FEBS Letters 582 (2008) 911–915

Toll-like receptors stimulate regulated intramembrane proteolysisof the CSF-1 receptor through Erk activation

Gary Glenn, Peter van der Geer*

Department of Chemistry and Biochemistry, San Diego State University, 5500 Campanile Drive, San Diego, CA 92182-1030, United States

Received 13 February 2008; accepted 13 February 2008

Available online 21 February 2008

Edited by Giulio Superti-Furga

Abstract The CSF-1 receptor is a protein-tyrosine kinase thatregulates the renewal, differentiation and activation of mono-cytes and macrophages. We have recently shown that theCSF-1 receptor undergoes regulated intramembrane proteolysis,or RIPping. Here, we report that RIPping can be observed inresponse to pathogen-associated molecules, which act throughToll-like receptors (TLRs). TLR-induced CSF-1 receptor RIP-ping is largely independent of protein kinase C, while maximalRIPping depends on Erk activation. Our studies show thatCSF-1 receptor RIPping can be activated by various intracellu-lar signal transduction pathways and that RIPping is likely toplay an important role during macrophage activation.� 2008 Federation of European Biochemical Societies.Published by Elsevier B.V. All rights reserved.

Keywords: Signal transduction; Protein-tyrosine kinase;RIPping; TACE; PKC

1. Introduction

CSF-1 regulates the proliferation and survival of cells of the

monocyte/macrophage lineage, their differentiation into mac-

rophages, and macrophage activation [1,2]. The receptor for

CSF-1 is a ligand activated protein-tyrosine kinase that utilizes

autophosphorylation sites to regulate kinase activity and to

control interactions with signaling proteins [1,2]. We previ-

ously reported that the CSF-1 receptor also undergoes regu-

lated intramembrane proteolysis, releasing the intracellular

domain from the plasma membrane, enabling it to move to

other compartments within the cell to participate in signaling

[3].

Regulated intramembrane proteolysis, or RIPping, is a con-

served process that consists of two consecutive proteolytic

cleavage events [4]. Integral membrane proteins are cleaved

within the extracellular region close to the plasma membrane,

resulting in ectodomain shedding, followed by cleavage within

the transmembrane region, yielding a soluble intracellular do-

main (ICD). Several metalloproteases, including TACE, have

been implicated in the first cleavage [5]. Intramembrane cleav-

age is often carried out by c-secretase [6]. In the case of Notch,

the ICD translocates to the nucleus where it alters gene expres-

sion [6,7]. Thus far, two protein-tyrosine kinases, ErbB4 and

*Corresponding author. Fax: +1 619 594 4634.E-mail address: pvanderg@sciences.sdsu.edu (P. van der Geer).

0014-5793/$34.00 � 2008 Federation of European Biochemical Societies. Pu

doi:10.1016/j.febslet.2008.02.029

the CSF-1 receptor, have been shown to undergo RIPping

[3,8]. RIPping is essential for the biological activity of Notch

and it may provide an alternate and still largely unexplored

signaling mechanism for receptor protein-tyrosine kinases.

Toll-like receptors (TLR), which are expressed on macro-

phages, recognize various pathogen-associated molecules,

including lipopolysaccharide, bacterial lipoproteins, double

stranded RNA, and bacterial DNA [9–11]. Upon ligand bind-

ing, TLRs initiate signaling pathways leading to the phosphor-

ylation of a variety of proteins, including the transcription

factors NF-jB, IRF3, and IRF5, culminating in increased pro-

duction of inflammatory cytokines. Thus, TLRs recognize

molecular patterns that are associated with pathogenic intrud-

ers and that function as essential activators of the innate

immune response [12].

We previously found that the CSF-1 receptor undergoes

RIPping induced by CSF-1, TPA, and LPS [3,13]. Here we

report that ligands for other TLRs also induce CSF-1 receptor

RIPping. RIPping in response to TLR activation is largely

independent of protein kinase C (PKC) or oxygen radical pro-

duction. Inhibition of MEK1/2 reduces CSF-1 receptor RIP-

ping, suggesting that stimulation of CSF-1 receptor RIPping

by TLRs involves Erk activation.

