Cytosolic phospholipase A2-driven arachidonic acid mobilization and PGE2 synthesis

within newly assembled lipid bodies of epithelial cells Luciana S. Moreira1; Bruno Piva1; Fabio Mesquita-Santos1; Clarissa M. Maya-Monteiro2; Patrícia T. Bozza2; Christianne Bandeira-Melo1; and Bruno L. Diaz1* 1Laboratório de Inflamação, Instituto de Biofísica Carlos Chagas Filho, Universidade

Federal do Rio de Janeiro, Rio de Janeiro, RJ, Brazil. 2Laboratório de Imunofarmacologia,

Departamento de Fisiologia e Farmacodinâmica, Instituto Oswaldo Cruz, Fundação

Oswaldo Cruz, Rio de Janeiro, RJ, Brazil.

Running title: Epithelial cells lipid bodies control AA mobilization.

This work was supported by Conselho Nacional de Pesquisa (CNPq, Brazil), PROFIX (to

CBM and BLD), Fundação de Amparo à Pesquisa do Rio de Janeiro (FAPERJ, Brazil) and

Howard Hughes Medical Institute (to PTB).

*Author for correspondence at: 1Laboratório de Inflamação, Instituto de Biofísica Carlos

Chagas Filho, Universidade Federal do Rio de Janeiro, Av. Carlos Chagas Filho room C1-

024, Rio de Janeiro, RJ, Brazil CEP 21941-902

FAX: (5521) 2280-8193 Phone: (5521) 2562-6509 E-mail: bldiaz@biof.ufrj.br

2

Abstract

Cytoplasmic lipid bodies are intracellular deposits of arachidonic acid (AA), which can

be metabolized for eicosanoid generation. PGE2 is a major AA metabolite produced by

epithelial cells and can modulate restoration of epithelium homeostasis after injury. We

studied lipid body biogenesis and their role in AA metabolic pathway in an epithelial cell

line derived from normal rat intestinal epithelium, IEC-6 cells. Lipid bodies were virtually

absent in confluent IEC-6 cells. Stimulation of confluent IEC-6 cells with unsaturated fatty

acids, such as AA or oleic acid (OA), induced rapid lipid body assembly that was

independent on its metabolism to PGE2 or direct incorporation of exogenous fatty acid into

nascent lipid bodies, but dependent on signaling through p38, PKC, and PI3K. Newly

formed lipid bodies compartmentalized cytosolic phospholipase (cPL)A2-α, facilitating AA

mobilization and synthesis of PGE2 within epithelial cells. Thus, both highly regulated

biogenesis of lipid bodies and their functions on cPLA2-α-driven enhanced AA

mobilization and PGE2 production may have key roles in epithelial cell-driven

inflammatory functions, and may represent relevant therapeutic targets of epithelial

pathologies.

Key Words: Lipid droplets, epithelial cells, cPLA2-α, PGE2, arachidonic acid, oleic acid, p38 MAP kinase

3

Introduction

Restoration of epithelium homeostasis after injury, as well as, cancer-related

cellular alterations depend on dynamic epithelial cell events ranging from apoptosis and

cell proliferation to migration that can be modulated by a variety of inflammatory

mediators, like arachidonic acid (AA) and its metabolites, the eicosanoids [7002]. It is now

recognized that non-myeloid cells, such as epithelial cells [7003], can mobilize esterified

AA and further metabolize it. AA is a key signaling molecule acting either as intracellular

second messenger, as paracrine mediator of cell activation or as substrate for enzymatic

conversion into eicosanoids [858,869,6999]. Therefore, autocrine/paracrine or even

intracrine activities of AA or its metabolites may also control epithelial cell functions.

However, the specific regulatory mechanisms controlling AA mobilization or its putative

agonistic effects within epithelial cells are poorly defined.

The first step in eicosanoid generation within any cell type is performed by a

superfamily of enzymes collectively known as phopholipase (PL)A2 due to their ability to

release fatty acids from sn-2 position of phospholipids [1670]. Of the more than twenty

different PLA2s described so far, the cytosolic PLA2-α, also know as group IVA PLA2 (for

a review of PLA2 nomenclature see Schaloske and Dennis, 2006 [6356]), is the most

extensively studied and was ascribed a central role in arachidonic acid (AA) mobilization

[6151]. cPLA2-α is a cytosolic enzyme that upon cell activation translocates to have access

to its selective AA-containing phospholipid substrate and to the vicinity of downstream

4

enzymes that will further metabolize the newly released AA [7005,456]. This is a highly

regulated process that depends on a rise in cytosolic calcium levels that allow cPLA2-α to

be targeted to cellular membranes mediated by its C2 domain [972]. Several stimuli that do

not induce calcium transients may also activate cPLA2-α through phosphorylation

providing an alternative regulatory pathway [7006,974,3395,770]. In addition to kinases-

driven regulation, there is now an evolving understanding of the key regulatory roles played

by distinct intracellular compartments involved in AA release and metabolism.

Lipid bodies (also known as lipid droplets) are osmiophilic organelles, recognized

as dynamic structures with key roles in regulating storage and turnover of lipids in virtually

all mammalian cells. These cytoplasmic structures are comprised of an outer monolayer of

phospholipids, have a neutral lipid-rich core and a unique and variable fatty acid and

protein composition depending on type and activating status of cell (for review see [6586]).

