Autoanti-phosphatidylinositide antibodies specifically inhibit noradrenaline effects on Ca2+ and Cl-channels in rat portal vein myocytes.

specifically inhibit noradrenaline effects on

(4,5)P2)' by phospholipase C with the consequent generation of two second messengers: inositol 1,4,54risphosphate (InsPJ and diacylglycerol (DAG).InsPs mobilizes intracellular Ca2+ whereas DAG activates protein kinase C isoforms to phosphorylate a specific set of proteins (1-5).Activation of aladrenoceptors in portal vein smooth muscle stimulates L-type Ca2+ channels (6).This effect is dependent on the activation of protein kinase C and may involve subsequent phosphorylation of intracellular proteins.InsPB mediates the release of Ca2+ from internal stores, initiating Ca2+-dependent activation of C1-channels.In addition, coupling between w-adrenoceptors and phospholipase C is mediated through activation of a G-protein insensitive to pertussis toxin (6).
Circulating autoantibodies (auto-Ab) directed against phosphatidylinositol (PtdIns) have been identified in highly diluted sera (1/15,000) of patients with malignant tumors using a reliable immunoenzymatic assay (7, 8).The statistically significant immunological response was found in sera of all patients with malignancies irrespective of type, grade, or organ localization, and histological origin of tissues.High levels of the auto-Ab were also found in sera of female Sprague-Dawley rats with chemically induced tumors (9).Immunochemical analyses of both human and rat auto-Ab with adsorption tests and competition experiments in an ELISA system have shown a higher avidity and specificity for PtdIns than for the other phosphatidylinositides (7-9).In view of the immunochemical properties of the autoanti-PtdIns Ab purified from human sera we tested their effects on the intracellular transduction pathway of vascular smooth muscle cells involving PtdIns as a second messenger precursor (6).We find that antibodies to phosphatidylinositides block specifically the noradrenaline-induced responses, i.e. activation of Ca2+-dependent C1-current and enhancement of Ca2+ channel current.In contrast, they alter neither the function and modulation of C1-and Ca2+ channels induced by caffeine or phorbol esters nor the transduction pathway after 8-adrenoceptor activation.These findings confirm that PtdIns or PtdIns metabolites are involved in the transduction processes activated by occupancy of al-adrenoceptors.
patients with malignant leukemia were supplied by the Centre Fran-Gois Magendie, Haut-LBv6que Hospital (Bordeaux).Control blood samples of healthy donors were given by the Centre de Transfusion Sanguine (Bordeaux).
Purification of Human Sera-One ml of each serum was precipitated with a 50% (NH4),S04 solution followed by dialysis against a lo-' M phosphate buffer containing 4 g/liter NaCl, pH 7.4, for 72 h at 4 "C.The sera were then Sephadex G-200 chromatographed through a column (50 X 1 cm, Pharmacia LKB Biotechnology Inc.) and eluted with the same buffer.Each fraction was subjected to an ELISA for evaluating its anti-PtdIns activity.Ab avidity and specificity were assayed by competition experiments as described below.
Zmmunoenzymatic Test for Detection and Characterization of Autoantibodies in Human Sera-The anti-phosphatidylinositide activity in sera of patients with malignant tumors (leukemia, breast cancer, gastrointestinal cancer) and of healthy donors was assayed with an ELISA (7,8).Human auto-Ab avidity and specificity were evaluated by competition experiments.They were performed as follows.Each compound (PtdIns containing either stearic and arachidonic acids or linoleic and palmitic acids, phosphatidylinositol 4-monophosphate (PtdIns(4)P), PtdIns(4,5)Pz, myo-inositol, InsPs, 1-oleoyl-2-acetylglycerol (OAG) and phosphatidylserine (PtdSer), all from Sigma) was preincubated, at various concentrations (from lo-' to lo-" M), with human sera diluted 15,000-fold or purified IgG fractions diluted 15,000-fold (7.10-' mg/ml) in a lo-' M phosphate buffer containing 0.15 M NaCl and 10% glycerol.After 18 h at 4 "C, each mixture was applied onto well plates coated with PtdIns or thyroglobulin.The absorbances in well plates were measured at 492 nm with a Multiskan spectrophotometer.Experimental values were corrected by subtracting blank values read on well plates coated with thyroglubulin.The specific binding was revealed with anti-human IgG Ab (diluted 20,000-fold) labeled with peroxidase and orthophenylenediamine, as described previously (7,8).
Adsorption of Purified Fractions-To determine anti-PtdIns activity, purified IgG fractions were preincubated or not with a M PtdIns solution for 16 h at 4 "C.Then both fractions were assayed in an ELISA for their anti-PtdIns activity.The binding was revealed as indicated above.
Electrophysiological Recordings-Single rat portal vein myocytes were isolated by enzymatic dispersion similar to that described previously (10) and maintained in short term primary culture for 40 h.Whole cell membrane currents were measured at 30 2 1 "C using the standard patch-clamp technique (11).For the recordings of C1-current, the extracellular medium (reference solution) contained (in mM): NaCl, 130; KCl, 5.6; CaC12, 2.5; MgCl,, 1; glucose, 11; HEPES, 8, pH 7.4, with NaOH.The basic pipette solution contained (in mM): CsCl, 130; HEPES, 10, pH 7.3, with CsOH.CsCl was used instead of KC1 in the pipette solution to block outward potassium currents.For the recordings of Caz+ channel current, 5 mM BaCl? was substituted for CaC1, in the reference solution, and 5 mM EGTA, 2.5 mM Na2ATP, 1 mM MgCl, were added to the basic pipette solution.
For the studies of Ca2+-activated C1-current, noradrenaline (Sigma) and caffeine (Merck) were applied with a pressure ejector from a glass pipette close to the cell.The effects of noradrenaline, isoprenaline, and phorbol 12,13-dibutyrate (Sigma) on Ca2+ channel currents were recorded during bath perfusion of these agents.Phentolamine, prazosin, and propranolol (Sigma) were used as a-and 8adrenoceptor blocking agents.
Currents were filtered at 1 kHz (-343 frequency), stored, and analyzed using an IBM PC computer (P-clamp system, Axon) as described previously (12).Capacitive and leakage currents were digitally subtracted.The values given are the means f S.E., with n the sample size.Significance was tested by means of Student's t test.