2. Materials and methods

2.1. ReagentsRecombinant human CSF-1 was obtained from R&D Systems.

A polyclonal antiserum was raised against the kinase insert domainof the murine CSF-1 receptor [14]. Lipopolysaccharide (Escherichiacoli O111:B4), lipid A (Salmonella minnesota Re 595), and U0126 werefrom EMD Biosciences (San Diego, CA). Lipoteichoic acid (Staphylo-coccus aureus), and polyinosinic–polycytidylic acid (pI:pC) wereobtained from Sigma (St. Louis, MO). DNA was isolated fromE. coli DH5a using the Easy-DNA kit according to the manufacturer�sdirections (Invitrogen, Carlsbad, CA). Phospho-specific antibodydirected against Erk1/2 (Thr202/Tyr204) was obtained from CellSignaling (Danvers, MA). An antibody directed against Erk1/2 wasobtained from Promega (Madison, WI).

2.2. Cell culture and treatmentsP388D1 mouse macrophages were grown, stimulated, and lysed as

described before [3]. Cells were seeded at 2–3 · 106 cells per 10 cm dishtwo days prior to stimulation.

2.3. Immunoprecipitation and Western analysisCleared lysates were incubated with a polyclonal anti-CSF-1 recep-

tor serum and proteins were collected using Protein A-sepharose. Sam-ples were fractionated by SDS–PAGE and analyzed by anti-CSF-1receptor immunoblotting as described previously [3].

blished by Elsevier B.V. All rights reserved.

912 G. Glenn, P. van der Geer / FEBS Letters 582 (2008) 911–915

3. Results

3.1. TLR agonists induce CSF-1 receptor RIPping

LPS is a potent and physiologically relevant activator of

macrophages. Our results confirm previous observations and

show that LPS induces CSF-1 receptor RIPping in a concen-

tration and time dependent fashion (Fig. 1). Proteins of 150,

130, and 55 kDa are seen on these blots. The 150 kDa protein

represents the mature receptor, present on the cell surface; the

130 kDa protein corresponds to a precursor protein, present in

the endoplasmic reticulum or Golgi apparatus; the 55 kDa

protein is produced by RIPping and represents the CSF-1

receptor ICD. To test whether CSF-1 receptor RIPping is an

integral part of macrophage activation, P388D1 cells were acti-

vated with lipid A, lipoteichoic acid, polyI:polyC, or bacterial

DNA, agonists of TLR4, TLR2, TLR3, and TLR9, respec-

tively [11]. Mammalian DNA was included as a negative con-

trol. Our results show that all TLR agonists that were tested

strongly induce CSF-1 receptor RIPping (Fig. 2).

Fig. 1. LPS induces CSF-1 receptor RIPping. P388D1 cells were stimulatedfor varying lengths of time (B) and CSF-1 receptor RIPping was analyzed byblotting.

Fig. 2. CSF-1 receptor RIPping in response to TLR activation. P388D1 cells(A), 10 lg/ml lipoteichoic acid (B), 25 lg/ml polyI:polyC (C), or 10 lg/ml bactwere treated for 20 min with increasing amounts of bacterial DNA or 10 lg

3.2. TACE inhibitors block TLR-induced CSF-1 receptor

RIPping

The metalloprotease, TACE, is important for TNFa release

in response to LPS [15,16]. Here we have shown that pretreat-

ment of P388D1 cells with the TACE inhibitor, TAPI-0, abro-

gated both receptor downregulation and the appearance of the

ICD in response to LPS or bacterial DNA (Fig. 3). Thus,

TACE, or a related protease, is essential for CSF-1 receptor

RIPping downstream of both TLR4 and TLR9.

3.3. LPS-induced CSF-1 receptor RIPping is independent of

PKC

Activation of PKC leads to increased TACE activity [17]. To

determine whether PKC plays a role in LPS-induced RIPping,

P388D1 cells were treated for 24 h with TPA to downregulate

PKC, and then stimulated with LPS. Both the disappearance

of mature receptors and the generation of the ICD were unaf-

fected by prolonged pretreatment with TPA (Fig. 4). As

expected, downregulation of PKC abrogated subsequent

with increasing amounts of LPS for 20 min (A) or with 100 ng/ml LPSCSF-1 receptor immunoprecipitation followed by anti-CSF-1 receptor

were treated for the indicated amounts of time with 100 ng/ml lipid Aerial DNA (D) and analyzed for CSF-1 receptor RIPping. P388D1 cells/ml human DNA and analyzed for CSF-1 receptor RIPping (E).

Fig. 3. TLR-dependent CSF-1 receptor RIPping is inhibited by the TACE inhibitor, TAPI-0. P388D1 cells were pretreated for 1 h with TAPI-0,stimulated with LPS (A), or bacterial DNA (B), and analyzed for CSF-1 receptor RIPping.