Characterization of the lipid bodies in different cell types has shown that they are

particularly active sites for metabolism of arachidonoyl lipids [7007,7008,7009,7010],

while Yu et al. (1998) [7010] demonstrated the compartmentalization of cPLA2 at

arachidonate enriched-leukocyte lipid bodies. Lipid bodies are also inducible organelles.

Rather than a simple cis-fatty acid esterification phenomenon, studies investigating the

mechanisms of lipid body biogenesis demonstrated that new lipid bodies are assembled in

ER membranes, by a complex biogenic process that is rapid, but highly regulated, and

stimulus- and cell-specific [6586].

At least for leukocytes, lipid body biogenesis and lipid/protein

compartmentalization appear to be mutually influenced. Based on that, it has been

postulated that for a specific cell type, distinct lipid body biogenic events may determine

5

lipid body composition, and therefore their functional outcomes. It has been reported that a

variety of epithelial cell types display lipid bodies within their cytoplasm [7051,7049],

Plotkwoski et al., in press). Here, we investigated the molecular mechanisms involved in

triggering lipid body assembly within epithelial cells and their functions in cPLA2-driven

AA-mobilizing activity.

6

Methods

Reagents

IL-1 pyrrolidine and PTX

Arachidonic acid (AA), Oleic acid (OA), non-esterifiable arachidonic acid analogue,

methyl arachidonyl trifluoromethyl ketone (ATK), and Prostaglandin (PG)E2 were

purchased from Cayman Chemical Co. (Ann Arbor, MI). Platelet activating factor (PAF);

Palmitic acid (PA); and Stearic acid (SA) were purchased from Sigma Chemical Co. (St.

Louis, MO). All lipids were diluted in absolute ethanol (Merck), with exception of PAF

that was diluted in PBS/BSA 0.1%. The inhibitors for COX-2, NS-398, and COX-1,

valeroyl salicylate were purchased from Cayman. MAP kinase inhibitors SB202190 (p38),

SP600125 (JNK), and U0126 (MEK1/2) were all from BIOMOL (Plymouth Meeting, PA)

and diluted in DMSO (Sigma). [5,6,8,9,11,12,14,15-3H] Arachidonic Acid with specific

activity of 214 Ci/mmol was obtained from Amersham Biosciences (Buckinghamshire,

England). Monoclonal antibodies for immunoblot assays were mouse IgG2b anti-cPLA2-α

(clone 4-4B-3C) at 1 µg/ml, and mouse IgG1 anti-GAPDH (clone 6C5) at 3 ng/ml, both

from Santa Cruz Biotechnology (Santa Cruz, CA). Rabbit polyclonal antibody against

phosphorylated p38 (pTpY 180/182) was purchased from Biomol and used at 1:2000

dilution. Secondary antibodies were goat polyclonal against mouse IgG (Santa Cruz) or

rabbit IgG (Jackson ImmunoResearch Laboratories; West Grove, PA). The goat HRP-

linked secondary antibodies were anti-mouse IgG (Santa Cruz) diluted at 1:4,000, and anti-

rabbit IgG from Jackson ImmunoResearch Laboratories diluted 1:10,000. Bodipy® 493/503

was from Molecular Probes.

Cell culture and treatments

7

IEC-6 [2291], a non-transformed rat intestinal epithelia-derived cell line (Rio de Janeiro

Cell Bank, Brazil), was maintained in Dulbecco Modified Eagle medium (DMEM),

supplemented with 5% fetal bovine serum (FBS, Cultilab, Brazil), 100 U/ml penicillin, and

100 µg/ml streptomycin (all cell culture reagents from Invitrogen, Carlsbad, CA) in culture

flasks (TPP, Switzerland). Cells were collected by 0.25% trypsin and 0.38 g/L EDTA in

HBSS without Ca++ and Mg++ and 1 x 105 cells/well were plated over glass coverslips in

24-well plates (TPP) for lipid body enumerations, and 2 x105 cells/well in 6-well plates for

all other experiments. Cells were grown in supplemented DMEM at 37oC in a 5% CO2

atmosphere and all experiments were performed after cultures reached confluency (2-3

days after plating). All pharmacological inhibitors were added after culture medium was

changed and 30 min before stimulation. Cells were stimulated for different periods with

PAF (1 µM), AA (1-30 µM), IL-1 (??), OA (10 µM), PGE2 (10 µM), PA (10 µM), SA (10

µM), in the presence of inhibitors or vehicle. In each experiment, DMSO or ethanol was

always below 0.1% and did not modify cell activation when compared to untreated cells.

Lipid Body Staining and Enumeration

Analysis of lipid body numbers was performed in osmium-stained cells. Briefly, while still

moist, confluent IEC cells onto slides were fixed in 3.7% formaldehyde in Ca2+/Mg2+- free

HBSS, pH 7.4 rinsed in 0.1 M cacodylate buffer 1.5% OsO4 (30 min), rinsed in dH2O,

immersed in 1.0% thiocarbohydazide (5 min), rinsed in 0.1 M cacodylate buffer, restained

in 1.5% OsO4 (3 min), rinsed in dH2O, and then dried and mounted. The morphology of

fixed cells was observed, and lipid bodies were enumerated by light microscopy with a

100X objective lens in 50 consecutively scanned cells.