RESULTS
Avidity and Specificity of Auto-Ab from Human Sera Purifications-The anti-PtdIns activity was expressed with the following ratio: R = optical density (OD) of cancer patients' sera or IgG fractions -OD of controls sera or control IgG fractions/OD of controls sera or control IgG fractions.For sera used in these experiments ratios were found between 1.2 and 3.25.After Sephadex G-200 chromatography, the anti-PtdIns activity was always found in the IgG elution peaks, as noticed previously (9).The ratio values of the selected fractions were increased to 2.05 and 4.05 in comparison with control purified IgG, indicating a high level of anti-PtdIns in sera of breast cancer patients ( n = 50).Fig. 1 shows the avidity and specificity of one of these IgG fractions of breast cancer serum, given by displacement curves established from competition experiments.The best avidity determined a t halfmaximal absorbance (BIBo = 0.5) was found for PtdIns.The avidity for PtdIns was identical when the fatty acids were either stearic and arachidonic acids or linoleic and palmitic acids ( n = 4).In contrast, the autoanti-Ab recognized about 15-fold less both PtdIns(4,5)Pz and PtdIns(4)P and 100-fold less myo-inositol and InsP3.Nonsignificant displacement was observed with OAG, a synthetic analog of DAG and with PtdSer.Other membrane phospholipids and lipids tested such as phosphatidylcholine, phosphatidylethanolamine, gangliosides, sphingosine, sphingomyelin, and cardiolipin were not recognized at all.All IgG fractions purified from 150 various cancer sera (breast, gastrointestinal tract, leukemia) showed the same Ab avidity and specificity whereas no effect was observed with purified control IgG (30 sera).Although the recognition of auto-Ab sites is mainly related to the presence of PtdIns molecules it is likely that several antibodies that specifically recognized PtdIns, PtdIns(4)P, or PtdIns(4,5)Pz are present in the IgG fraction of cancer sera.

