Src homology 2 domain-containing inositol-5-phosphatase 1 (SHIP1) negatively regulates TLR4-mediated LPS response primarily through a phosphatase activity- and PI-3K-independent mechanism

Lipopolysaccharide (LPS; endotoxin) is a cell wall component of Gram-negative bacteria. LPS activates immune cells and triggers production of proinflammatory cytokines and other mediators, which up-regulate host immune defense systems and eliminate the bacterial infection.¹  LPS hyporesponsive mice are susceptible to Gram-negative bacteria infection.²  On the other hand, excessive or prolonged production of LPS-induced mediators can result in septic shock, a life-threatening condition.¹  However, a process, endotoxin tolerance, in which exposure to LPS renders immune cells tolerant to subsequent LPS stimulation, helps to keep responses to LPS under control.^1,3  Endotoxin-tolerant macrophages, for instance, produce lower levels of the proinflammatory cytokines tumor necrosis factor α (TNF-α) and interleukin 6 (IL-6), both of which play important roles in septic shock.³  Little was known about the mechanism by which LPS interacts with immune cells until the Toll-like receptor (TLR) family was identified in mammalian cells.^4,5  TLRs recognize conserved microbial components and function as critical sentinel receptors to initiate immune responses against microbial infections. TLR2, TLR4, and TLR9 recognize lipoteichoic acid, LPS, and the cytosine-phosphate-guanosine (CpG) motif of bacterial DNA, respectively. Given the important physiologic and pathologic relevance of LPS response, the mechanism of LPS-induced TLR4 signaling has been intensively studied but remains incompletely defined.^4,5  Through recruitment of adaptor proteins such as myeloid differentiation factor 88 (MyD88) and TIR domain-containing adaptor inducing IFN-β (TRIF), TLR4 activates downstream mitogen-activated protein kinases (MAPKs), including extracellular signal-regulated kinase 1/2 (ERK1/2), p38, and c-Jun NH2-terminal kinase (JNK), and nuclear factor κ B (NF-κB). Activation of MAPKs and NF-κB represents the last events of cytoplasmic TLR4 signal transduction and is essential for LPS-induced production of proinflammatory cytokines.^4,5  IL-1 receptor-associated kinase M (IRAK-M) and suppressor of cytokine signaling-1 function as negative regulators of TLR signal transduction and are considered as key molecules in the development of endotoxin tolerance.^4-7  As such, these signaling molecules are targets for the modulation of LPS response. Identifying positive and negative regulators and the pathways of LPS signaling will provide insights into the mechanism of TLR signal transduction, rendering us the ability to finely orchestrate LPS- and other TLR agonist-induced responses so as to eliminate microbial infection while protecting against lethal host responses.

Src homology 2 (SH2) domain-containing inositol-5-phosphatase 1 (SHIP1) was initially identified as a 145-kDa protein that negatively regulated B-cell receptor signaling by catalyzing the phosphoinositide 3-kinase (PI-3K) product PtdIns-3,4,5P3 into PtdIns-3,4P2.^8,9  SHIP1 also inhibits PI-3K activation induced by immunoglobulin E (IgE) and cytokines.¹⁰  In addition, SHIP1 can inhibit activation of ERK1/2, JNK, and NF-κB pathways.^11-13  All of these pathways play important roles in LPS signaling^4,5,14,15 ; however little is known about the role of SHIP1 as an important immune inhibitory molecule in LPS-induced inflammatory response.

We have investigated whether SHIP1 participates in LPS signaling in TLR4-reconstituted COS7 cells and the macrophage cell line RAW264.7. Unexpectedly, in addition to phosphatase activity-dependently inhibiting LPS-induced PI-3K activation, SHIP1 also phosphatase activity- and PI-3K-independently inhibits LPS-induced activation of MAPKs and the production of TNF-α. Our present study provides the first insight into the mechanism by which SHIP1 negatively regulates TLR4-mediated proinflammatory LPS response in a PI-3K-independent manner, suggesting that SHIP1 might be a potent candidate molecule for therapies aimed at manipulating LPS-induced inflammatory response.

