Leukotriene D(4) triggers an association between gbetagamma subunits and phospholipase C-gamma1 in intestinal epithelial cells.

The proinflammatory mediator leukotriene D(4) (LTD(4)) binds to the seven-transmembrane receptor CYSLT(1). Although this leukotriene plays an important biological role, its intracellular signaling pathways are only partly known. In previous experiments, we found that LTD(4) induced tyrosine phosphorylation and translocation of phospholipase (PLC)-gamma1 to a plasma membrane fraction in a human epithelial cell line (Int 407). In the present study, we further examined these signaling events and found that LTD(4) induced a rapid interaction between Gbetagamma subunits and PLC-gamma1; results obtained with GST fusion proteins of PLC-gamma1 suggest that this interaction is mediated via the pleckstrin homology domain of PLC-gamma1. Moreover, LTD(4) induced an increased association of c-Src with PLC-gamma1, and the selective Src family tyrosine kinase inhibitor PP1 blocked both LTD(4)-induced tyrosine phosphorylation of PLC-gamma1 and the association of PLC-gamma1 with Gbetagamma subunits. The relevance of these observations in intracellular calcium signaling was investigated by microinjecting cells with anti-Gbeta, anti-PLC-gamma1, or anti-c-Src antibodies and by pretreatment with PP1. LTD(4)-induced calcium mobilization was blocked by each of the indicated antibodies (but not isotype-matched control antibodies) and by PP1. Our data suggest that Gbetagamma subunits can, directly or indirectly, serve as membrane-bound partners for PLC-gamma1 and c-Src and that each of these proteins is essential for LTD(4)-induced downstream PLC-gamma1 signaling.

Leukotrienes (LTs) 1 are metabolites of the fatty acid arachidonate, produced by the activation of 5-lipoxygenase, and are powerful inflammatory mediators (1). LTs are extremely potent and exert a large number of functional effects; for example, they increase vascular permeability and cause contraction of smooth muscle cells, trachea, parenchyma, and ileum (2,3).
It is generally assumed that LT-induced effects are mediated via specific plasma membrane receptors. Yokomizo et al. (4) have cloned and sequenced the LTB 4 receptor from a cDNA library of a gene that encodes an orphan seven-spanning receptor on human B-lymphoblasts (5). In studying Chinese hamster ovary LTB 4 receptor cells, Yokomizo et al. (4) also noted that the LTB 4 receptor, which belongs to the seven-transmembrane-spanning receptor group, can bind to two different Gproteins, one that is sensitive and one that is insensitive to pertussis toxin. The former G-protein has been identified as ␣ i and the latter as ␣ 16 , and both appear to signal via phospholipase C-␤ (PLC-␤) in COS-7 cells (6). The LTD 4 receptor was also recently cloned and shown to belong to the seven-transmembrane group (7,8), which is in accordance with previous observations reported from our laboratory (10) and by other investigators (9) showing that LTD 4 signaling occurs through heterotrimeric G-proteins. We have also demonstrated (10), as have others (11), that the cytosolic free Ca 2ϩ concentration is increased in cells stimulated with LTD 4 . Furthermore, we have concluded that the LTD 4 -induced mobilization of Ca 2ϩ involves PLC-␥1 in the human intestinal epithelial cell line Int 407 (12). Ligand binding to a seven-transmembrane-spanning receptor triggers an increased dissociation of heterotrimeric G-proteins into free G␣ and G␤␥ subunits, which in turn transduce external signals to intracellular responses by interacting with downstream effectors. However, it is well established that a certain degree of receptor-independent dissociation into free G␣ and G␤␥ subunits occurs within cells (13), and both of these subunits can separately regulate a variety of ion channels and intracellular enzymes, including different isoforms of PLC (13)(14)(15).
