Tyrosine Phosphorylation of a 94-kDa Protein of Human Fibroblasts Stimulated by Streptococcal Lipoteichoic Acid*

Lipoteichoic acid (LTA) is an amphipathic compo- nent of Gram-positive bacteria. Previous studies from this laboratory have shown that at low concentrations, ranging from 0.1 to 10.0 pglml, LTA binds to mam- malian cells and stimulates mitogenic responses as demonstrated by increased DNA and RNA synthesis. Tyrosine kinase appears to be involved in the action of a number of mitogens including epidermal growth fac- tor, platelet-derived growth factor, and insulin. In the present study, we report the novel finding that tyrosine protein kinase activity is increased in human fibro- blasts treated with LTA. Sodium dodecyl sulfate-po-lyacrylaide gel electrophoresis and autoradiography of the whole cell lysate of fibroblasts cultured with 32Pi showed increased phosphorylation of a 94-kDa polypeptide. Alkali treatment of the gel resulted in a de- creased intensity of the 94-kDa phosphorylated protein in control cells, but not in LTA-treated cells, sug-gesting the addition of phosphate groups to threonine or tyrosine residues. High voltage electrophoresis of the acid hydrolysate of the excised and eluted 94-kDa protein revealed that LTA stimulated the phosphorylation of tyrosine but not threonine residues. These results suggest that LTA acts on mammalian cells by phosphorylating tyrosine residues of certain proteins and thereby may regulate diverse functions of these cells. dium supplemented with 2% FCS. LTA was present throughout the labeling procedure, Incorporation of radiolabel into the insoluble fraction was measured by collection on filter paper discs using a Skatron cell harvester and quantitating in a scintillation counter. Isolation of LTA from Streptococci-LTA was extracted and puri- fied from the 1RP41 strain of group A streptococci, and the final product was determined to be pure by chemical analyses as described previously (32). Radiolabeling and LTA Stimulation of Cells-The cells were allowed to reach confluency (72 h) in the presence of various concentrations of LTA. The confluent cells were washed three times with 10 ml of phosphate-free minimum essential medium supplemented with 5% FCS dialyzed against 0.15 M NaCl. The cells were then labeled with 1 mCi of 32Pi/dish in phosphate-free minimum essential medium with 10% FCS for 16 h at 37 “C in the presence of LTA. In order to minimize the effect of phosphate ions present in FCS, one-half of the added FCS was dialyzed against 0.15 M NaCl. After the incubation, the plates were chilled at 4°C and washed three times with buffer A (120 mM NaC1, 4.75 mM KC1, 1.2 mM MgSO,, 2 mM CaC12, 24 mM NaHC03, 2.5 mM glucose). Cells were removed with a rubber police-man and pelleted at 10,000 X g in buffer A. Lysis of the cells was carried out in the presence of buffer B (10 mM Tris-HC1, pH 7.0, EDTA, 10 mM NaF, mM benzamidine mM 0.25% dimethyl sulfoxide, (v/v) min centrifuged further To determine the incorporation of 32Pi into cellular proteins, 10-20 111 of the cell lysate from the 178,000 X g supernatant was precip- itated with 10% t.richloroacetic acid. The precipitate was filtered a glass microfiber filter and washed three with 10% trichloro- acetic acid. Dried filters were counted in a scintillation mixture.

Lipoteichoic acid (LTA) is an amphipathic component of Gram-positive bacteria. Previous studies from this laboratory have shown that at low concentrations, ranging from 0.1 to 10.0 pglml, LTA binds to mammalian cells and stimulates mitogenic responses as demonstrated by increased DNA and RNA synthesis. Tyrosine kinase appears to be involved in the action of a number of mitogens including epidermal growth factor, platelet-derived growth factor, and insulin. In the present study, we report the novel finding that tyrosine protein kinase activity is increased in human fibroblasts treated with LTA. Sodium dodecyl sulfate-polyacrylaide gel electrophoresis and autoradiography of the whole cell lysate of fibroblasts cultured with 32Pi showed increased phosphorylation of a 94-kDa polypeptide. Alkali treatment of the gel resulted in a decreased intensity of the 94-kDa phosphorylated protein in control cells, but not in LTA-treated cells, suggesting the addition of phosphate groups to threonine or tyrosine residues. High voltage electrophoresis of the acid hydrolysate of the excised and eluted 94-kDa protein revealed that LTA stimulated the phosphorylation of tyrosine but not threonine residues. These results suggest that LTA acts on mammalian cells by phosphorylating tyrosine residues of certain proteins and thereby may regulate diverse functions of these cells.