LPS (0111:B4) was purchased from Sigma (St Louis, MO) and repurified as described previously.¹⁶  Antibodies specific to SHIP1, ERK1/2, p38, JNK, TLR4, MyD88, phospho-ERK1/2, phosphotyrosine, and horseradish peroxidase (HRP)-coupled secondary antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Antibodies specific to phospho-JNK and phospho-p38 were purchased from Promega (Madison, WI). LY294002, LY303511, wortmannin, and protease inhibitor cocktail were from Calbiochem (San Diego, CA).

The expression vector pBK-CMV/150.8SHIP1 (SHIP1) was a kind gift from Dr L. R. Rohrschneider.⁸  A phosphatase activity-disrupted mutant SHIP1 (mu-SHIP1) construct was generated from pBK-CMV/150.8SHIP1 by converting P671A, D675A, and R676G using a Mutanbest Kit (TaKaRa, Dalian, China).¹⁷  Mutagenic primers are 5′-GTGCGCCGGAGTCCTCTGGA-3′ and 5′-CAGGACGCCAAGTTGTACTTCATCC-3′. The pBS-U6 (pBluescript-U6) plasmid was a gift from Dr Y. Shi.¹⁸  An RNA-interfering plasmid, pU6-SHIP1, containing a sequence matching SHIP1 cDNA (5′-GGAAGTCATCAGGACTCTGCA-3′) was constructed. The control vector pU6 contains a scrambled sequence (5′-TCAGTCACGTTAATGGTCGTT-3′) without known specificity. The expression vectors pEF-TLR4 and pEF-MD2 were kind gifts from Dr K. Miyake.¹⁹  CD14 cDNA was amplified from THP1 cells by reverse transcriptase-polymerase chain reaction and cloned into pcDNA3.1 expressing vector (Invitrogen, Carlsbad, CA). The primers used for CD14 amplification were 5′-ATCCGCTGTGTAGGAAAGAACGT-3′ and 5′-GGAATTCAGACAGGTCTAGGCTGGTAA-3′. The pGL3.5XκB-luciferase reporter plasmid was kindly provided by Seamus J. Martin.²⁰  All constructs were confirmed by DNA sequencing.

The murine macrophage cell line RAW 264.7 and monkey kidney cell line COS7 were obtained from the American Type Culture Collection (Manassas, VA) and cultured as described previously.²¹  For Western blotting, 2 × 10⁶ cells, seeded into 60-mm dishes one day prior to transfection, were transfected with 5 μg DNA using Superfect Reagent (Qiagen, Hilden, Germany). For enzyme-linked immunosorbent assays (ELISAs), 2 × 10⁵ cells were seeded into 24-well plates and incubated overnight before LPS stimulation. In some SHIP1 knockdown experiments the cells were cotransfected with a green fluorescence protein (GFP)-expressing plasmid (pGFP), and GFP-expressing cells were sorted with a FACSDiVa system (Becton Dickinson, San Jose, CA).²²  SHIP1 stably transfected cells were selected with 300 μg/mL G418 (Invitrogen).

RAW264.7 macrophages were stimulated with 100 ng/mL LPS for indicated time periods. The concentrations of IL-6 and TNF-α in culture supernatants were measured using mouse IL-6 and mouse TNF-α ELISA Kits (R&D Systems, Minneapolis, MN) according to manufacturer's instructions.

Cells were lysed with M-PER Protein Extraction Reagent (Pierce, Rockford, IL) supplemented with protease inhibitor cocktail and protein concentrations of the extracts were measured by bicinchoninic acid (BCA) assay (Pierce). Forty micrograms of the protein was used either for immunoprecipitation or loaded per lane, subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), transferred onto nitrocellulose membranes, then blotted as described previously.²¹ 

PI-3K activity was assayed with PI-3K ELISA kit (Echelon Biosciences, Salt Lake City, UT) according to manufacturer's instructions. In brief, cell lysates were prepared and PI-3K was immunoprecipitated with antibody against the p85 subunit and incubated with PI(4,5)P2 (phosphatidylinositol-4,5-diphosphate). The reaction products were incubated with a PI(3,4,5)P3 detector protein then added to a PI(3,4,5)P3-coated microplate for competitive binding. A peroxidase-linked secondary detection reagent and colorimetric detection were used to detect PI(3,4,5)P3 detector protein binding to the plate. The colorimetric signal was inversely proportional to the amount of PI(3,4,5)P3 produced by PI-3K activity.

