Endotoxin Induces Rapid Protein Tyrosine Phosphorylation in 702 / 3 Cells Expressing CD 14 *

CD14, a glycosylphosphatidylinositol-anchored glycoprotein of leukocytes, binds endotoxin (lipopolysaccharide (LPS)) with high affinity. After the murine pre-B cell line 70W3 is transfected with DNA encoding human CD14 (hCD14), the resultant stably transfected cell line, 70W3-hCD14, responds to 1000-fold lower LPS concentrations than the parental CD14-negative line. We have used 70W3-hCD14 cells, RAW264.7 cells, and elicited murine peritoneal exudate macrophages (PEM) to study LPS-induced protein tyrosine phosphorylation. LPS induces the rapid tyrosine phosphorylation of a 38-kDa protein (p38) in 70W3-hCD14 cells, PEM, and RAW264.7 cells and of two isoforms of mitogen-activated protein kinases (MAPK) in only RAW264.7 cells and PEM. p38 can be distinguished from the MAPK isoforms based on differences in mobilities on SDS-polyacrylamide gel electrophoresis and the lack of reactivity of p38 with anti-MAPK antibody even after dephosphorylation with potato acid phosphatase. Synthetic lipid A induces p38 phosphorylation in 70W3-hCD14 cells, whereas phorbol 12-myristate 13-acetate and interferon-y fail to induce tyrosine phosphorylation of p38. Pretreatment of 70W3hCD14 cells with anti-hCD14 monoclonal antibody or the tyrosine kinase inhibitor herbimycin A inhibits LPSinduced tyrosine phosphorylation of p38. These results suggest that increased protein tyrosine phosphorylation occurs rapidly after LPS binds to CD14 and is likely to be an important event in mediating LPS-induced cell activation.

with MO involves a plasma protein, LPS-binding protein (LBP), that binds to LPS and a glycosylphosphatidylinositolanchored cell-surface glycoprotein, CD14, that efficiently binds LPS when it is presented as an LPSLBP complex (4-7). A variety of different experimental approaches have provided data to support the concept that binding of LPS to CD14 initiates LPS-induced MO stimulation. Thus, CD14 functions as a membrane receptor for LPS that enables transmembrane signaling (8).
Recognition of the importance of the LBPlCD14 pathway has provided new opportunities to study mechanisms of LPS-induced transmembrane signaling. For example, studies from our laboratory have shown that transfection of human CD14 into the murine pre-B cell line 70W3 results in surface expression of glycosylphosphatidylinositol-anchored CD14 (6). A consequence of CD14 expression in 70W3 cells is that LPS binding to CD14 can be directly demonstrated (6,9) and that the concentration of LPS required to stimulate the cells is up to 1000fold lower than that required to stimulate the parental CD14negative cell line (6). Thus, the 70W3-hCD14 cells resemble MO and provide a novel model system to study the effects of LPS binding to CD14 on intracellular changes involved in cell activation.
Ligand-induced protein tyrosine phosphorylation is a very rapid event that mediates subsequent intracellular changes for many different receptors (10). Recently, Weinsteinet al. (11,121 reported that LPS treatment of RAW264.7 cells, a murine MOlike cell line, rapidly increases tyrosine phosphorylation of multiple proteins including the 41-and 44-kDa isoforms of mitogen-activated protein kinase (MAPK), that these protein tyrosine phosphorylations occur rapidly after addition of LPS to cells, and that pretreatment of RAW264.7 cells with the protein-tyrosine kinase inhibitor herbimycin A blocks MAPK phosphorylation. However, these studies did not address the role of CD14 in protein-tyrosine kinase activation.
Herein, we show that stimulation of 702/3-hCD14 cells and two different sources of MO with LPS, but not other agonists, rapidly induces protein tyrosine phosphorylation of a 38-kDa protein (~3 8 ) . Phosphorylated p38 is distinct from the 41-and 44-kDa MAPK isofoms expressed in 70W3 and MO. Inhibition of LPS-induced tyrosine phosphorylation of p38 occurs when cells are pretreated with the protein-tyrosine kinase inhibitor herbimycin A or with anti-CD14 monoclonal antibody. These data provide evidence for LPS-specific protein tyrosine phosphorylation that is linked to cell stimulation via a CD14-dependent pathway.