8

Alternatively, the measurement of the area of lipid bodies used as a different

approach for lipid body quantification, was performed by staining IEC-6 cells with

BODIPY® 493/503 (4,4-difluoro-1,3,5,7,8-pentamethyl-4-bora-3a,4a-diaza-s-indacene).

BODIPY staining shows the accumulation of neutral lipids in lipid bodies. Images were

obtained using an Olympus BX-FLA fluorescence microscope equipped with a Plan Apo

60× objective (Olympus Optical CO., Japan) and CoolSNAP-Pro CF digital camera in

conjunction with Image-Pro Plus version 4.5.1.3 software (MediaCybernetics, San Diego,

CA). The images (at least four fields per slide), were transformed into black and white

pictures and analyzed with Image 2D (GE Healthcare). The BODIPY-labeled cytoplasmic

spots (lipid bodies) were determined by automatic spot detection, and the total area of

fluorescent spots was obtained for each field and divided by the number of cells in the

respective field. Values were expressed as lipid body fluorescent area/cell.

Arachidonic Acid Release

Determination of AA release was performed as previously described [770] with minor

modifications. Briefly, IEC-6 cells were collected and plated at 2 x 105 cells per well in 6-

well tissue culture plates in 2 mL of supplemented DMEM, as described above. After 2

days cells were labeled for 18 h with 1 µCi/well of [3H]-AA in 2 mL of supplemented

DMEM. After labeling, IEC-6 cells were washed twice with PBS and IEC-6 cells in fresh

medium were stimulated with OA (10 - 30 µM) for 4 h for induction of lipid bodies. After

cell culture medium was replaced, lipid body-bearing IEC-6 cells were then challenged for

1 h with 5 µM A23187 for stimulation of AA release. Culture supernatants were then

harvested and centrifuged at 200 g for 10 min, while cells were washed once with PBS and

lysed in 2 ml of 0.1% Triton X (Sigma). Samples were stored at –20oC until 0.2 ml of

9

sample were mixed with 1.6 ml of Ultima Gold scintillation cocktail (Perkin Elmer,

Shelton, CT) and then were analyzed in a LS6000 LL scintillation counter (Beckman,

Fullerton, CA). The percentage of arachidonic acid released for each condition was

calculated by supernatant radioactivity counts divided by total radioactivity (supernatant

plus cellular lysate) counts multiplied by 100.

Immunoblot analysis.

Cells were collected after stimulation by cell scraper (COSTAR) and addition of 120 µL of

sample buffer (1% Triton X; 0.5% sodium deoxycholate; 0.2% SDS; 150 mM NaCl; 10

mM HEPES; 2 mM EDTA; 2 mM sodium orthovanadate; 20 mM NaF; 5 µg/ml Pepstatin;

10 µg/ml Leupeptin; 1 µg/ml Aprotinin; 1mM PMSF). 40 µl of loading buffer (20% β-

mercaptoethanol; 370 mM Tris base; 160 µM Bromophenol blue; 6% glycerol; 16% SDS;

pH 6.8) was added to the cell lysate. Then, cellular lysates were immediately heated at

100oC for 5 minutes. Samples were resolved by electrophoresis on a SDS-PAGE 10%

polyacrylamide gel at 29 mA/gel for 1 hour. After electrophoresis, separated proteins were

transferred to nitrocellulose membranes (Santa Cruz) at 250 mA for 2 hours at 4oC.

Membranes were blocked with TTBS (10 mM Tris; 150 mM NaCl; 0.1% Tween 20; pH

7,4) and 5% of non-fat dry milk for 2 hours at room temperature. Membranes were washed

5X with TTBS and incubated with primary antibodies (above) diluted in TTBS for 16 h at

4oC under gentle agitation. After washing 5X with TTBS membranes were incubated with

secondary antibodies for 1 h at room temperature. ECL Western Blotting chemiluminescent

substrate (Amersham) was added after membranes were washed as before and bands were

visualized after exposure of Biomax Light autoradiography film (Kodak).

cPLA2 immuno-detection

10

Stimulated and non-stimulated confluent cultures of IEC-6 cells on glass coverslips

were fixed in 3.7 % formaldehyde for 10 min, permeabilized with 0.2 % Triton X-100/PBS

solution (5 min). After washing, slides were incubated for 1 h with 40 µg/ml mouse IgG2b

anti-cPLA2-α monoclonal antibody (Santa Cruz) diluted in 0.2 % Triton X-100/PBS

solution. Same concentration of irrelevant mouse IgG2b was used as control. After three

washes of 5 min, the preparations were incubated 1 h with 1:3,000 dilution of Cy3-labeled

donkey anti-mouse IgG secondary antibody. After washing with PBS for 10 min (3x),

BODIPY (4 µg/ml) was added for 5 min to label cytoplasmic lipid bodies within IEC-6

cells. Coverslips were then washed with HBSS, and an aqueous mounting medium

(Polysciences, Warrington, PA) was applied to each slide before cover-slip attachment.