Akt (protein kinase B) activity was detected using Akt kinase Assay kit (Cell Signaling Technology, Beverly, MA) according to manufacturer's instructions. Briefly, cell lysates were prepared and immobilized Akt primary antibody was used to immunoprecipitate Akt from the extracts. Glycogen synthase kinase 3 (GSK-3) fusion protein (1 μg) was incubated with the precipitated protein in kinase buffer and then detected with phospho-glutathione-S-transferase 3 (phospho-GST-3)-specific antibody by Western blot. The phosphorylation level of GSK-3 fusion protein was proportional to Akt activity.

Ras activation was detected using Ras Activation Assay Kit (Upstate Biotechnology, Lake Placid, NY) according to the manufacturer's instructions. In brief, RAW264.7 cells were transfected and then stimulated with LPS as indicated. The cells were lysed with lysis buffer MLB, and protein concentrations of the extracts were measured using BCA Protein Assay kit (Pierce). Then 40 μg of lysate was incubated with GST fusion protein containing the Ras-binding domain of Raf (GST-Raf-RBD) for one hour at 4°C. Ras-guanosine triphosphate (Ras-GTP)-binding beads were pelleted by centrifuging at 14 000g for 3 minutes at 4°C. After 3 washes with MLB, the beads were boiled in SDS sample buffer for 5 minutes and separated on 15% SDS-PAGE. Ras-GTP was detected with anti-Ras antibody and HRP-conjugated secondary antibody and visualized with SuperSignal West Femto Maximum Sensitivity Substrate (Pierce).

RAW264.7 macrophages were cotransfected with the mixture of indicated pGL3-luciferase reporter plasmid, pRL-TK-Renilla-luciferase plasmid, and indicated amounts of SHIP1 expression construct. After 24 hours, the cells were left untreated or treated with LPS. Luciferase activities were measured using Dual-Luciferase Reporter Assay System (Promega) according to the manufacturer's instructions. Data are normalized for transfection efficiency by dividing Firefly luciferase activity with that of Renilla luciferase. The relative values are presented as fold increase over indicated control.

Densitometric analysis was done with Labworks Image Acquisition and Analysis Software (UVP, Upland, CA). The background was subtracted, and the signals of the detected bands were normalized to the amount of loading control band. The relative values were presented as fold increase over control samples as indicated.

Statistical significance was determined by Student t test, with a value of P below .01 considered to be statistically significant.

To study whether SHIP1 was involved in LPS signal transduction, we detected the phosphorylation of endogenous SHIP1 in RAW264.7 macrophages following LPS stimulation. As shown in Figure 1A, LPS stimulation prompted an increase in tyrosine phosphorylation of SHIP1 and the phosphorylation of SHIP1 was enhanced over time, suggesting SHIP1 might play a role in LPS signaling. In addition, the expression of SHIP1 was also upregulated, which increased at 4 hours and peaked at 8 to 12 hours after LPS stimulation (Figure 1B).

The effects of SHIP1 knockdown on LPS-induced production of TNF-α and IL-6 in RAW264.7 cells were investigated. The cells were transfected with pGFP together with the SHIP1 RNA-interfering plasmid pU6-SHIP1 or control plasmid pU6. After they were cultured for 48 hours, GFP-expressing cells were sorted, and the expression level of SHIP1 in the cells was detected. As shown in Figure 1C, SHIP1 RNA interfering inhibited endogenous SHIP1 expression by approximately 65% in RAW264.7 macrophages. The pU6-SHIP1- or control plasmid pU6-transfected cells were stimulated with LPS, and production of TNF-α and IL-6 in the cells was detected. SHIP1 knockdown cells produced 1.7-fold TNF-α and 1.9-fold IL-6 of that produced by mock-transfected cells (P < .01 for U6 vs U6-SHIP1) upon LPS stimulation (Figure 1D), demonstrating that SHIP1 could inhibit LPS-induced inflammatory response in macrophages.