Various sub-isoforms of PLC-␤ have been identified as downstream effectors of G␤␥ subunits, and those subunits have been shown to interact directly with PLC-␤2 in vitro (16,17) and in response to various agonists in vivo (18 -20). However, there are no reports of a direct association in vivo between G␤␥ and PLC-␥ or PLC-␦. Rhee and Bae (21) originally described the PLC-␥ isoform as a calcium signaling enzyme involved in growth factor signal transduction, and these investigators noted that activation of PLC-␥ was associated with increased tyrosine phosphorylation and translocation of this enzyme to the plasma membrane (21). Although we have found that LTD 4 activated the heterotrimeric G␣ i3 -protein (22), we have also observed that LTD 4 stimulated PLC-␥1, detected as an increase in tyrosine phosphorylation and translocation of this isoform of the enzyme to the plasma membrane (23). Moreover, in the latter investigation, we noted that the PLC-␥1-mediated release of intracellular calcium in Int 407 cells included a Rho-dependent event. A similar role for Rho in PLC-␥-mediated calcium signaling was subsequently confirmed in other cells and for other types of receptors (24,25). Marrero et al. (26) cloned the angiotensin II receptor and found that it is a seven-transmembrane-spanning receptor that stimulates phosphoinositide hydrolysis upon ligand binding and, as activation of the LTD 4 receptor, leads to tyrosine phosphorylation of PLC␥1 in rat vascular smooth muscle cells. In a subsequent study (27), these investigators discovered that the tyrosine phosphorylation was inhibited by electroporation of anti-pp60c-src antibodies into such muscle cells. Later, the same research group (28) suggested that the G-protein-coupled angiotensin II AT1 receptor can associate with intracellular proteins other than G-proteins and thereby create a signal transduction complex similar to that observed for "classic" growth factor receptors.
The following findings provided the incentive for the present investigation. G␤␥ subunits reside in the plasma membrane; activation of PLC-␥1 is associated with translocation to the plasma membrane (21); G␤␥ subunits interact and activate PLC-␤2 in vitro (16,17); and G␤ subunits bind to proteins containing PH domains (29). Our aim was to investigate a possible role for G␤␥ subunits in the LTD 4 (30). The anti-G␤, anti-PLC-␥1, and anti-c-Src antibodies were obtained from Upstate Biotechnology Inc., Lake Placid, NY, and the anti-Src (Tyr 418 (P)) was from BIOSOURCE International, Camarillo, CA. LTD 4 was purchased from Cayman Chemical Co., Ann Arbor, MI, and fura-2/AM was from Molecular Probes, Eugene, OR. All other chemicals were of analytical grade and obtained from Sigma.
Cell Culture-Human embryonic intestinal epithelial cells (intestine 407 (31)), which exhibit typical epithelial morphology and growth, were cultured as a monolayer to approximately 80% confluence for 5 days. Cell cultures were kept at 37°C in a humidified atmosphere of 5% CO 2 and 95% air in Eagle's basal medium supplemented with 15% newborn calf serum, 55 international units/ml penicillin and 55 g/ml streptomycin. The cells were regularly tested to ensure the absence of mycoplasma contamination.
Cell Lysis and Membrane Fractionation-Cell stimulations (15 or 300 s) were terminated by adding ice-cold buffer containing 20 mM Na-HEPES (pH 8), 2 mM MgCl 2 , 1 mM EDTA, 2 mM Na 3 VO 4, 4 g/ml leupeptin, and 30 g/ml phenylmethanesulfonyl fluoride. Thereafter, the cells were scraped loose into the buffer and homogenized 10 times on ice and then centrifuged at 200 ϫ g for 10 min. The supernatant was collected, and after measurement of and compensation for the protein content, it was centrifuged at 1000 ϫ g for 5 min. The 1000 ϫ g supernatant was further centrifuged at 200,000 ϫ g for 30 min, and the resulting membrane-rich pellet was then washed and centrifuged at 200,000 ϫ g for another 30 min. The particulate (membrane) fractions were resuspended in 500 l of the described lysis buffer. The membrane and supernatant (cytosol) fractions were enzymatically characterized by measuring the presence of 5Ј-nucleotidase and lactate dehydrogenase; the activity of the membrane marker in the supernatant was negligible, and the cytosol marker showed no activity in the membrane fraction.
Immunoprecipitation-The solubilized membrane fractions (23) were incubated for 1 h at 4°C with 4 g of anti-PLC-␥1, 3.5 g of anti-c-Src, or 3 g of anti-G␤ antibodies. In the control experiments, the membrane fractions were incubated with isotype-matched antibodies directed against unrelated proteins or obtained from nonimmunized animals. The membrane fractions were subsequently incubated with gentle rocking for an additional hour at 4°C with 100 l of a 60 mg/ml solution of protein A-or 50 l of 3 mg/ml solution of protein G-Sepharose beads (Amersham Pharmacia Biotech). Immunoprecipitates were washed in buffer A (23) supplemented with 1% Tween 20.