MATERIALS AND METHODS
Reagents-Re595 LPS was isolated from lyophilized Salmonella minnesotu Re595 bacteria as described (13). LPS from Escherichia coli Olll:B4 was purchased from List Biological Laboratories (Campbell, CAI, and synthetic lipid A was purchased from ICN Biomedicals, Inc. (Costa Mesa, CAI. Recombinant murine interferon-y (IFN-y) was a gift of R. D. Schreiber (Washington University, St. Louis, MO). FB2, an anti-phosphotyrosine monoclonal antibody (mAb), was purified from 25009 FB2 hybridoma (ATCC CRL1891) culture supernatant by chromatography on a protein G column. 4G10 anti-phosphotyrosine monoclonal antibody was purchased from Upstate Biotechnology, Inc. (Lake Placid, NY). Rabbit anti-extracellular signal-regulated kinase (MAPK) polyclonal antibody was from Santa Cruz Biotech (Santa Cruz, CA). Goat anti-mouse IgG conjugated to horseradish peroxidase was purchased from Cappel (Durham, North Carolina). Enhanced chemiluminescence (ECL) Western blotting detection reagents and Hybond-ECL nitrocellulose membranes were purchased from Amersham Corp. and used as described by the manufacturer. MY4, an anti-CD14 monoclonal antibody to hCD14, was purchased from Coulter Diagnostics (Hialeah, FL), and IB4, an anti-hCD18 monoclonal antibody, was a gift from Dr. K. Arfors (Pharmacia LKB Biotechnology Inc.). An oligonucleotide specifying the NF-KB consensus sequence was purchased from Promega Biotec.
Cell Stimulation-Unless noted otherwise, cells were maintained in RPMI 1640 medium supplemented with heat-inactivated (56 "C, 30 min) 10% fetal bovine serum, 1 m M sodium pyruvate, and 2 m glutamine at 37 "C in a 10% CO$air mixture (RPMWCS). Experimental samples were prepared with 5 x lo6 cells resuspended in 1 ml of RPMI/ FCS in sterile 1.5-ml polypropylene microcentrifuge tubes, the cells were maintained at 37 "C, and then agonists or control diluents were added as noted. In some experiments, cells were suspended in serumfree RPMI 1640 medium supplemented with 500 ng/ml purified rabbit LBP prepared as described (15). In experiments in which the effects of inhibitors on LPS-induced responses were studied, the cells were pretreated for 4 h with herbimycin A (Calbiochem) or GF 109203X (bisindolylmaleimide; Calbiocheml. Preparation of Cell Lysates and Immunoblotting-After stimulation, cells were rapidly chilled on ice, washed twice with ice-cold washing buffer (10 m Tris-HC1,150 IILM NaC1,l m M Na3V04, pH 7.5), and then lysed in 0.25 ml of lysis buffer (20 m M Tris-HCI, 120 m M NaC1, 10% glycerol, 1 m M Na3V04, 2 m EDTA, 1% Triton X-100, 1 m phenylmethanesulfonyl fluoride, pH 7.5) for 10 min on ice. Insoluble material was removed by centrifugation (10,000 x g , 20 min, 4 "C). The proteins in the cell-free detergent lysate were separated on 12% SDS-polyacrylamide gels as described (16). Following electrophoresis, the separated proteins were transferred to a nitrocellulose membrane for 4 h at 350 mA. Subsequently, the nitrocellulose membrane was blocked with TBSbuffered 2% bovine serum albumin (BSA) (TBS = 0.01 M Tris, 140 m M NaCl, pH 7.5) for 4 h at room temperature, and after rinsing with TBS, the membrane was incubated with murine monoclonal anti-phosphotyrosine antibody, FB2 or 4G10 (1 pg/ml diluted with TBS containing 0.05% Tween 20 and 1% BSA), for 3 h at 37 "C with continuous shaking. The membrane was then rinsed with three changes of TBS (containing 0.05% Ween 20), once for 15 min and twice for 5 min. Asecond antibody, goat anti-mouse IgG coupled to horseradish peroxidase diluted with TBS (containing 0.05% Tween 20 and 1% BSA), was applied to the membrane. After 60 min at room temperature, the membrane was washed four times with TBS containing 0.05% Tween 20 and once with TBS for 5 min. The Western blot was analyzed by the ECL detection system. In some experiments, the membrane was stripped (2% SDS, 50 m M Tris-HC1, 100 m 2-mercaptoethanol, pH 6.5, at 50 "C for 30 min) and reprobed with rabbit polyclonal anti-extracellular signal-regulated kinase (MAPK) antibody (0.05 pg/ml in TBS containing 0.05% Ween 20 and 1% BSA for 3 h at room temperature), followed by goat anti-rabbit IgG coupled to horseradish peroxidase (15000 dilution in TBS containing 0.05% Tween 20 and 1% BSA at room temperature for 1 h). The immunoreactive proteins were then visualized on film by ECL.