The PI-3K pathway, the main target of SHIP1 in a variety of cell responses, plays an important role in LPS signaling. To investigate the role of SHIP1 in LPS-induced PI-3K activation, we established wild-type and mu-SHIP1 stably transfected cell lines. The mu-SHIP1 was constructed by introducing mutations of P671A, D675A, and R676G in the catalytic domain of SHIP1, which disrupted SHIP1's phosphatase activity.¹⁷  Western blot results showed that SHIP1 was overexpressed in both wild-type and mutant plasmid-transfected cells (Figure 2A). As shown in Figure 2B, wild-type SHIP1 inhibited LPS-induced activation of Akt, a PI-3K downstream molecule that is activated upon binding to PI(3,4,5)P3. On the contrary, mu-SHIP1 enhanced LPS-induced Akt activation. SHIP1 negatively regulates macrophage colony-stimulating factor (M-CSF)-induced PI-3K-dependent Akt activation through its phosphatase activity.^23,24  Consistent with previous studies, while overexpression of wild-type SHIP1 inhibited M-CSF-induced Akt activation, overexpression of mu-SHIP1 enhanced M-CSF-induced Akt activation (Figure 2B), providing further evidence that the mutations disrupted the phosphatase activity of SHIP1 in our study. Similar results were obtained in 3 independently established cell lines. These results demonstrate that SHIP1 negatively regulates the LPS-induced PI-3K pathway activation in macrophages in a phosphatase activity-dependent manner.

PI-3K promotes LPS-induced NF-κB activation in RAW264.7 macrophages.¹⁴  Our experiments demonstrate that this is also the case for LPS-induced ERK1/2 and JNK1/2 activation, which was strongly inhibited by the PI-3K-specific inhibitor LY294002 (Figure 3A). In contrast, LPS-induced p38 phosphorylation was enhanced by LY294002 treatment (Figure 3A), which was consistent with the previously reported inhibitory role of PI-3K in LPS-induced p38 activation in dendritic cells.¹⁵ 

We next investigated whether SHIP1 regulated LPS-induced activation of MAPKs through the PI-3K pathway. Unlike LY294002 treatment, SHIP1 overexpression not only inhibited LPS-induced ERK1/2 and JNK activation but also inhibited LPS-induced p38 MAPK activation (Figure 3B). The contrary effects of SHIP1 overexpression and LY294002 treatment on LPS-induced p38 activation suggested that SHIP1 might regulate LPS-induced p38 activation primarily via an alternative PI-3K-independent pathway. Consistently, phosphatase-inactive mu-SHIP1 could also inhibit LPS-induced ERK1/2, p38, and JNK activation, suggesting that SHIP1 regulates LPS-induced activation of MAPKs primarily via a phosphatase activity- and PI-3K-independent pathway. To confirm the conclusion, the PI-3K inhibitor LY294002 and wortmannin were used to inhibit PI-3K activation. As shown in Figure 3C-D, neither of them could efficiently diminish wild-type and mutant SHIP1-mediated inhibition of LPS-induced ERK1/2, p38, or JNK activation.

We further observed the effect of SHIP1 knockdown on LPS-induced activation of ERK1/2, p38, and JNK. Consistent with the above results, LPS-induced phosphorylation of ERK1/2, p38, and JNK was significantly enhanced in SHIP1 knockdown cells compared with that in control cells, in spite of the absence or presence of LY294002 (Figure 3E-F), demonstrating that SHIP1 regulates LPS-induced activation of MAPKs primarily by a PI-3K-independent mechanism in macrophages.

The NF-κB pathway plays an important role in TLR4 signal transduction. We investigated the role of SHIP1 in NF-κB activation by detecting NF-κB luciferase reporter gene expression. Unexpectedly, SHIP1 knockdown dose-dependently inhibited LPS-induced NF-κB activation (Figure 4A). Consistently, SHIP1 overexpression dose-dependently enhanced LPS-induced NF-κB activation (Figure 4B). We also detected LPS-induced NF-κB luciferase activity in SHIP1 or control vector stably transfected RAW264.7 macrophages. SHIP1 stable overexpression also increased LPS-induced NF-κB luciferase activity (data not shown). We further observed the effect of SHIP1 knockdown on LPS-induced IκB-α degradation. While SHIP1 knockdown enhanced LPS-induced IκB-α degradation (Figure 4C), SHIP1 overexpression inhibited LPS-induced IκB-α degradation (Figure 4D). These results suggest that SHIP1 positively regulates LPS-induced NF-κB activation through an IκB-α-independent mechanism.