GST Fusion Proteins and Binding Assays-The cDNA clones encoding the GST fusion proteins of the PLC-␥1 PH(N)-SH2-SH2-SH3-PH(C) domain (amino acids 483 to 936) and the PLC-␥1 SH2-SH2-SH3 domain (amino acids 550 to 850) in pGEX vectors were transformed into Escherichia coli and cultured at 30°C. Expression of the GST fusion proteins was induced with 1 mM isopropyl-1-thio-D-galactopyranoside, and the cells were subsequently collected by centrifugation at 3000 ϫ g for 15 min and sonicated in phosphate-buffered saline. Triton X-100 was added to the lysate (final concentration 1%), and particulates were removed by centrifugation for 15 min at 5000 ϫ g (30). The clarified lysate was incubated with glutathione-agarose beads (Sigma) for 1 h at 4°C and then washed three times with ice-cold phosphate-buffered saline. Thereafter, the bound protein was eluted with 10 mM reduced glutathione in 50 mM Tris-HCL (pH 8.0) and concentrated using a Centricon-10 (Amicon). For analysis of proteins binding to the GST fusion proteins, lysates of LTD 4 -stimulated or unstimulated cells were prepared in 1.0 ml of lysis buffer (25 mM Tris-HCL (pH 7.6), 0.15 M NaCl, 1% Triton X-100, 50 mM NaF, 2 mM Na 3 Vo 4 , 60 g/ml phenylmethanesulfonyl fluoride, 4 g/ml leupeptin, and 1 g/ml aprotinin). GST fusion proteins (5-10 g) or GST alone were prebound to agarose beads and then incubated with 1.0 ml of the different cell lysates (1 mg/ml of proteins) for 2 h at 4°C. Thereafter, the beads were washed with an ice-cold lysis buffer supplemented with 500 mM NaCl.
Immunoblotting-The separated proteins were electrophoretically transferred to a polyvinylidene fluoride membrane, and the membrane was then blocked overnight at 4°C with 3% (w/v) bovine serum albumin and subsequently incubated with a primary antibody for 1 h at 20°C (diluting 1:1000 for anti-G␤, 1:200 for anti-c-Src, and 1:250 for anti-PLC-␥1). The membrane was then washed extensively and incubated with a horseradish peroxidase-linked goat anti-rabbit or anti-mouse antibody for 1 h at 20°C (dilution 1:10,000). Thereafter, the membrane was again washed extensively, incubated with ECL Western blot detection reagents, and finally exposed to hyperfilm-ECL (Amersham Pharmacia Biotech) to visualize immunoreactive proteins.
Measurements of c-Src Autophosphorylation-An antiserum specific for Tyr 418 -phosphorylated Src was used to detect autophosphorylation of endogenous c-Src (32). The proteins in clarified whole-cell lysates of unstimulated and LTD 4 -stimulated cells were resolved by SDS-PAGE and transferred to polyvinylidene fluoride membranes. The presence of Src proteins with a phosphorylated Tyr 418 residue was detected using an anti-Tyr 418 -Src antibody (dilution 1:500) and a secondary horseradish peroxidase-conjugated goat anti-rabbit IgG antibody (dilution 1:5000).
Determination of Cytosolic Free Ca 2ϩ Levels and Microinjection-The cells were incubated with 4 M fura-2/AM in culture medium for 30 min at 37°C (10). After fura-2 loading, the cells were microinjected with antibodies against G␤, PLC-␥1, or c-Src or with an isotyped control antibody using an Eppendorf micromanipulator 5171/transjector at an injection pressure of 150 -200 hectopascals and an injection time of 0.7 s. The antibody solutions were prepared in phosphate-buffered saline (pH 7.3, 100 -200 g/ml) and clarified by centrifugation at 15,000 ϫ g for 10 min. The microinjected cells were allowed to rest for 15 min before beginning calcium measurement. The calcium signal was detected using a system consisting of a NIKON Diaphot microscope connected to a Photon Technology International (PTI) imaging system. Images of cells before and after stimulation with LTD 4 were acquired using an excitation wavelength rapidly alternating between 340 and 380 nm while recording the emission at 510 nm. The fluorescence intensity ratios (340/380 nm) of the images were subsequently analyzed using PTI Image Master Software.

LTD 4 -induced Association of PLC-␥1with G␤␥ Subunits and c-Src in Int 407
Cells-To investigate a possible agonist-induced association of PLC-␥1 with G␤␥ subunits in the Int 407 cells, anti-PLC-␥1 immunoprecipitates from membrane fractions were immunoblotted with G␤ antibodies. LTD 4 induced increased interaction between G␤␥ and PLC-␥1 that was detected after only 15 s (Fig. 1). To determine whether c-Src is involved in the activation of PLC-␥, as has been demonstrated in other cell types as well as in vitro (27,33,34), we also re-probed the membranes with a specific anti-c-Src antibody and observed an agonist-induced association of c-Src with PLC-␥1. For both G␤␥ and c-Src, the association with PLC-␥1 began to decrease after 300 s (Fig. 1). To confirm this finding, we immunoblotted anti-G␤ immunoprecipitates with anti-PLC-␥1 and anti-c-Src antibodies and immunoblotted anti-c-Src-immunoprecipitates with anti-PLC-␥1 and anti-G␤ antibodies. In both cases, we confirmed an increased association between the mentioned proteins after exposure to LTD 4 , which is shown in Figs. 2 and 3. In all experiments we used isotype-matched nonimmune IgG antibodies as a control to exclude the possibility of unspecific associations (designated N in Figs. 1-3).