Effects of Autoanti-Ptdlns Antibodies on Noradrenaline
Stimulation of Ca2+ and Cl-Channels-Autoanti-PtdIns Ab from breast cancer sera were used to prove directly the involvement of PtdIns breakdown in the noradrenaline effects in vascular smooth muscle cells.Incubation of the cells in external reference solution containing autoanti-PtdIns Ab a t concentrations ranging between 0.03 and 0.3 mg of IgG/ml for 2 or 24 h had no effect on either the maximal Ca2+ channel current elicited by a membrane depolarization to 0 mV from a holding potential of -60 mV or the C1-current elicited by ejection of M noradrenaline ( n = 8).In the presence of external autoanti-PtdIns Ab the Ca2+ channel current was increased by 55.5 f 6.1% ( n = 5) after the addition of M NA.This stimulatory effect was not significantly different from that obtained in the absence of autoanti-PtdIns Ab (65.1 & 6.5%, n = 6 , p > 0.05).It has to be noted that the stimulatory effects of noradrenaline were blocked in M phentolamine and M prazosin whereas, they remained unchanged in M propranolol, indicating that the responses to noradrenaline were mediated by al-adrenoceptors.
In contrast, alterations of the noradrenaline-induced responses were observed when the autoanti-PtdIns Ab were intracellularly applied.Fig. 2A shows a typical C1-current induced by noradrenaline ejection M) in a cell maintained at a holding potential of -60 mV, under control conditions.Noradrenaline was effective in 80% of the tested cells ( n = 20), and the mean amplitude of the noradrenalineinduced C1-current was -279.3 f 50.1 pA ( n = 16).When the purified IgG (0.012 mg/ml) from breast cancer sera was added to the basic pipette solution, the noradrenaline-induced C1-current was completely abolished in 80% of the tested cells (n = 20; Fig. 2B) whereas in the remaining cells noradrenaline was still able to activate a C1-current, but the mean current amplitude measured at a holding potential of -60 mV was strongly reduced to -50.8 f 30.6 pA ( n = 4, p < 0.01).In addition, when the cells were injected with control purified IgG from healthy patients at a concentration identical to that used for the autoanti-PtdIns Ab, no significant modifications of the mean amplitude of the noradrenaline-induced C1-current (Fig. 2C) were observed in comparison to control cells ( p > 0.05, n = 6).Finally, the intracellular application of the purified IgG fractions (0.012 mg/ml) from breast cancer sera preincubated with a lo-' M solution of PtdIns was also without effect (Fig. 2 0 ) .Six of the seven tested cells displayed a C1-current in response to noradrenaline ejection, and the mean amplitude of the C1-current measured at a holding potential of -60 mV was not significantly modified ( p > 0.05, n = 6).These results suggest that the inhibition of the noradrenaline-induced C1-current is caused by the anti-PtdIns activity of the IgG fraction of breast cancer sera.
We also tested the effects of the autoanti-PtdIns Ab on the noradrenaline-induced enhancement of the Ca2+ channel current.Under control conditions, the application of noradrenaline (lo-' M) produced an increase in the amplitude of the Ca2+ channel current evoked by depolarization to 0 mV from a holding potential of -60 mV (Fig. 3A).This increase was observed in 80% of the cells ( n = lo), and the mean percentage of current increase was 65.9 k 7.9% ( n = 8).Intracellular application of the purified IgG from breast cancer sera dose dependently reduced the stimulation induced by M noradrenaline.The effects of various concentrations of IgG from the same batch are illustrated in Fig. 3.In the presence of 0.0004 mg of IgG/ml (Fig. 323) and 0.0012 mg of IgG/ml (Fig. 3C) in the pipette solution, the noradrenaline-induced enhancement of the Ca2+ channel current was reduced to 40.7 f 2.4% ( n = 3) and 26.7 f 8.2% (n = 4), respectively.Complete  The noradrenaline-induced increase in Ca2+ channel current was dose dependently inhibited by autoanti-PtdIns Ab.In contrast, when 0.012 mg/ml purified IgG from healthy patients ( E ) or when 0.012 mg/ml IgG from breast cancer serum incubated with a lo-' M solution of PtdIns (F) was added to the pipette solution, the increase in the Ca2+ channel current induced by noradrenaline was similar to that obtained under control conditions.Experimental data were obtained from different cells.
inhibition of the noradrenaline effect on the Ca2+ channels was obtained by increasing the concentration of autoanti-PtdIns Ab to 0.012 mg of IgG/ml (n = 7; Fig. 30).Intracellular applications of either purified IgG from healthy patients (0.012 mg/ml; Fig. 3E) or purified IgG from breast cancer sera preincubated with a lo-' M solution of PtdIns (0.012 mg/ ml; Fig. 3F) had no effects on the increase of the Ca2+ channel current induced by Therefore, the inhibition of the noradrenaline-induced enhancement of Ca2+ channel current by IgG fractions from breast cancer sera seems to be linked directly to the autoanti-PtdIns Ab found in these sera.Similar inhibitory effects of the noradrenaline-induced responses were also obtained with IgG from leukemia and gastrointestinal cancer sera ( n = 10).
A possible direct effect of the autoanti-PtdIns Ab on C1channels, Ca2+ channels, or intracellular Ca2+ stores was checked by using both caffeine and phorbol-12,13-dibutyrate (PDBu).Caffeine is known to release Ca2+ from intracellular stores whereas phorbol esters can directly activate protein kinase C. Fig. 4 shows the relative amplitude of the Ca2+ channel current evoked by repetitive depolarizations to 0 mV from a holding potential of -60 mV as a function of time.Under control conditions, 10"j M PDBu produced an increase in the Ca2+ channel current amplitude of 65.9 f 3.4% ( n = 11; Fig. 44).In the presence of autoanti-PtdIns Ab (0.012 mg of IgG/ml) in the pipette solution, noradrenaline ( W 5 M) had no effect on the Ca2+ channel current amplitude (Fig. 4B).After 7 min of continuous recording, the current amplitude was similar to that obtained at the beginning of the experiment, suggesting that autoanti-PtdIns Ab did not accelerate the run-down phenomenon which was weak at the stimulation frequency used (0.05 Hz; 12).A subsequent application of PDBu induced an increase in the Ca2+ channel current (66.5 f 11.0%, n = 6, p > 0.05) similar to that obtained in control conditions (Fig. 4B).These results indicate that the protein kinase C remains functional and that Ca2+ channels still have the property to be stimulated through protein kinase C activation.Similarly, in the presence of the autoanti-PtdIns Ab, caffeine was still able to induce a Ca2+-activated C1-current whereas prior application of noradrenaline was ineffective (Fig. 5A).This observation reveals that the internal Ca2+ stores are not empty and that C1-channels can be opened by in cells dialyzed with a pipette solution containing 0.012 mg/ml autoanti-PMIns Ab.PDBu induced a sustained increase in the Ca2+ channel current whereas noradrenaline was ineffective.Inset, current traces corresponding to small letters on the curve.Currents are expressed as a fraction of their control values (Z/Zc).a rise in the cytosolic Ca2+ concentration.In addition, the presence of the autoanti-PtdIns Ab in the pipette does not alter the function and modulation of both C1-and Ca2+ channels by substances other than noradrenaline.