Images were obtained and analyzed as described above, but using a 100X objective. Images

were edited using Adobe Photoshop 5.5 software (Adobe Systems, San Jose, CA).

Statistical analysis.

Data are expressed as mean ± standard error of the mean (SEM) of at least three

independent experiments. Multiple comparisons among groups were performed by one-

way ANOVA followed by Bonferroni’s or Dunnett’s test. * and # represent p values < 0.05

when compared to control non-stimulated group or stimulated group respectively and were

considered statistically significant.

11

Results

Epithelial cells assemble new lipid bodies under AA stimulation

In order to investigate whether epithelial cells can mount an inducible lipid body

biogenic process during inflammatory conditions, we have directly stimulated in vitro a

non-transformed rat intestinal epithelial cell line (IEC-6 cells) with some lipid body-

relevant inflammatory mediators [ref]. As illustrated in Figure 1A (top panel), confluent

IEC-6 cells stained with osmium showed a translucent cytoplasm with virtually no

assembled lipid body. In clear contrast, confluent IEC-6 cells stimulated with AA (10 µM)

for 4 h, showed a cytoplasm packed with osmiophilic organelles (seen as dark punctate

structures), named lipid bodies (Figure 1A, bottom panel). Figure 1B shows a different

technical approach to visualize and quantify lipid body biogenesis. Similar to osmium-

stained cells, BODIPY-labeled non-stimulated IEC-6 cells showed a clear cytoplasm, while

cells stimulated with AA (10 µM) showed an increased numbers of cytoplasmic lipid

bodies that appeared as green fluorescent dots (Figure 1B).

Enumeration of osmium-stained dark inclusions within confluent IEC-6 cells

showed that cells stimulated with AA (10 µM) contain at least 15 times more lipid bodies

than non-stimulated cells (Figure 1C). Densitometric image analysis of BODIPY-labeled

cells, as a measure of lipid body area and neutral lipid accumulation per cell, confirmed the

stimulatory effect of AA in inducing lipid body biogenesis within IEC-6 cells (Figure 1D).

These findings indicate that epithelial cells can respond with lipid body biogenesis, leading

to an increased content of these organelles within their cytoplasm. AA (1 to 30 µM)-

induced increase in the numbers of cytoplasmic lipid bodies within confluent IEC-6 cells

was a dose-dependent (Figure 2A) and rapid phenomenon, which was apparent within 1 h,

12

significant within 2 h and maximum within 4 h at the concentration of 10 µM (Figure 2B).

Within 8 h of AA stimulation, newly formed lipid bodies started to disappear, even though

cell viability remained high (data not shown). Confirming that assembly of new lipid

bodies is a cell- and stimulus-specific phenomenon, PAF (1 µM) – a known agonist of lipid

body biogenesis in leukocytes (refs) – did not trigger lipid body formation within IEC-6

cells (Figure 1C). IL-1 e LPS

Molecular mechanisms involved in AA-driven assembling of new lipid bodies depend on

agonistic activity and p38 kinase activation.

The biogenic process of lipid bodies represents a complex cellular outcome

triggered by a variety of distinct signaling pathways. Here, our attempts to characterize the

molecular signals committed to AA-induced biogenesis of cytoplasmic lipid bodies within

epithelial cells revealed that the exogenous AA added to IEC-6 cells, rather than a substrate

for enzymatic conversion into eicosanoids or entirely an excess of incorporable fatty acid,

functions in part as a paracrine mediator of cell activation with downstream signaling

through PI3 kinase, PKC and p38.

First, we verified that lipid body formation induced by AA – a cis-unsaturated fatty

acid – did not appear to depend exclusively on direct incorporation of fatty acids into

nascent lipid bodies, but also on a specific agonistic activity that depend on fatty acid

structural characteristics, since: (i) stimulation with either palmitic acid (PA; 10 µM) or

stearic acid (SA; 10 µM) – two saturated fatty acids – did not trigger lipid body assembly

within IEC-6 cells (Table 1), even though they are available for esterification into newly-

assembling lipid bodies (Weller, 1991); (ii) IEC-6 cells stimulated with another cis-

13

unsaturated fatty acid (oleic acid, OA; 10 µM) showed increased numbers of cytoplasmic

lipid bodies (Table 1); and (iii) ATK (10 µM) – a non-esterifiable AA analogue – did

induce lipid body biogenesis within confluent IEC-6 cells (not shown).

Second, the hypothesis that endogenous PGE2 synthesized from stimulatory AA was

mediating AA-induced lipid body assembly within IEC-6 cells was ruled out since: (i) pre-

treatments with selective inhibitors of either COX-1 (valeroyl salycilate; 10 µM) or COX2

(NS-398; 1 µM) failed to affect AA-induced lipid body biogenesis within IEC-6 cells

(Figure 3A); (ii) stimulation with exogenous PGE2 (1 to 10 µM) was unable to increase

cytoplasmic numbers of lipid bodies within IEC-6 cells (Figure 3B); (iii) stimulation with

AA (10 µM) did not induce, per se, release of PGE2 within supernatants of IEC-6 cells (not

shown).