Dephosphorylation of p38 and MAPK-Detergent extracts were prepared as described above from either 70W3-hCD14 or RAW264.7 cells stimulated for 15 min with 1 ng/ml Re595 LPS and 0.1 p phorbol 12-myristate 13-acetate (PMA). The samples were electrophoresed by SDS-PAGE as described (12) and electrotransferred to polyvinylidene difluoride membranes (Bio-Rad) using transfer buffer lacking methanol (19). The resultant blots were stained with anti-phosphotyrosine mAb FB2 as described above, and aRer visualizing immunoreactive proteins using ECL, the region of the filter containing protein bands corresponding to p38 or MAPK2 was cut out. Elution of protein from the filter was accomplished by placing the filter in a small tube containing 0.5 mI of elution buffer/cm2 of polyvinylidene difluoride membrane (elution buffer = 50 m Tris-HC1,2% SDS, 1% Triton X-100, and 100 pg/ml BSA, pH 9.0) for 30 min with shaking at room temperature. The resultant solution was centrifuged for 10 min, and the eluted proteins were concentrated by precipitation with 4 volumes of acetone at -80 "C for 18 h. Precipitated proteins were recovered by centrifugation and dissolved in 10 m M HEPES, 0.1% Triton X-100, pH 6.0. A portion of the solubilized proteins were treated with potato acid phosphatase (600 unitdmg; Lot 20H 7215, Sigma) for 15 min at 30 "C; a ratio of 1 unit of potato acid phosphatase to 200 pl of original cell lysate was utilized. The dephosphorylation reaction was terminated by the addition of an equal volume of SDS-PAGE sample buffer (16). After heating at 100 "C for 5 min, the resultant samples were subjected to SDS-PAGE as described (16). After SDS-PAGE, the samples were electrotransferred to nitrocellulose and stained with either anti-phosphotyrosine or anti-MAF'K antibody as described above.

Characteristics of LPS-induced Protein Drosine Phosphorylation in CDI4-transfected Cells-Transfected CHO-K1 or 70Z3 cells expressing human CD14 (CHO-hCD14 or 702/3-hCD14 cells, respectively) or the same cells transfected with empty vector (CHO-RSV or 70Z3-RSV cells, respectively) (6)
were treated with 1 ng/ml Re595 LPS for 15 min, and detergent lysates were prepared from LPS-treated or -untreated cells. An aliquot of the cell-free lysate was subjected to SDS-PAGE, transfer to nitrocellulose membranes, and identification of tyrosine-phosphorylated proteins by staining with anti-phosphotyrosine mAb FB2 as described under "Materials and Methods." FB2 staining (Fig. l a ) revealed a complex pattern of phosphotyrosine-containing proteins in both CHO and 70U3 cells. The pattern of the constitutively tyrosine-phosphorylated proteins detected by FB2 differed in CHO-RSV and 70U3-RSV cells, but addition of 1 ng/ml LPS to these cell lines did not alter the cell-specific patterns. Expression of CD14 did not change the patterns of constitutively phosphorylated proteins in either cell type. However, LPS treatment of 70Z3-hCD14 cells resulted in tyrosine phosphorylation of a protein with an apparent molecular mass of 38 kDa, and we refer to this protein as p38 in the remainder of this report. In contrast, LPS addition (1 ng/ml) to CHO-hCD14 cells did not stimulate detectable changes in the patterns of protein tyrosine phosphorylations observed in unstimulated cells.