G. Glenn, P. van der Geer / FEBS Letters 582 (2008) 911–915 913

TPA-induced CSF-1 receptor RIPping. These observations

suggest that TLR-induced RIPping is independent of PKC.

3.4. MAP kinase dependent and independent pathways are

important for LPS-induced CSF-1 receptor RIPping

We have confirmed that LPS activates several protein

kinases in P388D1 cells, including IKK, p38 MAPK, JNK,

and Erk1/2 (data not shown). Pharmacological inhibitors were

used to assess the role of these kinases in LPS-induced CSF-1

receptor RIPping by pretreating P388D1 cells for 1 hour prior

to stimulation. With the exception of the MEK1/2 inhibitor,

U0126, none of the compounds tested had a significant effect

on LPS-induced CSF-1 receptor RIPping (data not shown).

In U0126-treated P388D1 cells there was a dose-dependent

decrease in LPS-induced CSF-1 receptor RIPping which corre-

lated with a dose-dependent inhibition of Erk1/2 activation

(Fig. 5). Thus, Erk activation is one of several pathways that

connect TLRs to CSF-1 receptor RIPping.

Fig. 5. Inhibition of the MAP kinase pathway reduces LPS-inducedCSF-1 receptor RIPping. P388D1 cells were pretreated for 1 hour withincreasing amounts of the MEK1/2 inhibitor, U0126, stimulated with100 ng/ml LPS for 20 min, and analyzed for CSF-1 receptor RIPping(A). Whole cell lysates were analyzed for Erk activation (B) and Erklevels (C).

4. Discussion

Macrophages are found in most tissues of the body. They

are indispensable for our defense against foreign intruders

and for tissue development. Macrophages engage in phagocy-

tosis, the production of cytokines, and communicate with the

adaptive immune system [18,19]. Macrophages have been

implicated in various maladies, including autoimmunity, can-

cer, atherosclerosis, and obesity [19–21]. CSF-1 regulates the

renewal and survival of macrophages, recruits macrophages

to specific sites within the body, and helps regulate macro-

phage activation [1,2].

CSF-1 binds to a receptor protein-tyrosine kinase and uti-

lizes autophosphorylation to initiate signal transduction [1,2].

In this classical paradigm, receptor autophosphorylation sites

function as docking sites for cellular signaling proteins that

are activated or inactivated as a consequence of their interac-

Fig. 4. Downregulation of PKC by chronic TPA treatment has little effect on24 h with 100 nM TPA or DMSO as a control, stimulated with 100 ng/ml LP

tion with the receptor. To terminate signaling, receptors are

internalized and degraded in the lysosomes [22]. Recently it

has become clear that the CSF-1 receptor is also subject to

RIPping, which results in ectodomain shedding and release

of the ICD into the interior of the cell [3]. In other systems

studied, the ICD moves to the nucleus where it interacts with

various transcription factors to alter gene transcription [4,6].

Thus, while RIPping contributes to the removal of CSF-1

receptors from the cell surface, it also generates a soluble

ICD capable of regulating gene expression (Fig. 6). The

removal of receptors from the cell surface is referred to as

downregulation. We now know that CSF-1 receptor

LPS-induced CSF-1 receptor RIPping. P388D1 cells were treated forS (A) or 100 nM TPA (B), and analyzed for CSF-1 receptor RIPping.

Fig. 6. Pathways leading to CSF-1 receptor RIPping. Activation of Toll-like receptors causes activation of Erk 1 and 2 and a second unidentifiedpathway leading to TACE activation. TACE mediates ectodomain shedding followed by intramembrane proteolysis carried out by c-secretase.Following release from the plasma membrane the CSF-1 receptor ICD can travel to the nucleus, is ubiquitinated, and degraded in the proteasome[3,13]. LPS-induced CSF-1 receptor RIPping is independent of PKC. However, activation of protein kinase C also results in CSF-1 receptor RIPping[3,13]. The TPA-induced CSF-1 receptor downregulation observed previously is a consequence of receptor RIPping.

914 G. Glenn, P. van der Geer / FEBS Letters 582 (2008) 911–915

downregulation can be mediated either by internalization fol-

lowed by lysosomal degradation, or by RIPping. The first pro-

cess acts exclusively to terminate signaling. In contrast,

RIPping results in the generation of a soluble ICD that is

likely to function as a signaling intermediate.