Lipoteichoic acids (LTAsl) are amphipathic molecules synthesized and excreted by many Gram-positive bacteria. As a surface component of certain bacterial cells such as Streptococcus pyogenes, LTA has been shown to mediate the adherence of these organisms to mammalian epithelial cells (1-3). In its free form, LTA exhibits other biological activities * These studies were supported by research funds from the United States Veterans Administration and by Research Grants AI-10085 and AI-13550 from the United States Public Health Service. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "uduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. including the stimulation of bone resorption in tissue culture (4), the inhibition of platelet aggregation by collagen ( 5 ) , and the production of nephrocalcinosis in rabbits (6). In addition, several reports have suggested that LTA may exert a mitogenic or cytotoxic effect on mammalian cells in culture, depending on the concentrations of LTA used (7-11). It is interesting to note that lipopolysaccharide (LPS), an amphipathic molecule from Gram-negative bacteria, also acts as a mitogen on bone marrow-derived mouse lymphocytes (12). LPS stimulates macrophages to release prostaglandins (13) and neutral proteinases (14,15); the biological actions of LPS have been attributed to the lipid A moiety (16,17). Recently, free lipid A has been shown to stimulate protein kinase C of macrophages (18). Thus, the mitogenic properties of LPS may reside in the capacity of the molecule to induce the phosphorylation of biologically active proteins.
Covalent modification of tyrosine residues of proteins by phosphorylation induced by tyrosine kinase is thought to play a central role in the action of a number of mitogens including epidermal growth factor, insulin, and platelet-derived growth factor (19-26). These factors bind to their respective membrane receptors and thereby stimulate the autophosphorylation of the receptor proteins at tyrosine residues (20, 27-29). Tyrosine kinase-induced phosphorylation also has been implicated in the transformation of infected cells by Rous and Fujinami viruses (30,31).
In this paper, we report studies of the effect of LTA in mitogenic doses on human foreskin fibroblasts in tissue culture. We show that LTA stimulates phosphorylation of the tyrosine residues of a 94-kDa protein produced by these cells. These studies provide new insights into the mechanism of the biological activity of streptococcal lipoteichoic acids.

EXPERIMENTAL PROCEDURES
Material~-'~Pi was obtained from ICN Radiochemical. [3H]Leucine and [3H]thymidine were purchased from New England Nuclear. All tissue culture reagents were from Gibco Laboratories (Grand Island, New York). Cellulose-precoated TLC plates (0.1 mm) were purchased from Merck. All other reagents were obtained from Sigma. Solvents were from Mallinckrodt Chemical Works and were of analytical grade.
Determination of DNA Synthesis and Protein Synthesis-DNA and protein syntheses were measured in stimulated as well as unstimulated cells by determining the incorporation of [3H]thymidine and [3H]leucine, respectively. Cells to be assayed were pulse-labeled for 18 h with 1 pCi of [3H]thymidine in the standard incubation medium or with 1 pCi of [3H]leucine in leucine-free minimum essential me-LTA-stimulated Tyrosine Phosphorylation 13343 dium supplemented with 2% FCS. LTA was present throughout the labeling procedure, Incorporation of radiolabel into the insoluble fraction was measured by collection on filter paper discs using a Skatron cell harvester and quantitating in a scintillation counter. Isolation of LTA from Streptococci-LTA was extracted and purified from the 1RP41 strain of group A streptococci, and the final product was determined to be pure by chemical analyses as described previously (32).
Radiolabeling and LTA Stimulation of Cells-The cells were allowed to reach confluency (72 h) in the presence of various concentrations of LTA. The confluent cells were washed three times with 10 ml of phosphate-free minimum essential medium supplemented with 5% FCS dialyzed against 0.15 M NaCl. The cells were then labeled with 1 mCi of 32Pi/dish in phosphate-free minimum essential medium with 10% FCS for 16 h at 37 "C in the presence of LTA. In order to minimize the effect of phosphate ions present in FCS, one-half of the added FCS was dialyzed against 0.15 M NaCl. After the incubation, the plates were chilled at 4°C and washed three times with buffer A (120 mM NaC1, 4.75 mM KC1, 1.2 mM MgSO,, 2 mM CaC12, 24 mM NaHC03, 2.5 mM glucose). Cells were removed with a rubber policeman and pelleted at 10,000 X g in buffer A. Lysis of the cells was carried out in the presence of buffer B (10 mM Tris-HC1, pH 7.0, 2 mM EDTA, 10 mM NaF, 8 mM benzamidine hydrochloride, 1 mM phenylmethylsulfonyl fluoride, 0.25% (v/v), dimethyl sulfoxide, 1.5% (v/v) Triton X-100 for 30 min in ice. The mixture was centrifuged in a Beckman Airfuge at 178,000 X g for 15 min. The pellet was discarded and the supernatant was used for further studies.