These and all subsequent results, with the exception of data shown in Fig. l b , were obtained from experiments in which the cells were stimulated in medium containing 10% FCS. As noted below in Fig. l b , LPS also stimulated p38 tyrosine phosphovlation in serum-free medium supplemented with purified LBP. We have previously reported that LPS-induced stimulation of 70m3-hCD14 cells is identical in serum-free medium supplemented with LBP or containing 10% heat-inactivated FCS (6).
Although the intensity of p38 staining on Western blots was somewhat variable from experiment to experiment, increased tyrosine phosphorylation of p38 was always noted when 70Z with anti-hCD14 mAb MY4 (10 pg/ml), anti-hCD18 mAb IB4 (IO pg/ml), actinomycin D (10 pg/ml), or cycloheximide (IO pg/ml) for 30 min in serum-free medium containing 500 ng/ml rabbit LBP. These cells were stimulated with (+) or without (-) 1 ng/ml Re595 LPS or with medium alone for 15 min, and the cell-free detergent lysates were analyzed for protein tyrosine phosphorylation as described under "Materials and Methods." 3-hCD14 cells were exposed to LPS. Analysis of cell lysates obtained from LPS-treated 70W3-hCD14 cells in more than 12 separate experiments failed to reveal additional changes in protein tyrosine phosphorylation; thus, for simplicity, in all subsequent figures, we will only show the areas of the Western blots containing p38. Essentially identical Western blot staining patterns were obtained using another anti-phosphotyrosine mAb, 4G10. Moreover, when anti-phosphotyrosine mAb was preincubated with a solution containing phosphotyrosine, phosphoserine, or phosphothreonine, only phosphotyrosine was able to block Western blot staining (data not shown).

33-
When 70W3-hCD14 cells were pretreated with anti-hCD14 mAb MY4, LPS-induced tyrosine phosphorylation of p38 was inhibited. In contrast, addition of a control mAb, IB4, did not prevent LPS-induced phosphorylation of p38 (Fig. lb). Pretreatment of cells with actinomycin D or cycloheximide did not prevent LPS-induced p38 tyrosine phosphorylation (Fig. lb). Thus, LPS-induced tyrosine phosphorylation of p38 is CD14dependent, but does not require ongoing RNA or protein synthesis. In studies to be described elsewhere, we show that MY4 25011 + + + + + + or other anti-CD14 mAbs also inhibit LPS-induced NF-KB activation. 2 We next determined complete time course and LPS doseresponse studies of p38 phosphorylation in the absence and presence of anti-hCD14 mAb MY4. When 1 ng/ml Re595 LPS was added to 70W3-hCD14 cells, the maximum increase in tyrosine phosphorylation of p38 was detected 15 min after LPS addition (Fig. 2a). Although the onset of p38 phosphorylation is rapid, we typically detected tyrosine-phosphorylated p38 up to 120-180 min after LPS addition; studies beyond this time period have not been performed. Pretreatment of the cells with MY4 prevented LPS-induced tyrosine phosphorylation of p38.
70W3-hCD14 cells maintained in the absence or presence of MY4 were also exposed to varying amounts of Re595 LPS (0.01-1000 ng/ml) for 15 min, and the extent of p38 tyrosine phosphorylation was evaluated. Maximum p38 phosphorylation was noted with 1 ng/ml Re595 LPS; the presence of MY4 completely inhibited p38 phosphorylation induced by 1 or 10 ng/ml LPS and nearly completely blocked p38 phosphorylation observed with 100 or 1000 ng/ml LPS (Fig. 2b). For all subsequent experiments, unless otherwise noted, we treated 70W3-hCD14 for 15 min with 1 ng/ml Re595 LPS.