Finally, lipid body biogenesis triggered by AA within epithelial cells appears to be

due a direct induction of cell activation by AA. AA is probably acting as a paracrine

mediator on specific receptors, initiating a rapid but specific intracellular signaling that

leads to lipid body biogenesis, since: (i) AA-induced lipid body biogenesis within IEC-6

cells was significantly inhibited by the pre-treatment with a specific MAP kinase p38

inhibitor (SB600125) (Figure 4A); (ii) such p38-dependent lipid body biogenesis was

preceded by an acute (apparent within 1 min) and transient phosphorilation of p38 detected

after stimulation with AA (10 µM) (Figure 4B); (iii) two other MAP kinases, ERK1/2 and

JNK, are not involved in AA-induced lipid body assembly, since inhibitors (U0126 and

SP600125, respectively) did not alter AA-induced lipid body formation; and (iii) while

inhibitors of tyrosine kinases (genistein) (Table 2) did not alter AA-induced rapid lipid

body biogenesis, inhibitors of PI3 kinase (LY294002) and PKC (calphostin C) blocked

14

formation of lipid bodies induced by AA within IEC-6 cells (Table 2). Of note, all

pharmacological inhibitors of kinases used (LY294002, calphostin C, genistein, U0126,

SP600125 and SB202190) did not alter the cytoplasmic basal numbers of lipid bodies

found in non-stimulated IEC-6 cells (not shown). Therefore, distinctively in AA-stimulated

IEC-6 cells, rapid intracellular signaling events, comprising at least p38, PI3 kinase and

PKC activation, are committed to lipid body biogenesis.

Newly formed lipid bodies function as intracellular platforms of cPLA2-driven mobilization

of endogenous AA within epithelial cells.

The potential functions of the newly assembled lipid bodies within epithelial cells were

then investigated. It is well established that at least three distinct intracellular domains may

compartmentalize the molecular organization for eicosanoid synthesis: the nuclear

membrane (22, 23), phagosomes (Arm) and lipid bodies (Pat JEM, Eu JBC, Fabio, Adriana,

Helo, MCP). Do newly formed lipid bodies have functional roles in AA metabolism within

epithelial cells? We hypothesized that in unsaturated fatty acids-stimulated epithelial cells

at least the first step of eicosanoid synthesis takes place within newly formed lipid bodies.

Therefore, we evaluated stimulatory capability of exogenous unsaturated fatty acids to

control the expression, location and function of AA-mobilizing enzyme cPLA2 within IEC-

6 cells. Stimulation with either AA (not shown) or OA (Figure 5B) was unable to increase

cPLA2 expression in IEC-6 cells. Although stimulation with lipid body-triggering

unsaturated fatty acids did not alter total protein content of cPLA2 in IEC-6 cells, AA (not

shown) or OA (Figure 5C) triggered the translocation of cPLA2 from the cytosol to the

newly formed lipid bodies. Figure 5C (top-left panels) shows that in non-stimulated IEC-6

cells immuno-fluorescent cPLA2 was found homogeneously distributed in the cytoplasm

15

that showed no BODIPY-labeled lipid bodies. In contrast, in OA-stimulated (10 µM) IEC-6

cells, which were packed with newly formed lipid bodies (Figure 5A), the cPLA2-α

immuno-fluorescenct staining shifted from IEC-6 cytoplasm to a clear punctate cytoplasmic

pattern, proximate to, but separate from the nucleus, and fully consistent in size and form

with lipid bodies of IEC-6 cells (Figure 5C, bottom panel). Lipid body

compartmentalization of translocated cPLA2-α within OA-stimulated IEC-6 cells was

ascertained by the co-localization with BODIPY (Figure 5B, middle panels). Differently

from the filled punctate immuno-staining of BODIPY, immuno-fluorescent cPLA2-α was

found as an external ring-like immuno-staining, placing cPLA2-α as a surface lipid body

protein. No immuno-reactivity was detected when an irrelevant mouse IgG2B was used as a

control to the anti-cPLA2 antibody in OA-stimulated cells, although the BODIPY-labeled

lipid bodies were strongly visualized (not shown). Of note, A23187 (a calcium ionophore; 5

µM) also induced partial translocation of cPLA2-α within IEC-6 cells. However, unlike in

unsaturated fatty acids-stimulated cells, immuno-fluorescent cPLA2-α shifted from the

cytoplasm to the perinuclear envelope of A23187-stimulated IEC-6 cells (Figure 5C, top-

right panel). Finally, if unsaturated fatty acids-driven newly formed lipid bodies of

epithelial cells are indeed involved in AA metabolism, then the AA esterified in these lipid-

rich structures should be mobilized first by the lipid body-compartmentalized cPLA2-α (as

shown above) to be released from cells. Figure 5D shows that, under proper stimulation,

IEC-6 cells bearing increased numbers of cytoplasmic lipid bodies were able to mobilize

AA. Of note, to avoid potential confounding effects of analyzing AA release in AA-

stimulated cells, in this set of experiments only OA was employed as stimulus. Stimulation

for 4 h with different concentrations of OA (10 – 30 µM) caused a dose-dependent increase

16

of lipid body numbers (Figure 5A) within cytoplasm of IEC-6 cells with stimulatory

potency similar to AA (Figure 2A). In parallel to the increased number of newly assembled

lipid bodies (4 h), OA was also able to prime IEC-6 cells in a concentration-dependent

manner for an enhanced AA mobilization (Figure 5D).