To determine the incorporation of 32Pi into cellular proteins, 10-20 111 of the cell lysate from the 178,000 X g supernatant was precipitated with 10% t.richloroacetic acid. The precipitate was filtered on a glass microfiber filter and washed three times with 10% trichloroacetic acid. Dried filters were counted in a scintillation mixture.
Gel Electrophoresis.-An aliquot of 20-30 111 of samples from the 178,000 X g supernatant was subjected to 6-15% gradient SDSpolyacrylamide gel electrophoresis (SDS-PAGE) according to the method of Laemmli (33). The gels were stained, destained, dried, and exposed to XAR-5 (Kodak) film for autoradiography as described (34). Densitometric scanning of the autoradiogram was performed with an LKB 2202 Ultroscan Laser Densitometer integrated with a Gelscan program.
Alkali Treatment of Gels-After SDS-PAGE, the gel was treated with 1 N NaOH as described previously (35). The dried gel was then subjected to autoradiography (34).
Phosphoamim Acid Analyses-The supernatants from the lysed cell samples centrifuged at 178,000 X g were precipitated with cold acetone and allowed to settle for 18 h for complete purification at -20 "C. The pellets were then washed with 5 ml each of acetone, dehydrated ethyl alcohol, ch1oroform:methanol (l:l), dehydrated ethyl alcohol, anhydrous diethyl ethermethanol (l:l), and anhydrous diethyl ether. The pellets were dissolved in 6 N HC1 and hydrolyzed under Nz at 110°C for 2 h. The HC1 was removed and the hydrolysates were subjected to one-dimensional high voltage electrophoresis on cellulose-precoated TLC plates (Merck) at pH 3.5 in acetic acidpyridine:water according to the method of Hunter and Sefton (36). The protein was eluted from the excised band from an SDSpolyacrylamide gel according to the method described previously (37). The eluted protein was hydrolyzed and subjected to high voltage electrophoresis (36).

RESULTS AND DISCUSSION
In order to study the mitogenic effect of LTA on human foreskin fibroblasts, DNA and protein synthesis were assayed in unstimulated and LTA-stimulated cells. In a typical experiment, the cells were treated with LTA at concentrations of 1, 10, and 100 pg/ml of culture medium for 72 h, and incorporation of [3H]thymidine and [3H]leucine was measured. The maximum incorporation of [3H]thymidine was observed in cells treated with 1 pg/ml LTA (Fig. 1); incorporation decreased as the concentration was increased to 10-100 pg/ml. The incorporation of C3H]leucine followed a similar pattern (Fig. 1). The incorporation of the radiolabeled amino acid reached a maximum at 1 pg/ml LTA but was suppressed to the untreated control level at 100 pg/ml. The total incorporation of 32Pi into the cellular proteins reached a maximum at a concentration of 5 pg/ml LTA (Fig. 1, inset). At higher concentrations of LTA, the uptake of 32P into the trichloroacetic acid-precipitable proteins was suppressed in parallel with the suppression of DNA and protein synthesis. The absolute level of phosphorylation varied from experiment to experiment probably because of variations in the state of confluency and the number of passages the test cells had undergone before each experiment.
Previous studies from this laboratory and other groups have demonstrated that at low concentrations, LTA acts as a mitogen, whereas at high concentrations it acts as a cytotoxic agent (7-9). The molecular mechanism(s) of mitogenicity of LTA were not explored in these reports. Several reports indicate that the mitogenic effects of various biological polypeptides including epidermal growth factor, platelet-derived growth factor, and insulin are associated with phosphorylation of cellular proteins at tyrosine residues (19-29). Since tyrosine-specific protein kinase activities are increased in concert with the phosphorylation of growth factor receptors, it is believed that the biological effects of these agents require the cdvalent modifications of certain polypeptides.