Agonist Specificity of p38 Tyrosine Phosphorylation-The previous experiments were performed with LPS isolated from the R-form mutant of S. minnesota, Re595. We next sought to determine whether LPS isolated from S-form bacteria or synthetic lipid A also induce tyrosine phosphorylation of p38 (Fig.  3). Because 70W3 cells are also stimulated by IFN-y and PMA (20, 21), we evaluated the effect of a single concentration of these substances on tyrosine phosphorylation of p38. The two LPS isolates (1 ng/ml) and synthetic lipid (10 ng/ml) all induced p38 phosphorylation in 70W3-hCD14 cells, but not in 70W3-RSV cells. Treatment of either 70W3-RSV or 70W3-hCD14 cells with IFN-7 did not result in tyrosine phosphorylation of p38. PMA also failed to induce p38 phosphorylation, but did induce tyrosine phosphorylation of a moiety clearly distinct from p38 with an apparent molecular mass of 41 kDa. In contrast to the results observed with LPS in which increased protein tyrosine phosphorylation only occurred in 70W3-hCD14, the effects of PMA on protein tyrosine phosphorylation were identical in 70W 3-RSV and 70W3-hCD14 cells.
LPS-induced Tyrosine Phosphorylation in Macrophages-Several recent reports describe LPS-induced tyrosine phosphorylation in macrophages and in a macrophage-like cell line (11,12,22). These studies showed that the 41-and 44-kDa isoforms of MAPK (MAPK2 and MAPK1, respectively) are Lee

CD14 Induces Protein
Tyrosine Phosphorylation prominent target,s among the proteins displaying increased ty-PAGE. Interestingly, in RAW264.7 cells, it appeared that LPSrosine phosphorylation after LPS addition. Thus, we next comor PMA-induced protein tyrosine phosphorylation reduced the pared the effects of LPS (1 ng/ml Re595 LPS) and PMA (0.1 PM) reactivity of MAPK2 with anti"APK antibody. on protein tyrosine phosphorylation in five groups of cells: 702/ Dephosphorylation of p38 and MAPK2: Effect on Reactivity 3-RSV and 70W3-hCD14 cells, RAW264.7 cells, and PEM obwith Anti-MAPKAntibody-We next performed experiments to tained from C3HeBReJ and C3wHeJ mice (Fig. 4a) and sperule out the possibility that p38 is a hypophosphorylated form cifically examined the effects of these agonists on tyrosine of MAPK2 that has anomalous behavior on SDS-PAGE and phosphorylation of MAPK and p38. As expected, this concenreacts poorly with anti-MAPK antibody. To do this, we treated tration of LPS induced p38 phosphorylation in 70Z/3-hCD14 702/3-hCD14 cells with 1 ng/ml Re595 LPS and 0.1 p~ PMA cells, but not in 70W3-RSV cells, whereas PMA also induced and recovered the tyrosine-phosphorylated bands that comephosphorylation of a 41-kDa protein in both cell lines. Addition spend to p38 or " K 2 from polyvinylidene difluoride memof LPS to RAW264.7 cells or PEM obtained from C3HeB/FeJ branes as described under "Materials and Methods." After mice induced tyrosine phosphorylation of p38 and two additreatment with potato acid phosphatase, the phosphorylated tional proteins with apparent molecular masses of 41 and 44 and dephosphorylated samples were subjected to SDS-PAGE, kDa, respectively. LPS failed to induce tyrosine phosphorylselectrotransfer to nitrocellulose, and immunoblotting with tion of any of these proteins in PEM obtained from C3HfHeJ anti-phosphotyrosine mAb FB2 or anti-MAPK antibody using mice. In PEM from C3wHeJ mice, P m i n d u c e d tyrosine phosthe procedures described under "Materials and Methods." Po-Phorylation of a Protein with an apparent size near that ofP38, tato acid phosphatase has broad specificity of action in dephosbut careful analysis showed that this protein band does not phorylating various phosphoproteins (23)(24)(25)(26). After treatment correspond to p38 (data not shown). Addition of PMA to all five of either MApK2 or p38 with potato acid phosphatase, these cell types resulted in tyrosine PhosPhoVlation of a 4l-kDa proteins failed to stain with FB2 (Fig. 5a). Dephosphorylation Protein. w e also Observed that PMA induced tyrosine Phosof p38 with potato acid phosphatase did not result in staining phorylation of a 44-kDa moiety in RAW264.7 cells. The effects with anti-MAPK antibody, whereas potato acid phosphataseof LPS and PMA on protein tyrosine phosphorylation in these treated ~~p m was still recognized by anti-MAPK antibody five groups of cells are summarized in Table I. (Fig. 5b). The same results were obtained with the protein TO confirm Previous repofis that Lps induces tyrosine phosbands corresponding to p38 and MAP= obtained from LpSphorylation of MAPKl and in MO cells (12,221 and to treated ~~~2 6 4 . 