17

Discussion

Lipid body biogenesis is a critical cellular event in whole body lipid metabolic

homeostasis, as well as, to several important human diseases (Murphy). In chronic

inflammatory pathologies, such as cancer (teresa) and Crohn's disease (Beil, 1995),

epithelial cells constantly show increased numbers of cytoplasmic lipid bodies. Advances

on the mechanisms behind the enhanced assembly of lipid bodies within epithelium are of

paramount importance for understanding the pathogenesis of epithelial diseases. Here,

employing confluent cultures of normal intestinal epithelium-derived cell line – IEC-6 cells

– we have unveiled that the biogenic process of lipid bodies within epithelial cells happens

very rapidly and in a highly regulated fashion triggered by relevant inflammatory

mediators. Unsaturated fatty acids, like AA and OA, appeared as potent biogenic stimuli of

cPLA2-α-bearing lipid bodies, placing epithelial lipid bodies as key organelles involved in

mobilization and release of endogenous AA. cPLA2-α-driven free AA was further

metabolized into PGE2 within the newly formed lipid bodies of OA-stimulated epithelial

cells, unveiling a functional impact of these organelles. The mechanism involved in AA-

elicited lipid body biogenic process appeared to be more complex than merely

incorporation of exogenous AA or an autocrine activity of PGE2, involving down-stream

signaling characterized by activation of PI3, PKC and p38 kinases.

IEC-6 cells appeared as an excellent epithelial cell model for lipid body biogenic

studies since, different from neutral lipids storing cells and similar to resting leukocytes,

confluent IEC-6 cells have virtually no lipid bodies. Notably, it has been recently described

that transformed epithelial cells, as in colon cancer, present highly increased lipid body

numbers through a yet not defined molecular biogenic process (Teresa). The current

18

knowledge on lipid body biogenic characteristics was developed mainly from studies with

leukocytes or pre-adipocytes. Specifically for these cell types, it is accepted that: (i) lipid

bodies are endoplasmic reticulum (ER)-derived organelles; (ii) genesis of new lipid bodies

can be evoked by a variety of inflammation-related stimuli; (iii) lipid bodies can be rapidly

assembled, thus can be detected within an hour after stimulation; (iv) the biogenic process

is cell- and stimulus-specific; and (v) specific signaling events, which may involve

activation of phosphatases and/or kinases, elicit stimulus-triggered assembly of lipid bodies

(for review see nossa).Whether such lipid body assembling characteristics are relevant to

cell types other than leukocytes and adipocytes needed to be investigated. We verified that

some, but not all, leukocyte-related features of lipid body biogenesis are shared by

epithelial cells. For instance, amongst AA and PAF – two well-known biogenic stimuli of

leukocyte lipid bodies (ref) – only the former was able to trigger assembly of lipid bodies

within IEC-6 cells, reinforcing the concept of lipid body biogenesis as a cell-specific

process.

It is often assumed that AA bioactivity is attributable to its conversion into

eicosanoids or other bioactive products. However, rather than its metabolites, AA itself

appeared to directly mediate epithelial lipid body biogenesis, although PGE2, a

cyclooxygenase product of AA metabolism, is known by its key role on maintenance of

epithelial cells homeostasis (ref). Here, we showed that the AA capability to evoke lipid

body assembly in IEC-6 cells was not mediated by AA-derived cyclooxygenase products,

even though prostanoids like PGE2 and PGD2 are able to trigger rapid lipid body biogenesis

in other cell types (Bandeira-Melo, unpublished data; Mesquita-Santos, 2006).

The notion that AA bioactivity in eliciting lipid body biogenesis is dependent on a

direct activity of AA on epithelial cells functioning as an agonist of lipid body assembly

19

was reinforced by a number of observations. First, AA-induced lipid body assembly in

IEC-6 cells was an acute phenomenon apparent within 1 h of stimulation. Second, such

rapid lipid body biogenesis elicited in AA-stimulated IEC-6 cells was a concentration-

dependent phenomenon, suggesting an agonistic activity for AA. Cell activation mediated

by AA functioning as agonist has been observed to other cell systems, although specific

receptors for free AA were yet not described. Of note, although with little specificity, a

variety of molecular targets to AA have been identified, including PPAR receptors and

multi-component protein complex, such as NAPDH oxidase and ion channels (for review

see Brash 2001). Third, AA-induced lipid body biogenesis within epithelial cells was

mimicked by another unsaturated fatty acid (OA) but not by saturated fatty acids (PA and

SA). The ability of AA and OA versus the inability of PA and EA to trigger lipid body

biogenesis indicates that, while exogenous fatty acids provides a source of lipids for

incorporation into newly forming lipid bodies, the structurally restricted capacities of

different fatty acids to induce lipid bodies depends on mechanisms other than simple

availability of lipid precursors. Fourth and finally, AA-driven activation of IEC-6 cells that

culminates in concentration-dependent lipid body biogenesis depended on rapid activation

of p38 MAP kinase, as well as, PI3 kinase and PKC. Altogether, AA direct bioactivity on

epithelial cells triggers a rapid but highly regulated process of lipid body biogenesis

comprising a specific signaling cascade.