To confirm the roles of SHIP1's phosphatase activity and the PI-3K pathway in SHIP1-mediated inhibition of LPS-induced inflammatory response, we compared the effects of wild-type SHIP1 and phosphatase-inactive mu-SHIP1 expression on LPS-induced TNF-α production. RAW264.7 macrophages stably transfected with SHIP1, mu-SHIP1 plasmid, or control vector were stimulated with LPS. As shown in Figure 5A, both wild-type SHIP1 and mu-SHIP1 expression inhibited LPS-induced TNF-α production. However, disruption of SHIP1's phosphatase activity partially blocked the decrease in TNF-α production. Similar results were obtained in 3 independently established cell lines. These results provided further evidences that SHIP1 inhibited LPS-induced proinflammatory response by both phosphatase activity-dependent and -independent mechanisms.

We also observed whether the PI-3K inhibitor LY294002 or wortmannin could affect SHIP1-mediated inhibition of LPS-induced TNF-α production. SHIP1 or mock-transfected RAW264.7 macrophages were stimulated with LPS in the presence of dimethyl sulfoxide (DMSO), LY303511, LY294002, or wortmannin, and TNF-α production in the supernatants was quantified (Figure 5B). Consistent with previous studies, LY294002 could partially inhibit LPS-induced TNF-α production, but wortmannin did not inhibit TNF-α production.¹⁴  However, neither of them could significantly diminish SHIP-mediated inhibition of TNF-α production. PI-3K activity assay showed that LY294002 and wortmannin used in the experiments were effective in blocking PI-3K activation (Figure 5C). These results suggest that the PI-3K-independent pathway functions in SHIP1-mediated inhibition of LPS-induced inflammatory response.

Ras is an important signal intermediator in the activation of immune cells and can be inhibited by SHIP1 in a PI-3K-independent manner.¹⁰  To further illuminate the mechanisms by which SHIP1 negatively regulates LPS response, LPS-induced Ras activation was detected in SHIP1 knockdown cells. As shown in Figure 6A, LPS treatment activated Ras over time. Compared with mock transfection, SHIP1 RNA interfering significantly enhanced LPS-induced Ras activation.

Recruitment of MyD88 to TLR4 is required for LPS signaling. We investigated the effect of SHIP1 knockdown on the LPS-induced TLR4-MyD88 combination by immunoprecipitation. As shown in Figure 6B-C, LPS treatment promoted the TLR4-MyD88 combination. Compared with mock transfection, SHIP1 RNA interfering significantly enhanced the LPS-induced TLR4-MyD88 combination.

As repurified LPS activates macrophages exclusively through TLR4,²⁵  the aforementioned results suggest that SHIP1 might be an inhibitory regulator of TLR4-mediated LPS response. We used TLR4-reconstituted COS7 cells to confirm that SHIP1 targeted to TLR4 signaling. LPS could not induce detectable phosphorylation of MAPKs in COS7 cells (data not shown), which was consistent with a previous report that COS7 cells did not express TLR4 and were irresponsive to LPS stimulation.²⁶  We established TLR4-reconstituted COS7 cells by transfecting the cells with constructs expressing TLR4, MD2, and CD14, all of which were required for TLR4-mediated LPS signaling, together with wild-type SHIP1 plasmid or control empty vector. As shown in Figure 7, LPS-induced phosphorylation of ERK1/2, p38, and JNK was significantly inhibited in the cells transfected with SHIP1 compared with the cells transfected with control empty vector, further demonstrating that SHIP1 negatively regulates TLR4-mediated LPS signaling.