A caceine 10-'M
To determine whether the autoanti-PtdIns Ab may interact with the same transduction process activated by muscarinic receptor activation but not with other membrane transduction processes, we tested their effects on acetylcholine and isoprenaline-induced responses.In smooth muscle, acetylcholine has been shown to produce a G-protein-mediated stimulation of phospholipase C activity (13).As illustrated in Fig. 5B, the acetylcholine activated C1-current was completely abolished when autoanti-PtdIns Ab (0.012 mg of IgG/ml) were applied intracellularly ( n = 4).On the other hand, it has been described recently that P-adrenergic receptor activation stimulates the L-type Ca2+ channel current of smooth muscle cells (14,15) and that a similar effect can be produced by forskolin which directly activates adenylate cyclase ( 14).Although the transduction coupling between P-adrenoceptors and Ca2+ channels is not completely understood in smooth muscle, it does not involve activation of phospholipase C. When M isoprenaline was applied in the presence of M phentolamine, the maximal Ca2+ channel current was increased by 47.4 f 12.6% ( n = 5; Fig. 6A).The increase in Ca2+ channel current induced by isoprenaline was suppressed after a 10min pretreatment with M propranolol, indicating that this effect was mediated through 8-adrenoceptors.Intracellular applications of autoanti-PtdIns Ab (0.012 mg of IgG/ ml) had no effect on the isoprenaline-induced stimulation of Ca2+ channel current, as illustrated in Fig. 6B, ( p > 0.05, n = 3).