G␤␥ Subunits Associate with GST Fusion Protein of the PLC-␥1 PH Domain-To further strengthen and explore the results described above, we employed two different PLC-␥1 GST fusion proteins for their ability to associate with G␤␥ subunits and c-Src. We noted that GST fusion proteins of the PLC-␥ PH-SH2-SH2-SH3-PH domain, but not of the PLC-␥ SH2-SH2-SH3 domain, caused increased precipitation of G␤␥ subunits from lysates of LTD 4 -stimulated cells (Fig. 4B). We also investigated the ability of c-Src from cells exposed to LTD 4 to bind to these PLC-␥ GST fusion proteins, and we noted that both of these fusion proteins caused increased precipitation of c-Src (Fig. 4A).

Effects of c-Src Inhibition on the LTD 4 -induced Association of PLC-␥1 with G␤␥ Subunits-
We also tested the effects of the specific Src tyrosine kinase family inhibitor PP1 on the ability of LTD 4 to induce an association between PLC-␥1 and G␤␥ subunits. Anti-G␤ subunit immunoprecipitates were obtained from membrane fractions of cells incubated for 15 min in the absence or presence of 10 M PP1. Pretreatment with PP1 blocked two LTD 4 -induced effects, namely the association between PLC-␥1 and G␤␥ subunits (Fig. 5A) as well as tyrosine phosphorylation of PLC-␥1 (Fig. 5B). These data suggest that Src kinase activity is involved in the activation of PLC-␥1 and required for the LTD 4 -induced association between PLC-␥1 and G␤␥ subunits. We subsequently investigated autophosphorylation of the Tyr 418 residue on Src as a means of determining the activity of the kinase in unexposed cells and cells stimulated with LTD 4 ; data establishing a correlation between such autophosphorylation and Src activity has previously been reported (32,35). In analyzing whole-cell lysates, we found that treatment with LTD 4 resulted in a 3-fold increase in phosphorylation of the Tyr 418 residue of c-Src, an increase that was blocked by PP1 (Fig. 5C).

Effects of Anti-G␤ Subunit, Anti-PLC-␥1, and Anti-c-Src Antibodies on the LTD 4 -induced Ca 2ϩ
Signal-To evaluate the significance of the above-mentioned observations for the downstream signaling capacity of PLC-␥1, we microinjected antibodies against G␤ subunits, PLC-␥1, and c-Src and assessed the effects on the LTD 4 -induced cytosolic free calcium signal by imaging the responses of single cells. As shown in Fig. 6 and Table I, separate injection of anti-G␤, anti-PLC-␥1 and anti-c-Src antibodies significantly impaired the LTD 4 -induced calcium signal, whereas the isotype-matched control antibodies had only a negligible effect. In agreement with the results reported above, we found that pretreatment of cells with 10 M PP1 for 15 min abolished the LTD 4 -induced calcium signal (data not shown). These findings suggest that each of the investigated signaling proteins is essential for LTD 4 -induced intracellular release of calcium.