In this study, we demonstrated that SHIP1 was phosphorylated and then up-regulated upon LPS stimulation and could inhibit LPS-induced activation of MAPKs and production of proinflammatory cytokines in macrophages. Kinetic studies demonstrated that the up-regulation of SHIP1 was much later compared with the phosphorylation of SHIP1, suggesting that the phosphorylation of SHIP1 could not be attributed to the up-regulation of SHIP1 expression. Recent studies suggested that SHIP1's 5-phosphatase activity appears not to change following extracellular stimulation and subsequent tyrosine phosphorylation.²⁷  LPS-induced phosphorylation might not influence the phosphatase activity of SHIP1. However, the change in the phosphorylation level of SHIP1 suggested that SHIP1 was involved in LPS signaling and might function as a negative regulator of LPS-induced inflammatory response in RAW264.7 macrophages.

SHIP1 overexpression inhibited Akt activity not only in resting cells but also in LPS-stimulated cells and the inhibition was dependent on the integrality of SHIP1's phosphatase catalytic domain, suggesting that SHIP1 negatively regulates LPS-induced activation of the PI-3K pathway through its phosphatase activity. However, disruption of the phosphatase activity of SHIP1 did not reverse its inhibitory regulation of MAPK activation and TNF-α production. Consistently, both LY294002 and wortmannin, which sufficiently inhibited LPS-induced PI-3K activity, could not significantly diminish SHIP1-mediated inhibition of LPS-induced MAPK activation and cytokine production, indicating that phosphatase activity of SHIP1 and the inhibition of PI-3K activation do not play predominant roles in SHIP1-mediated inhibition of LPS-induced inflammatory response in RAW264.7 macrophages.

LPS-induced Ras activation can be inhibited by SHIP1. In addition to functioning as a phosphatase, SHIP1 also competes with growth factor receptor-bound protein 2 (Grb2) for src homologous and collagen gene (Shc) and thereby PI-3K-independently inhibits Ras activation.¹⁰  The inhibition of Ras activation might be a possible PI-3K-independent pathway that SHIP1 employs to negatively regulate LPS response. However, LPS did not induce a detectable Shc-SHIP1 combination in our immunoprecipitation experiments to explore the upstream events that lead to the inhibition of Ras. Instead, SHIP1 knockdown enhanced the LPS-induced TLR4-MyD88 combination, a proximate signal event in LPS signaling. Upon LPS stimulation, TLR4 and MyD88 translocate into lipid rafts.^28,29  SHIP1 also translocates into lipid rafts following LPS treatment.³⁰  It is intriguing to suspect whether SHIP1 could inhibit the translocation of TLR4 and MyD88 into a lipid raft or whether SHIP1 could inhibit the combination between TLR4 and MyD88 in a lipid raft by a competitive mechanism. We sought to identify the association between endogenous SHIP1 and the intermediates involved in TLR4 signal transduction following LPS stimulation. However, we failed to detect the known intermediates including MyD88, IRAK1, and tumor necrosis factor receptor-associated factor 6 (TRAF6) in immunocoprecipitation experiments using a SHIP1-specific antibody (data not shown). Translocation to membrane was reported to be essential for the regulatory activity of SHIP1.^31,32  In this case, identification of specific and direct interactions between SHIP1 and the intermediates might be difficult.³² 

In TLR4 signal transduction, PI-3K is activated downstream of MyD88.¹⁴  SHIP1 might PI-3K-independently block LPS signaling through inhibiting the TLR4-MyD88 combination. In IL-1β signal transduction, Ras is activated downstream of MyD88 and IRAK1.³³  Our previous study showed that Ras activation promotes IRAK1-TRAF6 complex formation and thereby enhances TLR9 signal transduction in RAW264.7 macrophages.²¹  However, it remains unclear whether Ras activation is dependent on the TLR4-MyD88 combination in TLR signaling. Further study will determine whether the activation of Ras is MyD88 dependent in TLR4 and TLR9 signal transduction.