Macrophage activation by LPS is mediated by TLR4, one of

a family of cell surface receptors that recognize pathogen-asso-

ciated molecular patterns [9–11]. It has become clear that

TLRs are essential for activation of the immune response

against microbial pathogens [23]. Previous work has shown

that treatment of macrophages with LPS leads to the downreg-

ulation of the CSF-1 receptor from the cell surface [16,24]. Our

work confirms and extends these observations, showing that

the LPS-induced disappearance of the mature CSF-1 receptors

is mediated by RIPping (Fig. 3). Moreover, we have shown

that stimulation of other TLRs, including TLR2, TLR3, and

TLR9, also induces CSF-1 receptor RIPping (Fig. 2). These

observations reveal that physiologically relevant activators of

macrophages stimulate CSF-1 receptor RIPping, resulting in

the production of a soluble ICD. This is consistent with a mod-

el in which the CSF-1 receptor ICD functions in the signaling

network that triggers the innate immune response upon detec-

tion of pathogenic microorganisms.

What pathways do macrophages use to turn on RIPping?

We have shown previously that PKC activation can induce

CSF-1 receptor RIPping (Fig. 6). However, our current study

shows that PKC downregulation does not affect LPS-induced

RIPping. There is also evidence linking the production of oxy-

gen radicals during macrophage activation to TACE activa-

tion [25]. While we found that hydrogen peroxide can induce

RIPping, our experiments show no effect of free radical scav-

engers on LPS-induced CSF-1 receptor RIPping (data not

shown). We have used pharmacological inhibitors to show that

activation of Erk1 and Erk2 downstream of TLR4 plays a role

in the activation of RIPping (Fig. 5). This is consistent with the

observation that TACE can bind to and is a substrate for Erk1

and Erk2 [26]. Together these observations support the pre-

mise that Erk activation is one of several processes that act

in concert to stimulate maximal RIPping (Fig. 6). The other

players in this seemingly complex network remain to be iden-

tified.

Currently two receptor protein-tyrosine kinases have been

shown to undergo RIPping, ErbB4 and the CSF-1 receptor.

ErbB4 RIPping is induced by its natural ligand, neuregulin.

In contrast, CSF-1 appears to be a poor inducer of CSF-1

receptor RIPping [13]. Physiologically relevant inducers of

CSF-1 receptor RIPping include ligands for TLRs, which

are essential activators of the innate immune response. Preli-

minary results indicate that a stable version of the ICD local-

izes to the nucleus (unpublished results). Thus, our model

holds that the CSF-1 receptor ICD travels to the nucleus to

regulate gene transcription upon release from the plasma

membrane, thereby contributing to the rapid changes in gene

transcription that occur during the onset of the innate immune

response. Our research efforts are currently aimed at further

understanding the function of the CSF-1 receptor ICD in

the nucleus.

Acknowledgements: This work was supported by NIH Grants RO1AG025343 and R01 CA78629 and in part by the California MetabolicResearch Foundation.

References

[1] Bourette, R.P. and Rohrschneider, L.R. (2000) Early events inM-CSF receptor signaling. Growth Factors 17, 155–166.

G. Glenn, P. van der Geer / FEBS Letters 582 (2008) 911–915 915

[2] Pixley, F.J. and Stanley, E.R. (2004) CSF-1 regulation of thewandering macrophage: complexity in action. Trends Cell Biol.14, 628–638.

[3] Wilhelmsen, K. and van der Geer, P. (2004) Phorbol 12-myristate13-acetate-induced release of the colony-stimulating factor 1receptor cytoplasmic domain into the cytosol involves twoseparate cleavage events. Mol. Cell Biol. 24, 454–464.

[4] Ehrmann, M. and Clausen, T. (2004) Proteolysis as a regulatorymechanism. Annu. Rev. Genet. 38, 709–724.

[5] Arribas, J. and Borroto, A. (2002) Protein ectodomain shedding.Chem. Rev. 102, 4627–4638.

[6] Selkoe, D. and Kopan, R. (2003) Notch and Presenilin: regulatedintramembrane proteolysis links development and degeneration.Annu. Rev. Neurosci. 26, 565–597.

[7] Schweisguth, F. (2004) Regulation of notch signaling activity.Curr. Biol. 14, R129–R138.

[8] Ni, C.Y., Murphy, M.P., Golde, T.E. and Carpenter, G. (2001)gamma-Secretase cleavage and nuclear localization of ErbB-4receptor tyrosine kinase. Science 294, 2179–2181.