DISCUSSION
The results of this study are the first to show an association between PLC-␥1 and G␤␥ subunits after agonist stimulation in vivo. Furthermore, by using GST fusion proteins of different domains of PLC-␥1, we found that the LTD 4 -induced association between PLC-␥1 and G␤␥ subunits is most likely mediated by the PH domain of PLC-␥1. That possibility is supported by results obtained in vitro that demonstrate binding between a split PH domain from the PLC-␥1 molecule and purified G␤␥ subunits (29). In our experiments, the interaction between PLC-␥1 and G␤␥ subunits appeared to be regulated by the number of G␤␥ subunits that dissociated from their heterotrimeric complex, because LTD 4 increased the amount of G␤␥ subunits precipitated by the GST fusion protein of the PLC-␥ PH-SH2-SH2-SH3-PH domain. Such a signaling role of the PH domains of PLC-␥1 is also indicated by recent findings showing that the PH domain of the receptor kinase GRK2 is responsible for interaction with a G␤␥ subunit (36), which implies that G␤␥ subunits can interact with different signaling proteins provided their PH domain is properly exposed. We believe that such exposure of the PH domain of PLC-␥1 requires agonist stimulation and activation of a c-Src tyrosine kinase. This assumption stems from our findings that the interaction between PLC-␥1 and G␤␥ subunits was significantly increased in cells treated with LTD 4 and that this effect could be blocked by PP1, an Src family tyrosine kinase inhibitor. Several mechanisms have been proposed to explain how seven-transmembrane receptors activate Src tyrosine kinases. It was initially found that agonist-induced activation of the Src family kinases Src, Fyn, Yes, and Lyn in various types of cells is sensitive to pertussis toxin (37), which indicates that these kinases participate in G i -coupled receptor signaling pathways. Later, it was suggested that G␤␥ subunits induce coupling between heterotrimeric G-proteins and c-Src by mediating the formation of a Shc-Grb2-c-Src signaling complex (32). Alternatively, other investigators (38) have proposed that G␤␥ subunit-induced activation of Btk is caused by interaction with PH domains, which leads to a downstream activation of Src kinases via interaction with the SH3 domain.
The present results confirm the findings of Khare et al. (33), showing that agonist stimulation can induce a physical association between c-Src and PLC-␥; more specifically, the cited investigators treated colonocytes with 1,25(OH) 2 D 3 and observed an increased recovery of c-Src in PLC-␥1 immunoprecipitates. In addition, the present study extends these findings by suggesting that it is primarily the SH2 domain of PLC-␥1 that mediates the interaction with c-Src. It has been proposed that agonist-induced increase in the catalytic activity of PLC-␥1 is due to phosphorylation of tyrosine residues (21), and in our experiments that would have been mediated through activation of Src tyrosine kinase as indicated by an increased autophosphorylation of Src upon LTD 4 stimulation. Binding of Src tyrosine kinases to phosphatidylinositol 3-kinase has been found to increase the activity of the latter enzyme (39); in analogy, the interaction of c-Src and PLC-␥1 may represent an additional mechanism that increases the activity of PLC-␥1.
The importance of G␤␥ subunits for the agonist-induced PLC-␥1-mediated downstream calcium signal is best demonstrated by our observation that microinjection of G␤ antibodies before LTD 4 treatment abolished the mobilization of calcium induced by the agonist. We have previously found that such calcium mobilization is blocked by compactin, an inhibitor of protein isoprenylation (23). This finding can at least partly be due to decreased isoprenylation of G␤␥ subunits and to an inadequate association of these proteins with the plasma membrane (40). It has been suggested that an important function of G␤␥ subunits is to directly recruit proteins comprising a PH domain to the cell membrane (29,41,42), and our results support such a role for G␤␥ subunits in the LTD 4 -induced mobilization of calcium. However, it has also been shown that G␤␥ subunits can indirectly recruit Shc and Grb2 proteins to the membrane via activation of c-Src and tyrosine phosphorylation of epidermal growth factor receptors (43). Our data imply direct binding between G␤␥ subunits and PLC-␥ but indirect binding between G␤␥ subunits and c-Src.
We have previously demonstrated that LTD 4 induce activation of G i3 proteins, as indicated by an increase in the GTP/ GDP ratio (22). Considering our present results, at least one additional pertussis toxin-insensitive G-protein seems to be the source of the free G␤␥ subunits that take part in the LTD 4induced interaction with PLC-␥1. This assumption is based on the following observations: activation of the G i3 -protein is sensitive to pertussis toxin; pertussis toxin impairs the interaction between such G-proteins and seven-transmembrane receptors; and LTD 4 -induced mobilization of calcium is insensitive to pertussis toxin (10). This interpretation of our findings is in accordance with the well-established concept that seven-transmembrane receptors can activate several different types of heterotrimeric G-proteins (7,10,11). Moreover, in differentiated human U937 cells, the LTD 4 -induced Ca 2ϩ response is only partially blocked by pertussis toxin, suggesting that this receptor is coupled to pertussis toxin-sensitive as well as insensitive G-proteins (11). The identity of the indicated type of LTD 4 -activated pertussis toxin-insensitive G-protein(s) remains to be discovered.
Our study is the first to show an agonist-induced in vivo association between G␤␥ subunits and the downstream signaling protein PLC-␥1. This interaction requires both an increase in free G␤␥ subunits and proper exposure of the PH domain of PLC-␥1. These results along with the other findings of our investigation lend additional support to the recent suggestion that G␤␥ subunits can directly or indirectly interact with more than one protein and potentially act as docking proteins, leading to the formation of an intracellular signaling complex (44).