During the review of the manuscript, 2 groups reported that SHIP1 could regulate LPS-induced inflammatory response.^30,34  Sly et al³⁴  reported that LPS-induced ERK1/2 and Akt phosphorylation as well as cytokine production was enhanced in SHIP1^-/- BMmφ cells. Inconsistent with our results observed in RAW264.7 macrophages, the enhancement in TNF-α production could be efficiently blocked by LY294002 and wortmannin in SHIP1^-/- BMmφ cells.³⁴  The discrepancy might be due to the fact that the pharmacologic functions of LY294002 and wortmannin are different in the cells used in the 2 studies. Ly294002 and wortmannin completely blocked LPS-induced TNF-α production in SHIP1^-/- BMmφ cells rather than diminished SHIP1 deficiency-mediated enhancement of TNF-α production.³⁴  However, Ly294002 can only partially inhibit LPS-induced TNF-α production, and wortmannin cannot affect LPS-induced TNF-α production in RAW264.7 cells, as shown in our experiments and by another group.¹⁴ 

In another study, Fang et al³⁰  reported that SHIP1 up-regulated LPS-induced inflammatory response by showing that LPS-induced NF-κB-dependent gene transcription was up-regulated in SHIP1-overexpressing RAW264.7 cells, and LPS-induced ERK1/2 and Akt phosphorylation as well as TNF-α production was inhibited in SHIP1^-/- BMmφ cells. The controversial results concerning the activation of ERK1/2 and Akt and the production of TNF-α might be due to the fact that differently prepared LPSs were used in these studies and that contaminant agonists of other TLRs in commercially obtained LPS might lead to different results. In fact, we found that SHIP1 could positively regulate the signal transduction induced by some other TLR agonists in our experiments (data not shown). However, consistent with data by Fang et al,³⁰  we also observed that SHIP1 did not inhibit LPS-induced NF-κB activation in macrophages as revealed by increased luciferase reporter gene expression. In TLR signaling, MAPKs and IκB-α pathways lie downstream of the MyD88-IRAK1-TRAF6 cascade. The activation of the MyD88-IRAK1-TRAF6 cascade leads to MAPK activation and IκB-α degradation. In resting cells, IκB-α binds to the NF-κB p65 subunit and blocks its nuclear translocation. Upon activation, IκB-α is degraded and the NF-κB p65 subunit is released. As an NF-κB target gene itself, IκB-α is rapidly resynthesized and functions as a fast feedback down-regulator of NF-κB activation. We further investigated the effect of SHIP1 expression on LPS-induced IκB-α degradation. IκB-α was degraded within 20 minutes and significantly resynthesized within 40 minutes after LPS stimulation, as observed previously by another group in RAW264.7 macrophages.³⁵  Interestingly, SHIP1 expression inhibited LPS-induced IκB-α degradation in the cells, suggesting that SHIP1 enhances LPS-induced NF-κB activity by an alternative mechanism rather than IκB-α degradation. In addition to IκB-α degradation, several other pathways can regulate the activation of NF-κB. For example, the phosphorylation of the NF-κB subunit contributes to trans-activation of NF-κB.³⁶  A recent study demonstrated that proteasome-mediated NF-κB degradation IκB-α-independently regulates NF-κB activity.³⁷  Whether SHIP1 enhances LPS-induced NF-κB activation by affecting NF-κB phosphorylation or proteasome activity remains unclear. Our future study will illuminate whether SHIP1 enhances LPS-induced NF-κB phosphorylation or inhibits proteasome-mediated NF-κB degradation.

In conclusion, we provided the first evidence that SHIP1 can inhibit LPS-induced inflammatory response through a phosphatase activity- and PI-3K-independent mechanism. In addition to inhibiting LPS-induced PI-3K activation, SHIP1 inhibits the LPS-induced TLR4-MyD88 combination and Ras activation and subsequently inhibits LPS-induced activation of MAPKs, which is primarily in a phosphatase activity- and PI-3K-independent manner. These results suggest that RNA interfering and ectopic expression targeting SHIP1 might provide novel therapeutic strategies to modulate LPS response.

We thank Dr J. Rayner for critically reading this manuscript. Xuetao Cao and Huazhang An designed the research; Huazhang An, Hongmei Xu, Minghui Zhang, Jun Zhou, Tao Feng, Cheng Qian and Runzi Qi performed it; and Xuetao Cao, Huazhang An and Hongmei Xu analyzed data and wrote the paper.