7 cells (data not shown). W Y t a i n if ~3 8 is related to MAPK, we stripped the anti-Inhibition of Protein l"yrosine Phosphorylation Blocks Lpsphosphotyrosine mAb from the V k S t e r n blot shown in Fig. 4a induced N F Table 1). In contrast, no anti-MAPK reactivity in the determined the effect of a protein-tyrosine kinase inhibitor and 44-and/or 41-kDa protein phosphorylated by L p s or p m (Fig. CD14 mAb inhibits LPS-induced NF-KB activation.2 Here, we region corresponding to P38 was noted. Thus, P38 appears to be a protein kinase C inhibitor on LPS-induced protein tyrosine distinct from these MAPK isozymes based on the lack Of phosphorylation and NF-KB activation. 70Z&hCD14 cells were tivity with anti"APK antibody and its mobility on SDS-pretreated for h with the protein-tyrosine kinase inhibitor herbimycin A or with the protein kinase C inhibitor GF U a 109203X (27) and then stimulated for 15 min with either Re595 LPS or PMA. Cell-free detergent lysates were prepared for analysis of protein tyrosine phosphorylation (Fig. 6a) 15 min, and the cell-free detergent lysates were analyzed for protein tyrosine phosphorylation as described under "Materials and Methods" (a). NF-KB activity in nuclear lysates was determined by electrophoretic mobility shift assay as described under "Materials and Methods" ( b ) . Addition of the inhibitors alone had no effect (data not shown). endustin A (1.7 PM)) and the protein kinase A inhibitor H89 (30 PM) failed to block p38 tyrosine phosphorylation and NF-KB activation. DISCUSSION Herein, we show that LPS induces tyrosine phosphorylation of a 38-kDa protein (termed p38) in 70W3-hCD14 cells. Phos-phorylation of p38 is detected within 15 min after LPS addition and is blocked by pretreatment of cells with an anti-CD14 mAb or with a protein-tyrosine kinase inhibitor (herbimycin A). Because p38 has an electrophoretic mobility that differs from MAPKl and MAPK2 and because neither phosphorylated nor dephosphorylated p38 reacts with anti-MAPK antibodies, we conclude that p38 is distinct from the 41-and 44-kDa isofoms of MAPK. Although 70W3 cells contain both MAPK isofoms, LPS does not induce tyrosine phosphorylation of these proteins. We also observed LPS-induced tyrosine phosphorylation of a protein in the MO-like cell line RAW264.7 and in PEM obtained from C3HeBFeJ mice with an electrophoretic mobility comparable to p38. p38 phosphorylation is induced by Rand S-forms of LPS as well as by synthetic lipid A. In 70Z/3-hCD14 cells, the phosphorylation of p38 appears to be LPS-specific since other agonists for these cells such as IFN-.)I and PMA fail to induce tyrosine phosphorylation of p38. Pretreatment of 70W3-hCD14 cells with anti-hCD14 mAb MY4 nearly completely inhibited p38 phosphorylation when the cells were challenged with 0.01-1000 ng/ml LPS. Therefore, binding of LPS to CD14 initiates transmembrane signaling, resulting in tyrosine phosphorylation of p38. Although the mechanism whereby CD14 induces transmembrane signaling leading to protein tyrosine phosphorylation is not understood, several models have been proposed, which include the participation of additional membrane proteins that function together with CD14 to compose a high affinity heteromeric membrane receptor for LPS (8). These additional proteins may also act as a low affinity LPS receptor and mediate LPS-induced cell activation in the absence of CD14 expression.
We showed that addition of LPS to RAW264.7 cells or to PEM from C3HeBFeJ mice also resulted in tyrosine phosphorylation of a 38-kDa protein that we tentatively conclude is related to that observed in 70W3-hCD14 cells. PMA did not induce p38 phosphorylation in these cell types. Thus, p38 tyrosine phosphorylation appears to be a specific target of LPS action in three different LPS-sensitive cell types. Further evidence suggesting that p38 tyrosine phosphorylation is an LPS-specific event derives from the observation that LPS failed to induce p38 phosphorylation in PEM obtained from LPS-hyporesponsive C3WHeJ mice.
