramasamy R, Rahmat Z, Subramaiam H, Tan SW, Abdullah M,

Israf DA, Vidyadaran S: Bone marrow-derived mesenchymal stem cells

modulate BV2 microglia responses to lipopolysaccharide.

Int Immunopharmacol 2010, 10:1532–1540.

24. Rahmat Z, Jose S, Ramasamy R, Vidyadaran S: Reciprocal interactions of

mouse bone marrow-derived mesenchymal stem cells and BV2 microglia

after lipopolysaccharide stimulation. Stem Cell Res Ther 2013, 4:12.

Haw et al. Journal of Neuroinflammation 2014, 11:134 Page 9 of 10

http://www.jneuroinflammation.com/content/11/1/134

25. Vidyadaran S, Ooi YY, Subramaiam H, Badiei A, Abdullah M, Ramasamy R,

Seow HF: Effects of macrophage colony-stimulating factor on microglial

responses to lipopolysaccharide and beta amyloid. Cell Immunol 2009,

259:105–110.

26. Brown GC, Bal-Price A: Inflammatory neurodegeneration mediated by

nitric oxide, glutamate, and mitochondria. Mol Neurobiol 2003,

27:325–355.

doi:10.1186/1742-2094-11-134

Cite this article as: Haw et al.: A three-dimensional collagen construct to

model lipopolysaccharide-induced activation of BV2 microglia. Journal of

Neuroinflammation 2014 11:134.

Submit your next manuscript to BioMed Central

and take full advantage of:

• Convenient online submission

• Thorough peer review

• No space constraints or color ﬁgure charges

• Immediate publication on acceptance

• Inclusion in PubMed, CAS, Scopus and Google Scholar

• Research which is freely available for redistribution

Submit your manuscript at

www.biomedcentral.com/submit

Haw et al. Journal of Neuroinflammation 2014, 11:134 Page 10 of 10

http://www.jneuroinflammation.com/content/11/1/134