In order to determine whether or not LTA stimulates the phosphorylation of cellular proteins, human foreskin fibroblasts were incubated with 32Pi in the absence or presence (0.01-25 pg/ml) of LTA. SDS-polyacrylamide gel electrophoresis and autoradiography of the 178,000 x g supernatant of whole cell lysate showed that LTA stimulated an increased phosphorylation of two proteins migrating at 98 and 94 kDa (Fig. 2). Phosphorylation of the 94-kDa protein was maximal (2.5-fold) at 5 pg/ml LTA, as shown by densitometric scanning of the autoradiogram, whereas that of the 98-kDa polypeptide was random (data not shown).
It is known that the insulin receptor (350 kDa) is composed of two a-subunits (120-130 kDa) and two P-subunits (90-95 kDa) which are linked by disulfide bonds (38-40). In addition, . The 178,000 X g supernatant was boiled with SDS sample buffer containing 62.5 mM Tris-HC1, pH 6.8, 3% SDS, 5% 2-mercaptoethanol, 10% glycerol and electrophoresed in a 6-15% gradient SDS gel. The gels were stained, destained, dried, and exposed to XAR-5 (Kodak) film for autoradiography. The molecular weight standards used were: myosin, M , = 200,000; 0-galactosidase, 116,000; phosphorylase b, 94,000; bovine serum albumin, 67,000; ovalbumin, 45,000; carbonic anhydrase, 31,000; and soybean trypsin inhibitor, 21,000. the 0-subunit of the receptor undergoes phosphorylation in the presence of insulin (21,(41)(42)(43)(44). Like insulin, type 1 insulin-like growth factor is composed of two a-subunits (130 kDa) and two @-subunits (98 kDa) and are linked by sulfhydryl groups (45,46). Since the @-subunit of insulin receptor and the type-1 insulin-like growth factor have a molecular mass of approximately 94 kDa, it was of interest to explore the identity of the 94-kDa phosphorylated polypeptide present in LTA-stimulated fibroblasts. Electrophoresis of 32P-labeled whole cell lysate from fibroblasts under reducing and nonreducing conditions showed no discernible difference in the migration or intensity of the 94-kDa phosphorylated polypeptide (data not shown), indicating that the 94-kDa polypeptide is not part of the insulin receptor complex.

LTA-stimulated Tyrosine Phosphorylation
It has previously been shown that a 98-kDa cellular protein (NCP 98) from normal chicken bone marrow cell extracts cross-reacts and is functionally homologous with the Fujinami sarcoma virus transforming protein P140 (47). NCP 98 is a phosphoprotein with associated kinase activity capable of phosphorylating itself and other protein substrates specifically at tyrosine residues. Therefore, attempts were made to identify the phosphoamino acids of the 98-kDa as well as the 94-kDa protein stimulated by LTA.
In order to determine the degree of phosphorylation of serine as opposed to tyrosine or threonine residues, SDS gels containing the separated cellular proteins were exposed to alkali (35). The phosphate-ester linkages of serine are more labile in alkali than are those of tyrosine and threonine. Alkali treatment did not alter the relative intensities of the 98-kDa band either from unstimulated or from LTA-stimulated cells (Fig. 3). In contrast, the phosphorylated 94-kDa protein band in unstimulated control cells disappeared completely after alkali treatment, whereas that of the LTA-stimulated cells remained intense (Fig. 3). These results suggested that LTA stimulated phosphorylation of the 98-kDa protein mainly at serine residues, while stimulating that of the 94-kDa protein at threonine or tyrosine residues.