[9] Heine, H. and Ulmer, A.J. (2005) Recognition of bacterialproducts by Toll-like receptors. Chem. Immunol. Allergy 86, 99–119.

[10] Medzhitov, R. and Janeway Jr., C. (2000) The Toll receptorfamily and microbial recognition. Trends Microbiol. 8, 452–456.

[11] West, A.P., Koblansky, A.A. and Ghosh, S. (2006) Recognitionand signaling by Toll-like receptors. Annu. Rev. Cell Dev. Biol.22, 409–437.

[12] Cristofaro, P. and Opal, S.M. (2006) Role of Toll-like receptors ininfection and immunity: clinical implications. Drugs 66, 15–29.

[13] Glenn, G. and van der Geer, P. (2007) CSF-1 and TPA stimulateindependent pathways leading to lysosomal degradation orregulated intramembrane proteolysis of the CSF-1 receptor.FEBS Lett. 581, 5377–5381.

[14] Wilhelmsen, K., Burkhalter, S. and van der Geer, P. (2002) C-Cblbinds the CSF-1 receptor at tyrosine 973, a novel phosphorylationsite in the receptor�s carboxy-terminus. Oncogene 21, 1079–1089.

[15] Robertshaw, H.J. and Brennan, F.M. (2005) Release of tumournecrosis factor alpha (TNFalpha) by TNFalpha cleaving enzyme(TACE) in response to septic stimuli in vitro. Br. J. Anaesth. 94,222–228.

[16] Rovida, E., Paccagnini, A., Del Rosso, M., Peschon, J. and DelloSbarba, P. (2001) TNF-alpha-converting enzyme cleaves the

macrophage colony-stimulating factor receptor in macrophagesundergoing activation. J. Immunol. 166, 1583–1589.

[17] Shao, M.X. and Nadel, J.A. (2005) Neutrophil elastase inducesMUC5AC mucin production in human airway epithelial cells viaa cascade involving protein kinase C, reactive oxygen species,and TNF-alpha-converting enzyme. J. Immunol. 175, 4009–4016.

[18] Hoebe, K., Janssen, E. and Beutler, B. (2004) The interfacebetween innate and adaptive immunity. Nat. Immunol. 5, 971–974.

[19] Liang, C., Han, S., Senokuchi, T. and Tall, A.R. (2007) Themacrophage at the crossroads of insulin resistance and athero-sclerosis. Circ. Res. 100, 1546–1555.

[20] Sapi, E. (2004) The role of CSF-1 in normal physiology ofmammary gland and breast cancer: an update. Exp. Biol. Med. 22(Maywood), 1–11.

[21] Steinberg, G. (2007) Inflammation in obesity is the common linkbetween defects in fatty acid metabolism and insulin resistance.Cell Cycle 6, 888–894.

[22] Lee, P.S., Wang, Y., Dominguez, M.G., Yeung, Y.G., Murphy,M.A., Bowtell, D.D. and Stanley, E.R. (1999) The Cbl proto-oncoprotein stimulates CSF-1 receptor multiubiquitination andendocytosis, and attenuates macrophage proliferation. EMBO J.18, 3616–3628.

[23] Kawai, T. and Akira, S. (2005) Pathogen recognition with Toll-like receptors. Curr. Opin. Immunol. 17, 338–344.

[24] Sester, D.P., Beasley, S.J., Sweet, M.J., Fowles, L.F., Cronau,S.L., Stacey, K.J. and Hume, D.A. (1999) Bacterial/CpG DNAdown-modulates colony stimulating factor-1 receptor surfaceexpression on murine bone marrow-derived macrophages withconcomitant growth arrest and factor-independent survival. J.Immunol. 163, 6541–6550.

[25] Zhang, Z., Oliver, P., Lancaster, J.J., Schwarzenberger, P.O.,Joshi, M.S., Cork, J. and Kolls, J.K. (2001) Reactive oxygenspecies mediate tumor necrosis factor alpha-converting, enzyme-dependent ectodomain shedding induced by phorbol myristateacetate. FASEB J. 15, 303–305.

[26] Diaz-Rodriguez, E., Montero, J.C., Esparis-Ogando, A., Yuste,L. and Pandiella, A. (2002) Extracellular signal-regulated kinasephosphorylates tumor necrosis factor alpha-converting enzyme atthreonine 735: a potential role in regulated shedding. Mol. Biol.Cell 13, 2031–2044.