Functions of newly formed lipid bodies may vary according to cell type, specific

biogenic stimulus and subsequent intracellular signaling that set off its biogenic machinery.

Intracellular compartmentalization of AA metabolism has emerged as a key feature that

controls a variety of inflammatory cell functions (review). In leukocytes, lipid bodies

working in concert with other specific domains (e.g. perinuclear membrane and

20

phagosomes) have roles in regulating AA metabolism during inflammatory conditions.

Little is known about potential inflammatory functions of epithelial cell lipid bodies. If

epithelial cells lipid bodies have indeed roles in AA metabolism, then the AA present in

these epithelial lipid-rich structures should be readily hydrolysable from phospholipids by

PLA2s, which should be first compartmentalized within lipid bodies. In fact, cPLA2-α was

described to be associated with arachidonate-rich lipid bodies of activated leukocytes (Yu et

al., 1998). Stimulation of epithelial cells with unsaturated fatty acids, while failed to

increase the expression of cPLA2-α, triggered the translocation of constitutive cPLA2-α

from its resting cytoplasmic location to newly assembled lipid bodies within epithelial

cells.

Altogether, unsaturated fatty acids-elicited epithelial cells assemble intracellular

domains that compartmentalize both substrate and cPLA2-α. However, it is now well

recognized that successful AA mobilization is not merely determined by AA availability

and kinetic properties of the relevant enzyme cPLA2-α, but also depends on activity of

specific signaling responsible for activating cPLA2-α. Signaling through p38 MAP kinase

has emerged as an important activator of cPLA2-α activity (ref). Of note, co-

compartmentalization of cPLA2-α with MAP kinases (ERK1, ERK2, p85 and p38) within

lipid bodies of activated leukocytes has been described (ref). Therefore, besides their role in

AA-induced lipid body biogenesis within epithelial cells, activated p38 MAP kinase,

functioning as an up-stream regulator, may also modulate within the lipid body-

compartment other aspects of epithelial cell physiology, like AA mobilization by regulating

activity of lipid body cPLA2-α. Indeed, in our IEC-6 cells model of epithelial cell

activation, all these components of AA mobilizing molecular complex are elicited by

21

unsaturated fatty acid stimulation, including activation of p38 and lipid body co-

localization of phospholipids substrate and cPLA2-α. Translocated cPLA2-α appears to be

enzymatically active, inasmuch as in agreement to actively placing cPLA2-α close to the

lipid body content of esterified AA, OA-primed epithelial cells showed an enhanced

released of AA from epithelial cPLA2-α-bearing stores. Therefore, the first step of AA

metabolism – mobilization of free AA from phospholipids by p38-activated cPLA2-α –

may take place within newly formed lipid bodies of activated epithelial cells.

In conclusion, our results indicate that inflammatory relevant molecules, like AA

and OA, control lipid body biogenesis and functions in epithelial cells. Moreover, we

demonstrated that epithelial cells lipid body biogenesis is a highly regulated phenomenon

that culminates in p38-dependent lipid body assembly and cPLA2-α compartmentalization

leading to lipid body-driven enhanced AA mobilization. Based on that, we postulate lipid

bodies as key organelles involved in epithelial cell-driven inflammatory functions, and

therefore, inhibition of lipid body biogenesis may provide a novel target for anti-

inflammatory therapies of epithelial pathologies.

22

Legends

Figure 1. Lipid body biogenesis is triggered by AA within IEC-6 cells. Microscopy images

obtained from non-stimulated or AA-stimulated IEC-6 cells were stained with osmium (A)

or labeled with BODIPY (B). C shows lipid body counts after stimulation (4 h) with PAF

(1 µM) or AA (10 µM). Lipid bodies were enumerated using osmium staining. Results

were expressed as mean ± S.E. from at least three experiments. In D, BODIPY-labeled lipid

bodies were quantified by the measurement of the area of fluorescence per cell. Data on the

graph correspond to the mean ± S.E. from at least three experiments.

Figure 2. AA-induced lipid body assembly is a rapid and dose-dependent phenomenon. In

A, dose-response curve of lipid body biogenesis analyzed 4 h after stimulation with AA (1

– 30 µM). B shows kinetics of lipid body formation after stimulation with AA (10 µM).

Lipid bodies were enumerated using osmium staining. Results were expressed as mean ±

S.E. from at least three experiments.

Figure 3. AA-metabolite PGE2 does not mediate AA-induced lipid body biogenesis. In A,

IEC-6 cells were pre-treated with valeoryl salicilate (10 µM) or NS-398 (1 µM) 30 min

before stimulation with AA (10 µM). B shows that different concentrations of exogenous

PGE2 (1 – 10 µM) were unable to trigger lipid body formation within 4 h. Lipid bodies

were enumerated using osmium staining. Results were expressed as mean ± S.E. from at

least three experiments.