DISCUSSION
The in vivo administration of poly and monoclonal Ab directed against endogenous compounds has been described previously (16)(17)(18)(19)(20).A monoclonal Ab against PtdIns(4,5)P2, produced from mice, has been injected into NIH 3T3 cells transformed by transfection with cloned human activated c-Ki-rus and v-Ha-rus genes.These authors have thus shown that transformation by these types of oncogenes switches over the mitogenic signaling pathway to a pathway dependent on PtdIns(4,5)P2 ( 21 monoclonal Ab directed against PtdIns(4,5)P2 into neonatal rat vascular smooth muscle cells had been used to show that the rise in intracellular Ca2+ concentration induced by platelet-derived growth factor involved PtdIns(4,5)Pz or PtdIns(4,5)P2 metabolism ( 22).However, little has been reported with auto-Ab of human sera from patients with diseases.Obstacles encountered for the use of auto-Ab from human sera are: (i) the antigen recognition is not or poorly defined; (ii) the Ab levels are often low; and (iii) the Ab avidity and specificity are rather poor and often not known.
Evaluation and characterization of autoanti-PtdIns Ab in sera of cancer patients have shown that the level of these auto-Ab is high as the autoanti-PtdIns activity can be identified in highly diluted sera (1/15,000) and that both avidity and specificity have been found to be higher for PtdIns than for the other phosphatidylinositides (7-9; Fig. 1).
For the first time, we report in the present study the in vivo use of human autoanti-PtdIns Ab.The auto-Ab inhibited both the noradrenaline-induced enhancement of Ca" channel current and the noradrenaline-and acetylcholine-activated Ca2+-dependent C1-current in portal vein smooth muscle cells.In contrast, the effects of phorbol 12,13-dibutyrate and caffeine were unchanged in the presence of auto-Ab.These results indicate that Ca2+ channels, C1-channels, protein kinase C, and intracellular Ca2+ stores are not directly affected by the autoanti-PtdIns Ab which should act on a step of the transduction process between al-adrenoceptors and phospholipase C. As the avidity of the autoanti-PtdIns Ab for InsPs and OAG is rather low, it is unlikely that the inhibition of the noradrenaline-induced responses may be attributed to direct binding of the auto-Ab to either InsP3 and DAG inside the cell.In addition, when polyclonal IgG from healthy patients or IgG fractions from patients with malignant tumors incubated with PtdIns were added to the pipette solution, the noradrenaline-induced responses were not altered.
Inhibition of both noradrenaline and acetylcholine effects was only seen when the purified IgG from cancer sera was intracellularly applied, suggesting that it could result from the binding of the autoanti-PtdIns Ab to PtdIns or to PtdIns(4)P and PtdIns(4,5)Pz, which are bound to the inner side of the plasma membrane.Although the auto-Ab avidity for PtdIns is about 15-fold higher than for PtdIns(4)P and PtdIns(4,5)P2 and the PtdIns concentration in smooth muscle is estimated to be 10-fold higher than those of PtdIns(4)P and PtdIns(4,5)P2 ( 23), interaction of the autoanti-PtdIns Ab with these two phosphorylated forms of PtdIns cannot be excluded.Moreover, the IgG concentrations used for electrophysiological experiments are 10-100-fold higher than those used for immunological studies.However, it is likely that the IgG concentration reached in the cytoplasm may be lower than that of the pipette solution as diffusion of large molecules such as IgG may be limited.As the inhibitory action of the autoanti-PtdIns Ab inside the cell is similarly observed on the two electrophysiological responses induced by noradrenaline, i.e. activation of Ca'+-dependent C1-current and enhancement of Ca2+ channel current, these results suggest that both InsP3 and DAG production are prevented by the auto-Ab.In contrast, other transduction pathways such that mediated through P-adrenoceptor activation are not affected at all by intracellular dialysis of the autoanti-PtdIns Ab.As coupling between the P-adrenoceptor and the Ca2+ channel in smooth muscle may involve either a direct coupling by a G.type protein (15) or intracellular cyclic AMP (14) these results demonstrate the specificity of the autoanti-PtdIns Ab against the membrane transduction pathway involving phosphatidylinositol breakdown.
Use of human autoanti-PtdIns Ab provides further evidence for the involvement of phospholipase C and phosphatidylinositides in the intracellular transduction mechanism triggered by al-adrenoceptor activation in vascular smooth muscle cells.In several publications, neomycin, which is believed to prevent enzymatic action of phospholipase C by binding to PtdIns(4,5)P2, has been used to demonstrate the involvement of phospholipase C in coupling mechanisms between receptors and effectors (24-26).However, the effects of neomycin are not very specific as neomycin may produce an inhibition of Ca2+ influx ( 27) and changes in receptor-agonist binding affinity (28).Therefore, autoanti-PtdIns Ab appear to be a more specific tool to demonstrate the role of PtdIns in intracellular transduction processes.Although circulating autoanti-PtdIns Ab from cancer patient sera do not bind to both PtdIns and phosphoinositides in normal vascular smooth muscle cells when applied extracellularly, this observation does not exclude the possibility that auto-Ab may interact with PtdIns in transformed and injured cells.