Data supporting the contention that protein tyrosine phosphorylation is involved in LPS signaling in MO have been provided by Weinstein et al. (11,12). These investigators described LPS-induced protein tyrosine phosphorylation of several proteins in the murine macrophage-like cell line RAW264.7 and have identified the 41 (MAPK2)-and 44 (MAPKlI-kDa isoforms of MAPK as specific targets of LPS action. Inclusion of herbimycin A prevented tyrosine phosphorylation of these proteins, and since a protein kinase C inhibitor (compound 3) did not block LPS-induced tyrosine phosphorylations, Weinstein et al. concluded that the LPS effects are mediated by a herbimycin A-sensitive protein-tyrosine kinase and are not dependent on protein kinase C. The recent reports of Manthey et al. (22) and Dong et al. (28) also confirmed and extended these observations. Wosine phosphorylation of MAPK isoforms was noted with elicited PEM from LPS-responsive C3H mice, but not with PEM from the LPS-hyporesponsive mouse strain, C3wFeJ. Results described herein with PEM from C3HeBFeJ (LPSresponsive) and C3WHeJ (LPS-hyporesponsive) mice and with RAW264.7 cells demonstrating LPS-or PMA-induced tyrosine phosphorylation of proteins that correspond to MAF'Kl and MAPK2 isoforms are generally in agreement with the findings of these other investigators (11,12,22,28). Inspection of the published Western blots in the studies of Weinstein et al. (11,12) and Manthey et al. (22) reveals the presence of additional tyrosine-phosphorylated proteins present in lysates of LPSstimulated cells that might correspond to p38. It is difficult, however, to unequivocally assign a specific band corresponding to p38 in these previous studies. Nevertheless, because we have duplicated all of the published findings from these other investigators using comparable cell preparations, it is likely that tyrosine phosphorylation of a protein corresponding to p38 also occurred.
Western blot analysis established that 70W3 cells contain the two W K isoforms and that PMA induces the tyrosine stimulation of 70W3 cells. Although PMA can mimic some effects of LPS, PMA failed to induce p38 phosphorylation. Moreover, the effects of PMA on MAPK2 phosphorylation are blocked by the protein kinase C inhibitor GF 109203X (271, but not by herbimycin A. We show here that anti-CD14 antibody blocks LPS-induced p38 tyrosine phosphorylation and, in studies to be described elsewhere,2 that anti-CD14 antibody blocks LPS-induced NF-KB activation. Herbimycin A treatment also blocks LPSinduced tyrosine phosphorylation of p38 and LPS-induced NF-KB activation. Determining the exact relationship among p38 phosphorylation, NF-KB activation, and LPS-induced cell activation will require additional studies. However, our findings suggest that after LPS binds to CD14, transmembrane signaling occurs, which leads to activation of a protein-tyrosine kinase or, alternatively, reduced activity of a protein-phosphotyrosine phosphatase. Additional studies are needed to distinguish between these possibilities. We show here that three additional compounds reported to inhibit protein-tyrosine kinases failed to inhibit LPS-induced p38 tyrosine phosphorylation and NF-KB activation. Weinstein et al. (11,12) also reported that herbimycin Ainhibits LPS-induced protein tyrosine phosphorylation in RAW264.7 cells and that genistein fails to do so, but did not report findings with other protein-tyrosine kinase inhibitors. Others have demonstrated that the group of protein-tyrosine kinase inhibitors we tested may be highly selective in the inhibition of different protein-tyrosine kinases with distinct inhibitory effects on closely related enzymes (29). However, more rigorous interpretation of our data cannot be offered until the identification of the specific protein-tyrosine kinase activated by LPS is achieved.
Preliminary subcellular fractionation studies of LPS-stimulated 70W3-hCD14 cells indicate that tyrosine-phosphorylated p38 is not found in membrane-rich fractions, but is partitioned between cytosolic and nuclear fractions. These findings are of interest and suggest that a full understanding of the role of p38 in mediating LPS-induced cell stimulation will require completion of the characterization of the structure and function of this protein. This work is currently underway in our laboratory.