Figure 4. MAP kinase p38 mediates AA-induced lipid body formation in IEC-6 cells. In A,

IEC-6 cells stimulated in vitro with AA (10 µM) were pre-treated with inhibitors SB202190

(15 µM), U0126 (20 µM) or SP600125 (20 µM) of, respectively, three distinct MAP

kinases, p38, ERK1/2 and JNK. The treatments were performed 30 min before stimulation.

23

Analysis of lipid body formation was performed 4 h after in vitro stimulation. Results were

expressed as the means ± SEM from at least three experiments. * P ≤ 0.05 compared to

control groups. # P ≤ 0.05 compared to AA-stimulated cells. In B, a kinetic of AA-induced

phosphorilation of p38 is shown. Total EIC-6 cell lysates (2 x105 cells/lane) were separated

by SDS-PAGE and submitted to Western blotting for phosphorilated p38 or GAPDH as

indicated. The image is representative of at least three different blots.

Figure 5. OA-induced lipid bodies function as compartments of cPLA2-α-driven AA-

mobilization. In A, IEC-6 cells stimulated with 10 or 30 µM of OA assembled new lipid

bodies in a dose-dependent manner. B shows that 4 h stimulation with OA or A23187 did

not alter cPLA2 expression in IEC-cells. Total IEC-6 cell lysates (2 x105 cells/lane) were

separated by SDS-PAGE and submitted to Western blotting for cPLA2-α or GAPDH as

indicated. The image is representative of at least three different blots. In C, non-, A23187-

or OA-stimulated IEC-6 cells (4 h), as indicated, were labeled with BODIPY for lipid body

detection (green) and with anti-cPLA2-α (red). The merged image of AA-stimulated cells is

shown on the bottom right. Insert squares show enlarged images of cPLA2-α immuno-

labeling surrounding BODIPY-labeled lipid bodies. Images are representative of two

independent experiments. In D, dose-dependent effect of OA-induced enhanced

mobilization of AA. [3H]-AA was incorporated for 16 h within IEC-6 cells. Both non- and

OA-stimulated [3H]-AA-labeled IEC-6 cells were challenged with A23187 (5 µM) for 1 h,

and released [3H]-AA was measured in the supernatants. Each bar represents the mean ±

S.E. from at least three independent experiments.

24

Table 1. Unsaturated fatty acids (AA and OA), but not saturated fatty acids (PA and SA),

triggered lipid body biogenesis within IEC-6 cells.a

Condition Unsaturated Lipid body/cell

Control - 0.5 ± 0.2

Arachidonic Acid yes 12.5 ± 0.8*

Oleic Acid yes 9.1 ± 0.8*

Palmitic Acid no 0.7 ± 0.2

Stearic Acid no 0.4 ± 0.8

a Confluent IEC-6 cells were stimulated with AA, OA, PA or SA (10 µM). Analysis of lipid

body formation was performed 4 h after incubation with fatty acids in osmium-stained

cells. Results were expressed as the means ± SEM from at least three different experiments.

* P ≤ 0.05 compared to control group.

25

Table 2. Role of PI3 kinase and PKC on AA-induced lipid body biogenesis within IEC-6

cells.a

Condition Treatments Lipid body/cell

Control - 0.8 ± 0.1

Arachidonic Acid - 11.2 ± 3.0*

+ LY294002 1.9 ± 0.7#

+ Calphostin C 1.5 ± 0.4#

a Confluent IEC-6 cells were pre-treated for 30 min with PI3 kinase or PKC inhibitors,

LY294002 (10 µM) or Calphostin C (1 µM), and stimulated with AA (10 µM). Analysis of

lipid body formation was performed 4 h after AA stimulation in osmium-stained cells.

Results were expressed as the means ± SEM from at least three different experiments. * P ≤

0.05 compared to control group. # P ≤ 0.05 compared to AA-stimulated cells.

A B

DC

0

10

20

*

Lip

id b

odie

s/c

ell

Control AAPAF0

10

20

30

40*

Lip

id b

odie

sB

OD

IPY

flu

ore

scent

are

a/c

ell

Control AA

Figure 1

Control

AA

Control

AA

0

10

20

30

40

*

*

*

AA (µM) 1 3 3010_

Lip

id b

odie

s/c

ell

- 1 2 4 8 240

10

20

*

*

* *

(h)

AA (10 µM)

Lip

id b

odie

s/c

ell

Figure 2

A

B

0

10

20Lip

id b

odie

s/c

ell

**

*

AA (10 µM)

Treatments: _ _ VAL NS

0

10

20

PGE2 (µM) 1 3 10_

Lip

id b

odie

s/c

ell

Figure 3

A

B

A B

0

10

20

*

#

Lip

id b

odie

s/c

ell

AA (10 µM)

Treatments: SB SPU_ _

P-p38

GAPDH

AA (10 µM)

_ 1' 30'15'5' 60'

Figure 4

OA OA OA

Control Control A23187

A B

Figure 5

D

C

cPLA2

GAPDH

OA (1

0 µM

)

A23

187

OA (3

0 µM

)

Con

trol

0

10

20

30

40Lip

id b

odie

s/c

ell

OA (µM) 10 30_

*

*

0

5

10

[3H

]AA

rele

ase (

%)

OA (µM) 10 30_

*

*

BODIPY

BODIPY cPLA2cPLA2

cPLA2 overlay