FIG. 1 .
FIG. 1. Displacement curves estshlished by competition experiments in an ELISA system between PtdIns coated on well plates and each of the followinn Zompounds: PtdIns (O), PtdIns(4)P (A), PtdIns(4,5)Pz (V), Ins(1,4,6)Ps (O), Ins (W), PtdSer (X), and OAG (A) previously incubated, at various concentrations, with an IgG fraction (7.10-' mg/ml) purified from human breast cancer sera.Absorbance values are expressed as BIBo with BO = OD without competitors and B = OD with competitors.C = competitor molar concentration.Data points are the means of three determinations.Standard deviations are smaller than the symbol size.

FIG. 2 .
FIG. 2. Effect of noradrenaline ( N A ) on C1-current recorded with the whole cell patch-clamp technique in single vascular smooth muscle cells isolated from rat portal vein.Ca2+-activated C1-current was recorded at a holding potential of -60 mV in response to the application of noradrenaline M) for 200 ms in control ( A ) , in the presence of 0.012 mg/ml IgG from breast cancer serum ( B ) , and from healthy patients (C). in the pipette preincubated with a lo-' M solution of PtdIns (D) was added to the solution, and when 0.012 mg/ml IgG from breast cancer serum pipette solution.

FIG. 3 .
FIG.3.Effect of noradrenaline (lo-' M) on Ca2+ channel current.Ca2+ channel current was evoked by a depolarization to 0 mV from a holding potential of -60 mV before ( a ) and during the application of noradrenaline ( b ) in control conditions ( A ) or in the presence of purified IgG from breast cancer serum at 0.0004 mg/ml ( B ) , 0.0012 mg/ml (C), and 0.012 mg/ml (D) in the pipette solution.

FIG. 4 .
FIG. 4. Effects of PDBu on Ca2+ channel current.A, time course of the peak Caz+ channel current during the application of M PDBu.Inset, current traces corresponding to small letters on the curve.B, time course of the peak Ca2+ channel current during successive applications of noradrenaline (lo-' M) and PDBu FIG. 5. Effects of caffeine and acetylcholine on Ca2+-activated C1-current elicited from a holding potential of -60 mV.A, under control conditions, the application of caffeine (lo-' M) evoked a large transient inward current ( a ) .In the presence of autoanti-PtdIns Ab (0.012 mg/ml) in the pipette solution, caffeine (lo-' M) still elicited a large C1-current whereas noradrenaline ( N A ) ( M) was ineffective ( b ) .B, under control conditions, acetylcholine (ACh) ( W 5 M) induced activation of a C1-current ( a ) which was completely inhibited in the presence of autoanti-PtdIns Ab (0.012 mg/ml) in the pipette solution (b).The external solution contained 2.5 mM Ca2+,and there was no EGTA in the pipette solution.

FIG. 6 .
FIG. 6.Effects of isoprenaline on Ca2+ channel current.A,time course of the peak Ca2+ channel current during application of isoprenaline (10"j M).Inset, current traces corresponding to small letters on the curve.B, time course of the peak Ca2+ channel current during application of isoprenaline (1O"j M) in cells dialyzed with a pipette solution containing 0.012 mg/ml autoanti-PtdIns Ab.Inset, current traces corresponding to small letters on the curve.Currents are expressed as a fraction of control values (I/Zc).