In order to distinguish the degree of phosphorylation of serine and threonine as opposed to tyrosine in the 94-kDa protein, we performed high voltage electrophoresis of the acidhydrolyzed proteins precipitated with acetone from lysates of cells stimulated with LTA. LTA stimulated a dramatic increase in phosphotyrosine but only a slight increase in phosphothreonine residues (Fig. 4). It should be noted that the absolute stimulation .of tyrosine kinase activity varied slightly from experiment to experiment with a resultant change in the ratio of serine to tyrosine phosphate residues. The degree of simulation by LTA of tyrosine phosphorylation appeared to be related to the number of cell passages the fibroblasts had undergone before the experiments were performed (48). ,000 x g supernatants from LTA-treated and control cells labeled with 32Pi were electrophoresed in a 6-15% gradient SDS gel. The gel was stained with Coomassie Brilliant Blue R-250 and treated with 1 N NaOH for 1 h at 40°C. The gel was rinsed in 7.5% (v/v) acetic acid, 10% (v/v) methanol and subjected to autoradiography as described previously except that the exposure time was doubled. The standards used were the same as for Fig. 2. Procedures." The cells were lysed in buffer B and subjected to centrifugation. The supernatant from the 178,000 X g centrifugation was precipitated in the presence of cold acetone and washed as described under "Experimental Procedures." The precipitate was dissolved in 6 N HCl and hydrolyzed for 2 h at 110°C under Nz. The hydrolysate was dried and dissolved in water. The high voltage electrophoresis was carried out as described under "Experimental Procedures." The phosphoamino acids were identified by co-migration of the radioactivity with 10 pl of internal standards of phosphoserine (Ser (P)) (1 mg/ml), phosphothreonine (Thr(P)) (1 mg/ml), and phosphotyrosine (Tyr(P)) (1 mg/ml). The standards were detected by spraying with 0.2% ninhydrin in acetone. lysed as described under "Experimental Procedures," and the 178,000 X g supernatant was subjected to SDS-PAGE on a 6-15% gradient gel. In order to resolve the 94-and 98-kDa polypeptides, electrophoresis was continued for 15 min at 30 mA/1.5-mm slab gel after the dye front had reached the end of the gel. The 94-kDa polypeptide was localized by autoradiography and excised from the gel. The protein was eluted with 40 mM NH4C03, 0.1% SDS and 0.5% 2-mercaptoethanol. 100 pg of bovine serum albumin was added as a carrier, and the protein was then precipitated with 20% trichloroacetic acid and washed sequentially with ethanol and ethano1:ether (1:l). The pellet was acid hydrolyzed at 110°C for 2 h and electrophoresed for 45 min at 1 kV. Internal standards were used as for Fig. 4. To confirm that tyrosine was the major residue of the 94-kDa protein that was phosphorylated as a result of LTA stimulation, the 94-kDa band was excised from an SDS gel containing the separated cellular proteins. The polypeptide was eluted with 50 mM NH4C03, 0.1% SDS, hydrolyzed in tyrosine residues in the 94-kDa protein (Fig. 5); phosphotyrosine could not be detected in the proteins obtained from unstimulated control cells, and phosphothreonine was not detectable in this protein obtained from either unstimulated or LTA-stimulated cells (Fig. 5).

LTA-stimulated
In order to correlate increases in DNA synthesis in the presence of LTA with the phosphorylation of the cellular proteins, simultaneous time course experiments were performed (Fig. 6). Cells were stimulated with 5 pg/ml LTA since at this concentration the incorporation of 32Pi into proteins as evidenced by trichloroacetic acid precipitation was maximum (Fig. 1). Following the addition of LTA, cells were labeled with [3H]thymidine or 32Pi at 24,48, and 72 h (Fig. 6).
[3H]Thymidine uptake was maximum at 72 h, as was phosphorylation of the 94-kDa protein. The incorporation of 32Pi into fibroblast proteins was analyzed by lysing the cells and electrophoresing the 178,000 X g supernatant on a 6-15% gradient gel. The autoradiogram of the dried gel showed a consistent increase in the phosphorylation of the 94-kDa polypeptide with time, whereas in the control cells the phosphorylation of this protein remained the same as determined by densitometric scanning.
Preliminary studies indicate that the 94-kDa protein is located in the cell membrane. When LTA-stimulated and unstimulated cells were lysed by Dounce homogenization in 0.25 M sucrose, 10 mM Tris-HC1, 0.2 mM MgClz, 5 mg/ml bovine serum albumin, pH 7.4, followed by sucrose density gradient centrifugation, the 94-kDa protein was found in the membrane particulate fraction but not in the supernatant. Because of its membrane location, this protein becomes a candidate as a membrane receptor for the binding of LTA.
Although the mitogenic effects of amphipathic molecules, including LTA and LPS, have been investigated, no molecular mechanism of these actions has been elucidated. Recently, it has been reported that biologically active lipid moieties of LPS activate protein kinase C (18), indicating that the phosphorylation of certain polypeptides may be relevant to the biological activity of LPS. The studies of LPS were carried out by using partially purified protein kinase C. In the present study, enhancement of tyrosine kinase activity as well as phosphorylation of a cellular protein at tyrosine residues was observed in LTA-stimulated cells. Although there is no direct proof that enhancement of the phosphotyrosine level in the presence of this amphipathic molecule is sufficient to bring about the mitogenic effect of LTA, it is, nevertheless, possible that the covalently modified protein(s) plays an important role in the biochemical events associated with the biological properties of streptococcal lipoteichoic acid.